Professional Documents
Culture Documents
Gratton
Initial
Airworthiness
Determining the Acceptability of New
Airborne Systems
Second Edition
Initial Airworthiness
Guy Gratton
Initial Airworthiness
Determining the Acceptability of New
Airborne Systems
Second Edition
123
Guy Gratton
Aerospace Transport and Manufacturing
Cranfield University
Cranfield, Bedfordshire
UK
The first edition of this book was an adventure for both I and Springer, as it was the
first book describing the process of initial airworthiness evaluation, so nobody
really knew the response it would receive. Generally, response has been very
positive, with a great many practitioners and academics finding that it was worth
buying.
However, almost immediately it was published, I kept finding things I wanted to
improve in the book. There were inevitably a few things I felt could have been
executed better but also—particularly as reviewers and users became quite generous
with their feedback—increasingly I found ways I wanted to improve or expand the
book. So after a couple of years, I took a list to Springer, who agreed to my
proposal for a second edition.
A lot of small items have been added in, such as historical incidents, additional
biographies and some expanded detail about particular standards: particularly, the
new form of part 23 as it has come into being in the last few years; changes to
ETOPS; my home country, the United Kingdom, has made a number of changes to
deregulated flying opportunities including deregulated single-seat microlights and a
new set of ‘E-conditions’ for experimental flying, and the FAA followed by EASA
has substantially restructured part 23. All of these justified an update to the book.
I’ve also added three significant new sections that I hope will complement the
existing technical material in the book. One is on the growing an important topic
of the environmental impact of aircraft. The second may seem a little incongruous,
and is on the topic of professional ethics—this has become increasingly important
to all of us as aerospace professionals, and most countries are now including the
topic in engineering degree courses. The views here are mostly personal but built
vii
viii Preface to the Second Edition—2017
upon my own professional practice and I hope will be of some value. Further, I
have added a new chapter on the management of airworthiness programmes.
Also, many reviewers, quite rightly, criticised the first edition for its relatively
limited range of references; so, I have substantially increased the number of
pointers to further reading, mostly as footnotes within the main chapters. Second,
reviewers also praised the first edition for its use of real-world case studies; so I’ve
significantly expanded these within this second edition, from a mixture of my own
experiences—which hopefully will be mostly new to the reader, and better known
examples such as Nimrod XV230 and the A320 howl, with which readers may be
more familiar. I’ve tried to maintain a balance here between light and heavy, civil
and military: as I did with the first edition, reflecting both my own experiences, and
I hope the interests of the readership.
I’ve also enlarged a few sections. What was previously a section on stalling is
now on departures from controlled flight, so I’ve covered other departure modes
including the spin, tumble and spiral dive. The systems assessment section I’ve
expanded a little into electrical system specifics and some discussion of cyber threat
resistance. There are more illustrations, and also simply lessons I’ve learned in
recent years and wanted to pass on.
So I present the second edition of this book, which I hope that you will find at
least as useful and interesting as the 2014 edition, and provide new and useful
additions. As previously, I would be very glad to hear of any thoughts you have,
just in case we ever move to a third edition sometime in the future—I’m easily
contacted either via Springer or via Cranfield University.
For they had learned that true safety was to be found in long previous training, and not in
eloquent exhortations uttered when they were going into action.
—Thucydides, ‘The History of the Peloponnesian War,’ circa 404 BC.
The practice of airworthiness is a complex one, and I doubt that anybody anywhere
in the world can truly be said to understand the subject in its entirety. However,
there are many people who need to initially study, and then to practice various
trades within this professional umbrella.
