“fm” — 2003/3/11 — page i — #1
Aircraft Design Projects
“fm” — 2003/3/11 — page ii — #2
Dedications
To Jessica, Maria, Edward, Robert and Jonothan – in their hands rests the future.
To my father, J. F. Marchman, Jr, for passing on to me his love of airplanes and to my teacher, Dr Jim Williams, whose example inspired me to pursue a career in education.
“fm” — 2003/3/11 — page iii — #3
Aircraft DesignProjects
for engineering students
Lloyd R. Jenkinson
James F. Marchman III
OXFORD AMSTERDAM BOSTON LONDON NEW YORK PARIS
SAN DIEGO SAN FRANCISCO SINGAPORE SYDNEY TOKYO
“fm” — 2003/3/10 — page v — #5
Contents
Preface xiii xvi xvii
AcknowledgementsIntroduction
1 Design methodology 1
2 Preliminary design 6 2.1 Problem definition 6
7 8
2.1.1 2.1.2 2.1.3 Understanding the problem 8 2.1.4 Innovation 9 2.1.5 Organising the design process 10 2.1.6 Summary 11
The customers Aircraft viability
2.2 Information retrieval 112.2.1 Existing and competitive aircraft 112.2.2 Technical reports 122.2.3 Operational experience 12
2.3 Aircraft requirements 122.3.1 Market and mission issues 132.3.2 Airworthiness and other standards 132.3.3 Environmental and social issues 132.3.4 Commercial and manufacturing considerations 142.3.5 Systems and equipment requirements 14
2.4 Configuration options 142.5 Initial baseline sizing 15
2.5.1 Initial mass (weight) estimation 162.5.2 Initial layout drawing 19
2.6 Baseline evaluation 192.6.1 Mass statement 19 2.6.2 Aircraft balance 21 2.6.3 Aerodynamic analysis 22 2.6.4 Engine data 24 2.6.5 Aircraft performance 25 2.6.6 Initial technical report 25
2.7 Refining the initial layout 252.7.1 Constraint analysis 262.7.2 Trade-off studies 29
“fm” — 2003/3/10 — page vi — #6
vi Contents
2.8 Refined baseline design 312.9 Parametric and trade studies 32
2.9.1 Example aircraft used to illustrate trade-off andparametric studies 33
2.10 Final baseline configuration 392.10.1 Additional technical considerations 392.10.2 Broader-based considerations 39
2.11 Type specification 402.11.1 Report format 402.11.2 Illustrations, drawings and diagrams 41
References 41
3 Introduction to the project studies 43
4 Project study: scheduled long-range business jet 464.1 Introduction 474.2 Project brief 49
4.2.1 Project requirements 504.3 Project analysis 50
4.3.1 Payload/range 504.3.2 Passenger comfort 514.3.3 Field requirements 514.3.4 Technology assessments 524.3.5 Marketing 534.3.6 Alternative roles 544.3.7 Aircraft developments 544.3.8 Commercial analysis 55
4.4 Information retrieval 564.5 Design concepts 57
4.5.1 Conventional layout(s) 574.5.2 Braced wing/canard layout 584.5.3 Three-surface layout 594.5.4 Blended body layout 604.5.5 Configuration selection 61
4.6 Initial sizing and layout 624.6.1 Mass estimation 624.6.2 Engine size and selection 634.6.3 Wing geometry 634.6.4 Fuselage geometry 674.6.5 Initial ‘baseline aircraft’ general arrangement drawing 68
4.7 Initial estimates 704.7.1 Mass and balance analysis 704.7.2 Aerodynamic estimations 754.7.3 Initial performance estimates 764.7.4 Constraint analysis 784.7.5 Revised performance estimates 794.7.6 Cost estimations 80
4.8 Trade-off studies 824.8.1 Alternative roles and layout 824.8.2 Payload/range studies 85
“fm” — 2003/3/10 — page vii — #7
Contents vii
4.8.3 Field performance studies 864.8.4 Wing geometry studies 874.8.5 Economic analysis 91
4.9 Initial ‘type specification’ 964.9.1 General aircraft description 964.9.2 Aircraft geometry 974.9.3 Mass (weight) and performance statements 974.9.4 Economic and operational issues 98
4.10 Study review 99References 100
5 Project study: military training system 1015.1 Introduction 1025.2 Project brief 102
5.2.1 Aircraft requirements 1035.2.2 Mission profiles 104
5.3 Problem definition 1055.4 Information retrieval 106
5.4.1 Technical analysis 1085.4.2 Aircraft configurations 1105.4.3 Engine data 110
5.5 Design concepts 1105.6 Initial sizing 112
5.6.1 Initial baseline layout 1135.7 Initial estimates 115
5.7.1 Mass estimates 1155.7.2 Aerodynamic estimates 1175.7.3 Performance estimates 119
5.8 Constraint analysis 1295.8.1 Take-off distance 1295.8.2 Approach speed 1295.8.3 Landing distance 1305.8.4 Fundamental flight analysis 1305.8.5 Combat turns at SL 1305.8.6 Combat turn at 25 000 ft 1315.8.7 Climb rate 1315.8.8 Constraint diagram 131
5.9 Revised baseline layout 1325.9.1 Wing fuel volume 133
5.10 Further work 1345.11 Study review 137
5.11.1 Strengths 1375.11.2 Weaknesses 1375.11.3 Opportunities 1395.11.4 Threats 1395.11.5 Revised aircraft layout 140
5.12 Postscript 141References 141
“fm” — 2003/3/10 — page viii — #8
viii Contents
6 Project study: electric-powered racing aircraft 1436.1 Introduction 1446.2 Project brief 144
6.2.1 The racecourse and procedures 1446.2.2 History of Formula 1 racing 1456.2.3 Comments from a racing pilot 1466.2.4 Official Formula 1 rules 147
6.3 Problem definition 1496.4 Information retrieval 150
6.4.1 Existing aircraft 1506.4.2 Configurational analysis 1526.4.3 Electrical propulsion system 154
6.5 Design concepts 1576.6 Initial sizing 158
6.6.1 Initial mass estimations 1596.6.2 Initial aerodynamic considerations 1626.6.3 Propeller analysis 165
6.7 Initial performance estimation 1666.7.1 Maximum level speed 1666.7.2 Climb performance 1696.7.3 Turn performance 1716.7.4 Field performance 173
6.8 Study review 173References 174
7 Project study: a dual-mode (road/air) vehicle 1757.1 Introduction 1767.2 Project brief (flying car or roadable aircraft?) 1767.3 Initial design considerations 1777.4 Design concepts and options 1797.5 Initial layout 1817.6 Initial estimates 186
7.6.1 Aerodynamic estimates 1867.6.2 Powerplant selection 1897.6.3 Weight and balance predictions 1907.6.4 Flight performance estimates 1907.6.5 Structural details 1937.6.6 Stability, control and ‘roadability’ assessment 1967.6.7 Systems 1977.6.8 Vehicle cost assessment 198
7.7 Wind tunnel testing 1997.8 Study review 200References 201
8 Project study: advanced deep interdiction aircraft 2028.1 Introduction 2038.2 Project brief 203
8.2.1 Threat analysis 2038.2.2 Stealth considerations 2048.2.3 Aerodynamic efficiency 206
“fm” — 2003/3/10 — page ix — #9
Contents ix
8.3 Problem definition 2088.4 Design concepts and selection 2108.5 Initial sizing and layout 2138.6 Initial estimates 215
8.6.1 Initial mass estimations 2168.6.2 Initial aerodynamic estimations 217
8.7 Constraint analysis 2218.7.1 Conclusion 227
8.8 Revised baseline layout 2288.8.1 General arrangement 2288.8.2 Mass evaluation 2338.8.3 Aircraft balance 2338.8.4 Aerodynamic analysis 2348.8.5 Propulsion 241
8.9 Performance estimations 2428.9.1 Manoeuvre performance 2428.9.2 Mission analysis 2508.9.3 Field performance 254
8.10 Cost estimations 2598.11 Trade-off studies 2618.12 Design review 263
8.12.1 Final baseline aircraft description 2638.12.2 Future considerations 267
8.13 Study review 268References 268
9 Project study: high-altitude, long-endurance (HALE) uninhabited aerialsurveillance vehicle (UASV) 270
9.1 Introduction 2719.2 Project brief 271
9.2.1 Aircraft requirements 2729.3 Problem definition 2729.4 Initial design considerations 2759.5 Information retrieval 275
9.5.1 Lockheed Martin U-2S 2769.5.2 Grob Strato 2C 2769.5.3 Northrop Grumman RQ-4A Global Hawk 2779.5.4 Grob G520 Strato 1 2779.5.5 Stemme S10VC 277
9.6 Design concepts 2789.6.1 Conventional layout 2799.6.2 Joined wing layout 2809.6.3 Flying wing layout 2809.6.4 Braced wing layout 2819.6.5 Configuration selection 282
9.7 Initial sizing and layout 2839.7.1 Aircraft mass estimation 2839.7.2 Fuel volume assessment 2859.7.3 Wing loading analysis 2859.7.4 Aircraft speed considerations 286
“fm” — 2003/3/10 — page x — #10
x Contents
9.7.5 Wing planform geometry 2889.7.6 Engine sizing 2909.7.7 Initial aircraft layout 2929.7.8 Aircraft data summary 293
9.8 Initial estimates 2949.8.1 Component mass estimations 294 9.8.2 Aircraft mass statement and balance 297 9.8.3 Aircraft drag estimations 298 9.8.4 Aircraft lift estimations 299 9.8.5 Aircraft propulsion 300 9.8.6 Aircraft performance estimations 300
9.9 Trade-off studies 3059.10 Revised baseline layout 3059.11 Aircraft specification 307
9.11.1 Aircraft description 3079.11.2 Aircraft data 307
9.12 Study review 308References 309
10 Project study: a general aviation amphibian aircraft 31010.1 Introduction 31110.2 Project brief 311
10.2.1 Aircraft requirements 31210.3 Initial design considerations 31210.4 Design concepts 31210.5 Initial layout and sizing 313
10.5.1 Wing selection 31310.5.2 Engine selection 31410.5.3 Hull design 31410.5.4 Sponson design 31610.5.5 Other water operation considerations 31710.5.6 Other design factors 318
10.6 Initial estimates 31810.6.1 Aerodynamic estimates 31810.6.2 Mass and balance 31810.6.3 Performance estimations 32110.6.4 Stability and control 32310.6.5 Structural details 323
10.7 Baseline layout 32410.8 Revised baseline layout 32510.9 Further work 32510.10 Study review 328References 329
11 Design organisation and presentation 33111.1 Student’s checklist 332
11.1.1 Initial questions 33211.1.2 Technical tasks 332
11.2 Teamworking 33311.2.1 Team development 335
“fm” — 2003/3/10 — page xi — #11
Contents xi
11.2.2 Team member responsibilities 33611.2.3 Team leadership requirements 33611.2.4 Team operating principles 33711.2.5 Brainstorming 337
11.3 Managing design meetings 33811.3.1 Prior to the meeting 33911.3.2 Minutes of the meeting 33911.3.3 Dispersed meetings 341
11.4 Writing technical reports 34111.4.1 Planning the report 34211.4.2 Organising the report 34211.4.3 Writing the report 34311.4.4 Referencing 34411.4.5 Use of figures, tables and appendices 34511.4.6 Group reports 34611.4.7 Review of the report 347
11.5 Making a technical presentation 34811.5.1 Planning the presentation 34911.5.2 Organising the presentation 34911.5.3 Use of equipment 35011.5.4 Management of the presentation 35111.5.5 Review of the presentation 352
11.6 Design course structure and student assessment 35311.6.1 Course aims 35311.6.2 Course objectives 35411.6.3 Course structure 35411.6.4 Assessment criteria 35511.6.5 Peer review 356
11.7 Naming your aircraft 356Footnote 357
Appendix A: Units and conversion factors 359Derived units 360Funny units 360Conversions (exact conversions can be found in British Standards BS350/2856) 361Some useful constants (standard values) 362
Appendix B: Design data sources 363Technical books (in alphabetical order) 363Reference books 365Research papers 365Journals and articles 366The Internet 366
Index 367
“fm” — 2003/3/10 — page xii — #12
“fm” — 2003/3/10 — page xiii — #13
Preface
There are many excellent texts covering aircraft design from a variety of perspectives.1
Some of these are aimed at specific audiences ranging from practising aerospace engi-neers, to engineering students, to amateur airplane builders. Others cover specialized aspects of the subject such as undercarriage or propulsion system design. Some of these are quite detailed in their presentation of the design process while others are very general in scope. Some are overviews of all the basic aeronautical engineering subjects that come together in the creation of a design.
University faculty that teach aircraft design courses often face difficult choices when evaluating texts or references for their students’ use. Many texts that are suitable for use in a design class are biased toward particular classes of aircraft such as military aircraft, general aviation, or airliners. A text that gives excellent coverage of design basics may prove slanted toward a class of aircraft different from that year’s project. Alternatively, those that emphasize the correct type of vehicle may treat design fundamentals in an unfamiliar manner. The situation may be further complicated in classes that have several teams of students working on different types of designs, some of which ‘fit’ the chosen text while others do not.
Most teachers would prefer a text that emphasizes the basic thought processes of preliminary design. Such a text should encourage students to seek an understanding of the approaches and constraints appropriate to their design assignment before they venture too far into the analytical processes. On the other hand, students would like a text which simply tells them where to input their design objectives into a ‘black-box’ computer code or generalized spreadsheet, and preferably, where to catch the final design drawings and specifications as they are printed out. Faculty would like their students to begin the design process with a thorough review of their previous courses in aircraft performance, aerodynamics, structures, flight dynamics, propulsion, etc. Students prefer to start with an Internet search, hoping to find a solution to their problem that requires only minimal ‘tweaking’.
The aim of this book is to present a two pronged approach to the design process. It is expected to appeal to both faculty and students. It sets out the basics of the design thought process and the pathway one must travel in order to reach an aircraft design goal for any category of aircraft. Then it presents a variety of design case studies. These are intended to offer examples of the way the design process may be applied to conceptual design problems typical of those actually used at the advanced level in academic and other training curricula. It does not offer a step-by-step ‘how to’ design guide, but shows how the basic aircraft preliminary design process can be successfully applied to a wide range of unique aircraft. In so doing, it shows that each set of design objectives presents its own peculiar collection of challenges and constraints. It also shows how the classical design process can be applied to any problem.
“fm” — 2003/3/10 — page xiv — #14
xiv Preface
Case studies provide both student and instructor with a valuable teaching/learning tool, allowing them to examine the way others have approached particular design chal-lenges. In the 1970s, the American Institute of Aeronautics and Astronautics (AIAA) published an excellent series of design case studies2 taken from real aircraft project developments. These provided valuable insights into the development of several, then current, aircraft. Some other texts have employed case studies taken from industrial practice. Unfortunately, these tend to include aspects of design that are beyond the conceptual phase, and which are not covered in academic design courses. While these are useful in teaching design, they can be confusing to the student who may have diffi-culty discerning where the conceptual aspects of the design process ends and detailed design ensues. The case studies offered in this text are set in the preliminary design phase. They emphasize the thought processes and analyses appropriate at this stage of vehicle development.
Many of the case studies presented in this text were drawn from student projects. Hence, they offer an insight into the conceptual design process from a student per-spective. The case studies include design projects that won top awards in national and international design competitions. These were sponsored by the National Aeronautics and Space Administration (NASA), the US Federal Aviation Administration (FAA), and the American Institute of Aeronautics and Astronautics (AIAA).
The authors bring a unique combination of perspectives and experience to this text. It reflects both British and American academic practices in teaching aircraft design to undergraduate students in aeronautical and aerospace engineering. Lloyd Jenkinson has taught aircraft design at both Loughborough University and Southampton University in the UK and Jim Marchman has taught both aircraft and spacecraft design at Virginia Tech in the US. They have worked together since 1997 in an experiment that combines students from Loughborough University and Virginia Tech in interna-tional aircraft design teams.3 In this venture, teams of students from both universities have worked jointly on a variety of aircraft design projects. They have used exchange visits, the Internet and teleconference communications to work together progressively, throughout the academic year, on the conceptual design of a novel aircraft.
In this book, the authors have attempted to build on their experience in international student teaming. They present processes and techniques that reflect the best in British and American design education and which have been proven to work well in both academic systems. Dr Jenkinson also brings to this text his prior experience in the aerospace industry of the UK, having worked on the design of several successful British aircraft. Professor Marchman’s contribution to the text also reflects his experiences in working with students and faculty in Thailand and France in other international design team collaborative projects.
The authors envision this text as supplementing the popular aircraft design textbooks, currently in use at universities around the world. Books such as those reviewed by Mason1 could be employed to present the detailed aspects of the preliminary design process. Working within established conceptual design methodology, this book will provide a clearer picture of the way those detailed analyses may be adapted to a wide range of aircraft types.
It would have been impossible to write this book without the hard work and enthusi-asm shown by many of our students over more years than we care to remember. Their continued interest in aircraft design project work and the smoothing of the difficulties they sometimes experienced in progressing through the work was our inspiration. We have also benefited from the many colleagues and friends who have been generous in sharing their encouragement and knowledge with us. Aircraft design educators seem
“fm” — 2003/3/10 — page xv — #15
Preface xv
to be a special breed of engineers who selflessly give their effort and time to inspire anyone who wants to participate in their common interest. We are fortunate to count them as our friends.
References 1 Bill Mason’s web page: www.aoe.vt.edu/Mason/ACinfoTOC.html.2 AIAA web page: www.aiaa.org/publications/index.3 Jenkinson, L. R., Page, G. J., Marchman, J. F., ‘A model for international teaming in air-
craft design education’, Journal of Aircraft Design, Vol. 3, No. 4, pp. 239–247, Elsevier, December 2000.
“fm” — 2003/3/10 — page xvi — #16
Acknowledgements
To all the students and staff at Loughborough and Southampton Universities who have, over many years, contributed directly and indirectly to my understanding of the design of aircraft, I would like to express my thanks and appreciation. For their help with proof reading and technical advice, I thank my friends Paul Eustace and Keith Payne. Our gratitude to all those people in industry who have provided assistance with the projects. Finally, to my wife and family for their support and understanding over the time when my attention was distracted by the writing of the book.
Lloyd Jenkinson
I would like to acknowledge the work done by the teams of Virginia Tech and Loughborough University aircraft design students in creating the designs which I attempted to describe in Chapters 7 and 10 and the contributions of colleagues such as Bill Mason, Nathan Kirschbaum, and Gary Page in helping guide those students in the design process. Without these people these chapters could not have been written.
Jim Marchman
“fm” — 2003/3/10 — page xvii — #17
Introduction
It is tempting to title this book ‘Flights of Fancy’ as this captures the excitement and expectations at the start of a new design project. The main objective of this book is to try to convey this feeling to those who are starting to undertake aircraft conceptual design work for the first time. This often takes place in an educational or industrial training establishment. Too often, in academic studies, the curiosity and fascination of project work is lost under a morass of mathematics, computer programming, analytical methods, project management, time schedules and deadlines. This is a shame as there are very few occasions in your professional life that you will have the chance to let your imagination and creativity flow as freely as in these exercises. As students or young engineers, it is advisable to make the most of such opportunities.
When university faculty or counsellors interview prospective students and ask why they want to enter the aeronautics profession, the majority will mention that they want to design aircraft or spacecraft. They often tell of having drawn pictures of aeroplanes since early childhood and they imagine themselves, immediately after graduation, pro-ducing drawings for the next generation of aircraft. During their first years in the university, these young men and women are often less than satisfied with their basic courses in science, mathematics, and engineering as they long to ‘design’ something. When they finally reach the all-important aircraft design course, for which they have yearned for so long, they are often surprised. They find that the process of design requires far more than sketching a pretty picture of their dream aircraft and entering the performance specifications into some all-purpose computer program which will print out a final design report.
Design is a systematic process. It not only draws upon all of the student’s previous engineering instruction in structures, aerodynamics, propulsion, control and other subjects, but also, often for the first time, requires that these individual academic subjects be applied to a problem concurrently. Students find that the best aerodynamic solution is not equated to the best structural solution to a problem. Compromises must be made. They must deal with conflicting constraints imposed on their design by control requirements and propulsion needs. They may also have to deal with real world political, environmental, ethical, and human factors. In the end, they find they must also do practical things like making sure that their ideal wing will pass through the hangar door!
An overview of the book
This book seeks to guide the student through the preliminary stages of the aircraft design process. This is done by both explaining the process itself (Chapters 1 and 2) and by providing a variety of examples of actual student design projects (Chapters 3
“fm” — 2003/3/10 — page xviii — #18
xviii Introduction
to 10). The projects have been used as coursework at universities in the UK and the US. It should be noted that the project studies presented are not meant to provide a ‘fill in the blank’ template to be used by future students working on similar design problems but to provide insight into the process itself. Each design problem, regardless of how similar it may appear to an earlier aircraft design, is unique and requires a thorough and systematic investigation. The project studies presented in this book merely serve as examples of how the design process has been followed in the past by other teams faced with the task of solving a unique problem in aircraft design.
It is impossible to design aircraft without some knowledge of the fundamental the-ories that influence and control aircraft operations. It is not possible to include such information in this text but there are many excellent books available which are written to explain and present these theories. A bibliography containing some of these books and other sources of information has been added to the end of the book. To understand the detailed calculations that are described in the examples it will be necessary to use the data and theories in such books. Some design textbooks do contain brief examples on how the analytical methods are applied to specific aircraft. But such studies are mainly used to support and illustrate the theories and do not take an overall view of the preliminary design process.
The initial part of the book explains the preliminary design process. Chapter 1 briefly describes the overall process by which an aircraft is designed. It sets the preliminary design stages into the context of the total transformation from the initial request for proposal to the aircraft first flight and beyond. Although this book only deals with the early stages of the design process, it is necessary for students to understand the subsequent stages so that decisions are taken wisely. For example, aircraft design is by its nature an iterative process. This means that estimates and assumptions have sometimes to be made with inadequate data. Such ‘guesstimates’ must be checked when more accurate data on the aircraft is available. Without this improvement to the fidelity of the analytical methods, subsequent design stages may be seriously jeopardized.
Chapter 2 describes, in detail, the work done in the early (conceptual) design process. It provides a ‘route map’ to guide a student from the initial project brief to the validated ‘baseline’ aircraft layout. The early part of the chapter includes sections that deal with ‘defining and understanding the problem’, ‘collecting useful information’ and ‘setting the aircraft requirements’. This is followed by sections that show how the initial aircraft configuration is produced. Finally, there are sections illustrating how the initial aircraft layout can be refined using constraint analysis and trade-off studies. The chapter ends with a description of the ‘aircraft type specification’. This report is commonly used to collate all the available data about the aircraft. This is important as the full geometrical description and data will be needed in the detailed design process that follows.
Chapter 3 introduces the seven project studies that follow (Chapters 4 to 10). It describes each of the studies and provides a format for the sequence of work to be followed in some of the studies. The design studies are not sequential, although the earlier ones are shown in slightly more detail. It is possible to read any of the studies separately, so a short description of each is presented.
Chapters 4 to 10 inclusive contain each of the project studies. The projects are selected from different aeronautical applications (general aviation, civil transports, military aircraft) and range from small to heavy aircraft. For conciseness of presentation the detailed calculations done to support the final designs have not been included in these chapters but the essential input values are given so that students can perform their own analysis. The projects are mainly based on work done by students on aeronautical engineering degree courses. One of the studies is from industrial work and some have
“fm” — 2003/3/10 — page xix — #19
Introduction xix
been undertaken for entry to design competitions. Each study has been selected to illustrate a different aspect of preliminary design and to illustrate the varied nature of aircraft conceptual design.
The final chapter (11) offers guidance on student design work. It presents a set of questions to guide students in successfully completing an aircraft design project. It includes some observations about working in groups. Help is also given on the writing of technical reports and making technical presentations.
Engineering units of measurement
Experience in running design projects has shown that students become confused by the units used to define parameters in aeronautics. Some detailed definitions and con-versions are contained in Appendix A at the end of the book and a quick résumé is given here.
Many different systems of measurement are used throughout the world but two have become most common in aeronautical engineering. In the US the now inappropriately named ‘British’ system (foot, pound and second) is widely used. In the UK and over most of Europe, System International (SI) (metres, newton and second) units are stan-dard. It is advised that students only work in one system. Confusion (and disaster) can occur if they are mixed. The results of the design analysis can be quoted in both types of unit by applying standard conversions. The conversions below are typical:
1 inch = 25.4 mm1 sq. ft = 0.0929 sq. m1 US gal = 3.785 litres1 US gal = 0.833 Imp. gal1 statute mile = 1.609 km1 ft/s = 0.305 m/s1 knot = 1.69 ft/s1 pound force = 4.448 newtons1 horsepower = 745.7 watts
1 foot = 0.305 metres 1 cu. ft = 28.32 litres 1 Imp. gal = 4.546 litres 1 litre = 0.001 cubic metres 1 nautical mile = 1.852 km 1 knot = 0.516 m/s 1 knot = 1.151 mph 1 pound mass = 0.454 kilogram 1 horsepower = 550 ft lb/s
To avoid confusing pilots and air traffic control, some international standardization of units has had to be accepted. These include:
Aircraft altitude – feet Aircraft forward speed – knots∗
Aircraft range – nautical miles Climb rate – feet per minute
(∗ Be extra careful with the definition of units used for aircraft speed as pilots like to use airspeed in IAS (indicated airspeed as shown on their flight instruments) and engineers like TAS (true airspeed, the speed relative to the ambient air)).
Fortunately throughout the world, the International Standard Atmosphere (ISA) has been adopted as the definition of atmospheric conditions. ISA charts and data can be found in most design textbooks. In this book, which is aimed at a worldwide readership, where possible both SI and ‘British’ units have been quoted. Our apologies if this confuses the text in places.
“fm” — 2003/3/10 — page xx — #20
xx Introduction
English – our uncommon tongue
Part of this book grew out of the authors’ collaboration in a program of international student design projects over several years. As we have reported our experiences from that program, observers have often noted that one thing that makes our international collaboration easier than some others is the common language. On the other hand, one thing we and our students have learned from this experience is that many of the aspects of our supposedly common tongue really do not have much in common.
Pairing an Englishman and an American to create a textbook aimed at both the US, British and other markets is an interesting exercise in spelling and language skills. While (or is it whilst?) the primary language spoken in the United Kingdom and the United States grows from the same roots, it has very obviously evolved somewhat differently. An easy but interesting way to observe some of these differences is to take a page of text from a British book and run it through an American spelling check program. Checking an American text with an ‘English’ spell checker will produce similar surprises. We spell many words differently, usually in small ways. Is it ‘color’ or ‘colour’; do we ‘organize’ our work or ‘organise’ it? In addition, do we use double (“) or single (‘) strokes to indicate a quote or give emphasis to a word or phrase? Will we hold our next meeting at 9:00 am or at 9.00 am? (we won’t even mention the 24 hour clock!).
There are also some obvious differences between terminology employed in the US and UK. Does our automobile have a ‘bonnet’ and a ‘boot’ or a ‘hood’ and a ‘trunk’ and does its engine run on ‘gasoline’ or ‘petrol’? American ‘airplanes’ have ‘landing gear’ while British ‘aeroplanes/airplanes or aircraft’ have ‘undercarriages’, does it have ‘reheat’ or an ‘afterburner’. Fortunately, most of us have watched enough television shows and movies from both countries to be comfortable with these differences.
As we have pieced together this work we have often found ourselves (and our com-puter spell checkers) editing each other’s work to make it conform to the conventions in spelling, punctuation, and phraseology, assumed to be common to each of our versions of this common language. The reader may find this evident as he or she goes from one section of the text to another and detects changes in wording and terminology which reflect the differing conventions in language use in the US and UK. It is hoped that these variations, sometimes subtle and sometimes obvious, will not prove an obstacle to the reader’s understanding of our work but will instead make it more interesting.
Finally
All aircraft projects are unique, therefore, it is impossible to provide a ‘template’ for the work involved in the preliminary design process. However, with knowledge of the detail steps in the preliminary design process and with examples of similar project work, it is hoped that students will feel freer to concentrate on the innovative and analytical aspects of the project. In this way they will develop their technical and communication abilities in the absorbing context of preliminary aircraft design.
“chap01” — 2003/3/10 — page 1 — #1
1
Design methodology
The start of the design process requires the recognition of a ‘need’. This normally comes from a ‘project brief’ or a ‘request for proposals (RFP)’. Such documents may come from various sources:
• Established or potential customers. • Government defence agencies. • Analysis of the market and the corresponding trends from aircraft demand. • Development of an existing product (e.g. aircraft stretch or engine change). • Exploitation of new technologies and other innovations from research and
development.
It is essential to understand at the start of the study where the project originated and to recognise what external factors are influential to the design before the design process is started.
At the end of the design process, the design team will have fully specified their design configuration and released all the drawings to the manufacturers. In reality, the design process never ends as the designers have responsibility for the aircraft throughout its operational life. This entails the issue of modifications that are found essential during service and any repairs and maintenance instructions that are necessary to keep the aircraft in an airworthy condition.
The design method to be followed from the start of the project to the nominal end can be considered to fall into three main phases. These phases are illustrated in Figure 1.1.
The preliminary phase (sometimes called the conceptual design stage) starts with the project brief and ends when the designers have found and refined a feasible baseline design layout. In some industrial organisations, this phase is referred to as the ‘feasibil-ity study’. At the end of the preliminary design phase, a document is produced which contains a summary of the technical and geometric details known about the baseline design. This forms the initial draft of a document that will be subsequently revised to contain a thorough description of the aircraft. This is known as the aircraft ‘Type Specification’.
The next phase (project design) takes the aircraft configuration defined towards the end of the preliminary design phase and involves conducting detailed analysis to improve the technical confidence in the design. Wind tunnel tests and computational fluid dynamic analysis are used to refine the aerodynamic shape of the aircraft. Finite element analysis is used to understand the structural integrity. Stability and control analysis and simulations will be used to appreciate the flying characteristics. Mass and balance estimations will be performed in increasingly fine detail. Operational factors (cost, maintenance and marketing) and manufacturing processes will be investigated
“chap01” — 2003/3/10 — page 3 — #3
Design methodology 3
100
%
Process II
TimescaleA B C D
Cost Design flexibility
Cost expended
Process I
II
I
Region Task
A Defining requirements B Conceptual design phase C Project design phase D Detail design phase
Fig. 1.2 Design flexibility
layout are avoided or, at best, reduced. Such changes are expensive and may delay the completion of the project. Managers are eager to validate the design to a high degree of confidence during the preliminary and project phases. A natural consequence of this policy is the progressive ‘freezing’ of the design configuration as the project matures. In the early preliminary design stages any changes can (and are encouraged to) be considered, yet towards the end of the project design phase only minor geometrical and system modifications will be allowed. If the aircraft is not ‘good’ (well engineered) by this stage then the project and possibly the whole company will be in difficulty. Within the context described above, the preliminary design phase presents a significant undertaking in the success of the project and ultimately of the company.
Design project work, as taught at most universities, concentrates on the preliminary phase of the design process. The project brief, or request for proposal, is often used to define the design problem. Alternatively, the problem may originate as a design topic in a student competition sponsored by industry, a government agency, or a technical society. Or the design project may be proposed locally by a professor or a team of students. Such design project assignments range from highly detailed lists of design objectives and performance requirements to rather vague calls for a ‘new and better’ replacement for existing aircraft. In some cases student teams may even be asked to develop their own design objectives under the guidance of their design professor.
To better reflect the design atmosphere in an industry environment, design classes at most universities involve teams of students rather than individuals. The use of multi-disciplinary design teams employing students from different engineering disciplines is being encouraged by industry and accreditation agencies.
The preliminary design process presented in this text is appropriate to both the indi-vidual and the team design approach although most of the cases presented in later chapters involved teams of design students. While, at first thought, it may appear that the team approach to design will reduce the individual workload, this may not be so.
“chap01” — 2003/3/10 — page 4 — #4
4 Aircraft Design Projects
The interpersonal dynamics of working in a team requires extra effort. However, this greatly enhances the design experience and adds team communications, management and interpersonnel interaction to the technical knowledge gained from the project work.
It is normal in team design projects to have all students conduct individual initial assessments of the design requirements, study comparable aircraft, make initial estim-ates for the size of their aircraft and produce an initial concept sketch. The full team will then begin its task by examining these individual concepts and assessing their merits as part of their team concept selection process. This will parallel the development of a team management plan and project timeline. At this time, the group will allocate various portions of the conceptual design process to individuals or small groups on the team.
At this point in this chapter, a word needs to be said about the role of the computer in the design process. It is natural that students, whose everyday lives are filled with computer usage for everything from interpersonal communication to the solution of complex engineering problems, should believe that the aircraft design process is one in which they need only to enter the operational requirements into some supercomputer and wait for the final design report to come out of the printer (Figure 1.3).
Indeed, there are many computer software packages available that claim to be ‘aircraft design programs’ of one sort or another. It is not surprising that students, who have read about new aircraft being ‘designed entirely on the computer’ in industry, believe that they will be doing the same. They object to wasting time conducting all of the basic analyses and studies recommended in this text, and feel that their time would be much better spent searching for a student version of an all-encompassing aircraft design code. They believe that this must be available from Airbus or Boeing if only they can find the right person or web address.
While both simple aircraft ‘design’ codes and massive aerospace industry CAD programs do exist and do play important roles, they have not yet replaced the basic pro-cesses outlined in this text. Simple software packages which are often available freely at various locations on the Internet, or with many modern aeronautical engineering texts, can be useful in the specialist design tasks if one understands the assumptions and lim-itations implicit in their analysis. Many of these are simple computer codes based on
Output
Designyour ownairplanein 5 min
Fig. 1.3 Student view of design
“chap01” — 2003/3/10 — page 5 — #5
Design methodology 5
STAB &
CONT
PRO
PULS
ION
STRUCTURES
2 AM
AERO-
DYNAMICA/C
PERF LU
F
Fig. 1.4 The ‘real’ design process
the elementary relationships used for aircraft performance, aerodynamics, and stability and control calculations. These have often been coupled to many simplifying assump-tions for certain categories of aircraft (often home-built general aviation vehicles). The solutions which can be obtained from many such codes can be obtained more quickly, and certainly with a much better understanding of the underlying assumptions, by using directly the well-known relationships on which they are based. In our experience, if students spent half the time they waste searching for a design code (which they expect will provide an instant answer) on thinking and working through the fundamental rela-tionships with which they are already supposedly familiar, they would find themselves much further along in the design process.
The vast and complex design computer programs used in the aerospace industry have not been created to do preliminary work. They are used to streamline the detail design part of the process. Such programs are not designed to take the initial project requirements and produce a final design. They are used to take the preliminary design, which has followed the step-by-step processes outlined in this text, and turn it into the thousands of detailed CAD drawings needed to develop and manufacture the finished vehicle.
It is the task of the aircraft design students to learn the processes which will take them from first principles and concepts, through the conceptual and preliminary design stages, to the point where they can begin to apply detailed design codes (Figure 1.4).
At this point in time, it is impossible to envisage how the early part of the design process will ever be replaced by off-the-shelf computer software that will automatically design novel aircraft concepts. Even if this program were available, it is probably not a substitute for working steadily through the design process to gain a fundamental understanding of the intricacies involved in real aircraft design.
Reference 1 Mavris, D. et al., ‘Methodology for examining the simultaneous impact of requirements,
vehicle characteristics and technologies on military aircraft design’, ICAS 2000, Harrogate UK, August 2000.
“chap02” — 2003/3/10 — page 6 — #1
2
Preliminary design
Conceptual design is the organised application of innovation to a real problem to produce a viable product for the customer.
(Anon.)
As previously described, the preliminary design phase starts with the recognition of need. It continues until a satisfactory starting point for the conceptual design phase has been identified. The aircraft layout at the end of the phase is referred to as the ‘baseline’ configuration. Between these two milestones there are a number of distinctive, and partially sequential, stages to be investigated. These stages are shown in Figure 2.1 and described below:
2.1 Problem definition
For novice aircraft designers the natural tendency when starting a project is to want to design aircraft. This must be resisted because when most problems are originally presented they do not include all the significant aspects surrounding the problem. As a lot of time and effort will be spent on the design of the aircraft, it is important that all the criteria, constraints and other factors are recognised before starting, otherwise a lot of work and effort may be wasted. For this reason, the first part of the conceptual design phase is devoted to a thorough understanding of the problem.
The definition of conceptual design quoted above raises a number of questions that are useful in analysing the problem.
For example (in reverse order to the above definition):
1. Who are the customers? 2. How should we assess if the product is viable? 3. Can we completely define the problem in terms that will be useful to the technical
design process? 4. What are the new/novel features that we hope to exploit to make our design bet-
ter than the existing competition and to build in flexibility to cater for future developments?
5. What is the best way to tackle the problem and how will this be managed?
These questions are used to gain more insight into the definition of the problem as explained below.
“chap02” — 2003/3/10 — page 7 — #2
Preliminary design 7
Problem definition
Information retrieval
Aircraft requirements
Configuration options
Initial sizing
Baseline evaluation
Constraint analysis
Refined baseline
Parametric analysis
Final baseline design
Baseline analysis
Project brief
Aircraft type specification
Trade studies
Fig. 2.1 The preliminary design flowchart
2.1.1 The customers Who are your ‘customers’? They are not only the purchasers of the aircraft; many groups of people and organisations will have an interest in the design and their expectations and opinions should be determined. For example, it would be techni-cally straightforward to design a new supersonic airliner to replace Concorde. The operating and technical issues are now well understood. However, the environmental lobby (who want to protect the upper atmosphere from further contamination) and the airport noise abatement groups have such political influence as to render the project unfeasible at this time. For all new designs it is necessary to identify all the influential people and find out their views before starting the project.
Who are the influential people?
• Obviously at the top of the list are the clients (the eventual purchasers of the aircraft). • Their customers (people who fly and use the aircraft, people who operate and
maintain it, etc.). • Your technical director, departmental head and line supervisor (these have a
responsibility for the company and its shareholders to make a reasonable return on investments).
• Your sales team (they know the market and understand customers and they will eventually have to market the aircraft).
“chap02” — 2003/3/10 — page 8 — #3
8 Aircraft Design Projects
• As a student, your academic supervisors and examiners (what is it that they expect to see from the project work).
It is useful to make a list of those people who you think will be important to the project and then find out what views they have. In academic courses the available timescale and facility to accomplish this consultation fully may not be available. In this case, set up your own focus groups and role-play to try to appreciate the expected opinions of various groups.
2.1.2 Aircraft viability It will be impossible to make rational decisions during the detailed design stages unless you can clearly establish how the product/aircraft is to be judged. Often this is easier said than done, as people will have various views on what are the important criteria (i.e. what you should use to make judgements). The aircraft manufacturing company and particularly its directors will want the best return on their investments (ROI). Unfortunately, so many non-technical issues are associated with ROI that it is too complicated to be used as a design criterion in the initial stages of the project. In the early days aircraft designers solved this dilemma by adopting aircraft mass (weight) as their minimising criteria. They knew that aircraft mass directly affected most of the performance and cost aspects and it had the advantage of being easy to estimate and control. Without any other information about design criteria, minimum mass is still a valid overall criterion to use. As more knowledge about the design and its operating regime becomes available it is possible to use a more appropriate parameter. For exam-ple, minimum direct operating cost (DOC) is frequently used for civil transport aircraft. For military aircraft, total life cycle cost (LCC), operational effectiveness (e.g. lethality, survivability, dependability, etc.) are more appropriate. High performance aircraft may be assessed by their operating parameters (e.g. maximum speed, turn rate, sink rate).
Some time ago A. W. Bishop of British Aerospace observed:
The message is clear – if everyone can agree beforehand on how to measure the effectiveness of the design, then the designer has a much simpler task. But even if everyone does not agree, the designer should still quantify his own ideas to give himself a sensible guide.
The procedure is therefore relatively simple – ask all those groups and individuals, who you feel are important to the project, how they would assess project effectiveness. Add any weightings you feel are appropriate to these opinions and decide for yourself what criteria should be adopted (or get the project group to decide if you are not working alone). Remember that the criteria must be capable of being quantified and related to the design parameters. Criteria such as ‘quality’, ‘goodness’ and ‘general effectiveness’ are of no use unless such a description can be translated into meaningful design parameters. For example, the effectiveness of a fighter aircraft may be judged by its ability to manoeuvre and launch missiles quicker than an opponent.
2.1.3 Understanding the problem It is unusual if the full extent of the problem is included in the initial project brief. Often the subtlety of the problem is not made clear because the people who draft the problem are too familiar with the situation and incorrectly assume that the design team will be equally knowledgeable. It is also found that the best solution to a problem is always
“chap02” — 2003/3/10 — page 9 — #4
Preliminary design 9
found by considering the circumstances surrounding the problem in as broad a manner as possible. This procedure has been called ‘system engineering’. In this approach, the aircraft is considered only as one component in the total operating environment. The design of the aircraft is affected by the design of all the components in the whole system. For example, a military training aircraft is only one element in the airforce flight/pilot training process. There are many other parts to such a system including other air-craft, flight simulators and ground schools. The training aircraft is also part of the full operational activity of the airforce and cannot be divorced from other aircraft in the service, the maintenance/service sector, the flight operations and other airport man-agement activities. On the other hand, the training aircraft itself can be considered as a total system including airframe, flight control, engine management, weapon on sensor systems, etc. All of these systems will interact to influence the total design of the aircraft.
Such considerations may lead to conflicts in the realisation of the project. For exam-ple, although the airforce may have a particular view of the aircraft, the manufacturers may have a different perspective. The airforce will only be focused on their aircraft but the manufacturers will want the aircraft to form part of a family of aircraft, which will have commercial opportunities beyond the supply to the national airforce. Within this context the aircraft may not be directly optimised for a particular role. The best overall configuration for the aircraft will be a compromise between, sometimes competing, requirements. It is the designer’s responsibility to consider the layout from all the dif-ferent viewpoints and to make a choice on the preferred design. He therefore needs to understand all aspects of the overall system in which the aircraft will operate. Some of the most notable past failures in aircraft projects have arisen due to designs initially being specified too narrowly. Conversely, successful designs have been shown to have considerable flexibility in their design philosophy.
Part of the problem definition task is to identify the various constraints to which the aircraft must conform. Such constraints will arise from performance and operational requirements, airworthiness requirements, manufacturing considerations, and limita-tion on resources. There will also be several non-technical constraints that must be recognised. These may be related to political, social, legal, economic, and commercial issues. However, it is important that the problem is not overconstrained as this may lead to no feasible solution existing. To guard against this it is necessary to be forceful in only accepting constraints that have been fully justified and their consequences under-stood. For technical constraints (e.g. field performance, climb rate, turn performance, etc.) there will be an opportunity to assess their influences on the design in the later stages (a process referred to as constraint analysis). Non-technical restrictions are more difficult to quantify and therefore must be examined carefully.
In general, the problem definition task can be related to the following questions:
• Has the problem been considered as broadly as possible? (i.e. have you taken a systems approach?)
• Have you identified all the ‘real’ constraints to the solution of the problem? • Are all the constraints reasonable? • Have you thoroughly examined all the non-technical constraints to determine their
suitability? (Remember that such constraints will remain unchallenged after this time.)
2.1.4 Innovation The design and development of a new aircraft is an expensive business. The people who invest in such an enterprise need to be confident that they will get a safe and profitable
“chap02” — 2003/3/10 — page 10 — #5
10 Aircraft Design Projects
return on their outlay. The basis for confidence in such projects lies in the introduction and exploitation of new technologies and other innovations. Such developments should give an operational and commercial advantage to the new design to make it competitive against existing and older products. Innovation is therefore an essential element in new aircraft design. The downside of introducing new technology is the increase in commercial risk. The balancing of risk against technical advantage is a fundamental challenge that must be accepted by the designers. Reduction of technological risk will be a high priority within the total design process. Empirical tests and analytical verification of the effects of innovative features are the designer’s insurance policy.
Innovation does not just apply to the introduction of new technology. Novel business and commercial arrangements and new operational practices may be used to provide a commercial edge to the new design. Whatever is planned, the designer must be able to identify it early so that he can adjust the baseline design accordingly.
The designers should be able to answer the following questions:
• What are the new technologies and other innovations that will be incorporated into the design?
• How will such features provide an advantage over existing/competing aircraft? • If the success of the innovation is uncertain, how can the risk to the project be
mitigated?
2.1.5 Organising the design process Gone are the days, if they ever existed, of a project being undertaken by an individual working alone in a back room. Modern design practice is the synthesis of many dif-ferent skills and expertise. Such combination of talent, as in an orchestra, requires organisation and management to ensure that all players are using the same source of information. The establishment of modern computer assisted design (CAD) software and other information technology (IT) developments allows disparate groups of spe-cialists and managers to be working on the same design data (referred to in industry as ‘concurrent engineering’).
The organisation of such systems demands careful planning and management. Design-build teams are sometimes created to take control of specific aircraft types within a multi-product company. The design engineer is central to such activity and therefore a key team player. It is essential for him to know the nature of the team structure, the design methods to be adopted, the standards to be used, the facilities to be required, and not least, the work schedules and deadlines to be met. Such consid-erations are particularly significant in student project work, as there are many other demands on team members. All students will have to personally time-manage all their commitments.
Whether the team is selected by an advising faculty member or is self-selected, stud-ents will face numerous challenges during the course of a design project. In most student design projects the organisation of the work is managed by the ‘design team’. Good team organisation and an agreed management structure are both essential to success. These issues are discussed in detail in Chapter 11, with particular emphasis to teaming issues in sections 11.2 and 11.3 respectively. When working in a team environment, students are advised to consult these sections before attempting to proceed with the preliminary design.
“chap02” — 2003/3/10 — page 11 — #6
Preliminary design 11
2.1.6 Summary The descriptions above indicate that there is a lot of work and effort to be exerted before it is possible to begin the laying-out of the aircraft shape. Each project is dif-ferent so it is impossible to produce a template to use for the design process. The only common factor is that if you start the design without a full knowledge of the problem then you will, at best, be wasting your time but possibly also making a fool of your-self. Use the comments and questions above to gain a complete understanding of the problem. Write out a full description of the problem in a report to guide you in your subsequent work.
An excellent way for design teams to begin this process of understanding the design problem is the use of the process known as ‘brainstorming’. This is discussed in more detail in section 11.2.5. Brainstorming is essentially a process in which all members of a team are able to bring all their ideas about the project to the table with the assurance that their ideas, no matter how far-fetched they may at first appear, are considered by the team. Without such an open mind, a team rarely is able to gain a complete understanding of the problem.
2.2 Information retrieval
Later stages of the design process will benefit from knowledge of existing work pub-lished in the area of the project. Searching for such information will involve three tasks:
1. Finding data on existing and competitive aircraft. 2. Finding technical reports and articles relating to the project area and any advanced
technologies to be incorporated. 3. Gathering operational experience.
2.2.1 Existing and competitive aircraft The first of these searches is relatively straightforward to accomplish. There are several books and published surveys of aircraft that can be easily referenced. The first task is to compile a list of all the aircraft that are associated with the operational area. For example, if we are asked to design a new military trainer we would find out what training aircraft are used by the major air forces in the world. This is published in the reviews of military aircraft, in magazines like Flight International and Aviation Week.
Systematically go through this list, progressively gathering information and data on each aircraft. A spreadsheet is the best way of recording numerical values for com-mon parameters (e.g. wing area, installed thrust, aircraft weights (or masses), etc.). A database is a good way to record other textural data on the aircraft (e.g. when first designed and flown, how many sold and to whom, etc.). The geometrical and technical data can be used to obtain derived parameters (e.g. wing loading, thrust to weight ratio, empty weight fraction, etc.). Such data will be used to assist subsequent technical design work. It is possible, using the graph plotting facilities of modern spreadsheet programs, to plot such parameters for use in the initial sizing of the aircraft. For instance, a graph showing wing loading against thrust loading for all your aircraft will be useful in select-ing specimen aircraft to be used in comparison with your design. Such a plot also allows
“chap02” — 2003/3/10 — page 12 — #7
12 Aircraft Design Projects
operational differences between different aircraft types to be identified. Categories of various aircraft types can be identified.
2.2.2 Technical reports As there are so many technical publications available, finding associated technical reports and articles can be time consuming. A good search engine on a computer-based information retrieval system is invaluable in this respect. Unfortunately, such help is not always available but even when it is, the database may not contain recent articles. Older, but still quite relevant, technical articles might also be easily missed when a search relies on computer search and retrieval systems. All computer search systems are very dependent on the user’s ability to choose key words which will match those used by whoever catalogued the material in the search system database. Success with such systems is often both difficult and incomplete as the user and the computer try to match an often quite different set of key words to describe a common subject. It becomes somewhat of a game, in which two people with different backgrounds try to describe the same physical object based on their own experiences. Often, a manual search of shelves in a library will product far better results in less time. Manual search is more laborious but such effort is greatly rewarded when appropriate material is found. This makes subsequent design work easier and it provides extra confidence to the final design proposal.
An excellent place to start a technical search is with the reference section at the end of each chapter in your preferred textbooks. Start with a text with which you are already familiar and track down relevant references. Do this either by using computer methods, or in a manual search of the library shelves. This can rapidly lead to an expanding array of background material as subsequent reference lists, in the newly found reports (etc.), are also interrogated.
2.2.3 Operational experience One of the best sources of information, data and advice comes from the existing area of operation appropriate to your project. People and organisations that are currently involved with your study area are often very willing to share their experiences. How-ever, treat such opinions with due caution as individual responses are sometimes not representative of the overall situation.
The best advice on information retrieved is to collect as much as you can in the time available and to keep your lines of enquiry open so that new information can be considered as it becomes available throughout the design process.
2.3 Aircraft requirements
From the project brief and the first two stages of the design process it is now possible to compile a statement regarding the requirements that the aircraft should meet. Such requirements can be considered under five headings:
1. Market/Mission 2. Airworthiness/other standards 3. Environment/Social
“chap02” — 2003/3/10 — page 13 — #8
Preliminary design 13
4. Commercial/Manufacturing 5. Systems and equipment
The detail to be considered under each of these headings will naturally vary depending on the type of aircraft. Some general advice for each section is offered below but it will also be necessary to consider specific issues relating to your design.
2.3.1 Market and mission issues The requirements associated with the mission will generally be included in the original project brief. Such requirements may be in the form of point performance values (e.g. field length, turn rates, etc.), as a description of the mission profile(s), or as opera-tional issues (e.g. payload, equipment to be carried, offensive threats, etc.). The market analysis that was undertaken in the problem definition phase might have produced requirements that are associated with commonality of equipment or engines, aircraft stretch capability, multi-tasking, costs and timescales.
2.3.2 Airworthiness and other standards For all aircraft designs, it is essential to know the airworthiness regulations that are appropriate. Each country applies its own regulations for the control of the design, manufacture, maintenance and operation of aircraft. This is done to safeguard its pop-ulation from aircraft accidents. Many of these national regulations are similar to the European Joint Airworthiness Authority (JAA) and US-Federal Aviation Administra-tion (FAA) rules.1,2 Each of these regulations contains specific operational requirements that must be adhered to if the aircraft is to be accepted by the technical authority (ultimately the national government from which the aircraft will operate). Airworthi-ness regulations always contain conditions that affect the design of the aircraft (e.g. for civil aircraft the minimum second segment climb gradient at take-off with one engine failed). Although airworthiness documents are not easy to read because they are legalistic in form, it is important that the design team understands all the implica-tions relating to their design. Separate regulations apply to military and civil aircraft types and to different classes of aircraft (e.g. very light aircraft, gliders, heavy air-craft, etc.). It is also important to know what operational requirements apply to the aircraft (e.g. minimum number of flight crew, maintenance, servicing, reliabil-ity, etc.). The purchasers of the aircraft may also insist that particular performance guarantees are included in the sales contract (e.g. availability, timescale, fuel use, etc.). Obviously all the legal requirements are mandatory and must be met by the aircraft design. The design team must therefore be fully conversant with all such conditions.
2.3.3 Environmental and social issues Social implications on the design and operation of the aircraft arise mainly from the control of noise and emissions. For civil aircraft such regulations are vested in separate operational regulations.3 For light aircraft, some airfields have locally applied operation restrictions to avoid noise complaints from adjacent communities. Such issues are becoming increasingly significant to aircraft design.
“chap02” — 2003/3/10 — page 14 — #9
14 Aircraft Design Projects
2.3.4 Commercial and manufacturing considerations Political issues may affect the way in which the aircraft is to be manufactured. Large aircraft projects will involve a consortium of companies and governments (e.g. Airbus). This will directly dictate the location of design and manufacturing activity. Such influ-ence may also extend to the supply of specific systems, engines and components to be used on the aircraft. If such restrictions are to be applied, the design team should be aware of them as early as possible in the design process.
2.3.5 Systems and equipment requirements Aircraft manufacture is no longer just concerned with the supply of a suitable airframe. All aircraft/engine and other operational systems have a significant influence in the overall design philosophy. Today many aircraft are not technically viable without their associated flying and control systems. Where such integration is to be adopted the design team must include this in the aircraft requirements. This aspect is particularly significant for the design of military aircraft that rely on weapon and other sensor systems to function effectively (e.g. stealth). Regulations for military aircraft usually fully describe the systems that the airframe must support.
2.4 Configuration options
With a fully described set of regulations, knowledge of existing aircraft data and a complete understanding of the problem, it is now possible to start the technical design tasks. Many project designers regard this stage as the best part of all the design pro-cesses. The question to be answered is simply this: Starting with a completely clear mind, what configurational options can you suggest that may solve the problem? For example, a two-seat light touring aircraft could be: side-by-side or tandem seating, high or low wing, tractor or pusher engine, canard or tail stabilised, nose or tail wheeled, conventional or novel planform (e.g. box wing, joined wing, delta, tandem), etc.
The following stage of the design process will sort through the ‘weird and wonderful’ configurations to eliminate the unfeasible and uncompetitive layouts. At this point in the layout process a quantity of ideas is needed and a judgement on their suitability will be left until later. With this in mind it is unnecessary to elaborate on an option past the point at which its characteristics can be appreciated. A good starting point for this work is to list the configurations that past and existing aircraft of this type have adopted. A brief synopsis of the strength and weaknesses of each option may be written so that improvements to the designs can be identified. Such analysis will also help in the concept-filtering phase that will follow.
In the conceptual design stage, designers have two options available for their choice of engines. Namely a ‘fixed’ (i.e. a specified/existing or manufacturers’ projected engine), or an ‘open’ design (in which the engine parameters are not known). In most cases, and definitely at later stages in the design process, the size and type of engine will have been determined. The aircraft manufacturer will prefer that more than one engine supplier is available for his project. In this way he can be more competitive on price and supply deadlines. For design studies in which the engine choice is open, it is possible to adopt what is known as a ‘rubber’ engine. Obviously, such engines do not exist in practice. The type and initial size of the rubber engine can be based on existing or
“chap02” — 2003/3/10 — page 15 — #10
Preliminary design 15
4000
260
300
340
380
5000 6000
Number of seats (3 class)
7000 8000 9000
Aircraft range (with reserves) (nm)
Fig. 2.2 Aircraft development programme (Boeing 777)
engine manufacturers’ projected engine designs. Using a rubber engine, the aircraft designer has the opportunity to scale the engine to exactly match the optimum size for his airframe. Such optimisations enable the designer to identify the best combination of airframe and engine parameters. If an engine of the preferred size is not available, in the timescale of the project, the designer will need to reconfigure the airframe to match an available engine. Rubber engine studies show the best combination of airframe and engine parameters for a design specification and can be used to assess the penalties of selecting an available engine.
Aircraft and engine configuration and size is often compromised at the initial design stage to allow for aircraft growth (either by accidental weight growth or by intent (air-craft stretch)). Such issues must be kept in mind when considering the various options. Most aircraft projects start with a single operational purpose but over a period of time develop into a family of aircraft. Figure 2.2 shows the development originally envis-aged by Boeing for their B777 airliner family. For military aircraft such developments are referred to as multi-role (e.g. trainer, ground support, etc.). It is important that designers appreciate future developments at an early design stage and allow for such flexibility, if desired.
2.5 Initial baseline sizing
At the start of this stage you will have a lot of design options available together with a full and detailed knowledge of the problem. It would be impossible and wasteful to start designing all of these options so the first task is to systematically reduce the number. First, all the obviously unfeasible and crazy ideas should be discarded but be careful that potentially good ideas are not thrown out with the rubbish. Statements and comments in the aircraft regulations and the problem definition reports will help to filter out uneconomic, weak and ineffective options. The object should be to reduce
“chap02” — 2003/3/10 — page 16 — #11
16 Aircraft Design Projects
the list to a single preferred option but sometimes this is not possible and you may need to take another one or two into the next design stage. Obviously, the workload will be increased in the next stages if more options are continued. Eventually it will be necessary to choose a single aircraft configuration. This will mean that all the work on the rejected options may be wasted.
This can be a very difficult part of the design process for a student design team. At this point, it is common for each member of the team to have invested a lot of time and energy into his or her own proposed design concept. It is often difficult to get team members to release their emotional ties to their own proposals and begin to embrace those of others or even to find a viable compromise. To get through this stage of the process both good team management and an effective means of comparing and evaluating all proposed concepts are required. Some of these difficulties are discussed in Chapter 11 (section 11.2). All proposed solutions to the design objective need to be given a fair and impartial assessment during the selection of the final concept. Obvi-ously, a compromise solution which draws upon key elements of every team member’s contributions will result in a happier set of team players. On the other hand, it is important that the selected concept embodies the best design elements that the team has developed. These must be chosen for the benefit of the overall design and not just to keep each member of the team happy.
Once decisions have been made on the configuration(s) to be further considered it is necessary to size the aircraft. A three-view general arrangement scale drawing for each aircraft configuration will be required. Little detail will be known at this stage about the aircraft parameters (wing size, engine thrust, and aircraft weight) so some crude estimates have to be made. This is where data from previous/existing aircraft designs will be useful. Although the new design will be different from previous aircraft, such inconsistencies can be ignored at this stage. Use representative values from one or a small group of the specimen aircraft for wing loading, thrust loading and aircraft take-off weight. It is also possible to use a representative wing shape and associated tail sizes.
The design method that follows is an iterative process that usually converges on a feasible configuration quickly. The initial general arrangement drawing, produced to match existing aircraft parameters, provides the starting point for this process. Even though your design is relatively crude at this stage it is important to draw it to scale making approximations for the relative longitudinal position of the wing and fuselage and the location of tail surfaces and landing gear.
Most aircraft layouts start with the drawing of the fuselage. For many designs the geometry of the fuselage can be easily proportioned as it houses the payload and cockpit/flight deck. These parameters are normally specified in the project brief. They can be sized using design data from other aircraft. The non-fuselage components (e.g. wing, tail, engines and landing gear) are added as appropriate. With a reasonable first guess at the aircraft configuration, the aircraft can be sized by making an initial estimate of the aircraft mass. Once this is completed it is possible to more accur-ately define the aircraft shape by using the predicted mass to fix the wing area and engine size.
2.5.1 Initial mass (weight) estimation The first step is to make a more accurate prediction of the aircraft maximum (take-off ) mass/weight. (Note: if SI units are used for all calculations it is appropriate to consider aircraft mass (kilograms) in place of aircraft weight (Newtons).)
“chap02” — 2003/3/10 — page 17 — #12
Preliminary design 17
Aircraft design textbooks4,5,6 show that the aircraft take-off mass can be found from:
MULMTO = 1 − (ME/MTO) − (MF/MTO)
where MTO = maximum take-off mass MUL
∗ = mass of useful load (i.e. payload, crew and operational items) M ∗ = empty mass E MF = fuel mass
(*When using the above equation it is important not to double account for mass com-ponents. If aircraft operational mass is used for ME, the crew and operational items in MUL would not be included. One of the main difficulties in the analysis at this stage is the variability of definitions used for mass components in published data on existing air-craft. Some manufacturers will regard the crew as part of the useful load but others will include none or just the minimum flight crew in their definition of empty/operational mass. Such difficulties will be only transitional in the development of your design, as the next stage requires a more detailed breakdown of the mass items.)
The three unknowns on the right-hand side of the equation can be considered separately:
(a) Useful load The mass components that contribute to MUL are usually specified in the project brief and aircraft requirement reports/statements.
(b) Empty mass ratio The aircraft empty mass ratio (ME/MTO) will vary for different types of aircraft and for different operational profiles. All that can be done to predict this value at the initial sizing stage is to assume a value that is typical of the aircraft and type of operation under consideration. The data from existing/competitor aircraft collected earlier is a good source for making this prediction. Figure 2.3 shows how the data might be viewed. Alternatively, aircraft design textbooks often quote representative values for the ratio for various aircraft types.
Max. take-off mass (MTO)
Empty mass (ME)
Three engines Four engines
Slope (ME /MTO) = 0.55
More than two = 0.47
Two engines
Two engines
Fig. 2.3 Analysis of existing aircraft data (example)
“chap02” — 2003/3/10 — page 18 — #13
18 Aircraft Design Projects
a – take-off, b – climb, c – cruise, d – step climp, e – continued cruise, f – descent, g – diversion, h – hold, i – landing at alternate airstrip.
a
b
c d e
f g
h
i
Fig. 2.4 Mission profile (civil aircraft example)
(c) Fuel fraction For most aircraft the fuel fraction (MF/MTO) can be crudely estimated from the modified Brequet range equation:
MF
MTO = (SFC) · 1
(L/D) · (time)
where (SFC) = engine specific fuel consumption (kg/N/hr) (L/D) = aircraft lift to drag ratio (time) = hours at the above conditions
The mission profile will have been specified in the project brief. Figure 2.4 illustrates a hypothetical profile for a civil aircraft.
This shows how the mission profile consists of several different segments (climb, cruise, etc.). The fuel fraction for each segment must be determined and then summed. Reserve fuel is added to account for parts of the mission not calculated. For example:
(a) for the fuel used in the warm-up and taxi manoeuvres, (b) for the effects on fuel use of non-standard atmospheric conditions (e.g. winds), (c) for the possibility of having to divert and hold at alternative airfield when
landing.
The last item above is particularly significant for civil operations. In such applications designers sometimes convert the actual range flown to an equivalent still air range (ESAR) using a multiplying factor that accounts for all of the extra (to cruise) fuel.
When using the Brequet range equation it must be remembered that both engine (SFC) and aircraft (L/D) will be different for different flight conditions. These vari-ations arise because the aircraft speed, altitude, weight and engine setting will be different for each flight segment. Typical values for (SFC) can be found in engine data books7 or from aircraft and engine textbooks4,8 for the type of engine to be used.
The aircraft lift to drag ratio (L/D) will vary and be dependent on aircraft geometry (particularly wing angle of attack). Such values are not easily available for the aircraft in the initial design stage. However, we know that previous designers have tried to achieve a high value in the principal flight phase (e.g. cruise). We can use the fact that in cruise
“chap02” — 2003/3/10 — page 19 — #14
Preliminary design 19
‘lift equals weight’ and ‘drag equals thrust’. We can therefore transpose (L/D) into (W /T ). Both aircraft weight and engine thrust (at cruise) could be estimated from our specimen aircraft data. This value will be close to the maximum (L/D) and relate only to the cruise condition. At flight conditions away from this point the value of (L/D) will reduce. It must be stressed that the engine thrust level in cruise will be substantially less than the take-off condition due to reduced engine thrust setting and the effect of altitude and speed. This reduction in thrust is referred to as ‘lapse rate’. Engine specific fuel consumption will also change with height and speed. Values for (L/D) vary over a wide range depending on the aircraft type and configuration. Typical values range from 30 to 50 for gliders, 15 to 20 for transport/civil aircraft, 12 to 15 for smaller aircraft with reasonable aspect ratio and less than 10 for military aircraft with short span delta wing planforms. Aircraft design textbooks are a source of information on aircraft (L/D) if the values cannot be estimated from the engine cruise conditions and aircraft weight.
(Time) is usually easy to specify as each of the mission segments is set out in the project brief (mission profiles). Alternatively, it can be found by dividing the distance flown in a segment by the average speed in that segment.
2.5.2 Initial layout drawing Obviously, all the above calculations require a lot of ‘guesstimation’ but at least at the end we will have a better estimate of the aircraft maximum take-off mass than previ-ously. This value can then be used in conjunction with the previously assumed values for wing and thrust loading to refine the size of the wing and engine(s). The original concept drawing can be modified to match these changes. This drawing becomes the initial ‘baseline’ aircraft configuration.
2.6 Baseline evaluation
The methods used up to this point to produce the baseline aircraft configuration have been based mainly on data from existing aircraft and engines. In the next stage of the design process it is necessary to conduct a more in-depth and aircraft focused analysis. This will start with a detailed estimation of aircraft mass. This is followed by detailed aerodynamic and propulsion estimates. With aircraft mass, aerodynamic and engine parameters better defined it is then possible to conduct more accurate performance estimations. The baseline evaluation stage ends with a report that defines a modified baseline layout to match the new data. A brief description of each analysis conducted in this evaluation stage is given below.
2.6.1 Mass statement Since the geometrical shape of each part of the aircraft is now specified, it is possible to make initial estimates for the mass of each component. This may be done by using empirical equations, as quoted in many design textbooks, or simply by assuming a value for the component as a proportion of the aircraft maximum or empty mass. Such ratios are also to be found in design textbooks or could match values for similar aircraft types, if known. The list below is typical of the detail that can be achieved.
“chap02” — 2003/3/10 — page 20 — #15
20 Aircraft Design Projects
Generating a mass statement like this one is the first task in the baseline evaluation phase.
Wing (MW) Tail (MT) Body (MB) Nacelle (MN) Landing gear (MU) Control surfaces (MCS) ∑
total aircraft structure (MST)
Engine basic (dry) Engine systems Induction (intakes) Nozzle (exhaust) Installation ∑
total propulsion system (M P)
Aircraft systems and equipment (MSE) ∑ aircraft empty mass = M E = M ST + M P + M SE
Operational items (MOP) ∑ aircraft operational empty mass (M OE) = M E + M OP
Crew* (MC) Payload (MPL) Fuel (MF) ∑ ∑
aircraft take-off mass (M TO ) = M OE + M C + M PL + M F
(*For some military aircraft mass statements, the crew are considered to form part of the operational items and their mass is added to aircraft OEM.)
The main structural items in the list above (e.g. wing, fuselage, engine, etc.) can be estimated using statistically determined formulae which can be found in most aircraft design textbooks. (Note: if you are working in SI units be careful to convert mass values from historical reports, journals, and current US textbooks to kilograms (1 kg = 2.205 lb).) Many of these mass items are dependent on MTO, therefore estimations involve an iterative process that starts with the assumed value of MTO, as estimated in the initial sizing stage. Spreadsheet ‘solver’ methods will be useful when performing this analysis.
At the early design stages, the estimation of mass for some of the less significant (and smaller) components may be too time consuming to calculate in detail (e.g. tail, landing gear, flight controls, engine systems and components, etc.). To speed up the evaluation process, these can be estimated by assuming typical percentage values of MTO, as mentioned above. Such values can be found from existing aircraft mass breakdowns, if available, or from aircraft design textbooks.
At the final stages of the conceptual phase an aircraft mass will be selected which is slightly higher than the estimated value of MTO. This higher weight is known as the ‘aircraft design mass’. All the structural and system components will be evaluated using the value for the aircraft design weight as this provides an insurance against weight growth in subsequent stages of the design process. For aircraft performance estimation, the mass to be used may be either the MTO value shown above or some-thing less (e.g. military aircraft manoeuvring calculations are frequently associated
“chap02” — 2003/3/10 — page 21 — #16
Preliminary design 21
with the aircraft operational empty mass plus defensive weapons and half fuel load only).
2.6.2 Aircraft balance With the mass of each component estimated and with a scale layout drawing of the aircraft it is possible, using educated guesses, to position the centre of mass for each component. This will allow the centre of gravity of the aircraft in various load condi-tions (i.e. different combinations of fuel or payload) to be determined. It is common practice to estimate the extreme positions (forward and aft) so that the trim loads on the control surfaces (tail/canard) and the reaction loads on the undercarriage wheels can be assessed.
Up to this point in the design process, the longitudinal position of the wing along the fuselage has been guessed. As part of the determination of the aircraft centre(s) of gravity, it is possible to check this position and, by iteration, to reposition it to suit the aircraft lift and inertia force (i.e. mass × acceleration) vectors. This process is referred to as ‘aircraft balancing’. As moving the wing will affect the position of the aircraft centre of gravity and the wing lift aerodynamic centre from the datum, several iterations may be required. There are several methods that can be used to reduce the complications inherent in this iteration. The simplest method sets the position of the aircraft operational empty mass relative to a chosen point (per cent chord aft of the wing leading edge) on the wing mean aerodynamic chord line. To start the process the aircraft operational empty mass components are divided into two separate groups:
(a) Wing mass group (MWG ) (and associated components) – this will include the wing structure, fuel system (if the fuel is housed in the wing), main landing gear unit (even if it is structurally attached to the fuselage), wing mounted engines and all wing attached systems.
(b) Fuselage mass group (MFG) (and associated components) – this group will include the fuselage structure, equipment, cockpit and cabin furnishings and systems, operational items, airframe services, crew, tail structure, nose landing gear and fuselage mounted engines and systems.
Note: if the position of wing mounted engines is linked to internal fuselage layout requirements (e.g. propeller plane be in line with non-passenger areas) then these masses should be transferred to the fuselage group.
Obviously all the aircraft components relating to the aircraft operational empty mass must be included in either of the above groups (i.e. MOE = MWG +MFG). It is important to check that none of the component masses has been omitted before starting the balancing process.
It is possible to determine the centres of mass separately for each of the two mass groups above. The distance of the wing group centre of mass from the leading edge of the wing mean aerodynamic chord (MAC) is defined as XWG (see Figure 2.5a).
The next stage is to select a suitable location for the centre of gravity of the aircraft operational empty weight, on the wing mean aerodynamic chord. If the centre of gravity is too far aft or forward then the balancing loads from the tail (or canard) will be high. This will result in a requirement for larger tail surfaces and thereby increased aircraft mass and trim drag. For most conventional aircraft configurations, a centre of gravity position coincident with the 25 per cent MAC position behind the wing leading edge is considered a good starting position. If it is known that loading the aircraft from the operational empty mass will progressively move the aircraft centre of gravity forward,
“chap02” — 2003/3/10 — page 22 — #17
22 Aircraft Design Projects
then a 35 per cent MAC position would be a better starting point. Such cases arise on civil airliners with rear fuselage mounted engines. Conversely, a 20 per cent MAC would be chosen for designs with mainly aft centre of gravity movements. For aircraft flying at supersonic speed the centre of lift will be at about the 50 per cent MAC position. This must be carefully allowed for when selecting the operational mass position. The location of the chosen operational empty mass location with respect to the leading edge of the wing mean aerodynamic chord is defined as XOE.
It is possible to take moments of the aircraft masses shown in Figure 2.5a. By rear-ranging the moment equation, the position of the fuselage group mass relative to line XX can be calculated. The resulting equation is shown below:
XFG = XOE + (XOE − XWG )(MWG /MFG )
Overlays of the separate wing and fuselage layouts provide the best method of fixing the wing relative to fuselage. On a plan view of the wing, determine the position of the wing MAC and its intersection with the wing leading edge (line XX). Also, on this drawing show the position of the wing group centre of mass, see Figure 2.5b.
Measure the distance XWG from this drawing and use it in the formula above together with the selected value of XOE and the calculated wing and fuselage group masses (MWG and MFG), to evaluate the distance XFG. On a plan view of the fuselage, determine the position of the fuselage group mass centre (using any convenient datum plane) then draw a line XX at a distance XFG forward of this position, as shown in Figure 2.5c.
Overlay the wing and fuselage diagram lines XX. This is the correct location of wing and fuselage to give the aircraft operational centre of gravity at the previously selected position on the wing MAC. It is not unusual to discover by this process that the originally assumed position of the wing relative to the fuselage, on the aircraft layout drawing, is incorrect and must be changed.
With the aircraft balanced, it is now possible to determine the range of aircraft centre of gravity movement about the operational empty position and to assess the effect of this on the tail sizing. Obviously, it is preferable to design for small movements of the aircraft centre of gravity to ensure the control forces are small. To do this, the disposable items of mass (fuel and payload) should be centred close to the aircraft operational empty centre of gravity position as practical.
At this stage in the development of the aircraft geometry it is possible to position the undercarriage units. The process involves geometric and load calculations associated with the aircraft mass and centre of gravity range. The main units must allow for adequate rotation of the aircraft on take-off and in the landing attitude. When the aircraft is in the maximum tail down attitude, the aircraft rearmost centre of gravity position must be forward of the wheel reactions. This will ensure the aircraft does not stay in this position. The loads on the main and nose units can be determined by simple mechanics. Make sure that the nose wheel load is not excessive as this will require a large tailplane force to lift the nose on take-off. On the other hand, if the load is too small on the nose wheel it will not generate an effective steering force. The forces determined for each unit will dictate the tyre size commensurate with the allowable tyre pressure and runway point-load capability. Several aircraft design textbooks include undercarriage layout guidelines.
2.6.3 Aerodynamic analysis At the same time as the mass and balance estimation is made, or sequentially after if you are working alone, it is possible to make the initial estimations for the baseline aircraft aerodynamic characteristics (drag and lift). The aircraft drag estimation, like mass,
“chap02” — 2003/3/10 — page 23 — #18
Preliminary design 23
(a)
(b)
X XFG
MFG
MOE = L
M WG
XOE
X WG
X
Required position of aircraft MOE centre of gravity
C T
CT
XOE C T
X
X
C/2 Chord line
Wing MAC
Position of M WG centre of gravity
Assumed position of aircraft MOE behind MAC leading edge
Intersection of wing MAC with LE
Position of MFG centre of gravity
XFG As calculated from formulaX
X (c)
Fig. 2.5 Aircraft balance methodology (diagrams a, b and c)
“chap02” — 2003/3/10 — page 24 — #19
24 Aircraft Design Projects
can be broken down into individual components (e.g. wing, body, tail, etc.) and then summed. Allowance for interference effects between components must also be added to the value. Textbooks on aerodynamics and aircraft design provide several different methods for performing such calculations. The drag of the aircraft will eventually be used in the performance estimations, therefore it will be necessary to determine values at different flight conditions (e.g. take-off, climb, cruise, etc.). These calculations will involve the aircraft in different configurations with regard to the deployment of landing gear and flap extensions. The aircraft will also be at different speeds and altitudes for each condition. This affects the Reynolds number used in the drag calculations and other parameters.
You may find it useful to do the drag calculations during this stage in terms of ‘drag area’ rather than in the coefficient form. This effectively ‘dimensionalises’ the drag of each component, and ultimately the whole aircraft, in terms of the area of a flat plate that would have an equivalent drag to the component. As might be expected, this method is sometimes referred to as ‘the equivalent flat plate area’. Drag area gives a better visualisation of the effectiveness (or otherwise) of the various components and their contributions to the total aircraft drag. It also provides an indication of the influence of the geometrical parameters of the component to its drag. In the early design stages the selection of aircraft gross wing area (i.e. reference area) is very tentative, as it has not been checked against the performance requirements. Using it as a reference area in drag coefficient form may be regarded as premature. On the other hand, in the determination of aircraft lift many of the established methods are based on the manipulation of lift coefficients. It is therefore impossible to avoid the potential wing area confusions for the estimation of lift.
As with drag estimation, it is necessary to determine lift coefficients at different operating conditions (i.e. various flap deflection angles – e.g. take-off and landing settings). Use design data from existing aircraft to initially set values for flap deflections and wing planform (flap span ratio) geometry. At later stages, when more detailed aerodynamic analysis of wing and other aircraft components has been completed, it will be necessary to select specific flap angles to suit your particular aircraft operational requirements.
2.6.4 Engine data Before a detailed performance estimation can be made it is essential to have repre-sentative engine performance charts (or data) available. From the problem definition phase either the engine or the engine type may be known. The initial sizing work will have provided an estimate of the engine take-off thrust. To undertake aircraft perform-ance calculations it is necessary to know what thrust (and SFC) the engine will give at thrust settings other than take-off (e.g. at continuous climb, cruise, etc.). It will also be necessary to know the effect of aircraft altitude and speed on the engine parame-ters. For some military aircraft it is also necessary to understand what effect the use of reheat (afterburning) will have on engine performance. For existing engines, data may be available from the engine manufacturers but sometimes it is difficult to obtain this data. Engine manufacturers are reticent to release technical detail for commercial reasons. It is also often impossible for them to provide the data in the form that stu-dents can use as their engine performance is held in extensive databases that require flight data as input. For many new aircraft projects a new engine is required, therefore manufacturers’ data is not available. In these cases predictions based on similar engine types have to be made. Aircraft design and engine textbooks4,8 often contain data on which to make such predictions.
“chap02” — 2003/3/10 — page 25 — #20
Preliminary design 25
2.6.5 Aircraft performance With aircraft mass, drag, lift and engine characteristics known it is a relatively straight-forward process to make initial estimates of aircraft performance. This is done for each flight segment separately (climb, cruise, dash, loiter, descent, combat, etc.). The field performance (take-off and landing) is also required. Many textbooks are available on aircraft performance estimation.4,5,6,9 These can be used, with appropriate simplifying assumptions, to estimate performance values.
The results from the performance estimates are compared to the aircraft require-ments. It is now that the original estimates for wing area and thrust are re-evaluated. Changes in these values are often necessary to obtain aircraft performance to meet the requirements. It is essential that new values for wing area and engine thrust are selected that allow such compliance but not too much in excess as this will make the design inefficient. As aircraft mass, drag, lift and engine characteristics are directly affected by changes in wing and engine size it will be necessary to repeat all the pre-vious initial estimates for the baseline aircraft. This is a laborious task but the use of modern spreadsheet methods does assist in such iterative processes.
2.6.6 Initial technical report At the end of the baseline evaluation stage you should have a detailed knowledge of an aircraft configuration that will meet the original problem specifications. However, this configuration is unlikely to be ‘optimum’.
It is now possible to produce a report which contains a scale drawing of the modified baseline configuration, a detailed mass breakdown, drag and lift assessments for each operational configuration, and engine and aircraft performance predictions for all flight segments. Some examples of these calculations are shown in the project studies that follow (Chapters 4 to 10). Subsequent stages in the conceptual design process are aimed at improving the aircraft configuration to make a more efficient design and to address non-technical factors.
2.7 Refining the initial layout
At this point in the design process we have an aircraft layout that has been based mainly on crude estimates taken from previous aircraft designs. We also have not assessed the overall problem definition with regard to any of the aircraft design parameters. It is now time to improve this situation and provide more confidence in the aircraft layout. From the previous stage, we have enough geometric and configurational details to make detailed estimates. These include the mass of aircraft components (giving a mass statement), aerodynamic coefficient assessments (lift and drag) and some knowledge of engine performance (thrust and fuel flow) at various operating conditions. With this data, it is possible to undertake a more detailed analysis of the aircraft design. The following studies will allow us to progressively adjust the aircraft geometrical and layout features to better match the problem constraints and to improve the aircraft effectiveness as judged by the overall assessment criteria. These studies will also allow us to test the sensitivity of the problem constraints against the aircraft configuration. Two design processes are used:
• Constraint analysis • Aircraft trade studies
“chap02” — 2003/3/10 — page 26 — #21
26 Aircraft Design Projects
Although these methods are separately described in the following sections, they are often used in conjunction, and repeatedly, to refine the baseline layout and to assess the significance of the problem constraints.
2.7.1 Constraint analysis From our earlier work on understanding the problem (section 2.1.3) we identified several constraints that must be satisfied by the aircraft design. Many of the specified constraints are related to the performance of the aircraft (for example, minimum speeds or climb gradients). The requirement will be written in the form:
At a specified aircraft flight condition and configuration, the aircraft must demon-strate a performance no less than a specified value (e.g. climb gradient better than 0.024 with an engine failed on take-off with aircraft at max. weight).
The specified values are known as constraints. For most aircraft projects there are several constraints related to field performance (e.g. take-off, climb, balanced field length, stall speed, landing characteristics) and several linked to other mission seg-ments (e.g. minimum cruise speed, climb rates/times, turn rates, specific excess power, loiter speed, dash speed, etc.). Each aircraft project will have a different set of such con-straints. The design process known as constraint analysis is used to assess the relative significance of these constraints and their influence on the aircraft configuration.
It is common to display the constraint boundaries on a graph of aircraft thrust to weight ratio (T /W ) against wing loading (W /S) as illustrated in Figure 2.6.
The area of the graph is called the design space. Constraint lines can be plotted to show the relationship of the two aircraft parameters at the constraint value. These lines represent boundaries between the unacceptable and feasible regions of the design space. It is usual to attach hatching to the line on the unacceptable side of the constraint. In this way, the design space is divided into the two regions by each constraint. A com-bination of constraints graphically identifies the feasible design space (shown shaded on Figure 2.6). Any combination of (T /W ) and (W /S) values is possible within this region. The selection of a particular pair of values will dictate the size and nature of the aircraft layout. For example, given a reasonable estimate of the aircraft maximum weight (take-off mass), the engine thrust will be fixed by the T /W value and the gross wing area fixed by the W /S value.
Wing loading (W/S )
Feasible design area
ÂOptimumÊ design points
Climb
Landing
g
Thrust loading (T/W )
Take-off
Turnin
Fig. 2.6 Generalised constraint diagram
“chap02” — 2003/3/10 — page 27 — #22
Preliminary design 27
The original guesstimate of wing and thrust loading made in the initial sizing process can be checked to see if it lies in the feasible region. If not, the design point can be moved into the acceptable design space. The position of the design point in the feasible region, relative to the constraint boundaries, indicates the efficiency of the aircraft. It is desirable that the selected point lies close to the minimum value of T /W to reduce engine size and close to the maximum value of W /S to reduce wing size. Often such a point lies on the intersection of the constraint lines.