I’ve worked in the field of airworthiness in its many forms since about 1989
when as a young engineer at the Royal Aerospace Establishment (RAE),
Farnborough I was asked to look into the reasons why engine mounting bolts kept
failing on a prototype experimental aeroplane—I rapidly discovered that this was an
area in which every one of the subjects I’d studied—structures, materials, aero-
dynamics, writing, drafting, maths, etc. came together and still left my knowledge
base lacking. Over the years since, I’ve been called upon to consider whether a
wide variety of aircraft—both new and existing—were safe and fit for purpose, or
in other words airworthy. Mostly, this has involved in-depth consideration of
individual aspects of an aircraft design; a specific instrument or system, the landing
gear, the flying qualities or the flying controls for example. However, in 1997, I was
appointed by the British Microlight Aircraft Association as their Chief Technical
Officer, a post which I held until 2005 and required me to routinely consider not any
specific aspect but complete aircraft designs, often starting from little more than a
rough set of drawings—but with the ultimate objective of an approved and flying
aeroplane. This post was unusual in that the comparative simplicity of microlight
aeroplanes allowed me an oversight of the whole airworthiness process in a way
that the complexity of modern aircraft denies to most modern engineers.
Whilst I was employed at BMAA, I was asked by Sheffield University to teach a
basic course to their undergraduates in Aerospace Engineering in aircraft certifi-
cation; this was a great honour, but also concentrated my mind on how one can go
about breaking down the skills of the airworthiness engineer and teaching them.
About the same time, I was finishing a long-term personal project—a PhD thesis
ix
x Preface to the First Edition—2014
entitled ‘airworthiness evaluation techniques for small light aeroplanes’, which did
much the same thing. As one might expect, I referred regularly to many textbooks,
and to a greater extent to airworthiness standards published by various civil and
military authorities in the course of this work; however, mainly I found myself
using my increasing store of personal notes and experience.
Then, in 2005, I made the decision to change career and became a lecturer at
Brunel University in London, teaching aeronautical engineering. Unsurprisingly,
this brought a further request to teach the subject of airworthiness. About this point,
I finally realised that there simply was no coherent textbook describing the fun-
damentals of airworthiness practice, and in particular, initial airworthiness: the
methods by which the fitness for service of a new design is established: to me this
was a problem.
In 2008, I moved yet again, and at the present whilst I continue to engage with
Brunel, Sheffield and the BMAA, I now manage FAAM: the Facility for Airborne
Atmospheric Measurements which operates the UK’s BAe-146 Atmospheric
Research Aircraft. My previous experience has become invaluable in understanding
the complex airworthiness processes that sit behind Europe’s most complex
research aeroplane, but I’ve certainly learned a lot of new things to do with
managing a jet airliner, not to mention how you go about introducing a continuous
stream of new equipment onto one. Whilst at FAAM, I finally decided with a bit of
encouragement from Springer and several colleagues to knuckle down and finish
writing this book, which was started in 2006 at Brunel.
And so, 25+ years of my own experience and study has been brought into this
book, along with many centuries of other peoples. In writing it, I have not attempted
to produce something that an intelligent layman could use to launch into airwor-
thiness practice without any other knowledge—frankly, I don’t think that that is
reasonably possible. Like any other aeronautics specialist, my knowledge of air-
worthiness is built upon a much wider technical education, and I believe that this
will always remain essential. Nor have I tried to explain how to practice in any
single environment, since there are too many working environments: civil and
military, light and heavy, regulated and deregulated. What I have done, however, is
lay down what I consider to be the main subject areas of initial airworthiness in a
way that I hope will be usable by practicing engineers, students and teachers of the
subject. There are topics which might be included but have simply been omitted
through a need to constrain the length and complexity of the book; also many topics
are addressed in a depth below that which an experienced practitioner would
hopefully be familiar with that specialist topic—but nonetheless I have aimed to
achieve a competent overview, and hope that the reader will find it so.
Throughout this book, I have attempted to teach by example, and nothing in this
is here for the sake of it—every piece of theory has been used, by me, at some point
on real-world airworthiness problems.
This book is inevitably based upon the way in which I’ve myself taught and
practiced airworthiness, and as such is structured in a way which I hope is rea-
sonably sequential. This is the first such book on initial airworthiness (at least that I
know of) and I’d never claim it as good as it could be; so, if any reader would like
Preface to the First Edition—2014 xi
1 What Is Airworthiness? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 The Basic Principles of Certification . . . . . . . . . . . . . . . . . . . 2
1.3 Civil Aircraft Certification Practice . . . . . . . . . . . . . . . . . . . . . 4
1.3.1 Civil Design Codes . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4 Military Aircraft Certification Practice . . . . . . . . . . . . . . . . . . 15
1.5 Release for Flight Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.6 A Note on Modifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
1.6.1 The Case of the Leaky Oil Cooler . . . . . . . . . . . . . . . 19
1.7 Re-evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.7.1 The Case of the Paris Concorde Crash . . . . . . . . . . . . 21
2 The Atmosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.1 General Principles of the Atmosphere . . . . . . . . . . . . . . . . . . . 23
2.2 The International (or US) Standard Atmosphere . . . . . . . . . . . 26
2.2.1 Troposphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.2.2 Lower Stratosphere . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.2.3 Middle Stratosphere . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.3 Which Altitude Matters? . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.4 Variation in the Tropopause . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.5 The Effects of Atmospheric Conditions Upon Human
Survivability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.5.1 Pressure and Oxygen Supply . . . . . . . . . . . . . . . . . . . 34
2.5.2 Temperature Effects . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.6.1 Standard Atmosphere Exercise . . . . . . . . . . . . . . . . . 40
2.6.2 Life Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2.7 Sample Answers to Exercises, with Commentary . . . . . . . . . . 41
2.7.1 Standard Atmosphere Exercise . . . . . . . . . . . . . . . . . 41
xiii
xiv Contents
Chapter 1
The Case of the Leaky Oil Cooler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
The Case of the Paris Concorde Crash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Chapter 9
The Case of Jet Fuel Ice Cream . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
The Case of the Combusting Propeller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Chapter 12
The Case of the Sioux City Crash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254
Chapter 17
The Case of the Whining Airbus. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344
Chapter 19
The Case of the Mistral Stall Warner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
The Case of the Loss of a Nimrod . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372
xxi
Chapter 1
What Is Airworthiness?
Abstract This book is aimed at advanced students and professionals, who are
expected to have to make decisions regarding the fitness for service of new aviation
products, whether whole aircraft or subsystems. This chapter compares initial and
continued airworthiness and what goes into determining airworthiness, focusing
primarily on civilian aeroplanes, with a brief foray into military codes and speci-
fications. The ICAO process and main civilian design codes (both ICAO and
sub-ICAO) for aeroplanes are described, with how to decide which applies to what
aircraft. The differences between the civil and military initial airworthiness
approaches are briefly described.
1.1 Introduction
This definition1 is both a legal and an engineering one, and gives a starting point
from which we can start to understand what is meant by this word. The first part
defines initial airworthiness, the second continued airworthiness.
At the design and certification stages, engineering led teams must determine the
acceptability of a new product—whether that is a product as small as a new switch,
or as large as a whole aeroplane. They do this by applying a combination of
professional skill and judgment, and the use of standards built upon experience
which stretches back to before the Wright brothers. Eventually, they aim to have
established a safe and acceptable standard of product, and to have documented what
that standard is, and why it is acceptable. This is the process of initial airworthi-
ness, also often called certification.
1
Contained in the huge and very useful Airworthiness Circular AC43.13 “Acceptable methods,
techniques and practices—aircraft inspection and repair” which is used worldwide as a guide to
best technical practices in continued airworthiness.