Although the constraint analysis will give guidance on the selection of the design point as described above, it is principally used to show the significance of the specified constraints to the aircraft configuration. For example, it may show that a constraint is too demanding (over-riding all other constraints) and significantly reduces the allow-able design space. This will make the design excessively large and potentially not viable. In such cases, it is important to inform those who originally prescribed the constraint of the consequences. It is often possible to get a relaxation of the offending constraint to make the design more effective. The diagram will clearly show which of the constraints are inconsequential (i.e. falling well inside the unacceptable design space) and which are ‘active’ (i.e. forming the boundary of the feasible region). As described above, the ‘optimum’ design point will lie on the intersection of constraint lines, therefore the best set of constraints is that in which several intersect at roughly the same design point. This is sometimes referred to as a ‘well-balanced’ design.
As mentioned in the introduction to this section, the performance constraints are not the only ones to be imposed on the design. For example, for a naval aircraft operating from a carrier there will be geometrical limits that are set by the size and shape of the deck elevators. Such constraint may make it impossible to select the design point shown on the performance constraint diagram. However, the constraint analysis will provide a means of assessing the design penalty that has to be accepted for imposing the non-performance constraint.
The fundamental theory on which the constraint analysis is based involves the manip-ulation of the energy state (kinetic and potential) of the aircraft and its relationship to the available excess power, Ps:
Ps = [(T − D)/W ]V (where weight = W = Mg)
The excess power can be used to either climb (dh/dt) or accelerate ([V /g] · [dV /dt]) the aircraft separately, or in combination within the limit of the excess power:
Ps = [(T /W ) − (D/W )]V = dh/dt + ([V /g] · [dV /dt]) This equation can be rearranged to provide a relationship between aircraft thrust and wing loadings (T /W , W /S). This is done by expanding the drag term (D = qSCD) and the drag coefficient (CD = CDO +k1C2
L) and writing CL in terms of lift (CL = nW /qS):
(T /W ) = [qCDO/(W /S) + k1 · n2 · (W /S)/q] + (1/V ) · dh/dt + ([1/g] · [dV /dt]) where q = 0.5ρV 2 and n = L/W
The thrust to weight ratio (T /W ) can be normalised, for various different operating conditions, to the equivalent sea-level static thrust and maximum mass condition. This is done by the application of the following factors: thrust lapse rate = α = T /TSSL and aircraft weight fraction = β = W /WTO, where, T and W are the thrust and weight at the operating condition, TSSL is the static sea-level total engine thrust and WTO is the
“chap02” — 2003/3/10 — page 28 — #23
28 Aircraft Design Projects
aircraft take-off weight. Rearranging the (T /W ) equation above and introducing the parameters α and β we get:
(TSSL/WTO) = (β/α)[{(q/β)(CDO/(WTO/S)} + {[k1 · n2 · (WTO/S)]/(q/β)}] + (1/V ) · dh/dt + ([1/g] · [dV /dt])
Several textbooks explain the constraint analysis process in detail.6,8 Aircraft and engine characteristics in the form of drag coefficients (CDO and k1), weight (W ), speed (V ), altitude/density (ρ), normal acceleration (n) are set as constants for a par-ticular flight/operation segment. Obviously, this may be a somewhat crude assumption as many of the constants are affected by the aircraft and engine size, and operating characteristics but it is possible to use iteration to reduce the errors once the critical constraints have been identified.
Each of the aircraft constraints can be analysed with the above equation (e.g. the constant speed climb rate of 2.4 per cent mentioned above would set the last term to zero and the penultimate one to 0.024). The resulting expression can be solved for (T /W ) using increasing values of (W /S) as the variable input. The results can then be plotted to give a boundary to the feasible design region.
A graph of the form shown in Figure 2.7 can be drawn showing the extent of the specified constraints on a (WTO/S) versus (TSSL/WTO) graph. The area above and between the constraint lines defines the feasible design space. Any combination of (WTO/S) and (TSSL/WTO) in this area is allowable but the best design will lie towards the bottom and right-hand side of the diagram.
As mentioned above, constraint analysis is a crude analytical tool as it cannot easily take into account changes in the basic geometry of the aircraft (e.g. aspect ratio). It does, however, serve the principal purpose of identifying any constraint that is unduly
Thrust / Weight ratio
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0 0 100 200 300 400 500 600
Design point
Climb requirement
Combat turn @FL250
Combat turn @ SL
Take-off
V approach 62% MTOV approach 90% MTO Land run 62% MTO
L159
MiG AT
S212 L139
M339
T45
K8
BAE Hawk
Land run 90% MTO
Wing loading (kg/m2) Actual aircraft
Fig. 2.7 Constraint diagram (example)
“chap02” — 2003/3/10 — page 29 — #24
Preliminary design 29
or harshly influencing the design configuration. For example, if the vertical line in Figure 2.7 (the approach speed constraint) had been positioned too far to the left of the diagram it would have eliminated much of the available design space. It would also have made the two lower constraints ineffective.
A secondary purpose for the constraint diagram is to use it in conjunction with the wing and thrust loading data from your specimen aircraft list as shown in Figure 2.7. Note: plotting existing data on the diagram drawn for your aircraft configuration is somewhat misleading as other designs are constrained differently but the diagram does hold some general interest. It illustrates how existing aircraft have been constrained. If existing aircraft lie outside your predicted design area this raises a further discussion with your clients about the values chosen for your constraints.
As stated above, many of the aircraft and engine characteristics are held as constant values in the constraint analysis. It is important to conduct trade (sensitivity) studies on any assumptions used to generate these values (e.g. drag or thrust). Those shown to be critical should be more accurately predicted. Be careful not to get too involved in external discussions before such validation work has been completed.
The outcome of the constraint analysis may be the selection of a different geometry for the aircraft than currently specified for the baseline configuration. When you are ready to proceed a little backtracking will be necessary to re-evaluate the new configuration using methods that have been developed in the previous design stages.
2.7.2 Trade-off studies The outcome of the constraint analysis may be the selection of a different geometry for the aircraft than currently specified for the baseline configuration. Since it will be necessary to recalculate the baseline aircraft with these new values, it is worth considering other changes that might be beneficial to the design. For example, the wing aspect ratio may have been selected from arbitrary data from other aircraft. It would be appropriate to assess this decision as more detailed analysis of the design is now available. Such methods are referred to as trade-off or sensitivity studies. They generally investigate the variation of a single parameter while keeping all others constant. Multi-variable investigations are referred to as parametric studies, these are explained later in this chapter.
Trade studies can take many different forms. They can be used, as implied above, to assess the selection of aircraft geometrical features (aspect ratio, taper ratio, wing thickness/sweep combinations, etc.). They can be used to indicate the sensitivity of the design to assumptions made in the analysis (e.g. the affect of the assumed extent of laminar flow, the affect of high strength or composite materials, etc.). Alternatively, they may be used to show the effect of changing design requirements (e.g. field lengths, mass variations, engine characteristics, etc.). All of these have the objective of building confidence in the design predictions on which the aircraft configuration is based.
Some examples of trade studies based on the example aircraft (described in section 2.9) are shown in Figures 2.8 to 2.10. These figures are introduced here only as examples of the type of studies that are possible. The first example of a trade study (Figure 2.8) shows the sensitivity for the choice of wing taper ratio.
Various aircraft parameters are plotted which show variations with increasing wing taper ratio. A lower limit for taper ratio may be imposed to avoid wing tip stall resulting from low Reynolds number flow over the outer wing sections. (This provides a good example of considerations other than those of the trade-off study also being influential.) The second Figure (2.9) shows the effect of wing aspect ratio.
“chap02” — 2003/3/10 — page 30 — #25
30 Aircraft Design Projects
588
584
580
0.25 0.30 Wing taper ratio
0.35
0.5%
5430
5420
0.25 0.30 Wing taper ratio
Fuel wt (Ib)
0.35
0.5%
6.48
6.44
0.25 0.30 Wing taper ratio
Seat-mile cost (c )
0.35
0.5%
25700
25600
25400
0.25 0.30
Operational empty wt (lb)
0.35
0.5%
9.6
9.5
0.25 0.30
A/C price ($M )
0.35
1.0%
Wing gross area (sq. ft)
Wing taper ratio Wing taper ratio
Fig. 2.8 Wing taper ratio sensitivity study (example)
DOC per flight ($) Wing area (sq. ft)
Op. empty weight (lb)
2280 2570 620
2560 610
2260
2550 600
2240
2540 590
2220
2530 580
2000
1%
OEW
DOC
1%
1%Wing area
Optimum AR
8.5 9.0 9.5 10.0 10.5 Wing aspect ratio
Fig. 2.9 Trade-off study of wing aspect ratio (example)
The ‘kinks’ in the plots show the effect of two separate constraint boundaries on the problem (i.e. the influence of span loading on climb performance and the activation of the WAT (weight, altitude and temperature) restrictions at low values of aspect ratio). For this civil airliner the optimum aspect ratio for minimum direct operating cost (DOC) is seen to be about 9.32.
“chap02” — 2003/3/10 — page 31 — #26
Preliminary design 31
Gross wing area (sq. ft)
610
Seats = 44 Field = 5100 ft
590 Stage = 960 nm Wing AR = 9.27
570
550
4800 4900 5000 5100
Landing field length (ft)
Fig. 2.10 Sensitivity graph – landing versus wing area (example)
Varying an aircraft geometrical parameter, like taper and aspect ratio, is relatively easy as these are inputs to the aerodynamic, mass and performance equations. Altering a main operational parameter (e.g. landing field length, LFL) involves much more iteration. It is possible to show the sensitivity of such parameters using trade study methods as illustrated in Figure 2.10 (LFL versus aircraft wing area).
2.8 Refined baseline design
Up to this point in the design sequence the aircraft has retained many of the features assumed from the initial configuration stage. For example, the wing planform shape will not have been carefully considered. It is at this stage in the development of the baseline layout that more rational decisions can be taken on the size and shape of all the aircraft components. One of the first decisions to be considered is the aircraft balance. As we now have a detailed mass statement for the aircraft that shows the mass of all the major components and a scale drawing of the aircraft layout, it is possible to locate the centres of each component mass. It is then possible to predict the positions of the aircraft centre of gravity more accurately than previously. (Note: ‘positions’ is plural as the aircraft can be loaded and flown at different payload or fuel masses.) The final objective of this analysis is to assess the longitudinal position of the wing relative to the fuselage to achieve an acceptable static margin. (Refer to textbooks on aircraft stability and control or aircraft design for the definition of static margin.)
As mentioned earlier (Section 2.6.2), it is acceptable to position the aircraft centre of gravity for subsonic aircraft just ahead of the wing mean aerodynamic quarter chord position. For supersonic flight the wing lift moves back to about half chord but remem-ber that these aircraft also have to fly subsonically so a compromise or other technical measures (e.g. fuel management systems) may have to be introduced. The locations of the landing gear units is directly related to the aircraft centre of gravity, therefore repositioning of these units may also be considered at this time.
“chap02” — 2003/3/10 — page 32 — #27
32 Aircraft Design Projects
Detailed assessment of the ‘packaging’ of the aircraft is another task that can be considered at this point in the design process. An analysis of the space required and available for all components and systems must be made. This will include a check on the space available within the structure to hold all the fuel that is required to fly the mission. Will the landing gear have space for retraction? Will the pilot have adequate sight lines from the cockpit (these are mandatory)? Engine intake and exhaust flow requirements can be assessed to dictate the cross-sectional area distributions. Radar and other flight sensors must be suitably positioned. Will servicing maintenance and turn-round be easy to accomplish? In other words, it is possible at this stage to undertake a full review of all the installation and space aspects relating to the layout.
As mentioned earlier, the wing planform must be examined carefully. This will be reconsidered in work to be done later (e.g. parametric and trade-off studies) but for now some simple aerodynamic and mass analyses will provide evidence to make sensible judgements. For example, would an increase (or decrease) in wing aspect ratio be advisable. What about wing sweepback, thickness and taper? All these and other wing planform issues should be considered as thoroughly as technical methods, your ability and timescales allow.
Tail sizes can also be reassessed using typical values for tail volume coefficients, suit-ably adjusted to account for wing flaps, centre of gravity movement and aerodynamic efficiency. The planforms of the tail surfaces may also be altered to match the revised wing shape and rear fuselage geometry. It must be emphasised that using this method is relatively crude as each aircraft configuration produces different flow conditions over the tail surfaces. Also, some types may have uncharacteristic centre of gravity limits, offset thrust line, excessive aerodynamic moments and other features not representative of the general aircraft layouts. Such issues must be carefully taken into account when assessing a suitable value for the tail volume coefficients. As soon as enough detail is known it is important to conduct more precise control and stability calculations.
When all the geometric reviews have been completed it is necessary to recalculate the new aircraft configuration using the methods developed in the previous stage (baseline evaluation). However, it is now possible to incorporate more detail in the mass estima-tion methods used previously. Wherever possible the percentage MTO assumptions for components should be replaced by detailed estimates using geometrical formulae or suppliers’ data. Such information can be found in most aircraft design textbooks or in specialist technical papers on mass estimation (e.g. Society of Allied Weight Engineers, SAWE).
It is now possible to produce a report containing the revised baseline drawing and a summary of the mass, drag, lift and engine data, and aircraft performance predictions. This is sometimes regarded as the completion of the conceptual design phase. If deemed necessary it is possible to conduct more refined and extensive configurational studies using the methods described below.
2.9 Parametric and trade studies
The importance of clearly establishing the criteria upon which the effectiveness of the design will be judged was discussed earlier (section 2.1.2). A strong argument was made for this criteria to be used to direct the design process to arrive at the ‘optimum’ aircraft. This is done by the application of parametric and trade-off studies. For example, in many design projects aircraft costs figure highly in the assessment of aircraft effective-ness. The cost of buying the aircraft and operating it can be assessed against aircraft
“chap02” — 2003/3/10 — page 33 — #28
Preliminary design 33
design variables (e.g. wing area, aspect ratio, engine bypass ratio, thrust, etc.) and values selected that give the best overall configuration. For some aircraft the crite-ria on which the aircraft is judged may be related to aircraft performance (e.g. for fighters/manoeuvrability, racing aircraft/circuit speed, aerobatic aircraft/handling and control characteristics). In such cases, as with costs, parametric and trade studies will show the most effective choice of the aircraft design variables.
It is unusual for an aircraft design to be centred on one precise specification. Even if the original problem definition seems tight, the manufacturing company will be hoping to extend the design into other operating environments (e.g. a military trainer may be developed into a close-air support aircraft and civil aircraft are often stretched (or shrunk) to suit different markets). For most aircraft types some form of multi-tasking or even simple re-engining will be envisaged. To account for such considerations the baseline aircraft configuration defined in the previous stage will be assessed for its suitability to fulfil other roles. This may force a change to the baseline configuration so that future developments are easier to achieve. For example, a new design may have a wing area that is too large for the initial specification in order to allow for future heavier weight derivatives without the need for a new wing to be designed. However, this strategy must not make the original design too ineffective in its primary role. The design team must draw a fine line between existing and future requirements for the aircraft. Parametric studies are used to explore how successful the design can be made to meet all the potential developments.
Parametric design studies are often performed by adopting the classical nine-point method (see Figure 2.11a). This shows the variation of the objective function (e.g. MTO or direct operating cost) to changes in two of the aircraft parameters x and y (e.g. wing area and aspect ratio). The baseline aircraft, as developed in the previous stages of the design process, is often used as the centre point of the nine points. The parameters x and y are then taken as those for the baseline aircraft together with plus and minus steps each side of the central value (e.g. areas 54, 60, 66 and aspect ratios 7, 9, 11, where 60 and 9 are the values for the baseline design). The method can be extended to consider a third parameter (z) by producing a series of nine-point studies (see Figure 2.11b).
2.9.1 Example aircraft used to illustrate trade-off and parametric studies
To show how trade-off and parametric studies are employed in the design process, the following examples of typical studies are described. The design project used in these studies relates to a 40–50 seat regional jet operating over a 1000 nm single-stage length, flying out of a 5300 ft field on a hot day (ISA +25◦). The full study resulted in the selection of an aircraft configuration that was regarded as offering the minimum direct operating cost for the design mission. This configuration (the baseline aircraft layout) is shown in Figure 2.12 and some technical details are listed below:
Overall dimensions (m) Wing geometry Mass (kg) Overall length 24.5 Gross area 55.0 sq. m Maximum take-off 18 730 Overall width 22.6 Span 22.6 m Maximum zero fuel 17 006 Overall height 7.9 Aspect ratio 9.32 Payload 4790
Wing t/c 15% root, 11% tip Sweepback 20◦ at quarter chord
“chap02” — 2003/3/10 — page 34 — #29
34 Aircraft Design Projects
(a)
(b)
Objective function
x1
x2
x3
y1
y2
y3
Objective function
Variable z1
z2
z3
Trend line
Fig. 2.11 Classical nine-point carpet plots (diagrams a and b)
Fig. 2.12 Example aircraft layout
“chap02” — 2003/3/10 — page 35 — #30
Preliminary design 35
575
565
555
1.0 1.05 1.1 1.15
8600
8400
8200
1.0 1.05 1.1 1.15
Stage time (s)Wing area (sq. ft)
Engine scale Engine scale
10.0
9.6
6.00
5.90
Aircraft price ($M ) Seat mile cost (c )
1.0 1.05 1.1 1.15 1.0 1.05 1.1 1.15
Engine scale Engine scale
Fig. 2.13 Parameter study – engine size (example)
The results presented in the graphs (Figures 2.13 to 2.19) and described below are taken from a muli-variable optimisation (MVO) study in which all the aircraft design parameters were allowed to vary within predetermined limits. The aircraft parameter under investigation (e.g. wing taper ratio) was fixed at a selected value and the aircraft optimised for minimum aircraft direct operating cost. This process was repeated for other values of the parameter. The resulting optimised values of the design variables (e.g. wing area, aircraft empty mass, fuel mass, seat mile cost, etc.) were recorded for each value of the study parameter. The results when plotted show the sensitivity of the aircraft configuration to changes in the study parameter. As mentioned above, if a high sensitivity is indicated, more care must be taken in the selection of the parameter value for the final design configuration (and vice versa). The process described here is sometimes conducted without involving an MVO programme. This is less accurate as it requires some of the design variables to be held constant. It is easier and quicker to perform and, providing that the range of parameter variation is kept narrow, it is sufficiently accurate for the initial design stages. To avoid the complications associated with aircraft price and cost estimation it is possible to simplify the studies by adopting aircraft mass as the design objective function.
During the earlier description of the design process mention was made of the adoption of a ‘rubber’ engine in the aircraft to determine optimum engine size. Figure 2.13 shows the results of such a study. The penalties for including an oversize engine in the initial design for this project are quantified in this study. Of the parameters investigated, only stage-time benefits from the installation of the larger engine.
To study the stretch potential of the original baseline aircraft two parametric nine-point studies were completed. Figure 2.14 shows the effect on aircraft wing area and Figure 2.15 shows the effect on maximum take-off weight, for an aircraft with 56 seats (20 per cent increase) at various field length and stage distances.
With a newly designed 60-passenger baseline layout, further stretch potential was investigated by conducting a series of nine-point studies. Figure 2.16 shows the consequential effect on required engine size (still using the rubber engine).
“chap02” — 2003/3/10 — page 36 — #31
36 Aircraft Design Projects
Gross wing area (sq. ft)
720
680
640
600
560
520
56 seat aircraft
With 1.08 engine
5700 ft
900 nm
700 nm
1300 nm
6100 ft
5900 ft
1100 nm
5900 ft
1300 nm
1100 nm 6100 ft
Fig. 2.14 Parameter study – wing area (example)
Max. take-off weight (lb)
56 seat aircraft
5700 ft
51 500
49 500 5900 ft
900 nm
700 nm
1300 nm
1100 nm
6100 ft
47 500
45 500
Fig. 2.15 Parameter study – take-off mass (example)
Figure 2.17 shows the required wing area (indicative of aircraft size). Figure 2.18 shows the effect of stretch on aircraft maximum take-off mass.
Finally, the effect of all these changes on the aircraft seat mile cost (SMC) is shown in Figure 2.19. The diminishing improvement of SMC with aircraft stretch is clearly shown. Only the advantage of the initial stretch to 70 passengers (PAX) looks attractive.
“chap02” — 2003/3/10 — page 37 — #32
Preliminary design 37
Engine scale 1.75
1.65
1.55
1.45
1.35
1.25
1250
5800
6180
1000
80 PAX
72 PAX
60 PAX
1500 nm
6400 ft
Fig. 2.16 Operational design study – engine size (example)
Gross wing area (sq. ft)
760
680
600
520
1500 nm
1250 nm
1000 nm
80 PAX
72 PAX
60 PAX
5800 ft 6100 ft 6400 ft
Fig. 2.17 Operational design study – wing area (example)
The examples described above are only a brief selection of the types of investigation that can be conducted using parametric methods. Each aircraft project will raise specific types of study that are significant. By the time the project is developed to this stage the designers will be aware of the nature of the parametric studies that are of interest to them.
“chap02” — 2003/3/10 — page 38 — #33
38 Aircraft Design Projects
Max. take-off weight (lb)
76 000
1500 nm
68 000
1250
1000
6100 6400
80 PAX
5800 ft
72 PAX 60 000
60 PAX 52 000
Fig. 2.18 Operational design study – take-off mass (example)
Seat mile cost (cents) 5.15
5.0
4.95
4.85
4.75
4.65
4.55
4.45
5800 ft
6100 ft
6400 ft
1000 nm
1250 nm
72 PAX
60 PAX
80 PAX
1500 nm
Fig. 2.19 Operational design study – seat mile cost (example)
“chap02” — 2003/3/10 — page 39 — #34
Preliminary design 39
2.10 Final baseline configuration
The scope and depth of the trade-off and parametric studies undertaken during the conceptual design phase will depend on the time and effort available. The final base-line configuration will benefit from such in-depth studies but often decisions are required before such work can be completed. It is important that all decisions on the configuration are made with enough time left to perform the final analysis on the design.
2.10.1 Additional technical considerations As this stage in the initial design process represents the end of the technical work to be done, some extra details may be considered. For example:
• an appreciation of the structural framework for the aircraft, • consideration of the inboard and sectional profiles through the aircraft, • assessing the location and installation of the main systems and components
(i.e. engine including intake and nozzle, cockpit layout, fuel tankage, weapons, payload, services).
If time permits a first-pass analysis of the aircraft stability and control should be made to ensure that the tail control surfaces are adequate.
2.10.2 Broader-based considerations In contrast to the detailed technical analysis that has been the focus of much of the later stages of the design process, the final assessments should be concerned with broader-based aspects. In this respect each project will be different. The following list may help you in formulating the wider considerations for the project.
1. Manufacture • Repairs • How and where 5. Environmental issues • Required and available skills • Noise • Materials (availability and sizes) • Emissions • Timescales • Recycling • Developments • Handling potentially dangerous
2. Flying issues systems and substances • Pilot visibility and awareness 6. Safety • Handling and control • Airworthiness regulations • Training • Operational regulations • Developments • Manufacturing regulations
3. Operational issues • Certification procedures • Refuelling • Crashworthiness • Loading and unloading • Failure analysis • Provisioning • Reliability • Turn-round 7. Developments
4. Servicing • Stretch • Engines (accessibility) • Multi-roles • Stores • Improvements • Systems • Technical developments • Regular inspections • Flight testing
“chap02” — 2003/3/10 — page 40 — #35
40 Aircraft Design Projects
8. Programme management 9. Overall assessment • Stretch • (S) Strengths • Cost • (W) Weakness • Facilities • (O) Opportunities • Teambuilding • (T) Threats • Availability • Risk management
The prompts in the list above are not exhaustive. There may be other specific issues that apply to particular projects that are not mentioned above. Also, some of the topics mentioned might be irrelevant to your design but be careful not to dismiss any too hastily.
2.11 Type specification
At the completion of the initial design phase all details that are known about the aircraft are summarised in a report called the ‘Aircraft Type Specification’. It is the project manager who is responsible for this report. He is accountable for the accu-racy of the data and he will be expected to guarantee its validity. In a company, the sales and legal departments will use this document in contract negotiations. The technical specification therefore defines the guarantees the company will offer clients and thereby the liability it accepts in the contract to buy and use the aircraft and systems. Within this context the document is treated seriously in the design organisa-tion. It will not contain speculative statements or unsupportable data. The report consists of textural descriptions, drawings, diagrams, numerical data, graphs and charts. As the design of the aircraft progresses through later phases of the design process the document will be systematically reviewed and updated to include the latest information.
For student work it is good practice to simulate this procedure. Project management should require the production of a document which defines the aircraft characteristics. As the project matures more details can be added to the report.
In Chapter 11 (section 11.4) detail recommendations for creating a student design project report are presented. While a student design team report may not always cover all of the items suggested in the following professional report example, the list pro-vides suggestions for topics which could be considered for inclusion in the final team report.
2.11.1 Report format The type of information that is included in the document will vary depending on the nature of the aircraft project. The following list is representative of the sections included in a professional document:
1. Introduction 5. Weight and balance 2. General design requirements 6. Performance 3. Geometric characteristics 7. Airframe 4. Aerodynamic and structural criteria 8. Landing gear
“chap02” — 2003/3/10 — page 41 — #36
Preliminary design 41
9. Powerplant (and systems) 16. Interior accommodation 10. Fuel system 17. Environmental control 11. Hydraulic and pneumatic systems 18. Safety systems 12. Electrical system 19. Weapon systems (armament) 13. Avionics 20. Servicing 14. Instruments and communication 21. Exceptions to regulations 15. Flight controls 22. Definition and abbreviations
2.11.2 Illustrations, drawings and diagrams The Type Specification document contains several engineering drawings, schematic diagrams, system block diagrams, graphs, charts and general diagrams. The list below is not exclusive but provides a guide to the type of supporting illustrations to the text:
Aircraft three-view general arrangementInboard fuselage profileFuselage sectionsFuselage internal plan view (cabinarrangement)
Aircraft geometryMission profilesFlight envelopesFatigue spectrumUndercarriage vertical velocity spectrumRunway loadingWeight and C.G. diagramFuselage structural frameworkFuselage cross-sectionFloor loadingsCockpit view diagramWing structural frameworkWing/fuselage jointFlap detailsTailplane structure
Fin structure Undercarriage (main and nose) Nosewheel steering Engine installation Engine power off-takes Engine controls Fuel system and tankage Electrical system Antenna and sensor locations Avionics Hydraulics system Pneumatic system Flight control systems Environmental control system Cabin pressurisation schedule Instrumentation Ejector seat installation Auxiliary power unit Access panels Ground service
Obviously, some of the finer details contained in the above lists will not be known in the conceptual design phase. They have been included in the list to give a flavour of the type of detailed work that is still to be done on the aircraft design in subsequent phases.
References 1 Federal Aviation Administration (FAA-DOT) Airworthiness standards, FAR:
Part 1 Definitions and abbreviations Parts 11, 13, 15 Procedural rules Parts 21 to 49 Aircraft regulations
(details can be found on the FAA website www.faa.gov). 2 JAR (details can be found on the JAA website www.jaa.nl). 3 FAR part 36 – Noise standards: aircraft type and airworthiness certification. 4 Jenkinson, L. R., Simpkin, P. and Rhodes, D., Civil Jet Aircraft Design. AIAA Education
Series and Butterworth-Heinemann, 1999, ISBN 1-56347-350-X and 0-340-74152-X.
“chap02” — 2003/3/10 — page 42 — #37
42 Aircraft Design Projects
5 Raymer, D. P., Aircraft Design: A Conceptual Approach. AIAA Education Series, 1999, ISBN 1-56347-281-0.
6 Brandt, S. A. et al., Introduction to Aeronautics: A Design Perspective. AIAA Education Series, 1997, ISBN 1-56347-250-3.
7 Aviation Week Source Book, published annually in January. 8 Mattingly, J. D., Aircraft Engine Design. AIAA Education Series, 1987, ISBN 0-930403-23-1. 9 Eshelby, M. E., Aircraft Performance – Theory and Practice. Butterworth-Heinemann and
AIAA Education Series, 2000, ISBN 1-56347-250-3 and 1-56347-398-4.
“chap03” — 2003/3/10 — page 43 — #1
3
Introduction to theproject studies
The design process has been described in detail in the previous chapters. All the steps that are necessary to successfully complete the preliminary design stages have been identified. The amount of effort and time spent in each stage depends on the overall schedule for the project. It is essential to complete the process with a feasible baseline design, therefore it is necessary to programme and manage the work in association with all other commitments. Although the design method has been shown as a sequential process, it is possible to run some of the steps in parallel. It is also possible to do some preparation work (e.g. develop estimating methods and spreadsheets) ahead of the later stages. This is particularly useful if the project is to be done by a group, or team, of people. In such cases, it would be essential to allocate all tasks and to set a rigid timetable for the completion of the work (see Chapter 11 for more details on team working).
Some of the case studies that follow are laid out in the standard format shown below. This format mirrors the sequence of the work to be done in the preliminary design of any aircraft.
Introduction to the project
1. Project brief 2. Problem definition 3. Design concepts 4. Initial sizing and layout 5. Initial estimates 6. Constraint analysis and trade-offs 7. Revised baseline layout 8. Further work 9. Study review
Some of the projects in the following chapters have been included to illustrate design investigations into specific operational environments and therefore do not strictly follow the sequence above.
“chap03” — 2003/3/10 — page 44 — #2
44 Aircraft Design Projects
The first three projects (Chapters 4, 5 and 6) have been shown in more detail than some of the subsequent studies. They are chosen to illustrate the design aspects of different parts of the aeronautical industry; namely, civil, military and general aviation respectively. Each of the later projects is selected because of an unusual operational or design aspect.
Chapter 4 considers the design of a new type of civil aircraft. The expected devel-opment of exclusive executive/business scheduled services that provide small capacity, long range operation is the stimulus for the project. The design of the aircraft is not difficult but as such types of aircraft have not been built before, there is no information to use as the starting point for the design. The example illustrates the iterative process that is essential in such cases. The conclusion of the study raises questions that could stimulate several other design studies in this area.
The second project (Chapter 5) relates to the design of a new military trainer. This project has been selected as it shows how a systems approach to the solution of a design problem can offer substantial benefits. The aeronautical design of the aircraft is relatively straightforward once the operational issues have been decided. The aircraft in the context of the training environment represents only one element of the total system. Other parts of the training environment include pre- and post-flight simulation experience, ground-based instructor stations and modern electronic communication and data links. Adapting technologies developed in other aeronautical applications (in this case, flight testing) allows more efficiency and flexibility to the aircraft design and the total training system. The project definition for the aircraft is shown to be influenced by issues relating to the development of the aircraft family. Single and twin seat versions are eventually shown to be desirable. This complicates the design process but such considerations are not unusual in actual project work. The design process for this aircraft has been shown in more detail than for other studies in the book as it combines many interacting constraints.
General aviation is the largest sector in the aviation business. Chapter 6 shows how the design of a simple leisure aircraft can be combined with advanced technology developments. The project is set in the highly competitive field of air racing. Many racing aircraft have powerplants developed from automobile engines. Following this trend, this project postulates the introduction of electric propulsion to form a new type of racing formula. Developing and installing the current automotive fuel cell systems into an aircraft is investigated. Light aircraft design, from the Wright brothers onwards, has traditionally been used to test and develop new technologies. This project is chosen to simulate such situations.
The remaining four chapters each present a project that has unique operational requirements that significantly affect the basic layout of the aircraft. Such complica-tions are often part of novel design specifications. The first of these (Chapter 7) deals with an aircraft concept that has long stretched the imagination of aircraft designers, namely the roadable aircraft. To combine the attributes of an automobile and a light aircraft would offer a highly desirable mode of transport. It would possess the conveni-ence of the car for short journeys with the flexibility and time saving of an aircraft for long trips. The design problem concerns the matching of road and airworthiness regulations without compromising the operation in either transport mode. Although the completed preliminary design study lacks refinement in several technical areas, the student project won the NASA design prize for innovation in general aviation for the year 2000. A novel aspect of the work on this project was the integration of the design and analysis between student groups in UK and USA. This demonstrates that under-graduate design work does not have to be centred solely in one course, one department, one institution or even in one country. Dispersing the design team focused attention on
“chap03” — 2003/3/10 — page 45 — #3
Introduction to the project studies 45
the management and communication aspects of the design process, simulating modern industrial practice.
Many project studies arise from a request to consider an operational requirement outside existing experience. Such work can be classed as ‘Feasibility Studies’. Chapter 8 deals with such a study in the field of air offensive operations. This project formed the basis of the 2001/02 AIAA undergraduate design competition (see Chapter 8 for more details of this annual contest). Wars in the 1990s and since have demonstrated the need to gain air superiority over the war zone quickly. This leads to a requirement for aircraft to penetrate hostile territory in the early part of the offensive and ‘neutralise’ the air defensive capability of the enemy. Such initial strike aircraft are called interdictors. They must be stealthy and fast to avoid detection and have sufficient firepower to destroy heavily protected targets. This combination together with the long range required to attack deep inside an enemy country provides a challenging design problem.
The design study shown in Chapter 9 also has origins in the international conflicts in Europe and Asia in the 1990s. These demonstrated the need for improved local surveillance over potentially hostile territory. To provide this safely in an unstable area of conflict calls for the design of a high-altitude, long-endurance uninhabited aircraft. This defines the mission requirement for the project. This study illustrates the difficul-ties to be encountered in designing an aircraft to fly outside the normal operational environment. To add to the unorthodox mission requirements, the study also investi-gates an unusual aircraft configuration (i.e. a high aspect ratio, swept-forward, braced wing layout).
The final project (Chapter 10) returns to the problems faced by the early aircraft designers, namely, operating aircraft from water. In the more remote parts of the world, light aircraft provide the most convenient form of transport. In such places level-ground landing surfaces may not be available. Stretches of water (lakes and sheltered bays) provide a suitable alternative. Amphibious aircraft combine both aerodynamic and hydrodynamic requirements that must be met to produce a successful design. This project shows how these criteria are combined to produce a feasible aircraft to operate from either land or water.
From the studies described in Chapters 4 to 10, it can be appreciated that each design task is unique. Projects can take several different forms of investigation. Each one requires a different form of study. This is illustrated in the variation of work described. One of the tasks for the project management team in the early stages of the design process is to identify the type of work that is necessary to successfully complete the project.
The selected projects have intentionally covered unusual and difficult design prob-lems, set in civil, military and general aviation operating environments. The common theme in all the studies is the sequential nature of the preliminary design process. Work-ing through these projects will provide an understanding of the stages to be followed in other design studies. Some helpful guidance on the best way to handle such projects in an educational environment is given in Chapter 11.
“chap04” — 2003/3/10 — page 46 — #1
4
Project study: scheduledlong-range business jet
Bombardier Canadair Global Express, long-range bizjet
“chap04” — 2003/3/10 — page 47 — #2
Project study: scheduled long-range business jet 47
4.1 Introduction
Up to the events of 11 September 2001, all of the professional aeronautical industry market analysts predicted that scheduled airline business over the next 20 years was likely to increase at an annual growth rate of between 3.5 and 5 per cent. Such unex-pected and tragic events illustrate the vulnerability of airline market projections to influences outside the control of the industry. However, it is expected that after a period of industrial recession the previous projections will be resumed. Although this is welcome news for the aircraft manufacturers and airlines, as more passenger move-ments equate to a growth in business, it also means that existing airports and associated infrastructures will become increasingly inadequate to satisfy this expansion. Already many of the world’s international airports are working beyond capacity at peak operat-ing periods. The expected doubling of demand over the next 15 to 20 years is generally incompatible with the planning approval and building timescales for airport expansion. The political, social and economic factors that accompany airport building projects lie outside the control of the aeronautical industries. In the past, planning enquires and environmental pressure groups have delayed many of the proposed airport development projects. There is no evidence that this situation will improve in the future.
Some of the problems at airports may improve when the new, supercapacity aircraft are introduced but even this development will not solve the passenger capacity problems at airports. Moving airline operations to larger aircraft is not new. Most airlines now use larger capacity aircraft on services that smaller types satisfied a few years earlier. This trend is likely to continue. This development allows an increase in passenger movements without increasing aircraft movements (i.e. increases pas-sengers per flight ‘slot’). However, this practice does not solve the problems of increased passenger demand on the airport terminal facilities. Handling larger air-craft and greater numbers of passengers requires an associated expansion of airport infrastructure.
Analysis shows that although the main airports are working at full capacity, over 70 per cent of all aircraft movements involve relatively small aircraft. These aircraft do not need the service provided at the large airports (i.e. runway length for take-off and landing, and terminal lounge capacity). Many of these flights are related to regional ‘feeder’ services that provide linking flights to international scheduled services. The mixture of small and large capacity services at airports leads to an inefficient use of the facilities available. This inefficiency is the source of many of the delays and disruption currently endemic at large airports.
Business surveys show that delays at airports will increase as demand on the services increases in the future. Delays and disruptions in the service affect all passengers. Airlines provide exclusive facilities at airports for their business travellers but this does not pacify a customer who misses an important meeting because of a flight delay. Such passengers demand more certainty in their travel arrangements than can be provided by the current and future operations. An expensive alternative to the current situation is for the business traveller to use a small, exclusive business jet for the journey but this may not be within the budget of most commercial travellers.
It has been suggested by researchers that the current problems at large airports could be eased if the feeder services were transferred to satellite airports. Such developments would potentially increase the capacity provision at the larger airport without the need to make changes to the present runway or terminal facility. However, as the traveller will need to transfer to and from the large airport there will be a requirement to provide or improve the ground transport provision between the two airports. This type of development is slowly taking place at the main ‘hub’ airports. The downside to this
“chap04” — 2003/3/10 — page 48 — #3
48 Aircraft Design Projects
scheme is that the traveller is then subjected to extra potential delays from congestion at both airports and the ground interchange.
An essential element to any airline’s success is the ability to attract the ‘business’ traveller. Business travellers pay significantly higher prices for travel than tourist-class passengers and represent a more dependable source of income than travellers who opt for first class. Capturing the loyalty of the business traveller is high on the agenda of every major airline. This is often accomplished on longer-range flights, where there is sufficient space, by creating a separate ‘business-class’ cabin. In this the seat widths, seat pitch and cabin amenities are set between those of the first-class and tourist-class sections. Business-class passengers are allowed early boarding and a wider choice of in-flight movies and passenger services, etc. Business travellers, of course, pay for these advantages but ticket-pricing schemes are devised in which the business-class ticket costs little more than the ‘list’ price of one in economy class. And, unless one wishes to purchase his or her ticket a couple of weeks in advance of the flight and is willing to stay at his or her destination over the weekend, the ‘list’ price is the best available. These special business-class amenities are usually provided only on longer-range flights since it is assumed that on trips of a couple of hours or less the benefits of such service are questionable.
There have been a few attempts to create specialised, all business-class airlines using aircraft such as the Boeing 727 or 737 fitted with only about half to two-thirds the usual number of seats. However, these and similar aircraft like the Airbus 320 class are usually limited in range and are unsuited to the long transcontinental or international flights where business-class amenities can really make a difference to the target group of travellers. So far, these attempts have failed to attract enough customers to make a profit. This may be due to insufficient perceived advantage in the wider seats and better service against the higher price on shorter-range flights. Alternatively, it may be due to the insufficient flight frequency of the special services compared to flight schedules of the existing airlines.
Is there a market for a business-class only aircraft or airline? For success, it must meet the preferences of the business traveller, which include the following amenities:
• Larger, more comfortable seats with more leg room than those in tourist class. • Pemium in-flight service (better meals, free drinks, more selection of movies and a
wider choice of entertainment options). • Separation from tourist-class passengers in airport lounges during boarding, and on
board the aircraft (for mixed-class operations). • Faster flight check-in and post-flight luggage retrieval. • Direct flights without delays at airports, especially on longer journeys.
The first two of the above points are currently available to business-class travellers on larger, longer-range flights with most airlines. The next two preferences can only be achieved with special ‘business-class only’ flights. In addition, the last of these may require a reduced dependence on hub airports.
Most airlines depend on a mix of passenger classes and fare levels to operate prof-itably. First-class passengers or their company pay dearly for their extra comfort and amenities. However, on many flights, the first-class seats are filled with business nor tourist-class passengers who have used accumulated frequent flyer miles to ‘upgrade’ their seats. At the other end of the airplane, the tourist-class passenger may have paid anything from a couple of hundred dollars to over two thousand dollars for a transoceanic or transcontinental flight. The ‘list’ price for tourist class is often very near that for the discounted business class. This serves as an inducement for passengers
“chap04” — 2003/3/10 — page 49 — #4
Project study: scheduled long-range business jet 49
who cannot meet the requirements for discount tickets (typically 14 to 17 days’ advance purchase and travel which includes a weekend stay) to purchase business-class tickets. Those who can meet the discount requirements can often fly for very low cost. Airlines use sophisticated seat management software to optimise the price of each ticket; seats in ‘economy’ are sold at a different price depending on availability and demand. The goal is to fill every seat, and having a 90 per cent discount passenger is better than an empty seat. Like soft fruits, scheduled airline seats are perishable goods that must be sold before the ‘shop shuts’ or aircraft departs!
There is a question as to whether, an ‘all business class’ airline can fill enough seats on enough flights to make a profit. If not, it must become like every other airline and offer either managed discounts or some other ‘class’ of seating with a scheme to fill these in order to make each flight at least break even in cost. Such an airline would also have to come close to matching the flight frequency of regular airlines and it may need to offer at least slightly lower ticket prices to induce business travellers away from their frequent flyer club loyalties. Success would probably also require a new class of airliner which could fly transcontinental and transocean ranges with passenger capacities similar to today’s B-737s and A-320 class aircraft. Most existing airliners are designed as either a long-range/large-capacity, or a short-range/small-capacity operation, compatible with the current spoke and hub system of flight routing.
Meeting the last of the comments above will also require either a departure from the hub and spoke route model, or the use of aircraft designed to fly faster. The ideal airliner for this goal is probably a B-737, A-320 size aircraft with a range of 7000 nm and a cruise speed of Mach 0.9 or higher.
Any departure from the hub and spoke system may prove problematic given the satur-ated state of hub airports from which longer-range flights generally operate. A solution may be found in the use of other airports. Perhaps former metropolitan airports, which are currently used primarily for private and corporate aircraft operations, could be used. This would require the aircraft to be able to take off and land on shorter runways. It would also require the construction of high-speed ground transportation systems which could move business travellers between airports. This would further need the provision of gate-to-gate transfer of passenger and baggage without requiring additional baggage or security checks.
4.2 Project brief
From the analysis of existing and future air travel conditions above, it is possible to postulate a new type of airline service; one that is aimed at the profitable business travel market. This project study involves the development of a new scheduled business-exclusive international service from smaller airports.
An initial survey of the location of airports in the developed world was undertaken. This showed that within a radius of 50 miles around most current international airports, there is at least one regional airport that could be used for such a service. However, this may require the establishment of facilities to deal with international flights.
For existing airlines such a service would provide an improved and exclusive premium-class service and would allow an expansion of economy-class business at existing busy airports. New airlines may be set up to exploit the perceived market opportunity. Over the past decade, with the relaxed ‘deregulatory’ airline service, several new airlines have evolved to provide quick, easy and cheap (bus-type) alternative scheduled services. Some of these failed to achieve profitable operation but a significant number survived to compete with the older and larger established airlines. Such developments show that
“chap04” — 2003/3/10 — page 50 — #5
50 Aircraft Design Projects
the airline business is dynamic enough to respond to novel market opportunities. A new aircraft type would create a unique, convenient, exclusive high-class business service that would compete with the current business-class sections in existing mixed-class scheduled services.
4.2.1 Project requirements The following design requirements and research studies are set for the project:
• Design an aircraft that will transport 80 business-class passengers and their associ-ated baggage over a design range of 7000 nm at a cruise speed equal or better than existing competitive services.
• To provide the passengers with equivalent, or preferably better, comfort and service levels to those currently provided for business travellers in mixed-class operations.
• To operate from regional airports. • To use advanced technologies to reduce operating costs. • To offer a unique and competitive service to existing scheduled operations. • To investigate alternative roles for the aircraft. • To assess the development potential in the primary role of the aircraft. • To produce a commercial analysis of the aircraft project.
4.3 Project analysis
Project analysis will consider, in detail, each of the design requirements described in the previous section.
4.3.1 Payload/range With only 80 seats, the aircraft is considered as a small aircraft in commercial transport operations. This size of aircraft is normally used only on short-range, regional routes. Table 4.1 (data from a Flight International survey of world airliners) shows the existing relationship between aircraft size (number of passengers (PAX)), design range and field capability.
By considering the range requirement of 7000 nm, the new aircraft falls into the large aircraft category but the passenger capacity of only 80 defines it as a small aircraft. This contradiction defines the unique performance of the new aircraft. The closest comparison to the specification is with the corporate business jet. However, this type of aircraft has a much smaller capacity (usually up to a maximum of 20 seats).
To investigate the significance of the 7000 nm range requirement, an analysis of the 50 busiest international airports was undertaken. This compared the great circle distances between airport pairs. It showed that very few scheduled routes exceeded 7000 nm. The list below shows the exceptions:
US east coast – Sydney, Singapore, Thailand US central – Sydney, Thailand US west coast – Singapore Europe central – Sydney
All of the above routes could be flown with a refuelling stop at Honolulu for the US flights and Asia for the European flights. This analysis showed that 7000 nm was a
“chap04” — 2003/3/10 — page 51 — #6
Project stu
Table 4.1
dy: scheduled long-range business jet 51
PAX Range (nm) Field length (m)
Small aircraft 728Jet 70 1430 1676 CRJ 700 70 1690 1880 F 70 79 1870 1583 928Jet 95 1900 1950 RJ 100 112 2090 1275 F 100 107 1680 1715 B717-200 106 1460 1480 A318-100 107 2350 1400
Medium aircraft A310-300 218 5180 2300 B767-200 186 3000 2100 B767-300 245 3460 2550 A300-600R 266 4160 2280
Large aircraft A340-500 313 8504 3050 B777-200ER 305 7775 3020 MD11 285 6910 3110 B747-400 416 6177 3020 A380 555 7676 2900
reasonable initial assumption. This distance could be reduced if the design was shown, in subsequent trade-off studies, to be too sensitive to the range specification.
4.3.2 Passenger comfort Long-range flights obviously equate to long duration. At an average speed of 500 kt, the 7000 nm journey will take 14 hours. Increasing the speed by only 5 per cent (e.g. from M0.84 to M0.88) will reduce this by 45 minutes. Anyone who has travelled on a long flight will agree that this reduction would be very welcome. Business travellers may accept a premium on the fare for such a saving in time and discomfort. As flight duration and comfort are interrelated, it is desirable to provide a high cruise speed for long-range operations. Reduced aircraft block time will also provide an advantage in the aircraft direct operating costs providing that extra fuel is not required for the flight.
As journey time relates directly to perceived comfort level, airlines have traditionally provided more space for business-class travellers. In the highly competitive air transport industry many other facilities and inducements have been used to attract this high value sector of the market. A new aircraft design will need to anticipate this practice and offer, at least, equivalent standards. This will impact directly on the design of the aircraft fuselage and in the provision of cabin services and associated systems.
4.3.3 Field requirements The requirement to operate from regional airports effectively dictates the aircraft maxi-mum take-off and landing performance (see Table 4.1). Operation from smaller airports will also affect the aircraft compatibility to the available airport facilities.
“chap04” — 2003/3/10 — page 52 — #7
52 Aircraft Design Projects
100
90%
68%
80
60
40
20
1800 m
1400 m (4600 ft)
(5900 ft)
6000 ft 4000 ft
1000 m 2000 m 3000 m 4000 m
Field length
Fig. 4.1 Runway length survey
To understand this in more detail a survey is required to determine the available runway length of regional airports. Such a survey was undertaken in an aircraft design study1 for a conventional feederliner. Figure 4.1 shows the results of this survey.
The frequency distribution of major European, regional airport, runway lengths (mostly UK, France and Germany) indicates that the 90 percentile equates to a mini-mum field length of 1400 m (4600 ft). Many of the aircraft operating within this field requirement are general aviation types. For an aircraft of 80 or more seats, this short distance may be regarded as too demanding on the aircraft design. It would force the wing to be too large or require a complex flap system. Both, or either, of these would increase drag in the cruise phase and thereby the aircraft direct operating costs. A sens-itivity study on this aspect of the design could be conducted later in the design process when more details of the aircraft are available. Increasing the field length to 1800 m (5900 ft) will allow operation from 70 per cent of the airports surveyed. Comparing this choice to current, regional aircraft characteristics shows it is equivalent to the Avro RJ, Fokker and Boeing types. For this reason, the longer (1800 m) length will be specified for the design.
4.3.4 Technology assessments The requirement to incorporate advanced technology into the design raises several questions relating to commercial risk, technical viability and economics. A design study that included a detailed assessment of new technologies applied to regional aircraft was presented by a Virginia Tech (VT) team at an AIAA meeting in 1995.2 This considered some emerging technologies in propulsion, aerodynamics, materials and systems. In the final configuration of their aircraft, they selected ducted-direct-drive prop-fans as the powerplant. This showed substantial fuel saving over normal, high-bypass turbofans. They accepted the relatively slow cruise speed (M0.7) because their specification only
“chap04” — 2003/3/10 — page 53 — #8
Project study: scheduled long-range business jet 53
called for a 3000 nm range. As much of the flight duration on short stage distances is spent in climb and decent, a reduced cruise speed is not too critical. For our design, such a slow cruise speed would not be acceptable, as it would significantly compromise the performance (flight duration) against existing scheduled services. For this reason, the prop-fan engine is not a suitable choice for our aircraft. A conventional high-bypass turbofan engine that is already certified and in use on other aircraft types will be our preferred choice. Although this will not show the fuel savings identified in the VT study, it will be comparable to the competitive aircraft. In addition, adopting a fully developed engine will reduce commercial risk and lower direct operating costs.
From an aerodynamic standpoint, the VT study proposed the incorporation of natu-ral laminar flow aerofoil sections with boundary layer suction on the upper leading edge profile. Research results from NASA Langley were quoted to validate this approach. The hybrid laminar flow control system was shown to reduce aircraft drag and therefore fuel consumption. The study proposed the use of wing tip vortex turbines to power the boundary layer suction system. As such devices have not been developed in the time since the report was published, it is not considered wise to adopt this concept for our design. This will leave the wing tips clear for winglets to reduce induced drag in cruise. These are now well established on many long-range aircraft, therefore the technology is well understood. Boundary layer suction will need to be provided from bleeds from the engines. Later in the design process, a study will need to be undertaken to determine the effectiveness of the laminar flow system against the reduction in engine thrust in cruise caused by the demand from the air bleed system. On the turbulent flow parts of the aerofoil, it is proposed to incorporate the surface striation researched by Airbus and NASA in the late 1990s.
The use of new materials in the construction of civil aircraft is now becoming com-monplace. To continue this trend composite materials will be used for wing skins, control surfaces, bulkheads and access panels. Advanced metallic materials will be used in high load areas (landing gear, flap mechanisms, engine and wing attachment struc-tures). As proposed in the VT study, micro-perforated titanium, wing-leading-edge skins will be used for the boundary layer suction structure. A conventional, aluminium-alloy, fuselage pressure shell will be proposed as this is well proven and adds confidence to the aircraft structural framework. Filament wound composite structures may offer mass reductions for the pressure cabin but this technology is still unproven in airliner manufacture, so it will not be used on our aircraft.
Aircraft systems will follow current technology trends. This will include a modern flight deck arrangement. Aircraft system demand will increase due to the improvement in provision for the passenger services and comfort. This will include better air con-ditioning in the cabin to provide an increase in the percentage of fresh air feed into the system, more electronic in-flight passenger services and business (computing and communication) facilities. The aircraft will be neutrally stabilised to reduce trim drag in cruise and therefore require redundancy in flight control systems.
4.3.5 Marketing Our aircraft type lies between the conventional mixed-class scheduled service and the exclusive corporate jet. The aircraft and operator will be offering a unique service. A comparison to the old ‘Pullman’-class service operated by the railways at the begin-ning of the last century is appropriate. Avoiding major airports and the associated, and increasing, congestion and delays will be a significant feature of the service. Seg-mentation of the premium ticket passengers away from the low-cost travellers will be
“chap04” — 2003/3/10 — page 54 — #9
54 Aircraft Design Projects
another positive marketing feature. Providing commercial/office facilities and a quieter environment during the flight will be another improvement over the existing mixed-class operations. All of these advantages will need to be set against the premium fare that the service will need to charge to offset the higher cost of operating the aircraft compared to existing services. In an analysis of the pricing policy of the new service it may be difficult to assess the elasticity of the ticket price because the service is new and untried. In the past, a sector of the travelling public has been attracted to the Concorde service. The reason that the extra ticket price was accepted is not clear. Either the time saving from supersonic flight or the exclusivity of the service, or both, may have been the feature that the customer was attracted to. It is felt that a premium above the exist-ing business-class fare of 30 per cent is probably the limit of acceptance by the market sector. At this stage in the development of the project, this is only a ‘guesstimate’. Market research would be necessary to identify the exact premium. A more in-depth market analysis will be needed before confidence in this figure is possible. There will always be a number of people who would use such a service. But as the ticket price rises, this number reduces. The number of passengers willing to pay the extra price must be seen to be greater than the number required to make the service commercially viable. The price at which companies regard the airfare as excessive must be determined.
4.3.6 Alternative roles Developing an aircraft exclusively for a specialised role in civil aviation would be regarded as commercial madness. All aircraft projects should consider other roles the aircraft may fulfil. Our aircraft will have a fuselage size that is more spacious than normally associated with an 80-seat airliner. The long-range requirement will demand a high fuel load and this will make the aircraft maximum design weight heavier than normal for 80-seat aircraft. Both of these aspects suggest that the aircraft could be transformed into a conventional higher capacity, shorter-range airliner. A study will be required to investigate such variants. This type of investigation may result in recom-mendations to change the baseline aircraft geometry to make such developments easier to achieve. For example, increasing the fuselage diameter may allow a change from five to six abreast seating in the higher capacity aircraft to be made. Without such a change, six abreast seating may be unfeasible.
Other variants of the aircraft could be envisaged for military use. The long-range and small field features of the design are compatible with troop and light equipment transport operations. The ability to move military personnel without the need to refuel would avoid some diplomatic problems that have arisen in the past. The long endurance feature would make the aircraft suitable for maritime patrol, reconnaissance, surveil-lance and communication roles. The military variants should not be considered in the design of the baseline aircraft, as this would unduly complicate the conceptual design process. Such considerations should be left until the current design specification is better realised.
4.3.7 Aircraft developments All aircraft projects must consider future development strategies to avoid complicated and expensive modifications in the development process. In modern civil aircraft design, it is common practice to consider the aircraft type as a ‘family’. Airbus and Boeing use this approach successfully in their product lines. Stretching, and in some cases shrinking, the original design is now normal development practice. All new aircraft projects consider this in their definition of the initial design. It is essential to consider the
“chap04” — 2003/3/10 — page 55 — #10
Project study: scheduled long-range business jet 55
PAX 120
110
100
90
80
70
60
Max. cabin capacity
Increased PAX
Initial design pt.
reduced range
Increased range
reduced PAX
Development to increased MTOM
Max. fuel capacity
Initial design MTOM
7000 nm Design range
Fig. 4.2 Aircraft development (payload/range) options
consequences of this approach in the conceptual design phase. In this way, constraints to the development of the aircraft are reduced.
Apart from making geometrical changes around the initial, maximum design mass, it is common to expect a growth in this limit over the lifetime of the aircraft type.
Typically a 35 per cent growth in max. take-off mass may be expected over the lifetime of the type. Figure 4.2 shows how such developments are planned. The payload (PAX) – range (nm) diagram shows the initial design specification of the aircraft. The sloping maximum design mass line shows the initial layout options (trading passengers for range and vice versa). The dashed line represents a developed higher mass aircraft. This shows the growth (PAX and range) potential for an MTOM increase. Such investigations are required in the early conceptual design phase to guide the aircraft development path. It may be found necessary to slightly compromise the best layout of the initial aircraft to provide for such developments.
4.3.8 Commercial analysis This last topic in the analysis of the aircraft project considers the commercial viability of the whole project. Although this cannot be assessed in detail at the start of the project due to a lack of technical data, it is possible to prepare for a commercial analysis later in the design process.
This preparation will identify the potential market for the aircraft, the potential customers for the aircraft, and the main competitors. The design team will need to know what are the principal commercial parameters that potential customers (airlines
“chap04” — 2003/3/10 — page 56 — #11
56 Aircraft Design Projects
and passengers) will use to judge the attractiveness of the new service in the total market. One of the obvious issues to be considered is aircraft costs. This includes the purchase price and various direct operating cost (DOC) parameters.
Finally, assessment of the operating issues relating to the new service will need to be understood. This will include the customer service for both pre- and in-flight parts of the operation.
4.4 Information retrieval
As mentioned earlier in this chapter, this aircraft specification lies between long-range bizjets and regional feeder liners. The aircraft specified range is similar to the Gulfstream V but this bizjet only carries up to 15 passengers. The passenger capacity is similar to regional jets but they only fly about 1300 nm. To assess the design parameters that might be used in later sizing studies Table 4.2 has been compiled, which shows some of the details of these two different types of aircraft.
Table 4.2 shows that the thrust to weight ratios (T /W ) for the two types are signi-ficantly different. The reasons for this lie in the requirements for higher climb/cruise performance and short field performance for the bizjets. These are parameters that our aircraft should have, so a thrust/weight ratio of 0.32 (the lower value for bizjets and the upper one for regionals) will initially be assumed for our aircraft.
Wing loading (W /S) is also seen from the data in the table to be statistically different between the two aircraft groups. There may be a variety of operational criteria for this division but for the same reason as above, a value lying between the two sets will be selected. A value of 450 kg/sq. m, being low for regional jets but high for bizjets, will be used. This decision may mean that ‘high-performance’ flaps will be required.
Mass ratios are always difficult to assess from published data as there are often conflicting variations in the definition of terms. For example, empty weight ratio will be higher for smaller aircraft and smaller for long-range aircraft. It should be relatively
Table 4.2
Range MTO W/S PAX (nm) (kg, lb) T/W (kg/sq.m, lb/sq.ft) ME/MTO
Business jets Falcon 2000 19 3000 15 875, 35 000 0.327 323, 66.3 0.563 Gulfstream V 14 6500 40 370, 89 000 0.332 382, 78.3 0.526 Learjet 45 10 2200 8 845, 19 500 0.359 359, 73.6 0.600 Canadair RJER 50 2270 23 133, 48 800 – 478, 98.1 0.591 Beechcft 400A 8 1690 7 303, 16 100 0.360 326, 66.8 0.624 Hawker 100 10 3010 14 061, 31 000 0.340 404, 82.8 0.581 Citation 11 3300 15 650, 34 500 0.371 – 0.586
Commercial jets Fokker 100 107 1680 44 450, 98 000 0.308 475, 97.3 0.556 Romero 1-11 109 1480 47 400, 104 500 0.289 494, 101.2 0.500 RJ100 112 2090 44 000, 97 000 0.290 572, 117.2 0.573 B717-200 A318-100
106 107
1460 2350
49 895, 110 000 64 500, 142 200
0.291 0.330
536, 110.0 526, 107.8
0.614 0.627∗
∗ A derivative of a larger aircraft.
“chap04” — 2003/3/10 — page 57 — #12
Project study: scheduled long-range business jet 57
easy to reassess the selected mass ratio following the first detailed mass estimations. Until this data is available it is necessary to make sensible ‘guesstimates’. Values of 0.52 for the empty mass fraction and 0.35 for the fuel fraction seem reasonable, at this time.
For comparison, the values for these parameters for the VT study aircraft2 are quoted as: 0.32, 535, 0.42, 0.32. Some of these differences can be explained by the larger size (165 PAX), shorter-range design specification of the VT study.
4.5 Design concepts
The previous section has shown that all of the potential competitors to the new design are of conventional configuration. They have trapezoidal, swept, low-mounted wings, with twin turbofan engines and tail control surfaces. Obviously, one of the concepts to consider is to follow this arrangement. The conservative airline industry may prefer such a choice. An alternative strategy is to adopt a novel/radical layout.
The ‘new look’ would set the aircraft apart from the competition and offer a mar-keting opportunity. In adopting such a design strategy, care must be taken to reduce technical risk and to show improved operational efficiency over the conventional layout.
Four design options are to be considered:
• Conventional layout • Braced wing canard layout • Three-surface layout • Blended body layout
4.5.1 Conventional layout(s) (Figure 4.3) This must be regarded as a strong candidate for our baseline aircraft configuration as it is a well-proven, low-risk option. The technical analysis is relatively straightforward and has a high confidence level in the accuracy of the results. Its main advantage is that it is similar to the competitor aircraft and thereby with airport existing facilities and operations.
There are some drawbacks to choosing this layout. These relate to the geometrical difficulties of mounting a high-bypass engine on a relatively small aircraft wing (relat-ing mainly to ground clearance below the engine nacelle). This is illustrated in drawing A on Figure 4.3. There are two possible, alternative aircraft arrangements that could overcome this problem. Version B, shown on Figure 4.3, shows the engines mounted at the rear of the fuselage structure. This avoids the ground clearance problem but intro-duces other difficulties. Since a large component of aircraft mass is moved rearwards the aircraft centre of gravity also moves aft. This requires the wing to be moved back to balance the aircraft. The movement of the wing lift vector rearwards shortens the tail arm and consequently demands larger control surfaces. This increases profile drag and possibly trim drag in cruise. The second alternative layout is shown in version C on Figure 4.3. In this option the wing is moved to the top of the fuselage section (a high mounted wing). This lifts the engine away from the runway and provides adequate ground clearance. The high wing position, although used on some aircraft, is regarded as less crashworthy. The fuselage and therefore the passengers are not cushioned by the wing structure in the event of a forced landing. This is regarded as particularly signific-ant in the case of ditching into water, as the fuselage would be below the floating wing structure. For an aircraft that is likely to spend long periods over water, airworthiness considerations may deter airlines from this type of layout.
“chap04” — 2003/3/10 — page 58 — #13
58 Aircraft Design Projects
Version A (wing mounted engines)
Version B (fuselage engines)
Version C (high wing)
Fig. 4.3 Conventional layouts
A problem not necessarily restricted to the conventional layout is the potential lack of fuel tankage. A long-range aircraft will require substantial fuel storage and this may not be available in a conventional wing layout.
4.5.2 Braced wing/canard layout (Figure 4.4) Although this configuration looks radical, it is technically straightforward with well-proven, and understood, analysis that provides technical confidence.
The canard and swept forward wings offer low cruise drag possibilities. The rearward positioning of the engines reduce cabin noise. The bracing structure should reduce wing loads and allow a thinner wing section to be used. This, in combination, may reduce wing structural mass and aircraft drag. The main weakness of the layout lies in the uncertainty of the positioning of the canard, wing and engine components, and the
“chap04” — 2003/3/10 — page 59 — #14
Project study: scheduled long-range business jet 59
Fig. 4.4 Braced wing layout
interference effects of the airflow at the brace structure junctions. There is also some uncertainty about the effect of the brace on future stretch capability.
4.5.3 Three-surface layout (Figure 4.5) This configuration has the advantage of low trim drag in cruise. The combination of forward and rear control and stability surfaces can be used to trim the aircraft in cruise with an upward (lift) force which will unload the wing. Two different wing layouts can be considered – swept forward or swept back. These options are shown in Figure 4.5.
It is anticipated that the swept forward configuration will be more suited to the development of laminar flow but may be heavy due to the need to avoid structural divergence. The bodyside wing chord will need to be sufficient to permit laminar flow systems to be installed. This is easier to arrange on the swept back layout. The increased internal wing volume created by the larger root chord will also provide increased wing fuel tankage. This together with the better flap efficiency of the swept back wing makes it the preferred choice of layout. The rear mounted engines will reduce cabin noise and visual intrusion although increase aircraft structural mass.
This layout is a strong contender for the preferred layout of our aircraft as the techn-ical risks involved are low yet the configuration is distinctive. There may be a slight problem in positioning the forward passenger door due to the canard location but this should be solvable.
“chap04” — 2003/3/10 — page 60 — #15
60 Aircraft Design Projects
Optional fuel tank
Fig. 4.5 Three-surface layout
4.5.4 Blended body layout (Figure 4.6) There has been a lot of interest in this configuration for the new supercapacity (550–1000 seats) aircraft. It is not novel. Several previous aircraft designs (mostly mil-itary) have adopted the layout. Aerodynamically, this layout is very efficient and lends itself to the installation of laminar flow control systems. For a long-range aircraft this is a major advantage, as fuel consumption will be reduced. The large internal volume of the wing should provide sufficient fuel tankage.
The main disadvantages relate to the difficulty of providing cabin windows and ensur-ing passenger evacuation in the case of an accident. Some innovation will be necessary to overcome these problems and satisfy airworthiness authorities. Airlines may be cautious of making this ‘step into the unknown’ due to the uncertainty of passenger
“chap04” — 2003/3/10 — page 61 — #16
Project study: scheduled long-range business jet 61
Fig. 4.6 Blended body layout
acceptance. A further problem, inherent in this layout, is the difficulty of stretching the integrated aircraft structure during programme development.
4.5.5 Configuration selection From a narrow commercial viewpoint, the conventional layout should be chosen, as it is a low-risk, low-cost option. However, there are doubts regarding the adequate provision of fuel tankage and the lack of a new ‘aesthetic’ for the service. Of the conventional layouts, the best is version B (rear fuselage mounted engines). This makes the passenger cabin less noisy, which would be seen as an advantage for an executive-class aircraft operating long endurance flights.
Although the conventional design is the natural choice, we will select one of the more radical configurations. In this way, it should be possible to compare, in more detail, the strengths and weaknesses of the design relative to the alternative (competitive) strategy of using a modified version of an existing aircraft. This comparative study could form the conclusion to the study.
Of the three novel configurations described above, the most radical is the braced wing layout. This option presents a larger commercial risk therefore it will not be pur-sued. The blended body aircraft is potentially a strong contender as the integrated structure/aerodynamic concept provides a technically efficient layout. The generous internal volume of the aircraft will suit development potential and offer adequate fuel tankage. The main difficulty is the lack of understanding of the internal structural framework. The integration of the wing air loading and the fuselage pressurisation loads is not easy to envisage. For a larger aircraft, this problem is eased due to the internal space available. For our smaller aircraft, such separation of load paths may not be feasible. Even if the structural problems could be solved, the integration of
“chap04” — 2003/3/10 — page 62 — #17
62 Aircraft Design Projects
structure and aerodynamic designs would make it difficult to stretch the design into a family of layouts (e.g. Airbus A318, 319, 320, 321).
The three-surface configuration has been successfully used on other aircraft and has been shown to offer performance advantage over the conventional layout. The saving in fuel during cruise will reduce the tankage requirement. As mentioned previously, the canard will make the front fuselage design more complicated but this difficulty can be overcome by detail design. If necessary, a shortage of fuel volume may be avoided if external tanks are added to the wing or if fuselage tankage is allowed.
Based on the arguments above the three-surface layout will be selected for our baseline configuration.
4.6 Initial sizing and layout
At this point in the design process, we can begin to realise the aircraft geometrical configuration. It is necessary first to make an estimate of the aircraft mass. Using this value, the engine and wing sizes can be determined. The fuselage shape is determined from the internal layout and tail requirements. With the main components individually defined, it is then possible to produce the first scale drawing of the aircraft. Crude estimates from similar aircraft types are necessary to complete the layout (e.g. tail and landing gear).
4.6.1 Mass estimation From section 4.4 the following mass parameters were suggested:
Empty mass fraction = 0.52
Fuel mass fraction = 0.35
The mass estimations below are shown in kg only (conversion factor: 1 kg = 2.205 lb). The payload is specified as 80 business-class passengers and their baggage. For this
type of ‘premium-ticket’ operation the mass allowance per passenger (including bag-gage) will be larger than normal. We will allow 120 kg per passenger. The flight rules will dictate at least two pilots. Airlines will want to provide high-class service in the cabin so we will assume four cabin attendants are required. It is common practice1 to allow 100 kg for each flight crew and 80 kg for each cabin attendant.
Hence, the payload is estimated to be:
Mpay = (80 × 120) + (2 × 100) + (4 × 80) = 10 120 kg
To cover incidental flight services, allow an extra 5 kg per passenger. This adds 400 kg to the payload mass, making the aircraft ‘useful mass’ = 10 520 kg.
Using the equation described in Chapter 2 (section 2.5.1) with the values above gives:
MTOM = 10 520/(1 − 0.52 − 0.35) = 80 923 kg (178 435 lb)
“chap04” — 2003/3/10 — page 63 — #18
Project study: scheduled long-range business jet 63
The initial mass statement is:
Operational empty Extra services Crew (2 + 4) Passengers Fuel
kg % MTO 42 080 52
400 520
}
13 9 600
28 323 35
MTO = 80 923 100 (178 435 lb)
4.6.2 Engine size and selection The literature survey (section 4.4) indicated a thrust to weight ratio of 0.32 was appropriate.
Hence:
Engine total take-off thrust = 0.32 × 80 923 × 9.81 = 254 kN (57 100 lb)
With two engines this equates to 127 kN per engine (28 550 lb)
A choice of engines from different manufacturers is always the preferred commercial position for the airframe manufacturer. This ensures that the engine price and avail-ability is more competitive. It also provides the potential airline customer with more bargaining power when selecting the aircraft/engine purchase.
There are several available engines that would suit our requirement. All of them are currently used on civil aircraft operations therefore considerable experience is available. The engines below are typical options:
• CFM56-5B as used on the A320 • CM56-5C as used on the A340 • IAE-V2533 as used on the MD90 family • IAE-V2528 as used on the A321.
The details∗ below are representative of these engines:
Bypass ratio Thrust ISA-sea-level static Typical cruise thrust (max.) Cruise specific fuel consumption Length Diameter Engine dry weight (mass)
5.531 000 to 34 000 lb (138 to 151 kN)5840 to 6910 lb (26 to 31 kN)0.594 to 0.567102 in (2.6 m)68.3 to 72.3 in (1.7 to 1.8 m)5250 lb (2381 kg)
∗Note: engine manufacturers commonly quote values in Imperial units. These details will be enough for initial performance and layout purposes but as the
design progresses it will be necessary to periodically review the choice of engines to be used on the aircraft.
4.6.3 Wing geometry The recommended wing loading is 450 kg/sq. m, hence:
Wing gross area (S) = 80 923/450 = 180 sq. m (1935 sq. ft)
“chap04” — 2003/3/10 — page 64 — #19
64 Aircraft Design Projects
Selecting a high aspect ratio (AR) will lower induced drag in cruise and save fuel. A value of 10 is to be used. The choice of aspect ratio will need to be reviewed in a trade-off study later in the design process. Using the wing area and aspect ratio we can determine:
Wing span (b) = (AR × S)0.5 = 42.4 m (135 ft)
Mean chord (cm) = (b/AR) = 4.24 m (13.5 ft)
Selecting a taper ratio of 0.3 gives (approximately):
Tip chord = 2.0 m (6.6 ft)
Centre line chord = 6.52 m (21.4 ft)
Using published data on critical Mach number analysis, at our preferred cruise speed of M0.85, a sweepback angle of 30◦ will allow a maximum wing thickness of 15 per cent without incurring wave drag penalties. This wing thickness will allow space for the proposed laminar flow control system and offer extra fuel tankage. An unswept inboard trailing edge will make the flap more effective and provide space for the main undercarriage, retraction mechanism. This will add extra area so the wing chords will be changed from the values calculated above to 1.8 m and 6.0 m (5.9 and 19.7 ft) respectively to retain approximately 180 sq. m (1935 sq. ft) area.
The basic wing geometry is shown in Figure 4.7. This also shows the location of the wing fuel tanks and the position of the mean aerodynamic chord (MAC). As an initial assumption the longitudinal position of the wing on the fuselage will be arranged to line up the position of the MAC quarter-chord with the aircraft centre of gravity.
Fuel volume considerations
At this point in the design process it is necessary to determine the size of the required fuel volume estimated in the initial mass estimation and then to compare this to the available space in the wing. This involves transforming the fuel mass into a volume. Fuel volume/capacity is often quoted in terms of ‘gallons’. This must be converted into linear units (cubic metres or cubic inches) to relate the size to the aircraft geometry. To do this conversion it is necessary to understand the various systems of units used. The calculation is shown in detail below because it is seldom to be found in other aircraft design textbooks although it is always a significant consideration.
In SI units – one litre is defined as the volume required to hold one kilogram of water. It is further defined as 1 litre = 1000 cubic centimetres (i.e. 0.001 cubic metres).
Hence, 1000 kg of water occupies 1.0 m3
In USA – one US gallon equates to the volume required to hold 8.33 lb of water. For water at 62.43 lb/ft3 a US gallon therefore corresponds to a volume of 231 cubic inches.
Hence, 1000 lb of water occupies 120 US gallons (=16.02 ft3)
Warning: In the UK the definition of the gallon is different to the USA gallon! In the UK – the Imperial gallon is used to measure liquid capacity. This is defined
as the volume required to hold 10 lb of water. This makes the Imp gal. = 277.42 cubic inches.
Hence 1000 lb of water occupies 100 Imp. gallons (= 16.02 ft3)
Note: the density of a liquid (water) does not change with the system of units!
“chap04” — 2003/3/10 — page 65 — #20
Project study: scheduled long-range business jet 65
Main wing fuel tanks (hatched)
Wing area 2 (nominal)
Optional external fuel tank
25% MAC
Outer flap
Aileron
Space for landing gear
Inner flap
Airc
raft
cen
tre
line
Fuse
lage
dia
met
er
(a) = 180 m
Wing span = 42 m
and other systems
(b)
L/2
L/2
A2
Amid
Cross-sectional area: A1
LFuel tank volume = ( )(A1+A2+ 4Amid)6
Fig. 4.7 (a) Initial wing planform geometry (b) Fuel tank volume
Here are some useful conversion factors:
1 US gal = 0.833 Imp. gal 1 Imp. gal = 1.2 US gal 1 US gal = 3.79 litres 1 Imp. gal = 4.55 litres 1 cubic foot = 28.32 litres 1 cubic metre = 1000 litres 1 cubic metre = 35.3 cubic feet
Specific gravity is the unit that relates the density of a liquid to water. For aviation fuel, specific gravity varies with the type of fuel (e.g. JP1, 3, 4, 5, 6, or kerosene) between
“chap04” — 2003/3/10 — page 66 — #21
66 Aircraft Design Projects
values in the range 0.82 to 0.76. For civil aircraft fuel, a value of 0.77 can be assumed, hence:
1000 kg of fuel occupies 1.3 m3
1000 kg of fuel occupies 1300 litres 1000 lb of fuel occupies 155.8 US gallons 1000 lb of fuel occupies 129.9 Imp. gallons 1000 lb of fuel occupies 20.8 ft3
Estimation of wing fuel volume
To determine the usable capacity of a fuel tank it is possible to calculate the external volume and then reduce this value to account for internal obstructions caused by structural and system components within the tank. Typically, reduce the available internal volume to 85 per cent for integral tanks and to 65 per cent for bladder or ‘bag tanks’. Note: these factors do not account for landing gear or other significant intrusions into the available space. Such factors must be considered separately when deciding the overall location of fuel tanks.
For the aircraft in this project, we can consider the fuel to be held in wing tanks on each side of the aircraft fuselage as shown in Figure 4.7. Each tank occupies the space between the leading edge and trailing edge high-lift structure and associated mechanisms. The tanks will be of the integral type. The space ahead of the ailerons will not be used for fuel tankage, as the wing section here is too thin. Also, the space in front of the inboard flap is not used for fuel volume, as this is likely to be where the main landing gear will be stowed. The generalised geometry of the fuel tank is shown in Figure 4.7a.
The cross-section areas of the spanwise ends (A1 and A2) and mid-span (Amid) sections of each tank are determined by multiplying the average tank thickness (chordwise) (T ) by the distance between the front and rear spars (W ). The cross-sectional areas of each end of a tank are added and then multiplied by half the spanwise distance between the ends (L) to give the profile volume. This volume is then multiplied by the appropriate factor for the type of tank. Hence:
Average thickness T = k · (t/c) · c
where k = factor to relate the average tank depth to the max. wing profile depth. The value depends on the shape of the section profile and the allowance made for structure. Typical values lie between 0.8 and 0.5.
(t/c) = wing section profile thickness ratio (this will vary from thicker values near the bodyside to thinner values towards the tip).
c = the local wing chord length
Cross-sectional area (A) = T · W
where W is the width of the tank
Tank profile volume = (L/6)[A1 + A2 + 4Amid] where A1 and A2 are the cross-sectional areas of the ends of the tank
Amid is the cross-sectional area at the mid-length position L is the length of the tank
“chap04” — 2003/3/10 — page 67 — #22
Project study: scheduled long-range business jet 67
Table 4.3
Section A/C centre Mid. Outer
Wing chord (c) Wing thickness (t/c) Thickness factor (k) Thickness (T ) Widths (W )
Cross-sect. areas (A)
Tank length (L) = 16.0
8.5 5.0 3.0 0.15 0.13 0.11 0.6 0.65 0.65 0.765 0.422 0.214 4.2 2.6 1.7
3.06 1.52 0.32
Tank max. volume (profile) = 21.2 m3
The tank measurements (in metres) are quoted in Table 4.3. Total tank volume (both sides), including an 85 per cent factor for integral tankage:
Available tank volume = 2 × 21.2 × 0.85 = 36.1 m3
Required fuel mass (from section 4.6.1) = 28 323 kg (62 450 lb)
Using the volume conversion shown above, for typical aviation fuel:
Required tank volume = (28 323/1000) · 1.3 = 36.82 m3 (590 ft3)
Hence, within the accuracy of the calculation, the required fuel can be accommodated in the wing profile tanks. Extra fuel volume will be useful to extend the range of the aircraft for reduced payload operations. This could be provided by the optional external wing tanks but these would add extra drag.
4.6.4 Fuselage geometry For most aircraft, the fuselage layout can be considered in isolation to the wing and other control surfaces. The internal space requirements, set by the aircraft specification, are used to fix the central section of the fuselage. For civil aircraft, this shape is governed by the passenger cabin layout.
The fuselage width is set by the number of seats abreast, the seat width and the aisle width. The depth is set to accommodate the cargo containers below the floor and the headroom above the aisle. A circular section is preferred for an efficient structural pressure shell. This requirement may impose constraints on the preferred width and depth sizes. Although this aircraft is designed principally as an executive aircraft, we must make sure that the size is suitable for any other variants that we may want to consider as part of the aircraft ‘family’. For an aircraft of 80+ capacity, the conventional seating (mixed class) would be five abreast for economy and four abreast for business, with a single aisle. For our executive layout, four abreast would be sensible. As the aircraft mission is long range, it is necessary to provide a high comfort level. A typical maximum first-class seat is 0.7 m (27.5 in) wide. Providing a generous 0.6 m (24 in) aisle would make the cabin width 3.4 m (136 in). Adding 0.1 m (4 in) each side for structure makes the fuselage outside diameter 3.6 m (142 in). This width would allow five abreast ‘tourist’ seating with a seat width of 0.56 m (22 in). This is currently regarded as a very generous provision for this class. At six abreast the ‘tourist/charter’ seat width is 0.47 m (18.5 in). This is narrow for normal tourist provision but generous for the charter operation. The fuselage layout options are as shown in Table 4.4.
“chap04” — 2003/3/10 — page 68 — #23
68 Aircraft Design Projects
Table 4.4
Class Seats abreast Seat width Cabin internal width
Executive Tourist Charter
4 5 6
0.70 m 0.56 m 0.47 m
(4 × 0.7) + 0.6 = 3.4 m (5 × 0.56) + 0.6 = 3.4 m (6 × 0.47) + 0.58 = 3.4 m
Adding 0.2 m for the pressure cabin structure makes:
Total fuselage external diameter equal to 3.60 m (11.8 ft or 142 in)
The fuselage cross-section must also be considered in relation to the cargo pallet sizes to be accommodated below the cabin floor. This may require the fuselage profile to be altered to suit the geometry of standard containers. For example, the Boeing 757 fuselage section is 10 in deeper than the circular cabin shape. It is too early in the design process to consider such details but this aspect must be carefully studied later.
The length of the cabin is determined by the seat pitch. This varies as the class. Typical values are: executive class is 1.0 to 1.1 m (40 to 43 in), tourist is 0.8 to 0.9 m (31 to 35 in), and charter is 0.7 to 0.8 m (28 to 31 in). Using the longest executive seat pitch with four abreast seating requires a cabin length of 22 m (72 ft). With this length of cabin, the number of tourist passengers that can be accommodated is 140 and 120 for the short and longer pitches respectively. A similar calculation for the charter layout would provide 192 or 140 passengers respectively. It may not be possible for technical reasons (e.g. provision of emergency and other services) to accommodate the larger capacities calculated here. Nevertheless, the 22 m cabin length seems to offer a good starting point for the initial layout.
It is desirable to split the cabin into at least two separate sections. This makes the in-flight servicing easier and allows more options for the airline to segregate different classes. For the exclusive executive layout, this division will allow a quieter environment within the cabin. A service module (catering or toilets) is positioned at this location. External service doors and hatches are positioned here and these can act as emergency exits. The provision of service modules and the ‘wasted’ space adjacent to the doors will add about 4 metres (13 feet) to the cabin length.
The fuselage length is the sum of the cabin and the front and rear profile shaping. The front accommodates the flight deck and the rear provides attachment for the engines (in our case) and the tail surfaces. From an analysis of similar aircraft, the non-cabin length is about 15 metres.
Hence, the total fuselage length is (22 + 4 + 15) = 41 m (134 ft)
The resulting fuselage layout and geometry are shown in Figure 4.8.
4.6.5 Initial ‘baseline aircraft’ general arrangement drawing With details of the engine, wing and fuselage available, it is now possible to produce the first drawing of the aircraft. The control surface sizes are estimated from area and tail volume coefficients of other similar aircraft. The aircraft general arrangement (GA) drawing is shown in Figure 4.9. With the sizes of the major components of the aircraft available from the GA, it is possible to make the initial technical assessments.
“chap
04”—
2003/3/10—
pag
e69
—#24
BAGGAGE BAGGAGE
Executive 4 abreast Tourist 5 abreast
Executive class
Flight deckC = CloaksW = WCG = Galley
Charter 6 abreast
W W
C
W
CG
24 seats
G
= Access
36 seats 20 seats
CARGO
Fig. 4.8 Fuselage layout options
“chap04” — 2003/3/10 — page 70 — #25
70 Aircraft Design Projects
10 m (33 ft)
Scale
Fig. 4.9 Initial ’baseline’ layout
4.7 Initial estimates
Before we can estimate aircraft performance, we must more accurately determine theaircraft mass, drag, lift and engine characteristics. Detailed calculations are not shownin the sections below as they follow conventional methods. Where appropriate referenceis given to the methods used.
4.7.1 Mass and balance analysis
Each of the aircraft component masses will be estimated separately and then summed.For presentational convenience, the values below are quoted in kg only (conversion1 kg = 2.205 lb).
Wing structure
Using the estimation formula in reference 1, with:
MTOM = 81 000 kg Nult = 3.75
Wing area = 180 sq. m Wing aspect ratio = 10
“chap04” — 2003/3/10 — page 71 — #26
Project study: scheduled long-range business jet 71
Wing taper ratio = 0.3
Wing av. thickness = 15% Wing LE sweep = 30◦
R = (Mwing + Mfuel) = (8100 + 32 400) kg
Hence, the wing is calculated at: Mwing = 11 209 kg
This is 13.8 per cent of MTOM. This is uncharacteristically high for this type of aircraft.The formula is based on old and existing aircraft types. The average value of aspectratio for the source aircraft is about 6. The much higher aspect ratio of our design (10)seems to have caused a large increase in wing mass. In addition, the formula was basedon traditional metallic construction whereas our design will incorporate substantialcomposite structure. For these reasons, we will apply a reduction factor of 25 per centto the estimated mass:
Mwing = 11 209 × 0.75 = 8407 kg (10.4% MTOM)
As we have used the wing gross area, we will assume that this mass includes the flapweight.
Tail structure
With little knowledge of the tail design at this time, we will assume a representativepercentage. We know that the extra control surface (canard) will add some weight sowe will use a slightly higher percentage than normal for this type of aircraft (2.5 percent). As we will be constructing these surfaces in composite materials, we will apply atechnology reduction factor of 25 per cent.
Mtail = 0.025 × 81 000 × 0.75 = 1519 kg (1.9% MTOM)
Fuselage structure
The mass of the body will be estimated using a formula for body1 with the parametervalues shown below:
MTOM = 81 000 kg; O/A length = 40.0m;
max. diameter = 3.75 m; VD = 255 m/s
Gives:Mbody = 9232 kg
Increasing by 8 per cent for pressurisation, 4 per cent for tail engine location andreducing by 10 per cent for modern materials and construction gives:
Mbody = 9306 kg (11.5 MTOM)
Nacelle structure
Based on an engine thrust of 254 kN:
Mnacelles = 1729 kg (2.1% MTOM)
“chap04” — 2003/3/10 — page 72 — #27
72 Aircraft Design Projects
Landing gear
We will assume this to be 4.45 per cent MTOM:
Mlanding gear = 3604 kg
Surface controls
Using a typical value of 0.4 × (MTOM)0.684:
Ms/controls = 911 kg
Aircraft structure
Summing the above components gives the aircraft structural mass:
Mstructure = 25 475 kg
This is 31.5 per cent MTOM which is representative of this class of aircraft.
Propulsion system
Using the quoted engine dry weight of 5250 lb (each) and a system multiplying factorof 1.43, gives:
Mpropulsion = 6810 kg (8.4% MTOM)
Fixed equipment
A typical value for this type of aircraft is 8 per cent but as we will be providing morecabin services we will increase this to 10 per cent MTOM:
Mfix/equip = 8100 kg
Aircraft empty (basic) mass
Summing the structure, propulsion and fixed equipment masses gives:
Mempty = 40 385 kg (49.9% MTOM)
Operational empty mass (OEM)
Adding the flight crew (2 × 100 = 200 kg), cabin crew (4 × 70 = 280 kg), cabin serviceand water (@21.5 kg/pass. = 1724 kg) to the aircraft basic mass gives:
MOEM = 42 589 kg (52.8% MTOM)
This is close to the assumed value from the literature search.
Aircraft zero-fuel mass (ZFM)
This is the OEM plus the passengers (80 × 80 = 6400 kg) and the passenger baggage(40 × 80 = 3200 kg), giving:
MZFM = 52 189 kg (64.4% MTOM)
“chap04” — 2003/3/10 — page 73 — #28
Project study: scheduled long-range business jet 73
Maximum take-off mass (MTOM)
In the analysis above, the MTOM has been assumed to be 81 000 kg.
Fuel mass
The aircraft zero-fuel mass (ZFM) and the assumed maximum take-off mass define theavailable fuel mass:
Mfuel = MTOM − ZFM
Hence,
Mfuel = 81 000 − 52 189 = 28 811 kg (35.6% MTOM)
This is less than previously assumed so it will be necessary to recalculate the fuelmass ratio using a more detailed method. The Breguet range equation can be used ifassumptions are made for the aircraft (L/D) ratio and the engine fuel consumption (c).
Range = (V/c)(L/D) loge(M1/M2)
where V = cruise speed = M0.85∗ = 255 m/s = 485 ktsc = assumed engine fuel consumption = 0.55 N/N/hr(L/D) assumed to be = 17 in cruiseM1 = start mass = MTOM = 81 000 kgM2 = end mass = ZFM = 52 189 kg
∗the cruise speed is set to avoid incurring significant drag rise. Typically, a 20 pointdrag count (one drag count = 0.0001) rise sets this speed.
With the speed in knots, this gives:
Range = 6589 nm
Although this may seem close to the specified range of 7000 nm, it is necessary toaccount for the fuel allowances. Using the formula shown below,1 the required designrange can be used to calculate the equivalent still-air-range (ESAR). This includes thefuel reserves (diversion and hold) and other contingency fuel.
ESAR = 568 + 1.06 design range
Hence, the required ESAR (for our specified design range of 7000 nm) = 7988 nmReversing this process with the 6589 nm range calculated above would only give a
design range of 5927 nm. This shows that there is a substantial shortfall in the designrange. The original assumption of 0.35 for the fuel fraction seems to be in error for ourdesign. This is a major error as the aircraft is not viable at an MTOM of 81 000 kg.We can use the Breguet equation above to determine a viable fuel ratio for the 7988 nmESAR.
7988 = (485/0.55)17(loge(M1/M2))
(M1/M2) = 1.704
M2 = M1 − Mfuel
(Mfuel/MTOM) = 1 − (1/1.704) = 0.413
This is a big change to the value used in the initial MTOM prediction. We will usethe aircraft empty mass ratio of 0.495 determined in the component mass evaluation
“chap04” — 2003/3/10 — page 74 — #29
74 Aircraft Design Projects
above, in a new estimation of MTOM:
MTOM = 110 520/(1 − 0.495 − 0.413) = 114 348 kg (25 214 lb)
Note: the denominator in the expression above is only 0.092. This makes the evaluationvery unstable. For example, if the empty mass and fuel mass ratios are incorrect by only+/− 1 per cent the MTOM would change to 146 111 and 93 928 kg respectively. Thesevalues are 28 per cent more and 18 per cent less than the predicted value. This illustratesthe inappropriate use of the initial MTOM prediction method when the denominatoris small. However, as we do not have another prediction, we will have to use the 114 348value and, as quickly as possible, validate it with a detailed component mass prediction.
As we have still not evaluated the aircraft performance we will need to use the thrustand wing loading values (0.32 and 450) determined in the literature survey. The newvalue of MTOM will force a change in the engine thrust and wing area:
Engine thrust (total) = 359 kN (80 710 lb)
Wing area (gross) = 254 sq. m (2730 sq. ft)
As other alterations are likely to follow, changes to the engine selection caused by theabove will not be considered at this point in the design process.
The values above, together with the resulting heavier MTOM, will change the com-ponent mass predictions made earlier. Using the same methods, the aircraft massstatement (kg) is calculated as listed below:
Wing structure = 13 224 (11.6% MTOM)Tail structure = 2859 (2.5% MTOM)Body structure = 10 278 (9.0% MTOM)Nacelle structure = 2441 (2.1% MTOM)Landing gear = 5088 (4.45% MTOM)Surf. controls = 1153 (1.0% MTOM)
STRUCTURE = 35 043 (30.6% MTOM)Propulsion = 9625Fixed equip. = 11 435
A/C EMPTY = 56 103 (49.1% MTOM)Operational items = 1724Crew = 480
OEM = 58 307 (51.0% MTOM)Passengers = 6400Baggage = 3200
ZFM = 67 907 (59.4% MTOM)Fuel = 46 441 (40.6% MTOM)
MTOM = 114 348 (100% MTOM)(252 137 lb)
Note: the fuel ratio is still slightly under the requirement. The calculation should bedone again to obtain the correct ratio.
Applying the Breguet range equation with values determined above (M1 = 114 348and M2 = 67 068 kg) gives a range of 7812 nm. (A spreadsheet method with an iterativecalculation function is very useful in this type of work.) As we have still made somegross assumptions in the calculations above (e.g. if the aircraft L/D ratio is 18 insteadof 17 the range would increase to 8272 nm), we will continue the design process usingthe 114 348 MTOM value.
“chap04” — 2003/3/10 — page 75 — #30
Project study: scheduled long-range business jet 75
Before moving on to the aerodynamic calculations, it is necessary to redraw theaircraft with larger wings, control surfaces and engines. The fuselage shape will notchange. The overall aircraft layout will be similar to that shown later in Figure 4.11.Assuming that the internal wing volume increases as the cube of the linear dimensions,the wing will be able to hold 52 668 kg (116 134 lb). This will be large enough to holdthe extra fuel mass of the bigger aircraft.
4.7.2 Aerodynamic estimations
Conventional methods for the estimation of aircraft drag can be used at this stage inthe design process. As it is assumed that, with careful detail design, the aircraft can flyat speeds below the critical Mach number, substantial additions due to wave drag canbe ignored. Therefore, only zero-lift and induced drag estimations are required.
Parasitic drag is estimated for each of the main component parts of the aircraft andthen summed to provide the ‘whole aircraft’ drag coefficient. The component dragareas are normalised to the aircraft reference area (normally the wing gross area).
Component parasitic drag coefficient, CDo = Cf FQ[Swet/Sref ]where Cf = component skin friction coefficient. This is a function of local
Reynolds number and Mach numberF = component form (shape) factor which is a function of the geometryQ = a multiplying factor (between 1.0 and 1.3) to account for local
interference effects caused by the componentSwet = component wetted areaSref = aircraft drag coefficient reference area (normally the wing gross area)
Aircraft not in the ‘clean’ condition (e.g. with landing gear and/or flaps lowered, withexternal stores or fuel tanks) will also be affected by extra drag (�CDo) from theseitems. The extra drag values will be estimated from past experience. Several textbooks(e.g. references 1 to 5) and reports provide data that can be used.
Whole aircraft parasitic drag, CDo =∑
[component CDo] +∑
[�CDo]
From the previous analysis the reference area will be 254 sq. m (2730 sq. ft).
Cruise (at 35 000 ft and M0.85)
The component drag estimations for the aircraft in this clean configuration are shownin Table 4.5.
From reference 1, induced drag coefficient,
CDi = (C1/C2/πA)C2L + (0.0004 + 0.15CDo)C2
L
where (C1 and C2) are wing geometry factors (close to unity) and (A) is the wingaspect ratio.
For our design the equation above gives,
CDi = 0.035C2L
“chap04” — 2003/3/10 — page 76 — #31
76 Aircraft Design Projects
Table 4.5
Component R. No.∗ Cf F Q Swet (�CDo)
Wing 3.32 0.00234 1.50 1.0 432.0 0.00593H controls 1.86 0.00255 1.31 1.2 59.7 0.00094V control 2.99 0.00237 1.32 1.2 33.7 0.00050Fuselage 2.65 0.00175 1.07 1.0 437.4 0.00321Nacelles (2off) 3.42 0.00231 1.5 1.0 84.6 0.00116Secondary items 0.00192
∑(aircraft CDo) 0.01376
∗ R. No. = Reynolds number (×10−7)
4.7.3 Initial performance estimates
Cruise
Hence, at the start of cruise:
CD = 0.0137 + 0.035C2L and CL = 0.339,
Making,CD = 0.01 774
Therefore, at the start of cruise,
Aircraft drag = 54.3 kN
Assuming, at this point, the aircraft mass is (0.98 MTOM), then L/D ratio = 19.1
Engine lapse rate to cruise altitude = 0.197 (based on published data1)Hence, available engine thrust = 0.197 × 359 = 70.7 kN
This shows that the engine cruise setting could be 77 per cent of the take-off rating.At the end of the cruise phase, assuming that aircraft mass is (0.65 MTOM) the
aircraft CL reduces to 0.225 if the cruise height remains constant. This reduces theaircraft L/D ratio to 14.5. This would increase fuel use. To avoid this penalty the aircraftcould increase altitude progressively as fuel mass is reduced to increase CL. This is calledthe ‘cruise-climb’ or ‘drift-up’ technique during which the aircraft is flown at constantlift coefficient. At the end of cruise, the aircraft would need to have progressivelyclimbed up to a height of 43 600 ft. To reach such an altitude may not be feasible if theengine thrust has reduced (due to engine lapse rate) below that required to meet thecruise/climb drag.
Cruise/climb
At the initial cruise height, the aircraft must be able to climb up to the next flight levelwith a climb rate of at least 300 fpm (1.524 m/s).
This will require an extra thrust of 6758 N.Adding this to the cruise drag gives 61.1 kN. This is still below the available thrust at
this height (approximately 86 per cent of the equivalent take-off thrust rating).Performing a reverse analysis shows that an aircraft (T/W ) ratio of 0.276 would be
adequate to meet the cruise/climb requirement.
“chap04” — 2003/3/10 — page 77 — #32
Project study: scheduled long-range business jet 77
Landing
The two-dimensional (sectional) maximum lift coefficient for the clean wing is cal-culated at 1.88. The finite wing geometry and sweep reduce this value to 1.46.Adding simple (cheap) trailing edge flaps (�CL max = 0.749) and leading edge device(�CL max = 0.198) produces a landing max. lift coefficient for the wing of 2.41.
At this stage in the design process, it is sufficient to estimate the landing distanceusing an empirical function. Howe3 provides as simplified formula that can be used toestimate the FAR factored landing distance. The approach lift coefficient (CLapp) is afunction of the approach speed. This is defined in the airworthiness regulations as 1.3times the stall speed in the landing configuration. Hence CLapp is (2.4/1.69) = 1.42.Assuming the landing mass is (0.8 MTOM), the approach speed is estimated as 64 m/s(124 kt). This equates to a landing distance of:
FAR landing distance = 1579 m (5177 ft)
This is less than the design requirement of 1800 m.
Take-off
Reducing the flap angle for take-off decreases the max. lift coefficient to 2.11.As for the landing calculation, it is acceptable at this stage to use an empirical function
to determine take-off distance (TOD). For sea level ISA conditions, reference 3 givesa simplified formula for the FAR factored take-off distance. Assuming lift-off speed is1.15 stall speed, the lift coefficient at lift-off will be (2.11/1.152) = 1.59, with (T/W ) =0.32 and (W/S) = 450 × 9.81 = 4414 N/sq. m, the following values are calculated:
Ground run = 1292.6 m, Rotation distance = 316.1 m, Climb distance = 81.6 m,
FAR TOD = 1690 m (5541 ft)This easily meets the previously specified 1800 m design requirement.
Second segment climb with one engine inoperative (OEI)
For the second segment calculation the drag estimation follows the same procedure asdescribed above but in this case the Reynolds number and Mach number are smaller.The undercarriage is retracted and therefore does not add extra drag but the flaps arestill in the take-off position and will need to be accounted for in the drag estimation.The failed engine will add windmilling drag and the side-slip (and/or bank angle) ofthe aircraft will also add extra drag.
Using published methods to determine flap drag3 and other extra drag items1:
CL = 1.59, CDO = 0.0152, CDI = 0.0376,
�CDflaps = 0.015, �CDwdmill = 0.0033, �CDtrim = 0.0008
These values determine an aircraft drag = 116.1 kNThrust available (one engine), at speed V2 = 161.5 kNThis provides for a climb gradient (OEI) = 0.0405This is better than the airworthiness requirement of 0.024To achieve this requirement would demand only a thrust to weight ratio of 0.254
Later in the design process, it will be necessary to determine the aircraft balanced fieldlength (i.e. with one engine failing during the take-off run).
“chap04” — 2003/3/10 — page 78 — #33
78 Aircraft Design Projects
450
0.27
0.30
0.32
500
Wing loading (W /S ) (kg /m2)
550
Thru
st lo
adin
g (T
/W)
Original design point
New designpoint
Cruise/climb300 ft /m @ 37 000‰
Take-off1800 m
Landing1800 mVAP= 71.2
Fig. 4.10 Constraint diagram
4.7.4 Constraint analysis
The four performance estimates above have indicated that the original choice of aircraftdesign parameters (T/W, W/S) may not be well matched to the design requirementsas each of the design constraints was easily exceeded. The assumed thrust and wingloadings were selected from data on existing aircraft in the literature survey. It seemsthat as our design specification is novel, this process is too crude for our aircraft. As wenow have better knowledge of our aircraft geometry, it is possible to conduct a moresensitive constraint analysis. The methods described above will be used to determine theconstraint boundaries on a T/W and W/S graph. The results are shown on Figure 4.10.
Moving the design point to the right and downwards makes the aircraft more efficient.The constraint graph shows that it would be possible to select a design point at T/Wat 0.3 and W/S at 500 kg/sq. m (102.5 sq. ft). Recalculating the aircraft mass using thesame method as above and with these new values gives:
Wing structure = 11 387Tail structures = 2025Body structure = 10 050Nacelle structure = 2161Landing gear = 5088Surface controls = 1109
STRUCTURE MASS = 31 538 (29.2%)Propulsion mass = 8520Fixed equipment = 10 800
AIRCRAFT EMPTY MASS = 50 858 (47.1%)OPERTN EMPTY MASS = 52 862 (49.0%)ZERO FUEL MASS = 62 462 (57.8%)
“chap04” — 2003/3/10 — page 79 — #34
Project study: scheduled long-range business jet 79
Fuel mass = 45 538 kg (42%)MAX. MASS (MTOM) = 108 000 (100%)
(23 814 lb)
Using this mass and our new thrust and wing loading ratios gives:
• Total engine thrust (static sea level) = 317.8 kN (71 450 lb)• Gross wing area (reference area) = 216 sq. m (2322 sq. ft)
Assuming the wing tank dimensions are proportional to the wing linear size, the newwing area could accommodate 41 460 kg (91 400 lb) of fuel. This is less than predictedabove (by 9 per cent). As we have made several assumptions and have not made adetailed analysis of the geometry and performance, we will delay the effect of this onthe design of the wing until later in the design process.
4.7.5 Revised performance estimates
Range
With the cruise speed of 250 m/s (485 kt), assumed SFC of 0.55 force/force/hr, aircraftcruise L/D ratio of 17, initial mass (M1) = MTOM (108 000 kg), and final mass(M2) = ZFM (62 462 kg) gives:
Range = 8209 nm
This is slightly longer than the previously estimated ESAR of 7988 nm but is withinour calculation accuracy. The fuel ratio in the new design is 42.2 per cent whereas only41.3 per cent is required therefore we have about 900 kg slack in the zero fuel estimation.
Cruise
With the new mass and geometry, the drag polar (start of cruise, 35 000 ft @ M0.85) iscalculated as:
CD = 0.0148 + 0.0352C2L
At the start of cruise, the lift coefficient is 0.40, hence CD = 0.0204.This equates to a drag = 53.1 kN (11 938 lb), and hence a cruise L/D = 19.5
The engine lapse rate at cruise is 0.197. Therefore the available thrust at the cruisecondition = 0.197 × 317.8 = 62.6 kN (14 073 lb)This gives an engine setting in cruise of 85 per cent of the equivalent take-off rating
Cruise climb
Adding a climb rate of 300 fpm at the start of cruise makes the required thrust at thestart of cruise = 59.4 kN (13 354 lb). This is 95 per cent of max. take-off thrust rating.
Landing
The approach speed is 64.5 m/s (125 kt)This seems reasonable for regional airport operationsThe landing distance is calculated as 1594 m (5225 ft)This is well below the 1800 m design requirement
“chap04” — 2003/3/10 — page 80 — #35
80 Aircraft Design Projects
10 m
Scale
Fig. 4.11 Refined baseline layout
Take-offThe take-off distance is 1790 m (5869 ft)The balanced field length is 1722 m (5647 ft)These satisfy the design requirement of 1800 mThe second segment climb gradient (OEI) = 0.033This satisfies the airworthiness requirement of 0.024
All of the design requirements have been achieved with the new aircraft geometry.It is now possible to draw the refined general arrangement of our aircraft, Figure 4.11.
4.7.6 Cost estimations
Using the methods described in reference 1:
For an aircraft OEM = 52 862 kg, the aircraft purchase price will be $42M (1995)Assuming an inflation rate of 4 per cent per yearThis brings the 2005 aircraft price = $62M
For engines of about 40 000 lb TO thrust, the price would be $4.0M (1995)
“chap04” — 2003/3/10 — page 81 — #36
Project study: scheduled long-range business jet 81
For two engines (2005 prices) = $12MAirframe cost = $62M – $12M = $50M (i.e. aircraft price less engines)Assume 10% spares for airframe = $5.0MAssume 30% for engine spares = $3.6M
Total investment = $70.6M
Assuming depreciation to 10 per cent over 20 yearsAnnual depreciation = (0.9 × 70.6)/20 = $3.18 MAssume interest on investment cost of 3.5% per yr = $2.47 MAssume insurance 0.5% per yr of investment = $0.35 M
Total standing charges per year = $6.00 M
For the cruise range of 7000 nm at 485 kt, the flight time will be 14.4 hrAdd 0.75 hours to account for airport ground operations = 15.15 hr
Total block time = 15.15 hr
Some cost methods use this time in the calculation of DOC. Others use the flight timeonly. We will use the flight time in the calculations below.
Assume aircraft utilisation of 4200 hr per year (typical for long-range operations)
Standing charges per flying hour = $1429
Crew costs (1995) per hr = 2 × 360 for flight crew + 4 × 90 for cabin crew
= $1080 = $1594 per hr (2005)
Landing and navigation charges per flight = 1.5 cents/kg MTOM = $1620 per flightGround handling charge = $3220 per flightTotal airport charges = $4840 per flight = $336 per flight hrFrom the mass and range calculations: fuel used for ESAR (8209 nm) = 45 538 kgEstimated fuel used for the 7000 nm design range = 40 300 kgAssuming that little fuel is burnt in the ground,Fuel used per flight hour = 40 300/14.4 = 2798 kgFuel volume = 2798/800 = 3.5 sq. m = 3500 litres = 3500/3.785 = 924 US galAssuming the price of fuel is 90 cents per gal,
Fuel cost = $832 per flight hour
As maintenance costs are too difficult to assess at this time in the design process, wewill assume them to account for 15 per cent of the total operating cost.
Total operating cost $ per flight hourStanding charges = 1429Crew cost = 1594Airport charges = 336Fuel cost = 832Maintenance costs = (15%) (739)
= $4930 per flight hour
Hence DOC, Total stage cost = 4616 × 14.4 = $70 996Aircraft mile cost = 66 477/7000 = $10.14Seat mile cost (100% load factor) = 12.68 cents
“chap04” — 2003/3/10 — page 82 — #37
82 Aircraft Design Projects
Operators who lease the aircraft use ‘cash DOC’ to determine flight cost. They add thelease charges to their indirect costs as they are committed to this expense regardless ofthe aircraft utilisation. Cash DOC is calculated in the same way as above but without thestanding charges. As aircraft maintenance is unaffected by the ‘accountancy’ methodused to determine DOC, the cost is assumed to be the same as used above.
Cash DOC, Operational cost = $3504 per hrTotal stage cost = $50 451Aircraft mile cost = $7.21Seat mile cost = 9.01 cents
Assuming that the ticket price (LHR–Tokyo) is $4500 for the executive-class fare:Revenue per flight (assuming 65 per cent load factor) = $234 000This compares favourably with the direct operating stage cost of $70 996Even allowing for a 100 per cent indirect operating cost (IOC) factor added to DOC,the operation would be viable
The seat mile costs calculated above are substantially larger than those quoted forhigh-capacity mixed-class services in which about 75 per cent of the seats are assigned toeconomy-class travellers. The revenue from such customers is significantly lower thanfrom the executive class as they will be charged only about 20 per cent of the higherprice fare. Without a detailed breakdown of the financial and accounting practicesof an airline, it is impossible to determine the earning potential of the new servicecompared with the existing operation. However, the revenue assessment shown aboveis encouraging enough to continue with the project.
4.8 Trade-off studies
There are many different types of trade-off studies that could be undertaken at thisstage in the design process. These range from simple sensitivity studies on the effectof a single parameter or design assumption, to extensive multi-variable optimisationmethods. The studies shown below include trade-off plots that are used to determinethe best choice of aircraft geometry. Wing loading and wing aspect ratio are chosenas the main trade-off parameters. These are regarded as the most significant designparameters for the short field, long-range requirements of the aircraft operation.
The studies shown in this section are presented as typical examples of the type ofwork appropriate at this stage of aircraft development. Many other combinations ofaircraft parameters could have been selected and in a full project analysis would havebeen performed.
4.8.1 Alternative roles and layout
As mentioned in section 4.6.4 (fuselage layout), for all aircraft design studies it is neces-sary to consider the suitability of the aircraft to meet other operational roles. Althoughthe principal objective of the project is to produce an efficient large business exclusiveaircraft, we must also consider other mixed-class variants. In this way, a family of air-craft can be envisaged. This will increase the number of aircraft produced and reducethe design and development overhead per aircraft. Recognising this requirement, thefuselage diameter was designed to be suitable not only for the four abreast executiveclass seating but also five and six abreast layouts of higher capacity options. The cabinlength of 22 metres plus 4 metres for services and egress space is a fixed parameter and
“chap04” — 2003/3/10 — page 83 — #38
Project study: scheduled long-range business jet 83
Table 4.6
Rear cabin Centre cabin Front cabin Total seats
A Executive 24 32 24 80B Mixed∗ 35 econ. 50 econ. 24 exec. 85/24 = 109C Economy 35 50 35 120D Charter∗∗ 48 72 48 (168) = 150
∗ This provides 22 per cent business occupancy.∗∗ The maximum capacity is reduced by about 10 per cent to account for extra
spacing at emergency exits.
will control the layout and capacity of alternative roles. Within this length, variouscombinations of passenger layouts can be arranged. The position of doors and ser-vice modules (toilets, cupboards and galleys) is fixed but these can be used to providenatural dividers between classes.
From the previous fuselage layout drawing (Figure 4.8), the rear cabin is 6.5 metres,centre cabin 9.0 metres and front cabin 6.5 metres long. Using seat pitches of 1.1, 0.85,and 0.75 metres for executive/business, economy and charter classes respectively resultsin the layouts shown in Table 4.6.
For civil aircraft, it is common practice to stretch the fuselage in a later developmentphase. Typically, this may increase the payload by 35 per cent. Using this value (approx-imately), the single-class, economy version would grow to 160 passengers. At the 0.85metre seat pitch, this would equate to a lengthening of the fuselage by 6.8 metres. Tomaintain aircraft balance a 2.8 m plug would be placed in the rear cabin and a 4.0 mplug forward of the wing joint. In this version the capacity of the aircraft would increaseto the values shown below:
A Executive (single class) = 104 seatsB Mixed class = 141 seats (105 econ. and 36 exec.)C Economy (single class) = 160 seatsD Charter (single class) = 204 seats
The extra capacity would require more passenger service modules and extra emergencyexits to be arranged in the new cabin. This would reduce the space available for seatingand slightly reduce the capacities shown above. Alternatively, the fuselage stretch wouldneed to be increased by about a further 1.0 to 1.5 metres (40 to 60 in). Figure 4.12 showssome of the layout options described above.
Non-civil (military) versions of the aircraft could also be envisaged. With only 0.7 seatpitch for troop carrying a total of 186 soldiers could be carried in the original aircraftand 246 in the stretched version. The large volume cabin (for a small aircraft), the longendurance and the short field capabilities would be suitable for reconnaissance andelectronic surveillance roles. In such operations, the reduced payload mass could allowextra fuselage fuel tanks to be carried to extend the aircraft duration. The high-speed,long-range performance could be useful for military transport command. Using thisaircraft would avoid diplomatic complications caused by the need to refuel in foreigncountries in conflict scenarios.
Many other versions of the aircraft may be envisaged (e.g. freighter/cargo, corporatejet, and communication platform) but these would not significantly affect the designof the current aircraft configuration.
“chap
04”—
2003/3/10—
pag
e84
—#39
W W
C
W
CG
35 seats(econ.)
50 seats (econ.)
G
24 seats (business)
W W
C
W
CG
35 seats 50 seats
G
35 seats
W W
C
W
W
W CG
42 seats 66 seats
G
42 seats
An economy class (120 PAX)
All charter class (150 PAX)
Mixed class (85 econ. + 24 business = 109 PAX)
Fig. 4.12 Fuselage development options (see also Figure 4.8 for all executive (baseline) layout)
“chap04” — 2003/3/10 — page 85 — #40
Project study: scheduled long-range business jet 85
4.8.2 Payload/range studies
For any aircraft design, it is uncommon to consider just the capability of the aircraftat the design point. Trading fuel for payload with the aircraft kept at the max. designmass results in a payload range diagram (shown in Figure 4.13).
Point A (the design point) shows the aircraft capable of flying 80 executive-classpassengers over a 7000 nm range. Points B and C relate to the alternative cabin layoutsdescribed in the section above. At point B the payload is 11 380 kg. This means that1780 kg of fuel is sacrificed for payload. At point C, 2400 kg of fuel is lost. Assumingthe aerodynamic and engine efficiencies remain unchanged, the available range in thesetwo cases reduces to 6811 and 6675 nm respectively. The reduction of about 530 nmin range for a 50 per cent increase in passenger number is a result of the low valueof (Mpay/MTO) on this design. Stretching the design to accommodate 160 passengersas described above would therefore be relatively straightforward on this design. Thisdevelopment is also shown on the payload range diagram. Even after allowing for anincrease in structure and system mass of 1000 kg, the range in this configuration onlyreduces to 5625 nm.
As the aircraft is seen to be relatively insensitive to changes in payload, it is of interestto determine the effect of passenger load factor. Commercial aircraft do not alwaysoperate at the full payload condition. For this type of operation, an average loadfactor of 70 per cent is common. With less payload, the aircraft could increase fuelload (providing that space is available to accept the extra fuel volume). At 70 per centpassenger load factor with extra fuel, the Breguet equation gives an increase in rangeof 668 nm. If extra space is not available, the aircraft at 70 per cent load factor and withnormal fuel load would be able to fly a stage length of about 7500 nm. The sensitivity
16 000
15 000
14 000
13 000
12 000
11 000
10 000
9000
5000 6000
Design range (nm)
80 PAX (executive)
109 PAX (mixed class)
120 PAX (econ.)
150 PAX (charter)
160 PAX (econ.)
Payload (kg)
7000 8000
Stretched fuselage
MTOM = 108 000 kg
MTOM = 108 000 kg
A
B
C
D
Fig. 4.13 Payload/range diagram (developments)
“chap04” — 2003/3/10 — page 86 — #41
86 Aircraft Design Projects
of the range calculation to passenger load factor raises the question of the choice of therealistic design payload. Designing for the 0.7 × 80 passenger load would significantlyreduce the aircraft max. take-off mass and fuel load. This would considerably reducethe stage cost and aircraft price.
The payload/range study has shown that the loading conditions around the designpoint must be carefully considered. As the effect of range and associated fuel is unchar-acteristically sensitive on this aircraft it is important to reconsider the original designspecification to account for this aspect.
4.8.3 Field performance studies
The evaluation of field performance calculated earlier (section 4.7.3) was concernedwith the aircraft at the design condition only. The predictions can be recalculated forvariations in the aircraft parameters. In the earlier work, the take-off performance wasseen to be well matched to the 1800 m requirement. However, it is also necessary tounderstand the sensitivity of the calculation to changes in the main, aircraft designparameters (e.g. thrust and wing loadings). A carpet plot can be constructed to showthese effects (Figure 4.14).
Note: none of the study points achieves the 1400 m field length originally considered(section 4.3.3). The aircraft thrust loading is shown most influential in reducing take-off distance. To investigate this further, a second trade-off study has been conducted.Keeping the wing loading constant, the wing max. lift coefficient and thrust loadinghave been varied. The results are shown in Figure 4.15.
The carpet plot shows that increasing the thrust loading to 0.32 would allow areduction in take-off lift coefficient to 2.05. This would reduce wing structure com-plexity and thereby wing mass. Obviously, an increase in engine thrust would alsoinvolve a corresponding increase in propulsion group mass. A more detailed studywould need to be done later in the design process to draw a firm conclusion to thistrade-off.
In the revised aircraft layout, the landing performance was shown to be well withinthe 1800 m design constraint. This suggests that changes could be considered to theaircraft. For a fixed wing area the two parameters that have an effect on the landing
m
1900
1800
1700
1600
1500
1400
480
500
0.30 520
0.32
0.34
T/W Wing loading kg/sq.m
Constantat 2.2
CLmax (take-off)
Fig. 4.14 Take-off distance study (T/W and W/S)
“chap04” — 2003/3/10 — page 87 — #42
Project study: scheduled long-range business jet 87
m
2000
1900
1800
1700
1600
1500
2.2
2.1
2.0
T/W CLmax (take-off)0.30
0.32
1.9
0.34
Constant wingloading 500 kg/sq. m
Fig. 4.15 Take-off distance study (T/W and CLmax)
m2000
1900
1800
1700
1600
Mland/MTOCLmax (landing)
2.2
0.85
0.80
0.75
2.4
0.90
Constant wingloading 500 kg/sq. m
Fig. 4.16 Landing distance study
performance are aircraft max. lift coefficient (in the landing configuration) and thelanding mass ratio (Mlanding/MTO). The trade-off study results are shown in Figure 4.16.
This shows that the aircraft is capable of landing in the 1800 m field at 90 per centMTOM. The max. landing lift coefficient could be reduced to 2.2 and still allow a82 per cent MTOM landing mass. This lift coefficient is the same as previously usedfor the take-off condition therefore a further set of trade-off studies should be doneto select the best combination of lift coefficients for take-off and landing. This wouldfix the flap type and deflection angles to give the optimum design combination. To dothis, more detailed aerodynamic analysis is required than is available at this stage inthe development of the aircraft.
4.8.4 Wing geometry studies
To conduct a full and accurate analysis of the wing parameters (e.g. area and aspectratio) would involve a full, multivariate optimisation method. As most of the designparameters are interconnected this would be a complex process. At this early stage in
“chap04” — 2003/3/10 — page 88 — #43
88 Aircraft Design Projects
the design process and with limited resource and time available, it is not possible toundertake such a comprehensive study. Some simplifying assumptions are necessaryto enable a sensitivity study to be done. For example, we may assume that the engineparameters are kept constant and that the less sensitive mass components (e.g. surfacecontrols, systems, etc.) are held constant or considered to be directly proportional toMTOM. With such assumptions, the results of the study can only be used to indicatethe sensitivity of variations to the design and the direction of possible changes to theexisting configuration.
A study has been completed around the current aircraft configuration to investigatethe effect of changes to wing area and aspect ratio on some of the aerodynamic andmass parameters. A series of carpet plots (Figures 4.18 to 4.23) illustrate the resultsof the study. To allow comparison to the earlier work, the study used wing loading torepresent area variations. The resulting wing area values are plotted in the carpet plots.Wing loading (kg/sq. m) and aspect ratio values selected for the study are shown below:
Wing loading 400, 450, 500, 550Aspect ratio 8, 10, 12, 14
To appreciate the geometrical implications of these changes, the extreme layouts (400/14and 550/8) together with the existing baseline configuration (500/10) are illustrated inFigure 4.17. (Note: the drawings of these aircraft are illustrative only as they have notbeen balanced.)
As expected, the study shows that increasing the size of the wing (i.e. reducing wingloading) and/or increasing aspect ratio increases the wing mass. Figure 4.18 illus-trates these effects clearly and provides quantitative data of the mass changes aroundthe design point (500/10). The increasing slope of the aspect ratio lines shows theprogressive mass penalty, especially for the larger area wings.
Wing mass, although important, represents only a component of aircraft mass. Thecombined effect on aircraft empty mass is illustrated in Figure 4.19. Although a similarpattern is seen on this plot, the changes represent a smaller proportion (about a quarterof the previous percentage values). For example, moving from the design point to point550/8 is shown to reduce wing mass by about 35 per cent but the empty mass is reduced8 per cent. Note: the wing loading of 550 was shown to violate the original take-offconstraint (Figure 4.10). Making this move would require the take-off and possibly theclimb performance to be reconsidered.
Making the wing smaller and increasing aspect ratio has a significant effect on bothparasitic and induced drag. Both will be reduced. Figure 4.20, which plots aircraftlift/drag ratio, shows how the aerodynamic efficiency of the aircraft is improved. Notethat the design point shows a value higher than that originally assumed (i.e. L/D = 17).Over the range of geometrical changes investigated the L/D ratio varies between 16and 21. This is a significant variation that shows the sensitivity of choice of winggeometry.
The aircraft L/D ratio, and max. take-off mass (discussed below) are importantparameters in the calculation of the required fuel mass to fly the 7000 nm stage length.Assuming that the cruise speed and engine specific fuel consumption remain unchangedfrom their previous values, the resulting fuel mass calculations are shown in Figure 4.21.At each of the wing loading lines the ‘optimum’ aspect ratio value moves progressivelyfrom about 9 for the large wing to 14 for the smallest wing. At the design wing loadingof 500, there appears to be a small advantage to increasing aspect ratio from the designvalue of 10 to 12. Extending to 14 is not seen to be worthwhile.
“chap04” — 2003/3/10 — page 89 — #44
Project study: scheduled long-range business jet 89
Extreme(large)400/14
Extreme(small)550/8
Baselinegeometry
500/10
Fig. 4.17 Geometrical variations
Aircraft take-off mass (MTOM) is dependent on both structure mass and fuel mass.The studies above have shown that the wing geometrical changes may increase struc-ture mass but then reduce fuel mass. The combined effect is shown in the MTOMcarpet plots (Figure 4.22). Due to the reduced fuel mass, the significance of the struc-ture mass changes is eroded but the overall pattern remains similar to the emptymass plots discussed earlier. A move from the design point, to a wing loading of550 kg/m2
(112.6 lb/ft2) and aspect ratio of 8 (i.e. 550/8) would reduce MTOM by
about 5 per cent. This is a significant reduction and is worth investigating further if theeconomic studies described below confirm this advantage.
Wing area is a function of wing loading and MTOM. To show the dimensional effectsof the changes in these parameters the absolute values for wing area have been plotted(Figure 4.23). Note the significance of aspect ratio on the larger wings and the relativelylow sensitivity for small wings.
“chap04” — 2003/3/10 — page 90 — #45
90 Aircraft Design Projects
1000 kg
20
18
16
14
12
10
8
Aspect 8ratio
Wing loading 400(kg/sq. m)
14
12
450
500
550
10
Fig. 4.18 Trade-off study: wing mass
1000 kg66
64
62
60
58
56
54
52
50
48
14
12
10
8
400
450
500
550
Fig. 4.19 Trade-off study: aircraft empty mass
“chap04” — 2003/3/10 — page 91 — #46
Project study: scheduled long-range business jet 91
21550
500
450
40012
10
8
14
20
19
18
17
16
Fig. 4.20 Trade-off study: cruise L/D ratio
46
550
500
450
400
12108 14
44
42
40
38
32
34
36
1000 kg
Fig. 4.21 Trade-off study: stage fuel mass
4.8.5 Economic analysis
The results from the studies above can be used, together with operational data, toassess the economic viability and sensitivity to the aircraft geometrical changes. Theaircraft price is related to the aircraft empty mass and engine size. The cost of fuel isproportional to fuel mass. Other operational costs are related to aircraft take-off mass.Hence, changes to the aircraft configuration will affect both aircraft selling price andoperating costs. For civil aircraft designs, these two cost parameters are often selectedas the principal design drivers (optimising criteria). Although the aircraft configuration
“chap04” — 2003/3/10 — page 92 — #47
92 Aircraft Design Projects
127
123
119
115
111
103
107
1000 kg
8
10
12
14
400
450
500
550
Fig. 4.22 Trade-off study: aircraft max. TO mass
320
300
280
260
240
180
200
220
sq. m
810
12
14
400
450
500
550
Fig. 4.23 Trade-off study: wing area
may not be selected at the optimum configuration for these parameters, the design teamwill need to know what penalty they are incurring for designs of different configuration.
All of the cost calculations have been normalised to year 2005 dollars by applyingan inflation index based on consumer prices. Several separate cost studies have beenperformed as described below.
“chap04” — 2003/3/10 — page 93 — #48
Project study: scheduled long-range business jet 93
64
66
68
70
72
74
76
78
80
82
84
86
88
8
10
12
14400
450
500
550
$m (2005)
Fig. 4.24 Trade-off study: aircraft price
Aircraft price
Aircraft price is one component in the evaluation of total investment. This includesthe cost of airframe and engine spares. For this aircraft, the total investment is about12 per cent higher than the aircraft price.
Figure 4.24 shows the variation of aircraft price for the geometrical changes con-sidered previously. At the design point the price is estimated to be $70.5 m. The carpetplot shows that this price would fall by about 5 per cent if the aspect ratio was reducedto 8, and by about 9 per cent if the configuration was moved to point 550/8. The effect ofreducing wing loading progressively increases aircraft price (e.g. reducing wing loadingto 400 kg/sq. m increases the price by 9 per cent). Similarly, increasing wing aspectratio increases price (e.g. moving from 10 to 14 increases the price by over 10 per cent).Without consideration of other operating costs, the main conclusion of this study is tomove the design point to lower values of both wing loading and aspect ratio.
Direct operating cost (DOC) per flight
There are two fundamentally different methods of estimating aircraft DOC. The tradi-tional method includes the depreciation costs of owning the aircraft. On this aircraft,this would be about 33 per cent of the total DOC. If the aircraft operator leases theaircraft, the annual cost of the aircraft is regarded as a capital expenditure. This wouldbe considered as an indirect aircraft operating cost. In this case, the aircraft standingcharges (depreciation, interest and insurance) are not included in the calculation andthe resulting cost parameter is termed ‘Cash DOC’. It is important to calculate bothof the DOC methods. The results of the DOC calculations are shown in Figures 4.25and 4.26.
The DOC per flight at the design point (500/10) is $72 740. This figure would bereduced by 3 per cent if the design was moved to point 550/8 and still satisfy the technicaldesign requirements. Increasing wing area and/or aspect ratio from the design pointis not seen to be advantageous. At the design point the Cash DOC is estimated to be
“chap04” — 2003/3/10 — page 94 — #49
94 Aircraft Design Projects
72
70
74
76
78
80
82
8
10
12
14
400
450
500
550
$1000
Fig. 4.25 Trade-off study: DOC per flight
48
50
52
54
8 1012
14
400
450
500
550
$1000
Fig. 4.26 Trade-off study: cash DOC per flight
$49 470. In this case, curves are seen to be flatter than for the full DOC values. Thisresults in optimum points for aspect ratio. At the design point, the existing value ofaspect ratio is seen to be optimum. Moving to the higher wing loading (550), if feasible,would reduce Cash DOC by about 2 per cent.
It is of interest to note that the design conclusions from the two DOC methods aredifferent. This implies that the design strategy to be adopted is conditional on theaccounting practices used by the operator. This is a good example of the need for thedesigners to understand the total operating and business environment in order to selectthe best aircraft configuration.
“chap04” — 2003/3/10 — page 95 — #50
Project study: scheduled long-range business jet 95
Seat mile cost
The cost of flying the specified stage (design range) is dependent on the payload. Inthe DOC calculations above, the aircraft has been assumed to be operating at fullpayload. This is conventional practice as it allows the maximum seats to be used in theevaluation of seat mile costs. Flying at max MTOM, the DOC per flight does not varywith passenger numbers. The seat mile cost (SMC) shown in Figures 4.27 and 4.28(for DOC and Cash DOC respectively) are evaluated for the 80-seat executive versionof the aircraft.
Other versions have been evaluated at the design point and are listed in Table 4.7.Note the powerful effect of passenger numbers in reducing SMC and the substan-
tial reduction in the Cash SMC method. When using values from other aircraft it isimportant to know the basis on which cost data has been calculated.
12.6
13.0
13.4
13.8
14.2
14.6
cents
8
10
12
14
400
450
500
550
Fig. 4.27 Trade-off study: seat-mile cost (SMC)
8.6
8.8
9.0
9.2
9.4
9.6
cents
810
1214
400
450
500
550
Fig. 4.28 Trade-off study: cash SMC
“chap04” — 2003/3/10 — page 96 — #51
96 Aircraft Design Projects
Table 4.7
Layout Passengers SMC Cash SMC
Executive only 80 12.99 8.83Mixed class 107 9.74 6.62Economy only 120 8.70 5.91Charter only 150 6.85 4.68
The SMC and Cash SMC carpet plots show a similar pattern to the DOC figures.The SMC curves show an advantage to reducing aspect ratio for all values of wingloading but this advantage reduces with higher wing loadings. For the design point, areduction of aspect ratio to 8 would be recommended. For the Cash SMC calculation,the best aspect ratio varies from 8 at low loading to 12 at the high loading. For thedesign point, the value of 10 is seen to be about ‘optimum’.
4.9 Initial ‘type specification’
At the end of the initial concept stage it is important to record all of the known detailsof the current design. This document forms the initial draft of the aircraft type spec-ification. As the design evolves over subsequent stages, this document will be amendedand enlarged until it forms the definitive description of the final configuration. Theinitial draft will form the input data for the next stage of the design process. Thesections below are typical of a professional aircraft specification.
4.9.1 General aircraft description
This aircraft is designed for exclusive, business/executive, long-range routes fromregional airports. Although apparently conventional in configuration, it incorporatesseveral advanced technology features. These include natural laminar flow control,composite material and construction, enhanced passenger cabin services and com-fort standards, and three-surface control and stability. The single aisle cabin layout isarranged to accommodate four abreast seating for the baseline executive configura-tion. In other configurations it will provide five abreast economy class seating and sixabreast charter operations. In these versions the increased passenger numbers reducethe range capability of the aircraft. In the baseline executive layout, space for 80 sleep-erettes at 1.1 m (44 in) pitch is available. For the other layouts, 120 economy seats at0.8 m (32 in) pitch or 150 charter seats at 0.7 m (28 in) pitch are feasible. Toilet, galleyand wardrobe provision is adjusted to suit the layout using fixed service facilities inthe cabin. Emergency exits and other safety provisions meet FAR/JAR requirements.Underfloor cargo and baggage holds are positioned fore and aft of the wing/fuselagejunction structure.
To reduce engine operating noise intrusion into the cabin, during the long enduranceflights, the engines are positioned at the rear of the fuselage, behind the cabin pressurebulkhead. Several existing and some proposed new engine developments are suitableto power the aircraft. This provides commercial competitiveness and flexibility to the
“chap04” — 2003/3/10 — page 97 — #52
Project study: scheduled long-range business jet 97
potential airline customers. All the engines are modern, medium-bypass (typically 6.0)turbofans offering efficient fuel economy.
The modern, aerodynamically efficient, high aspect ratio wing layout provides goodcruise efficiency. A lift to drag ratio in cruise of 19 is partly achieved due to the aerody-namic section profiling and the provision of natural laminar flow. Leading and trailingedge, high-lift devices provide the short field performance required for operation fromregional airfields.
The three-surface (canard, main wing and tail) layout offers a reduction to trim dragin cruise and improved ride comfort. Integrated flight control systems with fly-by-wire actuation to multi-redundant electric/hydraulic controllers provide high levels ofreliability and safety.
Aircraft manufacture combines established high-strength metallic materials withnew composite construction techniques. The combination of conventional andnovel structural and manufacturing practices offers reduced structural weight withconfidence.
4.9.2 Aircraft geometry
Principal dimensionsOverall length = 43.0 m, 141 ftOverall height = 13.0 m, 42.6 ftWing span (total) = 48.0 m, 157 ft
Main wingGross (ref.) area = 216 sq. m, 2322 sq. ftAspect ratio = 10Sweepback (LE) = 22◦Mean chord = 4.65 m, 15.25 ftTaper ratio = 0.3Thickness (mean) = 11%
Control surfacesHorizontal tail area = 20.0 sq. m, 215 sq. ftVertical (fin) area = 15.5 sq. m, 167 sq. ftCanard area = 7.0 sq. m, 75 sq. ft
Fuselage/cabinFuselage length = 40.0 m, 131 ftCabin outside dia. = 3.6 m, 142 inPass. cabin length = 22.0 m, 72 ft
Landing gearWheelbase = 18.0 m, 59 ftTrack = 8.25 m, 27 ft
Engines (two)Various types, static SL thrust (each) = 160 kN, 35700 lb
4.9.3 Mass (weight) and performance statements
Mass statement
Aircraft empty mass 50 858 kg, 112 142 lbAircraft operational mass 52 862 kg, 116 560 lbAircraft max. (design) mass 108 000 kg, 238 140 lb
“chap04” — 2003/3/10 — page 98 — #53
98 Aircraft Design Projects
Baseline (executive) version (80 PAX)Zero fuel mass 62 462 kg, 137 729 lbPayload 9600 kg, 21 168 lbFuel load 45 538 kg, 10 0411 lbStill air range 7200 nm
Mixed-class version(107 PAX)Payload 11 380 kg, 25 093 lbFuel 43 758 kg, 96 486 lbStill air range 6810 nm
All economy version (120 PAX)Payload 12 160 kg, 26 812 lbFuel 44 138 kg, 97 324 lbStill air range 6675 nm
Charter version (150 PAX)Payload 13 450 kg, 29 657 lbFuel 41 688 kg, 91 922 lbStill air range 6350 nm
Stretched version (160 economy PAX)Payload 16 000 kg, 35 280 lbStill air range 5600 nm
Performance statement
(baseline version with 80 PAX and fuel):Cruise speed = M0.85Cruise altitude = 36 000 ftInitial cruise altitude climb rate = 300 fpmTake-off distance = 1790 m (5870 ft)Balanced field length = 1720 m (5670 ft)Second segment climb gradient = 0.033Approach speed = 125 ktLanding distance = 1594 m (5225 ft)
4.9.4 Economic and operational issues
Cost statement
(baseline aircraft, 2005 US dollars)Aircraft price = $62.0 MTotal investmt/aircraft = $70.6 MStanding charges/yr = $6.0 MStanding charges/flt hr = $1430DOC/fl hr = $4930Stage cost = $71 000Aircraft mile cost = $10.14Seat mile cost (100% PAX) = 12.7 cents
Cash DOC/flt hr = $3500Total stage cash cost = $50 450Aircraft cash mile cost = $7.20Cash seat mile cost (100%) = 9.0 cents
“chap04” — 2003/3/10 — page 99 — #54
Project study: scheduled long-range business jet 99
Operational statement
The aircraft is capable of stretching to accommodate up to 204 charter seats. In amilitary role the baseline aircraft can seat 186 soldiers and in the stretched version 246troops. In each of these versions it would be possible to fly 7000 nm (unrefuelled).
Other roles for the aircraft could include:
• Civil corporate jet• Freighter• Military refuelling tanker• Communication platform• Military surveillance aircraft• Military supply aircraft
These versions of the aircraft have not been considered in the overall geometrical layoutof the aircraft in the initial design process. A short study would be appropriate when theinitial baseline study has been completed to identify any small changes to the aircraftlayout to accommodate any of the above roles.
4.10 Study review
This aircraft project has shown how, for a relatively simple aircraft, the design processis taken from the initial consideration of the operational requirements to the end ofthe concept design phase. The intervening stages have shown how the aircraft designevolves during this process. This showed that the initial configurational assumptionsfor thrust and wing loadings, based on data from existing aircraft, were found to be inerror because of the unique operational performance of the aircraft. A more efficientaircraft layout was identified. Even the revised configuration was shown capable ofimprovement by the trade studies. For most aircraft projects, this iterative process iscommonplace.
The economic assessment of the aircraft indicated that the project was viable andtherefore worth taking into the next stage of development.
Due to time and resource restrictions in the conceptual stage, several technical aspectsof the design have not been fully analysed. These include:
• The stability and control analysis of the aircraft including the assessment of the effectof the three-surface control layout.
• The aerodynamic analysis of the laminar flow control system and the associatedstructural and system requirements.
• The aircraft structural analysis and the realisation of the combined conventional andcomposite structural framework.
• The aircraft systems definition and the associated requirements for the new executive-class communication and computing facilities.
• The special requirements for aircraft servicing and handling at regional airports.• The detailed trade-off studies applied to the field requirements (e.g. the definition
of aerodynamic (flap design and deflection), propulsion (T/W ), structures, systemsand costs).
• The assessment of the overall market feasibility of the project.
Each of the topics in the list above involves work that is either comparable with, orexceeds, the work that has already been done on the project. In industry, progress-ing to the next stage of aircraft development would involve a 20- to 50-fold increase
“chap04” — 2003/3/10 — page 100 — #55
100 Aircraft Design Projects
in technical manpower. To commit the company to this expenditure is a significantinvestment. A decision to proceed would only be taken after discussions with potentialairline customers.
If the type of operation envisaged by this project is seen to be attractive, it willstimulate competition. This may come from airframe manufacturers who could modifyexisting aircraft to meet the specification. It is essential that the design team of the newaircraft anticipate this threat. They will need to conduct their own studies on themodifications to the aircraft that may be used as competitors. These are studies thatrequire substantial effort, but in completing them, the advantages of the new designcan be identified. This information will be useful to the technical sales team of the newaircraft and used to counteract the threat from the ‘old-technology’, ‘modified’ existingtypes.
References
1 Jenkinson, L. R., Simpkin, P. and Rhodes, D., Civil Jet Aircraft Design, 1999, AIAA EducationSeries, ISBN 1-56347-350-X.
2 Mason, W. H. et al., Low-cost Commercial Transport – Undergraduate Team Aircraft DesignCompetition, 1995, Virginia Tech. AIAA 95–3917.
3 Howe, D., Aircraft Conceptual Design Synthesis, 2000, Prof. Eng. Pub. Ltd, ISBN 1-86058-301-6.
4 Fielding, J. P., Introduction to Aircraft Design, 2000, Cambridge University Press, ISBN0-502-65722-9.
5 Hoerner, S. F., Fluid Dynamic Drag, published by the author, Bricktown, NJ, 1965.
“chap05” — 2003/3/10 — page 101 — #1
5
Project study: military
training system
Yakovlev YAK–130 Aero Vodochody L–59
British Aerospace HAWK–100 Mikoyan MiG–AT
“chap05” — 2003/3/10 — page 102 — #2
102 Aircraft Design Projects
5.1 Introduction
A project similar to the one described below was the subject of a EuroAVIA designworkshop sponsored by British Aerospace. Undergraduate students from ten Europeancountries worked for three weeks in separate teams to produce specifications for newtraining systems. The study below represents a combination of the results from thisworkshop and some subsequent design work done on aeronautical courses in twoEnglish universities. Acknowledgement is given to all the students who worked onthese projects for their effort and enthusiasm which contributed to the study described.
In the following analysis general references are made to aircraft design textbooks.1–5
To avoid confusions in the text, a list of current popular textbooks, useful for thisproject, is included in the reference section at the end of this chapter. A fuller list ofinformation sources can be found in Appendix B towards the end of this book.
5.2 Project brief
All countries with a national airforce need an associated programme for their pilotselection and training; therefore the commercial market for military training aircraftand systems is large. Designing training aircraft is relatively straightforward as thetechnologies to be incorporated into the design are generally well established. Manycountries have produced indigenous aircraft for training as a means of starting theirown aircraft design and manufacturing industry. This has generated many differenttypes of training aircraft in the world. For many different reasons only a few of thesedesigns have been commercially successful in the international market. The BritishAerospace Hawk (Figure 5.1) family of aircraft has become one of the best selling typesin the world with over 700 aircraft sold. It is a tribute to the original designers that thisaircraft, which was conceived over 25 years ago, is still in demand. The maturity of theHawk design is not untypical of most of the other successful trainers. Only recentlyhave new aircraft been produced (mainly in East European countries) but these are stillunproven designs and not yet competitive with the older established products.
Since the early 1970s when the Hawk and other European trainers were designed,front-line combat aircraft operation has changed significantly. The introduction ofhigher speed, more agile manoeuvring, stealth, together with significant developments
Fig. 5.1 Hawk aircraft
“chap05” — 2003/3/10 — page 103 — #3
Project study: military training system 103
in aircraft and weapon systems generated a requirement for a new training system. Asairframe and system development is expensive it is essential that an overall systemsapproach is adopted to this project.
The project brief for a new training system covers pilot training and selection from theab-initio phase (assuming cadets have had 50 hours’ flight training on a light propelleraircraft) to the start of the operational (lead-in) training on twin-seat variants of combataircraft. This period covers the existing basic and advanced training phases covered byHawk type aircraft. To represent modern fighter capabilities the new training systemshould also include higher flight performance and weapon system training which is notfeasible on current (older) training aircraft.
The concepts to be considered are those associated with an integrated training sys-tem. This must account for the various levels of capability from the aircraft, synthetictraining systems (including simulators) and other ground-based facilities. It will benecessary to define the nature of the training experiences assigned to each componentof the overall training system.
The minimum design requirements for the aircraft are set out in the aircraft require-ments section below but consideration should be given to the development of thetraining programme to include flight profiles with transonic/supersonic performance.Also, as all commercially successful training aircraft have been developed into combatderivatives, this aspect must be examined. To reduce the overall cost of the projectto individual nations discussion must be given to the possibility of multinationalco-operative programmes. All the issues above will be influential in the choice of designrequirements for the aircraft.
5.2.1 Aircraft requirements
Performance
General Atmosphere max. ISA + 20◦C to 11 km (36 065 5 ft)min. ISA − 20◦C to 1.5 km (4920 ft)
Flight missions – see separate tables
Max. operating speed, Vmo = 450 kt @ SL (clean)Vmo = 180 kt @ SL (u/c and flaps down)
Turning Max. sustained g @ SL = 4.0Max. sustained g @ FL250 = 2.0Max. sustained turn rate @ SL = 14◦/sMax. instantaneous turn rate @ SL = 18◦/s
Field Approach speed = 100 kt (SL/ISA)TO and landing ground runs = 610 m (2000 ft)Cross-wind capability = 25 kt (30 kt desirable)Canopy open to 40 ktNose wheel steering
Miscellaneous Service ceiling > 12.2 km (40 000 ft)Climb – 7 min SL to FL250
(note: one flight level, FL = 100 ft)Descent – 5 min FL250 to FL20 (15◦ max. nose down)Ferry range = 1000 nm (2000 nm (with ext. tanks))Inverted flight = 60 s
“chap05” — 2003/3/10 — page 104 — #4
104 Aircraft Design Projects
Structural
• Flight envelope n1 = +7, n3 = −3• Max. design speed M0.8• VD > 500 kt CAS• Utilisation = 500 h/year• Fatigue life = 30 yr
Operational
• Hard points = 2 @ 500 lb (227 kg) plus 2 @ 1000 lb (453 kg), all wet• Consideration for fully armed derivatives• Consideration for gun pod installation• Provision for air-to-air refuelling
Cockpit
• Aircrew size – max. male 95 per cent, min. female 50 per cent• Ejection – zero/zero• All weather plus night operations• Cockpit temperature, 15–25◦C• Oxygen system
Systems
• Avionics to match current/near future standards• Consideration given to fly-by-wire FCS• Consideration given to digital engine control• Glass cockpit• Compatibility to third and fourth generation fast jet systems where feasible
5.2.2 Mission profiles
Mission profiles used in the design of the aircraft are to be defined by the design teambut they must not have less capability than described below:
1. Basic This is to represent early stages of the flight training. Two sorties are to beflown without intermediate refuelling or other servicing.
Phase Description Height Time (min)1 Start, taxi SL 42 Take-off FL20 13 Max. climb FL250 74 Cruise to training area FL250 65 High-speed decent FL20 56 General handling FL20 10
(Buffet control, etc.)7 Max. climb FL250 68 Manoeuvres FL250 4
(Turns, spin, etc.)9 Cruise to base FL250 5
“chap05” — 2003/3/10 — page 105 — #5
Project study: military training system 105
Phase Description Height Time (min)10 Descent FL20 511 Recover to base∗ FL20 3
– 100 nm fuel + 5% reserve or– 5 circuits + 10% reserves
12 Landing, taxi, shutdown SL 4Mission elapsed time 60
(∗reserve fuel is only applicable to the second sortie)
2. Advanced This mission is typical of fighter handling at the advanced training stage.
Phase Description Height Time (min)1 Start, taxi SL 42 Take-off FL20 13 Max. climb FL250 74 Cruise to training area FL250 65 Weapon training FL250 106 Aerobatics and high g FL50 107 Low-level flying 250 ft 108 Climb to cruise FL250 79 Cruise to base FL250 6
10 High-speed descent FL20 511 Recover to base∗ FL20 6
– 100 nm fuel + 5% reserve or– 5 circuits + 10% reserves
12 Landing, taxi, shutdown SL 4Mission elapsed time 76
(∗reserve fuel is only applicable to the second sortie)
Note: the times quoted in the above profiles are approximate and do not define aircraftperformance requirements. (FL = flight level, 1FL = 100 ft.)
3. Ferry This mission is required to position aircraft at alternative bases. The ferryranges are specified in section 5.2.1. The ferry cruise segment may be flown at besteconomic speed and height. Reserves at the end of the ferry mission should beequivalent to that for the basic mission profile.
5.3 Problem definition
The main difficulty with this project lies is the broad spectrum of training activitiesthat are expected to be addressed by the system. To cover all flight training frompost- ab-initio to pre-lead-in will include the basic, intermediate and advanced trainingphases (Figure 5.2). In most air forces this involves the use of at least two differenttypes of aircraft (e.g. a basic trainer like the Tucano and an advanced trainer likethe Hawk). There will be about 90 hours of training in the selection and elementaryphases. To reduce flight costs most of this will be done on modified light aircraft witha single piston/propeller engine and semi-aerobatic capability (e.g. Bulldog, Firefly).Such aircraft have a limited top speed of about 130 kt. The next phase (basic training)lasts for about 120 hours, using faster turboprop or light turbojet trainers (e.g. Tucano,L39). This includes visual flying experience (climbs, descents, turns, stall and spin)together with some aerobatics navigation training, instrument flying and formation
“chap05” — 2003/3/10 — page 106 — #6
106 Aircraft Design Projects
Selection Elementaryor
basic training
Piston-poweredtrainer aircraft
Piston/tproptrainer aircraft
Tprop/jettrainer aircraft
Jet/fast jettrainer aircraft
Conversionor lead-in
experience
Increasingcomplexity leadingto operationalposting
Advancedtraining
Basic ofintermediate
training
Fig. 5.2 Airforce flight training phases
flying. The advanced training phase is about 100 hours’ duration and takes the pilotup to the point of transfer to an operational conversion unit (OCU). This phase willinvolve using an advanced turbojet trainer (e.g. Hawk) to provide experience at higherspeeds (530 kt) and higher ‘g’ manoeuvres. The programme will include air warfare,manoeuvrability, ground attack, weapon training and flight control integration. Theoperational conversion unit will use two-seat derivatives of fast jets and provide theexperience for lead-in to operational type flying.
To devise a training system for both basic and advanced phases based on a singleaircraft type will present commercial opportunities to the manufacturer together withoverall cost and operational advantages to the airforce. If innovation can be harnessedto produce a system to meet all the through-training requirements it would offer sub-stantial advantages over all existing training aircraft and current projects which offerless capability. This is obviously a difficult task but the key to the successful solutionto this problem lies in the careful exploitation of new technologies that have been usedin other aeronautical applications.
Designing a new training system that introduces, develops and relies on innova-tion carries a commercial risk associated with the unpredictability of the technology.Although, as engineers we may have complete faith in new concepts, perhaps the prin-cipal drawback in using a novel, high-tech system lies in the conservative nature ofour proposed customers (i.e. training organisations). Any new system must possess theability to gradually evolve new features even if this means a temporary partial degradingof the overall concept in the early stages.
With the above considerations in mind we (the designers) are required to producea technically advanced system to meet the defined training requirements yet exhibitsufficient capability to avoid initial scepticism from established customers. The systemmust show technical and economic advantages over existing equipment and possess thepossibility to develop alternative combat aircraft variants based on the trainer airframe,engine and systems.
5.4 Information retrieval
Researching trade journals (e.g. the annual military aircraft reviews in aviation mag-azines, like Flight International and Aviation Week) provides data on existing andrecently proposed training aircraft. Clearly the market is saturated with training aircraftof various types. The list below shows aircraft that are available to potential customers.
“chap05” — 2003/3/10 — page 107 — #7
Project study: military training system 107
Aeromacchi (MB339/S211A/S260) ItalyAeromacchi/Alenia/Embraer/Aerospatial (AMXT) InternationalAero Vodachody (L39/L59/L139/L159B) CzechoslovakiaAIDC (AT-TC-3A/B) TaiwanAvionne (JARA/G-4M) RomaniaBAE Hawk UKBoeing (T2/T28/T43) USABoeing/McD (T45 Goshawk) USABombardier/Shorts (Tucano) UKCASA (C101DD) SpainCessna (T37) USADaewoo (KTX-1) South KoreaDassault-Breguet/Dornier, Alpha Jet InternationalDenel (MB326M) South AfricaFuji (T3/T5) JapanHAI (Kiran 1A/MK2) IndiaIsrael Aircraft (TC2/TC7) IsraelKawasaki (T4) JapanLockheed Aircraft (IA63) ArgentinaMAPO (MIG-AT) RussiaMitsubishi (T2) JapanNAMC (K8) ChinaNorthrop-Grumman (T38) USAPAC PakistanPolskie (WSK PZL M-93V, I-22) PolandRaytheon-Beechcroft (T1/JPATS) USARhein-Flugg. (Fantrainer) GermanySAAB (SK 60W) SwedenSamsung/Lockheed (KTX-2) South KoreaSocata (TB30/TB31) FranceUTVA (Soko G4) SerbiaYakovlev (Yak 130) Russia(Yak/Aeromacci) (Y130) International
The list above is a ‘mixed-bag’ of aircraft including propeller types, derivatives ofexisting non-training aircraft, and some purely national projects. It is necessary toreview the collection to select aircraft that we feel are more appropriate to this project.The following aircraft are regarded as significant:
1. B.Ae. Hawk (Mk60/100): this is one of the most successful training aircraft in theworld with more than 700 produced and sold internationally.
2. L139/159: are ‘westernised’ versions of the very successful earlier Czech trainingaircraft (L39/59) which were used by airforces throughout the old Eastern Bloc.When fully developed it may present a serious competitor in future trainer markets.
3. MB339: is a derivative of the very successful Italian trainer (MB326). It has beenextensively modernised with upgraded avionics and a modern cockpit.
4. MiG-AT: compared with the above aircraft this is a completely new design by thehighly competent Russian manufacturer. It is in competition with other aircraftfor the expected 1000+ order for the Russian airforce and their allies. It presents aserious competitor to this project.
“chap05” — 2003/3/10 — page 108 — #8
108 Aircraft Design Projects
5. Yak/AEM 130: this is a new subsonic trainer from a Russian/Italian consortium.It will compete with the MiG-AT for the Russian airforce order and could be aconsiderable challenge to the Hawk in future years.
6. KTX-2: is a new supersonic (M1.4) trainer from a South Korean manufacturer(in association with Lockheed Martin). It is expected to be sold in direct competitionwith all new trainer developments and with other light combat aircraft.
7. AMX-T: this is a trainer development of the original AMX attack aircraft. It isproduced by an international consortium and will be a strong contender in futureadvanced trainer aircraft markets.
5.4.1 Technical analysis
Details of the aircraft in the list above have been used in the graphs described belowto identify a suitable starting point for the design. Decisions on selected values to beused in the project are influenced by this data. To reduce format confusion the graphsare plotted in SI units only.
Empty mass data (conversion: 1 kg = 2.205 lb)
Figure 5.3 shows the empty mass plotted against maximum take-off mass for jet trainers.The graph also shows the constant ‘empty mass ratio’ radials. These radials can beseen to bracket 0.75 to 0.45. Our selected value of 0.6 lies between the higher values forthe Russian aircraft and the Italian MB338 but above those for the L159, Hawk andAlpha Jet.
Wing loading (conversion: 1 kg/sq. m = 0.205 lb/sq. ft)
Figure 5.4 is a graph of the maximum take-off mass versus wing reference area forexisting aircraft. The wing loading radials bracket 500 to 200 kg/m2. Our selected valueis 350 kg/m2. Most of the specimen aircraft have higher wing loading but our specifiedlow approach speed requirement will dictate a lower wing loading.
MTOM (kg)
OE
M (k
g)
45%
60%
75%
1000
2000
3000
4000
5000
6000
7000
8000
02000 4000 6000 8000 12 00010 0000
Fig. 5.3 Survey of empty mass ratio
“chap05” — 2003/3/10 — page 109 — #9
Project study: military training system 109
0 5 10 15 20 25 30
2000
4000
6000
8000
10 000
12 000
Ref. area (sq. m)
500 kg/sq. m
MTO
M (k
g)350 kg/sq. m
200 kg/sq. m
Fig. 5.4 Survey of wing loading (kg/m2)
MTOM (kg)0 12 00010 0008000600040002000
Asp
ect
ratio
1
2
3
4
5
6
7
Fig. 5.5 Survey of wing aspect ratio
Aspect ratio
Figure 5.5 plots the wing aspect ratios for the trainer aircraft. Most seem to lie in theregion of 5 to 6. A value of 5 will be used as an initial guide to the wing planformgeometry. In subsequent phases of the design process, it will be necessary to conductdetailed ‘trade-off studies’ to establish the technical ‘best’ choice of wing aspect ratio.At this stage in the development of the aircraft it is impossible to do such studies assufficient details of the aircraft are unknown.
Thrust loading (conversion: 1 N = 0.225 lb)
Figure 5.6 shows the installed thrust versus maximum aircraft take-off mass. The radi-als show that modern aircraft lie along the 40 per cent SLS thrust line. As might beexpected the manoeuvre and performance of these aircraft are similar. The new super-sonic aircraft are above this line and older aircraft substantially below. This reflects the
“chap05” — 2003/3/10 — page 110 — #10
110 Aircraft Design Projects
0 2000 4000 6000 8000 10 000 12 000
MTOM (kg)
20%
30%
40%
2000
4000
6000
8000
10 000
12 000
Tota
l thr
ust
(lb)
Fig. 5.6 Survey of thrust/weight ratio (lb/kg)
requirement for improved performance for newer aircraft. We will select a 40 per centbased on the SSL thrust rating.
5.4.2 Aircraft configurations
Looking in detail at the configuration of aircraft in the candidate list confirms theimpression that most of the existing trainers are conventional in layout. They all havetwin, tandem cockpits with ejector seats and large bubble canopies. Apart from thelatest Russian designs they are single-engined with fuselage side intakes. The sloweraircraft have thick (12 per cent) relatively straight wings. Some of the later designs havethinner swept wings to match the faster (supersonic) top speeds. The wing/fuselageposition is mostly low set but with some at mid-fuselage. The Alpha Jet has a shoulderwing position. Tail position for all aircraft except the MiG is conventional with thetailplane set on the aft fuselage with the fin slightly ahead to give protection for post-stall control. The MiG originally had a ‘T’ tail but this was later changed to a mid-finlocation.
5.4.3 Engine data
Engines suitable for trainer aircraft lie in the 9 to 29 kN (2000 to 6500 lb) thrust range.The engines shown in Table 5.1 are available.
Figure 5.7 shows the engine weight (mass) versus SSL thrust data.
5.5 Design concepts
To provide a stimulus for the design of the aircraft it has been decided that a radical(novel) solution to the problem should be investigated. This consists of specifying atotal training system to cover all the required phases. It comprises an advanced simula-tor, a single-seat aircraft (see later comment), ground-based instructor console(s) and amodern communication and data linking facility (Figure 5.8). Removing the instructor
“chap05” — 2003/3/10 — page 111 — #11
Project study: military training system 111
Table 5.1
Thrust SLS SFC Eng. massEngine (Manufacturer) Used on (kN/lb) (@ SLS) (–/hr) (kg/lb)
FJ44-2A (Williams/RR) – 10.20/2300 0.45 203/447JT15D (P&W Can) Citation 13.54/3045 0.55 284/627Larzac 04-C20 (TM/Snec.) MiG-AT 14.12/3175 0.74 302/666J85-21 (Gen. Elec.) F5/T38 15.57/3500 1.00 310/684Viper 680 (Rolls Royce) MB339 19.30/4339 0.98 –PW 545A (P&W Canada) Citation 19.79/4450 0.44 347/765DV-25 (PS/Russian) Yak 130 21.58/4852 0.60 450/992TFE 731-60 (Allied-Sig.) Citation 24.86/5590 0.42 421/929CFE 738 (GE/ASE) Falcon 2000 24.90/5600 0.37 601/1325PW 306A (P&W Canada) DO 328 25.35/5700 0.39 473/1043Adour 871 (Rolls Royce) Hawk/T45 26.81/6028 0.78 602/1328F124-100 (Allied-Signal) – 28.02/6300 0.81 499/1100AE 3007C (AEC) UAV 28.89/6495 0.33 717/1581
250 kg
500 kg
Eng
ine
wei
ght
(lb)
2010 30KN
1000
400
600
800
1000
1200
1400
1600
2000 3000 4000
Engine thrust (SSL) (tb)
5000 6000 7000
1
2
34 6
7
8
9
1012
11
13
(b) Trend line (?)
(Adour)
(a) Trend line:eng. wt (tb) =
50 + 0.175 thrust
Fig. 5.7 Survey of engine weight versus SSL thrust
from the training aircraft is regarded as feasible with the adoption of new technolo-gies that have been proven in other applications. Experience from flight test data linksand recording gives assurance that technically the systems are available and feasible onwhich to develop a remote instructor system. Without a second seat the aircraft will be
“chap05” — 2003/3/10 — page 112 — #12
112 Aircraft Design Projects
Ground-based systems Simulators and
post-flight reviewGround-based systems instructional consoles
Real-timedata transfer
Real-timecommunications
Real-timevideo link
Airborne systems(Training aircraft and equipment)
Grounddata transfer
Fig. 5.8 Proposed total training system diagram
simpler, lighter and cheaper and flying solo the pilot will be in a more realistic opera-tional environment. A further advantage lies in the development of the aircraft into acombat derivative. It will obviously be essential to carefully design the communicationslink and the instructor module to ensure reliable and safe operation. Modern electronicand video equipment should be capable of providing the necessary confidence. Thisdecision was later reviewed.
Development of a two-seat simplified version of the aircraft will be possible by sac-rificing some of the payload and performance capability. This may provide a meansof avoiding some of the apprehension centred on the use of the system in the basicand early parts of the intermediate training phases. The two-seat version represents arelatively straightforward development of the aircraft.
The single-seat aircraft strategy makes it possible to set the design point for the aircraftat the upper end of the advanced training spectrum. This will guide the definition ofthe critical performance, payload and systems specification. As previously mentionedsetting this specification will also provide a better baseline for the development of thecombat aircraft derivative.
5.6 Initial sizing
Using the assumed values for empty mass ratio (0.6), wing loading (350 kg/m2
(72 lb/sq. ft)), aspect ratio (5), thrust loading (0.40) and the specified useful load(pilot + operational equipment + 3000 lb weapon load, totalling an assumed 3308 lb(1500 kg)), it is possible to make estimates of the initial mass and sizes for the aircraft.
Using the equation from Chapter 2, section 2.5.1:
MTO = MUL
1 − (ME/MTO) − (MF/MTO)∗
∗Using a value of 0.15 for the fuel fraction (from Hawk data, this will need to be verifiedlater) and substituting the known and assumed values gives:
15001 − 0.6 − 0.15
= 6000 kg (13 230 lb)
“chap05” — 2003/3/10 — page 113 — #13
Project study: military training system 113
With this aircraft mass the assumed wing loading gives:
Wing reference area (S) = (6000/350) = 17.14 m2 (184 sq. ft)
Using an aspect ratio of 5 sets of wing span (b) = (5 × 17.14)−0.5 = 9.26 m (30.4 ft)
This sets the mean chord (cmean) = (9.26/5) = 1.85 m (5.9 ft)
Assuming a wing taper ratio of 0.25 sets the approximate values for the centre linechord of 3.0 m (9.8 ft) and tip chord of 0.75 m (2.5 ft)
In drawing the aircraft we will round off the measurements to give a span of 9.0 m(29.5 ft). This results in the slightly larger wing area of 16.88 m2 (181sq. ft)
The selected thrust loading of 0.4 in association with the estimated aircraft mass givesa required static sea level thrust of (0.4 × 6000 × 9.81) = 23.54 kN (5300 lb).
The choice of engines to provide this thrust rests between:
• the old and slightly overpowered Adour engine used in the Hawk,• a more modern higher bypass engine from Allied Signal/P&W Canada/GE (used on
business jets),• a slightly underpowered Russian engine as used on the Yak 130, or two TM/Snecma
engines as specified for the MiG-AT.
This presents a somewhat difficult choice as:
• the Hawk engine is thirsty,• the higher bypass engines are larger diameter and lose thrust at altitude (a major
disadvantage for the proposed combat variant),• the Russian manufactured engines may not appeal to established Western customers,• installing two engines will complicate the systems and cockpit (but would be
representative of modern fighter configurations).
After careful consideration it has been decided to use the Adour engine as it is wellrespected by established customers, is reliable and will add confidence to our noveltraining system. The Adour 861 provides 5700 lb of thrust so it would be possibleto derate the engine for the main specification (this would extend engine life). Thisstrategy would make extra thrust available for the faster trainer and combat variants.With maximum thrust available, the thrust loading would be increased to 45 per cent.
The initial sizing above has provided sufficient data to consider in more detail theinitial aircraft layout.
5.6.1 Initial baseline layout
With our understanding of the configurational options used on existing trainerstogether with representative sizes of components from our initial sizing, it is possibleto consider the detailed layout of the aircraft (Figure 5.9).
The following decisions on the aircraft geometry have been taken:
• The aircraft will be of conventional layout with rear fuselage-mounted tail surfaces.• A single RR Adour 861 engine will be mounted in the aft fuselage, length 2.0 m
(80 in), width 0.76 m (30 in), height 1.04 m (42 in), intake diameter 0.7 m (28 in),nozzle diameter 0.6 m (24 in).
• Single seat cockpit with ejector seat and enclosed canopy/windscreen providing therequired vision capability.
“chap05” — 2003/3/10 — page 114 — #14
114 Aircraft Design Projects
mac
mac
0 1 2
metres
Fig. 5.9 Initial aircraft layout
• The fuselage aft of the cockpit will be suitably configured to permit an easymodification to accommodate a twin-tandem layout for the basic trainer variant.
• The wing will be mid-to-high fuselage (shoulder) mounted to provide generousground clearance for underwing stores. It will also be manufactured as a single piece(tip to tip) structure which will be mounted above the upper fuselage longerons. Thiswill avoid complicated wing-to-fuselage structural joints.
• The wing will be trapezoidal in planform with at least 30◦ leading edge sweepbackand a thin (10 per cent) section (to provide for future higher-speed variants).
• Tail area ratios will match existing aircraft data (from a review of the aircraft datafile these values seem appropriate; SH/S = 0.255 giving SH = 4.3 m2 (46 sq. ft) andSV/S = 0.185 giving SV = 3.1 m2 (33 sq. ft)).
(Note: an initial layout drawing of the aircraft showed that the short fuselage lengthmakes a conventional fuselage-mounted tailplane suffer from a shortage of tail arm.Therefore a ‘T-tail’ arrangement has been adopted. This configuration improves boththe horizontal and vertical tail effectiveness which allows a reduction in tailplanearea to 22 per cent S (= 3.7 m2 (40 sq. ft)). A reduction in fin areas could also beanticipated but this was not adopted, as the proposed two-seat variant will requiremore fin to balance the increased fuselage nose length. The T-tail arrangement willenable the provision of a communication/video pad to be installed at the top of thefin (at the tailplane junction). This will be useful to accommodate some of the extraequipment necessary for the remote instructor facility.)
“chap05” — 2003/3/10 — page 115 — #15
Project study: military training system 115
• To assist in re-energising the airflow over the rear fuselage and fin in high-alphamanoeuvres, small wing leading edge extensions will be added to the planform.
• Underwing, fuselage-side intakes will be positioned below the leading edge exten-sions to ensure clean airflow into the engines in high-alpha manoeuvres.
• Conventional tricycle landing gear with wheel sizes representative of existing aircraft(from the existing aircraft data file, the main and nose-wheel diameters are 0.6/0.45 m(24/18 in) respectively).
5.7 Initial estimates
With a scale drawing of the baseline aircraft configuration and an understandingof the engines and systems to be used it is now possible to conduct a series ofdetailed calculations to estimate the aircraft mass, aerodynamic characteristics andperformance.
5.7.1 Mass estimates
The mass of each component of the aircraft can be calculated using methods describedin aircraft design textbooks. These are generally based on geometrical and aircraft loaddata and are often derived from analysis of existing aircraft configurations. Suitableadjustments need to be made in those areas where the proposed design is significantlydifferent from past designs. In our case there are two such considerations:
• much more composite material will be used than in the predominately aluminiumalloy aircraft built previously and,
• for this training system more sophisticated and extensive flight control and commu-nication systems will be installed (allowance will need to be made for the reducedmass and volume of new electronic/computer systems).
A design take-off mass of 6000 kg will be assumed for determination of the aircraftstructural components. Although this mass is likely to be higher than that estimated forthe maximum take-off mass of the aircraft it will provide an insurance against futuremass increase. If it is necessary to determine the minimum take-off mass for the aircraft,the estimation would need to be done iteratively using the calculated take-off mass asthe design mass input for components mass estimations.