The process of certifying a civil aircraft, whilst technically about proving that the
aircraft is safe, in practice is about proving that the aircraft/engine/system meets the
relevant certification standard. In most cases, this will be a standard declared to
ICAO, the International Civil Aviation Organisation which was established by the
1944 Chicago Convention to achieve overflight permissions across the world. The
standards used are usually declared to and approved by ICAO—meeting certain
minimum standards already published by the treaty organisation. An aircraft
complying entirely with an ICAO declared code, and entirely using sub-systems
(engines, avionics, etc.) that are also compliant with acceptable standards may be
issued with a document called a Certificate of Airworthiness,2 which allows it the
right of international overflight—at-least to other ICAO countries, and so long as
the authority certifying the aircraft is considered “competent” by ICAO.
Many countries also operate “sub-ICAO” certification systems, intended for
domestically operated aircraft only (i.e. those not normally requiring permission to fly
through other countries’ airspace). This may be used for example by amateur-built
aeroplanes, “warbirds” (aeroplanes formerly military operated but now passed into
civil hands) or former prototypes still flown for special purposes but incapable of
achieving full certification. Typically achievement of sub-ICAO certification will be
marked by something other than a Certificate of Airworthiness, for example in the UK
and under EASA across Europe it is reflected by issue of a Permit to Fly. Generally the
procedures by which a sub-ICAO document is issued will be reduced versions
compared to those which apply to those seeking an ICAO CofA, but are considered
acceptable for non-commercial use within the airspace of the issuing state only. The
operator of a sub-ICAO aircraft never has an automatic right of overflight over other
countries, although in practice many bi- and multi- lateral agreements exist to permit
this. But this is a fraught and complex subject, particularly because no two countries
have identical standards for sub-ICAO aircraft approvals and the level of assurance
can vary hugely from near-ICAO standards (for example the UK), to almost totally
hands-off (for example, the USA).
2
Airworthiness Certificate in a few countries.
1.3 Civil Aircraft Certification Practice 5
The basis of civil certification practice is the use of design codes (termed usually
“Airworthiness Requirements”, or more recently in Europe “Certification
Specifications”). At the start of ICAO certification from the late 1940s each major
aeronautical country had its own set of standards, creating an extremely complex
global environment. In subsequent years, whilst for more localised requirements,
many countries have maintained a small number of local requirements (for example
the UK maintains (the sub-ICAO) BCAR Section S for microlight aeroplanes,
(Sub-ICAO) Section T for gyroplanes, and (ICAO declared) BCAR-31 for hot air
balloons), most ICAO compliant airworthiness work has now polarised on two sets
of standards: these are the European requirements now maintained by EASA, the
European Aviation Safety Agency but formerly by JAA—the Joint Aviation
Authorities and secondly the American standards maintained by FAA, the Federal
Aviation Administration. Increasingly these two sets of requirements have been
converging and it is probably inevitable that at some point in the future they will
converge fully. Already, virtually all design codes use a common format and layout,
and standing US-European committees exist with the objective of achieving as
much standardisation as is acceptable to the two authorities and the industries that
they support.
When approaching a project, it is important to know what design code will be
applied to that project. For an existing aeroplane, it is likely to be defined in the
Type Certificate Data Sheet (TCDS), which is the document required by ICAO
defining the approved standard of the aircraft—for example the certification basis
for the Boeing 737 (Fig. 1.1) is shown in Fig. 1.2.
Fig. 1.1 Excerpt from FAA TCDS No. F16WE for Boeing 737
6 1 What Is Airworthiness?