Detailed calculations for mass estimations have not been shown below but the inputdata on which the calculation was based is given (for reference). To simplify presenta-tion, the data below is shown in SI units only. The completed mass statement is shownin dual units.
Wing: Area (S) = 16.88 m2 Aspect ratio = 5 Wing thickness (average) = 10%Sweepback (c/4) = 25◦ Taper ratio = 0.25 Control surface areas = 15% (S)
Conventional mass estimation = 462 kg, assuming 15 per cent reduction forcomposites = 392 kg
Fuselage: Length = 9.5 m Depth = 2.0 m Width = 0.75 m
Conventional mass estimation = 576 kg, assuming 10 per cent reduction forcomposites = 518 kg
Horizontal tail: Area = 22% (S) Tail span = 4.0 m Fuselage width = 0.75 m
“chap05” — 2003/3/10 — page 116 — #16
116 Aircraft Design Projects
Conventional mass estimation = 76 kg, assume 15 per cent reduction for composites= 65 kg
Fin: Area = 18% (S), T-tail structure
Conventional mass estimation = 39 kg, assume 15 per cent reduction for composites= 34 kg
Undercarriage: MLAND = 90% MTO nULT = 3 × 1.5 = 4.5Main: Length = 0.75 m → Mass = 185 kgNose: Length = 0.60 m → Mass = 52 kg Total = 237 kgNo reduction on this mass for new materials
Engine: Dry = 577 kg (from manufacturers)
Engine and fuel systems: Assume 50 per cent engine dry mass = 288 kg
Aircraft equipment: To account for new aircraft system requirements assume20 per cent of aircraft design mass → 1200 kg
Fuel: This will be checked by the performance estimates, for now we will still assume15 per cent of aircraft design mass → 900 kg
Crew: Assume one pilot with operational equipment → 300 lb = 136 kg
Weapon load: Specified at 3000 lb → 1360 kg
Hence the initial mass statement for the aircraft can be compiled:
Component kg/lb %MTOWing 392/664 6.9Fuselage 518/1142 9.1Horizontal tail 65/143)
}
→Fin 34/75) 1.7
Undercarriage 237/523 4.2Total structure 1246/2747 21.8
Engine (dry) 577/1272Engine and fuel systems 288/635
Total propulsion 865/1907 15.2Equipment (total) 1200/2646 21.0
Total aircraft empty 3311/7300 58.0Fuel 900/1984 15.8Crew 136/300Weapon load 1360/3000
Total Useful Load 2396/5283 42.0
MAXIMUM TAKE-OFF MASS∗ 5707/12584 100.0Note: aircraft design take-off mass 6000/13230
(∗This compares with 9000 to 7500 kg (20 000 to 16 000 lb) for older trainers and iscompetitive with the newer designs.)
At this stage all the input values used to predict component masses are somewhattentative, therefore the use of more extensive methods are inappropriate.
“chap05” — 2003/3/10 — page 117 — #17
Project study: military training system 117
The aircraft aerodynamic and performance calculations that follow will use the abovemass statement to determine aircraft masses for various stages in the flight profiles:
• At take-off, the aircraft can be loaded up to a variety of different conditions depend-ing on the operation to be performed. The most critical take-off condition will be atmaximum take-off mass. This will be used in the estimation of take-off distance.
• At landing, the aircraft could again be at different mass. The most critical flight casewould be to land immediately after taking off with full fuel and payload. This wouldbe an emergency condition and therefore it is not necessary to use this as a design case(in an emergency the pilot may jettison some of the weapons and/or part of the fuelload to reduce the landing mass). The minimum landing mass would relate to zeroweapon load and only a small percentage of fuel remaining (e.g. 10 per cent). Thismay also be a rare condition and therefore not applicable to design calculations.The landing calculations are often performed at a landing weight of 90 per centof the take-off weight. For the maximum take-off mass condition this relates to5707×0.9 = 5136 kg (11 325 lb). Another take-off condition that may be consideredis a take-off with zero weapon load (i.e. a basic training mission). Using 90 per centof the take-off weight in this condition gives (5707 − 1360) 0.9 = 3912 kg (8626 lb).These two cases will be used in the landing distance and speed estimations.
• For the manoeuvring calculations (turns and climb) a mean flying weight will beassumed. This relates to half weapon and half fuel load (= 4577 kg/10 092 lb).
• For the ferry case the aircraft take-off condition will be without weapon load butmay have carry external wing tanks.
5.7.2 Aerodynamic estimates
Using the geometrical data from the initial baseline layout and masses from the abovesection, it is possible to make initial estimates for the aircraft lift and drag in variousflight conditions.
Aircraft drag coefficients
Using standard equations from aeronautical textbooks and the data below, it is possibleto estimate the drag coefficients for the aircraft in different flight conditions.
In SI units:
Take-off mass (MTO) 5707 kg Wing aspect ratio (A) 5Wing ref. area (S) 16.88 m2 Wing LE sweep (�) 30◦Wing loading 3212 kN/m2 (337 kg/m2)
The following equations and parameters are appropriate:
• Induced drag factor K = (1/πeA)• Planform factor e = 4.61((1 − 0.045A0.68) (cos �)0.15) − 3.1• Assumed skin friction coefficient Cf = 0.0038• Profile drag coefficient CDO = Cf (Sw/S)
(assumed aircraft wetted area – estimated from the initial layout drawing (total)Sw = 78 m2)
• Dynamic pressure q = 1/2ρ.V 2
(ρ and V are the air density and aircraft speed at the flight condition underinvestigation)
• Aircraft lift = CLqS
“chap05” — 2003/3/10 — page 118 — #18
118 Aircraft Design Projects
0 50 100 150 200 250 300 350 400Aircraft speed (m/s)
Drag @ SL
Thrust @ FL360
Drag FL200
Drag FL360
10 000
20 000
30 000
40 000
50 000
60 000
Dra
g an
d th
rust
(N)
Thrust @ SL
M1.0 @ FL200
M1.0 @ SL
M1.0 @ FL360
Thrust @ FL200
Fig. 5.10 Aircraft drag polar
• Aircraft drag = CDqS• Aircraft lift/drag ratio = L/D = CL/CD
The above data was input to a spreadsheet program to determine the lift and dragparameters for a range of aircraft speed (V ), operating altitude (h) and load factor (n)for the aircraft in the ‘clean’ (no weapons, u/c and flaps retracted) flight condition. Therange of values used in the calculations is shown below:
V (m/s) 50/100/150/200/250/300/350
h (ft) 0 (SL)/25 000 (FL250)/36 000 (FL360)
n (g) 1/3/6
The spreadsheet results for sea level at n = 1 are shown in Figure 5.10. The thrustdisplayed in the graph is for the Adour engine (no variation of thrust with aircraftforward speed is typical for low bypass ratio engines).
The drag analysis described above assumes subsonic flow conditions but the uppervalue for aircraft speed is seen to be in excess of M1.0 at each altitude. Obviouslyaircraft drag will increase rapidly as supersonic flow is developed over the aircraft.Some allowance will need to be taken for the wave drag at higher speeds. Also, sinceit is intended to investigate the potential for increasing the aircraft top speed into thetransonic range (e.g. M1.2) it is necessary to make suitable corrections.
Textbooks quote the Sears–Haack wave drag coefficient as:
CDwave = (9π/2)(Amax/L)2(1/S)
where (Amax) is the maximum cross-sectional area of the aircraft(L) is the distance from the nose to the position where the area is maximum(S) is the reference wing area for the aircraft
From the aircraft drawing Amax = 2.0 m2 (21.5 sq. ft) and l = 4.5 m (14.75 ft) arereasonable estimates. This data gives CDW = 0.14. This value could be reduced toaccount for the swept wing but would need to be increased due to the poor area rulingof the design (due mainly to the effect of the canopy). At this early stage in the analysisof the aircraft it is assumed that these effects cancel leaving CDO = 0.14 at M1.05.
“chap05” — 2003/3/10 — page 119 — #19
Project study: military training system 119
Note: the increase due to wave drag at M1.05 approximately doubles the subsonicdrag previously calculated. The wave drag will be progressively felt from the dragdivergent Mach number, which for this wing sweep will be about M0.75. At this stagein the aircraft evaluation it will be sufficient to sketch a blend line from the subsonicdrag polar to account for the values above. A suggestion from one textbook assumeshalf the wave drag increase to have been achieved at M1.0, therefore the blend will befrom M0.75 and through the M1.0 and M1.05 points.
Aircraft lift coefficients
From an analysis of stall and approach speeds from existing aircraft it is acceptable toassume maximum lift coefficients for our aircraft of:
Landing Clmax = 2.10
Take-off Clmax = 1.70
5.7.3 Performance estimates
The calculation of the aerodynamic coefficients described above are relatively crude anddo not take into account many of the detailed factors that are known to be significant.However, they do provide ‘ballpark’ values which can be used to initially assess aircraftperformance. From these calculations it will be possible to identify the changes that arenecessary to the design to meet the specified design criteria. Later analysis can accountfor the more detailed aspects of the predictions.
Maximum speed
Applying the wave drag increase to the subsonic drag polars at the three altitudes (SL,FL250, FL360), shows (Figure 5.10) that a top speed of 320 m/s (621 kt) is possible atSL-ISA with the full rating of the Adour engine. If the engine was derated to 85 percent of maximum thrust the speed would reduce to 305 m/s (592 kt). Both of these arewell in excess of the 450 kt specified in the design brief. However, these calculationshave been done with the aircraft in the ‘clean’ condition. If an allowance for weapondrag (�CD = 0.01) is made, the speeds are reduced to 280 m/s (544 kt) and 260 m/s(505 kt) respectively. These are still faster than the specified speeds.
The conclusion to this part of the performance study is that the aircraft will easilymeet the specified top speed requirement.
Turn performance
Aircraft performance textbooks quote the general equations as:
• turn rate = χ = g(n2 − 1)0.5/Vwhere n = normal acceleration factor
V = aircraft speed• radius of turn R = V/χ• aircraft bank angle φ = cos−1(1/n)
These equations have been solved, using a spreadsheet, for values of:
V = (50/100/150/200/250/300/350 m/s) and
n = (1/2/3/4/5/6/7/8)
(A reminder: 1 m/s is approx. 2 kt (more precisely 1.94).)
“chap05” — 2003/3/10 — page 120 — #20
120 Aircraft Design Projects
Note that the turn equation above is independent of aircraft parameters. The resultingcurves are shown in Figure 5.11. This graph describes the overall manoeuvring designspace for the aircraft.
For our aircraft the boundaries to the manoeuvring space are:
• aircraft maximum speed (assumed to be M0.8),• stall speed (clean at mean flying weight with CLmax = 1.10),• structural load limit of n = 7.
These provide the limiting lines shown on the sea-level manoeuvring diagram(Figure 5.12). The corner speed gives the maximum instantaneous turn rate. A valueof 24◦/s is predicted for our aircraft.
Turn
rat
e (r
/s)
n = 8n = 7
n = 6
n = 3
n = 2
n = 5
5
10
15
20
25
30
35
n = 4
Aircraft speed (m/s)
0 50 100 150 200 250 300 350 400
Fig. 5.11 Manoeuvring design space
Point C
Point APoint B
Aircraft speed (m/s)
Turn
rat
e (r
/s)
Stall limit
Max. speed (sea level)
Corner point
Max. structural limit
10
15
20
25
30
35
5
Max. instantaneousturn rate
Speed at max. turn
0 50 100 150 200 250 300 350 400
SEP @ SL
Stall limit
SEP @ FL250
Fig. 5.12 Aircraft turn performance
“chap05” — 2003/3/10 — page 121 — #21
Project study: military training system 121
Speed (m/s)
Spe
cific
exc
ess
pow
er
n = 1
n = 2
n = 3
n = 4
n = 5
n = 6
Point A (V = 140 m/s, n = 4.2g)
100
50
–50
–100
–150
–200
–250
–3000 50 100 150 200 250 300 350 400
Fig. 5.13 Specific excess power at SL
20
40
0 100 150 200 250 300 350 400
Spe
cific
exc
ess
pow
er
Speed (m/s)
n = 3
n = 2
n = 1
50
–20
–40
–60
–80
–100
–120
–140
–160
–180
Fig. 5.14 Specific excess power at FL250
Determination of the sustained turn requires an analysis of the available specificexcess power (SEP) for the aircraft. The zero SEP boundary provides the value forsustained turn rate. From textbooks the equation for specific excess power is:
SEP = V (T − D)/W
SEP will therefore vary according to the aircraft speed, engine thrust, aircraft dragand weight (i.e. M.n.g). Drag (and thrust) will vary with aircraft altitude and speed.The values determined for SEP for a range of aircraft speeds (50 to 350 m/s) and loadfactors (1 to 6), for both sea level and FL250, have been calculated and are plotted inFigures 5.13 and 5.14. Cross plotting the aircraft speeds at zero SEP from these graphsonto the manoeuvring diagram and joining with a smooth line provides the boundaryfor sustained turn rate.
“chap05” — 2003/3/10 — page 122 — #22
122 Aircraft Design Projects
• From Figure 5.12: maximum instantaneous turn rate = 24◦/s at a flight speed of165 m/s (the specified requirement is 18◦/s).
• From Figure 5.13 with the results transferred to Figure 5.12 (point A): sustainedturn rate = 17◦/s at sea level at a flight speed of 140 m/s (the specified requirement is14◦/s).
• From Figure 5.12 with the zero SEP lines transferred from Figures 5.13 and 5.14:sustained turn g of 6 at sea level (point B) and just about 3 at (point C) and FL250(the specified requirements are 4 and 2g).
The above calculations were done at aircraft maximum take-off mass with a cleanconfiguration. Similar calculations were done with weapon drag added (�CD = 0.01)at mean flying mass (4577 kg/10 092 lb) but these were shown to be less critical than theclean-heavy case quoted above.
Note: the manoeuvre graph shows that an extra 5◦/s instantaneous turn rate wouldbe achieved if the aircraft structural limit was raised from the specified 7g to 8g.
Field performance
For take-off and landing calculations it is necessary to add the extra drag due to theextended undercarriage and flaps. The flap deflection for take-off is assumed to be 20◦and for landing 40◦. The following drag contributions are regarded as appropriate:
�CD undercarriage = 0.0075
�CD 20◦ flap = 0.015
�CD 40◦ flap = 0.030
Lift co-efficients are as quoted earlier (i.e. take-off 1.7 and landing 2.1).
Take-off estimationIn the early part of the conceptual design process simplified take-off calculations canbe done using Nicolai’s (reference 4) simplified take-off parameter (this includes W/S,T/W , Clmax). As we require only the ground run to be estimated, the take-off distanceattributed to the climb segment can be ignored. The calculation for total take-offdistance, assuming the aircraft at maximum take-off mass (5707 kg/12 584 lb) and withmaximum sea level static thrust (5700 lb), gives:
Total distance to 50 ft = 2408 ft (734 m)
Removing the distance covered in the climb gives the take-off ground run = 1856 ft(566 m) (the specified maximum take-off ground run is 2000 ft (610 m)).
Using a derated (85 per cent) engine thrust increases the ground run to 2184 ft (666 m)which is more than the required distance. In this case, to get off from a 2000 ft/610 mrunway, the maximum aircraft mass would need to be limited to 5300 kg (11 685 lb).
Landing estimationsFor landing, we will assume a landing mass of 90 per cent max. take-off mass. Therefore,MLAND = 5135 kg/11 323 lb which gives an aircraft stall speed of 46.7 m/s (90 kts). Theapproach speed for military aircraft is set at 1.2Vstall giving 108 kt which means thatthe specified value of 100 kt is not achieved. In the design brief the appropriate aircraftlanding weight for the approach speed requirement is not specified. If we assume thelow approach speed is applicable to the basic training role we can assume the weapon
“chap05” — 2003/3/10 — page 123 — #23
Project study: military training system 123
load to be zero. Assuming the landing weight is to be set at 90 per cent of the take-offweight gives a landing mass for the basic training role of:
MLAND = 0.9 (5707 − 1360) = 3912 kg (8626 lb)
At this mass the stall speed is 40.8 m/s (79 kt) making the approach speed = 1.2×79 =95 kt. This achieves the requirement but begs the question of the relationship of theapproach speed requirement to the aircraft role.
Assuming a constant deceleration of 7 ft/s2 (as assumed in the simplified Nicolai4
estimation) and a touch-down speed of 1.15Vstall. The landing distance is determinedfrom:
Landing ground run = V 2TD/(2 × 7)
(VTD = 1.15 × 46.7 = 53.7 m/s (176 ft/s)
∴ Landing ground run = 2212 ft (675 m)
This is also in excess of the specified distance of 2000 ft but the calculation was doneat maximum landing mass (5136 kg/11 325 lb). Performing the same calculation atthe lighter landing mass assumed above (3912 kg/8626 lb), using the same aircraftassumptions gives:
Landing ground distance = 1690 ft (515 m)
This easily achieves the specified distance of 200 ft (610 m) but again begs the questionof the definition of landing weight.
The heavy and light landing mass assumptions used in the landing calculations pro-vide a range of maximum/minimum values for the aircraft. Further discussion with theproject customers would be necessary.
Mission analysis
The mission analysis allows us to estimate the fuel requirements. The project briefspecifies three different mission profiles. At this stage, it is not obvious which one ofthese will be most critical, therefore each will be analysed to determine the requiredfuel. The calculations use the weight fractions suggested in Raymer’s book1 for the lesssignificant segments of the missions.
Using the aerodynamic analysis described earlier it is possible to determine thelift/drag ratio variation with aircraft speed for the aircraft cruising at 25 000 ft. Fromthis data a representative value of 9.0 will be assumed in the calculations. This decisionwill be verified later.
The specific fuel consumption of the Adour engine is 0.78 –/hr at the SSL thrust rating.This value will increase with aircraft altitude, speed and engine setting (e.g. cruise).Determining the exact extent of this rise requires engine performance data. At thisstage in the design process such data is not available, a value of 0.95 has been assumedin the calculations below.
Basic training profile (two sorties)The training profile is shown diagrammatically in Figure 5.15.
Analysis of the change in aircraft mass in each segment is shown in Table 5.2.As this is the basic training mission it is done clean (no weapon drag) and at a
lower take-off mass (no weapon load). The first of the two sorties will require the
“chap05” — 2003/3/10 — page 124 — #24
124 Aircraft Design Projects
23 6 7
4 5 89
10
6A
1
FL20 Repeat
FL0 (SL)
FL250
FL20
Fig. 5.15 Mission flight profile (basic training)
Table 5.2
Segment Description Parameters (M(n+1)/Mn)∗
0 to 1 Take-off 0.9701 to 2 Climb 0.9852 to 3 Cruise (out) 6 min 0.9903 to 4 Descent 0.9954 to 5 General handling 10 min 0.9805 to 6 Climb 0.9856 to 6A Manoeuvres 4 min6A to 7 Descent 5 min 0.9957 to 8 Recover 0.9958 to 9 Land 0.9959 to 10 Cruise (return) 0.980∑
0.877
Fuel fraction for the above sortie Mfuel/MTO = (1 − 0.877) = 0.123.∗(M(n+1)/Mn) is the ratio of the aircraft mass at the end of the segment relativeto that at the start.
following fuel:
∴ MTO1 = 5707 − 1360 = 4347 kg (to avoid duplication remember, 1 kg = 2.205 lb)
∴ Mfuel1 = 0.123 × 4347 = 535 kg
As there is no intermediate refuelling the second sortie will be flown at a lower take-offmass than the first:
MTO2 = 4347 − 535 + 3812 kg
The same mission will be flown, therefore the same fuel fraction will apply:
∴ Mfuel2 = 0.123 × 3812 = 469 kg
Total fuel required = Mfuel1 + Mfuel2 = 535 + 469 = 1004 kg/2234 lb
This is slightly more than the fuel load of 900 kg assumed in the mass calculations.
Advanced training profile (single sortie)As this profile is flown with full weapon load the aircraft take-off mass will be 5707 kg.To account for the weapon drag the lift/drag ratio will be reduced to 7.5.
“chap05” — 2003/3/10 — page 125 — #25
Project study: military training system 125
2 3
6 7 4
5
8 9
10
2A
FL20
FL250
1250 ft
11
FL20
FL0 (SL)
Fig. 5.16 Mission flight profile (advanced training)
Table 5.3
Segment Description Parameters (M(n+1)/Mn)∗
0 to 1 Take-off 0.9701 to 2 Climb 0.9852 to 3 Cruise (out) 6 min
Weapon training 10 min 0.9673 to 4 Descent 0.9954 to 5 Aerobatics 10 min 0.9705 to 6 Descent6 to 7 Low level 10 min 0.9707 to 8 Climb 0.9858 to 9 Cruise (return) 6 min 0.9909 to 10 Descent 0.99510 to 11 Land 0.995∑
0.835
∗(M(n+1)/Mn) is the ratio of the aircraft mass at the end of the segment relativeto that at the start.
The profile is shown diagrammatically in Figure 5.16 and the segment analysis isshown in Table 5.3.
Fuel fraction for the total mission = (1 − 0.835) = 0.165
∴ Fuel required Mfuel = 0.165 × 5707 = 942 kg/2077 lb
This is also slightly higher than originally assumed but less than that estimated for thebasic training profile above.
FerryThe quoted ferry range is 1000 nm. This will be flown from a maximum take-off massof 5707 kg and at optimum speed. It is assumed from Figure 5.17 that Vcruise is 150 m/s(291 kt). At this speed the aircraft lift/drag ratio is 9 for best lift/drag ratio (clean) asshown on Figure 5.18. The engine sfc (c) at this condition will be assumed to be 1.05.
To estimate the fuel fraction we will use the transformed Breguet range equation:
(M1/M2) = exp[−(R.c)/(V .(L/D))]= 0.72
“chap05” — 2003/3/10 — page 126 — #26
126 Aircraft Design Projects
Dra
g (N
)
5000
10 000
15 000
20 000
25 000
30 000
35 000
100 150 200 250 300 350 4000 50Speed (m/s)
Tangent line through origin
Fig. 5.17 Ferry mission drag polar
Speed (m/s)
Lifl
/ dra
g ra
tio
100 150 200 250 300 350 4005000
2
4
6
8
10
12
Fig. 5.18 Aircraft cruise lift/drag ratio
Multiplying the take-off, climb and landing mass fractions as used in the previousmission analysis (to account for the fuel used at the start and end of the ferry mission)to this gives the overall fuel mass fraction:
Mstart/Mend = 0.97 × 0.72 × 0.995 = 0.696
∴ Fuel required = 5707(1 − 0.696) = 1733 kg (3821 lb)
This is higher than the fuel that can be carried internally (assumed to be 900 kg),therefore external tankage will be necessary. This would reduce the (L/D) ratio used
“chap05” — 2003/3/10 — page 127 — #27
Project study: military training system 127
above due to the increased drag from tank profile. A rough estimate of this effect showsthat about another 140 kg of fuel would be required.
Reversing the analysis above and converting the entire payload (weapon) to fuelshows that:
Maximum fuel mass = 1360 + 900 = 2260 kg (4983 lb)
Subtracting the take-off, climb and landing fuel fractions reduces this to 2181 kg.Assuming 5 per cent of the external fuel mass is required for the external fuel tank
structure (i.e. 1360 × 0.05 = 68 kg) means that only 2181 − 68 = 2113 kg is availablefor fuel.
From the Breguet range equation:
Ferry range = [V × (L/D)/c]loge(M0/M1)
= (291 × 10/0.95)loge(5560/3447) = 1464 nm
(Note: this range estimation ignores any requirement for reserve fuel at the end of theflight.)
From the above calculation it appears that the requirement for 2000 nm ferry rangeis too difficult to meet. Keeping this requirement could seriously compromise the basicdesign of the aircraft. Within the accuracy of the calculations at this stage, it would bereasonable to request a reduction to this requirement down to a value of 1500 nm.
Climb and ceiling estimations
The rate of climb equation is:
(R of C) = V (T − D)/W
Note: this is equivalent to the specific excess power expression.Using the aerodynamic analysis from the earlier work it is possible to plot the rate of
climb against aircraft speed for flight at SL, FL250 and FL360. Figure 5.19 shows therate of climb at the mean mass condition with an allowance for an increase in drag forweapons (�CD = 0.01). The locus of maximum climb rate allows the data to be cross-plotted (Figure 5.20). This shows the service ceiling (i.e. the height at which the rate ofclimb falls to 100 fpm (0.5 m/s)) to be about 46 000 ft (against a specified requirementof 40 000 ft).
Figure 5.20 shows that the average rate of climb from sea level to FL250 is 34 m/s. Thisis used to predict that the time to climb to FL250 (7625 m) is 190 seconds (=3.2 min).This easily meets the specified requirement of 7 min. The same calculations were doneto check the climb rate and time for maximum take-off mass with full weapon drag(�CD = 0.017). This showed an average rate of climb of 30 m/s, which leads to a timeto climb to 25 000 ft of 4.2 min. At a similar condition but with engine derated to 85 percent thrust gives a time of 5.8 min.
All the calculations above show that the climb and ceiling requirements are notcritical.
Summary of initial performance analysis
• The initial estimate of take-off mass of 5707 kg will need to be revised to 5814 kg toaccount for the increase in fuel required for the basic training flight profile.
“chap05” — 2003/3/10 — page 128 — #28
128 Aircraft Design Projects
Clim
b ra
te (m
/s)
sea level
25 000 ft
36 000 ft
30
40
50
60
–20
–10
10
20
Aircraft speed (m/s)
0 50 100 150 200 250 300
Fig. 5.19 Aircraft rate of climb
0 5000 10 000 15 000 20 000 25 000 30 000 35 000 40 000
Aircraft altitude (ft)
Max
. clim
b ra
te (m
/s)
20
30
40
50
60
10
Fig. 5.20 Aircraft climb and ceiling evaluation
• The maximum speed even at 85 per cent thrust is 505 kt. This easily exceeds thespecified requirement of 450 kt.
• All the turn performance criteria are easily met.• Take-off ground run at 1856 ft is below the specified 2000 ft but with a derated engine
of 85 per cent thrust, this increases to 2184 ft.• The approach speed requirement of 100 kts cannot be met except by a lighter aircraft
(no weapon load). In this case an approach speed of 95 kt is achieved.• Landing ground run at 2215 ft also exceeds the specification of 2000 ft. Only with
aircraft at lighter landing mass can the specification be met.• The ferry mission of 1000 nm cannot be met with internal fuel but can be achieved
if 833 kg of fuel is carried externally. The maximum range that could be flown isestimated at 1464 nm. This is substantially less than the 2000 nm specified. It is
“chap05” — 2003/3/10 — page 129 — #29
Project study: military training system 129
suggested that this requirement be reviewed as in its present form it would seriouslycompromise the overall aircraft design.
• Climb and ceiling requirements are easily achieved.
5.8 Constraint analysis
From the project brief there are six separate constraints to be considered in thisanalysis:
1. Take-off distance less than 2000 ft.2. Approach speed no greater than 100 kt.3. Landing distance less than 2000 ft.4. Combat turn, at least 4g at sea level.5. Combat turn, at least 2g at 25 000 ft.6. Climb rate to provide for 7 min climb to 25 000 ft.
5.8.1 Take-off distance
The equations to be used to determine the effect of the take-off criterion can be foundin most textbooks (e.g. reference 4) as shown below:
(T/W ) = (constant) (W/S)/(Stake-off .CLtake-off )
Obviously this represents a straight line on the (T/W ) versus (W/S) graph. For ouraircraft the lift coefficient in the take-off configuration (CLtake-off ) is assumed to be1.7. The value Stake-off represents the total take-off distance (i.e. ground roll plus climbdistance to 50 ft). Assuming a climb gradient from zero to 50 ft of 5◦ gives a grounddistance covered of 571 ft. Adding this to the specified ground roll of 2000 ft givesStake-off = 2571 ft (784 m).
The constant in the above equation is assessed from Nicholi’s book4 as 1.27 (in SIunits with wing loading in kg/m2), so
(T/W ) = 1.27/(784 × 1.7)(W/S) = 0.00095(W/S)
5.8.2 Approach speed
Assuming the approach speed VA = 1.2VSTALL then:
(W/S)landing = β(W/S) = 0.5 × ρ(VA/1.2)2 × CLlanding/g
VA is specified at 100 kts (52 m/s). β is the ratio of landing mass to take-off mass. Ata maximum landing weight β = 0.9. At minimum landing weight (i.e. empty aircraftplus pilot plus 10 per cent fuel = 3311 + 136 + 90 = 3537 kg) β = 0.62.
Assuming the lift coefficient in the landing configuration (CLlanding) = 2.1
(W/S) = (0.5 × 1.225 × 52 × 52 × 2.1)/(1.2 × 1.2 × 9.81) = 273.6 @ β = 0.9
= 397.1 @ β = 0.62
Note: these constraints are constant (vertical) lines on the (T/W ) versus (W/S) graphs.
“chap05” — 2003/3/10 — page 130 — #30
130 Aircraft Design Projects
5.8.3 Landing distance
The approximate equation to determine ground run in landing can be rewritten asshown below:
(W/S) = (Slanding run × CLlanding)/(constant × β)
The landing ground run Slanding run is specified as 2000 ft (610 m). The lift coefficient inthe landing configuration (CLlanding) is assumed to be 2.1 (as above). The expressionwill be evaluated for the two landing mass fractions used above (i.e. β = 0.9 and 0.62).
The (constant) in the expression above (in SI units with W/S in kg/m2) is 5.0.
(W/S) = (610 × 2.1)/(5.0 × β) = 284.7 @ β = 0.9 and 413.2 @ β = 0.62
Note: these are also constant vertical lines on the constraint diagram.
5.8.4 Fundamental flight analysis
The fundamental equation used in the flight cases can be found in most textbooks. Interms of sea level, take-off thrust loading the equation is:
(T/W )TO = (β/α)[(q/β){CDO/(W/S)TO + k1(nβ/q)2(W/S)TO}+ (1/V )(dh/dt) + (1/g)(dV/dt)]
where (T/W )TO is the take-off thrust loadingα1 = T/TSLSTSLS = sea level static thrust (all engines)β = W/WTOCDO and k1 are coefficients in the aircraft drag equation, see belowD = qS(CDO + k1C2
L)
(W/S)TO is the take-off wing loading (N/m2)n is the normal acceleration factor = L/Wg = gravitational accelerationV is the aircraft forward speedq is the dynamic pressure = 0.5ρV 2
(dh/dt) = rate of climb(dV/dt) = longitudinal acceleration
5.8.5 Combat turns at SL
In this flight condition the aircraft is in ‘sustained’ flight with no change in heightand no increase in speed therefore the last two terms in the fundamental equation areboth zero.
At sea level α = 1
Assume that the turn requirement is appropriate to the mean combat mass (i.e. air-craft empty + pilot + half fuel + half weapon load = 3311 + 136 + 450 + 680 =4577 kg/10 092 lb)
Hence β = 4577/5707 = 0.8.
“chap05” — 2003/3/10 — page 131 — #31
Project study: military training system 131
From previous analysis (in SI units) the best speed for turning at SL is about 150 m/s.
∴ q = 0.5 × 1.225 × 1502 = 13 781
From the drag analysis done earlier (at 4577 kg with an increase in drag coefficient torepresent the stores on the wing) at a speed of 150 m/s, CD = 0.03 + 0.017C2
L.As specified, the aircraft is subjected to a normal acceleration n = 4 in the turn.
T/W = 13 781{(0.03/(W/S) + 0.017 × [4/13 781]2 × (W/S)}
5.8.6 Combat turn at 25 000 ft
This is similar to the analysis above but with α = 0.557/1.225 = 0.455.
At 25 000 ft the best speed for excess power is 200 m/s (in SI units)
∴ q = 0.5 × 0.557 × 2002 = 11 140
With β and CD values the same but with load factor n = 2 gives:
T/W = (0.8/0.445)[(11 140/0.8){(0.03/(W/S)+0.017×[(2×0.8)/11 140]2×(W/S)}
5.8.7 Climb rate
This criterion assumes a non-accelerating climb, so the last term in the fundamentalequation is zero but the penultimate term assumes the value relating to the specifiedrate of climb.
We will use an average value of climb rate of 18.15 m/s (i.e. 25 000 ft in 7 min) andmake the calculation at the average altitude of 12 500 ft, at a best aircraft speed of150 m/s.
At 12 500 ft α = 0.841/1.225 = 0.686At 150 m/s q = 0.5 × 0.841 × 1502 = 9461
Using the standard values for β at mean combat mass, and the drag coefficients (CDOand K ) previously specified, we get:
T/W = (0.8/0.686)[(9461/0.8){(0.03/(W/S) + 0.017 × [(1 × 0.8)/9461]2 × (W/S)}+ 18.15(1/150)
5.8.8 Constraint diagram
The above equations have been evaluated for a range of wing loading values (150 to550 kg/m2). The resulting curves are shown in Figure 5.21.
The constraint diagram shows that the landing constraints (approach speed andground run) present severe limits on wing loading.
To identify the validity of the constraints relative to other aircraft, values appropriateto specimen (competitor) aircraft that were identified earlier in the study have beenplotted on the same constraint diagram Figure 5.21. Some interesting conclusions canbe drawn from this diagram:
• The S212, T45, MiG, L159 and, to a lesser extent, the Hawk aircraft appear to fitclosely to the climb constraint line. This validates this requirement.
“chap05” — 2003/3/10 — page 132 — #32
132 Aircraft Design Projects
Thru
st/w
eigh
t ra
tio
Take-off run
Climb rate
FL250 turn
SL turn
Initial designpoint
New designpoint
Landing run andapproach speed62% MTOM
Landing run andapproach speed90% MTOM
0.000
0.100
0.200
0.300
0.400
0.500
0.600
0.700
100 200 300 400 500 6000Wing loading (kg/m2)
Fig. 5.21 Aircraft constraint diagram
• None of the existing aircraft satisfy the landing conditions at MLAND = 0.9MTO.This suggests that this requirement is too tight.
• The turn requirements do not present critical design conditions for any of the aircraft.The 25 000 ft turn criteria is seen to be the most severe. Some further detailed analysissuggests that the aircraft is capable of a 3g turn rate at this altitude.
Warning: The constraint analysis described above is a very approximate analytical toolas it does not take into account some of the finer detail of the design (e.g. detailedchanges in engine performance with speed). It can only be used in the form presentedin the initial design phase. Later in the development of the layout more detailed analysisof the performance will enable the effect of the various constraints on the aircraft designto be better appreciated. However, with this consideration in mind it is possible to usethe constraint diagram to direct changes to the original baseline layout as discussedbelow.
5.9 Revised baseline layout
The main conclusion from the constraint analysis and aircraft performance estimationsis that the aircraft landing requirements are too tight and should be renegotiated withthe customers. To provide evidence on the effects of the landing constraints, the revisedbaseline layout will ignore them. The new design can be analysed to show what landingcharacteristics are feasible.
With the above strategy in mind the design point for the aircraft will be moved closerto the intersection of the take-off and climb constraint lines, i.e.:
(T/W ) = 0.38 and (W/S) = 390 kg/m2(80 lb/sq. ft)
Anticipating the need to increase aircraft mass to allow more fuel to be carried, the max-imum take-off mass is increased to 5850 kg (and the structural design mass increased
“chap05” — 2003/3/10 — page 133 — #33
Project study: military training system 133
to 6100 kg). Using the new values for (T/W ) and (W/S) the new thrust and wing areabecome:
T = 0.38 × 5850 = 4900 lb (SSL)
S = 5850/400 = 14.65 m2(136 sq. ft)
For an aspect ratio (AR) of 5, the new area gives a wing span (b) = 8.56 m and a meanchord = 1.71 m. For an aspect ratio of 4.5 the wing geometry becomes b = 8.12 m andmean chord = 1.80 m. Rounding these figures for convenience of the layout drawinggives:
cmean = 1.75 m (5.75 ft) and b = 8.5 m (28 ft)
∴ gives, AR = 4.86 and S = 14.87 sq. m/160 sq. ft
This geometry will be used in the new layout.Also, since the tip chord on the previous layout seemed small, the taper ratio will be
increased to 0.33.
Hence Cmean = (Ctip + Croot)/2 = 1.75 m (assumed)
With, (Ctip/Croot) = 0.33
This gives Croot = 2.63 m/8.6 ft, Ctip = 0.87 m/2.8 ft
5.9.1 Wing fuel volume
It is now possible to check on the internal fuel volume of the new wing geometry.Assume 15 per cent chord is occupied by trailing edge devices and 33 per cent span istaken by ailerons (assume no fuel in the wing tips ahead of the ailerons).
Although previously the wing thickness was assumed to be 10 per cent, it has nowbecome clear that the aircraft will require substantial internal volume for fuel storage.To anticipate this, the wing thickness will be increased to 15 per cent in the expectationthat supercritical wing profiles can be designed to assist in the transonic flow conditionsparticularly for the high-speed development aircraft.
With the above geometry (see Figure 5.22) and assuming 66 per cent of the enclosedvolume is available for fuel, gives an internal wing fuel capacity of 0.5 m3. A totalfuel load of 1050 kg equates to a volume of 305 Imp. gal. This requires a volume of1.385 m3. It is therefore necessary to house some fuel in the aircraft fuselage (namely1.385 − 0.5 = 0.885 m3). This is not uncommon on this type of aircraft. The preferredplace to keep the fuel is in the space behind the cockpit and between the engine airintakes. This is close to the aircraft centre of gravity, therefore fuel use will not causea large centre of gravity movement. For our layout it would be preferable to keep thefuel tank below the wing structural platform to make the wing/fuselage joint simpler.From the original aircraft layout this fuselage space would provide a tank volume ofabout 1 × 2 × 0.5 = 1 m3. This is satisfactory to meet the internal fuel requirement.Using all of this space for fuel may present a problem for the installation of aircraftsystems. To anticipate the need for extra space in the fuselage to house the electronicand communication systems an extra 0.5 m will be added to the length of the fuselage.Moving the engine and intakes back to rebalance the aircraft will also provide a cleanerinstallation of the intake/wing junction (i.e. moving the intake behind the wing leadingedge).
“chap05” — 2003/3/10 — page 134 — #34
134 Aircraft Design Projects
Wing LE extension
Wing LE fuel tank
25% MAC
Front sparline
MAC
25%C
50%C
Rear sparline
FLAPAILERON
Aircraft centre line
Fuselage bodyside
Fig. 5.22 Revised aircraft wing planform
Lengthening the fuselage has the effect of increasing the tail effectiveness. This maypermit either a traditional low tailplane/fin arrangement, or more likely, a twin fin/tailbutterfly layout. Subsequent wind tunnel tests and CFD modelling would be necessaryto define the best tail arrangement. In the revised layout a butterfly tail will be shownto illustrate this option.
It is now possible to redraw the baseline layout to account for the above changes. Atthe same time it is possible to add more details to the geometry (Figure 5.23).
5.10 Further work
With the new baseline aircraft drawing available and increased confidence in the aircraftlayout it is possible to start a more detailed analyses of the aircraft.
We start this next stage by estimating the mass of each component using the newaircraft geometry as input data for detailed mass predictions. Such equations can befound in most aircraft design textbooks. These formulae have, in general, been derivedfrom data of existing (therefore older) aircraft. As our aircraft will be built using mate-rials and manufacturing methods that have been shown to provide weight savings itwill be necessary to apply technology factors to reduce the mass predicted by theseolder aircraft related methods. The factors that are applied must correspond to theexpected degree of mass reduction. Different structural components will require indi-vidual factors depending on their layout. For example, the wing structure is more likelyto benefit from a change to composite material than the fuselage. The fuselage has manymore structural cut-outs and detachable access panels than the wing which makes itless suitable. The mass reduction factors for composite materials may vary between 95and 75 per cent. The lower value relates to an all-composite structure (e.g. as used forcontrol surfaces and fin structure).
“chap05” — 2003/3/10 — page 135 — #35
Project study: military training system 135
metres
Extra equipmentExtra fuel tank
17°
15°
15°
Conformal systems pack
0 1 2
fuel
Fig. 5.23 Revised baseline aircraft layout
Aspects other than the choice of structural material may also influence the estimationof component mass. Such features may include the requirement for more sophisticationin aircraft systems to accommodate the remote instructor concept, the requirementsrelated to the proposal for variability in the flight control and handling qualities of theaircraft to suit basic and advanced training, and the adoption of advanced technologyweapon management systems. All such issues and many more will eventually need tobe carefully considered when finalising the mass of aircraft components.
When all the component mass estimations have been completed it will be possible toproduce a detailed list in the form of an aircraft mass statement. Apart from identifyingvarious aircraft load states, the list can be used to determine aircraft centre of gravitypositions. As the aircraft will be used in different training scenarios (e.g. basic aircrafthandling experience to full weapon training) it is necessary to determine the aircraftcentre of gravity range for different overall loading conditions. With this information itwill be possible to balance the aircraft (see Chapter 2, section 2.6.2) and to accurately
“chap05” — 2003/3/10 — page 136 — #36
136 Aircraft Design Projects
position the wing longitudinally along the fuselage. Up to this point in the designprocess the wing has been positioned by eye (i.e. guessed).
With the wing position suitably adjusted and a knowledge of the aircraft massesand centre of gravity positions, it is now possible to check the effectiveness of the tailsurfaces in providing adequate stability and control forces. Until now the tail sizes havebeen based on the area ratio and tail volume coefficient values derived from existingaircraft. It is now possible to analyse the control surfaces in more detail to see if theyare suitably sized.
The previously crude methods used to determine the aircraft drag coefficients cannow be replaced by more detailed procedures. Using the geometry and layout shown inFigure 5.23 it is possible to use component drag build-up techniques or panel methodsto determine more accurate drag coefficients for the aircraft in different configurations(flap, undercarriage and weapon deployments). Aircraft design textbooks adequatelydescribe how such methods can be used. Likewise, more accurate predictions can nowbe made for the aircraft lift coefficient at various flap settings.
Before attempting to reassess aircraft performance it is necessary to produce a moreaccurate prediction of engine performance. If an existing engine is to be used it maybe possible to obtain such data from the engine manufacturer. If this is not feasible itwill be necessary to devise data from textbooks and other reference material. It maybe possible to adapt data available for a known engine of similar type (e.g. equivalentbypass and pressure ratios) by scaling the performance and sizes. Design textbookssuggest suitable relationships to allow such scaling.
More detailed aircraft performance estimations will be centred on point performance.The results will be compared to the values specified in the project brief and subsequentconsiderations. The crude method used previously will be replaced by flight dynamiccalculations (e.g. the take-off and landing estimations will be made using step-by-steptime methods).
It is also possible at this stage to use the drag and engine performance estimations toconduct parametric and trade-off studies. These will be useful to confirm or adjust thevalues used in the layout of the aircraft geometry (for example, the selection of wingaspect ratio, taper, sweepback and thickness).
Further detailed work on the aircraft layout will include:
• The identification and specification of the aircraft structural framework.• The installation of various aircraft system components. This will require some
additional data on the size and mass of each component in the system (e.g. APU).• A more detailed understanding of the engine installation. This will include the
mounting arrangement and access requirements. It will also be necessary to considerthe intake and nozzle geometry in more detail.
• Investigate the landing gear mountings and the required retraction geometry.• Make a more accurate evaluation of the internal fuel tank volumes (wing and fuselage
tanks).• Detailed considerations of the layout requirements for wing control surfaces
including flap geometry.
It is obvious that the above list of topics requires a great deal of extra work. All of this isnecessary in order to draw the final baseline layout. It would be wasteful to do all of thiswork without first reviewing the project and considering the overall objectives againstthe predicted design. The following section outlines the nature of such a review process.
“chap05” — 2003/3/10 — page 137 — #37
Project study: military training system 137
5.11 Study review
There are several different ways in which a design review can be conducted. At thehigher level a technique known as a SWOT (strengths, weaknesses, opportunities,threats) analysis can be used. At a lower (more detailed) level an analysis similar tothat described in section 2.10.2 could be followed. In this study we will adopt the SWOTmethod as this will illustrate the use of this technique in a design context. It must beemphasised that the low- and high-level methods of review are not mutually exclusiveand that in some projects it is advisable to use both.
Before starting the review it must be mentioned that the descriptions below do notconstitute a complete analysis. A project of this complexity has many facets and itwould be too extensive to cover all of them here. The intention is to provide a guide tothe main issues that have arisen in the preceding work.
5.11.1 Strengths
The most obvious advantage of this project lies in the overall life cycle cost (LCC)savings that are expected from introducing a new advanced technology, training system,approach. If such savings cannot be shown it will be difficult to ‘sell’ the new system toestablished air forces. The savings will accrue from the lighter modern aircraft. The useof composites will increase the purchase cost of the aircraft based on the price per unitweight. This would also require extra stringency in inspection of the structure. Moreelaborate systems will also increase the aircraft first cost. However, the new conceptwould avoid duplicity of aircraft types in the basic to advanced phase and this willreduce life cycle costs. In addition, the aircrew will have received a higher standardof training from the advanced training system, a consequential reduction of OCRtraining cost.
The second most powerful advantage for the new concept lies in the ability of theaircraft to more closely match modern fast-jet performance than is currently possiblewith training aircraft that were originally conceived and designed in the 1970s.
Another strength of the new system is the total integration of modern flight andground-based systems into a total system design approach. Upgraded older aircrafttypes are not capable of achieving this aspect of the training system.
Many more advantages could be listed for the system. How many can you identify?
5.11.2 Weaknesses
There are three principal weaknesses to the project as currently envisaged. To reducethese deficiencies, if at all possible, it will be necessary to devise strategies ormodifications to our design.
The main and intrinsic difficulty lies in the conservative nature of all flight train-ing organisations. This is a natural trait as they take responsibility of human life andnational security. As such they will be highly sceptical of the potential advantages ofconducting advanced training in a single seat aircraft with a remote instructor. Forour concept, as we currently envisage it, this difficulty is insurmountable. Therefore achange of design strategy must be considered to save the credibility of the project. Itwill be necessary to extend the design concept to encompass a two-seat trainer through-out the full (basic to advanced) training programme. The remote instructor conceptcan be developed as a separate part of the aircraft/system development programme(i.e. flight testing the aircraft without the instructor present as a proof of concept).
“chap05” — 2003/3/10 — page 138 — #38
138 Aircraft Design Projects
This would allow the design and validation of the ground-based instructor system andassociated communication and data linking without jeopardising the success of the tra-ditional design. As we had already accepted that the basic training role would requirethe development of a two-seat variant, the new strategy will only involve an upgradeto the design to allow the full payload to be carried in this version. Initial calculationssuggest that the new aircraft will be about 500 kg (1100 lb) heavier than the existingdesign (i.e. approximately 10 per cent increase in MTOM). At this point in the develop-ment of the project it is obvious that significant changes to the baseline aircraft wouldbe required. Therefore, the work on the present design must be delayed until a revisedbaseline layout is produced.
The second weakness is associated with the risk involved in the development of newtechnologies on which the whole system is reliant. If the changes described above areaccepted this risk to the project will be avoided. The remaining technologies usedin the design can be assured by their current adoption in new aircraft projects (e.g.Eurofighter, F22 and JSF).
The third area relates to the selection of engine for the existing design. From theprevious work there are two aspects that require further consideration. First, the Adourengine is shown to be too powerful for our design. The original suggestion (to deratethe engine) would only seem to be sensible if the full-rated engine was to be used infuture aircraft variants. For the existing trainer aircraft, incorporating an engine largerthan necessary effectively adds about 100 kg to the aircraft empty mass. A secondpropulsion issue relates to fuel usage. Previous calculations showed that the requiredferry range was not feasible without seriously penalising the aircraft MTOM. Evento accommodate the fuel required to fly the training sorties it was shown necessaryto extend the fuselage to house a larger fuel tank behind the cockpit. For each of thethree missions investigated it was found necessary to increase the fuel load that hadbeen previously assumed. As the fuel requirements are directly related to the enginefuel consumption, and thereby to operational cost, it would be advantageous to use amore fuel efficient engine.
Selecting a modern higher-bypass engine with slightly less static sea-level thrust wouldoffer a better design option than using the Adour. Although the engine will be of largerdiameter and therefore increase the size of the rear fuselage, it will be lighter and useless fuel. Overall, the change will lead to a lighter and potentially cheaper aircraft.
From the engine data collected earlier (section 5.4.3) there are three possible enginesfrom which to choose (specific fuel consumption (sfc) in lb/lb/hr or N/N/hr):
1. TFE 731-60 manufactured by Allied Signal and used on the Citation and Falconbusiness jets (SSL thrust = 5590 lb, sfc = 0.42, L = 1.83 m, dia. = 0.83 m, depth =1.04 m, dry mass = 448 kg).
2. CFE 738 (General Electric/ASE) used on the Falcon business jet (5725 lb, sfc = 0.38SSL, sfc = 0.64 cruise, L = 2.5 m, W = 1.09 m, depth = 1.2 m, mass = 60 1kg).
3. PW 306A (Pratt & Whitney of Canada) used on the Dormier 328 regional jet(5700 lb, sfc = 0.39, L = 2.07 m, W = 0.93 m, depth = 1.15 m, mass = 473 kg).
Aircraft manufacturers prefer to have a choice of available engines as this adds com-petition on price and delivery. The three engines above are all used on civil aircraftand this may further provide a cost advantage as engine manufacturers will identifyan additional market for their product. This should result in a competitive commercialadvantage. Approval for military applications will require some extra certification workbut this extra cost will be negligible compared to that required to design and developa completely new engine.
“chap05” — 2003/3/10 — page 139 — #39
Project study: military training system 139
Selecting the PW306A engine would reduce the current dry engine mass by 130 kg(287 lb). This would also reduce the propulsion group mass, thereby reducing the air-craft empty mass. Assuming a cruise specific fuel consumption of 0.64 (as quoted forthe equivalent CFE engine) reduces the fuel required to fly the 1000 nm ferry range fromthe previously estimated 1733 kg for the Adour engine to 1099 kg. This is close to the900 kg (1985 lb) initially assumed for the fuel mass. The 2000 nm ferry range (assumingexternal tankage) would require 2328 kg of fuel. This is close to the combined fuel andweapon load (900+1360 = 2260 kg/4984 lb) originally specified. Therefore, it appearsthat by installing this type of engine it would not be necessary to request a reductionin the specified ferry range from originators of the design brief.
The design penalty for installing the higher-bypass type engine lies in the requirementfor a larger rear fuselage diameter. The PW306 engine is 0.17 m (7 in) larger in diameterthan the Adour. The extra fuselage mass required to house the fatter engine wouldbe more than offset by the reduction in fuel tank weight. The higher bypass ratioengine will also suffer greater loss of thrust with altitude and speed than a pure jetengine.
For designers, the selection of an engine is always a difficult decision as many non-technical factors may intrude into the process (e.g. political influences, offset cost andmanufacturing agreements, national manufacturing preference). Without a knowl-edge of these influences on this project it is recommended that the PW306A engine isinstalled. This decision will still allow the other competitor high-bypass engines listedabove to be used if commercially advantageous. Alternatively, the Adour engine couldbe used but this would involve a substantial reduction in aircraft range capability unlessexternal tanks are fitted.
5.11.3 Opportunities
Most of the successful training aircraft were originally designed over 20 years ago.Although many have subsequently been ‘modernised’ they still present old technologiesfor structure, engines and some systems. The capability of modern fast-jets in the sameperiod has substantially changed and the nature of air warfare which has developedwith these improved capabilities. This situation opens a wide gap in the effectivenessof old trainers to meet current demands. Here lies the major opportunity for a newtrainer design.
Nearly all of the existing successful trainers have been developed into light combatvariants for local area defence and ground attack. However, many of these aircraft areof limited capability due to the age of their systems and their inadequate performance.Our new trainer could be developed into an effective combat aircraft to compete withthese existing older trainer aircraft variants.
There is therefore substantial worldwide potential for marketing a new trainer andits derivatives.
5.11.4 Threats
We are not alone in identifying the need for a new trainer. Two other countries havestarted to manufacture and develop new trainer aircraft over the past few years. Thesecould present a serious commercial challenge to our project unless we can exploitour advanced technologies to produce a more effective and technically capable designsolution.
“chap05” — 2003/3/10 — page 140 — #40
140 Aircraft Design Projects
5.11.5 Revised aircraft layout
The result of the study review has proposed significant changes to the existing baselinelayout. These include:
• a two-seat cockpit,• a change of engine,• a requirement for less internal fuel volume.
Each of these changes will effect the aircraft mass and geometry. A revised gen-eral arrangement drawing of the new baseline layout is shown in Figure 5.24. Initialcalculations showed that the increase in aircraft structural mass resulting from theaddition of the second seat and larger diameter engine has been offset by the reductionin mass from the lighter engine and the reduced fuel requirement.
The single-seat derivative of the new aircraft would benefit from either a 230 kg/507 lbincrease in weapon load, or by an increase in range from the equivalent 230 kg increasein fuel load. The single-seat variant is shown in Figure 5.25.
The detailed analysis of the new aircraft follows the same methods as outlined earlierin this chapter. To avoid repetition these calculations have not been included in thischapter.
metres
1 20
Fig. 5.24 Post-design review layout (two seat)
“chap05” — 2003/3/10 — page 141 — #41
Project study: military training system 141
+Extra systems
Enlarged fuel tank
0 1 2
+
metres
Fig. 5.25 Single-seat aircraft variant
5.12 Postscript
This study has demonstrated how project design decisions may change as the aircraftis more thoroughly understood. This demonstrates the iterative nature of conceptualdesign. It is possible for students to continue this project into the next iterative stageusing the final aircraft drawings (Figure 5.25) as the starting point.
References
Textbooks for military aircraft design and performance:
1 Raymer, D. P., Aircraft Design: A Conceptual Approach, AIAA Education Series, 1999, ISBN1-56347-281-0.
“chap05” — 2003/3/10 — page 142 — #42
142 Aircraft Design Projects
2 Brandt, S. A. et al., Introduction to Aeronautics: A Design Perspective, AIAA EducationSeries, 1997, ISBN 1-56347-250-3.
3 McCormick, B. W., Aerodynamics, Aeronautics and Flight Mechanics, Wiley and Sons, 1979,ISBN 0-471-03032-5.
4 Nicolai, L. M., Fundamentals of Aircraft Design, METS Inc., San Jose, California 95120,USA, 1984.
5 Mattingly, J. D., Aircraft Engine Design, AIAA Education Series, 1987, ISBN 0-930403-23-1.
The following publication is also useful in collecting data on existing aircraft:
Aviation Week Source Book, published annually in January.
This handbook is a useful source of general aeronautical data:AIAA Aerospace Design Engineers Guide, 1998, ISBN 1-56347-283X 1.
“chap06” — 2003/3/10 — page 143 — #1
6
Project study:
electric-powered racing
aircraft
Existing Formula 1 racers
“chap06” — 2003/3/10 — page 144 — #2
144 Aircraft Design Projects
6.1 Introduction
This project is the direct result of collaboration between aeronautical and automotiveresearch teams. Government requirements aimed at reducing the detrimental effects ofemissions from automobiles on the environment have stimulated the automotive indus-tries into investigating and developing alternative power sources for mass producedcars and light vans. Various types of electric propulsion systems have been studiedin detail. These produce near-zero, harmless emissions. Future automotive legislationmay require a substantial and increasing proportion of motor vehicles to be environ-mentally ‘friendly’. It is expected that this will result in the development of lightweightand cheap electric propulsion systems. Such systems could be adapted for aircraft use.Although the reduction of emissions is not too significant for the short duration of arace, the development flights for this aircraft and the use of such propulsion systemsin other applications must be considered. Investigating this possibility in a competitiveenvironment that will stimulate rapid technical development is the main objective ofthis project. And, of course, the design of a fast racing aircraft should also be fun!
6.2 Project brief
From the earliest beginnings of powered flight, general/light aviation has modifiedautomotive engines for powering aircraft. Even the famous Wright Brothers followedthe principle in their epic first flights about a hundred years ago. As in the developmentof any new technology and innovation, it is necessary to introduce new concepts slowlyand in a controlled environment. Sport aviation has traditionally been a suitable wayof developing such technologies into commercial opportunities. Air racing is currentlyreported to be the fastest growing motor sport in the USA. Commercial sponsorshipand television sports coverage of weekend race meetings have generated renewed inter-est in the sport. This environment offers the means by which we could gain flyingexperience with a new propulsion system in a highly controlled environment.
As we will be designing a new racing aircraft, it is important to investigate the currentair-racing scene. At present, there are several classes of air racing. The two most closelycontrolled pylon-racing organisations are Formula 1 and Formula V (vee). The maindifference between these lies in the specification of the engine type. Formula 1 relates tothe 200 cu. in. Continental (0–200) engine and for Formula V to a converted Volkswagenengine (hence the significance of the vee). Using this pattern, we should project a newFormula (E) to relate to the electric propulsion.
Apart from the engine details, all other requirements should match the Formula 1rules. In this way, the new formula will benefit from the many hours of successfulracing experience. It will also ensure that the race organisers accept the new formula.The rules and procedures are available from the Formula organisers and are publishedon the Web.1 The main features, and a brief history of air racing, are described below.
6.2.1 The racecourse and procedures
The race starts with a field of six to eight aircraft on the ground (runway) for a simul-taneous take-off. The normal formation consists of three aircraft in front, two in themiddle position, and three at the rear. As in motor racing, the positions on the startinggrid are related to previous race performance. The fastest aircraft/pilots are at the front
“chap06” — 2003/3/10 — page 145 — #3
Project study: electric-powered racing aircraft 145
1.75 miles
0.5 mile
1.25 miles
4
3
2
1
5
6
Start /Finish
Scatter
Eightaircraftstart
Fig. 6.1 Racecourse geometry
of the grid and have a 150-yard advantage over those at the back. As in earlier motorracing, the racing team ground crew assist in starting the engine, securing the pilot (etc.)and preparing the aircraft for the race but must leave the take-off area no later thanone minute prior to the start. A green flag is raised about ten seconds before the ‘off’at which point the pilots apply full throttle. When the green flag drops the race begins.
The racecourse consists of a two turn, three-mile oval as shown in Figure 6.1. Theseven marker pylons that define the course are typically 60-gallon oil drums fastenedon the top of short poles. The first pylon (outside the oval track) is called the scattermarker. Although the aircraft are racing from the take-off, the lap that includes thescatter pylon is not included in the race. The racing time starts when the first aircraftpasses the start/finish line. Races usually last for eight laps (sometimes six dependingon the number of heats that are required to sort out the field). Overtaking is the ‘nameof the game’ but pilots should pass high and outside the flight path of the slowercompetitor. Stewards are positioned at each pylon to ensure that pilots do not ‘cut’the track. Such indiscretions earn the pilot penalty time. This is two seconds per lap,which is more than can be won back in the race. It is therefore important to have clearvisibility to ensure that such penalties are avoided. The 24 fastest aircraft/pilots fromthe heats are split into three groups. The slowest group competes for the bronze, thenext for the silver and the fastest for the gold. The winner of the gold race is crowned thechampion of the race. These victories build up points for the national championship.Prize money is earned in proportion to the success in the heats and, more profitably,in the finals.
6.2.2 History of Formula 1 racing (further reference can be
found on the Formula 1 web site1)
Prior to 1945, racing aircraft were mostly original designs specifically aimed at racing.They were unique creations that often advanced the field of aeronautics. Innovativedesigners of air racers consistently produced aircraft that outperformed the best mil-itary aircraft of the day. In the early days, these aircraft led to the development ofmonoplane wing layouts and introduced materials and construction methods that werelighter and more reliable. After World War II, there was a surplus of high-powered,
“chap06” — 2003/3/10 — page 146 — #4
146 Aircraft Design Projects
mass-produced, ex-military aircraft. These were introduced into air racing but thepurists argued that this was not in the spirit of the original sport. Such aircraft wereraced in an open classification category that still exists. To return to the original rac-ing concept and to make the sport more pilot-centred, the Formula rules mentionedearlier were created. In 1946, these special racing criteria were introduced to continueaerodynamic refinement within the financial means of many people.
The first racing class was referred to as the ‘Midget’ class due to the relatively smallsize of the aircraft. Initial rules limited the aircraft to a minimum wing area of 66 sq. ft,fixed-pitch propeller, and engines limited to a maximum displacement of 190 cu. in.The aircraft also had to have good pilot visibility, nose-over pilot protection and weighat least 500 lb (227 kg) empty. These constraints forced designers to concentrate onminimising drag and structural weight. Sponsorship was initially with the GoodyearCompany, hence the name ‘Goodyear racers’ was formed.
The first competition of the Midget class was held at Cleveland in 1947. Thirteenaircraft competed in this event. By 1949, the field had grown to 25 and the prizemoney to $25 000. This attracted the famous names of aviation enthusiasts and modelbuilders. Unfortunately, it was also during this meeting that Floyd Odom, a renownedracer, crashed his ex-military fighter aircraft into a home killing a young mother andher baby. This accident and the sudden realisation that racing over populated areas waspotentially very dangerous caused many in the population to call for a ban on all airracing and air shows. The multi-classes Cleveland National Air Races were abandoned.Fortunately, Continental Motors sponsored the 190 cu. in. class and kept the sport alive.During the 1950s and early 1960s, the Midget class air races were held throughout thecountry. Multi-class racing was finally resumed in 1964 with the first Reno NationalChampionship races. Reno offered a remotely located site that minimised the risk tonon-participants. The 1964 races were the first major racing event in 15 years. Six190 cu. in. racers took part in this event with famous pilots such as Steve Wittman andAir Scholl participating.
Formula 1 racers have continued to evolve, reflecting the spirit for which the sport wasformed. The most significant change in the rules occurred in 1968 when the maximumengine size was increased to 200 cu. in. and the name changed from ‘Midget’ to ‘1’.Radical designs such as the Miller Special pusher configuration and the all-compositeNemesis racer have recently ‘pushed the envelope’ further. Formula 1 and the morerecent Formula V races continue to attract a healthy mixture of new enthusiasts andseasoned veterans each year.
6.2.3 Comments from a racing pilot
This section is added to offer guidance to prospective designers who have not hadexperience in air racing. Collecting such information from people who have operatedthe type of aircraft you are designing is a good strategy if it is easy to find.
Constant vigilance is the number one rule when racing. You always have to belooking and thinking. It is essential to always be looking around for other aircraftand to be keeping abreast of the current situational awareness. In fact, you shouldstart this process on the ground at the start of the race and understand whichaircraft are likely to be accelerating to gain the early advantage. Know whichway the wind is blowing as this will affect the racing line to be taken by yourcompetitors. Watch out for shadows when you are turning around the pylon.Who is above you and likely to overtake you in the turn, and who is below andperhaps will be over-run by you.
“chap06” — 2003/3/10 — page 147 — #5
Project study: electric-powered racing aircraft 147
In the race, you want to fly a smooth line at constant altitude but this will notbe entirely feasible if you want to overtake and avoid contact with a competitor.I like to fly with my lowest wing tip about 20 feet above the top of the pylon. Inthis way, if I get blown off line or misjudge a turn I still avoid the pylon. It alsoleaves me with a way out if I think that I might get hit. Normally I like to see abouthalf a circle looking through the pylon. If you try to see more than this you runthe risk of a cut if the wind or turbulence suddenly alters your course. In a turn,we do not throttle back, as in motor racing, we just bank and add a little loadfactor on!
When it comes to overtaking, it is important to take care. This means seeingthem and the course ahead at all times. The best way is to pass them above and onthe outside. This requires good airmanship. You have to fly carefully but if you gotoo wide or high you will erode your speed advantage and you could interfere withthose aircraft that are following and hoping to overtake. Passing a plane that iswell piloted is fun. One that is flown erratically, or by someone who tries to climbin the turn, can be hazardous. You must know your opponents’ flying abilities.Once you get past a slower aircraft some pilots wave, salute or make other gestures.You are close enough to see facial expressions and hand signals easily!
Turning is the fun part of the race. It is here that the skill in piloting and thetechnical advantage of the aircraft design are most clearly apparent. I try to lookat the next two pylons at all times and keep updating my flight path to get thebest out of the situation. Once a pylon is passed, I forget it and start lining up forthe next one. Since the course is only about three miles long (mile straights andtwo half-mile turns) you can pick up the corner pylons as you enter the straight.Using big landmarks like hangars, road and mountain features help in quicklyidentifying the pylons from the ground clutter.
(with grateful acknowledgements to the Formula 1 web site,1 June 2000)
The description above provides an excellent insight into the enthusiasm and skill ofracing pilots. Once the race is over and the times have been adjusted for any ‘cuts’, thewinner is paraded in front of the cheering crowd (often in a specialist automobile).
6.2.4 Official Formula 1 rules
Current Formula racing rules can be found on the Internet.1 Most of the rules relateto the design and modification of the engine. The list below summarises the aircraftrelated aspects of the rules as they appeared in June 2000:
• wing area, minimum size = 66 sq. ft (6.132 sq. m),• cockpit height, minimum = 30 ins (0.75 m),• pilot visibility, minimum 25◦ aft, 5◦ over nose, 45◦ upwards,• aircraft empty weight (mass) = 500 lb (227 kg) minimum,• aircraft centre of gravity = between 8 per cent and 25 per cent of the wing mean
aero. chord,• main landing unit fixed (nose may retract),• main wheels = 5.00 × 5 tyre size,• fixed pitch propeller (metal props not allowed),• structural design limits = +/−6g minimum.
There are several rules that relate to pilot experience and medical condition. The racecommittee has the right to revoke the licence if unsafe flying habits are demonstrated.
“chap06” — 2003/3/10 — page 148 — #6
148 Aircraft Design Projects
The flight test requirements include:
• perform aileron rolls in each direction with no more than 50 ft loss of height,• perform a half roll to the left then recover with a half roll to the right with less than
50 ft height loss,• fly in formation,• fly the race course safely without climbing in turns,• properly overtake on the racecourse,• complete normal and aborted starts without deviating more than 10 and 20 ft
(respectively) from the straight ground track,• complete normal and engine-out landings.
Reference to the full set of rules should be made if unconventional concepts are to beconsidered.
Apart from the constraints described in the official rules there are a number of otherdesign considerations that must be taken into account when configuring a new racerlayout:
• It is common practice to tow the race aircraft to the air show in a specially builtand prepared trailer. This trailer must comply with national construction and useregulations for highway vehicles. The aircraft will need to be partially dismantled tofit into the trailer. Reassembly of the aircraft at the race site must be easy and quick.More importantly, the aircraft reassembly must be reliable and safe. The descriptionof the aircraft must detail how this is to be done and what special equipment andskills are required.
• Flying at speeds in excess of 200 kt at only 50 to 100 ft above the ground, and incompetition with equally enthusiastic racing pilots, is not regarded as the safest ofactivities! If an accident happens it is important that the pilot is adequately protected.Crashworthiness is a significant consideration in the overall design philosophy of theaircraft.
• Even with our best endeavours, it may be necessary to make changes to the aircraftdesign after the first few races. This ‘tweaking’ of the aircraft is regarded as anessential part of the development process that matches the aircraft, pilot and coursecharacteristics. A design that offers some flexibility will be more easily ‘tuned’ toachieve higher performance and therefore be more competitive.
• Although vision and cockpit dimensions are specified in the rules, these may be con-sidered as the minimum safety standard. Pilots win races by overtaking competitoraircraft, often in a turn. To do this effectively they need to be able to see suitableopportunities quickly and as they arise. For example, a prone position for the pilotmay be better aerodynamically as it reduces frontal area, and this position resistshigher ‘g’ loading, but the pilot would find it too restrictive in racing conditions.
• Although formula racing is becoming ‘commercial’, enthusiastic amateurs stilllargely dominate the sport. Many aircraft are built from kits by such people. Theyoften personalise them to suit their preferences and competitive spirit. Aircraft mustbe designed to be ‘easy to make’ and with the freedom to allow for small changes ifrequired.
• Race duration is only about 10 minutes but, to account for possible delays andemergencies, a flight time of at least 30 minutes must be possible.
• Consideration must be given in the design for flight trials prior to the commerciallaunch of the project. This also extends to the need for pilots to become familiarwith the aircraft prior to the racing season. Flight test programmes lasting up to2 hours should be possible, perhaps with a modified engine if the electric system is
“chap06” — 2003/3/10 — page 149 — #7
Project study: electric-powered racing aircraft 149
not feasible for this duration. Such a change must be carefully considered to avoidpenalising the fundamental configuration of the aircraft.
6.3 Problem definition
For this project, the problem definition phase is relatively straightforward. Put simply,the aircraft is designed to win Formula E races. However, there are several associatedcriteria that must also be taken into account. As with most projects it is importantto identify who is likely to be the ‘customer’ for the aircraft. Although the race pilotwill be the ultimate ‘user’ of the aircraft, he may not be the owner or purchaser. Aswith automobile racing, it is possible to envisage that a corporate/sponsor involvementis desirable. Development of advanced technologies, like electric propulsion, requiressubstantial investment. It would be preferable to design the aircraft for a ‘professional’team rather than an enthusiastic amateur. Such considerations would allow a moreambitious approach to the design and manufacture of the aircraft. This would per-mit more sophistication in the systems to be used. However, such possibilities areaccompanied by the need to be ultra-confident in the performance and operationalcapabilities of the aircraft. Sponsors will soon evaporate if the aircraft is not providingthe success that they will demand. Equally, safety (or more precisely the lack of safety)would be a paramount consideration. Successful companies will not want to be associ-ated with failure of any kind. Such thoughts lead to a number of essential requirementsfor the aircraft:
• The aircraft must have performance that is at least as good as existing Formula 1racers, or it will not be taken seriously by the existing competitor and viewing public.
• Crashworthiness must be a priority in the design layout.• Excellent pilot visibility is essential.• The aircraft must have significant development possibilities.• Aircraft flying characteristics will be of particular significance. It is often difficult to
achieve a suitable blend between racing trim and development flying characteristics.• Aircraft development will require significant effort in flight testing.
Although the main intention of this project is to work alongside a major industrialsponsor, it must be appreciated that air racing is largely dominated by amateurs. In theevent that a sponsor cannot be found, the design should be capable of development foramateur construction and operation.
Assessment of the ‘viability’ of the aircraft at the project stage for a racing aircraftis difficult as it obviously involves, first, the attraction of sponsors and then successin racing. Obviously both of these are interlinked but the main technical criterion isdirected at superiority in competitive racing.
The initial brief, the mandatory airworthiness requirements, and the compulsoryFormula racing rules describe the main definition of the problem. However, as describedabove, there are several other design considerations that must be included in thedevelopment of the aircraft specification:
• Consideration must be given to the methods to be used in the pre-race flight trials.• The specification must allow for 2-hour flight duration in development aircraft.• In racing trim, 30-minute flight duration is required.• Better than average pilot visibility is required to allow tight turns around the course
pylon to be executed with precision.• Pilot crash-survivability must be built into the fuselage/cockpit structure.
“chap06” — 2003/3/10 — page 150 — #8
150 Aircraft Design Projects
• The aircraft must be capable of been ‘trailered’ to the race site.• The aircraft must be capable of reassembly at the race site within 30 minutes and
must not require special facilities, tools or skills.• Safety of the aircraft structure, controls, trim and systems must be assured following
reassembly.
For this aircraft, the principal area of innovation lies in the propulsion system. It maybe possible to distribute the components of the electric system more advantageouslythan with traditional engines. For example, avoiding the blunt nose profile and coolingair intakes of a conventional internal combustion tractor engine layout. An electricpropulsion system could be arranged to have the fuel cell components behind thecockpit yet still adopt a tractor propeller layout. Because of the technical uncertaintyof the propulsion system, it may be a better overall design strategy to avoid develop-ments in other technical areas (e.g. novel aerodynamic devices and complex structures).Such considerations may be delayed until after the integration of the electric propul-sion system has been successfully achieved. This will ensure that the development ofthe new power system is separated from other, potentially troublesome, innovations.A consequence of this approach may be that the aircraft layout and structure followsimilar lines to existing successful racing aircraft, except for the detail positioning ofthe propulsion system components.
6.4 Information retrieval
The sections above have set out the foundations for the initial aircraft specification.Before we move on to selecting a configuration for the aircraft, it is preferable to spendsome time conducting a literature search to review competitor aircraft, understandingracing operations and researching the current electric propulsion systems.
Many of the aspects relating to air racing have been described above. These haveincluded a description of the racing environment and a racing pilot’s views on his flyingrequirements. More can be found from the various web sites related to air racing. Nomore information will be discussed here. The other two aspects of the informationretrieval phase are concerned with existing aircraft and electric propulsion.
6.4.1 Existing aircraft
Existing aircraft can be used to guide us in the choice of configuration for our design.Many of these are home-built designs. Some of them were designed by their enthu-siastic owners. A feature of this collection of aircraft is the uninhibited selection ofunconventional layouts and novel details. The list below is not intended to be a recom-mendation of the ‘best’ designs or of preferred configurations. It is intended to give aflavour of the variety of aircraft to which our design will need to compete. (Apologies tothose who feel that we have not included their favourite designs in this survey!) A briefintroduction to each aircraft is given below and a list of the main characteristics iscompiled for use in the initial sizing stage.
Nemesis
We start with an aircraft that is regarded by many as the ‘state of the art’ in Formularacing aircraft. It has been designed and developed by professional aeronautical engin-eers. It encompasses many features that have contributed to its outstanding successin national championships. The mid-fuselage mounted wing uses a natural laminar
“chap06” — 2003/3/10 — page 151 — #9
Project study: electric-powered racing aircraft 151
flow aerofoil. The wing tips are a noticeable and unique detail of the aircraft. Theyare of a novel ‘cusp’ design. They have been geometrically sheared rearward to reducewing tip vortex power, thereby reducing aircraft induced drag. Recognising the need toreduce wing/fuselage interference effects, the wing is permanently fixed to the fuselagestructure. To transport the aircraft, a fuselage joint behind the wing detaches the rearfuselage and tail controls. The aircraft structure is mainly constructed in mouldedgraphite composite (carbon/foam sandwich). To reduce weight and drag, the Nemesisdesigners have paid great attention to detail design (e.g. one piece wheel fairings). This isa good overall strategy that is worth considering by all new racing aircraft design teams.
AR-5
This aircraft is another example of a modern low drag design. Although powered byonly a 65 hp motor it set a world record in 1992 of 213 mph (which gave a remark-able 3.28 mph per hp). It has a completely, all-composite, home-built structure. Theconventional configuration achieves its outstanding performance by the designers’ andmanufacturers’ attention to detail. This is particularly evident in the treatment of thewing (and tail) to fuselage junctions to reduce interference drag. The lower wing sectionprofile is flat and arranged to be coincident with the bottom fuselage section. The uppersurface of the wing has been carefully ‘filleted’ to the fuselage side profile notably at thenose and tail intersections. The wheels are faired and located in the airstream outsideof the propeller wash. It is regarded as a tribute to the designers that the relatively oldNACA 65 series aerofoil performs so well. The wing achieves almost 50 per cent lam-inar flow. The design strategy for the aircraft is of a simple (uncomplicated) layout, veryclean aerodynamic shape, high-quality (smooth) surface manufacture and an almostobsessive attention to the avoidance of interference drag. As in the Nemesis aircraft,any successful racing aircraft is worth careful study by new aircraft teams.
Monnett Sonerai
This aircraft, a Formula V racer, shows how the configuration is affected by the require-ment to fold the wings to lie along the side of the fuselage for transport (in this casetail-first towing). The resulting low aspect ratio wing planform will be ‘draggy’ in high-gturn manoeuvres but set against this deficiency is the low wing weight from the shortspan composite structure. This presents the classical aeronautical dilemma – savingsin drag from low weight but increased drag from lower wing aspect ratio.
Option Air Reno Acapella
This aircraft illustrates a different configuration for the engine/airframe integration.The small sized fuselage houses a rear engine ‘pusher’ propeller. Twin booms mountedfrom the wing structure provide support for the rear fins, rudders, tailplane and ele-vator. The advantage of the rear propeller lies in the avoidance of prop-wash over thefuselage profile. This should reduce drag, but as the propeller is positioned in a moredisturbed airflow, it will be less efficient. (Another aeronautical dilemma!)
Alpha Macro J-5
The Alpha is another example of a ‘pusher’ layout. The reason for the configuration isthe desire to avoid the increased airspeed from the propeller flowing over the fuselageand increasing aircraft parasite drag. To avoid the reduction in propeller efficiencyfrom the blockage of airflow into the propeller, the engine in this aircraft is mounted
“chap06” — 2003/3/10 — page 152 — #10
152 Aircraft Design Projects
high on the cockpit ‘capsule’. This layout allows a low mounted single tail boom to bepositioned below the propeller disc. A ‘butterfly’ tail is positioned at the rear of the tailboom. Although overcoming some of the problems normally associated with a pusherlayout, care must be taken with this configuration to avoid aircraft trim variation withpower changes.
Holcomb Perigree
Although this is not a racing aircraft, it has been included in this review to illustrateanother pusher propeller layout. In this case, the propeller is mounted on an extensionto the rear fuselage. The extension is long enough and suitably shaped to reduce thepropeller ‘blanking’ and to ensure that the tail surfaces are not too much affected by theinflow to the propeller. The fin is arranged in a ventral (downward) position to providethe mounting structure for the tail wheel. This gives the aircraft an approximatelyparallel attitude to the ground at take-off and avoids the high-drag, tail-down fuselageand propeller attitude of more conventional designs. It is expected that this will reducedrag during take-off but has the disadvantage of reducing wing incidence, therefore liftgeneration. However, it is claimed that the aircraft take-off performance is improvedover the tail-down attitude of more normal configurations. Without the use of airbrakes, the landing distance will be increased due to the streamlined layout of theaircraft on the ground run.
FFT SC01B: Speed Canard
Although this aircraft is also not a racing aircraft or even a single seater, it is aninteresting example of the canard configuration that is used on some aircraft. Thisaircraft is based on the successful Rutan VariEze tourer aircraft. Its structure is acomposite of GRP and foam materials. The wing is mid-mounted and swept back toprovide wing tip mounted vertical control surfaces. The rear engine, pusher layoutcomplicates the shape of the rear fuselage but has been carefully contoured to providea good aerodynamic profile. This layout requires a tricycle undercarriage arrangement.The nose unit retracts into the front fuselage profile. The main units are fixed. Toprovide the required balance to the aircraft, forward swept wing root extensions areused as fuel tanks.
In summary, several aircraft were researched in the literature search. Some of theseare potentially competitor aircraft while others are included to illustrate alternativeconfigurational options. Some selected technical details of the aircraft are shown inTable 6.1.
The details in Table 6.1 will be useful when making the initial estimates for ouraircraft.
6.4.2 Configurational analysis
In reviewing all the different types of aircraft that are similar to our expected design, it isclear that the main configurational decision to be made rests between the choice of trac-tor or pusher propeller position. Both have advantages and disadvantages associatedwith airflow conditions over the aircraft profile. As neither configuration has emergedin the preferred layout for modern racing aircraft, there seems to be no over-ridingtechnical (racing efficiency) reason for the choice.
From the review, the conventional tractor layout is seen to have less variation in theoverall aircraft layout. The traditional two-surface layout prevails with the mainplane
“chap
06”—
2003/3/10—
pag
e153
—#11
Table 6.1 Survey of existing aircraft
Long Span Area ME MTO W/S PName Description (m) (m) (sq. m) AR (kg) (kg) ME/MTO (N/m2) (kW) T/W
Nemesis Formula 1 racer 6.71 6.41 6.22 6.6 236 340 0.69 536 75 0.16AR-5 Formula 1 racer 4.42 6.4 5.12 8.0 165 290 0.57 556 49 –Monnett Sonerai Formula V racer 5.08 5.08 6.97 3.7 199 340 0.59 479 45 0.15Optiori AirReno Formula V – twin boom pusher 5.03 8.08 6.06 10.8 295 473 0.62 766 88 0.18Alpha Macra J-5 Home-built – twin boom pusher – 8.16 6.28 10.6 165 290 0.57 453 19 0.12Perigree Kit – pusher – high wing 4.78 8.53 7.57 9.6 172 326 0.53 422 26 0.13FFT Speed Canard Two seat – canard – sport 7.79 7.79 7.88 7.7 440 715 0.62 890 120 0.17Cassult Special Formula 1 – Home-built 4.88 4.57 6.27 3.3 227 363 0.63 568 64 0.17Pottier P70s Home-built – sport a/c 5.15 5.85 7.21 4.7 215 325 0.66 442 45 0.2Monnett Money Blended canopy sport a/c 4.67 5.08 4.27 6.0 191 295 0.65 678 60 0.19Aerocar Micro Pup High wing – pusher 4.57 8.23 7.49 9.0 118 238 0.50 312 22 0.15
“chap06” — 2003/3/10 — page 154 — #12
154 Aircraft Design Projects
ahead of the control surfaces. On the other hand, the pusher layout offers severaloptions. These include either tail or canard control surfaces. If the tail arrangement isselected, this presents difficulties at the rear fuselage. Using a twin boom layout avoidsthe tail surfaces/propeller interference but complicates the wing and fuselage structure.Lifting the propeller line above the fuselage may cause trim changes with power andalso complicates the rear fuselage profile.
The choice of landing gear geometry lies between the nose (tricycle) and the tail (taildragger) arrangements. The tail wheel layout is lighter but introduces the possibilityof ground looping. Current formula rules prohibit retraction of the wheels but ourproposed Formula E rules will allow the auxiliary wheel to be retracted as this doesnot seem to overcomplicate the design yet improves aerodynamic efficiency.
In selecting the aircraft configuration, the most significant criterion is the requirementfor high aerodynamic efficiency (i.e. low drag). This implies:
• smooth profiling of the external shape of the aircraft,• avoidance of the canopy/windscreen discontinuity,• fairing of the landing gear and other structural details,• reduction of airflow interference areas (e.g. mid-mounting of the wing to fuselage),
and• avoidance of engine/propulsion system cooling drag.
Many of the low drag features would be considered during the manufacturing (surfacesmoothness and preparation) and operational (gap taping and surface cleaning) phases.