Part 25 is the “catch all standard” and includes FAR-25, CS.25 and JAR-25, and
covers the same ground as a now obsolete British standard BCAR Section D; apart
from very small differences these three main standards are virtually identical at any
contemporary issue state, although because of the continual progress in the
understanding of best practice in airliner design, these standards change issue state
fairly frequently, so the practitioner must take care with issue states and latest
amendments that may, or may not, apply to any particular project or any particular
1.3 Civil Aircraft Certification Practice 7
3,00,000
UK Microlight standards
GLiders, motorgliders,
2,50,000
2,00,000
1,50,000
Deregulated
ultralights
VLAs
1,00,000
part 23
50,000
0
100 1000 10000 100000 1000000
Maximum permiƩed mass, kg
Fig. 1.3 Number of words versus MTOM for aircraft covered by major civil airworthiness
standards
4,50,000
Words in main airworthiness standard
A380 / part 25
4,00,000
Microlights GLiders, motorgliders,
3,50,000
3,00,000
2,50,000
2,00,000
1,50,000
VLAs
1,00,000
part 23
50,000
0
1 10 100 1,000
Number of passenger seats
Fig. 1.4 Relationship between number of words and number of passenger seats for common
airworthiness standards
time. Part 25 is a large and complex standard, which encompasses virtually all
aeroplanes, although because of its complexity and the rigour with which it must
necessarily be applied, it is only generally used when unavoidable. However, it is
also common that sections of part 25 may be used as a special condition added into
8 1 What Is Airworthiness?
Part 23 has historically been the baseline standard for most light aircraft, and
smaller transport aeroplane approvals. In the USA FAR-23 has existed in various
slowly iterating forms since the 1950s, whilst the substantially similar JAR-23
which became CS.23 came into being in the 1990s. Broadly, part 23 therefore
applies to aeroplanes for which (Fig. 1.6):
However, between about 2008 and 2016, the FAA led a review of part 23 which
reflected a developing concern that part 23 had ceased to be fit for purpose. In
particular it was seen as failing to recognise the potential for aircraft to be in service
for many years and potentially degrade with age—which can require life extending
modifications, that some systems—particularly avionics might be regularly updated
and needed to be able to do so affordably, and that a single “one size fits all”
1.3 Civil Aircraft Certification Practice 9
Fig. 1.6 Typical part 23 Aeroplanes (PA28-161 Warrior II and dHC-6 Twin Otter)
standard had ceased to be fit for purpose over a wide range of aeroplanes which
might stretch from a simple 2-seat piston-single training aeroplane to a 20 seat
pressurised multi-turboprop aeroplane. Therefore from 30 December 2016, a new
form of FAR-23 was issued with four certification levels and two performance
levels, and a few months later a matching CS.23 (titled Amendment 5). These are
shown in Tables 1.1 and 1.2. The effect of this is to create an extremely simple
basic standard, but then to move the previous requirements into AMC (Approved
Means of Compliance) so that they can still be followed, but opening potential for
alternative means to also be applied.
This book hereafter will refer to the pre-restructured version of part 23 as “le-
gacy part 23” and the new versions as “New part 23” (or FAR/CS as applicable).
Where no distinction is made (as is the case for most matters still, although this may
change with future amendments), this implies that the legacy and new versions are
substantially the same. Paragraph numbers of the new and legacy part 23 standards
don’t co-incide—any paragraph number between 1 and 1999 is from the legacy
standards, and from 2000 upwards is from the new standards.
In practice the main implications of the new part 23 standards are a shift from
clearly defined minimum standards, to a set of very minimally defined “outcomes”,
that mostly sum up to a requirement to prove to the authority that the aeroplane is
safe. It is likely that continued compliance with the legacy standard will continue to
be acceptable as a basis for approval against the new standard; an alternative view
might be that it is a move towards the military approach of “fitness for purpose”
10 1 What Is Airworthiness?
described later in this section, rather than the historical civil standard approach of a
“pass/fail” requirement to demonstrate compliance with a detailed standard.
CS.VLA was developed in the early 1990s as a simple, but ICAO compliant, code to
allow relatively inexpensive development and certification of non-aerobatic light
aeroplanes with no more than two seats and a single engine, designed for use in VMC
only. For companies developing aircraft in this class it was a welcome development in
certification practice. For the student of this subject, it also provides an excellent
learning tool, since it contains most of the major elements of any fixed wing air-
worthiness standard, but remains relatively simple and therefore accessible when
trying to understand the format and use of civil design codes. Whilst a European code
(and in fact originally developed from the UK microlight standard, BCAR Section S,
described below), this standard is also accepted in the USA with the same operational
restrictions. CS.VLA applies to aeroplanes for which (Fig. 1.7):
• MAUM 750 kg
• VS0 45 kn CAS
• No. Seats 2
Whilst unclear at this time, it appears likely that in coming years, EASA may
subsume CS.VLA into the new multi-level part 23 described above.
CS.22 is derived from a domestic German airworthiness standard for gliders, which
was in use during the 1970s and 1980s before being modified as a common
European standard which subsequently has been accepted worldwide as a basis for
issue of a Certificate of Airworthiness for gliders and motorgliders. Differentiation
may be made locally (particularly within pilots’ licensing regulations) between true
sailplanes (those which always require a winch or tow launch and have no other
power source), self-sustaining sailplanes (those which contain small “get you
home” powerplants), self launching sailplanes (those containing larger “get airborne
but little else” powerplants) and finally touring motorgliders (possessing sufficient
power and fuel for a take-off and sustained powered flight). However, whilst some
internal distinctions exist, this single airworthiness standard may be applied to all
such aeroplanes. CS.22 applies to aeroplanes for which (Fig. 1.8):
CS.22 has historically been accepted by the FAA in the USA, although domesti-
cally FAR-23 is the more normal standard for glider and motorglider certification.
BCAR Section S is not formally an ICAO standard, but it is based upon several
(particularly JAR-22 and JAR-VLA), and is recognised as an acceptable standard
12 1 What Is Airworthiness?
Fig. 1.9 Typical Section S Aeroplanes (Flight design CTSW and Pegasus QuikR)
for this class of aeroplane in many countries. It refers to the class of lightweight
aeroplane commonly called “microlights” or “ultralights” (although the latter term
has many definitions depending upon country). It applies to aeroplane for which
(Fig. 1.9)…
• VH 100 kn
• No. Seats 2
Table 1.3 A selection of deregulated airworthiness categories (all single pilot with no passengers,
unless stated otherwise)
National Category Upper limits Regulation
regime
Australia “Low momentum – MTOW 300 kg (+20 kg with an Civil
ultralight emergency parachute, +35 kg for aviation
aeroplanes” seaplanes and amphibians) order
– Wing loading 30 kg/m2 at MTOW 95.10
United Powered – Empty mass 70 kg (75 kg with an ORS4
Kingdom paragliders and emergency parachute) No. 1224
hang-gliders – Vs 20 kts CAS
United Foot launched – Empty mass 60 kg UK air
Kingdom microlights navigation
order
United “SSDR” single – MTOW 300 kg (315 kg with an UK air
Kingdom seat deregulated emergency parachute, 330 kg for navigation
microlights seaplanes and amphibians) order
– VSO 35 kts CAS
United Powered and – ZFW 254 lb for powered aircraft, or FAR-103
States of Unpowered 155 lb for unpowered aircraft. (Plus
America Ultralightsa 24 lb for a parachute system, 20 lb for
each primary water landing float and
10 lb for each outrigger float)
– VS0 24 kn CAS
– VH 55 kn CAS
– Fuel capacity 5 US gal
Whilst FAR-103 applies to single place aircraft, including balloons, hang-gliders and aeroplanes,
there are also various exemptions permitting 2-place operation for training under certain
conditions, but adding little to the genuine airworthiness oversight
for example. At the same time, the limits shown in Table 1.3 were not developed
because it was considered that such aircraft presented low risk to the aircraft’s
occupants, but because they present low risk to third parties. As reflected in the
quote at the start of this section, the potential for loss of life on board is no less for
the aircraft being deregulated, and any wise engineering team will apply the same
engineering principles as they would be required to for any regulated aeroplane—
whilst of course taking advantage of the removal of the substantial regulatory
burdens which inevitably come with formal oversight and regulatory compliance. It
is the view in some sectors of the light aviation community that standards such as
BCAR Section S or CS.VLA are excessive and that this provides a welcome
opportunity to build aircraft to lower (and thus in their view more appropriate)
engineering standards or, more likely, simply ignore large areas of airworthiness
assessment. The author strongly disagrees with this view, has refused to support
projects applying that philosphy, and would strongly encourage his readers to do
likewise (Figs. 1.10 and 1.11).