For this project, the most significant difference in configuration compared with con-ventional designs is the location of the various components of the propulsion system.Whereas conventional designs have the propeller and engine closely positioned, in anelectric system only the electric motor is linked to the propeller. This motor is muchsmaller than a conventional internal combustion engine and can therefore be stream-lined into the fuselage profile. All other components in the electrical system can belocated in convenient positions in the aircraft. These options will create an installationthat has potentially less drag and higher propeller efficiency. It is also envisaged thatthe electrical system will require less cooling than the equivalent internal combustionengine. This will also reduce aircraft drag.
6.4.3 Electrical propulsion system
Powering an aircraft with an electrical power system is not new. Several attempts weremade in the early days of aviation to incorporate electrical power into aircraft. However,the main difficulty rested with the storage of electrical energy in bulky and heavybatteries. Even now, batteries are still too heavy to be used in general aviation aircraft asthe main power source, except for aircraft with very short operating time. One aeronaut-ical field in which electrical propulsion has become established over the past 20 yearsis for model aircraft flying. Progressive miniaturisation of electrical and electroniccomponents has shown the advantages of the technology. The aircraft are environ-mentally cleaner, more reliable and easier to integrate the operating systems. Morerecently, a number of large aircraft have been designed and flown using solar panels asthe power source. Reference to the Pathfinder, Solar Challenger and Penguin aircraftshould be made to understand these developments (namely, NASA and LawrenceLivermore National Laboratory).
As described earlier in this chapter, the automotive industry has again become inter-ested in electrical propulsion for environmental reasons. The search for clean power
“chap06” — 2003/3/10 — page 155 — #13
Project study: electric-powered racing aircraft 155
systems has led to the development of alternatives to the traditional lead–acid battery.Many of the leading companies have displayed prototype vehicles using fuel cells.
A fuel cell is a chemical and mechanical device to convert chemical energy stored ina source fuel into electrical energy without the need to burn the fuel. Theoretically, ifthe device can be made to work, it is potentially highly efficient, has almost harmlessemissions, and is quiet. Such devices are not new. They have been successfully usedin spacecraft and submarines for many years. There have also been some small-scaleapplications used in some industrial, power generation units.
The fundamental operation of a fuel cell matches that of a traditional battery.Electrons are freed from one element in order to create an electrical potential. Theessential difference between a battery and a fuel cell lies in the ability of the fuel cellto perform the process of dissociation of the chemical components continuously, pro-viding fuel is supplied to the cell. The fuel cell is fed with hydrogen. After the electronshave been removed, the spent hydrogen protons pass through an electrolyte to combinewith oxygen to form pure water, an environmentally acceptable emission. Several typesof electrolyte could be suitable for our application. As the ‘solid polymer’ type has beensuccessfully developed for automotive applications this is the one that is recommended.Knowledge of the precise details of the construction of a fuel cell is not necessary forthis project. For general interest, such details can be found in textbooks, technicalpapers and various automotive companies’ web sites.
The fuel cell requires hydrogen and oxygen. The latter is easily obtained from theambient atmosphere. Hydrogen is more difficult to feed to the system. In its pure form(as a gas or a cryogenic liquid), it would be a very efficient fuel. The problem withhydrogen in this form is the need to store and transport it in pressurised gas tanksor refrigerated liquid containers to reduce tank volume. Both of these options areheavy and bulky, seriously eroding the chemical efficiency of the system. Althoughnot pure, a more convenient source of hydrogen supply for our application would bea hydrocarbon fuel. Methane is the preferred fuel as it is rich in hydrogen and canbe easily reformed. Reformation is the process by which methane is mixed with waterand vaporised to split it, with the assistance of heat and a catalyst, into hydrogen andcarbon dioxide. Any carbon monoxide present in the emissions can be converted tocarbon dioxide using another catalyst. While supplied with the two input gases, the fuelcell process is continuous. It is therefore not suitable for changes in energy demand. Toprovide higher power, for example on take-off and climb or some emergency condition,it would be necessary to supplement the fuel cell energy with a battery. The batterycould be recharged by the fuel cell during low-energy flight periods. This feature maybe less appropriate for a racing aircraft that continually uses full power.
Several components are required for a fuel cell system. These are shown diagram-matically in Figure 6.2, and described below.
The basic principles of the reformer (2) and fuel cell (5) have been described pre-viously. The oxidiser unit (3) is more accurately referred to as the ‘flue gas clean-upunit’. This reduces harmful (NOX) gases. The air compressor (4) provides the supplyof oxygen to the oxidiser unit and fuel cell. The compressor is powered by the bat-tery (8). The battery also provides the start-up energy for the system. This may take upto 15 minutes. It could be provided by an external power source. The internal batterywill be recharged during flight. The DC/DC converter (6) transforms the low-voltageDC supply from the fuel cells to the inverter (7) and then to the electric motor (9). Thecontroller (10) provides the overall system control. This includes:
• the input to the motor (via the converter and inverter),• the start-up sequence,
“chap06” — 2003/3/10 — page 156 — #14
156 Aircraft Design Projects
Methane
Fuel reformer
Fuel cell
DC/DC converter
Battery
Inverter
Motor
Gear box
Water
Reformedgas
oxidiser
Aircompresserand dryer
Fluegas
Propeller
Cont rol
1
2
3
4
10
8
5
6
7
9
H
O
Gas/Liquid
Electric
Fig. 6.2 Electric propulsion system
• the power-down control, system condition monitoring, and• management of the various subsystems.
In a fast moving technology there is a risk that developments not envisaged at the timemay arise to change the details described above. For example, the reformation processto create hydrogen from the methane is heavy and complex. It would be a substantialimprovement if the methane could be fed directly to the fuel cell. Considerable researchwork is currently being conducted to make this possible. If this works, the system willbe changed from that described but the fundamental process will be unaffected. Withsuch thoughts in mind, it is wise to treat the details of the system presented below witha degree of scepticism. The technical details (e.g. weight, volume requirements andefficiencies) have been extrapolated from experimental automotive applications andsuitably modified to include improvements expected in the timescale of the project. Thisprocess is not untypical when designing aircraft that incorporate advanced technologyfeatures.
“chap06” — 2003/3/10 — page 157 — #15
Project study: electric-powered racing aircraft 157
Table 6.2
Component Mass (kg) Dimensions (m)
Fuel cell 28.8 0.44 × 0.20 × 0.13Reformer 53.6 0.61 × 0.31 × 0.31Compressor 4.5 0.08 dia. × 0.04Inverter 5.0 0.24 × 0.24 × 0.12Motor 26.0 0.23 dia. × 0.20Battery 3.0 0.15 × 0.15 × 0.15Total 120.9 112 litres
Table 6.3
Component Mass (kg) Dimensions (m)
Fuel cell 37.5 0.64 × 0.30 × 0.20Compressor 4.5 0.08 dia. × 0.04Inverter 5.0 0.24 × 0.24 × 0.12Motor 26.0 0.23 dia. × 0.20Battery 3.0 0.15 × 0.15 × 0.15Total 76.0 65 litres
Technical specification of the fuel cell system
When assessing the performance of a rapidly advancing technology (e.g. fuel cells), it isnecessary to define the year of adoption. In our project, we will assume 2010. We alsoneed to define the size (power) of the propulsion system. We will assume 75 kW. For themethanol reforming system described above, the projections shown in Table 6.2 weremade.2
This compares to a mass of about 65 kg for a conventional internal combustionengine. The estimated system mass of 121 kg will be used in the design of the aircraft.
In contrast, the potential mass and size of a direct methanol system are also given inthe same reference which quotes the data shown in Table 6.3.
Notice the considerable savings in mass and volume to be gained if such a systemcan be made to work efficiently. However, the system is very speculative and thereforecannot be assumed for our design.
The methanol used in the half-hour duration of the race and warm-up period hasto be added to the system mass. Estimates suggest that 60 kg will be required. Thiswill need a tank size of 0.44 × 0.44 × 0.44 m. Allowing about 13 per cent for propeller,system installation mass and other contingencies brings the propulsion group massto 205 kg.
6.5 Design concepts
At this stage in the development of the project, it is difficult to decide between the tractoror pusher layouts. Both will benefit from the cleaner and more streamlined shape madepossible by the fuel cell system and the electric motor. To avoid this decision it is
“chap06” — 2003/3/10 — page 158 — #16
158 Aircraft Design Projects
Fig. 6.3 Initial concept sketches
proposed to consider both configurations in the initial stages of the project and tocompare them to establish the ‘best’ design.
The tractor layout is associated with the conventional monoplane and tail config-uration. The Nemesis aircraft is the best example of this configuration for a racer.For the pusher layout, either a canard or a tail control position is possible. For thelatter choice either a twin boom or a high propeller position must be selected. To avoidthe complications involved in the design of the rear fuselage to accommodate a tailsurface we will select a canard layout for our aircraft. Initial concept sketches of bothdesigns are shown in Figure 6.3. The appearance of the conventional design resultsfrom the elliptical wing and tail planforms, the tail-dragger landing gear, the mid-fuselage cockpit, and the mid-fuselage wing mounting. It resembles the conventionalaircraft of the 1940s. The canard layout follows the FFT/VariEze configuration but isscaled down to suit the racer requirements. The aircraft has a swept back wing, wingtip fin/rudder control surfaces and a semi-retractable undercarriage (assuming that thiswould be permitted in the Formula E rules).
6.6 Initial sizing
Unlike most aircraft projects, the selection of wing area and engine power is not aproblem as they are part of the Formula rules. The wing area is set at the minimumallowed by the rules (66 sq. ft/6.132 sq. m) and the engine power at 75 kW to the pro-peller is set to match the fuel cell performance. A constraint analysis could be conductedlater in the design process to show the influence of these restrictions on the design.Cockpit size will be kept as small as practical within the formula rules. The shape ofthe canopy will be significantly different on the two aircraft due to the mid-fuselageposition on the conventional and the forward position on the canard layouts.
Although structurally heavier, the ‘canard’ wing will have a higher aspect ratio (7.5)to provide for a greater fin arm. The ‘conventional’ will have a low aspect ratio (5.5) to
“chap06” — 2003/3/10 — page 159 — #17
Project study: electric-powered racing aircraft 159
reduce aircraft span and roll inertia. This will make the wing structure lighter givinga lower aircraft empty mass. Aerodynamically the high aspect ratio of the canard willprovide a lower induced drag but the sweepback will reduce lift capability. Detailedanalysis will be needed to determine the absolute effects of the layout differences.
In order to undertake the necessary performance calculations it is essential to estimatethe aircraft mass (and balance) and the aerodynamic characteristics of the two aircraft.These calculations require an accurate (to scale) drawing of the aircraft (see Figure 6.4for the conventional and Figure 6.5 for the canard layouts). The dimensions from thesedrawings are input into the mass and aerodynamic spreadsheets.
6.6.1 Initial mass estimations
As the aircraft propulsion system does not use conventional fuels, and since the flightduration of the race is short compared with non-racing aircraft, conventional methodsof predicting take-off mass are not appropriate. In this case, it is necessary to make theinitial calculations using known data to validate an acceptable estimation.
Data from the two closest existing aircraft to our designs, namely the Nemesis racerand the FFT Speed Canard, will be used to verify the estimation formula. Adjust-ments will need to be made in the case of the FFT, as this is a two-place touringaircraft.
For our aircraft the mass analysis will be considered in the following breakdown:
Mgross = Mstructure + Mpropulsion + Mfixed equipmt + Mcrew + Mdisposables
In this breakdown, any fuel (methane) will be considered in the ‘disposable’ masssection.
Aircraft dataO/A length 16.2 ft /5 mO/A height 4.0 ft /1.25 mWing span 20 ft /6.1Wing area 66 sq. ft /6.1 sq. mWing AR 6Wing taper 0.3 Wing thickness 13%Prop. diam. 4.0 ft /1.2 m
0 1.5
5
metres
feet
Fig. 6.4 ‘Conventional’ initial general arrangement
“chap06” — 2003/3/10 — page 160 — #18
160 Aircraft Design Projects
Aircraft dataO/A length 12 ft /3.7 mO/A height 4.5 ft /1.4 mWing span 22.5 ft /6.9 mWing area 66 sq. ft /6.1 sq. mWing AR 7.7Wing sweep LE 23°Wing thickness 15%Canard area 9 sq. ft /0.84 sq. m Canard span 10 ft /3 m Prop. diam. 4.0 ft /1.2 m
0 1.5
5
metres
feet
Fig. 6.5 ‘Canard’ initial general arrangement
Table 6.4
Mass Actual Method 1 Method 2 Method 3
Structure 104.7 93.7 90.9 117.7Empty 235.8 237.0 206.1 251.7Gross 340.1 341.3 310.4 356.0% change <1% −8% +5%
Details of the Nemesis aircraft were found in professional journals (e.g. SAWE paper2343, June 1996) and from the Nemesis web site.3 These include a detailed mass state-ment that can be used to compare the predictions from mass formulae, etc. in varioustextbooks. These comparisons are shown in Table 6.4.
Method 1 is a well-publicised general aviation method,4 method 2 is a method used forcivil aircraft design5 and method 3 is mainly associated with military aircraft.6,7 All themethods give reasonable results that generally bracket the actual values. The structuralresults seem to be more variable. A word of caution is appropriate here as the Nemesisis mainly built from composites and the formulae used are based on conventional metalstructures. Method 3 allows the incorporation of ‘technology factors’ that accountedfor composites. Although research shows that composite structures can reduce massby 20 to 30 per cent, it has been found that such improvements are only achievable inlarge aircraft structures due to the connections required to feed loads into and out ofthe shell. Applying factors of 0.85 for the wing, 0.83 for the control surfaces, 0.90 forthe fuselage and 0.95 to the landing gear result in the estimations (per cent MTOM)for three specimen aircraft shown in Table 6.5.
Method 3 appears to underestimate the wing structure and overestimates the fuselageof the Nemesis aircraft. Because of the difficulty of accounting for the wing to fuselage
“chap06” — 2003/3/10 — page 161 — #19
Project study: electric-powered racing aircraft 161
Table 6.5
ComponentNemesisActual
Nemesismethod 3
FFT Canardmethod 3
VariEzemethod 3
Wing 12.3 10.0 12.3 9.4Fuselage 10.0 12.5 5.9 6.9Horiz. tail 1.3 1.1 1.0 1.0Vert. tail 0.4 0.4 2.0 1.0Main u/c 7.1 8.0 7.9 6.8Aux. u/c – 1.0 1.2 1.1
Structure 31.1 33.0 30.3 26.2Engine 29.0 28.9 21.0 17.9Propeller 1.2 1.2 0.8 0.9Fuel system 1.6 0.9 1.8 2.1
Propulsion 31.8 31.0 23.6 20.9Fixed equip. 6.4 6.7 3.7 4.3A/C empty 69.3 70.7 57.6 51.4
Crew 26.7 25.5 28.2 32.4Fuel 4.0 3.8 14.2 16.2
A/C gross 100 100 100 100Gross (kg) 340 356 643 560
joint structure into either the fuselage or wing mass component, uncertainty alwaysexists to the precise division between wing and fuselage components. The above analysisdoes give a reasonable estimate for the combined (wing+ fuselage) structure ratio. Thismay be due to the design of the landing gear (small and with perhaps no brakes) forracing aircraft. The method seems to overestimate the mass of the landing gear. Notethat the Nemesis empty mass fraction is much higher than the other two general aviationaircraft. At about 70 per cent, this is typical for the short range/duration, single-seatracer aircraft.
The analysis above was also done to indicate any variations in mass fractions due tothe canard layout. Although both of these aircraft are much heavier than our proposeddesign, some general conclusions can be drawn. The wing structure for the FFT isseen to be about 2 per cent heavier than the Nemesis. This is largely due to the needto sweep the wing planform back to provide an acceptable fin control arm. For bothof the canard aircraft, the fuselage mass is substantially less than the conventionallayout. This is because the pusher layout shortens the fuselage length. In addition, theengine is mounted close to the wing/fuselage joint making all the heavy loading on thefuselage concentrated in the same area. The control surface mass is slightly higher forthe canard designs.
The propulsion system for the two aircraft layouts will be assumed the same. Thepredicted electric propulsion mass (205 kg) is twice as heavy as a conventional petrolengine.
For all small aircraft, the landing gear represents a substantial weight penalty. Thetricycle arrangement will be slightly heavier than the tail dragger type. The retractionsystem, which is to be used on the canard aircraft, will also add a little extra mass.
The estimated8 mass statement for the two layouts is shown in Table 6.6.The estimating method seems to correctly identify the higher wing mass and lower
body mass for the canard layout compared to the conventional design. As both estim-ates for the maximum take-off mass are so close, the aerodynamic and performanceanalyses that follow will assume that both designs are at 470 kg.
“chap06” — 2003/3/10 — page 162 — #20
162 Aircraft Design Projects
Table 6.6
‘Conventional’ ‘Canard’
Component Mass (kg/lb) %MTOM Mass (kg/lb) %MTOM
Wing structure 68.3 14.7 71.9 15.3Body structure 41.5 8.9 37.3 7.9Control surfaces 9.0 1.9 14.3 3.0Landing gear 32.0 6.9 33.5 7.1Propulsion group 205.0 44.2 205.0 43.6Fixed equipment 18.0 3.9 18.0 3.8Pilot 90.0 19.4 90.0 19.1TO mass (MTOM) 463.8 100 470.0 100
6.6.2 Initial aerodynamic considerations
Aircraft speed is one of the most significant factors in racing. Therefore, the mainaerodynamic analysis for racing aircraft focuses on the reduction of drag.
The layout details of the two aircraft will affect the aerodynamic calculations. Thepusher propeller configuration will reduce the size, and therefore the wetted area, ofthe fuselage. The clean flow conditions over the nose will help to maintain laminar con-ditions over the forward fuselage profile. The smooth contours on the front fuselageand the forward position of the cockpit will allow the windscreen and canopy to beblended into the fuselage profile. This will substantially reduce drag. All these featuresare an advantage for the canard layout. The conventional layout will conversely sufferdrag penalties from the disturbed airflow, from the propeller, over the front fuselage.The mid-mounted cockpit will force the adoption of a bubble canopy. This will be‘draggy’.
The conventional aircraft wing will produce a clean and efficient aerodynamic result;a low drag coefficient and good lift generation. The canard layout will suffer aero-dynamic penalties due to the swept wing planform and the wing tip ‘fins’. The canardsurface will be ‘flying’ therefore producing lift that will help to off-load the main wing.The relatively close coupling of the fins will mean that larger surface areas are necessaryand this will add to the aircraft drag.
Assuming a flight (racing) speed of 200 kt at sea level, the drag of each compon-ent of both aircraft has been calculated using classical aerodynamics. Based on thedescriptions above the following assumptions have been made:
‘Conventional’ wing
• Reference area 6.14 sq. m (66 sq. ft).• A modern, high-performance, general aviation wing section.• Average thickness 14 per cent.• No twist.• Aspect ratio 6.0.• Taper ratio 0.5.• Smooth surface.• 50 per cent chord laminar flow.• Oswald efficiency factor 0.9 due to the elliptical planform.• Wetted area twice the wing ref. area as the small penetration into the fuselage will
compensate for the wing section curvature.
“chap06” — 2003/3/10 — page 163 — #21
Project study: electric-powered racing aircraft 163
‘Canard’ wing
• Reference area 6.14 sq. m (66 sq. ft).• A modern, high-performance, general aviation wing section.• Average thickness 16 per cent.• 23◦ sweep at quarter chord.• Aspect ratio 7.7.• Smooth surface.• 25 per cent laminar flow due to the effect of spanwise drift caused by the sweep and
the 30 per cent max. thickness of the wing section.• Oswald efficiency 0.8 due to the wing tip/fin interference.
‘Conventional’ fuselage
• The complex fuselage profiles increase wetted area and interference factors.• Tractor propeller position makes fuselage flow totally turbulent.• Mid-fuselage wing position reduces interference factor.• Fuselage wetted area 6.9 sq. m (74 sq. ft) with planform area 2.54 sq. m (27.3 sq. ft).• Equivalent fuselage diameter 0.7 m (28 in).• No fuselage base drag as rudder extends below the fin.• Canopy drag effects calculated separately.
‘Canard’ fuselage
• Smooth profile.• No base drag.• 10 per cent laminar flow assumed (this is considered as conservative).• Wetted area 4.5 sq. m (48 sq. ft).• Width of fuselage 0.64 m (26 in), depth 0.88 m (35 in).• Wing/fuselage interference factor 1.07.• Turbulent flat plate coefficient 0.00245.• Zero-lift drag reduced by 7 per cent due to the pusher propeller position.
‘Conventional’ empennage
• Thickness 10 per cent throughout.• Wetted areas 1.52 sq. m (16.3 sq. ft) horizontal, 0.69 sq. m (7.4 sq. ft) vertical.• Fin sweep at quarter chord 27◦.
‘Canard’ control surfaces
• Wetted areas 1.68 sq. m (18 sq. ft) horizontal, 0.88 sq. m (9.5 sq. ft) vertical (total).• Thickness 18 per cent horizontal, 15 per cent vertical.• Flat plate skin friction coefficient 0.00375.
Canopy (both aircraft)
• ‘Conventional’ wetted area 0.057 sq. m (0.62 sq. ft).• ‘Canard’ blended profile therefore no extra drag.
Trim (both aircraft)
• A value of about 6 per cent of total drag is common but as the flight duration isshort and the aircraft can be pre-race adjusted for trim reduction, no extra drag isassumed in the race condition.
“chap06” — 2003/3/10 — page 164 — #22
164 Aircraft Design Projects
Table 6.7 summarises the detailed drag calculations:8
Table 6.7
Component Parasite Induced Total Per cent
‘Conventional’Wing 0.00463 0.00052 0.00515 26.8Body 0.00564 0.00007 0.00571 29.8Controls 0.00177 0.00006 0.00183 9.5Canopy 0.00284 zero 0.00284 14.8L/gear 0.00366 zero 0.00366 19.1Total 0.01854 0.01919 100.0
‘Canard ’Wing 0.00694 0.00038 0.00732 38.4Body 0.00359 zero 0.00357 18.8Controls 0.00425 0.00002 0.00427 22.3Canopy zero zero zero zeroL/gear 0.00391 zero 0.00391 20.5Total 0.01869 0.01907 100.0
For this class of aircraft, interference drag will be kept low in the racing trim.A contribution has been added to each of the component drag calculations shownin Table 6.7.
The effect of the tractor propeller on the fuselage skin-friction drag can clearly beseen by the fact that this is the largest drag component on the conventional aircraft.(It has been reported elsewhere that a 7 per cent increase can be expected.) Also, theinfluence of the blister canopy on drag is seen to add about a further 10 per cent tothe total drag. For the canard design, drag is seen to be predominantly affected by thewing and the large contribution from the control surfaces (wing tip fins and canard).As expected, on both configurations the landing gear represents a substantial dragpenalty (about 20 per cent in both cases). Fairing the main gear would seem to be asensible option for these aircraft.
It is interesting to note that the predicted drag of both aircraft is approximately thesame. This confirms the view that a choice of the preferred configuration cannot bemade on the basis of aerodynamic and mass efficiency (a view borne out by the factthat both types of aircraft are currently used in formula racing).
For both wing layouts, a non-flapped lift coefficient of 1.0 can be assumed. Theswept wing of the canard design will suffer a reduction in lift generation but the canardsurface will contribute to the overall lift and so reduce this disadvantage. A simple stallspeed calculation can now be done:
Stall speed = [aircraft weight/(0.5ρSCLmax)]0.5
Stall speed = [470 × 9.81/(0.5 × 1.225 × 6.14 × 1.0)]0.5
Stall speed = 35 m/s (64 kt)
This is regarded as a little too high for a light aircraft. Either the wing area needsto be increased (this will increase aircraft drag) or a flap will be required. To reducethe stall speed to 60 kt (making the approach speed 1.3 × 60 = 78 kt) will demanda lift coefficient of 1.29. This could be easily achieved with a simple plain or splitflap. Careful detail design of the wing trailing edge and flap hinges, to minimise dragincreases, should be possible.
“chap06” — 2003/3/10 — page 165 — #23
Project study: electric-powered racing aircraft 165
As the aircraft will be pulling g in the tight turns, it is necessary to determine the stallspeeds in relation to the load factor (n). Using the equation above the following resultsare obtained:
Load factor (n) Aircraft stall speed (m/s)2 49.53 60.64 70.0
6.6.3 Propeller analysis
For light aircraft, propeller performance is the most difficult parameter to accuratelyassess. The diameter of the propeller must be limited to avoid sonic flow over the tips.This would generate noise and be aerodynamically inefficient. Most prototype lightaircraft have to be refitted with a different propeller after the initial flights because it isvirtually impossible to predict accurately the aircraft drag and thrust values. A fixed-pitch propeller produces its best performance at a specific combination of aircraftforward speed and engine rotational speed. The lower the number of blades, the betteras the preceding blade disturbs the airflow for the following blade. One blade wouldbe aerodynamically best but . . . ! The formula rules dictate the use of a propeller withfixed pitch. This creates a problem for racing aircraft, as a fine pitch propeller will bemost efficient at low aircraft forward speeds and a coarse pitch at high speed. In a race,it is important to have good take-off performance in order to achieve a good positionat the first turn on the circuit. Being ahead of the field allows the pilot to choose hisracing line (height and position) and avoids flying in the turbulent airstream from otheraircraft. A clear view with a preferred racing line is a significant advantage. However,the take-off and early race represents only a small proportion of the total competition.As airspeed builds up during the race, a fine pitch propeller will be a serious handicap.Aircraft with a coarse pitch propeller with the same engine will fly faster and mayeventually overtake the early leaders.
The choice of propeller size (diameter) may be dictated by the geometric constraintsof the layout. If the diameter is too large the landing gear will need to be longer andthe aircraft ground clearance high. This will make it more difficult to climb into thecockpit and the increased height of the aircraft centre of gravity above the groundmay make ground manoeuvring over rough ground unstable. If the diameter is small,the inefficient hub area will form a larger proportion of the total disc area reducingthe propeller overall efficiency. To make towing easier, a two-blade layout is best. Theblades can be stopped in a horizontal position, parallel to the road.
For the electric propulsion system, the electric motor speed can be varied to bettermatch the propeller requirements. This is not as easy to achieve with a conventionalinternal combustion engine. This feature is potentially very useful and should beinvestigated in more detail in later stages of the project (after the preliminary designphase has been completed). For example, it may be possible to adopt a higher motorspeed for the take-off phase than used in the race condition. This would effectivelyproduce a thrust boost for take-off if the propeller geometry/performance can accountfor such a change.
For initial design considerations, typical propeller details would be:
• Tip diameter 1.2 m.• Spinner diameter 0.24 m.• Rotational speed 2000 rpm (racing), 3000 rpm (take-off ).
“chap06” — 2003/3/10 — page 166 — #24
166 Aircraft Design Projects
• Advance ratio 2.0 (racing), 0.5 (take-off ).• Efficiency 82 per cent max.
6.7 Initial performance estimation
For racing aircraft, performance is the key issue in the design. As there is little differencein mass, drag and thrust between the two proposed configurations, their performancewill be similar. At this initial stage, it will not be possible to distinguish between the twoaircraft and identify the best design. It may be necessary to build, test and then raceboth types to decide which is the best! Very small differences in performance are alwaysto be expected between competitive racing aircraft. Pilot ability will be exaggerated andthe best flyers will be successful.
Notwithstanding the above comments, it is necessary to determine the overall per-formance to establish the viability of the aircraft and the new racing formula. Thefollowing estimates are required:
• maximum level speed,• climb performance,• turn performance,• field performance.
6.7.1 Maximum level speed
As the drag and propeller parameter estimates are made with several crude assumptions(e.g. extent of laminar flow over the surfaces), and as the aircraft profile and induceddrag coefficients are similar for both aircraft, an average between the two aircrafttypes will be used. To reflect the variability in the estimation of the coefficients, a+/−5 per cent range will be applied to show the sensitivity of optimistic and pessimisticestimates. We will also apply the same variability to the propeller efficiency.
The values used in the analysis are shown in Table 6.8.Two curves (fine and coarse pitch) for propeller efficiency against aircraft forward
speed are shown in Figure 6.6. Aircraft drag and thrust curves are shown in Figure 6.7.The effect of propeller pitch selection on aircraft performance is clearly seen in thisgraph. The extra thrust provided at low speed by the fine pitch propeller is eroded asspeed increases. The aircraft maximum level speed is seen to be 96 and 102 m/s forthe fine and coarse propellers respectively. The +/−5 per cent variation shown aboveresults in a +/−2 to 3 m/s change in maximum speed. Although seemingly not verymuch this change would result in either a ‘dog’ or a ‘pearl’ of a racing aircraft. Thisconfirms the essential requirement to get the aircraft parameter estimation as accurateas possible in the early stages.
Table 6.8
Pessimistic Mean Optimistic
Profile drag coeff. 0.0195 0.0186 0.0177Induced drag factor 0.0488 0.0465 0.0442Aircraft gross mass 470 kg 470 kg 470 kgProp. efficiency 0.78 0.82 0.86
“chap06” — 2003/3/10 — page 167 — #25
Project study: electric-powered racing aircraft 167
30 40 50 60 70 80 90 100Aircraft speed (m/s)
Eff
icie
ncy
Course pitch propeller
Fine pitch propeller
Optimistic
Pessimistic
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Fig. 6.6 Propeller efficiency versus aircraft forward speed
Aircraft speed (m/s)
Dra
g an
d th
rust
(N)
Aircraft drag
Course pitch propeller
Fine pitch propeller
Optimistic thrust
Pessimistic thrust
30 40 50 60 70 80 90 100 1100
500
1000
1500
2000
2500
Fig. 6.7 Drag and thrust versus aircraft forward speed
The difference between the thrust and drag curves shows the energy available foraircraft manoeuvre. For the sea level, straight and level, flight performance the (thrust–drag) versus aircraft forward speed is shown in Figure 6.8.
Dividing the aircraft drag at a given speed into lift (=Mg for straight and level flight)gives the aircraft lift to drag ratio (L/D) variation. Figure 6.9 shows the L/D ratio withspeed. For economical flight it is necessary to fly at the speed close to maximum L/D.For our aircraft, this speed is very slow due to the very low drag characteristics but fueleconomy in racing aircraft does not have a high priority.
All the above calculations have assumed that the aircraft is not manoeuvring(i.e. structural load factor (n) = 1.0). Pulling extra ‘g’ will increase the lift on the
“chap06” — 2003/3/10 — page 168 — #26
168 Aircraft Design Projects
30 50 70 90 110
Aircraft speed (m /s)
(T–D
) new
tons
Fine pitch propeller
Course pitch propeller
–500
500
1000
1500
2000
2500
Fig. 6.8 (Thrust–drag) versus aircraft forward speed
Lift coefficient
L/D
rat
io
Aircraft speed (m/s)
30
40
50
60
7080
90100110
0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.400
2
4
6
8
10
12
14
16
18
Fig. 6.9 Lift/Drag ratio versus lift coefficient
wings and thereby increase induced drag. Figure 6.10 illustrates the change of aircraftdrag with manoeuvring load factor. Notice how the minimum drag speed progressivelyincreases with load factor. The pilot will not want to fly the aircraft ‘on the back sideof the drag curve’ as this results in unstable and difficult handling and will prefer toonly pull ‘g’ at higher speeds. The extra drag will slow the aircraft. The relationshipbetween aircraft speed and manoeuvre is considered further under the climb and turnperformance below.
“chap06” — 2003/3/10 — page 169 — #27
Project study: electric-powered racing aircraft 169
20 30 40 50 60 70 80 90 100 110
Aircraft speed (m/s)
Airc
raft
dra
g (N
)
Load factor = 4
3
2
1
500
1000
1500
2000
2500
Fig. 6.10 Drag and load factor versus aircraft forward speed
30 50 70 90 110Aircraft speed (m/s)
(T–D
) new
tons
Load factor n = 1
2
3
4
123
4
Course pitch prop.
Fine pitch prop.
Intersection = max. speeds–500
500
1000
1500
2000
2500
Fig. 6.11 (T–D) and load factor versus aircraft forward speed
6.7.2 Climb performance
As mentioned above, the difference between the thrust and drag curves, at a specificspeed, represents energy that is available for the pilot to either accelerate (kinetic energyincrease) or climb (potential energy increase) the aircraft. The excess force available(thrust–drag) at various aircraft speed, and with the aircraft pulling ‘g’, is shown onFigure 6.11. This figure also shows the advantage of fine pitch at low speed and coarsepitch at high speed. Using all the available extra energy to gain height provides themaximum rate of climb. Multiplying (T – D) by aircraft speed and dividing by aircraft
“chap06” — 2003/3/10 — page 170 — #28
170 Aircraft Design Projects
weight gives the max. climb performance of the aircraft at constant aircraft forwardspeed (i.e. with zero acceleration).
The term [V (T −D)/W ] is referred to as the specific excess power (SEP). At sea levelthe maximum rate of climb versus aircraft speed is shown in Figure 6.12. Drag increasein manoeuvring flight, as mentioned above, has a significant effect on the aircraft SEP.Figures 6.13 and 6.14 illustrate the effect of choice of propeller pitch.
–30.0
–20.0
–10.0
0.0
10.0
20.0
30.0
40.0
30 40 50 60 70 80 90 100 110Aircraft speed (m /s)
RoC
=V
(T–D
)/Mg
Fine pitch
Course pitch
Max. speed
Max. speed
Fig. 6.12 Rate of climb versus aircraft forward speed
20 30 40 50 60 70 80 90 100Aircraft speed (m /s)
V (T
–D)/M
g
Max. speed
Load factor n = 4
1
23
–30.0
–20.0
–10.0
0.0
10.0
20.0
30.0
40.0
Fig. 6.13 Specific excess power (SEP) versus aircraft forward speed (fine pitch)
“chap06” — 2003/3/10 — page 171 — #29
Project study: electric-powered racing aircraft 171
30 40 50 60 70 80 90 100 110
Aircraft speed (m/s)
Load factor n = 1
23
4
Max. speeds
–15.0
–10.0
–5.0
0.0
5.0
10.0
15.0
20.0
25.0
30.0
V (T
–D
)/Mg
Fig. 6.14 Specific excess power (SEP) versus aircraft forward speed (coarse pitch)
6.7.3 Turn performance
Racing aircraft fly an oval circuit; it is therefore necessary to investigate the aircraftturn performance in some detail to establish the optimum racing line. Good turningperformance will allow the aircraft to fly a tighter turn and therefore cover less distancein the race. The pilot faces a dilemma. Pulling a tight turn will increase drag andtherefore reduce aircraft forward speed. This loss of speed will have to be made upalong the straights. Alternatively, flying gentle (larger radius) turns will maintain speedbut extend the race distance. Figure 6.15 shows the basic relationship between aircraftforward speed, manoeuvring load factor (n) and aircraft turn rate. Tight turns (high ‘g’)are achieved at low speeds. Race pilots do not like high ‘g’ and slow speed. They liketo fly fast and gentle.
To achieve a balance of forces on the aircraft in a turn, it is necessary to bank theaircraft. The angle of bank is related to the aircraft load factor as shown in Figure 6.16.Although the loads on the aircraft in a correctly banked turn are balanced, it is necessaryto instigate the turn from a straight and level condition and then to return to it. Theapplication of the control forces required to change these flight conditions createsextra drag. To avoid these complications, a race could be flown in a fully balanced andconstant attitude if a circular, or near circular, path outside of the pylon was selected.This would result in a much longer flight distance that would penalise the pilot unlessa higher average race speed could be achieved to offset this disadvantage. The beststrategy to adopt for the race is not obvious. Here lies the essence of good racingtechnique.
Not all of the aircraft parameters can be considered in the performance analysis. Forexample, sighting and aligning the pylons is an important element in successful racing.The mid-fuselage cockpit position of the conventional layout may be regarded as lesseffective than the forward position on the canard. Also, the canard control surfacemay offer the pilot a reference line to judge his position more accurately. ‘Cutting apylon’ carries a substantial time penalty but flying a line that is too wide may present anopponent with a passing opportunity. These are features that are difficult to assess in the
“chap06” — 2003/3/10 — page 172 — #30
172 Aircraft Design Projects
30 40 50 60 70 80 90 100 110Aircraft speed (m /s)
Turn
rat
e (°
/s)
n = 2
n = 3
Load factor n = 4
Stall boundary
10
20
30
40
50
60
70
80
Max. speed limits(depending on aerodynamic dragand prop. efficiency)
Fig. 6.15 Turn performance
1.5 2.0 2.5 3.0 3.5 4.0Aircraft load factor (g )
Ban
k an
gle
(°)
40
45
50
55
60
65
70
75
80
Fig. 6.16 Aircraft bank angle (balance turn) versus load factor
initial design stage. The combination of turn performance and flight path strategy offersa good example of the application of computer flight simulation in the early designstages. In this way, it is possible to test the external (visual) and internal (handling)features of the aircraft in a synthetic racing environment. Unfortunately, the initialaerodynamic, mass, propulsion and performance predictions do not hold sufficientfidelity to make accurate judgements from such simulations. However, some crudeassessments are possible.
“chap06” — 2003/3/10 — page 173 — #31
Project study: electric-powered racing aircraft 173
6.7.4 Field performance
As described previously, Formula racing starts with a grid of eight aircraft that havewon the previous heats. The pole positions are awarded to the fastest aircraft in previousraces. Take-off performance is therefore a significant aspect of the race. Obviously, thereis an advantage to the first aircraft to reach the scatter pylon and avoid the congestionof other competitors. As mentioned in the propeller section, the designers must make adifficult choice between compromising race speed for take-off advantage, or vice versa.Short take-off performance and initial climb ability demands good lift generation atlow speed. This implies a thick wing section profile, a cambered chord line, a low wingloading, efficient flaps and a fine pitch propeller. Conversely, maximum race speed willbe achieved with high wing load, thin unflapped wing section and a coarse pitch prop.This is a difficult choice for the designers that will involve compromises to be made. Ofall the parameters mentioned, the propeller selection is the easiest to change after theaircraft is built. In the early stages of the design all that can be done is to analyse theaircraft in a generalised method.
Estimation of field performance comprises both take-off and landing manoeuvres.In race conditions, the aircraft will not follow generalised procedures. For example,a racing pilot may hold the aircraft down in ground effect to build up energy beforestarting the climb. Disregarding such aspects, we will analyse the field performanceusing established design methods. Using average values for the aerodynamic coeffi-cients, a sectional max. lift coefficient of 1.0, simple landing flaps, and aircraft gross(race) mass gives:
Take-off to 50 ft at 1.2 Vstall (with max. lift coeff . = 1.0)Ground run = 340 m (1114 ft)Climb to 50 ft = 136 m (446 ft)Total take-off distance = 476 m (1560 ft)
Landing from 50 ft at 1.3 Vstall (with flapped max. lift coeff . = 1.3)Approach distance = 406 m (1330 ft)Ground distance = 117 m (384 ft)Total landing distance = 523 m (1714 ft)
These values appear to be acceptable for this type of aircraft.
6.8 Study review
Design of racing aircraft is different to most design projects in that the main object-ive is simply to win competitive races. As these are set in a highly controlled designand operational environment, the design process is made easier. For the designer, theFormula rules and the racing conditions provide a very narrow focus to the selectionof the design criteria and a simplification of technical decisions. Some of the normaldesign procedures (e.g. constraint analysis and overall operational trade-off studies)are not appropriate. The ‘rules’ set the wing area, engine type and power so the maindesign drivers become:
• reduction of aircraft mass (down to the specified minimum allowed by the rules),• making the configuration aerodynamically efficient (reducing drag and generating
lift),
“chap06” — 2003/3/10 — page 174 — #32
174 Aircraft Design Projects
• selecting a propeller geometry that is ‘matched’ to the race requirements,• ensuring that the aircraft is easy to fly in the competitive racing environment,• ensuring that the aircraft is reliable and serviceable at the race location,• enabling the aircraft to be transported to the racecourse and easily reassembled.
Many of the detailed developments involved in the above will only be possible duringthe racing season. The ‘fine tuning’ of the aircraft is an established feature of a successfulrace team. Such late changes to the aircraft arise because it is not possible to model theaircraft using the analytical methods that are available in the design stages. Races arewon by very small margins in aircraft performance between aircraft. These differencesare much smaller than the accuracy of our design calculations. All that can be done inthe design stages is to provide the best starting point for the race development process.
This illustrates a tenet of aircraft design:
Analytical methods will only provide a starting point for the aircraft design whichwill subsequently only be improved by detailed design, empirical trimming andflight test work.
However, this should not be used as an excuse to avoid quality in the preliminary designphase, as subsequent improvements will not overcome inherent weaknesses in the basicdesign.
This project has provided a good example of the strengths and limitations of theconceptual design process. It should serve as a reminder that good design relies onexcellence in each phase of the total design and development process. Ineptitude in anyof the parts of the design work will only produce a poor quality aircraft.
References
1 Formula 1 web site (www.if1airracing.com/Rules).2 Warner, F., ‘An investigation into the application of fuel cell propulsion for light aircraft’,
Final-year project study, Loughborough University, May 2001.3 Nemesis web site (www.nemesisnxt.com).4 Stinton, D., The Design of the Aeroplane, Blackwell Science Ltd, 2001, ISBN 0-632-05401-8.5 Jenkinson, L. R. et al., Civil Jet Aircraft Design, AIAA Education Series and Butterworth-
Heinemann Academic Press, 1999, ISBN1-56347-350-X and 0-340-74152-X.6 Raymer, D., Aircraft Design – A Conceptual Approach, AIAA Education Series, ISBN 1-
56347-281-0, third edn, 1999.7 Brant, S. A. et al., Introduction to Aeronautics: A Design Perspective, AIAA Education Series,
1997, ISBN 1-56347-250-3.8 Tully, C., ‘Aircraft conceptual design workbooks’, Final-year project study, Loughborough
University, May 2001.
“chap07” — 2003/3/10 — page 175 — #1
7
Project study: a dual-mode
(road/air) vehicle
Taylor Aerocar
Convair Aircar (prototype)
Existing and proposed roadable aircraft
“chap07” — 2003/3/10 — page 176 — #2
176 Aircraft Design Projects
7.1 Introduction
‘Flying car’, ‘roadable aircraft’, ‘dual-mode vehicle’ and other terms are used todescribe the all-purpose vehicle that can fly like an airplane and drive on the highwaylike an automobile. Make it amphibious and we have the perfect all-purpose vehicle!Nevertheless, this might be taking our ideas a bit too far.
It has long been the dream of aviation and automobile enthusiasts to have a vehiclethat will bring them the best of both worlds. Many drivers stuck in rush hour traffichave fantasies about being able to push a button and watch their car’s wings unfurlas they lift above the stalled cars in front of them. Just as many pilots who have beengrounded at an airport far from home by inclement weather have wished for someway to wheel their airplane out onto the highway and drive home. This yearning hasresulted in many designs for roadable aircraft since as early as 1906.1
A designer of a flying car will encounter many obstacles, including conflicting regula-tions for aircraft and automobiles. As an automobile, such a vehicle must be able to fitwithin the width of a lane of traffic and pass under highway overpasses. It must be ableto keep up with normal highway traffic and meet all safety regulations. It must alsosatisfy vehicle exhaust emission standards for automobiles. (Note: these regulationsare easier to meet if the vehicle could be officially classed as a motorcycle.) Therefore,the wings must be able to fold (or retract) and the tail or canard surfaces may have to bestowable. The emission standards and crashworthiness requirements will add weightto the design. The need for an engine/transmission system that can operate in the stopand go, accelerate and decelerate environment of the automobile will also add systemcomplications and weight.
For flight, the roadable aircraft must be lightweight and easy to fly. It must havea speed range at least comparable to existing general aviation airplanes. Conversionfrom aircraft to car or vice versa must be doable by a single person and the engine mustbe able to operate using either aviation fuel or auto fuel. Ground propulsion must bethrough the wheels and not via propeller or jet which would present a danger to nearbypeople, animals or other vehicles.
7.2 Project brief (flying car or roadable aircraft?)
While some people use the above terms interchangeably, or use the latter term to bypassthe science fiction connotations of the former, they are explicitly two quite differentconcepts. One wishing to design such vehicles must first decide which approach isappropriate. The ‘flying car’ is primarily a car in which the driver has the option oftaking to the air when desired or necessary. The ‘roadable aircraft’ is an airplane thatalso happens to be capable of operation on the highway.
In the past, most designs1 have actually been for roadable aircraft. They started outlooking like conventional airplanes but with wings and possibly with tails that couldbe retracted or folded. Alternatively, they may be removed and towed in a trailer whenthe vehicle is operated on the road. Several such vehicles have been designed and built.A few, such as the Taylor Aerocar1 or the Fulton Airphibian,2 have been certified foruse in flight and on the highway. Both types of vehicle have been sold to the public. Theroadable aircraft is meant to be primarily an airplane but with the capability of beingdriven on roads to and from the airport. It must also be capable of getting the pilot andpassengers to their desired destination on the highway when the weather prevents flight.As such, it is a vehicle primarily sold to licensed pilots. They would use its on-roadcapabilities in a limited manner, and not as a substitute for the family automobile for
“chap07” — 2003/3/10 — page 177 — #3
Project study: a dual-mode (road/air) vehicle 177
everyday trips to the supermarket. Typical problems with such designs have been theirpoor performance both in the air and on the road. Also, there has been in the past areluctance of insurance companies to write policies which will cover their operation inboth environments.
The ‘flying car’, unlike the roadable aircraft, has proved to be more of a fantasy thanan achievable reality. A key element in the development of a successful flying car isdesigning a control system that will enable a ‘driver’ who may not be a trained pilot tooperate the vehicle in either mode of travel. This virtually necessitates a ‘category IIIcapable’ automated control system for the vehicle. This must provide a ‘departure-to-destination’ flight control, navigation and communication environment. Many expertsfeel that such a design is possible today, but only at high cost. Ideally, if the ‘flyingcar’ is to become the family car, it must have a price that is at least comparable to aluxury automobile (preferably less than 25 percent of the cost of the cheapest currentfour passenger general aviation aircraft).
Both the flying car and the roadable aircraft concepts usually assume a self-containedsystem capable of simple manual or even automated conversion between the car andairplane modes. A third choice is the dual-mode design which is capable of operationon the road or in the air but does not necessarily carry all the hardware needed for bothmodes with it at all times. One such vehicle was the Convair/Stinson CV-118 Aircar.2
Designed in the 1940s, it combined a very modern looking fiberglass body car with awing/tail/engine structure that could be attached to the roof of the car for flight. Thisdesign successfully flew, and operated well on the highway, but was a victim of highcost and changing corporate goals for its manufacturer.
Another decision facing the designer of any airplane/automobile hybrid vehicle iswhether to attempt to meet government standards for both types of vehicles. Unlessone wishes to go to the extreme of developing a very light weight flying motorbike whichwill operate under ultra-light regulations, one must meet FAR or JAR requirementsfor general aviation category aircraft. On the other hand, there is a choice when oneconsiders the automotive aspects of the design.
Automobile safety and emission control requirements necessitate structural andengine designs that are heavier than one would ordinarily need for an aircraft. Thereis, at least under United States law, a ‘loophole’ in the regulations under which anyroadable vehicle with fewer than four wheels can be classified as a motorcycle andnot an automobile. This allows those who wish to avoid the extra weight and expenseof meeting automobile design standards to develop a three-wheeled vehicle and clas-sify the resulting design as a flying motorcycle, a vehicle that officially is an airplanein the air and a motorcycle on the road. Motorcycles have very few safety or emis-sion design requirements beyond the specification of lighting, horn and engine muffler.Three-wheeled road vehicles do have operational speed restrictions in the United States.
Another decision that must be made is the extent to which the vehicle will meet the‘luxury’ standards of automobile buyers that are not normally seen in general aviationaircraft. A typical modern American automobile lists in its ‘standard’ equipment pack-age air-conditioning, electric window controls and door locks, automatic transmission,CD/tape players and similar items. None of these are usually found in most generalaviation aircraft and all add (sometimes considerable) weight to the aircraft.
7.3 Initial design considerations
This design was developed by a single team of students from two universities in theUnited States and in Britain to satisfy the requirements for an aircraft design class.
“chap07” — 2003/3/10 — page 178 — #4
178 Aircraft Design Projects
The final design was to be entered in an American design competition sponsored byNASA and the FAA. As such, there were no initial customer requirements other thanthe above-mentioned regulations for the design of aircraft and automobiles in both theUS and the EU. The student team had to decide which of the above design approachesto take and had to determine their own specifications for things like range, endurance,rates of climb, and cruise (on land and in the air) speed. For this study, the designersselected a ‘roadable aircraft’. This is defined as a vehicle which is primarily meant forair travel but which, when pressed into duty in its automobile mode, will be able to fullymeet the requirements for travel on high-speed motorways as well as city streets. It wasdesigned to meet all EU and US requirements for both automobiles and aircraft. Theinitial assumptions were that the vehicle, as an aircraft, had to match the performance ofcurrent four-place, piston-powered, general aviation. As a car, it must have performancesimilar to a family sedan type of vehicle.
The general goals agreed upon at the start of the design process were for an aircraftwith a cruise speed of 150 knots and a range of between 750 and 1000 nautical miles(1388 to 1850 km) at a cruise altitude of about 10 000 ft (3048 m). It must be able totake off and land in less than 2000 ft (610 m) and carry four people. As an automobile,it must be able to cruise at 70 mph (113 km/hour), have a reasonable acceleration capa-bility, a range at highway speed of at least 300 miles (482 km), and handling qualitiescomparable to a family sedan. In addition, the design had to meet all FAR (JAR)regulations for airworthiness and meet both American and EU requirements for auto-mobiles. There was considerable discussion about opting for a three-wheel design inorder to eliminate many of the automotive design constraints but this was rejected. Theteam accepted the challenge of meeting US and EU automobile safety and emissionrequirements in order to have a vehicle that would handle like a car on the highway.
Additional challenges noted by the team at the beginning of the project included:
• the need to have acceptable in-flight wing aerodynamics while being able to retract,fold, or detach and stow the wing for road travel,
• the need to ‘rotate’ on take-off,• the need to find an engine/transmission combination which could meet the conflicting
demands of ground and air travel,• the need for dual-mode control systems, and the need to meet rigorous stability and
performance requirements in both modes of travel.
The design of a satisfactory wing is a dominant part of any roadable aircraft layout. Asa ‘car’ the vehicle must fit into standard roadway widths. The resulting vehicle footprint(aspect ratio) is less than unity. This is regarded as inefficient for an aircraft wing plan-form. A wing of reasonable aspect ratio must then be capable of being extended fromthe body (fuselage) for flight and somehow stowed for highway use. There are manyways to do this including folding wings, rotating wings, telescoping wings, and detach-able wings. These could be stored in, under, or over the car configuration. Alternatively,they could be towed behind the car.1 All such designs impose structural compromiseand weight penalties. The use of the wing for a fuel tank location would also be ruledout.
The take-off problem reflects the differing stability requirements of automobiles andairplanes. Most modern aircraft are designed with a tricycle landing gear arrangementwith the rear or main wheels placed only slightly behind the center of mass (center ofgravity). This allows easy rotation in pitch to a reasonable take-off angle of attack afterground acceleration. Placement of the rear wheels in the optimum location for the maingear of an aircraft would result in a very unstable car. It would have a tendency for its
“chap07” — 2003/3/10 — page 179 — #5
Project study: a dual-mode (road/air) vehicle 179
front wheels to lift off the road at highway cruise speeds near the desired take-off speedsfor the aircraft. Cars are designed to minimize the likelihood of the wheels lifting offthe road at highway speeds! Some roadable aircraft designs have attempted to solve thisproblem by having a conventional aircraft tail section that is removed for road travel.This effectively moves the center of mass further forward between the front and rearwheels. Others have employed a car type suspension with wheels or axles that can beextended or retracted to give the needed angle of attack for take-off.
Further complications arise due to the need for the wing on the airplane to developsome lift during the take-off run while the automobile must produce as little lift aspossible at highway cruise speed. Removing or retracting the wings for the car layoutwill obviously solve most of the highway lift problem.
Aircraft piston engines are designed to be run at constant rpm for long periods oftime. Automobile engines are designed to operate over a wide range of rpm and arecoupled to a transmission to make possible combinations of torque and power suitablefor a variety of operational needs. Aircraft engines must also be capable of efficientoperation over a wider range of altitude than car engines. Air-cooling is normally usedwith aircraft engines while water-cooling is usually used for automobile engines. Botha water-cooling system and a transmission system will add extra weight not commonin most aircraft designs. Some flying car designs have proposed using separate enginestailored to each mode of travel. This is on the assumption that two optimized enginesmay not weigh much more than a single dual-mode engine and drive train, and thatthe improved efficiencies may allow lower fuel consumption. Other designers havesuggested the use of an engine and transaxle from a small 4WD automobile with thedrive for one set of car wheels attached to the wheels and the other to the propeller.
The extent to which the controls for flight and ground operation can be merged isalso a design concern. Do the in-flight rudder pedals become the accelerator and brakepedals on the road? Does the car steering wheel, with a release to allow it to movetoward and away from the driver/pilot, become the in-flight control yoke, or can a‘stick’ replace the wheel and be used in both modes of travel? Moreover, how are thesecontrols coupled to the rest of the vehicle? Can a fly/drive-by-wire system work in bothmodes or must the controls be mated to two separate mechanical or hydraulic systems?
Finally, there is the question of ‘roadability’. Beyond the question of tip-over angles(or ground loops during taxi, take-off, and landing), this is an issue that does notnormally face the aircraft designer. The vehicle’s wheel placement and suspension sys-tem and even the choice of tires must take into account the need for comfortable,stable handling on the highway as well as be able to absorb the sudden shock oflanding.
7.4 Design concepts and options
Given the above challenges, constraints, and goals, the design process employed inthis study was different from the usual systematic approach to aircraft design. It wasnot possible to generate a specific set of aircraft performance targets from studies of‘comparable’ aircraft designs, or to initially ‘size’ the vehicle and refine it by the use ofconstraint diagrams and plots.
The design process began with each member of the team proposing an initial concept.The perceived merits of each concept were evaluated and compared and three generalconfigurations from each of the two collaborating universities were brought to the table
“chap07” — 2003/3/10 — page 180 — #6
180 Aircraft Design Projects
+
(a)
(b)
Fig. 7.1 Three initial concept sketches
at the first formal meeting of the complete team. Figure 7.1 shows sketches of three ofthese ‘intermediate’ concepts:
• a gyrocopter,• a lifting body design with telescoping wings, and• a car with ducted fans and folding wings.
“chap07” — 2003/3/10 — page 181 — #7
Project study: a dual-mode (road/air) vehicle 181
(c)
Fig. 7.1 Continued
The team then developed a ‘decision matrix’ with which to evaluate these sixproposals. The decision matrix included assessments of the following features:
• the structural design,• performance, and control aspects,• propulsion system(s),• ‘roadability’,• cost,• complexity of manufacturing, and• ergonomics and human factors considerations.
They then divided themselves into six smaller groups, each rating all six concepts basedon one of the above criteria. These ratings were subjective in nature since none of thedesigns had been developed beyond the stage of an initial sketch and concept. Theresulting matrix is shown in Table 7.1.
Based on this matrix analysis, a decision was made to merge some elements from thesecond and fifth of the preliminary concepts. This resulted in a design with a liftingfuselage, a dual ducted fan propulsion system, and retractable (telescoping) wings asillustrated in the sketch in Figure 7.2.
The design employed a conventional four-place seating arrangement. The ductedfans were felt to solve the problems presented by using either a conventional tractor orpusher propeller, either of which would probably have to be removed for road travel topreclude accidental damage. This design had the engine located in front of the passengercabin to provide protection to passengers in a crash. However, this led to the need fora complex drive train to couple the engine to the drive wheels and propulsive fans.
7.5 Initial layout
The concept initially selected employed an inner wing section that blended, to somedegree, with the fuselage. It also had outboard wings which could be retracted into the
“chap
07”—
2003/3/10—
pag
e182
—#8
Table 7.1 Aircraft concept selection
1. GA craftwithtransmissionin wings
2. Liftingbody withtelescopingwings
3. Liftingbody withfoldingwings 4. Gyrocopter
5. Car withducted fansand foldingwings
6. Cessnawithrotatingwings
Structures1. Wing position −2 −1 −2 0 2 22. Aspect ratio 1 2 1 2 1 13. Sweep/Taper −2 2 1 0 1 24. Number of moving parts 0 1 0 1 0 15. Size of moving parts −2 1 −2 1 −2 −16. Wing loading 1 0 −1 −1 2 17. Weight distribution (moments) 0 1 −1 −1 1 18. Crashworthiness −1 0 0 2 1 1
Subtotal 1/6 −5 6 −4 4 6 8
Stability and control/aerodynamics/performance1. Control in all aspects (air) 0 1 1 1 1 12. Aft CG for rotation (air) 2 0 1 0 −1 13. Crosswind effects (road) −2 0 −1 1 −1 04. Low CG (road) 1 0 1 1 2 −25. Central longitudinal CG (road) 2 1 1 −1 2 26. Reduced lift (road) 1 −1 0 2 1 27. Clean flow over surfaces and props (air) 1 2 1 0 0 28. Streamlined frontal cross-section (air) 1 1 0 −1 1 19. High aspect ratio (air) 1 −2 1 0 2 −1
10. Wing placement-mid wing (air) 2 0 2 0 0 111. Low profile drag – after conversion (road) 1 1 1 1 1 2
Subtotal 1/6 10 3 8 4 8 9
“chap
07”—
2003/3/10—
pag
e183
—#9
Propulsion1. Power −1 0 0 −1 −1 −22. Size, weight of engine and transmission −1 −2 0 0 −1 13. Engine type (fuel) 0 0 0 0 0 −14. Fuel efficiency and range −1 0 1 0 0 05. Cost 0 0 0 −1 0 16. Location of engine and transmission. Easy access 0 0 0 0 0 −1
Subtotal 1/6 −3 −2 1 −2 −2 −2Car1. Stability −2 1 −2 2 1 02. Crashworthiness −2 0 −2 1 1 −13. Driver visibility 0 2 2 2 2 14. Road friendly −2 1 2 2 2 −15. Ease of conversion −2 1 −2 0 0 16. Aesthetics −2 0 0 2 2 −27. Access −1 1 2 2 1 2
Subtotal 1/6 −11 6 0 11 9 0Cost/manufacturing1. Market 0 1 0 1 0 02. Development (outsourcing) 0 −1 −1 −1 0 23. Simplicity of design −1 1 −1 −2 −1 14. Service and running costs −1 −1 1 −1 1 0
Subtotal 1/6 −2 0 −1 −3 0 3Human factors1. Safety −2 0 1 1 0 −12. Ingress/Egress −2 −1 2 2 2 23. Visibility −2 2 0 2 1 14. Conversion ease −2 2 0 2 1 25. Aesthetics/Noise −2 1 −1 0 2 −1
Subtotal 1/6 −10 4 2 7 6 3Total −3.50 2.83 1.00 3.50 4.50 −3.50
“chap07” — 2003/3/10 — page 184 — #10
184 Aircraft Design Projects
Fig. 7.2 Sketch of agreed configuration
inner wing when in the road configuration. This would give a vehicle width of less than8 ft (2.44 m) for highway use. This is within normal highway lane-width limits for mainroadways and is no wider than some automotive vehicles already in use. The vehicle’sheight and length were 8 ft (2.44 m) and 17 ft (5.18 m) respectively; both selected toenable roadway travel without restriction due to length or height. The weight (mass)was estimated to be 3500 lb (1591 kg).
The design of the wing was a crucial part of the concept. The inner wing was tohave a chord of 11.32 ft (3.45 m) with its 8 ft (2.44 m) span, giving an aspect ratio of0.707. An end plate that expanded into a vertical stabilizer/winglet was to be used atthe tip of these inner wings in expectation of improving performance and providing aclean separation between inner and outer wing sections in flight. A horizontal stabilizerconnected the vertical stabilizers. The 6 ft (1.83 m) chord outer wings were each to bemade of either three or four sections that would telescope out of the inner wings to afinal span of 23 ft (7.01 m) for flight. The telescoping mechanism was to be similar tothat patented in the United States by Branko Sarh.1
For highway stability, the center of gravity (mass) is located midway between thefront and rear wheels. This makes rotation for take-off difficult, if not impossible. Toovercome this problem, the inner wing was positioned on the vehicle such that it is at anegative angle of attack when operating on the highway. For take-off and landing, thefront wheels were to be extended to raise the nose of the vehicle. This gives a positiveangle of attack for both the inner and outer wings and allows take-off in a reasonabledistance without rotation.
The twin vertical tails and the horizontal stabilizer are used to provide pitch and yawstability and control. Due to the relatively short moment arm, all these surfaces arelarger than normal. Flaperons on the outer, telescoping wing sections are used for rollcontrol and to provide extra lift in landing.
As noted above, the vertical tails were extended around the inner wing tip to givean end-plate effect. This was done to improve the performance of its very low aspectratio planform. It was later decided to twist these slightly to provide a winglet effect,
“chap07” — 2003/3/10 — page 185 — #11
Project study: a dual-mode (road/air) vehicle 185
providing slightly more thrust. The aerodynamic analysis of the wing included anoptimization of the winglet angle for these vertical tail sections.
During the aerodynamic analysis of the vehicle, it was found that the thrust from thetwin-ducted fans was insufficient for the desired cruise speeds. The design was there-fore changed to employ a single, large, unducted, pusher propeller placed behind thefuselage over the trailing edge of the inner wing. This change was accompanied by arelocation of the engine to a position aft of the cockpit. This placed it closer to thepropeller and rear drive wheels. This change necessitated the addition of a rear firewalldesigned to force the engine downward, under the cabin, in case of an accident. Thisshift in engine location moved the center of gravity (mass) aft, requiring an increasedsweep of the vertical stabilizer to provide an additional moment arm for the horizontaltail. The resulting inner wing, tail, propeller configuration represents a variation on a‘channel’ wing where the large propeller enhances the flow over the top of the innerwing and thus increases lift. The winglet-like capabilities of the vertical stabilizers arealso enhanced by the repositioning of the propeller. This configuration is shown inFigure 7.3.
The baseline configuration had a slightly smaller width on the highway with theinboard wing-span now at 7.48 ft (2.28 m) and had a chord of 8.2 ft (2.5 m). Theouter wings were redesigned to telescope in four sections with a chord averaging5.74 ft (1.75 m). They extend to a span of 27.16 ft (8.28 m), giving a gross wingarea of 174.4 ft2 (16.2 m2) and an aspect ratio of 4.23. The resulting unusual wingplanform with its partial span located ‘winglets’ (vertical stabilizer/end plates) wouldrequire careful analysis and testing to ensure that the vehicle performs as required inflight.
The housings for the wheels were also modified from the initial concept to make thefront wheel enclosures integral with the body/fuselage of the vehicle. The drive wheelswere located at the rear of the inboard wing section and enclosed in housings thatprojected from the inside corner formed by the inboard wing and the vertical stabilizers.The front wheels were designed to retract tightly into their housings in flight and toextend to both a ‘highway’ position and a take-off position. Observers noted that theresulting vehicle, in its highway configuration, looked like a propeller-powered, turn of
Four-parttelescopic
wing
Fig. 7.3 Baseline aircraft layout
“chap07” — 2003/3/10 — page 186 — #12
186 Aircraft Design Projects
the twenty-first century Volkswagen ‘Beetle’ sitting on a stubby wing and employing alarge racing spoiler.
7.6 Initial estimates
Let us now further examine the technical details and performance (flight and roadway)of this unusual vehicle.
7.6.1 Aerodynamic estimates
This design incorporated a unique combination of aerodynamic concepts including:
• a lifting fuselage,• an inboard ‘channel’ wing,• ‘inboard’ winglets, and• a telescoping outboard wing.
These made the analysis of vehicle performance a challenging prospect.The aerodynamic examination needed to consider both in-flight and highway modes.
These operating conditions presented contradicting aerodynamic requirements. Theanalysis was described in detail in the final project report.3 A summary of the mainfindings is given below.
The analysis of the in-flight aerodynamics required detailed examination of the newconcepts incorporated into the design. The vehicle body/fuselage, the inboard wing,the vertical tails and the propeller ducting accounted for a substantial part of the liftand drag. The fuselage was shaped with a flat bottom and curved top in the hope ofproducing some lift in flight. The flow over the fuselage was enhanced by the pusherpropeller. These combined with the inboard wing and vertical tails to form a modified‘channel wing’. The channel wing was proposed by Custer4 in the 1940s. In the take-offphase, the channel wing design enables the inboard wing to develop extra lift at lowspeeds due to flow augmentation from the propeller. The performance of the low aspectratio inboard wing can be improved by as much as 15 percent by using the vertical sta-bilizers as winglets. This effect should also be enhanced by the propeller if the ‘winglets’are properly designed. In the aerodynamic analysis, the outboard wing sections can betreated as separate, low aspect ratio surfaces. However, attention needs to be given toassessing the effect of the inboard/outboard wing junction on the spanwise loading.
Seven airfoil profiles were considered for the wing: the NACA 2412, 4412, 631-412,632-415, 652-215, the NASA LS 0417 (GA (W) 1) and LS 0413 (GA (W)-2). The initialselection was to use the 4412 airfoil with its almost flat lower surface for the inner wingand the LS 0417 for the telescoping outer panels. The outboard wing was to be mountedwith its chord set at an angle 2◦ higher than the inboard wing to give enhanced lift ontake-off, but this produced stall control problems in flight. Consequently, the designwas altered to use the LS 0417 for both parts of the wing with both at the same angle.To gain the needed lift on take-off the front undercarriage legs are extendable to givean angle of attack of 8◦ prior to the take-off run. At the desired 150 kt (77.2 m/s) cruisespeed the ideal angle of attack for the airfoil was essentially 0◦, hence, that angle wasselected as the mounting angle of the wing to the fuselage. The telescoping outer wingswere given a dihedral angle of 5◦ for roll stability in yaw. In order to counteract thenose-down pitching moment inherent in the LS 0417 airfoil, the horizontal stabilizerwas moved rearwards.
“chap07” — 2003/3/10 — page 187 — #13
Project study: a dual-mode (road/air) vehicle 187
90Velocity(m/s)
Width position
Chord position
75
600
1.0
2.00
1
2
3
Fig. 7.4 Velocity distribution over ‘scoop wing’
The NACA 0012 airfoil profile was selected for the horizontal tail. Based on wingletthickness recommendations in Raymer5 an NACA 0008 was used for the verticalstabilizers/inboard winglets.
The aerodynamic characteristics of the channel wing were analyzed using the con-ventional actuator disk model for propellers. This simple momentum theory assumesa continuous acceleration of the flow forward and aft of the propeller disk. Since thewing was flat and not wrapped around the bottom of the propeller disk (as in the truechannel wing configuration), this design was termed a ‘scoop’ wing. This term arosedue to the resemblance to a rectangular scoop in frontal profile. The analysis of the‘scoop’ wing performance produced the plots of velocity and pressure coefficient overthe wing upper surfaces shown in Figures 7.4 and 7.5. Using this data, it was estimatedthat the propeller could induce a lift of 522 pounds (2248 N) at zero forward speed.As the speed of the vehicle increases this lift enhancement decreases.
The extent to which two vertical stabilizers serve as winglets and a fence betweenthe outboard and inboard wing sections was analyzed using a model of the vortexgenerated at the intersection of two semi-infinite wings of different chords. A wingletworks by using the cross flow from the tip (or in this case the junction) vortex combinedwith the free stream flow to create a ‘lift’ with ‘spanwise-inward’ and forward (thrust)components. This vortex-induced cross flow decreases with increasing distance fromthe vortex core. The angle of the local flow to the vertical stabilizer/winglet can be foundas a function of that distance, allowing the calculation of the optimum local ‘angle ofattack’ or twist angle of the winglet for ‘thrust’ production. This result is shown inFigure 7.6.
Based on the above it was found that at cruise conditions each winglet could generate9.32 pounds (41 N) of thrust. Although the winglet only makes use of the inner/outerwing-junction vortex and not the full wing tip vortex, this corresponds to a 5.5 percentincrease in L/D for the entire wing (including the outer wing lift and drag). This boostin performance may seem small but on a vehicle with such a low aspect ratio, any suchhelp is welcome.
“chap07” — 2003/3/10 — page 188 — #14
188 Aircraft Design Projects
Pres. coeff.
Width position
Chord position
–10
1.0
2.0 0
1
2
3
Fig. 7.5 Pressure coefficient distribution over ‘scoop wing’
1.0
0.8
0.6
0.4
0.2
0 0.5 1.0Distance from centre (m)
Twis
t fr
om f
rees
trea
m (°
)
1.5 2.0
Fig. 7.6 Optimum wing twist distribution
The unique shape of the wing planform with the outer wing extended requireda detailed three-dimensional aerodynamic analysis. Using methods in Raymer5 andaccounting for the effect of the winglets, lift curve slopes of 4.97 and 4.92 were calcu-lated for the cruise and take-off/landing phases of flight respectively. The associatedmaximum lift coefficients were estimated at 1.818 and 1.755 respectively.
Using methods outlined in Raymer5 and Torenbeek6 to calculate flap effectiveness,it was found that full span, 20 percent chord flaps on the outer extended wing wouldincrease CLmax at landing to 2.092 with 10◦ deflection and to 2.534 with 60◦ of flap.
There was some concern about the possible need to twist the wing to assure that theoutboard sections of the wings stalled later than the inboard wing to provide adequateroll control in post-stall flight. Calculations using methods in Torenbeek6 indicated thattwist would not be needed. It was also found that the vortex at the inboard/outboardwing junction would help assure that stall progressed outboard along the wing-span.Later wind tunnel tests confirmed that no twist was needed for acceptable spanwisestall progression.
Similar thorough analyses were performed on the vertical and horizontal stabilizersto assure their effectiveness in providing the required aircraft trim and control.
“chap07” — 2003/3/10 — page 189 — #15
Project study: a dual-mode (road/air) vehicle 189
Drag was estimated using methods of Torenbeek.6 This involved totaling the drag ofall aircraft components (wings, flaps, winglet/stabilizer, horizontal stabilizer, fuselage,engine cooling, windscreen, and the undercarriage). Corrections for fuselage/winginterference, the low mounting of the wing, surface roughness, and control gap effectswere applied. The drag coefficient based on the extended wing planform (gross) areawas found to be 0.0241 in cruise, 0.0718 at take-off, and 0.0830 in landing. The increaseddrag in take-off and landing configurations comes from flap deflection and the extensionof the landing gear.
Finally, a similar aerodynamic analysis was made for the vehicle in the highwaymode to determine its highway aerodynamic drag characteristics. The outer wing beingretracted to assure that the lift would not endanger controllability at a cruise speed of70 mph (31.3 m/s). These results were used in the vehicle road handling analysis.
7.6.2 Powerplant selection
Propulsion represented a unique problem for this design since the selected engine mustprovide power for both airborne and highway use. Alternatively, two separate enginescould be used. Based on the drag calculations above with a cruise at 150 kt (77.2 m/s) at3000 m (9843 ft) altitude, and an estimated maximum gross weight (maximum take-offmass) of 3308 lb (1500 kg), it was found that a cruise engine power of 207 hp (155 kW)was required.
The use of two engines was considered, one to power the propeller in flight and theother to drive the wheels on land. Weight and the capability of the propulsion systemto operate efficiently in both modes of travel are the critical issues. For road travel,an engine plus a transmission is required to provide the needed power over a widerange of wheel speeds and the needed acceleration in stop and go traffic. For flight, notransmission may be needed and the engine would operate at a constant speed. While itis conceivable that separate, optimized systems could be found which would weigh lessthan a single, perhaps more complex engine and transmission system, it is not likely.The design team made the decision to employ a single, aircraft type engine, optimizedfor constant rpm operation, and to combine it with a transmission/drive train thatwould allow operation in highway traffic. The decision to design around the aircraftengine was made because the primary mode of travel for the vehicle was to be flight.The automobile role was considered as secondary; i.e. it would not be a vehicle used inday-to-day commuter highway use.
Several engines were considered as shown in Table 7.2. Two of these are standardgeneral aviation piston engines, one is an automobile engine which has been popularas a ‘conversion’ engine for home-built general aviation aircraft and two are engineswhich were in development at the time of the design project.
Table 7.2 Engine properties
Engine Current use Power, hp (kW) Weight (mass), lb (kg)
Continental IO-520 Beech C-55 310 (231) 436 (198)Lycoming IO-540 Cessna 182 290 (216) 437 (198)Subaru EJ-22 Conversion 160 (119) 295 (134)Dyna-Cam In development 200 (149) 300 (136)Wilksch In development 250 (186) 287 (130)
“chap07” — 2003/3/10 — page 190 — #16
190 Aircraft Design Projects
The selected engine was the Wilksch five-cylinder turbo diesel with intercooler. Itwas the closest to the required power output, gave an excellent thrust to vehicle weightratio (0.3), and could run on both diesel and jet fuel. Turbocharging gives the enginesea-level performance at the desired cruise altitude of 3000 meters (9843 ft). This enginehad a small size, 3.45 × 1.48 × 2.1 ft (1.05 × 0.45 × 0.64 m), and at maximum powerit produced 486 ft-lb (660 N-m) torque at 2700 rpm. Its specific fuel consumption was0.441 lb/hp-hr (0.270 kg/kW-hr). The engine had an internal fuel pump that would pullthe fuel from an 80 US gallon (303 liter) tank in the fuselage below the engine.
Aircraft engines such as the selected Wilksch diesel are designed for hour after hour ofconstant rpm operation while automobiles, especially in urban traffic, may encountercontinuously varying rpm requirements. Automobile transmissions use a variety ofgear ratios to provide a set range of drive shaft speeds and torque within the rpmrange of the engine. This utilizes the torque capability of the auto engine over a rangeof rotational speeds. While some car engines, like the Subaru engine in Table 7.2,have been used in aircraft, they tend to be heavier than comparable power aircraftengines. The solution to this dilemma was found in the use of a ‘continuously variabletransmission’ or CVT that permits constant rpm operation of the engine while varyingthe amount of power transmitted to the wheels in road use. The selected system was theAudi Multitronic CVT that has a weight (mass) of 220 pounds (99.8 kg) and measures1.3×0.82 ft (0.40×0.25 m). The quoted weight (mass) includes the needed transmissionfluid and is slightly lower than standard automobile transmissions with comparablecapabilities.
A gearing system was also incorporated into the drive train design to allow couplingof the propeller and wheel drives. A ‘dog’ clutch is used to switch power between thewheels and the propeller and is designed to prevent simultaneous operation of bothsystems.
The 6.56 foot (2 meter) diameter, three-bladed propeller was designed using a pro-peller design program from the Internet web site of Leonard Newnham.7 The programrequired the input of engine power, engine rpm, aircraft speed, number of blades, andthe propeller diameter and blade angle of attack in order to output an ‘optimum’ bladedesign. The result promised to give a high propulsive efficiency.
The location of the liquid-cooled engine requires a cross-flow radiator and an elec-tric fan similar to that used in modern automobile engines. Size constraints of thedesign limited radiator area to 1.5 ft2 (0.14 m2) and calculations showed that a 0.5 hp(373 W) electric motor would be more than adequate for cooling under the most adverseconditions.
7.6.3 Weight and balance predictions
Initial sizing (weighing) of the vehicle was attempted using published statistical curvefit methods (Roskam8 and Raymer5) but these were of questionable value, given theunconventional nature of the ‘aircraft’. Hence, a combination of traditional sizingmethods and actual system weight (mass) data was used to produce Table 7.3.
Based on the estimated weights (masses) the balance of the aircraft could be analyzedand the center of gravity (mass) excursion determined. This is shown in Figure 7.7.
7.6.4 Flight performance estimates
Based on the above aerodynamic analysis and powerplant performance estimates, theflight performance of the vehicle in cruise could be calculated using the following
“chap07” — 2003/3/10 — page 191 — #17
Project study: a dual-mode (road/air) vehicle 191
Table 7.3 Component weight (mass) estimates
Component Weight (lb) Mass (kg)
StructureWing 215 97.7Horizontal tail 30 13.6Vertical tail 50 22.3Fuselage 350 159.1Main gear/rear wheels 85 38.6Nose gear/front wheels 85 38.6
Subtotal 815 370.5PropulsionEngine 400 181.8Transmission 305 138.6Propeller 50 22.7Fuel system 45 20.5
Subtotal 800 363.6SystemsControls 20 9.1Electrical 190 86.4Avionics 100 45.5Anti-icing system 80 36.4
Subtotal 390 177.3Cabin furnishings 115 52.3Variable weights (masses)Fuel 480 218.2Front passengers 320 145.5Rear passengers 320 145.5Luggage 60 27.3
Subtotal 1180 536.4Total 3300 1500
characteristics:
• a wing aspect ratio of 4.46,• a calculated Oswald efficiency factor of 0.92,• an aircraft parasitic drag coefficient CD0 = 0.025,• a propeller efficiency, ηp, of 88 percent giving a constant thrust of 1012 lb (4500 N),• a specific fuel consumption of 0.441 lb/hp-hr (0.270 kg/kW-hr), and• a maximum gross weight (mass) of 3308 lb (1500 kg).
The power plot, Figure 7.8, shows a cruise speed (at 80 percent power, at 9843 ft(3000 m) altitude) of 157.5 kt (81 m/s) and a maximum speed at this altitude of 179 kt(92 m/s).
Using take-off at 1.2 stall speed from a hard surface gave a take-off ground rollof 689 ft (210 m) and a 50 ft (15.24 m) obstacle clearance take-off distance of 920 ft(280 m). With touchdown at 1.3 stall speed, which can be achieved with less than 10◦flap deflection, and braking at 80 percent of touchdown speed, the landing groundroll was calculated at 755 ft (230 m). This gave a total distance of 1148 ft (350 m) afterclearing a 50 ft (15.24 m) obstacle at an approach sink rate of 787 ft/min (4 m/s). With30◦ flap deflection, this distance is reduced to 1066 ft (325 m).
“chap07” — 2003/3/10 — page 192 — #18
192 Aircraft Design Projects
–25900
1000
1100
1200
1300
1400
1500
1600
–20 –15 –10Static margin (% MAC)
–5 0
Max. take-off mass
Add rear passengers
Less fuel
Less bags
Less rear passengers
Operational empty
Empty
Less front passengers
Add fuel
Add front pass
Mas
s (k
g)
Add bags
Fig. 7.7 Aircraft center of gravity excursion diagram
200
150
100
50
20 40 60Aircraft forward speed (m/s)
Power available 100%
Power available 80%
Power required
Pow
er (k
W)
80 100
Fig. 7.8 Performance envelope at 3000 meters
The aircraft maximum rate of climb at sea level was found to be 1460 ft/min(445 m/min), and 755 ft/min (230 m/min) at the cruise altitude of 9842 ft (3000 m). Theabsolute ceiling was determined as 21 650 ft (6600 m). In normal 80 percent power cruiseconditions at 9842 ft (3000 m) the range was calculated to be 825 nm (1528 km) witha 5.7 hour endurance. Flying at minimum drag conditions gave a maximum range of960 nm (1778 km). At the speed for minimum power required the maximum endurancewas found to be 9.5 hours. The design had proved to exceed all performance goals inthe aircraft operation. It would be possible to re-optimize the aircraft configuration tobetter match the operational specification at this point but time was not available todo this in this project.
“chap07” — 2003/3/10 — page 193 — #19
Project study: a dual-mode (road/air) vehicle 193
7.6.5 Structural details
There is an essential difference in structural design considerations for aircraft and cars.For aircraft, low weight with strength is paramount, while automobile designers needto add a focus on structural stiffness to improve handling and suspension performance.For this project the structure was designed to meet both general aviation aircraft andautomobile requirements (FAR 23 and US National Highway Transportation SafetyAdvisory respectively).
The aircraft loads and their distributions over the lifting surfaces were developedbased on the information shown in the flight envelope (V -n diagram), Figure 7.9.
The general structural layout of the vehicle is shown in Figure 7.10 with the majorstructural members numbered on the figure and identified in Table 7.4.
The structural design was evaluated in three parts:
1. at the fuselage/inner wing combination,2. at the telescoping outer wings, and3. at the tail.
The fuselage/inner wing structure consists of four regions:
1. the crumple zone forward of the cockpit,2. the passenger compartment,3. the wing box, and4. the engine compartment.
The crumple zone was designed with an aluminum substructure covered by a com-posite skin. The skin is only lightly stressed and the aluminum frame is designed forcontrolled deformation in a crash using v-shaped indentations, termed ‘fold initiators’.The forward wheels (landing gear) and their structure are mounted to the first bulkheadat the rear of the crumple zone. The aluminum substructure continues through the pas-senger and engine compartments. The passenger compartment skin is fabricated withcarbon composite for stiffness and deformation resistance. The aluminum bulkhead atthe rear of the passenger compartment transfers the loads between the forward sparof the inboard wing and the fuselage. Attached to this bulkhead is a fiberglass firewallcoated with sperotex and phenolic resin. The firewall is mounted to the bulkhead at a
20
Gust + 15.24 m/s
+7.62 m/s
–3
–2
–1
1
2
3
4
5
40
n max
Load
fac
tor
(n)
60 80 100Ve (m/s)
Fig. 7.9 Aircraft structural flight envelope
“chap07” — 2003/3/10 — page 194 — #20
194 Aircraft Design Projects
11
18
19
12
13 1415
16
17
109
754,
6
8
32
1
Fig. 7.10 Structural framework
Table 7.4 Location and identification of major structural members
Component Member Number in fig. Fuselage sta. Wing sta.
Crumple zone Rib 1 1 107.87 n/aBulkhead 1 2 128.35 n/a
Passenger compartment Rib 2 (doorframe) 3 164.96 n/aBulkhead 2 4 202.36 n/a
Engine compartment Firewall 5 202.36 n/aBulkhead 3 6 275.20 n/a
Inboard wing spars Forward spar 7 202.36 n/aRear spar 8 275.20 n/a
Telescoping wing spars Forward spar 9 213.19 n/aRear spar 10 229.96 n/a
Horizontal tail Forward spar 11 350.79 n/aRear spar 12 362.20 n/a
Telescoping wing Rib 1 13 n/a 44.09Rib 2 14 n/a 73.62Rib 3 15 n/a 103.15Rib 4 16 n/a 132.48Rib 5 17 n/a 161.81
Horizontal tail Rib 1 18 n/a 14.72Rib 2 19 n/a 44.33
slight angle. This configuration, combined with fold initiators, is designed to drive theengine below the passenger compartment in a collision.