1.4 Military Aircraft Certification Practice 15
Fig. 1.11 The reality and innovation of modern deregulated ultralights. The Polish built Ekolot
KR-010 Elf
Fig. 1.12 T-50 Golden Eagles of the Republic of Korea Air Force
budget in the tens of billions. Political will is such that a project may virtually never
be permitted to fail—yet this political and financial determination does not detract
from the vital importance of delivering to the customer an aircraft that has been
proven to do the job for which it was designed, within a sensible and useable set of
operating limits. Whilst the scale may reduce with smaller projects, the importance
and priorities of successful military procurement do not change.
This state of affairs has become more pressing since the 1970s, so that now even
the USA has largely ceased to be able to afford competing indigenous projects, with
the knowledge that if an aircraft failed to meet the customers specification, there
would be another which would—the only countries who can afford true competition
are those with little local aerospace industry. Although the governments of such
countries at-least still have sufficient competition that they can afford to “shop
around”; when they do, this shopping usually includes a substantial element of
reliance upon the manufacturing countries own military airworthiness system, so
relatively little new certification work will be done by the customer nation.
However, so far as indigenous procurement projects are concerned, now, and for
the foreseeable future, a company or consortium will be selected. But, all real
development and testing is done in the (near) certain knowledge that the aircraft
must, and will, be purchased and approved. [Notwithstanding the even more certain
knowledge that the budget and specification will be amended at-least annually
throughout the 10++ years of the project timescale].
1.4 Military Aircraft Certification Practice 17
Military aircraft design codes exist, and in-fact they are considerably more
detailed, than civil design codes (most countries also still maintain their own to a
greater or lesser extent—for example US Mil-Stds or British Def-Stan documents).
However, these documents do not have the high degree of imperative that a civil
standard does. The primary document that is considered in the design of a military
aircraft (or aeronautical system) is the specification.
The specification is a large and complex document, which in the twenty-first
century contains considerably more than only the requirement for an aeroplane to
fulfil a certain set of design requirements—it is likely to include requirements for
interoperability, training, logistic support, upgradeability, compatibility with
existing ground and airborne equipment, maintenance man-hour requirements,
fatigue life—and of-course cost, which may be either purchase or (particularly for
transport or training aircraft) rental costs. This is all in addition to the more obvious
requirements of range, speed, payload, endurance and the basic ability to perform
specific airborne tasks.
Nonetheless, it would appear logical that the military airworthiness system(s)
could operate on a similar basis to their civil equivalents—requiring production of a
clear and unambiguous set of compliance reports, showing how (and that) each
specific point of the specification is met. In practice however the sheer complexity
of any such declarations, together with the vital importance of any such project to a
country’s national defence, has forced most governments to be “expert purchasers”
and led to a substantially different approach to such approvals—this approval work
rotates around an the “Official Test Centre”, or OTC as previously mentioned.
No aircraft can be certified without flying it, and even once a type has received
certification, testing of modifications, new equipment, and envelope expansion will
be required throughout the service life of the aircraft. This is true irrespective of
whether an aircraft is civil or military, new or modified, and equally irrespective of
the role of the aircraft. At the same time, sometimes it is appropriate to fly an
uncertified (and possibly uncertifiable) experimental aircraft so that new data can be
obtained to further aeronautical knowledge.
Therefore, uncertified aircraft must be able to be flown—in ways that meet three
requirements, in descending order or importance:
1. Without significant risk of 3rd party damage or injury during testing.
2. Without significant risk of damage or injury to the test aircraft and its occupants.
3. In a manner which allows adequate data to be brought back about the state of the
aircraft and its equipment.