The most complex part of the structural framework is the telescoping wing. Theloads on the telescoping outer wing section were first approximated by examiningthe aerodynamic behavior of the combined inner and outer wings. The discontinuity
“chap07” — 2003/3/10 — page 195 — #21
Project study: a dual-mode (road/air) vehicle 195
1
1 2
Planform
Schrenk
Ellipical
Span (m)3 4
2
3
Eff
ectiv
e ch
ord
(m)
Fig. 7.11 Schrenk spanwise lift distribution
Inboard sectionand fuselage here
Endplate here
Rotating spar
Fixed spar
Axial movement
Axial movement
Rotation
Fig. 7.12 Diagram of telescopic wing mechanism
in wing chord at the inner/outer wing junction makes load analysis a challenge. Anapproximation based on Schrenk’s method5 was used to estimate the loads over theentire span. The result is shown in Figure 7.11.
Each of the outboard wings consists of four sections. These telescope outward fromtheir stowed position inside the inboard wing. The mechanism used to deploy andretract the outer wings is based on a patented design9 as illustrated in Figure 7.12.Each of the telescoping outer sections from tip to root is slightly larger than the innerones, allowing it to slide in over its neighbor. The telescoping sections are driven bythreaded, rotating spars supported by bearings and powered by a 12-volt motor in thecentral wing box. To prevent accidental deployment/retraction of the outboard wings,the motor can only be operated when the wheels/landing gear are in their extendedposition supporting the weight of the vehicle, and when the wheels are not turning.
“chap07” — 2003/3/10 — page 196 — #22
196 Aircraft Design Projects
Table 7.5 Structural material selection
Structure Component Material
Crumple zone Rib 1 Al 7075Bulkhead 1 Al 7075
Passenger compartment Rib 2 (doorframe) Al 7075Bulkhead 2 Al 7075
Engine compartment Firewall Fiberglass coated with sperotexand phenolic resin
Bulkhead 3 Al 7075Engine mounts SteelFuselage skin Carbon fiberWindows Plexiglas
Inboard wing Forward spar Al 7075Rear spar Al 7075Top skin Al 7075Bottom skin Al 2024
Telescoping wing Rotating spars Stainless steelNon-rotating spars Carbon fiberSpar attachments Al 7075Ribs Carbon fiber sandwichSkin Carbon fiber sandwich
Horizontal tail Forward spar Al 7075Rear spar Al 7075Skin Glass/Carbon fiber hybrid
Vertical tail Skin Glass/Carbon fiber hybridLanding gear Struts, supports, etc. Steel
Wheels Al 7075
The rotating spars are made of stainless steel for strength and stiffness. The rest of theoutboard wing is mostly manufactured in carbon fiber composite construction.
The twin, vertical tail sections are designed to be manufactured entirely of carbon-glass-epoxy resin, composite materials. Material thickness is greater toward the root ofthe vertical stabilizer/winglets where the greatest bending moments would exist. Thenumber of composite fiber layers will be reduced toward the horizontal stabilizer. Thespars in these elements will also be composite in construction. The horizontal tail hasaluminum spars.
The structural analysis included an extensive investigation of materials, strengths,and certification requirements for the composite structures. Table 7.5 lists the materialsused in the various parts of the vehicle.
7.6.6 Stability, control and ‘roadability’ assessment
A wide range of factors must be considered when examining the stability and controlneeds of a vehicle that operates as either a car or an airplane. These include:
• the sizing and design of aircraft control surfaces and the resulting static and dynamicflight stability,
• the ease and predictability of handling in the automobile operating mode, and• the internal systems needed to operate both the automotive and flight control
systems.
“chap07” — 2003/3/10 — page 197 — #23
Project study: a dual-mode (road/air) vehicle 197
Despite the somewhat unusual configuration of this vehicle, its flight control systemand the requirements placed on that system are fairly conventional. The design isdifferent from most general aviation aircraft in its use of a twin vertical tail and in itstelescoping wing. The adoption of the large twin vertical tails resulted in the need forrelatively small rudder size, as a percent of tail chord. The telescoping wing design ledto the need for simplicity in flap/aileron systems and, ultimately, to the use of a plain‘flaperon’ system, combining the role of conventional flaps and ailerons.
The static and dynamic control and stability requirements were calculated usingmethods of Raymer,5 Thurston,10 Etkin and Reid,11 and Render.12 The resulting tailvolume coefficient was 0.35 and both rudder and elevators were sized at 35 percent oftheir respective stabilizer areas. Full span, 25 percent chord flaperons were used on theouter, telescoping wings. Calculations showed that with these controls the aircraft wasable to meet Military Specification 8785C, level-one dynamic stability requirements forall cases except Dutch roll mode, which met level-two requirements. A complete analysisof the flight stability is presented in the final design report3 but is not included here.
In highway use, this vehicle was not designed to be a high performance automobile.The emphasis was on handling and control, safety and predictability, and passengercomfort. All US and EU transport regulations related to safety and environmentalimpact had to be met. An added consideration was the requirement that a vehicledesigned to fly does not do so on the highway!
7.6.7 Systems
One of the major decisions in the design process was to integrate the car and airplanecontrol systems as much as possible. This has been achieved by using electronic ratherthan cable or hydraulic actuation of both automotive and aeronautical control systems.In this fly/drive-by-wire system, a joystick would replace both the automobile steeringwheel and the aircraft yoke or stick. On the road the vehicle would have an automaticrather than a manual transmission and thus would have two foot controls, the brakeand the accelerator pedals. In the air, these would serve as conventional rudder pedals.Both of these controls (floor pedals and joystick) would be attached to a fully electronic,fly/drive-by-wire control system. This would include a feedback to the pedals andjoystick designed to give normal feel in both flight and the highway operation.
The instrument panel would have a large liquid crystal display (LCD) which wouldshow a conventional automotive instrument array on the road and a modern aircraftflight control system display in the air. Required mechanical back-up instruments wouldbe placed on the perimeter of the LCD panel. Switching from aircraft to automotive(or reverse) control and instrument display systems would be accomplished manuallywith system locks that would prevent any changeover when the vehicle was in motion.
The joystick controls are side-mounted, simulating the practice in many moderntransport and military aircraft. The throttle control when in the aircraft mode ismounted on a center panel. Numerous studies of joystick type control systems forautomobiles have shown that such systems are easy to use for most drivers and otherstudies of drive-by-wire automobile control and steering systems have proven theirfeasibility. Table 7.6 illustrates the way in which the driver/pilot would use the joystickand pedals for control of the vehicle in both operational modes.
There will also be a four-way toggle switch on top of the joystick. This will operateeither the elevator trim or the headlight beam position when moved forward and aft,and either the rudder trim or the turn signals when moved left or right.
“chap07” — 2003/3/10 — page 198 — #24
198 Aircraft Design Projects
Table 7.6 Control system actions
Action Aircraft mode Automobile mode
Left rudder/brake pedal depressed Yaw to left Four wheel brakingRight rudder/accelerator pedal depressed Yaw to right Vehicle acceleratesMove joystick to left Roll to left Steer to the leftMove joystick to right Roll to right Steer to the rightStick pushed forward Lower aircraft nose No actionStick pulled back Raise aircraft nose No action
The wheels/tires and suspension system represented a unique challenge. Thesuspension system had to meet requirements for all three modes of operation:
• highway use (normal extension),• flight (full retraction into wheel wells),• take-off and landing (normal extension of rear wheels, full extension of front wheels
for increased take-off roll angle of attack).
The system had also to be designed to absorb the vertical and horizontal impact forcesencountered in landing and to handle the side force loads associated with cornering inthe automotive mode. This required a careful specification of tire type and size as wellas a good design of the suspension system itself.
The tires will need to possess characteristics that represent a hybrid of normal aircraftand car tires in terms of cornering stiffness and impact deflection. These propertiesare primarily a function of the tire aspect ratio (height to width). Low aspect ratiogives increased cornering stiffness and high aspect ratio gives better impact deflection.Different tire widths were specified for front and rear units to provide greater corneringstiffness at the rear (main) gear location. The front suspension uses an upper wishboneconfiguration with the lower arm attached to a longitudinal torsion bar. A screw jack isused with a damper (shock absorber) to attach the suspension wishbone to the vehicleframe, allowing extension or retraction of the wheel into the wheel well. The rearsuspension is a trailing arm configuration with a spring/damper unit between the wheeland the vehicle frame. An extensive analysis of this suspension system and its behaviorunder all conditions was undertaken using methods of Gillespie.13 This was presentedin the design final report.3
7.6.8 Vehicle cost assessment
An analysis of the projected cost of an airplane is always difficult and such an eval-uation for a combination automobile/airplane is necessarily based more on guessesthan technical methods. Cost estimation began with standard methods outlined byRoskam.14 Such methods are heavily based on past experience of general aviation air-craft. There are few, if any, vehicles comparable to this design. However, based onan admittedly optimistic production estimate of 1000 vehicles per year over a ten-year period and on assumptions of modern manufacturing techniques, an estimatedcost per vehicle is $276 627. This figure is based on the cost components outlined inTable 7.7.
“chap07” — 2003/3/10 — page 199 — #25
Project study: a dual-mode (road/air) vehicle 199
Table 7.7 Summary of estimated costs per vehicle
Research, development, testing, and evaluation cost $15 000Program manufacturing costs
Airframe engineering and design $1688Aircraft production $215 935Flight test operations $400Overhead and indirect costs $21 802Profit $21 802
Total $261 627Aircraft estimated price $276 627
Fig. 7.13 Wind tunnel test model
This projected cost is at the high end of a range of four-place aircraft with comparableperformance. However, our aircraft provides a ‘roadable’ option. It would be interestingto see if there is a viable market for such a design.
7.7 Wind tunnel testing
An eighth scale model of the vehicle was constructed of wood, plastic foam with alu-minum wing spars. It was tested in a wind tunnel with a 6×6 (1.83 m×1.83 m) test areacross-section. The model was mounted in the wind tunnel on a six-component straingauge balance and tested through a range of angle of attack (from −6 to +16◦). Testresults consisted of force and moment data as well as photographic flow visualizations.Figure 7.13 shows the model being tested with wool tufts for flow visualization.
Although, due to time constraints, testing was limited in scope, the results did confirmthe viability of the design. Stall was quite manageable and the outboard wings were
“chap07” — 2003/3/10 — page 200 — #26
200 Aircraft Design Projects
shown to have attached flow after the inboard wing stalled, allowing control in stall.The horizontal tail also exhibited attached flow after stall of the inboard wing. Despitethe somewhat unusual design of the vehicle, there was no evidence of separated flowareas at the rear of the fuselage, even with the propeller not operating. The tests alsoconfirmed a rather broad range of angle of attack for near maximum lift to drag ratioshowing that cruise efficiency is not very sensitive to angle of attack.
Tests were also run with the outboard wings removed from the model, simulating theon-road configuration. These confirmed that this gave a lift coefficient low enough toavoid unintended ‘lift-off’ while in use on the road.
7.8 Study review
The design of the roadable aircraft proved a challenging but successful student project.The design report was entered in the 2000 NASA/FAA General Aviation Design Com-petition and won first prize. Details of the final design are given in Table 7.8. While itmay remain unlikely that a truly roadable aircraft will ever be successfully marketed,this exercise, like several designs for ‘flying cars’ that have been built and introduced inthe past, shows that such a vehicle is feasible. There continues to be strong interest insuch vehicles among inventors and dreamers. In the future, a design with many of thefeatures described here may finally fulfill these dreams. As illustrated in Figure 7.14, acar/plane that will give its owners and operators a freedom of transport that does notexist with present-day aircraft or automobiles must one day be a reality.
Table 7.8 Aircraft description
Aircraft type: General aviation four-place radable aircraftPropulsion: Wilksch 250 hp (186 kW) diesel engineAircraft mass: Empty = 1568 kg 3457 lb
Max. fuel = 480 kg 1058 lbPayload = 800 kg 1764 lbMax. TO = 2848 kg 6280 lb
Dimensions: Overall length = 4.25 m 14.0 ftOverall height = 1.30 m 4.2 ftSpan (wing extended) = 4.14 m 13.6 ftSpan (wing retracted) = 2.16 m 7.1 ftWing area (total) = 15.88 sq. m 170 sq. ftAspect ratio (total) = 4.46Wing taper ratio = 1.0Wing profile NASA GAW-1Wing thickness = 17%Wing sweep = 0◦Wing dihedral (outbd) = 5◦Horizontal tail area = 2.85 sq. m 30.6 sq. ftVertical tail area = 3.18 sq. m 30.6 sq. ftTail profile NACA 0012
Performance: Stall speed = 28 m/s 54 kts(at max. TO mass) Cruise speed = 77 m/s 150 kts
TO speed = 33.6 m/s 65 kts
“chap07” — 2003/3/10 — page 201 — #27
Project study: a dual-mode (road/air) vehicle 201
Fig. 7.14 Computer simulation of vehicle in flight
References
1 Stiles, Palmer, Roadable Aircraft, From Wheels to Wings, Custom Creativity, Melbourne,FL, 1994.
2 Mertins, Randy, Closet Cases, Pilot News Press, Kansas City, MO, 1982.3 Gassler, R. et al., Pegasus, the First Successful Roadable Aircraft, Virginia Tech Aerospace &
Ocean Engineering Dept., Blacksburg, VA, 2000.4 Anon. ‘The wing that fooled the experts’, Popular Mechanics, Vol. 87, No. 5, May 1947.5 Raymer, D. P., Aircraft Design: A Conceptual Approach, 2nd edition, AIAA, Washington
DC.6 Torenbeek, Egbert, Synthesis of Subsonic Aircraft Design, Delft University Press, Delft, 1981.7 Newnham, L., http://helios.bre.co.uk/ccit/people/newnhaml/prop.8 Roskam, Jan, Airplane Design, Part IV, DARcorporation, Lawrence, KS, 1989.9 Czajkowski, M., Clausen, G. and Sahr, B., ‘Telescopic wing of an advanced flying
automobile’, SAE Paper 975602, SAE, Warrendale, PA, 1997.10 Thurston, David B., Design for Flying, 2nd edition, McGraw-Hill, New York, 1994.11 Etkin, Bernard and Reid, Lloyd, Dynamics of Flight, Stability and Control, Wiley & Sons,
New York, 1995.12 Render, Peter M., Aircraft Stability and Control, Aeronautical & Automotive Engineering
Dept., Loughborough University, UK, 1999.13 Gillespie, Thomas D., ‘Fundamentals of Vehicle Dynamics’, SAE, Warrendale, PA, 1992.14 Roskam, Jan, Airplane Design, Part VIII, DARcorporation, Lawrence, KS, 1989.
“chap08” — 2003/3/10 — page 202 — #1
8
Project study: advanced deep
interdiction aircraft
Northrop Grumman B-2A Spirit Stealth Bomber
“chap08” — 2003/3/10 — page 203 — #2
Project study: advanced deep interdiction aircraft 203
8.1 Introduction
This project formed the basis of the American Institute of Aeronautics and Astronautics(AIAA)1 annual undergraduate team aircraft design competition in 2001/02. Teamsof three to ten students from the best aeronautical courses compete for prestige andcash prizes. The Request for Proposal (RFP) published by AIAA is based on recentindustrial project work. Judges look for a thorough and professional submission fromthe team, which demonstrates a specific and complete understanding of the problem.This competition provides a useful source of current projects and operational data thatcan form the basis of undergraduate design projects even if the designs are not to besubmitted for the competition.
The background to the project, as described in the original RFP,2 is given below:
When the F111 was retired from service in 1996 it was partially replaced by theF-15E. The balance of USAF deep-interdiction capabilities are provided by theF-117, B-1 and B-2 aircraft. All of these aircraft are expected to reach the endof their service lives in or before the year 2020. The need exists for a new aircraftwhich can effectively deliver precision guided tactical weapons at long range andwhich can rapidly deploy with minimum support to regional conflicts world-wide.Improved threat capabilities dictate that this new aircraft have signatures in allspectra comparable to or less than those of the F-117. The capability to super-cruise (fly supersonically without the use of afterburner) will allow these aircraftto respond to crises around the world in half the time required for current strikeassets. Approximately 200 aircraft are needed to replace the F-15E, F-117, B-1 &B-2 aircraft.
The complete AIAA description of the problem2 includes some detailed operationalrequirements, mission profiles and some engine and weapon design data. These areincorporated and discussed in the problem analysis and aircraft specification sectionsbelow.
8.2 Project brief
Recent conflicts in the Middle East, Eastern Europe and Central Asia have displayedthe military strategy for modern warfare. The first objective of a new offence is to‘neutralise’ the command and control centres of the enemy and to degrade their airdefence facility. This is termed ‘interdiction’. For the Airforce, this is a difficult anddangerous mission. In the initial attacks, the aircraft are expected to engage well-defended targets lying deep inside enemy territory. The range of the mission may bebeyond the operational range of protective fighter aircraft and other support. Theinterdictive-role aircraft must therefore be self-supporting and able to evade, or protectthemselves against, all the defensive systems of the enemy.
8.2.1 Threat analysis
Interdictive strike aircraft are expected to operate early in the conflict. This is at atime when the enemy’s defensive systems have not yet been degraded. To avoid threats,the traditional tactic relied on fast, low-level approach under the protective screen ofthe enemy radar. Improvements in radar technology and the introduction of relatively
“chap08” — 2003/3/10 — page 204 — #3
204 Aircraft Design Projects
cheap surface-to-air missiles (SAM) eventually made this tactic ineffective. Modernpractice relies on aircraft stealth and high-altitude penetration. This avoids low andmedium height threats from small-arms fire and low-technology SAM which nowmakes flight at altitudes below 20 000 ft very dangerous. A high-altitude mission profileensures that the aircraft can only be attacked with much more sophisticated defensiveweapons. The development of effective precision guided munitions and accurate tar-get designation makes the high-altitude operation effective. Providing the aircraft witha high-speed capability, reduces the duration of the mission over the target area andthereby lowers the exposure to enemy defensive systems. The adoption of stealth meansthat the aircraft is more difficult to detect. However, this means that it must act withoutclose air support that is any less stealthy. Defensive missile systems are becoming moreeffective at high altitude and such threats are also getting harder to detect and coun-teract. To rely on self-defence weapons and systems in future manned aircraft may beregarded as too optimistic. It is likely that even small countries will be able to afford suchdefence systems. Stealth, speed and height, which will make the defensive task moredifficult, are likely to be the best forms of protection in future interdictive operations.
In order to strike deep inside enemy territory, from friendly airfields, requires along operational range capability. The AIAA specification called for a combat radiusof 1750 nm (3241 km) without refuelling. This long-range, high-altitude performancedemanded an aerodynamically efficient aircraft configuration.
The two most significant design drivers for this project are identified as ‘stealth in allspectra’ and ‘high aerodynamic efficiency at supersonic speed and high altitude’.
8.2.2 Stealth considerations
In recent years, the technical and popular press has focused so much attention onradar detection (radar cross-section, RCS) that it would be easy to forget that there areseveral other ways to identify and target an intruding aircraft. These include, infraredemissions (IR), electronic radiation, sound (aural signature) and sight (visual signa-ture). Traditionally, the last of these led to the development of camouflage (the originalstealth solution!). In modern warfare, it is important to make sure that each identifier isreduced to a minimum. None of the signatures should be more significant than the oth-ers. For example, we all are aware that in civil aviation the noise is much more intrusivethan the visual characteristic. Similarly in military aircraft, the RCS or the IR char-acteristic must not dominate. Detailed technical information on stealth can be foundin textbooks3 and in technical papers. These textbooks and papers give advice on theanalysis methods used to design for stealth. The methods used to predict RCS fromthe geometry of the aircraft are complex and beyond the scope of undergraduate pre-liminary design projects. However, generalised guidance on the selection of layout andprofiling of the aircraft to minimise RCS is available.
Stealth issues influence the design of our aircraft in several different ways.
Radar
The AIAA specification required the RCS to be less than −13 dB. It is felt that withthe expected technical improvements in radar performance in the period up to firstflight (2020) this RCS may be too large. A value of −30 dB, if achievable, may be abetter target for this aircraft. To achieve this figure will require as much help as possiblefrom new technologies and the development of existing techniques. Existing methodsinclude ‘edge alignment’, avoidance of shape discontinuities, elimination of flat sur-faces, using radar absorbing structures (RAS), coating the external profile with radar
“chap08” — 2003/3/10 — page 205 — #4
Project study: advanced deep interdiction aircraft 205
absorbing material (RAM), and hiding rotating engine parts from direct reflection ofradar waves. Attention must also be given to the avoidance of radar scattering causedby the aircraft profiles and from the edges of access panels. All of these methods havebeen demonstrated and proved on the B-2 aircraft. However, the main objective of suchtechniques is to reduce radar reflectivity. This is important when the radar transmitterand receiver are at the some location.
New defensive radar systems now displace the two parts of the system. This makesit more important to absorb the radar energy into the structural framework and thematerials covering the aircraft profile.
Passive stealth techniques are currently being developed. These use plasma genera-tion to ‘assimilate’ the radar energy. Another method attempts to displace or disguisethe returning radar signature. This is intended to confuse defensive systems and maketargeting more difficult. Obviously, for security reasons, published information onthese developments is scarce. Therefore, little account can be taken of these new meth-ods when currently deciding on our aircraft configuration. It is encouraging to notethat research is identifying methods to reduce the radar threat. These are likely to beoperationally mature for this next-generation aircraft.
Infrared
Infrared radiation is a natural consequence of heat. It is more pronounced at highertemperatures therefore the best way to reduce the exposure is to lower the temperatureof the hot parts of the aircraft. The engine exhaust gases and surrounding structure giverise to the main source of IR radiation. A pure-turbojet engine exhaust is obviouslyeasier to detect than that of a bypass engine. In the bypass engine, the hot core airflow ismixed with the cooler bypass air before leaving the engine. This substantially reduces theexhaust stream temperature and therefore the IR signature. Another way of reducingthe IR signature is by shielding the hot areas from the potential detector. For example,if the IR detector is likely to be below the aircraft (a good assumption for our highflying aircraft) it would be possible to use the colder aircraft structure to hide theengine nozzle location. Positioning the engine exhaust forward and above the rearwing structure would provide this protection.
For aircraft travelling at supersonic speeds for long duration, the disturbed airflowwill cause kinetic heating of the structure. The stagnation temperature resulting fromaerodynamic heating is directly related to aircraft speed and ambient air parameters.For an aircraft in the stratosphere, travelling at M1.6 the stagnation temperature isestimated at over 100 ◦F (38 ◦C). It is difficult to estimate the actual skin temperaturesthat would result from this heat input as this will be dependent on the conductiveproperties of the structure and the heat radiation to the surrounding airflow. The tem-perature will be higher at positions of flow stagnation. Because of this, the leading edgesof the flying surfaces and the nose of the aircraft will be affected more than the rest ofthe structure. This could present a potential problem as infrared radiation will naturallyoccur. If this is regarded as a serious problem, it would be necessary to cool these areas.In the case of the wing structure, it may be possible to use the fuel from the wing tanksto conduct the heat from the structure into the cold fuel mass. In other areas, it may benecessary to use ceramic coatings or other materials to improve the conductive path.
Other observables
For most of us, aircraft noise is the most noticeable characteristic. Exhibitionists at airshows try to make as much noise as possible to attract attention. For missions over
“chap08” — 2003/3/10 — page 206 — #5
206 Aircraft Design Projects
enemy territory, the opposite strategy is advised! For our aircraft, there are mainly twosources of noise emission: the sonic boom and the engine.
For many years, researchers, mostly working on civil supersonic airliners, have beentrying to reduce or eliminate the noise from the sonic boom. For example, it maybe possible to mitigate the intensity of the noise energy by subtle shaping of theaircraft configuration, or by system innovations. However, the sonic boom is a nat-ural consequence of the pressure changes as the ambient airflow is accelerated andthen decelerated over the aircraft profile. The double-boom ‘explosion’ heard on theground will alert the enemy defences to the presence of the aircraft. They will lookupwards to confirm the sound. Hence, visual and aural observations are intrinsicallylinked.
Whereas noise heard on the ground from the sonic boom is transient, that from theengines is constant. This gives the observer more time to identify the aircraft visually.Engine noise is generated mainly from the impingement of the internal airflow on therotating machinery and by the intermixing of the exhaust airflow into the atmosphere.The exhaust noise is affected by the jet velocity to the seventh power. Mixing exhaustvelocity core airstream with a slower bypass stream before leaving the nozzle reducesengine generated noise significantly. The overall effect is to change the noise spectrumto increase higher frequency sound waves. As these are more rapidly dissipated withdistance, there is a reduction of noise heard on the ground compared to a pure jetengine. The bypass engine will also provide better fuel consumption. From both ofthese aspects, this makes it a good choice for the interdiction mission.
Human sight is a very effective sensor. During day time, we can all see airliners highin the sky against a clear blue sky. If condensation trails or reflections (glints) from theaircraft are present the observation is even easier. Avoidance of visual detection wasthe first stealth technology. Camouflage has now become a natural strategy in warfare.Research has shown that it is possible to reduce the condensation and reflections fromaircraft. At night, the main source of light comes from the exhaust glow. As the observerwill usually be below the aircraft, shielding this glow from the observer, as previouslyrecommended to reduce IR signature, will be an appropriate countermeasure.
The descriptions above provide some useful guidelines for the choice of configurationfor our aircraft for stealth, but this is not the only requirement to be considered.
8.2.3 Aerodynamic efficiency
For the specified mission, the aircraft will spend nearly all of the flight time at supersonicspeed. Therefore, it is important that the aerodynamic design concentrates on thereduction of wave drag. For a given size of aircraft, the longitudinal distribution of thecross-sectional area of the aircraft volume has a considerable influence on wave drag.Several aerodynamic and design textbooks (e.g. reference 4) describe the Sears–Haackanalysis. They show that a smooth progression (i.e. following a statistically normaldistribution) produces the minimum wave drag. The minimum increase in drag areadue to wave drag is calculated using the formula below:
(S × CDwave) = 14.14[Amax/L]2
where S = aircraft reference (usually gross) wing areaAmax = maximum aircraft cross-sectional areaL = aircraft overall longitudinal length less any constant section segments
“chap08” — 2003/3/10 — page 207 — #6
Project study: advanced deep interdiction aircraft 207
The equation shows that wave drag can be minimised by reducing (Amax) and increasing(L). However, care must be taken to avoid adding significantly to wetted area andthereby increasing parasitic drag.
Shaping the aircraft body profile to try to achieve the Sears–Haack distributionis called ‘area ruling’. Many existing aircraft exhibit this design strategy. It involves‘waisting’ the fuselage of the aircraft to reduce the cross-sectional area at the intersectionof the maximum wing profile and bodyside. This area ruling is shown clearly in the shapeof the Northrop F-5 aircraft. The most cited case in aeronautical history books for theadvantages of area ruling is that of the Convair F-102 which was originally designedwith a straight fuselage but could not achieve supersonic speed until the shape waschanged.
No practical aircraft can achieve an exact match to the Sears–Haack recommendationeven by area ruling. The ratio of the actual wave drag to the minimum drag from theSears–Haack prediction can vary from about 1.2 to 3.0. The lowest value would relate toa pure blended body with an area-ruled profile. The higher value would be for an aircraftnot principally designed to minimise wave drag (e.g. a supersonic fighter that needsgood pilot visibility and combat turn manoeuvrability). In this case, the main designdriver would probably be combat effectiveness. For our project aerodynamic efficiencyis paramount so every effort must be made to reduce wave drag. The adoption of ablended body configuration looks attractive. However, it must be realised that the maincontributor to wave drag is the value of maximum aircraft cross-sectional area (Amax).This term is squared in the Sears–Haack equation. The aircraft layout must focus onreducing this to the minimum value possible to hold the payload. Alternatively, or incombination, as the aircraft length is also squared, an increase will reduce wave drag.
At supersonic speeds, a Mach wave is formed which surrounds the aircraft. Theangle of this wave cone relative to the longitudinal axis of the aircraft is known as theMach angle (µ). This angle is a function of the aircraft forward speed (Mach number)5
such that:
µ = sin−1 (1/M) With the specified cruise speed of M1.6: µ = 38.7◦
To avoid discontinuity in airflow regions, it is desirable to keep the aircraft geometry,particularly the wing planform, within the Mach cone (i.e. keeping the wing leadingedge sweepback angle greater than (90−µ)◦). For our aircraft this dictates a wing lead-ing edge sweep angle greater than 51.3◦. As the air velocity in this region is substantiallylower than free-stream, this also reduces wave drag.
The wing planform will be designed to fit within the Mach cone therefore the wingspan will be restricted. This will increase lift induced drag but at the high cruising speedthe lift coefficient will be relatively low which will make induced drag less significantfrom this effect.
The wing section profile will need to be of the ‘supercritical’ type to reduce the strengthof shock in transonic flight. As the wing planform will be within the shock cone it wouldbe possible to use a rounded wing leading edge profile. This will improve low-speed liftgeneration over the wing especially if a leading edge flap is used. Effort must be madeto generate laminar flow over as much of the profile as possible to reduce parasitic drag.There is a potential conflict here between the preferred sharp wing leading edge profilefor minimisation of radar signature and the rounded profile for aerodynamic efficiency.A choice will have to be made.
The body will need to be contoured to suit the area ruling mentioned above. Inthe region of the cockpit there are conflicting requirements. A smooth cross-sectiondistribution in the forward part of the body may not provide the visibility requirements
“chap08” — 2003/3/10 — page 208 — #7
208 Aircraft Design Projects
demanded of a strike aircraft. Good pilot visibility is also an advantage for the landing.Systems, including artificial vision and computer controlled imagery, will offer scopefor innovation to overcome this problem in an aircraft designed for 2020. This aspectof layout and systems integration will require careful consideration.
8.3 Problem definition
The project description specifies a two-place advanced deep interdictor aircraft. Theentire long-range mission will be flown at supersonic speed. The exact mission definitionis shown in Figure 8.1. The long-duration, high-intensity flight conditions, much ofwhich is over enemy territory, demands the security of twin-pilot operation. The longwork periods and high manoeuvre load environment imposed on the pilots requirescareful design of the cockpit. The workload related to flight safety and weapon deliverymust be reduced by system design. Such systems must be made reliable and safe.
1
2
3
4
5
67
8
9
10
11
12
Take-off
Climb
Climb
Dash(in)
Dash(return)
Descend
Descend
Land (with reserves)
Manoeuvreturn
Supercruise(out)
Supercruise(return)
1–2 Warm-up, taxi and take-off Sea level NATO 8000 ft, icy 2–3 Climb to best supercruise alt. 3–4 Supercruise to conflict area Opt. alt. M1.6 1000 nm 4–5 Climb to 50 000 ft 5–6 Dash to target 50 000 ft M1.6 750 nm 6–7 Turn and weapon release 50 000 ft 180° 7–8 Dash out 50 000 ft M1.6 750 nm 8–9 Descend to supercruise alt. 9–10 Supercruise return 50 000 ft M1.6 1000 nm10–11 Descend to base11–12 Land (with reserve fuel*) NATO 8000 ft, icy
Segment Description Height Speed Distance/duration
*Diversion and hold at sea level with 30 min fuel at economical flight conditions.
Fig. 8.1 Mission profile
“chap08” — 2003/3/10 — page 209 — #8
Project study: advanced deep interdiction aircraft 209
The aircraft must be capable of ‘all-weather’ operation from advanced NATO andother bases. Aircraft shelter dimensions may impose configurational constraints onthe aircraft. Aircraft servicing and maintenance at austere operational bases demandminimum support equipment and skill. Easy access to primary system componentsmust be provided.
Closed-loop, static and dynamic stability and handling flight characteristics mustmeet established military requirements. A digital flight control system will be necessaryfor a longitudinal unstable aircraft configuration. All systems must be protected againsthostile damage and inherent unreliability.
In addition to strict stealth criteria, the AIAA problem description sets out severalrequired design capabilities and characteristics. These include:
• The aircraft must accommodate two pilots but should be capable of single pilotoperation. For such a long-range mission, pilot workload must be reduced by suitabledesign and specification of flight control and weapon delivery systems. Crew safetysystems must be effective in all flight modes.
• The design layout should allow for easy maintenance. Minimum reliance on supportequipment is essential for off-base operations.
• Structural design limit load factors of +7 to −3g (aircraft clean and with 50 per centinternal fuel) are required. An ultimate design factor of 1.5 is to be applied. Thestructure must be capable of withstanding a dynamic pressure (q) of 2133 lb/sq. ft(i.e. equivalent to (q) at 800 kt) and be durable and damage tolerant.
• All fuel tanks must be self-sealing. Aviation fuel to JP8 specification (6.8 lb/US gal)is to be assumed.
• Stability and handling characteristics to meet MIL-F-8785B subsonic longitudinalstatic margins to be no greater than +10 per cent and no less than −30 per cent.
• The aircraft must be ‘all-weather’ capable. This includes operation from and on toicy 8000 ft runways.
• The aircraft must operate from austere bases with minimum support facilities. Onthese bases the aircraft will be required to fit into standard NATO shelters.
• The flyaway cost for 200 aircraft purchase must not exceed $150 M (year 2000dollars).
In addition to the high-altitude, supercruising mission shown in Figure 8.1 anddescribed in section 8.2 above, the design specification sets the following manoeuvringtargets (specific excess power, SEP, is defined as PS in Chapter 2 (section 2.7.1)):
• SEP (1g) military thrust (dry), 1.6 M at 50 000 ft = 0 ft/s.• SEP (1g) maximum thrust (wet), 1.6 M at 50 000 ft = 200 ft/s.• SEP (2g) maximum thrust (wet), 1.6 M at 50 000 ft = 0 ft/s.• Maximum instantaneous turn rate, 0.9 M at 15 000 ft = 8.0◦/s.
(all the above performance criteria are specified at aircraft manoeuvre weight (definedas 50 per cent internal fuel with two AIM-120 and four 2000 lb JDAM)).
The design specification calls for five separate weapon capabilities:
• Four Mk-84 LDGP + two AIM-120.• Four GBU-27 + two AIM-120.• Four 2000 lb JDAM + two AIM-120.• Four AGM-154 JSOW + two AIM-120.• Sixteen 250 lb small smart bombs.
“chap08” — 2003/3/10 — page 210 — #9
210 Aircraft Design Projects
(the AIAA specification gave details of the size, weight and cost of all governmentfurnished equipment. This data is used in the layout, mass and cost estimations).
When details like those shown above are not provided with the initial specification,it is always necessary to spend time gathering the data before moving on to the nextstage. In this case, we are now ready to consider initial aircraft design concepts.
The details below suggest several potential design requirements:
• The field take-off requirement, particularly with regard to the icy runway conditionswill require a high thrust/weight ratio.
• Initial climb performance will require good specific excess power to reach thesupercruise altitude and speed in reasonable time.
• Supercruise will require low overall drag to give a good lift/drag ratio and thereby alower fuel requirement.
• The rear movement of the centre of lift in supersonic flight may require fuel transferto balance the aircraft and reduce trim drag.
• The climb from supercruise altitude to 50 000 ft for the dash phase may require aburst of afterburning to offset the low SEP at high/fast operation. Stealth may becompromised by either the use of afterburning or from the long-duration climb fromsupercruise altitude to dash without the extra thrust.
• The aircraft must be able to drop the weapons without significant trim changes.• The SEP requirements and the turn performance may require the use of manoeuvring
flaps although this may compromise stealth.• Landing will require low wing loading to avoid high approach speed and to reduce
aircraft energy on the ground.• Icy conditions may demand aerodynamic braking assistance (parachutes and lift
dumping).• Compatibility with NATO shelter size will limit the aircraft to a span of less than
20 m (65 ft) and length to less than 30 m (98 ft).
8.4 Design concepts and selection
Although initially many design layouts were envisaged, the three design conceptsdescribed below were selected for investigation.
• Conventional, straight wing• Pure delta/diamond• Blended delta
The conventional, tapered-wing layout (Figure 8.2) was selected as this offers lesstechnical risk to the project. The design processes for this layout are well understoodand the configuration can be easily developed for alternative roles.
The pure arrow-wing layout (Figure 8.3) results from considerations of stealth andaerodynamic efficiency. The main drawbacks of the diamond planform centre onthe unorthodox control arrangement and the difficulty of developing the layout toaccommodate alternative roles.
The blended arrow-wing configuration (Figure 8.4) can be regarded as either offeringthe best of the other options, or the worst of both types! The blended body can beconfigured to give lower wave drag than the straight wing and could be more easilydeveloped than the pure delta.
“chap08” — 2003/3/10 — page 211 — #10
Project study: advanced deep interdiction aircraft 211
Fig. 8.2 Design concept – conventional straight wing
Fig. 8.3 Design concept – delta/diamond
A decision matrix method was used to analyse the different options on a consistentbasis. The criteria used to assess the options in the selection process are listed belowtogether with (in brackets) the significance (weighting) to the overall assessment.
Effectiveness of incorporating stealth technology into the layout (5)Aerodynamic efficiency (mainly L/D ratio) of the layout (5)Potential for low-weight design (4)Technical difficulties (ease of analysis) and risk (3)
“chap08” — 2003/3/10 — page 212 — #11
212 Aircraft Design Projects
+ + +
Fig. 8.4 Design concept – blended delta
Field performance and rough ground handling (2)Maintainability and operational dependability (2)Survivability and ease of repair (2)Multi-role capability (1)
Naturally, the choice of criteria and the relative weightings is highly subjective but agroup response tends to smooth the assessment process. The result of the ‘voting’ onthe criteria above is shown below:
Conventional option (56), Delta/diamond layout (72), Blended body (58)
The necessity for high L/D ratio and improved stealth were the key issues in the selectionof the delta/diamond layout. It was also agreed that as much effort as possible shouldbe given to the use of blending the profiling of the body (as on the B-2 aircraft). It wasalso decided that an advantage would be gained if the aircraft length was increased.These issues led to changes in the original configuration. To reduce aircraft maximumsectional area and effectively lengthen the aircraft, tandem seating and tandem weaponstowage was employed. This resulted in the concept sketch shown in Figure 8.5.
The basic structural framework consists of a continuous (tip-to-tip) wing box. Theweapons and main landing gear are suspended below this and housed in profiled fairingswith radar reflective door and hinge edgings. Forward of the weapon bay, the profileis extended to accept the engine intakes which sweep up in S-bends to the top wingsurface. This duct-profiling protects the intake profile against radar reflections fromthe engine compressor face. It also ensures clean airflow to the engines with the aircraftat high-incidence attitude. The nose landing gear is retracted into the space betweenthe separate intake ducts. The twin engines are supported in cradles above the wing
“chap08” — 2003/3/10 — page 213 — #12
Project study: advanced deep interdiction aircraft 213
Fig. 8.5 Selected and revised concept sketch
structure. Nozzle exhaust ducts terminate forward of the wing trailing edge to shieldthe aircraft from downward infrared emissions. Fuel tankage is provided between theengine support cradles and intake ducting. The pilot and equipment bays are locatedin the aircraft centre line fuselage profile forward of the fuselage fuel tanks. The upperbody is profiled to blend smoothly into the wing surface and to give an advantageousSears–Haack volume distribution.
8.5 Initial sizing and layout
The initial sizing of the preferred configuration requires estimates of the main aircraftparameters. Instead of just guessing these values it is a good idea to investigate the values
“chap08” — 2003/3/10 — page 214 — #13
214 Aircraft Design Projects
Table 8.1
Parameter Fighters Strikers Bombers
Empty mass ratio (ME/MTO) 0.45–0.60 0.41–0.54 0.37–0.42Fuel mass ratio (MF/MTO) 0.21–0.33 0.17–0.33 0.40–0.62Payload ratio (MPAY/MTO) 0.21–0.28 0.18–0.37 0.14–0.19Wing loading (MTO/S) kg/sq. m 262–467 315–544 447–516Thrust/Weight ratio (dry) 0.65–1.29 0.56–0.88 0.26–0.40
associated with existing aircraft of the same type. It is possible to compile a list of designdata for existing military aircraft using published data.6 The problem with using thisapproach for our project is the unique nature of the specified mission requirements ofthe design. It does not follow the ‘fighter’ class of aircraft because of our need to flya longer range and carry a heavier weapon load than is normal for fighters. It doesnot fit into the ‘bomber’ class due to the higher speed and lower weapon load of ouraircraft. ‘Multi-role’ and ‘strike’ aircraft may have some comparable features but theseusually have much better manoeuvring ability and are not expected to supercruisefor long periods. Using data on appropriate military aircraft from reference 6 (withextreme values ignored), it is possible to assess the variation of some design parameters(Table 8.1).
It is clear from this analysis that there is wide variation in the aircraft used in thestudy. Also, as with all published data, the definition of aircraft parameters (e.g. emptyweight) may not be consistent from each manufacturer. The data therefore only providesa crude guide to the selection of parameters for use in the initial estimates. This impliesthat the initial estimates will be unreliable. It will be necessary to adopt a more refinedanalysis as quickly as possible.
Some thoughts about our design that might help us to select suitable starting values:
• Most of the aircraft in the survey are not supersonic in dry thrust so our design islikely to require a higher thrust to weight ratio than the bomber values.
• Travelling for long distance at supersonic speed will require more fuel than is seenin the fighter and strike classes above but not as much as the max. bomber (B-52)value.
• The fuel capacity required will be larger than on equivalent size aircraft so it may beadvantageous to have a larger wing area to provide extra tankage.
• A large wing size (low wing loading) will help in meeting the icy runway requirements.• The payload carried by our design, as defined in the specification, will give a relative
low useful load ratio and the range flown at supersonic speed will give a high fuelmass ratio.
• The empty mass ratio would also be reduced due to the large fuel mass but to accountfor the stealth requirement extra structure mass (radar absorbent materials) will berequired. With no better information these two effects will be assumed to cancel eachother, giving a conventional empty mass ratio.
With these thoughts in mind, our initial estimates are shown below:
• Empty mass ratio = 0.44(this is low for fighters but high for bombers).
• Fuel mass ratio = 0.46(this is outside the range for fighter/striker aircraft but about average for bombers).
“chap08” — 2003/3/10 — page 215 — #14
Project study: advanced deep interdiction aircraft 215
• Using the above assumption would make the payload ratio = 0.1(as predicted, this falls below all aircraft classes).
• Wing loading = 390 kg/sq. m (about 80 lb/sq. ft)(which is low for bombers, high for fighters and about average for strike aircraft).
• Thrust loading = 0.60(this is low for strike and fighter aircraft but it is not clear from the collected data howmany of the sample have quoted afterburning (wet) thrust. It is outside the range forbomber aircraft).
It is now possible to use the assumed values to make our first ‘rough’ predictions ofthe size of the aircraft:
• From the problem specification we can predict that the payload (including two crew)is 6600 kg (14 550 lb). As we assume above that this represents 0.1MTO, the aircraftmaximum take-off mass must be 66 000 kg (145 500 lb).
• With an empty mass ratio of 0.44MTO,the empty mass = 29 000 kg (64 000 lb).
• With a fuel mass ratio of 0.46 MTO,the fuel mass = 30 360 kg (67 000 lb).
• With a wing loading of 390 kg/sq. m = 3826 N/sq. m (about 80 lb/sq. ft),the gross wing area = 170 sq. m (1827 sq. ft).
• With a thrust loading of 0.6WTO,the total engine thrust (sea level, static, dry) = 388 kN (87 300 lb).This equates to 194 kN (43 700 lb) per engine.
This makes our aircraft heavier and larger than any of the fighter and strike aircraftsurveyed but much smaller than the existing bombers.
The diamond planform (area, S = 170 sq. m, 1830 sq. ft) which is limited in span(b) to 18.3 m (60 ft) (to keep within the hangar width) will have a centre line chord =(2S/b) = 18.6 m. For a symmetrical planform (90◦ at the tip) the wing sweep is onlyabout 45◦ and we must have at least 51.3 (see section 8.2.3). It is also advantageous tomaintain a long overall length to reduce wave drag. Both of these requirements can bemet by reducing aircraft span to 17 m (55.7 ft). In this case the centre line chord willbe increased to 20 m (65.6 ft). Providing a 90◦ angle between the leading and trailingedges at the tip gives a 60◦ wing leading edge sweep angle.
This geometry may need to be changed later in the design process if more fuel tankageis required.
Using the concept sketch (Figure 8.5) and the values above we can now produce ourfirst scale drawing of the aircraft (Figure 8.6).
8.6 Initial estimates
With an accurate drawing of the aircraft (Figure 8.6) it is possible to estimate thecomponent masses and drags (and lift). The predicted thrust will allow us to select asuitable engine or scale an existing design to provide engine performance data at allflight conditions. With mass, aerodynamic and propulsion data it will be possible toperform initial performance calculations and draw our first constraint diagram.
“chap08” — 2003/3/10 — page 216 — #15
216 Aircraft Design Projects
Fuel
O/length = 20 m
O/height = 4.8 mSpan = 17 m
Area (ref.) = 170 m2LE = 60°
Fuel
Fuel
Eng.
Eng.
Equip.
Cockpit
Eng.Cockpit
Intakes
Eng.
10 m
30 ft
Thrustvector
Fig. 8.6 Initial baseline aircraft general arrangement
8.6.1 Initial mass estimations
The initial mass estimates can be calculated by using published empirical equationsbased on existing aircraft designs.4 As our aircraft has a unique operating envelope,such methods may be regarded as crude. At this stage in the design process, the analysisis likely to be more accurate than the ‘guesstimates’ made from the survey used above.Using our knowledge of the aircraft specification, some corrections to the method canbe applied. All the required input data for the method can be gleaned from the initiallayout drawing, the project specification and common sense. Applying such data to theequations in reference 4 gives the mass statement shown in Table 8.2.
This initial estimate of MTO is substantially less than previously predicted. Themain reason for this reduction is due to the lower prediction of aircraft empty mass.Although, as expected, the propulsion system mass is large due to the high thrust
“chap08” — 2003/3/10 — page 217 — #16
Project study: advanced deep interdiction aircraft 217
Table 8.2
Mass
Component (kg) (lb) % MTO
Wing (inc. controls) 1 888 (4 163) 3.0Body (inc. engine cowls) 7 070 (15 589) 11.3Undercarriage (all units) 1 138 (2 509) 1.8
Total structure 10 096 (22 262) 16.1Propulsion system 12 077 (26 630) 19.2Fixed systems 3 673 (8 100) 5.8
Aircraft empty 25 846 (56 990) 41.1Crew (two pilots) 500 (1 100)Weapons 6 100 (13 450)
Zero fuel mass 32 446 (71 543) 51.7Fuel∗ 30 360 (66 944) 48.3
Max. take-off mass 62 806 (138 487) 100.0
∗The fuel load is retained at the value estimated from the higher MTO massoriginally predicted. This will need to be checked when the mission analysisis completed
to weight ratio, the aircraft structure and fixed systems masses are low. This couldhave been expected as the compact and stiff structure framework will provide a lightstructure. However, for our high-tech, modern weapon system, the low systems massmust be treated as suspicious. As the project develops, and more detail is known aboutthe aircraft systems, it will be necessary to reassess this estimate.
As the aircraft empty mass estimation was based mainly on the original value ofMTO it is expected that the aircraft mass and size could be reduced. Before any changesare contemplated, it is advisable to continue with the aerodynamic and performanceestimations using the original design. In this way, all the design modifications can beassessed at the end of the initial estimation process.
8.6.2 Initial aerodynamic estimations
The initial aerodynamic estimations concern the prediction of aircraft drag and lift.For this aircraft the main focus of drag will be on the supersonic wave drag (CDw)estimation. Using the wave drag equation in reference 4, with the following inputvalues, gives the first estimation of CDw:
Aircraft cruise Mach number, M = 1.6Aircraft max. cross-sectional area, (Amax) = 10.06 sq. mReference wing area, Sref = 170 sq. mWing LE sweep = 60◦Aircraft overall length (less any constant sections), L = 20 mAn adjustment factor to relate the actual cross-section distribution to the Sears–Haack perfect shape, EWD = 1.4 (assuming a smooth distribution from theblended body)
Gives, C Dw = 0.02104
This is a very large drag increment that will substantially penalise the design. Somehow,we will need to either reduce the cross-sectional area or increase the aircraft length.The area cannot be changed significantly unless we alter the internal requirements. It
“chap08” — 2003/3/10 — page 218 — #17
218 Aircraft Design Projects
is relatively easier to increase the length (see later aircraft drawings). Assuming that itis possible to stretch the aircraft to 28 m (92 ft) the calculation above would change to:C Dw = 0.01408.
The parasitic drag will be estimated by using an equivalent skin friction coefficientof 0.0025 (representative of a smooth fast transport aircraft).
Hence, with an estimated aircraft wetted area of 400 sq. m (4300 sq. ft),
CDO = 0.0025 (400/170) = 0.00588
This gives a total ‘clean aircraft’ zero-lift drag coefficient of (0.02692), in cruisecondition.
At the start of the initial dash phase, the aircraft weight will be less than the take-offvalue due to the fuel used during take-off, climb and supercruise. As we do not knowthe fuel used yet we will assume that weight is at 80 per cent of the take-off value:
66 000 × 9.81 × 0.8 = 518 kN (116 450 lb)
Therefore, the cruise CL = 518 000/(0.5 × 0.1864 × [295 × 1.6]2 × 170) = 0.147From reference 4, at M1.6, the induced drag factor (K ) = 0.3Hence, induced drag coefficient, CDi = 0.3 × 0.1472 = 0.00648
Therefore the total drag coefficient at start of cruise = 0.02644Hence, drag at 50 000 ft and dash speed of M1.6,
= 0.5 × 0.1864 × (295 × 1.6)2 × 170 × 0.02644 = 93.3 kN (20 982 lb)
Hence, the lift to drag ratio will be (518/99.5) = 5.56
The reciprocal of (L/D) is equal to the (T/W ) required in the initial dash. In this case(1/5.56) = 0.18. This has to be multiplied by the engine thrust lapse rate appropriateat the cruise condition (height and speed) and the weight reduction, to obtain anequivalent static, sea-level (SSL) value. Although a high bypass ratio engine wouldhave a cooler exhaust temperature which would give a lower IR signature, it wouldbe substantially larger. This would make the aircraft much bigger which would beless stealthy in other ways. Therefore, the engine we will select will be of the lowerbypass type. For this type of engine the effect of speed on thrust lapse rate can beignored for initial estimates. Reference 7 quotes the following expression to determinethe lapse rate:
Thrust at altitude/SLS thrust = σ x
where SLS denotes sea-level static condition, σ is the relative ambient air density and theexponential x has the value of 0.7 in the troposphere and 1.0 in the lower stratosphere.
At 50 000 ft the lapse rate is (0.428 × 0.51 = 0.22)As above, the weight ratio (Wdash/WTO) is 0.8Therefore, to achieve a cruise T/W of 0.18, requires an SLS value of 0.8 (0.18/0.22) =
0.654This is higher value than the 0.6 value originally assumed
The above calculations have highlighted a potential problem area for the design. Thehigh drag in cruise reduces the aircraft L/D ratio which will have a direct effect on the
“chap08” — 2003/3/10 — page 219 — #18
Project study: advanced deep interdiction aircraft 219
fuel required to complete the mission. The high T/W value will require a larger engineand corresponding propulsion system. Both of these effects will seriously compromisethe effectiveness of the design. Hence, it is important to reduce the aircraft drag.
To illustrate the overall effect, if we could reduce wave drag by 20 per cent, a similaranalysis to that above would yield:
Drag in cruise = 83.0 kN (18 668 lb)
Lift/Drag at the start of cruise = 6.24
SLS thrust/weight = 0.58
This looks more encouraging but a 20 per cent drag reduction may be difficult to achievewithout a significant changes in aircraft layout. Obviously, the influence of wave dragis paramount in the drag estimation. For example, if the (Amax) value could be reducedfrom 10.06 to 9.8 and the length increased from 28 to 30, the wave drag coefficient wouldlower to the 20 per cent required. This shows the significance of the parameters input tothe equations and the need to carefully assess the values used. From the standpoint oflayout, it is clear that in reviewing the existing design we must reconfigure the aircraftshape to reduce Amax and increase vehicle length, if this is possible.
In the evaluation of wave drag, the equation shows a direct proportionality to thefactor (EWD) which, as defined above, relates the actual aircraft longitudinal cross-sectional area distribution to that of the perfect Sears–Haack distribution. A valuetypical of fighter aircraft optimised for supersonic flow has been assumed (i.e. 1.4). Itis impossible to achieve a value of 1.0 with realistic shapes but it may be possible4 toachieve 1.2 for an optimum blended-fuselage, delta-wing configuration. Our aircraftlayout fits into this category so we could consider reducing the originally assumedvalue. Lowering the 1.4 value to 1.2 generates a 14 per cent reduction in wave drag. Thegraph for cross-sectional volume for our current configuration is shown in Figure 8.7.
To avoid the profiling problems at the front of the aircraft associated with pilotwindscreen and canopy shaping, artificial vision systems are proposed for the cockpitlayout. This means that the cockpit could then be positioned away from the noseprofile. With a reduction of the maximum cross-sectional area it should be possible toreduce the cruise drag coefficient to 0.0232, giving L/D = 6.34 and (T/W )TO = 0.58.Although this may be optimistic, it does represent a good initial estimate as it indicatesthe direction that future design decisions must take.
Drag of the aircraft in other flight conditions and in different configurations must alsobe estimated. The most significant of these are the take-off and landing phases. In addi-tion to the aircraft clean condition we must add landing gear, flap and any aerodynamic
Aircraft length
Nose Tail
Canopyand
cockpit
Sears–Haack idealdistrubution
Rear engineinstallation
Fig. 8.7 Sears–Haack cross-sectional area distribution
“chap08” — 2003/3/10 — page 220 — #19
220 Aircraft Design Projects
retarding devices (e.g. lift dumpers and braking parachutes). In this initial stage it is suf-ficient to make sensible guesses, in terms of ‘drag area’ (drag/dynamic pressure = D/q)for each addition.
For the undercarriage (including interference effects) we will assume D/q = 0.5 sq. m.For our reference area of 170 sq. m this gives a �CDO of (0.5/170) = 0.003.
At this stage in the evolution of the aircraft it is not known if conventional flapswill be required on the aircraft. Flaps may only be required for landing as at take-offthe afterburner (if fitted) could be used. To assess the effect of flaps on aircraft drag,plain flaps with only 20◦ deflection will be assumed. (If it is possible to avoid flaps, theinboard trailing-edge surfaces could act as pitch control surfaces.) We will assume D/qfor flaps if used to be 2.0 sq. m, giving �CDO = 0.0118. In later drag estimations it willbe necessary to take into account changes in drag from the cruise condition. These arisefrom the reduced Reynolds number at the lower speeds in take-off and landing phases,and the effect on lift-induced drag due to the disturbed spanwise lift distribution causedby the flap lift. At this time, we can assume the clean drag coefficient is unchanged fromthe cruise value shown above, i.e. 0.00589. At take-off, we will assume that there is noflap deflection. The zero lift drag coefficients, based on the reference area of 170 sq. m,is shown below:
CDO for take-off = 0.00889 and CDO for landing = 0.02069
Using reference 8, the position of the aerodynamic centre can be calculated. For ouraircraft (λ = 0, LE sweep = 60◦, wing aspect ratio = 1.7) the supersonic position(Xac/Croot) is 0.55 and subsonic 0.45. As this aircraft will spend most of the missionat supersonic speed, it would be sensible to trim the aircraft for this condition. Theforward movement of the lift in the take-off and landing phases will need to be balancedby fuel mass transfer or control surface trim forces.
For the determination of the lift capabilities of the aircraft there are two principalcharacteristics; the lift curve slope and the maximum lift. For any aircraft configura-tion these are notoriously difficult parameters to predict accurately. In practice, manyaircraft have required ‘fixes’ after flight tests, to correct lift characteristics that were notpredicted.
The prediction of the lift curve slope is required to determine the best (low drag) anglebetween the fuselage and wing (this is not necessary on our blended body layout). It isalso used in the prediction of drag due to lift, and in the stability and control analysis.When more accurate geometrical information on the aircraft layout is available, itwill be possible to use computational fluid dynamic (CFD) methods to provide moreaccurate estimations. At this stage expending such effort would be inappropriate as theaircraft shape will be under continuous revision. For a bi-convex section of aspect ratioless than 2, the shape of the CL versus angle of attack (α) graph is shown in Figure 8.8.
For moderately swept, high aspect ratio wings (typical of transport aircraft) theCLmax of the unflapped wing will be close to the infinite span (2D) aerofoil value butour aircraft is not of this planform. For highly swept, very low aspect ratio planforms,the airflow over the wing surfaces will be significantly affected by vortex generationover the leading edge. These leading edge vortices add both lift and drag and ensurethat the flow over the upper surface remains attached well above the normal stall angleof higher aspect ratio trapezoidal planforms. This vortex formation (see Figure 8.9) willbe most prominent for wings with a sharp leading edge. These conditions are expectedto be found on our wing planform.
Vortex-generated lift will stay attached to the wing up to the point of vortex burst athigh angles of attack. The traditional stall characteristic by which we predict maximum
“chap08” — 2003/3/10 — page 221 — #20
Project study: advanced deep interdiction aircraft 221
10°α
0.5
1.0
1.5
20° 30°
CL
TheoryExp
eriment
Source:McCormick10
Aspect ratio 1.5
Without vortex lift
Fig. 8.8 Section lift coefficient versus angle of attack
Source:McCormick10
Angle ofattack
V
Spiral vortexsheet aroundleadingedge
Fully developed spiralvortex flow
Fig. 8.9 Vortex induced flow
lift capability is not appropriate in this situation. As shown on Figure 8.8, the max. liftcoefficient will not be reached until exceptionally high nose angles have been pulled.When the aircraft is on or close to the ground (e.g. on take-off and landing) such highangles will cause the aircraft rear fuselage structure to scrape the runway. The geometryof the aircraft will limit the max. attitude to about 15◦. At this angle Figure 8.8 showsthat the CL is approximately 0.52. Therefore, the limit of lift generation on or close tothe runway will be set by the aircraft tail scrape angle of 15◦. For flight away from theground, the max. lift coefficient will be set by the limit of controllable angle of attack.
8.7 Constraint analysis
As the initial mass (weight) and aerodynamic estimates have now been made, it ispossible to conduct a constraint analysis to determine if the original choice of thrustand wing loading values are reasonable. As these were derived from data on otheraircraft, it is likely that a better selection can improve the design. This process will alsoindicate which of the constraints on the problem are most critical.
“chap08” — 2003/3/10 — page 222 — #21
222 Aircraft Design Projects
The equation below, as developed from the specific excess power relationship inChapter 2, is the general form of the constraint function:
(TSSL/WTO) = (β/α)[{(q/β)(CDO/(WTO/S)} + {[k1 · n2 · (WTO/S)]/(q/β)}+ (1/V ) · dh/dt + ([1/g] · [dV/dt])
In order to draw the constraint diagram (thrust versus wing loading) it is necessary todetermine the values of the coefficients (etc.) to be used in the equation.
These are defined below:
List 1TSSL = engine static sea-level thrustWTO = aircraft take-off weightS = aircraft reference wing areaW = aircraft weight at the condition under investigationT = engine thrust at the condition under investigation
List 2β = aircraft weight fraction for the case under investigation = (W/WTO)α = thrust lapse rate at the altitude and speed under investigation = (T/TSSL)q = dynamic pressure at the altitude and speed under investigation = (0.5ρV 2)V = aircraft speed at the condition under investigationh = aircraft altitude at the case under investigationρ = air density at height hCDO = aircraft zero-lift drag coefficientk1 = aircraft-induced drag coefficientn = aircraft normal load factor = L/Wdh/dt = aircraft rate of climb at the case under investigationg = standard gravitational acceleration = 32.2 ft/s2 (or 9.81 m/s2)dV/dt = aircraft acceleration at the case under investigation
For each constraint case, the analysis requires all the values for the parameters in thesecond list above to be substituted into the equation for (TSSL/WTO) above. Selectedvalues of wing loading (WTO/S) are then used to determine corresponding values forthrust loading (TSSL/WTO). These values are then plotted to indicate the constraintboundary for the case. This process is repeated for all constraints.
In the design proposal, there are several performance requirements:
• Take-off from 8000 ft (2440 m) runway, on standard day with icy runway.• Climb to optimum supercruise altitude.• Supercruise at optimum altitude at M1.6 for 1000 nm (less climb distance).• Dash at M1.6 at 50 000 ft (min.).• Manoeuvre with specific excess power (SEP), at specified weapon load and 50 per cent
fuel:– at 1g, M1.6, alt. = 50 000 ft with SEP = 0 ft/s with no afterburning– at 1g, M1.6, alt. = 50 000 ft with SEP = 200 ft/s with afterburning– at 2g, M1.6, alt. = 50 000 ft with SEP = 0 ft/s with afterburning
• Land onto 8000 ft runway, on standard day with icy runway.
Before the analysis can be made there are several assumptions that must be made:
• Take-off from icy* conditions will be with afterburning (called maximum thrust).• Take-off in normal conditions will be with no afterburning (called military thrust).
“chap08” — 2003/3/10 — page 223 — #22
Project study: advanced deep interdiction aircraft 223
• As some of the constraints are related to military thrust, it is necessary to define theincrease in thrust from afterburning. We will initially assume (Tmax/Tmil) = 1.5.
• Initial climb to supercruise with final rate of climb of 1000 fpm (our requirement).• Supercruise starts with 90 per cent MTOM.• Dash starts with 80 per cent MTOM.• Manoeuvres are at aircraft mass empty + crew + weapons + 50 per cent fuel
(25 846 + 500 + 4000 + 15 180 = 45 526 kg (100 385 lb)).Basing all of the constraint analysis on our original mass estimate of66 000 kg (145 530 lb) gives βmanoeuvre = (W/WTO) = 0.69.
• Landing approach speed less than 160 kts (82 m/s) at 95 per cent MTOM.• Landing on an icy∗ runway with fuel dumping and possibly emergency braking
parachute.• Landing in normal conditions will be determined at 95 per cent MTOM with
emergency braking (µ = 0.5).
(∗Operation from icy runways requires directional control that is not reliant ontyres.)
Aerodynamic surfaces and engine thrust mechanisms are the only alternatives.Lateral thrust vectoring will be available for take-off but not for landing. Operationfrom icy runways may be difficult unless other solutions can be found.
The last three constraints dictate maximum vales for (W/WTO). The approach speedis only affected by the aircraft minimum speed. As we will not have a reverse-thrustcapability on the aircraft, the landing distance calculations will be independent ofengine thrust. The appropriate calculations are shown below and the results plotted inFigure 8.10.
0.3
0.4
0.5
0.6
0.7
0.8
Thru
st lo
adin
g (S
SL)
2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000
Wing loading (N/sq. m)
0.9
1
1.1
F-117
F-16XL F-15EX
F-14DX
Normal landing
Approach speed
Originaldesignpoint
Revised design point
B-2
Normal TO
Concorde Supercruise
Dash + Climb
F-22
F-23
F-32
Manoeuvre 200 ft /s
100 ft /s
150 ft /s
Fig. 8.10 Constraint diagram
“chap08” — 2003/3/10 — page 224 — #23
224 Aircraft Design Projects
(a) For the approach speed
(WTO/S) = 1/β{0.5ρ(Vapproach/1.2)2CLapproach}
where β = 0.95ρ = 1.225 kg/m3
Vapproach = 82 m/sCLapproach = 0.52
We are assuming that the approach speed is 1.2Vstall. This is slower than normal.This gives (WTO/S) = 1566 N/sq. m (max.) (For reference 1000 N/sq. m =20.9 lb/sq. ft.)
This is much too low. It will create a large wing area which will be inefficient in thecruise phases. (For reference, the initial estimate for wing loading is 3880 N/sq. m.)It will be necessary to generate more CL from the wing. The value used above wasconsistent with an unflapped delta wing limited to a maximum angle of attack of15◦. For an aircraft of our layout it may be possible to adopt a high angle of attack(HAA) approach (see Figure 8.11) as demonstrated by the X31 vector technologydemonstrator.9 This uses an HAA to provide a slow speed flight at low decent rate formost of the approach. Obviously, the aircraft must be stable in such a flight attitudeand must be capable of maintaining its heading. As the aircraft gets near to the groundthe angle of attack is raised to the maximum value to slow the aircraft. Just priorto the tail scrape, the incidence is rapidly reduced (nose-down). This will cause theaircraft to effectively have a controllable crash landing onto the runway threshold.This manoeuvre will demand an extra strong landing gear to withstand the high loadsrequired to absorb the vertical energy. This flight profile requires automation as thepilots will not be capable of reacting to such a landing manoeuvre. (Most large civilaircraft landings are automatic these days, although not like this profile!)
Touchdownrollout
Runway threshold
Rapidde-rotationand level-out
Curved slow speedlanding trajectory
Max. HAAstabilisedaltitude
HAA approachwith low decent
rate
Transition tohigher angle of
attack
Source: S. W. Kandebo9
Fig. 8.11 High angle of attack approach profile
“chap08” — 2003/3/10 — page 225 — #24
Project study: advanced deep interdiction aircraft 225
As we have already decided that artificial vision and automatic landing systems wouldbe incorporated into the aircraft to avoid the forward cockpit profiling, the unusualaircraft attitude should not present a problem. Increasing the angle of attack to 35◦would raise the CL to 1.3. The wing could be fitted with a leading edge (vortex) flap toincrease CL to 1.5. Even when not deployed the additional mechanisms and systemsneeded to deploy the flaps would affect the stealth image of the aircraft so may not bedesirable. Vortex flaps will not be included but will provide some insurance if the flightprofile is seen in flight tests to require extra lift capability.
We could also assume some fuel dumping or burn-off before landing. This wouldreduce β to 0.8.
These changes would increase the maximum wing loading (N/sq. m) as shownbelow:
(a) baseline = 1566,(b) with HAA profile = 3914,(c) HAA plus fuel dumping = 4648.
The aircraft conditions to be adopted will be decided when all the constraints havebeen assessed.
(b) For normal landing
(WTO/S) = (sL · ρ · CLlanding · g · µ)/(1.69 · β)
where sL = available runway length = 2440 m (8000 ft)CLlanding = 0.52 (see above)µ = 0.5β = 0.95
This gives a maximum value of (WTO/S) = 4748 N/sq. m.Note that an approach speed of 1.3 times minimum speed has been assumed above
(factor 1.69). This is typical of conventional aircraft to protect from stall due to suddenchanges in atmospheric conditions. As the delta planform flying at 15◦ angle of attackis well away from the max. lift angle it may be argued that this factor could be ignored.If so, the maximum value of wing loading would be 8024 N/sq. m.
(c) For icy landingThe same formula as above is applicable if braking parachutes (etc.) are not used withinput values of:
sL = available runway length = 2440 m (8000 ft)CLland = 0.52µ = 0.1β = 0.95Giving: (WTO/S) = 950 N/sq. mOr without the factor = 1605 N/sq. m.
These are obviously too low, therefore extra retardation is required. As we are likely toneed thrust-vectoring and afterburning on the engine, it is unfeasible to expect thrustreversal to be available. Braking parachutes, air brakes, runway-retarding devices andice removal offer some possibilities. As all of these devices complicate the analysis, itis not appropriate to get too involved in detail design at this early stage in the designof the aircraft.
“chap08” — 2003/3/10 — page 226 — #25
226 Aircraft Design Projects
(d) Normal take-offFor take-off conditions the constraint equation reduces to:
(TSSL/WTO) = [(1.44 · β2)/(α · ρ · CLto · g · sTO)] · (WTO/S)
where sTO = available runway length = 2440 m (8000 ft)α and β = 1.0CLto = 0.52
As the equation is a straight line through the origin, it is only necessary to evaluateit for one value of wing loading. For (WTO/S) = 5000 N/sq. m, giving (TSSL/WTO) =0.472.
The same argument as outlined above for landing can be made for the avoidance ofthe 1.44 factor in the take-off equation. In this case, the (TSSL/WTO) reduces to 0.328.
(e) For icy take-offThe calculation requires the estimation of the balanced field length using the maxi-mum thrust for the flight condition but not for the braking condition to determine thedecision speed. The braking part of the calculation involves the same difficulties asdescribed in the icy landing description above. As with landing, it is too early in thedesign process to perform these calculations in sufficient detail. We will need to returnto this subject later in the design process.
(f) Supercruise at optimum altitudeFor a parabolic drag polar the condition for maximum range can be shown4 to be:
CDo = (3 · k1 · C2L)
For our aircraft:
CDo = 0.01996 and k1 = 0.3 Hence, CL for max. range is 0.149
Using the definition of lift:
L = W = 0.5 · ρ · V 2 · S · CL
With W = 0.9 · 66 000 · 9.81, V = M1.6 = 1.6 · 295 = 472 above 11 000 m, S =170 sq. m, gives ρ = 0.2295. From ISA tables this density occurs at 14 000 m (46 000 ft)altitude, this is the initial supercruise height. This calculation involves the initial guessfor the wing loading (i.e. 3808 N/sq. m). The equation above can be solved in termsof other values for wing loading to indicate the sensitivity of (WTO/S) against initialoptimum altitude:
Wing loading (N/sq. m) 3 000 4 000 5 000Optimum altitude (m) 16 000 13 000 11 600Optimum altitude (ft) 52 500 42 000 38 030
As fuel is used and the aircraft gets lighter the wing loading will reduce and the optimumcruise height will rise. On the return supercruise phase (and for the dash manoeuvres)when the aircraft is lighter the cruise height will be increased providing that the enginethrust is large enough to reach these altitudes.
“chap08” — 2003/3/10 — page 227 — #26
Project study: advanced deep interdiction aircraft 227
Artificially fixing the supercruise height for the initial calculation at 14 000 m it ispossible to determine the relationship of thrust to wing loading using the constraintequation above. The result is shown below (assuming β = 0.9):
Wing loading (N/sq. m) 2000 3000 4000 5000 6000 7000Thrust loading (TSSL/WTO) 0.905 0.656 0.548 0.496 0.473 0.465
(g) Initial climb to supercruise altitudeThe required thrust to achieve the supercruise condition, as calculated above, mustinclude sufficient climbing ability at the start of cruise. The minimum for this typeof aircraft is 1000 ft/min (5.08 m/s). The thrust loading to give this rate of climb iscalculated by the climb term [(1/V )·dh/dt] in the constraint equation, suitably adjustedto the take-off condition (i.e. multiplied by β/α). Hence,
�(TSSL/WTO) = [(1/V ) · dh/dt] · β/α = (1/472) · 5.08 · (0.9/0.3) = 0.0323
(h) Dash at 50 000 ft altitudeThis is similar to the supercruise case except that the starting mass will be lowerdue to the fuel used in the previous sector. We will assume β = 0.8. The calcula-tion is performed with and without the climb requirement. The results are shown inFigure 8.10.
(i) ManoeuvresThere are three separate manoeuvres that have to be investigated (as described insection 8.3 as cases (a) to (c)). The constraint equation is used with the afterburningthrust ratio (1.5) for cases (b) and (c).
Case (a) is similar to the initial dash phase described above except that the aircraftweight is lower (β = 0.69). This will make it uncritical and therefore not worthinvestigating for the constraint analysis.
Case (b) is very critical as shown by the results plotted in Figure 8.10. This require-ment overpowers all other constraints and will solely dictate the aircraft layout. Foraircraft design this is an undesirable situation and calls into question the validity ofthis requirement. The specified climb rate of 200 ft/s (12 000 fpm) at the high altitudeand high weight may be desirable for avoidance of threats but seems excessive in viewof the stealth characteristics of the aircraft. It would be sensible to discuss this prob-lem with the originators of the RFP to establish how ‘firm’ they are on retaining therequirement. Requirements often fall into two categories: ‘demands’ and ‘wishes’. Partof the constraint analysis is concerned with distinguishing between these two types forthe critical design requirements. To assist with the discussion it is worth showing thesensitivity of the climb requirement by performing the analysis for different values; inthis case, for 100 and 150 ft/s. These extra cases are shown on Figure 8.10. The 100 ft/scase seems to offer the most ‘balanced’ design and still provide a respectable 6000 fpmclimbing ability.
Case (c) is similar to case (a) but with the normal acceleration value (n) increased to 2,and with afterburning applied. As seen on the constraint diagram the case fits well withthe other requirements.
8.7.1 Conclusion
The constraint analysis has shown that, in general, the aircraft requirements are wellbalanced. The exceptions to this optimism are concerned with the manoeuvre climb
“chap08” — 2003/3/10 — page 228 — #27
228 Aircraft Design Projects
requirement and the airfield performance onto icy runways. Both of these present prob-lems for the design. As discussed above, the climb requirement should be reduced to100 ft/s. In the following work we will assume that this concession has been made bythe customer. Operation from and onto icy runways is not avoidable so some extraretardation systems will have to be introduced or some other possibilities consid-ered. Incorporating reversed thrust into the already complex, engine-nozzle systemappears to be unfeasible. Braking parachutes will have to be used together with areduction in the touchdown speed to lower the energy to be dissipated. The best solu-tion would be the installation of an arrester-hook on the aircraft and some form of wirepick-up on the runway for those airstrips that are susceptible to icing. Such a conceptis outside the remit for our design.
It should be remembered that constraint analysis is a very crude process. It is basedon potentially inaccurate data that has been generated from the initial ‘guesstimates’of mass, aerodynamic and propulsion values and characteristics. Nevertheless, it offersthe first tests of the initial layout and provides a direction to first revision of the aircraftgeometry.
8.8 Revised baseline layout
The most efficient aircraft layouts on the constraint diagram are those with lower valuesfor thrust loading and higher values for wing loading. The original design point was setat a wing loading of 3826 N/sq. m and thrust loading of 0.6. From Figure 8.10, it is seenthat this point violates the modified manoeuvre constraint. Moving to 4500 N/sq. mand 0.58 brings the design into the feasible region. This reduces the wing area by about15 per cent and the engine by about 4 per cent. This should result in a reduction of theaircraft MTOM.
Using the detailed mass estimate calculated earlier (section 8.6.1) and assuming asaving of 2000 kg in empty mass (about 8 per cent) to reflect the new aircraft parame-ters above, provides an initial value of MTOM of 60 806 kg (134 077 lb).
This makes the new wing area = (60 806 × 9.81)/4500 = 133 sq. m approx.(i.e. 1425 sq. ft)The static sea-level military thrust (both engines) = (60 806 × 9.81) × 0.58 = 346 kN(77 782 lb)SSL thrust per engine = 173 kN (38 900 lb)
It is now possible to modify the original aircraft general arrangement drawing and tomake some detailed estimates for the aircraft mass, aerodynamic characteristics andengine performance.
8.8.1 General arrangement
The new general arrangement drawing of the aircraft forms the basis of the input to thedetailed technical analysis for the next stage of the design process. The basic layout ofthe aircraft will not be changed from that devised previously but some of the principaldimensions will be different. We now realise the importance of reducing the aircraftmaximum cross-sectional area and lengthening the ‘fuselage’. This will be achieved bystretching the planform.
“chap08” — 2003/3/10 — page 229 — #28
Project study: advanced deep interdiction aircraft 229
Table 8.3
TE sweep (forward)
LE sweep 30 25 20 15 10
60 15.18∗ 15.56 15.93 16.31 16.7017.53 17.10 16.70 16.31 15.93
65 13.98 14.27∗ 14.56 14.85 15.1419.03 18.63 18.27 17.91 17.57
70 12.66 12.87 13.08∗ 13.28 13.4921.03 20.67 20.34 20.03 19.72
30°
30°
20°
10°
LE sweep60°
LE sweep70°
17.53 m15.18 m
21.03 m12.66 m
20.34 m13.08 m
19.72 m13.49 m
Centre line chord16.70 m
span15.93
15.93 m16.70
20° 10°
Fig. 8.12 Planform shapes
(a) WingFor a diamond wing planform of a specified wing area (133 sq. m) it is possible to applysimple geometry to determine the span and centre line chord for various leading andtrailing edge sweep angles (degrees) (Table 8.3).
In Table 8.3, the upper value is the wing span and the lower the centre line chord.Both values are in metres. The ∗ values represent a wing tip angle of 90◦. It is alwayswise to visualise geometric data. To appreciate the wing planform shapes the optionsin the table are drawn in Figure 8.12.
There are several considerations to take into account in making a choice ofplanform:
• To reduce wave drag a long centre line chord is desirable.• Less TE sweep makes the trailing edge controls more effective.
“chap08” — 2003/3/10 — page 230 — #29
230 Aircraft Design Projects
• Less TE sweep pushes the centre of lift further aft. This could lead to aircraft balanceand trim problems and demand larger control surfaces.
• A long centre line chord will give a deeper wing profile (for a given profile thicknessratio).
• A long centre line chord will reduce wing span and thereby reduce roll control butimprove roll inertia.
Such concerns require a number of design compromises to be made. At this stage inthe design process, the sensitivity of the control responsiveness and the aircraft balanceissues are unknown. As our previous analysis has highlighted the need to reduce wavedrag, our choice will be towards this aspect. We will select the 70◦ sweep with the 90◦wing tip angle. Rounding the exact values shown above gives:
Wing span = 13 m, Centre line chord = 20.5 m, Wing area = 133.2 sq. mWing span = 42.6 ft, Centre line chord = 67.2 ft, Wing area = 1432 sq. ft
(b) WeaponsIn order to arrange the fuselage shape it is necessary to identify the requirementsfor weapon storage. All the weapons are carried internally to reduce drag and radarreturns. The design brief defined the type and combination of weapons to be carried.There are two categories of weapon listed: air-to-surface munitions and an air-to-airdefence missile. The latter is the only form of self-defence on the aircraft. There arefive different types of munitions specified:
• general purpose (GP) guided bombs,• cluster bomb units (CBU),• direct attack penetrators (JDAM),• stand-off weapons (JSOW),• small smart bombs (SSB).
The air-to-air missile is the AIM-120/AMRAAM which is commonly carried on otherUS military aircraft.
Descriptions and dimensions for the weapons are easily found in the aeronauticalpress6 and web sites.
Most of these weapons are already used on other aircraft but some are externallymounted, therefore some detailed modifications for internal storage will be required.
The largest weapon in the list (Table 8.4) is the GBU. This will define the requiredweapon bay dimensions. The internal measurements will depend on the choice betweenfour abreast or two abreast layouts (m/ft):
Layout Length Width Depth4 abreast 4.7/15.4 3.1/10.2 0.76/2.52 abreast 9.5/31.1 1.6/5.25 0.76/2.5
(c) LayoutMaking some assumptions with regard to the engine size and installation, it is nowpossible to draw the revised baseline general arrangement (GA). This is shown inFigure 8.13.
To avoid the possibility of unstable flow conditions at the apex of the delta wingplanform, a fuselage extension has been added (agreeing with the assumed lengthincrease assumed in section 8.6.2). This will provide a separation of the airflow atthe nose of the aircraft, it will add length which will reduce wave drag, and it will
“chap08” — 2003/3/10 — page 231 — #30
Project study: advanced deep interdiction aircraft 231
Table 8.4
Weapon name Guidance Number Size, length × dia. (m/ft) Configuration
GBU-27 Laser 4 4.7/15.4 × 0.76/2.5 2 × 2Mk-84 LDGP None 4 3.6/11.8 × 0.45/1.5 2 × 2JDAM GPS 4 3.0/10.0 × 0.50/1.6 2 × 2AGM-154 JSOW Internal 4 4.3/14.0 × 0.33/1.1 2 × 2SSB N/A 16 2.0/6.50 × 0.15/0.5 4 × 4
Scale
Aero centresMAC
A/cSears–Haack
sections
Mach coneangle M1.6
Optionalaileron
0 3 6m
0 10 20
C-C
IntakeIntake
B-BA-A
D-D
E-EF-F
Fuel
B
B
C
C
D
D
E
A/cCG
E
F
F17°
A
A
ft
C
Fig. 8.13 Revised baseline GA
provide a useful storage volume to house sensitive sensors and flight instruments. Theincreased centre line chord provides sufficient volume to accommodate the weapon bayin a four-across configuration and part of the engine depth. This produces a blendedbody shape for the aircraft. For stealth and to avoid flow problems at high angles of
“chap08” — 2003/3/10 — page 232 — #31
232 Aircraft Design Projects
attack, the intakes have been positioned on the underside of the wing profile. Withoutdetails of the aircraft centre of gravity position but knowing that the centre of lift isslightly forward of the 50 per cent MAC in the low-speed flight cases, the landing gearhas been positioned relative to the centre of area of the wing planform. With the mainunits in this position there is sufficient tail-down angle to allow for tail-down rotationon take-off. The position of the nose unit may present intake ingestion problems thatmay demand wheel-debris shielding. Moving the nose units behind the intakes doesnot seem feasible.
Although the layout looks practicable, there are three areas of concern. The firstrelates to the reduced roll control that has arisen due to the smaller wing-span and area.To overcome this potential deficiency it may be necessary to modify the wing tip toextend the aileron surface. This is shown on the new aircraft GA as a surface with a 60◦degree LE sweep and 20◦ TE sweepback. The second potential problem area involves thecomplication of geometry at the engine nozzle/wing trailing edge. The nozzle will needto have some vectoring capability to provide the aircraft with pitch control (particularlyat high angles of attack). How this will be provided, without excessive complication tothe wing structure and flow conditions at the wing intersection, has still to be realised.Finally, the third problem area concerns the positioning of the intakes relative to thewing leading edge vortex flow, particularly at low speeds. This is a complicated layoutproblem which will require some detailed computational fluid dynamic investigationand, at later stages, wind tunnel testing.
(d) Cross-sectional area distributionsBefore starting the detailed technical analysis tasks, it is possible to use the aircraftGA to assess the volume distribution for wave drag evaluation and to determine theavailable fuel tank capacity relative to the estimated fuel load.
Some of the normal (90◦) sections are shown on the GA and can be used to determinethe cross-sectional areas at each station. These are plotted in Figure 8.14 and representthe values for the Mach-one case.
For higher Mach number cases, the sectional areas must be calculated as the sur-face area on the Mach cone intersection projected forward. For our aircraft, wewill determine the distribution at Mach 1.6. In this case, the Mach cone semi-angleα = (90 − 51.3) = 38.7◦. We can approximate the cone section areas by applying the
12
10
8
6
4
2
0Nose Tail
X (Fuselage station)
A(s
q. m
)
M1.6
M1.0
Fig. 8.14 Sears–Haack area distribution
“chap08” — 2003/3/10 — page 233 — #32
Project study: advanced deep interdiction aircraft 233
ratio of the ‘normal’ section to the cone surface area:
πrl/πr2 = l/r and l = r/ sin α
where r is the base radius of the cone and l is the sloping length. For a cone angle of38.7◦, l/r = 1.6.
Using this ratio to adjust the M1.0 areas gives the curve also shown in Figure 8.14.The actual area distribution, although smooth, shows a skew profile resulting fromthe arrow-wing planform. It is impossible to correct this without making impracticalchanges to the internal layout. Our current layout is not able to achieve the perfectSeers–Haack distribution but is regarded as relatively good compared to other aircraft.A distribution factor (EWD) of between 1.2 and 1.4 seems reasonable.
(e) Fuel volumeAs we have not yet performed a detailed mass prediction on our new layout we do notknow the exact fuel load to be carried. We will have to use the value from previouscalculations (i.e. a fuel load of 30 360 kg (66 944 lb)). Assuming a specific gravity of 0.8for the fuel gives a required volume of (30 360/800 =) 38 cubic metres, or 10 300 US gal(or 66 944/6.8 = 9845 US gal if the specified JP8 fuel is used). From the aircraft GA,the central fuel tank = 2.0 × 1.0 × 7.5 = 15 m3 and each wing tank = 7 × 2.5 × 1.0 =17.5 m3, making a total volume of 50 m3. Allowing 10 per cent for structural intrusionand internal systems still leaves 45 m3. This is more than adequate for the estimatedrequirement which we suspect is slightly high.
8.8.2 Mass evaluation
With the aircraft geometry determined in the general arrangement drawing it is nowpossible to make a detailed assessment of the aircraft mass. Using empirical formulaein design textbooks slightly modified to suit the particular features of our aircraft, eachaircraft component can be evaluated. The results are shown in Table 8.5.
The blended profile of our aircraft layout makes it difficult to distinguish betweenthe wing and body mass components. The total mass of these components roughlyequals that expected for the wing alone of a traditional design. The compact structuralarrangement of the blended body is likely to give substantial savings in structural massso the result above can be accepted until more detailed analysis can be attempted.As mentioned in the initial mass estimation (8.6.1), the engine mass on our aircraftis unusually large. This can be explained by the need for the supercruise/dash speedand the high altitude performance. Extra allowance has been given to the avionics andcockpit systems mass to account for the sophisticated nature of the aircraft operation.Anticipating a lower MTOM from this analysis, the fuel mass has been reduced fromthe earlier value but still gives about a 48 per cent fuel mass ratio. The overall resultshows the new design MTOM to be substantially lower than previously anticipated.This should prompt a review of the aircraft geometry and a further mass iteration.Without the confidence of the detailed aerodynamic and performance analysis thismight be presumptuous, so we will continue the design process without altering ourrevised baseline details and accepting the above mass values.
8.8.3 Aircraft balance
Although the aircraft control limits have not yet been determined, it is still worthwhileto check if the aircraft configuration provides a sensible location for the centre ofgravity excursions. Using the aircraft GA, Table 8.5 can be extended by adding the
“chap08” — 2003/3/10 — page 234 — #33
234 Aircraft Design Projects
Table 8.5
Component lb kg % MTO Arm (m)
Wing 6 157 2 800 5.4 18.0Control surfaces 832 378 0.7 23.5Body 5 210 2 368 4.6 16.5Main gear 2 835 1 289 19.7Nose gear 868 394 3.2 (u/c) 6.7Intakes 2 633 1 197 2.3 13.5∑
STRUCTURE 18 535 8 406 16.2
Dressed engine 15 600 7 091 13.7Installation 769 349 0.7Engine system 1 054 478 0.9∑
PROPULSION 17 423 8 728 15.3 18.8
Fuel system and tanks 1 723 783 1.5 19.5Aircraft systems 1 546 701 1.4 14.0Avionics 2 370 1 077 2.1 9.5Cockpit systems 1 440 653 1.3 10.0Weapon systems 1 500 682 1.3 17.0∑
FIXED EQUIP. 8 579 3 891 7.6∑ ∑
EMPTY 44 537 21 025 39.1
Crew and op items 1 100 500 1.0 10.0Weapons 13 448 6 113 11.8 17.5∑ ∑
ZERO FUEL 59 082 27 623 51.9
Fuel 55 000 25 000 48.1 18.0 (central)20.5 (wing)
Max take-off 114 082 51 739 100.0
estimated positions of the component masses. With the aid of a spreadsheet, variouscombinations of aircraft loading can be assessed. The main results are shown on theexcursion graph (Figure 8.15). Values of around 50 per cent MAC are acceptable fora supersonic delta wing and should prove feasible. The centre of gravity range for theflight cases is seen to be only about 6 per cent MAC (0.83 m or 2.7 ft). For the landinggear layout, the centre of gravity range (44 to 51 per cent MAC) should present noconfigurational difficulties.
8.8.4 Aerodynamic analysis
At this stage in the development of the project, the aerodynamic analysis must focuson a more accurate estimate of drag, an assessment of lift and lift-curve slope, andthe determination of the aerodynamic centre of the wing. As the design matures itwill be possible to refine these estimates using more sophisticated computational panelmethods. As these require detailed geometric definitions of the full aircraft profile,which at this time is not fixed, it would be presumptuous to start such work now.
Drag
In the earlier part of the project, when the geometry of the aircraft was unknown, estim-ates of the aircraft subsonic drag were based on an equivalent skin friction coefficient.
“chap08” — 2003/3/10 — page 235 — #34
Project study: advanced deep interdiction aircraft 235
50
Kg×
1500
40
30
2044 46 48
% MAC
Empty
(full payload)ZFM
Light (reserve fuel, zero weapons)
Combat (1/2 fuel 4000 kg weapons)
MTO Full payloadfull fuel
50 52
Fig. 8.15 Aircraft centre of gravity excursion plot
This is a crude method which is open to substantial inaccuracies as it cannot take intoaccount any subtleties in the aircraft configuration. With more detail available aboutthe aircraft layout, it is now possible to use the component build-up method to predictaircraft parasitic drag. Most design textbooks outline this process which combines theflat plate skin friction drag coefficient, a form factor to account for viscous effects,and an interference factor. Each component part of the aircraft is assessed separately,normalised to the aircraft reference area, and then summed to give the principal partof the zero-lift drag. Additional drag producing items (e.g. landing gear, flaps, externalstores and fuel tanks, aerials and sensors, etc.) are analysed and their drag included, ifappropriate to the flight case under investigation.
A word of warning and some adviceAerodynamic analysis frequently involves the use of coefficients. These are con-venient as it allows aerodynamicists to non-dimensionalise their parameters. Forexample, drag (which is a force) is divided by dynamic pressure (q = 0.5×ρ×V 2)and by a reference area (e.g. S = gross wing area) to give the drag coefficient (CD).For a particular flight case, the value of a drag coefficient is useless without theknowledge of the accompanying reference area. As both of these numbers arealways necessary, it is often better to quote drag in terms of drag area (CD × S).Drag area retains the definition of measurement units as it will need to be quotedin square feet, or square metres. This definition therefore allows a visualisationof the magnitude of the drag of components.
Providing that a consistent set of units is used, the aerodynamic coefficient willbe the same value in each measuring system. Moving between the use of SI and‘British’ units may cause errors unless care is taken with regard to the definitionof the reference area units when calculating drag area.
“chap08” — 2003/3/10 — page 236 — #35
236 Aircraft Design Projects
For our blended-body layout, without the need for extra control surfaces, the com-ponent build-up involves only the wing surfaces and some allowance for the centralbody profile. Using a spreadsheet method, it is possible to determine parasitic drag forvarious combinations of altitude and speed. The following constants were used in theevaluation:
• ISA conditions• Reference area = 133 sq. m (1430 sq. ft)• Wing surfaces wetted area = 268 sq. m (2929 sq. ft)• Body wetted area = 160 sq. m (1720 sq. ft)• Wing reference length (MAC) = 13.8 m (45.2 ft)• Body reference length (overall length) = 25.5 m (274 ft)• Wing laminar flow = 10 per cent• Body laminar flow = 1 per cent• Interference factor = 1.0
As an example the values used to determine the drag coefficients (in SI units, for sealevel and M0.2) are shown below:
• Reynolds number (×106) = 64.3 wing, 118.8 body• Form factor = 1.0 wing, 1.07 body• Skin friction coefficient (turbulent) = 0.00245 wing, 0.00225 body• Skin friction coefficient (laminar) = 0.00165 wing, 0.00122 body• Zero-lift drag coefficient = 0.00448 wing, 0.00263 body
Extra (miscellaneous) drag (base, up-sweep, leaks, protuberances drag) is assessedas �Cdo = 0.00061, and landing gear drag (when appropriate) as �Cdo = 0.00376(including interference). As we shall see in the lift section the vortex lift increase onthis configuration is sufficient to avoid the inclusion of flaps and associated drag andsystem complications. Adding the extra drag to that for the wing and body above gives:
CDO = (0.00448+0.00263+0.00061) = 0.00772 (clean), and 0.01148 (with u/c down)
The results for other flight cases are shown in Figure 8.16.Using methods described in reference 4, the critical Mach number for our wing is
calculated as M0.849. Just prior to and beyond this speed, supersonic flow over theaircraft creates a substantial increase in drag (wave drag).
As described earlier (section 8.6.2) reducing wave drag is a vital objective in thisdesign. It is important to make changes to the aircraft layout to bring this about.The revised GA, even with the smaller wing, was analysed and found to generate anunacceptable wave drag. To reduce the drag the wing thickness was reduced to 6 percent. This was acceptable as the fuel volume was found previously to be more thanadequate. The Sears–Haack volume distribution graph for the thinner wing design andwith an extension to the nozzle length is shown in Figure 8.17 together with the idealSears–Haack distribution.
The cross-sectional areas shown in Figure 8.17 have been calculated around a Machcone angle of about 38◦. This corresponds to the cruise speed of M1.6. To improve theshape of the distribution some area ruling can be applied to the fuselage profile aroundthe forward wing junction. The effect of such a modification to the shape of the aircraftis shown in Figure 8.18. The resulting area distribution can be seen in Figure 8.19.
The advantage of the blended-body configuration together with area ruling has cre-ated a volume distribution that is as close to the ideal ‘Sears–Haack’ profile as can beexpected. It should be possible to have confidence in using an efficiency factor EWD of
“chap08” — 2003/3/10 — page 237 — #36
Project study: advanced deep interdiction aircraft 237
SL20 k
36 k
50 k
0.010
0.011
CD
o
0.009
0.008
0.007
0.006
0.2 0.4 0.6 0.8
Mach number
1.0 1.2 1.4
65 000 ft Assumedcontinuationwithoutwave drag
Ment
Fig. 8.16 Subsonic zero-lift drag coefficient
Sears–Haackidealdistribution
Exposedwing only
Body plus wingless internalflow paths
Fig. 8.17 Revised Sears–Haack distribution
1.4 for our aircraft. The wave drag area is calculated using the formula below, with theinputs shown:
D/q = EWD(1 − 0.386(M − 1.2))0.57(1 − (π × LE sweep0.77/100))
× (4.5π(Amax/length)2)
CDwave = (D/q)/S
where EWD = 1.4LE sweep = 70◦Amax = 8.6 sq. m (92.4 sq. ft)length = 25.5 m (83.6 ft)
Giving, for S = 133 sq. m (1430 sq. ft):
CDwave = 0.01691 at M = 1.2, = 0.01646 at M = 1.4
= 0.01624 at M = 1.6, = 0.01607 at M = 1.8
“chap08” — 2003/3/10 — page 238 — #37
238 Aircraft Design Projects
Body arearuling
Thinner wing profile
Intakes
Fig. 8.18 Area ruling
Reduction involume dueto area ruling
Old profile distribution
Finalbaselineimprovedprofile
Sears–Haackidealdistribution
Fig. 8.19 Final Sears–Haack distribution
Adding these to the zero-lift values and sketching the transonic region allow us to drawthe zero-lift drag versus speed graph, Figure 8.20.
This shows that at the initial dash condition (M1.6, 50 000 ft), the total aircraft CDo =0.0205. The wave drag contributes nearly 80 per cent of the drag. This illustrates that thesignificance of reducing wave drag is obvious. It can be seen from the equation for wavedrag that it is inversely proportional to aircraft length squared. As we have anticipateda potential problem area in the flow conditions at the rear of the aircraft, a futuremodification may be required to extend the nozzle region (as shown in Figure 8.21).
This change, together with a small stretch to the aircraft nose, could increase thelength from 25.5 m (85 ft) to 27.5 m (90 ft). This increase to the aircraft length wouldreduce wave drag by 14 per cent. A very small penalty would be incurred in the parasiticdrag but the overall effect would be to lower the zero-lift drag coefficient to 0.0183.Although these changes will not be taken into account at this time, the calculationsshow how sensitive the drag estimate is to aircraft configuration.
Lift and lift-curve slope
To determine the drag due to lift requires the estimation of the induced drag factor(k1). A vortex-lattice method is the most appropriate way to estimate the induceddrag factor but this is too time consuming and involved to be considered at this time.
“chap08” — 2003/3/10 — page 239 — #38
Project study: advanced deep interdiction aircraft 239
0.005
0.010
0.015
0.020
0.025
0.00.2 0.4 0.6
65 k20 k
SL
65 k
SL
0.8 1.0
Mach number
1.2 1.4 1.6 1.8
CD
o Clean aircraft
Fig. 8.20 Zero-lift drag versus Mach number
Engine C
A/c C
Engine C
Engine C
Chord line
Thrustvectors
Fig. 8.21 Modified engine nozzle geometry
“chap08” — 2003/3/10 — page 240 — #39
240 Aircraft Design Projects
Simple methods, however, may lead to substantial inaccuracy. To add confidence tothe estimation two different methods will be tried.
The Oswald span efficiency method, with A = 1.27 and LE sweep = 70◦ givesa value of e = 0.493. Using the formulae for k1(=1/πAe) gives k1 = 0.508. Thisvalue seems high when compared to conventional trapezoidal, long span wings but forour low aspect ratio, delta planform, the spanwise air loading is far from optimum.McCormick10 and other aerodynamic textbooks provide methods of predicting (k1)that are specific to supersonic delta wings. As the planform will produce zero leadingedge suction k1 = (1/CLα), where CLα is the wing lift curve slope.
For our wing, the two-dimensional lift curve slope = 4/(M2 − 1)0.5
For LE sweep of 70◦ at M1.6, the method in reference 10 determines the finite wingCLα = (0.6CLα2D) = 0.6 × 4/(1.62 − 1)0.5 = 1.92This gives (k1) = 0.521.
There are several other methods that could be used to determine (k1) but as these giveapproximately the same value an average will be used (i.e. 0.514).
The combination of zero-lift and induced drag provide the aircraft drag equation.
At the initial dash phase, CD = 0.0205 + 0.514C2L
With the aircraft at 80 per cent of the take-off mass, CL = 0.146This gives the aircraft drag coefficient = 0.03146Therefore drag = 0.5 × 0.1864 × 4722 × 133 × 0.03146 = 86.88 kN (19 531 lb)The aircraft lift/drag ratio = 0.146/0.03146 = 4.64
McCormick10 also provides a method of predicting the lift characteristics for deltawings. The lift from the delta planform is augmented by the leading edge vortex flows(called vortex lift). The results are shown in Figure 8.22. Note that at higher angles ofattack the lift-curve slope increases and this will reduce the lift-induced drag factor.We will assume an average value of 0.15 for the subsonic drag evaluations.
Raymer4 presents some data that recommends a maximum angle of attack of about35◦. To avoid flying too close to this condition we will limit our design to 30◦ whichgives a maximum lift coefficient of 1.4. At these high angles of attack the aircraft dragis roughly proportional to (sin α) times the lift. This high drag will be useful to offsetthe horizontal component from the vectored thrust that will be required to control thehigh pitch attitude.
Centre of lift
Nicolai12 provides data on the position of the aerodynamic centre for a symmetricprofile, delta wing configuration. The chart gives the following results relative to thedistance behind the leading edge of the mean aerodynamic chord (MAC):
M0.2 = 35% MAC, M0.8 = 39% MAC,
M1.0 = 50% MAC, M1.6 = 48% MAC
This data will be useful to enable us to balance the aircraft for each of the flight andloading (weight) cases.
“chap08” — 2003/3/10 — page 241 — #40
Project study: advanced deep interdiction aircraft 241
Angel of attack (α)
5
0.2
0.4
0.6
0.8
1.0
1.2
1.4
10 15 20 25 30 35
Airc
raft
lift
coe
ffic
ient
With vortexcontribution
Withoutvortexflow
Fig. 8.22 Aircraft lift coefficient versus angle of attack
8.8.5 Propulsion
Before aircraft performance can be estimated it is essential to have data on the availablethrust and fuel flow from the engine. In this design project, the AIAA provided somegeneralised performance data for a low bypass supersonic cruise turbofan engine. Thequoted static sea-level thrust is 26 356 lb per engine.
Our constraint analysis indicated a (T/W )SSL of 0.58. For our predicted MTOM(51 739 kg: 114 082 lb) this relates to two 33 080 lb thrust engines. We therefore need toscale the given data by a factor of 1.25. This represents about the maximum feasiblelimit on scaling engine data. As the engine is a current design and as we are scaling itinto a larger engine, it should be possible to envisage a new engine that is more fuelefficient than this design. Mattingly11 provides programs that allow the design of anew engine. This method could be used to optimise an engine specifically suited toour specification. Although this has not been done in the present study, this type ofaircraft-engine design integration can make an interesting team project as it involvesdesign compromises that need to be made between airframe and engine requirements.
Until the available engine data has been assessed and some aircraft mission analysisconducted, it is not appropriate to include any factors to account for a new engineconfiguration. The data supplied by AIAA although corrected for installation effects(including intake and nozzle flow losses and off-takes) was in the generalised (non-dimensioned) format typical of data from engine manufacturers. Both the net thrustand the fuel flow values were divided by the pressure ratio (δt) which relates the airpressure at the engine fan face to the ambient air pressure at sea level. This is calculatedas the product of the local total, or stagnation, pressure ratio (Pt/P) and the ambient,far-field ratio (P/PSL). Both of these are functions of air properties as shown below:
(Pt/P) = 1 + ((γ − 1)M2/2)n
where M = flight Mach number and n = (γ /(γ − 1)).
“chap08” — 2003/3/10 — page 242 — #41
242 Aircraft Design Projects
For air at ISA conditions γ = 1.4, giving [(γ −1)/2] = 0.2 and thereby, n = 3.5. Thevariation of (P/PSL) (commonly denoted as (δ)) with altitude can be found in mostISA tables.
The value of fuel flow provided by engine manufacturers is also divided by totaltemperature ratio (θt) raised to the power of 0.6. Where (θt) is calculated as the productof the local total, or stagnation, temperature ratio (Tt/T ) and the ambient, far-fieldratio (T/TSL).
These are defined below:
(Tt/T ) = 1 + ((γ − 1)M2/2)
where M = flight Mach number and as above [(γ − 1)/2] = 0.2.The variation of (T/TSL) (commonly denoted as (θ)) with altitude can be found in
most ISA tables.Using the functions above with the AIAA engine data it is possible to determine the
installed net thrust and specific fuel consumption (= fuel flow/thrust) of the engineagainst aircraft Mach number and altitude. This data, scaled to a TSSL thrust of33 080 lb, is shown cross-plotted in Figures 8.23a to 8.23c. These graphs are drawnin ‘British’ units as this is how the original data was given (note: 1 lb = 4.448 N andsfc values are the same in N/N.hr).
The AIAA engine data is based on an axi-symmetric, translating, centrebody intakedesign with an ejector nozzle. Although the nozzle design on our aircraft is likely tobe more complex, it is not expected to be any less efficient. Our intake should bedesigned to provide lower losses at the operating conditions. This will involve the useof a moveable ramp and tip at the front of the intake duct, to better match the low- andhigh-speed requirements. This may provide up to about 10 per cent improvement butwith the mechanical complexity it will be less stealthy. Will the propulsion advantageoutweigh the disadvantage in stealth (a classic design decision)?
8.9 Performance estimations
With the detailed analysis of the aircraft mass, aerodynamic coefficients and enginecharacteristics, it is now possible to assess the aircraft performance in all flightconditions. This involves analysis of three different operational modes:
1. Manoeuvring performance2. Mission analysis3. Field performance
8.9.1 Manoeuvre performance
(Note: more detailed explanation of the methods used in this section can be foundin design textbooks, e.g. reference 4 or 7. In order to avoid confusion on the variousgraphs in this section they have been drawn in ‘British’ units only.)
Assessment of the manoeuvring capabilities of military fighting aircraft is an essentialpart of the performance analysis. The methods used for this work are based on thecalculation of the specific excess power (SEP) available in various flight conditions(speed, height, aircraft weight and load factor). SEP has the units of rate of climb (ft/sor m/s) as defined in Manoeuvring below.
From the previous sections on propulsion, aerodynamics and mass we have sufficientinformation to calculate SEP at any point in the aircraft flight envelope.
“chap08” — 2003/3/10 — page 243 — #42
Project study: advanced deep interdiction aircraft 243
0 10 000 20 000 30 000 40 000 50 000 60 000 70 0000
5000
10 000
15 000
20 0000.45
1.1
1.31.5
1.61.8
2.0 Mach number0.6 0.9
25 000
30 000
35 000(b)
(a)
0 0.2 0.4 0.6 0.8 1
Mach number
Altitude (ft)
1.2 1.4 1.6 1.8 2
70656055
50
43
36
302520
Altitude (ft × 103)
5000
10 000
15 000
20 000
25 000
30 000
35 000
40 000
SL 1.5
5 10
(c)
0.600
0.700
0.800
0.900
1.000
1.100
1.200
1.300
1.400
1.500
0 10 000 20 000 30 000 40 000
Altitude (ft)
StaticM0.2
M0.6
M0.9
M1.3M1.6
M2.0
50 000 60 000 70 000
Inst
alle
d th
rust
(lb)
Inst
alle
d th
rust
(lb)
Inst
alle
d TS
FC (p
er h
our)
Static
Fig. 8.23 Engine performance data (graphs a, b and c)
“chap08” — 2003/3/10 — page 244 — #43
244 Aircraft Design Projects
The original problem definition (section 8.3) defined four manoeuvrability criteria:
(a) SEP (1g) military thrust (dry), 1.6 M at 50 000 ft = 0 ft/s.(b) SEP (1g) maximum thrust (wet), 1.6 M at 50 000 ft = 200 ft/s.(c) SEP (2g) maximum thrust (wet), 1.6 M at 50 000 ft = 0 ft/s.(d) Maximum instantaneous turn rate, 0.9 M at 15 000 ft = 8.0◦/s.
Each of these criteria must be assessed at the aircraft manoeuvre weight (defined as50 per cent internal fuel with two AIM-120 and four 2000 lb JDAM) = 81 797 lb(37 100 kg).
The manoeuvring and turning performance estimates are based on the evaluation ofthe available specific excess power. However, the flight condition to be considered inmanoeuvre and turning cases are different so the analysis will be done in two separatesections.
Manoeuvring
The best way to approach the manoeuvring analysis is by generating data matricesof altitude and aircraft speed. These provide the values for aircraft drag and enginethrust at the aircraft weight and load factor to be considered. Computer programs orspreadsheet applications are the best way to perform the calculations repeatedly fordifferent flight cases.
The specific excess power (SEP) is calculated at each point (height and speed) of thedata matrix by the formula:
SEP = [(T − D)/W ]VT = thrustD = dragW = aircraft weightV = aircraft forward speed
For the aircraft at the manoeuvre weight with 1g loading and dry thrust, the SEP valuesare shown plotted in two different formats in Figures 8.24a and 8.24b.
These data maps show that at M1.6 and 50 000 ft, the aircraft exceeds the zero SEPrequirement (see (a) above). Figure 8.24b shows how the introduction of wave dragaffects the SEP. Similarly, a set of curves can be drawn for the wet (afterburning) thrustcases for 1g and 2g loading. These are not shown in full but the SEP versus heightcurves for an aircraft speed of M1.6 have been reproduced on Figure 8.25. This showsthat the zero and 200 ft/s SEP requirements ((b) and (c) above) have been exceeded.This, more detailed analysis, seems to have shown that the original problem with the‘wet/200 fps’ requirement is now overcome. This is a good example of not rushing tomake changes too early in the design process because the predicted SEP is very sensitiveto changes in the drag and thrust values.
The process of evaluating SEP developed for the manoeuvre assessments can alsobe used to specify the optimum climb profiles for the aircraft. The most critical caseis the initial climb to the cruise height (assumed to be 50 000 ft) following take-off atmaximum weight. Using an assumed average weight in the climb of 0.9 MTOW anda load factor of one, SEP maps for dry and wet thrusts can be drawn. These can becross-plotted to provide SEP contours as shown on Figure 8.26.
It is possible to draw lines of constant energy height (i.e. potential and kineticenergy = altitude + aircraft speed squared divided by 2g) onto this graph. The com-bination of the energy height and SEP contours are used to identify the quickest time
“chap08” — 2003/3/10 — page 245 — #44
Project study: advanced deep interdiction aircraft 245
(a)
20 000 25 000 30 000 35 000 40 000 45 000 50 000Altitude (ft)
Mach number
55 000 60 000 65 000 70 000–300
–200
–100
100
200
300M2.0
M1.8M1.6
M1.5
M1.3
M1.1
400
500
SE
P (f
t/s)
SE
P (f
t/s)
(b)
Design requirement M = 1.6
ConditionsÛ 1gÛ Dry engine thrustÛ W = manoeuvre
–300
–200
–100
100
200
300
400
500
600
700
0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
70
70
65
60
55
50Design req. 50 000 ft
43
36
Alt. (ft × 10–3)
30
6560 k
55 k
50 k43 k
36 k
30 k
20 k
10 k
5000
1500SL
Conditions Û 1g Û Dry engine thrust Û W = manoeuvre
Fig. 8.24 Specific excess power (SEP) (graphs a and b)
to altitude. This is achieved by following a line drawn through the 1g SEP lines, per-pendicular to the energy height contours. This is shown on Figure 8.26 as line A. Theproblem with this profile is that, as the aircraft is relatively underpowered in the drythrust condition, the line extends past the zero SEP contour (i.e. obviously an unfeas-ible criteria). This difficulty arises due to the penalty imposed on the SEP by the wavedrag at transonic speeds. The classical way of overcoming this problem is to performa zoom-climb flight profile. In this, the pilot climbs the aircraft until the climb rate
“chap08” — 2003/3/10 — page 246 — #45
246 Aircraft Design Projects
20 000 30 000 40 000 50 000Altitude (ft)
60 000 70 000
n = 1
n = 2
–1000
–800
–600
–400
–200
200
400
600
800
1000
1200
SE
P (f
t/s)
Design requirementsn = 1 n = 2
Fig. 8.25 SEP for M1.6 versus altitude
0.0 0.2 0.4 0.6 0.8 1.0 1.2Mach number
10
20
1.4 1.6
50100
150
200250
300Maxdynamicpressure(2133 lb/sq. ft)
1.8 2.00
5
10
15
20
25
30
35
40
45
50
55
60
65
Alti
tude
(ft×
10–3
)
50
1g min. speedboundary
He = 60 × 10–3
40
30
300
250200
150
100
50
300150
SEP dry
SEP wet
AC
B
SEP dry
Fig. 8.26 Energy height plot
deteriorates then dives (through the initial sound barrier) and then at an increasedspeed and high-energy returns to the climb condition. The dive profile follows theconstant energy height contour. With dry thrust, a possible zoom-climb profile couldfollow that shown as line B. With afterburning the climb could be quicker if profile Cwas used. Profile B is shown to overshoot the final cruise speed to make the final sectionperpendicular to the SEP lines (i.e. transferring kinetic energy for potential energy).
“chap08” — 2003/3/10 — page 247 — #46
Project study: advanced deep interdiction aircraft 247
Detailed evaluation of the climb performance would involve a step-by-step calculationfollowing the height and speed profile defined in Figure 8.26. This has not been donein this study.
Turning
Turning capability is relatively easy to assess providing that a database of aircraft dragand engine thrust against altitude and aircraft speed is available. The turn diagram isdrawn for a specified aircraft weight and altitude (Wman. and 15 000 ft in our case). Asshown below, several significant manoeuvring parameters can be determined from theturn diagram. This makes it a useful device for comparing the effectiveness of differentdesign configurations in a quantifiable way. For example, in trade-off studies, differentvariants of a baseline design can be assessed. Also, using this diagram at the conclusionof the project, the new aircraft can be directly compared to known competitor aircraftor other threats.
The analysis starts by drawing the generalised turn diagram. This is a graph of turnrate against aircraft speed. The formula relating these two parameters is:
Turn rate (◦/s) = 57.3 [g(n2 − 1)0.5/V ]
where g =gravitational acceleration (32.2 ft/s2, or 9.81 m/s2)n =manoeuvre load factor = (L/W )
V = aircraft speed
Note, the turn rate formula is not specific to a particular aircraft design.From the formula above a series of curves can be drawn for each value of load factor
(n), as shown in Figure 8.27.
0.200 0.300 0.400 0.500 0.600
Mach number
Turn
rat
e (°
/s)
0.700 0.800 0.9000
5
10
15
20
25
30
35
R = 1000 ft
R = 2000 ft
R =3000 ft
n = 7
Cn = 6
n =5
n =4
n =3
n =2
n =2
Vmin.
speed
Ps = 0 (WET)
Ps = 0 (Dry)
W = 81 797 lb (37 100 kg)(Manoeuvre wt.)
DB
A
D +
Fig. 8.27 Turn performance graph
“chap08” — 2003/3/10 — page 248 — #47
248 Aircraft Design Projects
As the radius of the turn equals the turn rate divided by the aircraft speed, it ispossible to construct a set of straight lines on the graph that represent specific turnradii.
The turn diagram is made specific to a particular aircraft by introducing threeboundaries:
1. Maximum positive structural normal acceleration factor (nmax). For our design, thisis set in the design requirements at +7g.
2. Maximum structural dynamic pressure (qmax). The speed at which this limit isreached at various altitudes is determined from the equation: q = 0.5ρV 2. For thespecified requirement of 2133 lb/sq. ft the calculated values are shown in Table 8.6.
3. The aircraft minimum speed boundary at the flight condition (e.g. Wman. and15 000 ft). This can be determined for each load factor (n) from:
Vmin = Wn/(0.5ρSCLmax)
Assuming that for our aircraft CLmax = 1.4 (limited by angle of attack) (seeTable 8.7).
The maximum g and minimum speed boundaries are plotted on Figure 8.27 but thedynamic pressure boundary falls outside the scope of the graph. All that is now neededare the contours of the zero SEP (0 ft/s) for the dry and wet engine thrust.
This is done by plotting the full SEP contours (as shown on Figures 8.28a and b)and transferring the (0 ft/s) intersection speeds onto the appropriate (n) curves onFigure 8.27. Several points on the turn diagram are of interest:
(A) The maximum instantaneous turn rate for M0.9 is limited by the nmax line at avalue of 13.4◦/s (the associated turn radius is 4069 ft). The turn rate is in excess ofthe requirement of 8.0◦/s.
Table 8.6
Altitude (ft) V (ft/s) Mach no.
SL 1340 1.25 000 1443 1.29
10 000 1559 1.4020 000 1835 1.6430 000 2190 1.96
Table 8.7
n Vmin (ft/s) Mach no.
1 234 0.2212 331 0.3133 405 0.3834 467 0.4425 523 0.4946 573 0.5247 618 0.585
“chap08” — 2003/3/10 — page 249 — #48
Project study: advanced deep interdiction aircraft 249
–100
100
200
300
400
500
0.000 0.600 0.800 1.000 1.2000.400
Mach number
0.200
(a)
n = 1
n = 2
n = 3
Mach number
(b)
–400
–200
200
400
600
800
1000
0.200 0.400 0.500 0.600 0.700 1.000 1.100 1.2000.300 0.800 0.900
n = 1
n = 2
n = 3
n = 4
SE
P (f
t/s)
SE
P (f
t/s)
Fig. 8.28 Specific excess power at 1500 ft (graphs a and b)
(B) The sustained turn rate (dry) at M0.9 is limited by the zero SEP boundary at 5.4◦/s.This gives a radius of turn of 10 097 ft. The wet thrust, zero SEP boundary givesa turn rate of 8◦/s (6815 ft radius).
(C) The intersection of the minimum speed and nmax boundaries defines the cornerpoint. This gives the highest (instantaneous) turn rate for the aircraft∗ of 21.0◦/sat a corner speed of 365.7 kts. The turn radius is only 1686 ft which explains whyfighter pilots attempt to get to the corner point in dog fights!
(D) The peak value on the zero SEP curves give the highest sustained turn rate. Forthe dry engine the rate is 8◦/s at 407 kts (note how flat the sustained rate againstspeed is on this aircraft). The corresponding values for the wet thrust are 12◦/s atM0.35 (4921 ft radius). This curve is also very flat.
“chap08” — 2003/3/10 — page 250 — #49
250 Aircraft Design Projects
(E) Although point D above gives the max. sustained turn rate, the tangent of a radialfrom the origin to the zero SEP curve gives the smallest sustained turn radius. Inthe dry thrust case, the values are 7.5◦/s at 200 kts with a radius of 2582 ft. Thewet thrust intersection coincides with the minimum speed boundary.
∗ The minimum speed boundary in this calculation assumes that the high angle of attackrequired to achieve the max. CL value is controllable. It is likely that in our design, thismay only be possible with a contribution from vectored thrust. The component of forcefrom the thrust vectoring has not been included in the calculations because this wouldrequire more aircraft and propulsion details than are available at this stage. As we arerelying on assistance from thrust vectoring for landing control, it may be possible todesign the system to provide an integrated aerodynamic and propulsion control systemin the turn manoeuvre without much additional complexity. We have easily met theinstantaneous turn requirement so we will assume that the minimum speed boundaryis achievable and not critical.
Recommendations
All of the specified manoeuvre and turn requirements have been easily met with thecurrent design but a word of caution is appropriate. As the value of SEP at a particularflight condition is dependent on the difference between two relatively large numbers(thrust and drag), small percentage changes in either will result in large variations inSEP. At this stage in the design process, when only crude estimates have been madeabout aerodynamic and propulsion characteristics, this must concern us. For example,when considering flight at high manoeuvre load factors, the lift-induced drag becomesa significant component of drag. As this is dependent on the estimation of the induceddrag factor, which is difficult to predict accurately for our planform, there could beuncertainty in the ‘high g’ performance. Also, the engine performance is affected by thedetail layout and control mechanisms in the intake. A poor estimate of intake efficiencywill significantly affect the net thrust available. For these reasons it is important toobtain better (higher confidence) estimates of these parameters in the next phase of thedesign process.
8.9.2 Mission analysis
There are four cruise stages to be assessed: outbound supercruise, outbound dash,return dash and return supercruise. Each of these stages is to be flown at M1.6. It isnecessary to determine the optimum (based on minimum fuel burn) cruise height foreach stage. From the engine data it is possible to extract the performance (net thrustand sfc) versus altitude for the cruise speed of M1.6, see Figure 8.29.
This provides the variation of sfc to be used to predict fuel burn by assuming thatthe thrust required equals the aircraft drag. From section 8.8.4, the aircraft drag polarat M1.6 is:
CD = 0.0205 + 0.514C2L
The dynamic pressure q (=0.5ρV 2) in the stratosphere (above 36 089 ft) where thespeed of sound is constant at 986 kts, can be determined for each height using the ISAformula for relative density multiplied by the sea-level value (0.002378 slugs/cu. ft) asshown below:
q = 0.5(0.2971e−x)0.002378(1.6 × 986)2
“chap08” — 2003/3/10 — page 251 — #50
Project study: advanced deep interdiction aircraft 251
TSFC (1/hr) and net thrust (lb) for M1.6
5000
10 000
15 000
20 000
25 000
30 000
35 000 1.40
1.30
1.20
1.10
20 000 25 000 30 000 35 000 40 000 45 000 50 000 55 000 60 000 65 000 70 000
Altitude (ft)
Net thrust (lb) TSFC/hr
SFC/hr
(a)(b)
(c) (d)
Net thrustper engine (lb)
Fig. 8.29 Engine performance at M1.6
where x = (H − 36089)/20806.7H = altitude (ft)
For our aircraft, the reference area is 1430 sq. ft (133 sq. m).As we are unaware of the fuel burnt in each segment at this stage in the design process,
it will be necessary to make some assumptions regarding the weight of the aircraft atthe start of each stage, as shown below:
(a) Outbound supercruise = 0.9 MTOW(b) Outbound dash = 0.8 MTOW(c) Return dash = 0.7 MTOM(d) Return supercruise = 0.6 MTOW
And if the return stages follow the release of munitions:
(e) Return dash = 0.7 MTOM – 8000 lb(f) Return supercruise = 0.6 MTOW – 8000 lb
where, from section 8.8.2, MTOM = 114 082 lb (51 739 kg).The aircraft weight defines the CL which in turn defines the CD from which the aircraft
drag is calculated. This is multiplied by the engine sfc to obtain the fuel used per hour.This procedure is easily performed using a spreadsheet method.
The results are shown in Figure 8.30.This clearly shows an optimum altitude for each stage. The optimum heights are
cross-plotted against aircraft weight in Figure 8.31. The associated fuel consumptionis also plotted on this graph.
At this point it is possible to use the fuel consumption results to determine theoverall fuel burnt on the mission (assuming that the fuel consumption in each stageis the average between the start and end values). The time spent on each stage is thestage distance divided by the aircraft speed. As the speed is constant (933.5 kt), the
“chap08” — 2003/3/10 — page 252 — #51
252 Aircraft Design Projects
10 000
15 000
20 000
25 000
30 000
35 000
40 000
35 000 40 000 45 000 50 000 55 000 60 000 65 000 70 000
Cruise altitude (ft)
(a) Outward supercruise �= 0.9
(b) Outward dash � = 0.8
(c) Return dash � = 0.7
(c) Less weapons
(d) Return dash � = 0.7
(d) Less weapons
Fuel
bur
n(lb
/hr)
Fig. 8.30 Aircraft fuel burn versus altitude
Best alt
50 000
55 000
60 000
65 000 25 000
20 000
15 000
70 000
65 000 70 000 75 000 80 000 85 000
Aircraft weight (lb)
90 000 95 000 100 000 105 000
Optimum (minfuel burn) altitude
Fuel burn optimum altitude
Altitute (ft) Fuel burn (lb/hr)
Fig. 8.31 Optimum cruise versus aircraft weight
supercruise stages of 1000 nm take 1.07 hr and the dash stages of 750 nm take 0.80 hr.The analysis is shown in Table 8.8.
This is much larger than originally estimated due to the lower lift to drag ratio(4.87 compared to 5.56 assumed earlier). This would mean that the aircraft MTOW
“chap08” — 2003/3/10 — page 253 — #52
Project study: advanced deep interdiction aircraft 253
Table 8.8
Fuel burn per hour
Stage Start End Average Time Fuel used (lb)
(a) 25 521 22 742 24 131 1.07 25 820(b) 22 742 19 968 21 355 0.80 17 148(c) 19 968 17 178 18 573 0.80 (14 914)(d) 17 178 14 000∗ 15 589 1.07 (16 680)(e) 18 012 15 219 16 620 0.80 13 346(f) 15 219 13 000∗ 14 109 1.07 15 097
71 411 lb
∗ Guessed values.
5000
10 000
15 000
20 000
25 000
15 000 20 000 25 000 30 000 35 000 40 000 45 000 50 000 55 000 60 000 65 000
Altitude (ft)
Thru
st a
nd d
rag
(lb)
Thrust (single eng.) @ M0.9
Thrust (single eng.) @ M1.6Aircraft drag (50%): (a) 1st supercruise (b) 1st dash (c) 2nd dash (d) 2nd supercruise
a
b
cd
Fig. 8.31optimum
Fig. 8.32 Engine thrust at M0.9 versus altitude
must be increased. As the above mission assumed that only 8000 lb of weapon loadwould be dropped and about 13 000 lb was used in the mass statement to account fordifferent missions, we could substitute 5000 lb into the fuel load. This would meanthat the aircraft weight would need to be raised by 11 400 lb. However, some of thispenalty could be set against potential improvements in engine design as mentioned insection 8.8.5. (For example, if the sfc could be reduced from 1.2 to 1.1 the fuel loadwould reduce by 6000 lb.)
The cruise analysis predicted the drag. This can be compared to the available thrustwhich has been extracted from the engine data and shown on Figure 8.32.
The analysis shows that the engine needs to be slightly more powerful to fly atoptimum (minimum fuel burn) altitudes. However, as the aircraft L/D ratio is lowerthan expected no change should be made to the engine until a more accurate estimationof aircraft drag is available.
“chap08” — 2003/3/10 — page 254 — #53
254 Aircraft Design Projects
Recommendations
Assuming a 5 per cent reduction in engine sfc is possible from a new design, it issuggested that a fuel load of 68 000 lb (30 840 kg) should be provided in the nextreview of the aircraft mass. The review should also reduce the weapons load to 9400 lb(4263 kg).
8.9.3 Field performance
There are four operational issues to be investigated in this performance section:
1. Normal take-off distances to the point at which the aircraft achieves lift-off.2. Balanced field length and the decision speed, for single engine operation.3. Approach speed.4. Landing distances from aircraft touchdown.
The calculations for each of the cases above require an analysis of the forces on theaircraft (weight, lift, drag, thrust and ground friction). Our previous estimations ofmass, aerodynamic and propulsion characteristics are sufficient to use as input tothe analysis. The prediction of take-off and landing distances requires a step-by-stepcalculation which can be done using a spreadsheet application method.
Normal take-off distances
The take-off distance is the sum of the ground distance (sG) and the rotation distance(sR). The ground distance is that travelled along the runway up to the point at whichthe rotation speed is reached. The rotation distance is a nominal distance to accountfor the rotation of the aircraft to achieve the initial lift-off manoeuvre, prior to theclimb from the runway. Take-off speed (VTO) is defined as that reached at the pointthat the aircraft leaves the runway. To avoid inadvertent instabilities in the initial climbphase, this speed must be faster than that related to the lift coefficient at rotation in thetake-off configuration. The allowance above stall on conventional aircraft is typicallyset at 15 to 20 per cent. As we are well away from the max. CL angle in our aircraft, wecan either no factor is applied or that the factor is small:
Take-off speed (VTO) = [WTO/(0.5ρSCLto)]0.5
where for our aircraft:
WTO = 51 739 kg (114 082 lb)ρ = ISA sea level air densityS = reference area = 130 sq. m (1340 sq. ft)CLto = maximum lift coefficient∗
∗As the aircraft does not have any flaps, this is taken as the lift coefficient at themaximum aircraft tail-down attitude of 15◦. From Figure 8.22 this is seen to be 0.52.
Using the values above gives:
VTO = 36.3 m/s (118 ft/s, 70 kt)
The numerical integration of the ground distance covered is calculated in steps ofaircraft speed from brake release (zero speed) to VTO. Although the aircraft wouldaccelerate during the rotation phase, which would reduce the calculated distance, we
“chap08” — 2003/3/10 — page 255 — #54
Project study: advanced deep interdiction aircraft 255
will concede this small advantage (inaccuracy) to make the calculation simpler. Theformula to estimate ground distance is available in most textbooks and repeated below:
sG = 0.5∫
(1/a)d(V 2)
where a = aircraft acceleration = [T − D − µ(W − L)]/MT = take-off thrust. As there is only a small variation of thrust
during the take-off speed change, we will assume that thethrust remains constant at the average energy speed(i.e. 0.707VTO). From engine data, this relates to a thrust of32 950 lb per engine.
D = drag in the take-off configuration. This is calculated from thezero-lift drag coefficient estimated as 0.01148 in section 8.5.7,and the induced drag coefficient at subsonic speeds is assumedto be 0.15. With the wing at a 4◦ angle of attack on the ground,the lift coefficient (from Figure 8.22) is 0.15. Hence the aircraftdrag coefficient is 0.01148 + (0.15 × 0.152) = 0.01486.
µ = the coefficient of ground friction without braking. Designtextbooks suggest this is 0.04 for dry runways and 0.02for icy ones.
(W − L) = the ground reaction force. Where W is the aircraft take-offweight (114 082 lb, 507.44 kN) and L is the lift generated withthe lift coefficient of 0.15 mentioned above.
M = aircraft mass = W/g.
The ground distance, calculated by the step-by-step integration, is 583 ft for the dryrunway and 563 ft for the icy one. In this case, the ice reduces ground friction and istherefore not critical except that the aircraft may be less directionally stable (see furthercomments in the landing section below).
The time spent in the rotation phase is assumed to be 3 seconds. Hence, at the take-offspeed of 118 m/s, the distance covered during rotation (sR) = 354 ft.
For normal take-off, at maximum take-off weight, the max. total take-off distance is937 ft. Even if the usual 1.15 factor to account for pilot and atmospheric variability isapplied to this figure, it is still within the 8000 ft specified in the design brief. Therefore,the all-engines take-off distance is shown to be not critical.
Balanced field length
If an engine fails during the take-off run, the pilot must make a decision either tocontinue the take-off with only one engine working, or to abort and bring the aircraftto rest further down the runway. If the failure occurs late in the take-off run he wouldnaturally continue and vice versa if it happened earlier. The aircraft speed at which itis better to continue the take-off is called the decision speed. The pilot will be awareof this speed from the aircraft flight manual before starting the take-off manoeuvre.To determine this speed, it is necessary to calculate separately (for each of the possiblespeeds at which an engine might fail) the distances required to effect an ‘accelerate-go’and an ‘accelerate-stop’ manoeuvre.
For the accelerate-go case, the calculation includes 1 second of travel after the enginefailure to recognise the fault and to take the necessary actions. After this period, thefailed engine is assumed to be shut down and a ‘normal’ take-off performed with only
“chap08” — 2003/3/10 — page 256 — #55
256 Aircraft Design Projects
the remaining engine producing thrust. During this time, no changes to the aircraftconfiguration are allowed. An increase in drag is applied to account for the drag fromthe failed engine and the trim forces from the control surfaces required to stabilise theasymmetric flight condition.
For the accelerate-stop case, again a 1 second delay is applied before any action istaken. After this time, a 3 second allowance is given to account for the applicationof brakes and the deployment of other drag devices (e.g. air brakes, reverse thrust,drag chutes). As our engine is relatively complex due to the reheat and vector thrustmechanisms, it is unlikely that thrust reversal will be available. For reasons of stealthand aerodynamic efficiency, the smooth wing profiles will not be disturbed by theinstallation of air brakes. Hence, the deployment of braking parachutes seems to be thepreferred method of providing extra retardation at high speeds. Reference 12 provides avalue for the drag area of parachutes. Using the figure of 1.4 times the canopy maximumarea gives a �CD of 0.076 for two 2 m (7 ft) diameter drag chutes.
The results of the calculation using the previous aircraft characteristics and the opera-tional assumptions above for the accelerate-go and accelerate-stop cases, for dry andicy conditions, are shown in Figure 8.33.
The intersection of the lines for the go and stop cases define the decision speed andthe balanced field length. These distances are again substantially less than the required8000 ft specified. In fact, the high thrust to weight ratio of the aircraft means that, ifnecessary, the take-off could be achieved with only one engine operating from the start(this is not a common feature on most aircraft).
Calculations show that single-engine take-off can be achieved in 1688 ft for a dryrunway and 1596 ft for the icy condition.
Approach speed
The approach speed is dependent on the value of the maximum CL in the approachcondition and the maximum aircraft landing weight. Using the high angle of attack
250
500
750
1000
1250
1500
1750
2000
30 40 50 60 70 80
Decision speed (ft /s)
90 100 110 120 130
Dry runway: accel-go
To d
ista
nce
(ft)
Ice runway:accel-go
Ice runway:accel-stop
with brake-chute
Ice
Dry
Dry runway:accel-stop
Fig. 8.33 Balanced field length curves
“chap08” — 2003/3/10 — page 257 — #56
Project study: advanced deep interdiction aircraft 257
on approach as described in section 8.7 and the lift data in Figure 8.22 at an assumedangle of attack of 30◦, provides a CLland of 1.4. The maximum landing weight is setby the operational requirements of the aircraft. If it is necessary to allow for a landingimmediately following take-off (e.g. emergency due to system or engine failure) thelanding weight could be up to 95 per cent of the take-off weight. If it was possible toburn or dump fuel before landing then a lower landing weight could be set. To avoidpenalising the aircraft for the exceptional emergency case we will assume the moreconventional landing weight of MTOW less 50 per cent of fuel. For our aircraft, thisdefinition makes the landing weight:
Wland = 114 082 − (0.5 × 55 000) = 86 582 lb (39 266 kg)
For many conventional aircraft, the minimum approach speed is set at 1.3 times stallspeed. As our aircraft must be fully automated for landing (due to the poor pilotvisibility) and will have precision positioning systems we can assume this safety factorto be reduced to 1.2. In this landing case (as compared to the take-off), the aircraft isflying close to its maximum CL so a factor is still appropriate.
Therefore:
Vland = 1.2[86 582/(0.5 × 0.002377 × 1340 × 1.4)]0.5
= 236.5 ft/s (72.1 m/s, 140 kts)
This seems reasonable compared to estimates of the approach speeds for similar mil-itary aircraft (F-14 = 134, F-117 = 144, Su-33 = 194!, B-2 = 140, B-52 = 140 kts).However, the analysis for our aircraft was based on assumptions for the landing weightand CLmax for each aircraft which may be in error, so a sensitivity study was undertaken.The result is shown in Figure 8.34.
50 000 60 000 70 000 80 000
Aircraft weight (lb)
App
roac
h sp
eed
(ft/
s) (=
1.2V
stal
l)
1.0
1.2
1.4
90 000 10 0000 110 000 120 000160
180
200
220
240
260
280
300
320
340
CLmax
Fig. 8.34 Approach speed versus aircraft weight
“chap08” — 2003/3/10 — page 258 — #57
258 Aircraft Design Projects
Landing distance
Landing distance is computed in a similar method to that for take-off except that thrustis set to zero. To stabilise the aircraft on the ground and to apply maximum braking, afree-roll on touchdown of 3 seconds is assumed. In conventional landing procedures,the touch-down speed is lower than the approach speed due to the drag produced in theflare phase. In our design the high angle of attack on approach will be reduced priorto landing to avoid scraping the rear fuselage. This may suggest that the touchdownspeed will be higher than the approach speed. However, to simplify the calculation wewill assume that the touchdown speed is equal to the approach speed.
The detailed landing calculation shows that, at the landing weight assumed above,the unfactored distance is 2535 ft (773 m) on a dry runway. On an icy runway, thedistance increases to 9273 ft (2828 m). This is beyond the available runway length of8000 ft specified in the project brief. It will therefore be necessary to use brake chutes toreduce the distance. Braking parachutes are particularly useful devices as they are mosteffective at higher speeds when the aircraft brakes are less powerful (due to the unwantedlift reducing the ground reaction force). Using the two 7 ft diameter chutes describedpreviously, the landing distance on an icy runway is reduced to 7047 ft (2150 m). Thisbrings the distance within the available length. In fact the aircraft would be able to landat 95 per cent MTOW within the 8000 ft allowance. Figure 8.35 shows the variation ofunfactored landing distance against landing weight.
Although the results above look acceptable, it must be remembered that the landingmanoeuvre may not be as precise as we have assumed in the analysis. For example, theapproach speed may be higher than expected or the aircraft may overshoot the runwaythreshold due to gust disturbance just prior to touchdown. To guard against suchpossibilities it is common practice to apply a factor to the calculated landing distance.Typically, this is set at 1.67. Applying this to the dry distance of 2535 ft and the icydistance of 7074 ft increases them to 4233 ft and 11 768 ft. The normal, dry runwaylanding is still acceptable but clearly the icy one is still much too long.
200070 80 90
Dry
MLW
60% Wet
Ice
Max. factored
distance
100
Weight (lb×10–3)
110
MTOW
120
3000
4000
6000
8000
Unf
acto
red
hour
ly d
ista
nce
(ft)
Fig. 8.35 Landing distance versus aircraft weight
“chap08” — 2003/3/10 — page 259 — #58
Project study: advanced deep interdiction aircraft 259
As military airfields are fully serviced, it is not unreasonable to expect that in icyconditions the runway surface will be treated to dissolve the ice (as on highways).Recalculating the landing distances using the accepted runway friction coefficient forwet surfaces (0.3) over the last 60 per cent of the runway length, instead of that for ice(0.1), reduces the landing distance to 4715 ft (unfactored) and 7874 ft (factored). Thisis within the allowable runway length. Treating the runway to avoid ice contaminationwill also avoid potential directional instabilities and skidding problems.
8.10 Cost estimations
Estimating the costs of future aircraft has always been seen as an inexact science.Evidence from previous design programmes show that even the seasoned professionalsin industry do not have a good track record at making such estimates. For students, andeven faculty, in an academic environment it is impossible to predict the absolute costsassociated with a new project. Too many of the factors that are needed are only availablewithin a commercial organisation. Such factors relate to the accountancy practicesused, the organisation of the company (or more likely the consortia of companies thatare formed to share the design and manufacturing tasks), the interrelationship betweengovernment and industry, and many more non-technical issues.
For military projects, the need to incorporate modern and advanced technologies isparamount. The timescales involved in the development of such technologies oftenoverlaps the aircraft development period. This leads to uncertainties in the costsincurred. For our project there are at least six technological areas (e.g. stealth, propul-sion, aerodynamic design, structures and materials, and systems) which need to bematured before an exact cost can be assumed. Notwithstanding these difficulties, it isoften financial parameters that are used to choose between different design options.It is therefore essential to be able to determine relative costs to create a frameworkfor such decision making and to be able to compare our design with competitoraircraft.
Fortunately, historical data shows that many of the cost parameters are related toaircraft design variables (e.g. aircraft empty weight, installed engine thrust, numberof engines, aircraft operational speed, and the overall system complexity). Otherfactors are related directly to manufacturing variables (e.g. labour rates, number ofaircraft produced and the production rate). Due to the variability of the value ofa currency with time, it is always essential to ‘normalise’ the quoted cost numbersto a specific date (year). This means that inflation rates for the currency must beapplied to any data used. Cost estimates must always state the year to which they areindexed.
Several aircraft design textbooks provide details of cost estimation methods but inthis study the method published by the Society of Allied Weight Engineers (SAWE)13
is used. This paper describes fully all the details required to estimate the significantcost values at the preliminary design stage. It also provides a spreadsheet methodand example. The method is based on regression of historical data from aircraft ofspecific types. As new designs will be more technically complex than older aircraft itis necessary to apply factors to account for the increase in costs associated with thesenew features. Our aircraft has many new technical features including new structuralmaterials and construction processes, a sophisticated flight and weapon control system,vectoring engine nozzles, efficient high altitude and fast flight, and advanced stealthfeatures. Each of the technical factors in the SAWE method will need to be set at high
“chap08” — 2003/3/10 — page 260 — #59
260 Aircraft Design Projects
Table 8.9
Phase Development Production
Engineering cost 5 738 4 678Development support 1 174Flight testing 943
Tooling labour 1 629 1 623Manufacturing labour 2 369 12 030Quality and security 313 1 564
Materials and equipment 265 6 615Propulsion systems 173 4 336Avionics 189 4 735
Total programme 12 792 35 585
Acquisition cost per aircraft ($M) 232.6Recurring flyaway cost per aircraft ($M) 178.0Recurring cost/lb empty weight ($) 3995.0
values to match these innovations. Details of the factors used in the analysis are shownbelow:
Factor1. Advanced technology features (ATF) 2.02. Flight test requirements to prove ATF 1.33. Application of advanced materials 1.54. Incorporation of stealth technologies 1.35. Cost burden of project security 1.3
Each of these factors is equal to, or higher than, the advanced fighter example used inthe report.
Applying the method to our aircraft, with the factors above, and assuming a pro-duction run of 200 aircraft, gives the following cost breakdown ($M, FY2000) (seeTable 8.9).
Clearly, the recurrent flyaway cost exceeds the $150M mentioned in the design brief.There are several strategies that can be used to reduce the cost to the specified target:
• To accept a reduction in the capability of the design. This is probably the worst of theoptions for the military to take. It is unlikely to be acceptable unless the operationalrequirements placed on the Defence Department by the government are altered.
• To reduce the number of aircraft to be produced to match the available budget. Ifall the overhead costs could be held proportional, this would mean that only 168aircraft could be afforded.
• To produce more aircraft than is needed for the US military by supplying aircraft tofriendly (NATO) countries. This may not be feasible for political and national secur-ity reasons. However, many modern military programmes (including the Eurofighterand the JSF aircraft) are produced by international consortia. To investigate theeffect on costs of increasing the production volume, the cost method used above wasapplied to the production of 500 and 1000 aircraft. As the development overhead is
“chap08” — 2003/3/10 — page 261 — #60
Project study: advanced deep interdiction aircraft 261
200100
110
120
130
140
150
160
170
180
Number of aircraft produced500
430 aircraftrequired
Targetprice
Rec
urre
nt f
lyaw
ay u
nit
cost
($20
00M
)
1000
Fig. 8.36 Aircraft recurrent cost versus production run
shared by the increased number of aircraft produced the flyaway cost is substantiallyreduced, providing that additional costs due to the collaboration can be avoided.Figure 8.36 shows the results of the investigation. From this graph, it would bepossible to reach the recurrent unit cost target of $150M if 430 or more aircraft aremanufactured (and sold!).
In an attempt to judge the accuracy of the cost method, details of the F-22 aircraftwere input and analysed. This showed that, at FY2000 prices, the aircraft would costabout $141M. Investigating published data from the US National Audit Office andother government reports suggests that the aircraft actual unit cost is about $94M.This suggests that the published figures have been misinterpreted, the costs may havebeen inaccurately extrapolated to FY2000, or that part of the development cost couldhave being transferred to a different accounting record. Alternatively, the method maysimply overestimate the cost of the F-22 aircraft. The price does seem to be high relativeto our aircraft which is larger and more capable than the F-22. This leaves the accuracyof the method under suspicion but does provide us with a ‘ballpark’ figure to use insubsequent trade-off studies. The value of weight saving ($/lb), as defined in reference13 and shown above, reduces to 3221 for 500 and 2784 for 1000 aircraft. This type ofdata will be very useful in subsequent trade-off work as it links cost changes to aircraftweight.
Estimation of aircraft life cycle costs (LCC) for the aircraft are considered to bemuch too speculative at this stage in the design process, so this calculation has not beenattempted.
8.11 Trade-off studies
As we now have developed all the necessary techniques to analyse the aircraft con-figuration, we can investigate if the aircraft characteristics are the best choice for our
“chap08” — 2003/3/10 — page 262 — #61
262 Aircraft Design Projects
purposes. This is done by sequentially making small changes to the aircraft para-meters and comparing the results to the baseline values. These investigations are called‘trade-off studies’. They may take different forms depending on the purpose of theinvestigation. For example, to determine the best choice of wing and thrust loading, toidentify any constraint that is imposing a critical design penalty on the aircraft, to testthe sensitivity of assumptions that have had to be made to complete the performanceanalysis, and to make a more informed selection of geometric and other characteristics.In some reports and textbooks, such investigations may be referred to as ‘parametricstudies’ or ‘sensitivity analyses’. Examples of such studies are given in references 4and 14.
The list of possible trade-off studies that can be undertaken on a project is obviouslylarge. The selection of which to choose is dependent on the type of aircraft and thepurpose of the study. Here are some suggestions relating to our aircraft:
• To review the selection of aircraft wing loading and associated thrust loading. Thischoice was made previously in the constraint analysis using very crude assumptions.
• To understand, with more accurate analysis, the influence of each of the designconstraints and to recommend changes to these if appropriate.
• To investigate the trade-offs between aircraft parameters (e.g. wing aspect ratio,thickness, sweepback, etc.) and aircraft weight or performance. These parameterswere previously chosen to be similar to existing layouts. This type of trade-off willprovide a more rational basis for the values selected and provide a more efficientconfiguration.
• Test the sensitivity of the assumptions made in the aerodynamic and propulsionanalyses (e.g. drag and lift assessments, engine performance). These results will allowus to focus subsequent work on improving the estimation of those characteristics thatare seen to be most critical to the design.
• To investigate the influence of known critical design drivers. For example, in ourdesign the engine specific fuel consumption translates to the fuel mass and thento the aircraft performance. Making changes to the engine design to improve sfcwill affect several other design parameters (e.g. drag and weight). There must be anoptimum choice of engine configuration to minimise aircraft weight and cost.
As the aircraft system and weapon cost are fixed by the design specification, the mainvariables contributing to aircraft cost are the aircraft empty weight and engine size(thrust). The cost estimation has provided a value for the value of weight saving ($ perpound) and the price of engines. It is therefore possible to translate changes in aircraftweight and thrust directly to aircraft cost.
Many of the choices made in the trade-off studies require a definition of the objective(or goal). In some cases, this may be stated simple as ‘minimum wing weight’, ‘minimumfuel used’, ‘minimum aircraft price’. Sometimes a combination of parameters is used(e.g. weight and size, or structure weight and fuel weight). The ability to use the costtrade-off value in such cases will be very useful.
The difficulty of using trade-off studies lies in the assumptions used in their analysis. Itwould be very time consuming to have to individually analyse the various combinationsof configurations in the detail that has been used to study the baseline design in theprevious sections of this chapter. Trade-off studies at this point in the design processdo not make substantial changes in the basic aircraft layout. They concentrate onrelatively small modifications (e.g. 5, 10, of 20 per cent variations), therefore some of theaircraft parameters may not change significantly in the pursuit of the overall answers.Recognising such parameters allows us to hold them constant, or make them change
“chap08” — 2003/3/10 — page 263 — #62
Project study: advanced deep interdiction aircraft 263
relative to some other variable, and thereby reduce the amount of work. (For example,the aircraft wetted area that is used in the drag calculation can be somehow related tochanges in wing area.) Choosing the assumptions to make at the start of the trade-offstudies is the most difficult part of the process.
As trade-off work involves the repeated calculation of similar types of analysis, itis appropriate to use some form of computer assistance. This may be in the form ofspecifically written computer programs or the use of spreadsheet application software.In this way, and by making suitable assumptions as mentioned above, small variationsin aircraft parameters can be quickly assessed and graphs produced to illustrate thetrends. The use of such methods must be tailored to the specific aircraft configurationand the type of study to be followed. Unless one is fully conversant in the use ofcommercial programs and aware of their limitations (i.e. their validity to the problem),it is unwise to simply ‘turn the handle’ to get results to specific types of study.
It is not possible within the limits of this chapter to perform any trade studies insufficient detail. However, there are plenty of opportunities for students who have fol-lowed the development of the aircraft this far to continue with their own investigations.The question that is still unanswered in this chapter is ‘what is the best (not optimum)configuration for this aircraft?’ This leaves plenty of scope for coursework!
8.12 Design review
From the analysis above, we have shown that the aircraft meets all of the design require-ments apart from the specified range. As the aircraft will be analysed in more depthwith respect to aerodynamic (drag) and propulsion (sfc) characteristics in the followingphases of the design process, it would be unwise to make any substantial changes to theconfiguration at this time. The suggestion to increase the aircraft length by extendingthe engine nozzles made previously will reduce wave drag and this may rectify the rangedeficiency.
It is now appropriate to redraw the aircraft general arrangement to include the minoralterations suggested in the previous design process. This drawing together with a moredetailed internal arrangement drawing and an initial specification of the structuralframework can be seen in Figures 8.37, 8.38 and 8.39.
At this stage in the design process, it is advisable to compile a detailed descriptionof the aircraft so that the work that follows (often by different specialists) will havea common basis. The section below is typical of the detail that should be included insuch a description.
8.12.1 Final baseline aircraft description
Aircraft description
Aircraft type: Two-seat, high altitude, supersonic, low-observable, deepinterdiction aircraft.
Design features: Mid-wing, diamond planform, blended body, tailless,twin-engine layout. All weapons stored internally in a centralbomb bay below the engine and equipment compartments.Side-by-side, high mounted, low-bypass engines with 2Dvariable geometry, under-wing intakes positioned close to the
“chap08” — 2003/3/10 — page 264 — #63
264 Aircraft Design Projects
Scale
Optional
0 2
0 5 10 15 20
4 6
ft
m
+
Fig. 8.37 Final baseline aircraft GA
SensorsCockpitcapsule Equipmt
Intake
Engine Main u/cA/burner
Vectorednozzles
Wing fuel tank
Avionics andnose u/c
FuelTurbofanengine
Munitionsbays
Thrust vectorsAvionics
and nose u/c
Amraams
Cockpit
Sensors
Intakes
EngineNozzles
Weapons bay Undercg
Equipmt
Intake
Body fuel
Wing fuel tank
C
C
Engine u/c
.
Fig. 8.38 Final baseline aircraft internal arrangement
“chap08” — 2003/3/10 — page 265 — #64
Project study: advanced deep interdiction aircraft 265
Frames Longeron
Y20 Y20
Stifn
r
Spar
R15
Longerons
L.E. structure(cooled)
Integral fuel tanksMoveable T.E.control surface
Cockpit slopingbulkhd
Section X95
Access shearpanel
Longerons
Section X165
Root rib
Fuel tank
Root rib
Doors
X95 X165
X165X95Ground line
Section Y20
.
Intake structurenot shown
Fig. 8.39 Final baseline aircraft structural framework
wing leading edge. Afterburning and vectoring rectangularnozzle positioned to the rear of the wing trailing edge.Mid-fuselage, side-by-side, twin pilot cockpit with limitedexternal view. Access to the cockpit is through the forward bombbay bulkhead. Artificial pilot vision and automatic flight controlsystem. Cockpit capsule-escape system. Conventional tricycleretractable landing gear.
Stealth features: Very low radar cross-sectional area, achieved by the blendedprofile with aligned external geometry and structure, and theapplication of radar absorbent materials and structure. Structurecooling to reduce kinetic heating. Shielded and intercooled engineexhaust flow. Polymer coatings to reduce infrared signature, andsound-profiling to reduce the sonic boom.
Structure: Integrated wing and body internal and profiled structuralframework. Extensive use of composite structural materialswith RAM and RAS applied to reduce observable signature.Design limits +7/−3g, VD = M2.0 and max. dynamicpressure = 2133 lb/sq. ft (equivalent to 800 kt at SL).
“chap08” — 2003/3/10 — page 266 — #65
266 Aircraft Design Projects
Weapons: Common racking for combination weapon loads as defined below:(4) Mk-84 LDGP + AIM-120(4) GBU-27 + AIM-120(4) 2000 lb JDAM + AIM-120(4) AGM-154 JSOW + AIM-120(16) 250 lb small smart bomb
Aircraft dataDimensions: Overall length 87.0 ft 26.5 m
Overall span 42.6 ft 13.0 mOverall span (option) 52.5 ft 16.0 mOverall height 11.5 ft 3.5 mWing aspect ratio 1.27Wing taper ratio 0Wing LE sweep 70◦U/C wheelbase 42.3 ft 13.5 mU/C track 14.8 ft 4.5 m
Areas: Wing planform (ref) 1430 ft2 133 m2
Exposed wing 700 ft2 65.0 m2
Total wetted 2472 ft2 230 m2
Max. cross-section 91.5 ft2 8.6 m2
Elevators 56.0 ft2 5.2 m2
Ailerons (normal) 56.0 ft2 5.2 m2
Ailerons (option) 134 ft2 12.5 m2
Weight (mass) Max. TO (design) 114 082 lb 51 739 kgEmpty 44 537 lb 21 025 kgManoeuvre 81 797 lb 37 100 kgLanding (90% MTO) 102 674 lb 46 565 kgFuel load 55 000 lb 25 000 kgFuel (US gals) 10 300Weapons (max.) 13 448 lb 6113 kg
Loadings: Wing loading (max.) 80 lb/sq. ft 3815 N/sq. mThrust/Weight (TO) 0.58 DryThrust/Weight (combat) 0.18 Dry at 50 000 ft
Engines (each): Thrust (ssl, dry) 33 080 lbSFC dry (TO) 0.85SFC dry (cruise) 1.20Weight: mass 7800 lbBypass ratio 0.6
Aerodynamics: Supersonic cruise CDo = 0.0205CL = 0.146CD = 0.0315L/D = 4.65
“chap08” — 2003/3/10 — page 267 — #66
Project study: advanced deep interdiction aircraft 267
Subsonic (clean) CDo = 0.0077(TO and land) CDo = 0.0115(approach) CL = 1.4 (HAA)
Performance: Mission:Cruise speed M1.6Cruise height 54 to 63 000 ftRange 3500 nmManoeuvre (SEP at M1.6 at 50 000 ft):1g dry thrust 60 ft/s1g afterburning 370 ft/s2g afterburning 50 ft/sTurning (M0.9 at 15 000 ft):Instantaneous 13.4◦/s (4069 ft radius)Sustained 5.4◦/s (10 097 ft radius)Sustained (A/B) 8.0◦/s (6815 ft radius)Max. instantaneous 21.0◦/s at 366 kt (1686 ft radius)Field:TO run @ MTO 1009 ft dry (unfactored)Speed V2 70 ktBalanced field 1150 ft normal, 2320 ice (unfactored)Approach speed 140 ktLanding roll (dry) 2530 ft (unfactored)
(ice) 7000 ft (unfactored)(wet) 4700 ft (unfactored)
Recurrent flyaway unit cost (FY2000) $178 M for 200 production$150 M for 430 production$123 M for 1000 production
8.12.2 Future considerations
Although the aircraft layout appears feasible, there are a number of outstanding issuesthat must be resolved. The main concern relates to the directional control and stabilityof the configuration. There are several examples of this type of tailless aircraft flyingto give confidence but these do not fly as fast and are not expected to approach andturn at such a high angle of attack. The design of the vectored thrust and the flightcontrol systems are intrinsically integrated into such analysis. Other issues relate tothe technologies required for stealth. These include the performance of new radarabsorbent materials and structure, the dissipation of the sonic boom, infrared reductionof the nozzle area and of the kinetic heating of the structure. Finally, little has beendone so far to define the systems integration on the aircraft. Many of the expectedimprovements in the aircraft capability are linked to the system design. Hence, thereare plenty of opportunities for further individual studies.
Obviously, this aircraft is a very sophisticated weapon system that relies on the totalintegration of many diverse technology developments. Working in such a design envir-onment, it is difficult to accurately determine the effect of the technologies on aircraftmass, supportability, efficiency, costs and timescales. These problems are common inadvanced aircraft design and add to the interest and fascination of the work.
Before starting the subsequent detailed design stages, it is worth considering futuredevelopments for the aircraft. This may colour future decisions on the layout andcapabilities of the aircraft. Most existing aircraft have not been limited to their initial
“chap08” — 2003/3/10 — page 268 — #67
268 Aircraft Design Projects
design specification. They have been developed from the original concept by extendingtheir capabilities (speed, manoeuvre, range, payload, weapons, etc.) or by adaptingthem for other roles. If such developments could be anticipated, the aircraft would bemore versatile and easier to modify later in its operational life span. For example:
• What new weapons might the aircraft need to carry?• Could an uninhabited version be envisaged?• Would more self-defence be necessary in the future?• Can future threats be anticipated?• Will stealth features need to be reconsidered as offensive systems are improved?• Could a shorter range version utilise the reduction in fuel load by increasing payload?
And what would this mean in respect to weapon storage and delivery systems?• Would flight refuelling be a useful extra feature? And how would this affect the
aircraft payload range and stealth issues?
The application of a SWOT (strength, weakness, opportunity and threat) analysis ofthe aircraft and the operational environment in which it functions provide a moreprocedural method of arriving at the sort of questions listed above.
8.13 Study review
This study has provided an example of the design of an advanced-technology, mil-itary aircraft. It has demonstrated some of the methods and techniques needed toanalyse modern high performance, stealthy configurations. The principal design driverfor the project has been a combination of stealth technology and efficient supersonicaerodynamic performance. Balancing the demands of two such significant require-ments together with combat effectiveness is common in the design of military aircraft.Layout considerations that provide the necessary stealth characteristics and the min-imisation of wave drag have led to the unusual profiling of the aircraft. The analysisof the manoeuvring performance has provided a good example of the ‘energy height’and ‘specific excess power’ methodologies. The assumptions made with respect to lowobservables, minimised wave drag and thrust vectoring in the project may be regardedas somewhat optimistic and unachievable in the ‘real’ world but they are intended tooffer a forward vision for combat aircraft design. Although incomplete, the study pro-vides a useful starting point for several continuation projects that would be suitable forunder-post-graduate coursework.
The author acknowledges the work of previous students on this project15,16 in com-pleting their submissions to the AIAA design competition. Their groundwork in datacollection and the development of analytical methods has assisted in the writing of thischapter.
References
1 www.aiaa.org.2 www.aiaa.org/education/undergradaircraft.pdf.3 Jenn, D., Radar and Laser Cross Section Engineering, AIAA Education Series, 1995, ISBN
1-56347-105-1.4 Raymer, D. P., Aircraft Design: A Conceptual Approach, AIAA Education Series, 1999, ISBN
1-56347-281-0.5 Brandt, S. A. et al., Introduction to Aeronautics: A Design Perspective, AIAA Education
Series, 1997, ISBN 1-56347-250-3.
“chap08” — 2003/3/10 — page 269 — #68
Project study: advanced deep interdiction aircraft 269
6 Aviation Week Source Book, published annually in January.7 Eshelby, M. E., Aircraft Performance – Theory and Practice, Butterworth-Heinemann and
AIAA Education Series, 2000, ISBN 1-56347-250-3 and 1-56347-398-4.8 AIAA Aerospace Design Engineers Guide, 1998, ISBN 1-56347-283-X 1.9 ESTOL aircraft landing profile for the X31 demonstrator (www.aviationnow.com).
10 McCormick, B. W., Aerodynamics, Aeronautics and Flight Mechanics, Wiley and Sons, 1979,ISBN 0-471-03032-5.
11 Mattingly, J. D., Aircraft Engine Design, AIAA Education Series, 1987, ISBN 0-930403-23-1.12 Nicolai, L. M., Fundamentals of Aircraft Design, METS Inc., San Jose, California 95120,
USA, 1984.13 Society of Allied Weight Engineers Inc., J. Wayne Burns, ‘Aircraft cost estimation methodo-
logy and value of a pound derivation for preliminary design development applications’,SAWE Paper No. 2228 Cat. No. 29, May 1997.
14 Jenkinson, Simpkin and Rhodes, Civil Jet Aircraft Design, AIAA Education Series andButterworth-Heinemann, 1999, ISBN 1-56347-350-X and 0-340-74152-X.
15 Southampton University, UK, ‘AIAA undergraduate team design competition – Groupreport’, June 2002.
16 Loughborough University, UK, ‘AIAA undergraduate team design competition – Groupreport’, June 2002.
“chap09” — 2003/3/10 — page 270 — #1
9
Project study: high-altitude,
long-endurance (HALE)
uninhabited aerial
surveillance vehicle (UASV)
Global Hawk UAV
Existing American reconnaissance aircraftExisting American reconnaissance aircraft
Lockheed Martin U–2R
“chap09” — 2003/3/10 — page 271 — #2
High-altitude, long-endurance (HALE) uninhabited aerial surveillance vehicle (UASV) 271
9.1 Introduction
This project study was stimulated by a technical paper1 presented by European defencestaff in 2000. The study has been used in this book to highlight the unique problemsassociated with operations at high altitudes and for long periods. A novel configura-tion has been investigated to allow comparison with conventional designs. The systemand communication problems associated with remote and autonomous control of theuninhabited vehicle has not been addressed as this is outside the scope of the ini-tial/conceptual aircraft design task. Such an investigation could form a suitable topicfor system/electronic/electrical/avionics design courses.
9.2 Project brief
Aerial reconnaissance has always been an essential feature of military intelligence. Thefirst use of aircraft in a military context was as artillery spotter planes at the start ofWorld War I. At this time, airships were used for reconnaissance but they soon becametoo vulnerable to ground fire. High-altitude surveillance was perfected during the startof the Cold War. At this time, anti-aircraft munitions were unable to reach high flyingaircraft. The incident in which a US pilot (Gary Powers) flying a U2 ‘spy plane’ at65 000 feet was shot down over Russia curtailed such operations over hostile territory.The exploitation of the pilot by his captives and the ensuing political and diplomaticconsequences has given rise to the requirement for unmanned flights in dangerousmissions. Surveillance has subsequently been more safely undertaken by sophisticatedsatellite systems.
Following the end of the Cold War, many national airforces have been deployed ininternational peacekeeping roles for the UN and other bodies. Part of such activitiesinvolves the monitoring of ‘no fly’ and demilitarised zones. This requires continuous(day and night), all-weather surveillance over large areas. Although, in such operations,there is only a small chance of a threat to the aircraft, the political consequences ofdealing with unfriendly governments holding a pilot as hostage (like Gary Powers andGulf War prisoners) sets a requirement for unmanned autonomous operations. Thereare few such aircraft in existence. Many airforces have piloted surveillance aircraftbut these are used for tactical military support (e.g. target designation and damageassessment). They are mostly operated over short range, at modest altitude and forshort duration.
An aircraft possessing the capability to monitor for long periods and operate fromremote bases would also be appropriate for some civil or quasi-civil operations. Suchroles may include:
• Maritime patrol• Drug law enforcement• Remote high-value facility protection• Civil disorder• Border control and police surveillance• Traffic intelligence• Environmental protection• Disaster management
“chap09” — 2003/3/10 — page 272 — #3
272 Aircraft Design Projects
As a research vehicle for environmental studies, such an aircraft could be used tomonitor and report on atmospheric/climatic conditions, weather intelligence (provid-ing near-earth observations) to supplement satellite data. The ability to fly for longendurance at high altitudes could also be used to provide communication links whereradio or satellite facilities are unavailable or inadequate.
9.2.1 Aircraft requirements
The design requirements for the aircraft are based on data from a design/research paperpresented at a conference in 2000.1 They are relatively straightforward:
• payload (reconnaissance systems) 800 kg (1760 lb),• ability to easily reconfigure the systems on the aircraft,• ferry range, from the home airfield = 6000 nm,• operational radius from advanced base = 500 nm,• patrol duration 24 hours,• all weather, day/night capability,• short, rough field (unspecified) capability without the need for specialised ground
launch or recovery systems (e.g. catapult/rocket launch, arresting wires),• quick operational readiness,• quick turnaround between missions,• cost efficient (not necessarily minimum first cost) system,• safety systems to avoid or reduce co-lateral damage in the event of an aircraft failure,• structural loading +2.5/−1.25g.
For comparison, the US Global Hawk HALE-UASV is reported to be capable of flying1200 nm, spending 24 hours on patrol at 60 000 ft (about 18 km) and returning to base.This is a reputed 32 hour mission!
9.3 Problem definition
We all know that if we stand on the shoreline and look out to sea, the horizon is onlyabout five kilometres away. If we climb up to the top of the cliff, we can see muchfurther out to sea. From a study of observation geometry, we understand that themaximum observation range is a function of height above sea level and the radius ofthe earth (Figure 9.1). The earth is not a pure sphere. It is flattened at the polar regionsrelative to the equator. The exact values2 for the earth radii are 6356.9 km at the polesand 6378.4 km at the equator. In the analysis below, the often-quoted mean value of6378.1 km (3444 nm) is assumed.
When we are flying, if we look down below the horizon the observation range becomesshorter and the view is clearer. This improved vision is partly due to reduced atmo-spheric contamination and ground clutter that adversely affect longer viewing distances.Simple trigonometry can be used to determine the observation range at various alti-tudes and for different downward-viewing angles (known as slant angles). These resultsare shown in Figure 9.2.
At an altitude of 20 km (about 65 000 ft), the horizon (zero slant angle) is about500 km (270 nm) away. Looking down 5◦, the distance to the ground reduces to about200 km (108 nm). The corresponding figures for 10 km altitude are 350 km (190 nm)and 100 km (54 nm) respectively. If the aircraft reconnaissance equipment is powerfulenough, it is obviously an advantage to operate at high altitude to observe a greater
“chap09” — 2003/3/10 — page 273 — #4
High-altitude, long-endurance (HALE) uninhabited aerial surveillance vehicle (UASV) 273
Observationheight
HSlant
angle ±
Radius ofthe Earth
Tangent toEarth's surface
Max. observation range Rmax
Fig. 9.1 Observation geometry
24
20
16
Hei
ght
(km
)
12
8
4
00 100 200 300 400 500
Max. ground observation range (km)
10°5° 3°
2° 1° 0°
Increasinglypoorvisibility
Slant angle ±
Fig. 9.2 Observation range
surface area. Published photographs of ground surveillance from satellites show thatsuch reconnaissance equipment is feasible and available. Operating at high altitudesmeans that a greater ground area can be observed. Therefore, fewer missions or aircraftare required to sweep a territory. High-altitude operation also provides a stealthiermission which reduces aircraft vulnerability. In the situation where the territory to beobserved is ‘hostile’, the aircraft could stand off from a border and still be capable ofobserving over the border for a considerable distance (Figure 9.3).
The aircraft is required to be capable of operating in all weather conditions. In thisrespect, one of the most important factors to be considered is the strength of winds.The average wind speed on a statistical basis varies with altitude as shown in Figure 9.4.Flying in constant wind has little influence on the aerodynamic characteristics of theaircraft as the forces on the aircraft are related to the relative/local air movements.It does, however, affect the perceived speed over the ground. This may influence theperformance of the search equipment. Winds also affect the aircraft climb performancerelative to the ground. In any event, it is desirable to operate in the calmest atmosphericconditions available. Figure 9.4 shows that average wind speeds are less at 18 to 20 kmaltitude. This region is selected as the best patrol altitude.
“chap09” — 2003/3/10 — page 274 — #5
274 Aircraft Design Projects
Friendlyterritory
Hostilenation
territory
Nation
al bo
unda
ry
Communicatio
n
Aircraft
Surveillance
Fig. 9.3 Stand-off surveillance
405
10
15
Alti
tude
(km
)
20
25
30
50 60 70Wind speed (m/s)
80 90 95
Wind minimum 18–20 km
95% wind90% wind
80% wind
Fig. 9.4 Wind speeds1
Winds are seldom present in a stable direction or at a constant speed. This variationleads to the generation of vertical wind loading on the aircraft (gust effects). Althoughgusts are less pronounced at higher altitudes they must still be considered. Such gustactivity leads to dynamic changes in aircraft lift. Aircraft with a low wing loading orwith a high value of ‘lift-curve slope’ are more susceptible to gust disturbance. Aircraftdesigned for high altitude operation are likely to require both of these aerodynamiccharacteristics. The influence of gusts on the aircraft flight behaviour will need to becarefully assessed especially when operating at low altitudes (e.g. take-off and landing).Figure 9.4 indicates that at low altitude, average wind speed can be almost doublethat at the operating height. To mitigate the effects of gusts on the aircraft dynamicbehaviour, it will be necessary to equip the aircraft with powerful gust load alleviationand some form of auto-stabilising systems. If the surveillance equipment is affected bygust activity it will be necessary to fit the aircraft with ride stabilising control systems.
“chap09” — 2003/3/10 — page 275 — #6
High-altitude, long-endurance (HALE) uninhabited aerial surveillance vehicle (UASV) 275
Flying at high altitude is affected by an increase in stall speed due to the reduction inair density. There is also a reduction in the speed of sound due to the reduced air tem-perature and this will reduce the speed at which aerodynamic buffet will be felt. Thesetwo effects combine to severely reduce the available aircraft speed range (i.e. minimumto maximum operating speeds). This consequently increases the pilot/system workloadto avoid the stall or buffet regions. The U2 aircraft was notoriously difficult to handlein this context. At a later stage in the development of the aircraft a detailed analysis ofthis problem will need to be undertaken to establish the best operating speed and theconsequences on the design of the aircraft flight control system. At this early stage inthe project, these issues are not considered in any more detail.
9.4 Initial design considerations
As the observation area increases with operational height and the vulnerability ofthe system is reduced, a satellite-based system may be regarded as a natural choice.However, such systems are known to be expensive and are inflexible in operation. Forexample, changing the orbit of a satellite to observe a new territory is often impossibleand always time-consuming, complex and expensive. Together with these difficulties,it is often impossible to change the reconnaissance equipment to match the new oper-ational requirements. For these reasons, a satellite-based system is not consideredsuitable for the current brief.
A lighter-than-air aircraft (airship or balloon) could achieve the long endurancespecified for the system but these are, in general, large and slow-moving vehicles.This makes them unsuitable for quick response. They are also vulnerable to hostileaction and incapable for all-weather operations. Such aircraft are therefore consideredunsuitable options for this project.
A specifically designed aircraft would seem to offer the best option to meet the designbrief. The design could be relatively cheap, very flexible in operation, quickly refittedto suit different missions and be able to be sent rapidly to new areas of strategic signifi-cance. Such aircraft could be uninhabited to avoid pilot/observer vulnerability in poten-tially hostile areas. The design could also be designed to house a pilot for operationsthat were deemed to be safe and benefit from human observation and on-board control.
9.5 Information retrieval
Reviewing significant reconnaissance aircraft, found in the published literature,3,4
produces the following information. Most air forces in the world have aircraft in theirfleets that are used for reconnaissance. Many of these are converted civil transportaircraft (mainly turboprop regional airliners). The list below is representative of suchtypes:
Embraer BraziliaDe Havilland Dash 8BAE(Hindustan) HS748Boeing 707, 747Fokker 50Ilyushin 76, 18, 86Tupolev 134Saab 34Pilatus BN2TFairchild Metro
“chap09” — 2003/3/10 — page 276 — #7
276 Aircraft Design Projects
None of these aircraft would be suitable for the specified missions as they are inca-pable of flying at high altitude and would not be able to achieve the specified range orendurance.
Three special-purpose aircraft were found in the review and are described below.
9.5.1 Lockheed Martin U-2S
This is a re-engined and newly equipped/upgraded version of the aircraft type that waspiloted by Gary Powers and shot down on that fateful mission over Russia in the 1960s.The new version was first delivered into US service in 1994. The more fuel-efficient newengine has increased the aircraft range, endurance and operating capability. The highaspect ratio, large area wing is of traditional design and construction. The slenderfuselage accommodates a single pilot and carries conventional tail surfaces and rearair brakes. Below are some of the aircraft details:
Wing span 31.4 m 111 ftOverall length 19.2 m 63 ftWing area (ref. gross) ∗sq. m ∗sq. ftWing aspect ratio ∗Empty mass 8074 kg 17 800 lbTO mass 18 145 kg 40 010 lbCruise speed 192 m/s 373 to 470 ktCeiling (operational) 22.4 km 70 to 73 400 ftCeiling (absolute) 27.4 km 90 000 ftRange 3800 nmDuration 12–15 hoursStructural limit 2.5gEngine: F188-GE-101 turbofanMax. thrust (total) 84.5 kN 19 000 lb
(∗As some of the details of the new U-2S aircraft are unavailable, the table above doesinclude figures from the earlier model.)
New sensors installed in the updated equipment include electro-optical multi-spectral cameras and advanced synthetic aperture radar systems. Equipment is housedin a forward interchangeable nose bay, behind new fuselage hatches and in wingmounted pods.
9.5.2 Grob Strato 2C
This German aircraft first flew in 1995 (as a proof of concept vehicle). To maintainsufficient power up to an operating height, reputed to be 24 km (78 700 ft), it has twocompound piston engines incorporating a two-stage turbocharger with intercooler. Thetwo engines drive large five bladed pusher propellers mounted aft of the wing trailingedge. The engines are rated at 300 kW (400 hp) each. The large area, high mounted,straight, laminar flow wing has a high aspect ratio. The conventional pressurised fuse-lage carries tail surfaces, an outrigger mounted main undercarriage and nose unit. Theaccommodation comprises two flight crew and two observers, work stations, rest areasand associated galley and toilet provision. Details are shown below:
Wing aspect ratio 21.3Empty mass 6650 kg 14 663 lbTO mass 13 350 kg 29 440 lbPayload 1000 kg 2200 lbCruise speed 100 m/s 194 ktCeiling (operational) 16–26 km 52 500–85 250 ftCeiling (design) 24 km 78 700 ft
“chap09” — 2003/3/10 — page 277 — #8
High-altitude, long-endurance (HALE) uninhabited aerial surveillance vehicle (UASV) 277
Range 9773 nmDuration 48 hours @ 18 km 59 000 ftMax. duration hours @ 24 km 78 700 ft
Equipment includes various atmospheric monitoring instruments, observation systemsand radar.
9.5.3 Northrop Grumman RQ-4A Global Hawk
This aircraft is a high-altitude, long-range, uninhabited aircraft designed to providemilitary field commanders with high-resolution, near-real time imagery over large geo-graphical areas. It is therefore close to the same operational specification as our designproject. At the time of writing, little technical data on the aircraft was available due toUS national security.
Wing span 116.16 ftOverall length 44.33 ftOverall height 15.16 ftMax. TO weight 25 600 lbPayload 2000 lbOperational height 65 000 + ftMax. range 13 500 nmDuration 36 hours
Due to their potential for long endurance flights, powered gliders have also beenconverted for reconnaissance operations. Two examples are described below.
9.5.4 Grob G520 Strato 1
First flight, as a derivative of the Egrett, was in 1995. This was a proof of conceptproject as a high altitude ‘research’ vehicle. It is powered by an Allied Signal turboprop(560 kW). The aircraft has a large area, high aspect ratio wing and a conventionalfuselage with rear control surfaces. It carries one pilot and the fuselage accommodatessix interchangeable equipment modules. Data is listed below:
Wing span 33 m 108 ftOverall length 12 m 39 ftWing area (ref. gross) 40 sq. m 430 sq. ftWing aspect ratio 27.5Empty mass 2700 kg 5950 lbTO mass 4700 kg 10 360 lbCruise speed 50 m/s 97 ktCeiling 16 km 52 500 ftRange 1930 nmDuration 13 hoursStructural limits 3.3/ − 1.3g
9.5.5 Stemme S10VC
This aircraft is also a derivative of an existing powered glider. Its unique feature consistsof a propeller designed to fold into the fuselage nose fairing. It first flew in 1990as a sensor platform for atmospheric research. It is a classically configured, twin-seat, composite-constructed glider with extra equipment pods mounted on each wing.Details are shown below:
Wing span 23 m 75.3 ftOverall length 8.4 m 27.5 ft
“chap09” — 2003/3/10 — page 278 — #9
278 Aircraft Design Projects
Wing area (ref. gross) 18.7 sq. m 200 sq. ftWing aspect ratio 28.2Empty mass 670 kg 1477 lbTO mass 980 kg 2161 lbCruise speed 46 m/s 89 ktLift/drag ratio 50 @ 30 m/s 58 kt
Powered glider derivatives employ a power-off ‘glide down’ technique to save fuel andextend flight duration.
Another aircraft project to note, because of its unorthodox configuration, is theBoeing project study for a Common Support Aircraft. Though this aircraft is notdirectly related to our design brief it houses an active phased 360◦ scanning radarmounted conformally in the unconventional wing structure. This is possible due to theadoption of the ‘joined wing’ layout. The 40◦ swept back wings are joined towardstheir tips to tail mounted, swept forward wings. The radar system is mounted on thewing surfaces to provide nearly all-round scanning. The aircraft, which is designedto fly from/to aircraft carriers, accommodates one pilot and three system operators.Few details of the aircraft are released but it is reported to have a wing-span of19.3 m (63.3 ft), an overall length of 15.5 m (51 ft), max. mass of 25 535 kg (56 300 lb),max. speed of M0.8 and a patrol speed of M0.38. Although not directly comparable tothe design brief under consideration, the aircraft design illustrates how technical inno-vation can be used to provide an efficient alternative configuration to conventionallayouts.
9.6 Design concepts
As the project brief defines a unique mission, the literature search did not reveal anyaircraft that could match the requirement. The nearest are seen to be the U-2S, Strato-2and the Global Hawk.
In order to fly at high speed for quick transfer onto operational stations and tobe able to fly at the high Mach number needed for patrol at high altitudes, propellerpropulsion systems are not feasible. A turbofan engine with a modest bypass ratio(BPR) that balances low fuel consumption with acceptable thrust loss at high altitudeis the preferred choice. An engine with a BPR of between three and five will be compactenough (i.e. a medium size fan diameter) to give a feasible installation on the aircraft.Two engines will be selected to provide improved flight safety, although there will bean increase in engine failures and maintenance.
The fuselage will need to house a variety of different operational equipment/systems.It will also be desirable to be able to quickly change the equipment fits between sorties.This suggests a modular design. Downward and forward sensor viewing will be essen-tial, therefore the fuselage structural frame must consist of a beam stretching fore andaft above the equipment bays. This will allow different modules to be suspended belowthe beam.
The choice between manned and uninhabited/autonomous operation needs to becarefully considered. The development and flight testing of a new aircraft concept withuntried and sophisticated characteristics may be most easily achieved using a humantest pilot. Conversely, operating in a hostile environment will demand an unmannedvehicle to avoid the political and diplomatic difficulties that can arise if the aircraft iscaptured. The best compromise between these requirements is to consider the cockpitto be a ‘unique system’ module that can be suspended at the front of the fuselage beam
“chap09” — 2003/3/10 — page 279 — #10
High-altitude, long-endurance (HALE) uninhabited aerial surveillance vehicle (UASV) 279
structure. As confidence in the aircraft flying characteristics and remote control systemperformance is gained during flight testing, the cockpit module could be replacedby the autonomous flight system. In service, variants of the aircraft could be eitheruninhabited or piloted as suits the mission.
Reviewing the layout options for the configuration of the aircraft leads to fourpossibilities:
1. Conventional (glider type development)2. Joined wing (Boeing CSA example)3. Flying wing (USA B2 example)4. Braced wing (NASA research)
Each of these layouts will be described and considered for the design brief.
9.6.1 Conventional layout (Figure 9.5)
In many ways, this option offers an attractive choice as the wing is aerodynamicallyefficient and the fuselage acts as a central beam structure which supports the wing, tailsurfaces, engines, landing gear, fixed equipment and mission system modules. The aero-dynamic and structural features and the associated analyses are well developed. Thus,this option provides low commercial risk with regard to the technological development.
There are two main concerns. First, the long wing-span, and its low height abovethe ground may make remote piloting difficult during crosswind landings. In fact itwould be essential to use a sophisticated auto-land system. The small angle from themain undercarriage contact point to the wing tip will make accurate approach attitudeimperative. The aircraft has a low wing loading that will make it very susceptible togust disturbance. This will make precise landing manoeuvres difficult in all-weatherconditions. Touching the wing tip on the ground during landing will inevitably lead tothe classical ‘ground looping’ phenomena.
Second, there may be difficulty in producing the wing thin enough to delay the criticalMach number up to the anticipated operating speed. With a long span, the root bending
Fig. 9.5 Conventional layout option
“chap09” — 2003/3/10 — page 280 — #11
280 Aircraft Design Projects
Fig. 9.6 Joined-wing layout option
moments will require a deep section to avoid excessive wing structural mass. Even if itwere feasible to design a thin wing, its internal volume would not be sufficient to holdthe required fuel load.
9.6.2 Joined wing layout (Figure 9.6)
Although this configuration has been promoted in new civil aircraft projects5 it has notyet been fully validated in a flying prototype. Even the Boeing CSA project was nottaken into the manufacturing phase. There is serious concern about the flow conditionat the wing junctions and the interference of the flow field between the two surfaces.These problems could disturb the critical transonic flow conditions and reduce thecritical Mach number.
The joined wing is also seen to have difficulty in the positioning of the main under-carriage unit. The wings to fuselage attachment points lie far ahead and well behindthe aircraft centre of gravity (cg). The main landing gear unit must be located onlyslightly behind the cg position. This results in a heavy and complicated fuselage struc-ture. Alternatively a bicycle undercarriage layout could be employed but this is heavyand makes it difficult to rotate the aircraft on take-off and landing.
Notwithstanding these disadvantages, the layout may offer a novel solution to theinstallation of the otherwise cumbersome antenna required on the aircraft (see theBoeing design).
9.6.3 Flying wing layout (Figure 9.7)
Although several aircraft of the pure flying wing configuration have been proposedand some designs have been flown in the past they have not yet been fully exploited in
“chap09” — 2003/3/10 — page 281 — #12
High-altitude, long-endurance (HALE) uninhabited aerial surveillance vehicle (UASV) 281
Fig. 9.7 Flying-wing layout option
production aircraft (with the exception of the B-2 stealth bomber). Many enthusiastsfor the type have claimed the layout to be aerodynamic and structurally efficient but itseems that such expectations have not yet been realised (the B-2 layout is selected forstealth reasons). The reason may be due to the linking of stable torsional deflections tothe aerodynamic forces and the consequential requirement to modify the wing planformand sectional geometry to avoid this problem. The layout is regarded as efficient at onedesign point but seriously compromised away from this condition. The flying wing con-figuration was considered in the German study1 but dismissed on these technical issues.
9.6.4 Braced wing layout (Figure 9.8)
The structural bracing of the wing to the fuselage was a common feature in historicaircraft layouts. This was done to reduce the loads in the wing to match the relativelypoor structural properties of the materials used in the construction. The developmentof stronger and more consistent materials allowed such bracing and the associated dragpenalty to be eliminated. The traditional monoplane wing layout has been the preferredchoice over the past several decades. As wing aspect ratio is increased, the benefit ofbracing becomes more attractive as it significantly reduces wing bending moments.In recent years, some NACA funded research6 has shown that wing bracing couldprovide advantages to the design of long-range civil jet transport aircraft. The purelytensile loaded brace reduces the shear, bending and torsion on the wing structure. Thiscorrespondingly allows either a thinner wing or a larger aspect ratio to be used onthe wing geometry. A thinner wing would allow the wing to be less swept for a designcritical Mach number. All of these effects reduce aircraft drag and consequently fuelburn. Positioning the brace attachment to the wing ahead of the sectional structuralaxis also provides a reacting nose-down moment to stabilise the divergence tendencyassociated with a swept forward wing planform.
Mounting the wing on the fin structure and adding dihedral to counter the unstableyaw coupling from the swept forward planform places the wing well above the ground
“chap09” — 2003/3/10 — page 282 — #13
282 Aircraft Design Projects
Fig. 9.8 Braced-wing layout option
plane during landing. This provides the aircraft with adequate bank angle to protectagainst disturbed landing manoeuvres.
The main drawback with the configuration is associated with the novelty of the layoutand its potential for technical risk.
9.6.5 Configuration selection
Of the four options, only the conventional and braced wing seems to be worth furtherconsideration. As the German design study1 selected a conventional layout for theirbaseline design we will investigate the braced wing layout. This will provide a usefulcomparison with the previous study.
Having selected the braced wing layout there are several detail design considerationsto be made:
1. The engine mountings, fuselage brace attachments and the main undercarriagemounting will be combined into a central fuselage structural framework. Thiswill leave the forward fuselage structure uncluttered and capable of holding theequipment modules as conformal containers below the fuselage structural beam.
2. To avoid the difficulty of attaching the brace to the wing structure, and the possibilityof complex airflows at the junction, pylon mounted equipment/fuel pods will beinstalled on the wing. The brace will be attached to the pod support structure (seeFigure 9.9).
3. The brace structure will need to be streamlined and this will provide the opportunityto run equipment service lines or fuel supply pipes directly between the wing andfuselage.
4. It may be possible to use the wing and brace structures to house conformal radarantenna (as proposed by Boeing on their CSA).
5. To reduce trim drag (an important feature on long-range/endurance aircraft) theforward fuselage could support a small canard surface mounted above the equip-ment modules. Care will need to be taken on the position of this surface relative
“chap09” — 2003/3/10 — page 283 — #14
High-altitude, long-endurance (HALE) uninhabited aerial surveillance vehicle (UASV) 283
A A
(a) Connection without attachment pod. Flow interference at wing joint A.
(b) Connection using attachment pod. Improved flow at
junction A.
Fig. 9.9 Wing to brace interconnection detail
to the engine intakes. Wind tunnel tests will need to be done to finalise the exactgeometry.
6. It will be necessary to incorporate wing inboard control surfaces to provide pitchcontrol.
7. Although main wheels will be required, it may be preferable to use skids for thethird (nose or tail) unit.
9.7 Initial sizing and layout
While the aircraft is of an unconventional configuration, the initial sizing and lay-out process will follow the normal procedure. This will involve estimating the aircrafttake-off mass, wing loading, some airspeed predictions, wing layout and powerplantsizes. These are described in the following subsections. Finally, all of the componentstudies are linked together to produce the initial baseline aircraft layout.
9.7.1 Aircraft mass estimation
In order to size the aircraft it is necessary to estimate the maximum take-off mass.The formula below is often used for this purpose (see Chapter 2, section 2.5.1 for thedefinition of terms):
MTO = (MUL)/{1 − (ME/MTO) − (MF/MTO)}For the HALE aircraft there are some difficulties that arise from the definitions ofaircraft systems to be included in the aircraft empty mass ratio. Many of the systems onthe aircraft are directly related to the type of operation. Some of the equipment may bechanged to suit the mission (reconnaissance, communication, surveillance, atmosphericresearch and monitoring). To resolve the lack of knowledge of these systems and thevariability with the mission, the equipment mass will be assumed to be 800 kg. Thisvalue will be attributed to the ‘useful load’ in the above equation. At a later stage in thedevelopment of the design it will be appropriate to conduct sensitivity analyses aroundthis assumption.
A second difficulty arises due to the expected, unusually large, fuel ratio. An aircraftwith a duration of 24 hours is almost unique. Therefore data from other, shorter-range aircraft may be misleading. For this reason it will be essential to check the fuelrequirements as soon as the aircraft mass, lift, drag and engine characteristics areknown with reasonable accuracy.
“chap09” — 2003/3/10 — page 284 — #15
284 Aircraft Design Projects
The main conclusion from these observations is that the estimation of aircraft max-imum mass from the above expression must be treated with suspicion and regardedas tentative. Several iterations of the subsequent analysis will be necessary beforeconfidence in the results can be realised.
Analysis of the technical descriptions of the main aircraft types described insection 9.5 shows the following values for empty mass ratios:
U-2S 0.445Stratos 1 0.570Stratos 2 0.500Stemme 0.680
These contrast with a value of 0.254 for the German EADS project.1 Our aircraft designbrief is closer to the U-2S and EADS aircraft. The U-2S is recognised as being an oldaircraft design with 1950/1960’s materials and construction methods. Application ofmodern materials and methods would be expected to reduce the structural mass. TheEADS aircraft data may have linked more of the equipment mass to the useful loadcomponent than assumed in our case. Without more information, it is difficult tochoose between these two values for empty mass ratio, therefore an average figure of38 per cent will be initially used for our design.
The fuel fraction can be estimated using the Breguet range equation:
(MF/MTO) = (engine cruise sfc) · [1/(L/D)] · (flight time)
Note: engine sfc varies with cruise altitude. For a typical medium bypass ratio turbofanengine, the following relationship is quoted7:
(sfc)altitude/(sfc)sea level = θ0.616
where θ is the ambient air temperature ratio (TA/TSL). In the stratosphere the ISAtemperature is constant at 216.76 K. ISA sea-level temperature is 288.16 K. This makesθ = 0.75.
Hence,(sfc)altitude = 0.84(sfc)sea level
A medium BPR engine is likely to have a sea-level sfc of 0.55 (lb/lb/hr or N/N/hr).Therefore using the above formulae gives an engine sfc in the stratosphere of 0.46.
With a high aspect ratio wing and slender fuselage the aircraft lift to drag ratio (L/D)in cruise could safely be assumed to be better than the value of 17 which is typical ofmodern civil airliners. Due to the forward swept wing and the interference arising fromthe brace structure it will not be possible to achieve the value of 40 which is typical ofhigh performance gliders. Being conservative, we will assume a value of 25 but this willneed to be checked and adjusted when detailed drag estimations are available later inthe design process.
The duration of the patrol is specified as 24 hours in the design brief. It is unclear ifthis is to include the time needed to reach the patrol area, so an extra two hours will beadded to this time. A design duration of 26 hours will be used in the analysis below:
Hence,(MF/MTO) = (0.46) · (1/25) · (26) = 0.48
We will add 10 per cent for contingencies to give a design value of 0.53.Using the above values in the initial aircraft take-off mass equation gives:
MTO = 800/(1 − 0.38 − 0.53) = 8888 kg (19 600 lb)
To provide for some design flexibility in the subsequent work a design (max.) mass of9200 kg (20 280 lb) will be assumed.
“chap09” — 2003/3/10 — page 285 — #16
High-altitude, long-endurance (HALE) uninhabited aerial surveillance vehicle (UASV) 285
9.7.2 Fuel volume assessment
The calculation above predicts a fuel mass of 0.53 × 8888 = 4693 kg (10 350 lb). Witha specific mass for aviation fuel of 0.8 (fuel varies between 0.76 and 0.82), this masswill need 5833 litres (5.866 m3, 207 ft3) tankage volume to hold the fuel. When the winggeometry has been defined, a check will be necessary to establish if the volume can beaccommodated in the integral wing tanks. If not, the size of other storage tanks to beincluded in the aircraft layout will need to be determined.
9.7.3 Wing loading analysis
Flying at high altitude where the air is thin requires either a fast airspeed or a high valueof lift coefficient (or probably a combination of both) to reduce the required wing area.Maximising these parameters for a chosen altitude sets the value for the maximumwing loading as shown below:
L = 0.5ρV 2SCL
Substituting ρ = p/RT , and using a = (γ RT )0.5, where (ρ) is air density, ( p) is airpressure and (a) is the speed of sound.
Assuming γ = 1.4 and using M = Mach number (=V/a), gives the equation:
L/S = 0.7pM2CL
Typical values of the parameter (M2CL)max, range from 0.05 for gliders to 0.6 for mili-tary jets. Conventional civil transports lie in the range 0.2 to 0.4 (i.e. M0.8 @ CL = 0.4gives (M2CL) = 0.24). Figure 9.10 shows the distribution of maximum wing loadingagainst altitude for various values of (M2CL)max. The areas marked for each type rep-resent the common values. Obviously some aircraft are designed to operate away fromthese regions. Figure 9.11 shows the portion of the previous figure relating to highaltitude operations. As discussed in section 9.3, calmer wind conditions are found at
00 5 10 15 20
2000
4000
6000
8000
10 000
12 000
Cruise altitude (km)
Win
g lo
adin
g (N
/m2 )
0.2 0.3 0.4
(Ma2CL)max
Gliders
Civil airlines
Military jets
UASVs
Fig. 9.10 Max. wing loading versus altitude
“chap09” — 2003/3/10 — page 286 — #17
286 Aircraft Design Projects
015 16 17 18 19 20 21 22
500
1000
1500
2000
2500
3000
Cruise altitude (km)
Win
g lo
adin
g (N
/m2 )
0.2
0.3
0.4(Ma2CL)m
Project design point
UASV operating range
..
Fig. 9.11 UASV wing loading selection
altitudes around 18 km (59 000 ft). This sets the design point shown in Figure 9.11.Therefore, the selected wing loading is 1800 N/m2 (183.5 kg/sq. m, 37.6 lb/sq. ft). Forour chosen aircraft design mass of 9200 kg (20 280 lb), this equates to a minimum wingarea of 50 m2 (537 sq. ft).
9.7.4 Aircraft speed considerations
The aircraft operating envelope is bounded at slow speed by the aircraft stall perfor-mance. At high speed, the operating envelope is restricted by the available engine thrust,the rise in transonic wave drag and the effects of the associated buffet on the aircraftstructure. The effect of high altitude operation affects both of these speed boundaries.For a given wing area and sectional CLmax value, the stall speed will increase as airdensity reduces as defined below:
Vstall = [L/(0.5ρSCLmax)]0.5
As air temperature reduces with altitude (up to the start of the stratosphere) the speedof sound and thereby the aircraft speed at the onset of transonic flow will reduce. Thespeed of sound is determined by the relationship:
a = aoθ0.5
where (ao) is the speed of sound at sea level = 340.29 m/s, 661 kt.These effects are shown diagrammatically in Figure 9.12.It is advisable to fly at a speed greater than the stall speed to allow a margin of
safety to protect against gusts. This margin will avoid inadvertent stalling and reducepilot/system control demands. This defines the minimum speed boundary. For manytypes of aircraft the margin is set by applying a factor of 1.3 to the stall speed in thelow speed, approach to landing, phase. High wind speeds may demand an increasein this margin. Although wind speed does not affect the aerodynamic parameters ofthe aircraft (the aircraft travels with the ambient air and only relative changes aresignificant) it does alter the perceived (relative to the ground) climb and descent flight
“chap09” — 2003/3/10 — page 287 — #18
High-altitude, long-endurance (HALE) uninhabited aerial surveillance vehicle (UASV) 287
Feasible flightregime
Flightsafetyzone1.3 Vs
Aircraft stall (Vs)boundary
Aircraft speed (m/s TAS)
24
20
16
12
8
4
0100 200 300 400
Alti
tude
(km
)
Troposphere
Stratosphere
Speed of soundM1.0 boundary
Transonic flowzone: buffetboundary
Fig. 9.12 Operating speeds constraints (diagrammatic)
1050 100 150 200 250
12
14
16
18
20
22
24
Aircraft speed (m/s TAS)
Alti
tude
(km
)
Selectedaircraftdesign/operationalpoint(H =18 kmV = 210 m/s)
Transonicflow andbuffet
Flight safety speedboundary = 1.3 Vs
Aircraft stall boundary (Vs)(M = 0.9MTO, CLM
= 1.4)
Feasibleoperating
region
80% wind vector addedto Vs boundary
Fig. 9.13 Aircraft speed envelope
paths. This will influence the time needed to get onto, and from, the operating station.The aircraft high-speed boundary is directly affected by the aerodynamic (transonic)characteristics of the aircraft (mainly the wing geometry and pressure interferenceeffects). Smooth aircraft cross-section area shaping, supercritical wing profiles andincreased sweep are methods to delay the onset of transonic effects. Civil transportspush the boundary to about M0.85 but this increases drag by about 3 to 5 per cent.Reducing the operating speed to less than M0.8 should avoid this penalty.
Figure 9.13 shows the absolute (stall and Mach1.0) boundaries for the aircrafttogether with the 80 per cent average wind speeds at various heights. The wind speedat altitude is important as it will add or subtract to the ground speed and therefore the
“chap09” — 2003/3/10 — page 288 — #19
288 Aircraft Design Projects
search pattern. Note that at about 21 km, the minimum and maximum boundaries arenearly coincident. To fly above this height would require the aircraft to reduce weight,have an increased wing area or an increase in CLmax, or combinations of these changes.For a given aircraft geometry, a cruise-climb technique, in which height is gained asthe aircraft mass reduces with fuel use, could be considered.
The design point selected on the above considerations and the earlier discussion(section 9.3) is:
Operating altitude (initial) = 18 km (59 000 ft)
Operating speed = 210 m/s (408 kt), representing M0.71 at 18 km
At the above condition the aircraft lift coefficient is 0.604 at the start of patrol(mass = 0.9MTO). This reduces to 0.332 at the end of the patrol (mass = 1.3Mempty).At the end of patrol, the aircraft stall speed will have reduced from an initial value of138 m/s (268 kt) to 102 m/s. This change would allow the aircraft to fly progressivelyeither slower or higher.
The discussion above has concentrated on the cruise performance; it is also neces-sary to check the approach speed to determine if it is acceptable. Assuming ISA-SLconditions with an aircraft mass on approach of 1.15Mempty and a CLmax of 1.4(i.e. no flaps):
V 2s = [1.15 (0.38 · 9200) 9.81]/[0.5 · 1.225 · 50 · 1.4]
Vs = 30.3 m/s (59 kt)
If Vapproach = 1.3Vstall, then
Vapproach = 1.3 × 30.3 = 39.4 m/s (75.6 kt)
This speed should be slow enough to allow for automatic/remote landing control inthe UAV version. For emergency landing at higher weight, consideration may need tobe given to the provision of fast fuel dumping.
9.7.5 Wing planform geometry
Selection of the wing planform is the most significant design decision with regard tothe aircraft performance. This aircraft will spend most of its time on long-durationpatrol missions. It is therefore important to choose the wing geometry to ‘optimise’this part of the operating envelope. In this case, drag reduction forms the main basisfor the selection of wing characteristics. In the search phase, the aircraft induced dragwill form a significant component of drag. Selecting a high aspect ratio for the wingis an effective method of reducing induced drag. On conventional monoplane designs,high values for aspect ratio lead to a substantial increase in wing structural mass. Thispenalty arises due to the outboard movement of the centre of lift (away from the wingroot/bodyside attachment). The bracing structure selected on our design avoids thispenalty as part of the outboard lift is reacted by the brace structure. Higher values ofwing aspect ratio than normally seen on conventional designs are therefore feasible.For a given wing area, high aspect ratio corresponds to a large span, it may there-fore be necessary to impose a limit to ensure that the aircraft is easy to handle onor near the ground. In this respect, an aspect ratio of 25 will be selected. This com-pares to values in the range 7–9 for civil transports and 20–30 for higher performancegliders.
“chap09” — 2003/3/10 — page 289 — #20
High-altitude, long-endurance (HALE) uninhabited aerial surveillance vehicle (UASV) 289
Airload
(a) Traditional monoplane
(cantiliver)
Max.BM
Max. bending moment diagram
Note the wing progressivelyreacts an increasing bendingmoment from tip to root chord.
Airload
Brace load
Max. bending moment diagram
Note the significant reduction inmaximum bending moment reactedby the wing structure.
(b) Braced wing structure
Max. BM
Fig. 9.14 Wing bending-moment diagrams
To delay the onset of transition the wing will be swept forward by 30◦. A thin wing(8 per cent) thickness will be adopted. These characteristics should provide a criti-cal Mach number above M0.84 and therefore avoid transonic wave drag penalty andstructural buffeting in the cruise/search phases.
The high aspect ratio wing will produce a small chord and correspondingly a lowvalue for the airflow Reynolds number. This will encourage the retention of laminarflow over the wing section. A transition at 70 per cent chord may be possible if thewing profile skins are smooth, continuous (no gaps or junctions) and the surfaces arekept clean. This should be possible with a composite construction and normal militaryservice care.
The reaction force on the wing from the brace will alter the wing bending momentdistribution. This will cause an unusual distribution of wing taper. In a traditionalunbraced design the maximum bending moment occurs at the wing to fuselage attach-ment section (see Figure 9.14). This is the position where the deepest wing thicknessis required and therefore the widest chord. A straight tapered wing planform withthe largest chord at the root is the usual configuration. For the braced wing layoutin which the relative stiffness of the wing structure and the brace can be selected, thelargest wing bending moment may be at the brace attachment section. To reflect thischange of bending moment distribution the wing taper will be unconventional. Thelargest chord will be at the brace attachment position, as shown in the initial aircraftlayout drawing.
The aircraft wing geometry is now defined:
Wing area 50 sq. m 537 sq. ftWing aspect ratio 25Wing span (=(S · A)0.5) 35.4 m 116 ftWing sweep 30◦ forwardPhysical span (=35.4 cos 30) 30.6 m 100 ftWing thickness ratio 0.08Wing max. CL 1.4Wing taper to match brace geometryWing section supercritical cambered
“chap09” — 2003/3/10 — page 290 — #21
290 Aircraft Design Projects
9.7.6 Engine sizing
As already described, the engine type will be a medium BPR turbofan. The requiredthrust will be dependent on the climb performance at the cruise altitude. If the climb rateat high altitude is too small then the time required to reach the search height will be toolong. Without a specified value for climb rate, we will assume the civil transport criterionof 300 ft/min (1.53 m/s). This assumption will need to be checked when more detailedmass, aerodynamic, propulsion and performance analysis is possible. A sensitivityanalysis will be required to show the inter-relationship of this assumption with thecommonly specified 100 ft/min service ceiling criterion. The fundamental analysis isshown below:
From basic flight mechanics:
(dh/dt) = (T − D) · V/W (9.1)
To obtain the required sea level, take-off thrust/weight ratio, equation (9.1) is modifiedto:
(T/W )SL = (β/α)[(D/Wo) + (1/V )(dh/dt)]A (9.2)
where subscripts SL is sea level, o is initial (take-off) value and A refers to altitudevalues. (β) represents the mass fraction at the start of cruise (which, in this case we willassume to be 0.85). (α) represents the engine thrust reduction with altitude (TA/TSL).
From Eshelby’s book8: (TA/TSL) = σ 0.7 in the troposphere (up to 11.02 km) and σ 1.0
in the stratosphere (σ is relative density = (ρaltitude/ρSL)).This assumes a constant engine rating. For this type of aircraft the loss of thrust at
this high altitude will be large therefore it is likely that the static sea thrust (TSL) atthe climb (or cruise) rating will be suitable to meet the take-off requirement. For thisreason, a constant (climb or cruise) rating will be assumed. As our cruise will be in thestratosphere:
(TA/TSL) = (ρ11.02/ρSL)0.7 · (ρA/ρ11.02) = 0.24
The values for ρ (air density) can be found in ISA tables(in SI units, ρSL = 1.225, ρ11.02 = 0.364 kg/cu. m)(in Imp. units, = 0.002378, 0.000707 slug/cu. ft)Aircraft drag at the cruise condition = 0.5ρAV 2SCDwhere V = 210 m/s (408 kt), S = 50 sq. m (537 sq. ft),and CD assumed to be 0.022.
Aircraft take-off weight Wo = 9200 × 9.81 = 90.25 kN = 20 280 lb
Equation (9.2) is computed and plotted in Figure 9.15. At our selected design altitudeof 18 km (see Figure 9.11) the required thrust to weight ratio (known as the thrustloading) is 0.24 for a 300 ft/min climb ability. The corresponding line at 100 ft/minindicates a service ceiling, with this thrust ratio, of 24 km (78 700 ft). From our previousdiscussions, this value appears to be satisfactory.
This thrust loading gives a required take-off thrust of:
To = 0.24 · 9200 · 9.81 = 21.6 kN = 4870 lb
With two engines this equates to 10.8 kN (2434 lb) per engine.It is necessary to check that this thrust is adequate for safe single-engine take-off in an
emergency. The civil aircraft airworthiness requirement sets a climb gradient of 0.024at 50 ft height and speed V2 (undercarriage retracted but take-off flaps still deployed).
“chap09” — 2003/3/10 — page 291 — #22
High-altitude, long-endurance (HALE) uninhabited aerial surveillance vehicle (UASV) 291
0.2015 16 17 18
Altitude (km)
19 20 21 22
0.22
0.24
0.26
RoC = 100 ft/min
RoC=300 ft/min
Thru
st lo
adin
g (T
/Wo)
Fig. 9.15 Thrust/weight requirement for climb
Aircraft drag at take-off with one engine inoperative will be affected by an increasedue to the flow blockage on the failed engine, extra drag from the asymmetric flightattitude (yaw) and extra trim drag from the control surfaces. To account for these effectswe will assume additional drag to increase the aircraft drag coefficient to 0.04.
Assuming an aircraft speed of 1.2Vstall and a CLmax of 1.4 at the take-off mass of9200 kg (20 280 lb) gives:
(Vstall)2 = (9200 · 9.81)/(0.5 · 1.225 · 50 · 1.4)
Vstall = 45.9 m/s (89 kt)1.2Vstall = 55.1 m/s (107 kt)
Aircraft drag (in SI units) = 0.5×1.225×55.12 ×50×0.040 = 3719 N (=836 lb)Climb gradient = (T − D)/W = (10 800 − 3719)/(9200 · 9.81) = 0.078Climb rate = 0.078 · 55.1 = 4.32 m/s (850 ft/min)
This result seems to provide acceptable initial, single-engine climb, performance.The engine will need to provide the power to drive all of the electrical equipment and
sensors on the aircraft. For example, when flying for long periods at high altitude it isnecessary to warm the electronic and sensor equipment to protect it against the coldambient temperature. This additional ‘load’ on the aircraft engine system is likely tobe significantly more than on other types of aircraft. It is therefore necessary to installan engine with more thrust than the 10.8 kN predicted above.
From the literature, the Pratt & Whitney of Canada PW530 engine used on the CessnaCitation business jet looks suitable for our aircraft:
Take-off thrust = 2900 lb (12.9 kN)Fan max. diameter = 27.3 in (0.69 m)Length = 60 in (1.5 m)Basic engine weight (mass) = 632 lb (286 kg)Specific fuel consumption = 0.55By-pass ratio = 3.3
“chap09” — 2003/3/10 — page 292 — #23
292 Aircraft Design Projects
This engine will give about 20 per cent extra thrust than required for aircraftperformance so should be adequate to meet the aircraft service needs.
9.7.7 Initial aircraft layout
The previous sections have set out the geometrical requirements for the aircraft. It isnow possible to produce the first general arrangement drawing (Figure 9.16).
As prescribed, the layout is very unorthodox. Investigating the technical featuresshows that the configuration is logical. The high mounted wing provides good bank-ing stability when the aircraft is on or near the ground. The high aspect ratio, thinsupercritical wing section and swept forward design should reduce drag. The planformtaper matches the spanwise loading distribution. The configuration should have goodpendulous stability, which will help with low-speed manoeuvrability.
The unobstructed front fuselage provides suitable housing for the observation, recon-naissance and communication systems. These systems are undefined in the project briefbut the length and volume provided on the aircraft is consistent with other aircraft ofthis type. The rear fuselage provides the main structural framework for the attach-ment of engines, main landing gear, brace connection and the fin/wing mounting.The internal volume in this area provides the main fuel tank. The enclosed volume ofthe tank is 3 m long × 1.5 m deep × 0.7 m wide, giving a capacity of 3.15 m3. More
Cg
Equip. modulesCg Fuel
Max. bankangle 28°
Optionalcanards
c4
0 5 m
HALE-UASV
Wing span 30 mWing sweep 30°LEWing area 50 m2
Wing AR 25/18U/A length 15 mEmpty mass 3500 kgTO mass 9200 kgEngine 2 × PW530Thrust 2 × 12.9 kN (TO SSL)
35°Tip
angle
Fig. 9.16 Initial aircraft layout drawing
“chap09” — 2003/3/10 — page 293 — #24
High-altitude, long-endurance (HALE) uninhabited aerial surveillance vehicle (UASV) 293
fuel is housed in the central wing boxes. The capacity of the wing tanks is 0.72 m3.This combined capacity of the tanks (fuselage and wing) (3.87 m3) is substantiallysmaller than the fuel volume requirement estimated in section 9.7.2. At this stage inthe design process no modifications will be made as later calculation of aircraft massmay reduce this early estimate. If it is found later that more fuel is required, the wingmounted ‘equipment/brace’ pods could offer another 0.64 m3. However, this wouldreduce equipment/sensor positioning flexibility. All of the fuel tanks are positionedclose to the aircraft centre of gravity (estimated at the wing mean aerodynamic quar-ter chord position). This will ensure that fuel used in the mission does not lead tosignificant increase in trim drag.
The outboard wing control surfaces will act as conventional ailerons. The inboardcontrol surfaces will provide pitch control and aircraft stability. Due to the relativelyshort tail arm on the aircraft, it may be found necessary to add canard surfaces to thefront fuselage to complement the rear controls. Although such an arrangement couldreduce aircraft trim drag; the interference of flow over the wing sections may affect thelaminar flow condition. The net result could be an aerodynamic inefficiency and a lesseffective layout. Wind tunnel tests would need to be done to quantify the overall flowcondition.
9.7.8 Aircraft data summary
The initial baseline aircraft layout may be summarised as shown in Table 9.1.
Table 9.1
SI units Imperial units
Wing Span 30 m 98 ftAspect ratio 18Sweep 30◦Area 50 sq. m 537 sq. ft
Fuselage Length 15 m 49 ftDepth 2 m 6.6 ftWidth 0.7 m 28 in
Mass Empty 3500 kg 7717 lbMax. TO (design) 9200 kg 20 280 lbPayload 800 kg 1760 lbFuel load 4700 kg 19 360 lb
Engine PW530/545TO thrust 12.9 kN 2900 lbBypass ratio 3.3Cruise sfc 0.54Fan diameter 0.7 m 28 in
Performance Cruise/patrol 210 m/s 408 kt@ 18 km @ 59 000 ft
Duration (gross) 26 hrsApproach 40 m/s 76 ktTO climb (OEI) 7.8%
“chap09” — 2003/3/10 — page 294 — #25
294 Aircraft Design Projects
9.8 Initial estimates
With a fully dimensioned general arrangement drawing of the aircraft available it ispossible to undertake a more detailed analysis of the aircraft parameters. This willinclude component mass predictions, aircraft balance, drag and lift estimations in var-ious operational conditions, engine performance estimations and aircraft performanceevaluations. The results from these studies will allow us to verify the feasibility of thecurrent layout, and our earlier assumptions, and to make recommendations to improvethe design.
9.8.1 Component mass estimations
The geometrical and layout details allow us to estimate the mass of each aircraft com-ponent. This will provide an initial aircraft mass statement that we can use to checkon our initial empty mass ratio and maximum mass estimates. The new mass predic-tions will be used in the following performance predictions. It is necessary to estimateeach of the mass components in the aircraft mass statement described in Chapter 2,section 2.6.1. These component mass calculations are set out below.
Wing structure
Available wing mass estimation formulae are based on conventional cantilever trape-zoidal wing planforms. This presents difficulties in using them to predict our highaspect ratio, braced wing layout. When more details of the wing structural frameworkare known it will be possible to roughly size the main structural elements and therebyto calculate the mass of the structure. This method will give a reasonable estimate ofthe wing mass. Until this is possible, we will need to ‘improvise’!
Using established wing formula for civil jet airliners results in a mass of about10 per cent MTO for our geometry. Such formulae are based on much larger aircraftthan our design. Therefore, the calculation was repeated using general aviation formu-lae. This resulted in a prediction of about 18 per cent MTO. This is also regarded as toohigh and not representative of our aircraft. The high value of the estimate may be dueto the sensitivity of the formulae to the high value for aspect ratio. The difficulties thatarise from the prediction of aircraft mass for unusual/novel designs are not untypicalin advanced project design studies. In the early design stages, all that can be done toovercome these difficulties is to make relatively crude assumptions and to rememberto check these as soon as more structural details are available.
Without better guidance, we will average between the two results that have beenproduced. As the bracing structure will reduce the wing internal loading and as weexpect to use high strength composite construction, we will reduce the estimate by30 per cent as shown below:
Civil aircraft prediction 879 kg (1938 lb)GA aircraft prediction 1720 kg (3597 lb)
Average value 1299 kgLess 30% 390 kg
Predicted wing structure 909 kg (2004 lb)
Add to this an allowance for surface controls and winglets (10 per cent) = 91 kgAdd 20 kg for each mid-span pod structure = 40 kg
“chap09” — 2003/3/10 — page 295 — #26
High-altitude, long-endurance (HALE) uninhabited aerial surveillance vehicle (UASV) 295
The wing brace structure mass can be estimated by assuming a tube (100 mmdiameter×1 mm thick) and measuring the brace length from the layout drawing (8 m).Note: with these sizes for the brace it may be impossible to avoid the strut buckling fromloads in a heavy landing. An aluminium alloy material with a density of 2767 kg/m3
gives:
Brace mass (each) = (π · 100 · 1 · 8) 2767/(1000 · 1000) = 7 kgAdd 10 kg (22 lb) for fairing and support structure and add a contingency of 25per cent:
Total brace mass (both) = 2 · (7 + 10) · 1.25 = 42 kg
Hence, total wing mass (including surface controls, pods and brace):
Structure. 909 / 2004Controls, etc. 91 / 201Pods 40 / 88Brace 42 / 92
1082 kg / 2385 lb (11.8 MTO)
At 11.8 per cent MTO this is slightly higher than modern conventional wing structuresbut the high aspect ratio and large wing area probably are correctly represented.
Tail surfaces
The mass of the vertical tail is estimated using a typical civil aircraft mass ratio of28 kg/m2 (of exposed area). The fin and rudder areas on our aircraft are larger thannormal due to the short tail arm and long forward fuselage. Scaling from the aircraftlayout drawing gives an area of 6 m2. Using the same mass ratio as conventional designspredicts the mass at 168 kg (370 lb).
This represents a mass of over 2 per cent MTO. This is larger than normal but reflectsthe large area. As the wing is mounted on top of the fin structure, a penalty of 10 percent will be added. The vertical tail mass is therefore estimated as 185 kg (408 lb).
The tailplane/elevator structure (i.e. horizontal tail surfaces) on our aircraft is inte-grated into the wing. To allow for an increase in structural complexity and for theoptional canard control a mass of 1 per cent MTO (=92 kg) will be added to the tailstructure mass:
Tail mass = 185 + 92 = 277 kg (611 lb)
This represents 3 per cent MTO, which is typical of many aircraft
Body structure
The mass of the body is estimated using civil aircraft formulae reduced by 8 per centto account for the lack of windows, doors and floor. For the body size shown on thedrawing, the civil estimate is 808 kg. Therefore, our estimate is 743 kg (1638 lb). Thisrepresents 8 per cent MTO which seems reasonable.
The body structure on our aircraft is complicated by a number of special features.These must be taken into account in the estimation:
• add 4 per cent for fuselage mounted engines,• add 8 per cent for the fuselage brace/undercarriage attachment structure,• add 10 per cent to allow for the modular fuselage equipment provision.
Hence, body mass = 1.04 × 1.08 × 1.10 × 743 = 883 kg (1947 lb).
“chap09” — 2003/3/10 — page 296 — #27
296 Aircraft Design Projects
This is 9.6 per cent MTO which is higher than normal but accounts for the complexnature of the fuselage structural framework.
Nacelle mass
Engine nacelle mass is estimated using civil aircraft formulae related to the predictedthrust of 21.6 kN (4856 lb) (i.e. 2 × 12.9 = 25.8 kN (5800 lb)). This is acceptable as theinstallation is comparable to rear mounted engines on civil business jets. The nacellemass prediction is 147 kg (324 lb) (i.e. 1.65 MTO).
Landing gear
The undercarriage on the aircraft is expected to be straightforward and relatively simpletherefore a value of 4.45 per cent MTO, which is typical of light aircraft, is proposed:
Landing gear mass = 0.0445 × 9200 = 409 kg (902 lb)
For aircraft balance, it will be assumed that 15 per cent of this mass is attributed to thenose unit (61 kg/135 lb), leaving 348 kg/767 lb at the main unit position.
Flying controls
This item has been included in the wing structural mass estimation.
Propulsion group mass
For large turbofan engines with BPR of 5.0 the basic (dry) mass ratio is predicted frompublished engine data to be 14.4 kg/kN. This would give a mass of (14.4 × 21.6 =311 kg/686 lb). Smaller engines with lower BPR would not achieve this value due tothe effects of descaling. Data from the suggested engine gives a dry weight for eachengine of 632 lb (287 kg). With two engines this gives a total dry-engine mass of 573 kg(1263 lb). There is a substantial difference between these estimations but as the largestone is from an existing engine this will be used. The engine services and systems willincrease the dry mass. Typical civil aircraft incur an additional 43 per cent:
Propulsion group mass = 1.43 × 573 = 820 kg (1808 lb)
Fixed equipment mass
For conventional aircraft, this mass group would fall within the range 8 to 14 percent MTO. Our aircraft is not typical as the equipment forms a major subsystem.Observation, monitoring, communication and intelligence gathering equipment will beused on the aircraft on different missions. Versatility of equipment installations will bean essential feature on the aircraft. As discussed earlier, this operational flexibility hasbeen addressed by allowing 800 kg (1764 lb) of equipment mass to be assumed as ‘usefulload’. However, to support the operational equipment modules the aircraft will need tohave some fixed equipment services (e.g. power supplies). It will also require systems toallow the aircraft to function (e.g. hydraulic, electrical, fuel supply, etc.). Some of thesystems found on conventional aircraft will not be necessary due to the absence of thecockpit and pilot (e.g. instruments, controls, environmental controls and protection,safety, furnishings). Until more details are available on the systems to be installed wewill assume that the fixed equipment accounts for 8 per cent MTO (=736 kg/1623 lb).
“chap09” — 2003/3/10 — page 297 — #28
High-altitude, long-endurance (HALE) uninhabited aerial surveillance vehicle (UASV) 297
Equipment requirements above this figure will be transferred to the previously described(800 kg/1764 lb) ‘useful load’ component.
Fuel mass
Until a more detailed aerodynamic and performance analysis is done, the previouslyestimated fuel load of 4693 kg (10 348 lb) will be assumed. As this presents a substantialcomponent to the overall aircraft mass (51 per cent MTO) it is important to carefullyestimate the fuel requirements as soon as possible.
9.8.2 Aircraft mass statement and balance
From the sections above it is now possible to compile the detailed aircraft massstatement (see Table 9.2).
The empty mass fraction at 44 per cent is higher than assumed (38 per cent) in theinitial sizing. This has increased the aircraft MTO to a value above the 9200 kg (21 168 lb)design mass. A further iteration should have been done to estimate more accuratelythe component masses and ultimately the MTO. However, as several of the componentmasses and the fuel mass are based on crude assumptions it is not appropriate to gointo such detail at this stage.
The mass statement can be used to determine the position of the aircraft centreof gravity (as described in Chapter 2, section 2.6.2). This will confirm, or otherwise,the assumed longitudinal position of the wing relative to the fuselage as shown onthe aircraft layout drawing. The component masses are located around the aircraftstructure as shown in Figure 9.17.
These are used to predict the position of the aircraft centre of gravity for differentloading conditions. With a datum set at one metre ahead of the aircraft nose theresults are:
• at MTO: xcg = 10.05 m 33.0 ft (51 per cent MAC)• at MTO less body fuel: xcg = 9.61 m 31.5 ft (40 per cent MAC)• at empty mass: xcg = 10.04 m 32.9 ft (51 per cent MAC)• at MTO − useful load: xcg = 10.3 m 33.8 ft (58 per cent MAC)
Table 9.2
kg lb % MTO
Wing structure 1082 2 386 11.0Tail structure 277 611 2.8Body structure 883 1 947 9.0Engine nacelles 147 324 1.5Landing gear 409 902 4.1Total structure 2798 6 170 28.4Propulsion group 820 1 808 8.3Fixed equipment 736 1 622 7.5Aircraft empty 4354 9 600 44.2Useful load 800 1764 8.1Fuel load 4695 10 352 47.7Aircraft MTO 9849 21 716 100.0
“chap09” — 2003/3/10 — page 298 — #29
298 Aircraft Design Projects
+ +
Wing structure 1082and Wing fuel 939
Noseundercarriage61
Bodystructure883
Mainundercarriage348
Predictedaircraftcentre of gravity
Predicted aircraftcentre of gravity
Datum(side)
Datum (plan)
Forward fixedequipmt 368and Canard 17
Usefulload800
Fuel inbody3754
Tailstructure260
Engines 445Nacelles 147Fixedequipment 368
Fig. 9.17 Aircraft balance
The values quoted in parentheses above are the cg positions as percentages of the meanaerodynamic chord (aft of the leading edge). This analysis shows that the wing meanchord position should be moved rearward with respect to the datum. On our design,this is most easily achieved by reducing the sweep angle. Due to the lack of confidencein the component mass estimation at this stage, no changes will be made (yet). It isreassuring to note that even in the present unbalanced configuration the cg range isacceptable and that ballasting to reduce the range does not seem to be necessary.
Although a number of small changes to the aircraft initial layout have been suggestedin the mass and balance analysis, it has confirmed the feasibility of the design andprovided data for subsequent calculations.
9.8.3 Aircraft drag estimations
The initial drag evaluation will be done using the conventional component dragbreakdown and applying the equation below:
CD = CDo + CDi + �CDw
Aircraft cruise speed is set at subtransonic flow conditions. This makes the wave dragcomponent zero. The parasitic drag coefficients (CDO) are evaluated, for each aircraftcomponent, by estimating the terms in the following equation:
CDo =∑
(CDo)i =∑
{(Cf )i · Fi · Qi(Swet/Sref )}
“chap09” — 2003/3/10 — page 299 — #30
High-altitude, long-endurance (HALE) uninhabited aerial surveillance vehicle (UASV) 299
where Cf = skin friction coefficientFi = form (shape) factorQi = interference factor
Swet = component wetted areaSref = wing reference areaSref = 50 sq. m (537 sq. ft) for our aircraft
Formulae used for the above estimation can be found in most aerodynamic or aircraftdesign textbooks (e.g. reference 7). Geometrical inputs are scaled from the layoutdrawing. The results (with a reference area of 50 m2/537 sq. ft) are shown in Table 9.3.
9.8.4 Aircraft lift estimations
To reduce complexity and to avoid drag increases in cruise, the aircraft will be manu-factured without conventional flaps. If it is found necessary to increase CL for landingor take-off, the aileron surfaces could be drooped or a simple leading edge device used.These possibilities will not be considered in the initial layout. Assuming a camberedsupercritical wing profile is used, the two-dimensional max. lift coefficient may be 1.65for our high aspect ratio clean wing.
The three-dimensional value is determined below:
(CLmax)3D = 0.9(CLmax)2D · cos
Assuming quarter chord sweep = 22o gives (CLmax)3D = 1.4.This confirms our original assumption.
Table 9.3
Flight cases
Cruise Take-off OEI climb∗ Landing
Airspeed m/s/kt 210/408 37/72 55/107 40/78(0.7V2) (V2)
Altitude km/1000 ft 18/59 SL SL SLMass kg/lb 7820/17 243 9200/20 280 9200/20 280 4976/10 970CDo (×104) fuselage 26.9 26.7 25.2 26.4
wing 58.3 56.1 51.6 55.1braces 16.8 16.4 15.2 16.2tail 7.2 7.6 7.2 7.5nacelles 9.7 9.6 9.0 9.4
CDo total basic 119.4 116.4 108.2 114.6Add undercarriage — 104.0 — 104.0Add trim 2.0 5.0 65.0 5.0Note: no flaps on this aircraftCDo total (incl. contingency) 129.7 237.1 173.1 235.2CL 0.592 0.974 0.976 0.996Induced drag factor 0.022 0.022 0.022 0.022CD total (×104) 206.9 445.8 381.9 453.5Lift/Drag (initial cruise) 28.6CL end of cruise (M = 4976) 0.376Lift/Drag (final cruise) 23.4 (with no height gain)
∗OEI = one engine inoperative at the start of climb, i.e. emergency take-off case.
“chap09” — 2003/3/10 — page 300 — #31
300 Aircraft Design Projects
At the initial cruise speed and height, the design lift coefficient will be 0.59, as shownin Table 9.3.
Much more work would need to be done in designing the best wing section profile.This would entail the application of sophisticated CFD methods that are not practicalin the initial design stages. However, the calculations above seem to be reasonable andwill provide values for use in the performance calculations that follow.
9.8.5 Aircraft propulsion
The previous initial sizing work identified the required engine parameters and a can-didate powerplant. Reference 8 provides formulae to determine engine performanceat specified operating conditions (speed and height). The manufacture’s quoted values(per engine) for the selected engine at static, sea-level, take-off conditions are:
Thrust = 2900 lb (12.9 kN)
Specific fuel consumption = 0.55 lb/lb/hr or N/N/hr
Howe’s formulae9 applied to the cruise condition (M0.7, 18 km) with BPR of 3.3estimates thrust at:
T/To = 1[0.88 − (0.016 · 3.3) − (0.3 · 0.7)]0.9850.7 = 0.166Hence, T = 12.9 × 0.166 = 2.15 kN (48 lb)
Using the same formula, the thrust per engine at the end of take-off (and for OEI) iscalculated as T = 11.37 kN (2556 lb).
And specific fuel consumption (C):
C/Co = [1 − (0.15 · 3.3)0.65][1 + 0.28(1 + 0.063 · 3.32)0.7]σ 0.08
Giving, C = 0.493
9.8.6 Aircraft performance estimations
Initial estimates of aircraft performance are based on methods described in mostaircraft design textbooks (e.g. references 7 to 10). Point estimates are required to deter-mine the suitability of the aircraft layout to the operational requirements. Three flightphases are investigated:
• field performance,• climb performance,• cruise.
Field performance
Although it could be possible to assess the take-off and landing performance using stepintegration of the aircraft path, it is sufficient in these early stages to use generalisedformulae.10 Four calculations will be made:
(a) stall and operating speeds,(b) take-off distance,(c) second segment climb,(d) landing distance.
(a) As the aircraft wing has been simplified by the avoidance of flaps, the CLmax is thesame for the take-off and landing configurations. The take-off will be calculated at
“chap09” — 2003/3/10 — page 301 — #32
High-altitude, long-endurance (HALE) uninhabited aerial surveillance vehicle (UASV) 301
the maximum mass (9200 kg) and the landing at a reduced mass (10 per cent fuel plusfull payload) of 4976 kg. An emergency landing calculation will also be done for theaircraft at MTO less 10 per cent fuel. The operating speeds are determined below:
Vstall = (W/S) (2/(ρ · CLmax))0.5
where S = 50 sq. m (537 sq. ft)ρ = 1.225 kg/m3
(0.002378 slug/cu. ft)CLmax = 1.4
Giving: at take-off, Vstall = 45.9 m/s (89.1 kt)take-off speed V2 = 1.2Vstall = 55.1 m/s (107 kt)at landing, Vstall = 33.7 m/s (65.4 kt)landing approach speed VA = 1.3Vstall = 43.9 m/s (85.2 kt)at emergency landing, Vstall = 44.7 m/s (86.8 kt)emergency approach speed VA = 1.3Vstall = 58.1 m/s (112.8 kt)
The normal take-off and approach speeds seem reasonable. As commented on previ-ously, the high speed that is required for the emergency landing case could be reducedif fuel dumping was included in the fuel system.
(b) Take-off distance can be calculated by the formula10 below (note: the formula in thisbook is derived in ft-lb units, therefore some conversion will be needed to transform toSI units (see Appendix A)):
STO = 20.9[(W/S)/(σ · CLmax · (T/W )] + 87[(W/S) (1/(σ · CLmax)]0.5
The two terms in square brackets are for the ground roll (with a rolling frictioncoefficient of 0.03) and the climb to 50 ft obstacle clearance respectively:
(W/S) = 20286/537.5 = 37.74 lb/sq. ft
(T/W ) = 4856/20286 = 0.239
σ = 1, CLmax = 1.4
Hence, STO = 2357 + 452 = 2809 ft (857 m)
(c) The second segment climb calculation is a check on the ability of the aircraft toclimb away from the ground after an engine failure on take-off. The aircraft ‘rate ofclimb’ (RoC) is calculated by:
RoC = (V/W ) (FN − D)
where V = 1.2VstallFN = emergency thrust from the remaining engineD = aircraft drag with the landing gear retracted but with an asymmetric flight
attitude to counteract the adverse yaw from the engine thrust/drag
In our case: V = 55.1 m/s (107 kt)FN = 11.37 kN (2556 lb)D = (0.5ρV 2) SCD = 1853 × 50 × 0.03819 = 3538 N (795 lb)
Hence, RoC = [55.1/(9200 · 9.81)] (11370 − 3538) = 4.78 m/s (940 fpm)
“chap09” — 2003/3/10 — page 302 — #33
302 Aircraft Design Projects
We can also calculate the aircraft climb gradient = sin−1(RoC/V ) = 0.08 (i.e. 8
per cent which is satisfactory) (note: the minimum value for civil transport aircraft is2.4 per cent).
(d) The landing distance is calculated using standard formula10 (in ft-lb units) is:
SL = 118(W/S)/(σ · CLmax) + 400
For normal landing: W/S = 10972/537.5 = 20.4 lb/sq. ftFor emergency landing: W/S = 19252/537.5 = 35.8 lb/sq. ftWith: σ = 1 and CLmax = 1.4:Hence: SL (normal) = 2119 ft (646 m)
SL (emergency) = 3417 ft (1042 m)
It may be desirable to add airbrakes/lift dumpers to the aircraft to increase drag onlanding and thereby reduce the landing distances shown above.
Climb performance
The climbing ability of aircraft is a function of the aircraft airspeed, weight and enginesetting (for our aircraft the engine power/setting is constant). The best rate of climb for agiven weight and thrust will be at the speed for minimum drag. For most operations thisspeed is too slow. The aircraft is normally flown at a higher speed and often acceleratedduring climb. This acceleration sacrifices some climbing ability but has the advantageof gaining ground distance and, at the top of climb, matching the climb to cruise speed.
When a full performance estimation is produced it will show the time to climb tospecific heights and the associated ground distance covered. With our current degreeof knowledge and confidence with the aircraft parameters, such detailed analysis isnot appropriate. To predict the climb performance a point analysis will be all thatis necessary. This will use the mass, drag and engine thrust data, as described in theprevious sections, and an assumed flight speed:
RoC = (V/M · 9.81) (FN − D)
The climbing speed profile is assumed to be 250 kt EAS (129 m/s) up to about 30 000 ft(9 km) then Mach 0.7.
Aircraft mass is set at 0.95MTO = 8740 kg (19 270 lb) throughout the climb (a sim-plifying assumption as fuel is continuously burnt). The results of this calculation areshown below and the rate of climb variation plotted in Figure 9.18.
Altitude km 1.4 4.6 9.0 14.0 18.0ft 4600 15 080 29 500 45 900 59 000
True airspeed m/s 138 163 209 206 206kt 268 316 406 400 400
Drag N 5362 5486 5552 3457 2930lb 1205 1233 1248 777 659
Thrust per engine N 8147 6289 4075 2440 1572lb 1832 1414 916 549 353
RoC (both engs) m/s 17.6 13.5 6.3 3.4 0.51fpm 3458 2645 1244 674 101
“chap09” — 2003/3/10 — page 303 — #34
High-altitude, long-endurance (HALE) uninhabited aerial surveillance vehicle (UASV) 303
20
16
12
8
Rat
e of
clim
b (m
/s)
4
020181510
Altitude (km)50
10
20
30Ti
me
to c
limb
(min
)
40
50
60
70
Rate ofclimb
Time toclimb
300 fpm
Fig. 9.18 Rate-of-climb and time-to-climb
Time to climb can be roughly calculated using the average values from the RoC graph:
Stage 0–5 km 0–10 km 0–15 km 0–18 km(Stage to 16 400 ft 32 800 ft 49 200 ft 59 000 ft)Time min. 5.2 14.5 35.3 68.6
(note that it will take over one hour to climb up to the 18 km operation altitude).From these calculations it is clear that the cruise altitude of 18 km represents the air-
craft service ceiling (normally defined as 100 fpm), at the aircraft conditions assumed.In addition, the calculations show that the time to climb the final 3 km almost doublesthe time to reach cruise height. This does not seem to be a sensible operational practice.Either the aircraft weight or drag must be reduced, or the available thrust increased. Forexample, similar calculations show that when the aircraft mass is reduced to 7500 kg aclimb rate of 300 fpm will be possible at 18 km. Alternatively, a different operationalpractice may be used (e.g. start the mission at a lower altitude and increase this as theaircraft weight is reduced through fuel burn).
Cruise
Several operational strategies can be adopted for the cruise phase. The one to be used inour analysis is to fly the aircraft at a constant angle of attack (constant Mach number).This implies that as the fuel is used and the aircraft becomes lighter, the aircraft gainsheight. This is known as a cruise-climb profile.
The usual Breguet range equation can be written as:
R = (V/c)(L/D) loge(M1/M2)
This gives a maximum value when aircraft speed is 1.316 times the speed for minimumdrag. As mentioned above, this speed may be too slow for the aircraft at high altitudewhere the allowable speed range is narrow. We will cruise at a speed of M0.7. Usingthe above equation, it is seen to be operationally desirable to start the cruise at a loweraltitude than the originally specified 18 km. This confirms the climb result. The aircraftlift to drag ratios for cruise at 15 and 18 km is calculated to be 26.5 at 15 km and 28.6at 18 km cruise height.
“chap09” — 2003/3/10 — page 304 — #35
304 Aircraft Design Projects
In the cruise phase the true airspeed at M0.7 is 206.5 m/s (400 kt), and cruise sfc =0.493 per hour. We will calculate the aircraft range taking no account of the fuel usedin the climb and descent phases. To correct this assumption we will add 10 per cent tothe calculated fuel mass. Hence:
Starting mass = M1 = MTO = 9200 kg (20 280 lb)End mass = M2 = M(operational empty) + Mpayload = 4354 + 800
= 5154 kg (11 365 lb)
Hence range, R = (206.5/[0.493/3600]) × 26.5 loge(9200/5154)= 23 154 km (12 500 nm)
It is also possible to rearrange the Breguet equation to calculate endurance (E):
E = (1/c)(L/D) loge(M1/M2) = 31.14 hours
This flight time is more than specified; therefore we can determine the mass fractionfor the required 24 hour endurance:
(M1/M2) = ex, where x = (E . c)/(L/D) = (24 · 0.493/26.5)
(M1/M2) = 1.563(M1/M2) = MTO/M2, which assuming MTO = M2 + Mfuel gives:Mfuel = 0.563, M2 = 2902 kg (6400 lb), adding 10 per cent to this value gives:Mfuel = 3192 kg (7038 lb)
This is substantially less than estimated very crudely in section 9.7.1 (4693 kg). Thisreduction will make the fuel tankage easier to provide as only 3.99 m3 will be necessary.We can now estimate the new value of MTO and the mass fractions:
MTO = 4354 + 800 + 3192 = 8346 kg (18 400 lb)
Mfuel/MTO = 38.5 per cent, M(operational empty)/MTO = 52.2 per cent
This empty mass fraction is much higher than previously assumed but lies between theStratos 1 and Stratos 2 values.
We can now determine the engine cruise rating (as a percentage of the max. available)at the operating condition, as shown below.
The aircraft drag at the start and end of the cruise is estimated to be 3236 N and2086 N (727 lb to 470 lb). The engine thrust at 15 and 18 km is estimated at 4352 N and3126 N (978 to 703 lb) (both engines operating). Hence the engine rating (aircraft dragdivided by available thrust) will be:
75 per cent at 15 km (49 200 ft) at start of cruise67 per cent at 18 km (59 000 ft) at end of cruise
These seem to be reasonable cruise ratings from maximum and will extend the life ofthe engine (between overhauls) due to the associated lower operating temperature.
If, however, the aircraft is flown at constant altitude these become:
62 per cent at end of 15 km (49 200 ft) cruise85 per cent at start of 18 km (59 000 ft) cruise
“chap09” — 2003/3/10 — page 305 — #36
High-altitude, long-endurance (HALE) uninhabited aerial surveillance vehicle (UASV) 305
7000
8000
9000
10 000
Max
. tak
e-of
f m
ass
(kg)
18 hr
Payl
oad
800
kg
650
kg
500
kg
30 hr
24 hrA
Fig. 9.19 Parameter trade-off study
9.9 Trade-off studies
At this stage in the development of the project, we have sufficient detail to conductsome trade-off studies. Using the engine and aerodynamic equations, it is possible toshow the relationship between the two main operational parameters (namely payloadand endurance) on the aircraft maximum mass. A classical nine-point carpet plot willbe constructed. The endurance will be varied between 18 and 30 hours and the payloadreduced from 800 to 500 kg (1764 to 1102 lb). The results are shown in Figure 9.19.
The study shows that the aircraft is relatively insensitive to changes in payload butthat endurance is a very influential parameter. A horizontal line drawn across the plotprovides an indication of the trade-off between payload and endurance. For example,moving from the design point (A) to a lower payload (500 kg/1102 lb) and substitutingextra fuel to replace the lost payload (providing that sufficient tankage is available)allows two extra hours of flight.
It is possible to conduct similar studies to illustrate the effects of varying the followingparameters:
• wing area versus aspect ratio,• cruise altitude versus aircraft maximum mass (MTO),• system mass versus MTO,• introduction of advanced technologies (e.g. laminar flow, composite materials, etc.),• variation in mission requirements.
Such studies provide a detailed appreciation of the factors and parameters affecting theaircraft design space. The results of such studies are used to revise the aircraft layoutand specification to produce a solution better matched to the design brief.
9.10 Revised baseline layout
The main changes to the initial aircraft layout, from the work done so far, are associatedwith the provision for adequate lateral (weathercock) stability and control. The long-span, forward-swept wing with winglets, and the long forward fuselage with deep sidearea, will generate destabilising moments in cross-wind conditions. Balancing these
“chap09” — 2003/3/10 — page 306 — #37
306 Aircraft Design Projects
moments is difficult due to the relatively short tail arm. Two modifications are proposedto ease this problem. The forward fuselage length is to be reduced by 2 metres and‘finlets’ are to be placed on the wing outboard of the inner wing trailing edge controlsurfaces. These finlets could be made large enough to double the original fin area ifrequired. Some of the loss of equipment volume resulting from the reduction of thefuselage length could be regained by moving the fuselage fuel tank further back andincreasing the amount of fuel held in the wing. These two proposals together shouldprovide sufficient flexibility into the layout to overcome the perceived stability problem.
A second concern relates to the layout of the landing gear. The large wing-span,high aircraft centre of gravity and the narrow main-wheel track combine to makethe aircraft potentially unstable in taxi, take-off and landing conditions. The reducedlength of the forward fuselage mentioned above will improve the landing gear geometrybut this will not be sufficient. It will be necessary to increase the track of the mainwheels. This can only be done by adding fuselage sponsons at the main undercarriagemounting positions. Increasing the track to 4 m will provide an overturning angle ofabout 52◦ (convention suggests that an angle greater than 60◦ is unsafe or twitchy inoperation). The sponsons will need to be extended fore and aft to provide aerodynamicblending. These extensions will provide extra storage. This new arrangement will alsoimprove the attachment geometry of the braces at the side of the fuselage.
Following the calculations of the component masses and the associated aircraft centreof gravity assessment the wing leading edge sweep will be reduced from 30 to 25◦.
The above changes have been included into a revised aircraft general arrangementdrawing, see Figure 9.20.
Scale5 m
10 ft
Overturningangle55°
Cg
Fig. 9.20 Revised aircraft general arrangement
“chap09” — 2003/3/10 — page 307 — #38
High-altitude, long-endurance (HALE) uninhabited aerial surveillance vehicle (UASV) 307
9.11 Aircraft specification
9.11.1 Aircraft description
Aircraft type: Manned or uninhabited high-altitude, long-endurance,reconnaissance vehicle.
Design features: The novel aircraft layout, with a high aspect ratio,multi-tapered, swept-forward braced wing planform,provides a platform for the mounting of alternativepayloads and systems. The wing profile employs asupercritical section. The clean wing is unflapped withoutboard ailerons and inboard elevators. The fuselage isconfigured to allow equipment modules to be quicklychanged, giving unique flexibility in operation. One of theforward modules offers the alternative of either a mannedcockpit capsule or an autonomous unmanned flight controlsystem. The twin turbofan engines are developments from asimilar type currently used on business jets (e.g. P&W ofCanada PW-530/545). The tricycle retractable landing gearis of conventional design.
Operational features: The mission profile includes 24 hour flights at 45 000 to65 000 ft altitude at Mach 0.7. Operation into and fromconventional military airfields. Optional detachment ofwings and brace structure for rapid deployment tooperational theatre.
Structure: Conventional glider-technology composite structuralframework with rapid access to interchangeable fuselageequipment modules. Fuel held in integral wing tanks andcentral fuselage bladder tanks.
Equipment: Space provision for reconnaissance and communicationpackages to suit variable operational missions.
9.11.2 Aircraft data
SI units Imperial unitsDimensions: Overall length 12.7 m 41.6 ft
Overall span (incl. winglets) 30.0 m 98.4 ftOverall height (incl. winglets) 5.6 m 18.4 ftWing aspect ratio 18Wing taper ratio 0.5 outer, 0.75 innerWing LE sweep 25◦ forwardWheelbase 6.6 m 22.6 ftTrack 4.0 m 13.1 ft
Areas: Wing (ref.) 50 sq. m 540 sq. ftElevator 4.0 sq. m (total) 43 sq. ftAileron 0.7 sq. m (each side) 7.5 sq. ftFin 2.7 sq. m (upper) 29 sq. ft
0.5 sq. m (under) 5.4 sq. ftMass/Weight Max. take-off 8665 kg 19 100 lb
Empty 4354 kg 9600 lbLanding 5505 kg 12 140 lb
“chap09” — 2003/3/10 — page 308 — #39
308 Aircraft Design Projects
SI units Imperial unitsFuel 3511 kg 7742 lbFuel (US gal) 1200Equipment/useful load 800 kg 1764 lb
Loadings: Max. wing loading 1637 N/sq. m 331 lb/sq. ftThrust (SSL)/weight 0.315
Engines (each): SSL take-off thrust 12.9 kN 2900 lbSFC take-off 0.54/hrSFC cruise 0.49/hrBypass ratio 3.3Dry mass/weight 287 kg 633 lb
Aerodynamic: CDo (cruise) 50 000 ft @ M0.7 0.01297CDo (landing) SL @ VA 0.02352CL (cruise) 50 000 ft @ M0.7 0.60 (initial), 0.38 (final)CLmax 1.4CL (landing) 1.0L/D (cruise) 28.6
Performance: MissionCruise speed (M0.7 @ 50 000 ft) 207 m/s, 400 ktService ceiling 45 000 ft (initial), 65 000 ft (final)Endurance 26 hoursFieldStall speed (VS) 45.9 m/s 89.1 ktTake-off speed (V2) 51.1 m/s 107 ktApproach speed (VA) 43.9 m/s 85.2 ktTake-off distance 857 m 2809 ftSecond segment climb OEI 0.08Landing distance 646 m 2119 ftClimbRate of climb (initial) 3458 fpm @ 268 ktTime to climb, to 10 km 14.5 min(from TO) to 15 km 35.3 min
to 18 km 68.6 min
9.12 Study review
There are many further design considerations to be studied in the development ofthis project. The aircraft is a complex combination of advanced technologies in aero-dynamics, structures, materials, stability and system integration. This represents asubstantial challenge which reflects the nature of future aircraft project work. As aero-nautical design matures it will become harder to make significant improvements tocurrent designs. This will force aeronautical engineers to introduce innovation into newdesigns. The ability to handle the necessary analysis methods to reduce technical riskwill form a major feature of future design teams. These teams will include many morespecialists from disciplines that have not been traditionally included in aircraft projectdesign. Organising, managing and controlling these teams will demand skills other thanthose conventionally related to aeronautical engineering. A more ‘system-orientated’approach will become the new practice.
“chap09” — 2003/3/10 — page 309 — #40
High-altitude, long-endurance (HALE) uninhabited aerial surveillance vehicle (UASV) 309
As well as dealing with the integration of new technologies and methods, this projecthas involved the analysis of aircraft operating in the higher atmosphere. In this envi-ronment, the stall and buffet flight boundaries begin to converge to make control moredifficult. High-speed, high-alpha must be carefully considered to ensure that the aircraftis dynamically stable yet, as in this case, aerodynamically efficient. This combinationoffers a serious test to the aerodynamic and structural disciplines.
This project has demonstrated the unique features of designing an aircraft toaccount for:
(a) uninhabited/autonomous missions,(b) fast and high operation,(c) system and airframe integration,(d) the introduction of new technologies and methods.
Not many new projects incorporate such a mixture of challenges to the design team.However, if the difficulties of meeting such demands can be successfully achievedwithout jeopardising aircraft operational integrity, then we will be in the enviableposition of ‘pushing the envelope’.11 Good luck!
References
1 Kampf, K. P., ‘Design of an unmanned reconnaissance system’, ICAS 2000, Harrogate UK,August 2000.
2 AIAA Aerospace Design Engineers Guide, AIAA Publications, ISBN 0-939403-21-5, 1987.3 Brassey’s World Aircraft & Systems Directory, Brassey Publications, ISBN 1-57488-063-2.4 Jane’s All the World’s Aircraft, Jane’s Annual Publication, various years. See www.janes.com
for list of publications.5 Lange, R. H., ‘Review of unconventional aircraft design concepts’, Journal of Aircraft 25, 5:
385–392.6 Ko, A. et al. ‘Effects of constraints in multi-disciplinary design of a commercial transport
with strut-braced wings’. AIAA/SAE World Aviation Congress 2000/1, paper 5609. See alsoGundlach, J. T. et al. ‘Concept design studies of a strut-braced wing, transonic transport’.AIAA Journal of Aircraft, Vol. 137, No. 6, Nov-Dec 2000, pp. 976–983.
7 Jenkinson, L. R. et al., Civil Jet Aircraft Design, Butterworth-Heinemann, 2000,ISBN 0-340741-52-X.
8 Eshelby, M. E., Aircraft Performance, Theory and Practice, Butterworth-Heinemann, 2001,ISBN 0-340758-97-X.
9 Howe, D., Aircraft Conceptual Design Synthesis, Professional Engineering Publishing, UK,ISBN 1-86058-301-6.
10 Nicholai, L. M., Fundamentals of Aircraft Design, METS Inc., San Jose, California 95120.11 Rabinowitz, H., Pushing the envelope, Metro Books, 1998 (www.metrobooks.com).
“chap10” — 2003/3/10 — page 310 — #1
10
Project study: a general
aviation amphibian aircraft
“chap10” — 2003/3/10 — page 311 — #2
Project study: a general aviation amphibian aircraft 311
10.1 Introduction
Many early aircraft designs were developed for take-off and landing on water. Withthe absence of readily available level and mowed fields, lakes and even the ocean werelooked upon as ideal choices for a place to land or take-off. This allowed operation ina wide range of headings to accommodate wind direction. It also did not require anypreparation for landing or take-off other than a quick look to make sure that boats ordebris were not in the fight path. This provided a decided advantage over land-basedoperations where real estate had to be purchased or rented, obstacles (tree stumps androcks) cleared, and grass cut to a reasonable height or a hardened earth or macadamsurface prepared. In emergencies, a lake was also more likely to be clear of obstaclesthan a farmer’s field that might be filled with cattle or bisected by a fence. Hence, manyearly airplane designers opted for a seaplane configuration. In the event that landoperation was sought, an amphibian design offered the capability of water or landoperation. In fact, due to the public’s lack of confidence in airplane engine reliability, itwas not until almost the mid-twentieth century that long, overwater passenger flights(transatlantic, transpacific, Caribbean, etc.) were routinely attempted in anything otherthan seaplanes or amphibians.
With extensive use of land-based aircraft to transport military personnel duringWorld War II and with improvement in engine reliability, the flying public gained theconfidence needed for such aircraft to replace their water-based counterparts. Thisallowed inland airports to replace coastal sites as ports of entry and exit for overseasflights and the large amphibians and seaplanes of the 1930s and 1940s were retiredfrom service.
In the general aviation (GA) field, seaplanes and amphibians have always occupied asmall but important niche in the marketplace, used primarily for operations into and outof remote areas where lakes were more plentiful than airports. Today, most such aircrafttend to be ‘floatplanes’, aircraft originally designed for land operation to which havebeen added rather large floats to replace the conventional wheeled undercarriage. Suchaircraft are usually considerably slower in flight and more limited in performance thantheir original designs due to the added weight and drag of the floats. In attempts to getbetter overall performance, a few specialty aircraft have been designed as amphibianswith a hull fuselage. However, the compromises required to allow both land and wateroperations have still resulted in added weight and complexity, and a lower cruise speedthan conventional land-based aircraft designs.
In the following summary of the design process, emphasis will be placed on thefactors unique to amphibian aircraft. Consideration of aspects of the process that arecommon to all aircraft designs will be given more cursory coverage.
10.2 Project brief
The design of a modern, general aviation airplane for operation on both land and waterproved an interesting challenge for a group of aerospace engineering students. Theywanted to enter their design in the National General Aviation Design Competitionsponsored by NASA and the Federal Aviation Administration in the United States inthe late 1990s.
In this case, the ‘customers’ for the aircraft being designed consisted of a group ofjudges in a design competition and the original ‘specifications’ for the design werethe competition guidelines. Some of these guidelines were rather broad. They included