Inevitably therefore, a system is required by which an aircraft can be released for
safe and productive flight testing. This requirement is universal; the approach to
meeting the requirement not however: every country, every authority and to a large
18 1 What Is Airworthiness?
extent every organization and project has its own approach and philosophy to
meeting these needs. This is unsurprising—it would be totally inappropriate to
apply the same approach to the first flight of a new airliner versus a new microlight,
nor for testing a new engine on a single engine fighter versus a multi-engined
transport aeroplane. The only absolute is that any system for releasing an aircraft for
flight test must rest primarily upon the judgment and experience of airworthiness
(and in particular flight test) specialists.
A mature release to flight test system may be typified by that in use in Britain
called “B-conditions”, which is the term used within British civil aviation to define
the mechanism by which an organization may fly an aircraft without a CofA or
other equivalent formal release to service.
B-conditions firstly requires the organization intending the flight testing to put in
place the infrastructure to be able to make a reasoned judgment about whether an
experimental or prototype aircraft should be flown. This needs to be appropriate to
the nature of the aircraft and project, and will generally be defined in a manual
approved by the national airworthiness authority, but the essential elements will be:
• A design organization, managed by an accountable chief designer.
• An airworthiness review organization, or in smaller organizations competent
co-signatories to the Chief Designer.
• A maintenance and inspection organization, capable of ensuring that the aircraft
and equipment are in the state laid down by the design organization.
• A flight test organization, capable of planning and performing the actual testing.
(And usually also encompassing safety and training organizations.)
The scale of this will depend upon the nature of the task. A small light aircraft
manufacturer or independent flight test consultancy may well consist of only one
very competent specialist (in some cases combining all of the above roles) supported
by 2 or 3 “as required” internal or external consultants; a large aircraft company such
as Boeing, Airbus or AgustaWestland will have several hundreds of specialists in
these roles, and an extremely large and complex management structure.
In 2015 the UK introduced a variation upon B-conditions, termed E-conditions.
This has permitted competent persons to be given full authority over a flight test
programme on any manned aircraft up to 2,000 kg. As a set of processes, the UK
community is still learning how to use E-conditions, but they are already providing
a useful mechanism for prototype testing and airborne research, and both the
published procedures and growing body of experience may prove a useful resource
to flight testers in other countries also.3
3
E-conditions procedures are contained in UK CAA Publication CAP1220 “Operation of exper-
imental aircraft under E conditions”. The definition of a “competent person” normally means a
Chartered Engineer who is a Member or Fellow of the Royal Aeronautical Society. Whilst it’s
understood that not all Chartered Aeronautical Engineers are flight test specialists, their code of
conduct requires that they will always work to their personal limitations and enlist specialist
support where required—in the same way that a physician would, if presented with a medical
condition outside of their personal specialisms.
1.5 Release for Flight Test 19
Because they are the baseline of airworthiness approval, in all regimes, this book
will concentrate on the approval of whole aircraft. However, throughout most
aeroplanes at various stages in their service life can expect to be modified. The
process of approving modifications requires a good understanding, always, of the
basis on which the base aircraft was originally approved. The engineer asked to
manage the approval of modifications to that aircraft must obviously understand the
design, or proposed design, of the modification also, and then must identify the
portions of the original approval basis which will be affected by these changes to
the aircraft. In the majority of cases it will be the original approval basis that is used
for the approvals, but in some cases it will be acceptable to all concerned to use a
current standard. Often an acceptable approach is an analysis of the old form of the
aircraft, and the post modification form and to demonstrate “no worse than before”,
which is a particularly helpful approach when considering old aircraft, potentially
approved one or more generations prior to the modification, and for which the
evidence of the original approval basis may be extremely limited. The term often
used for such old, known good, but poorly documented designs is “legacy”.
The author was co-owner of a 1947 Stinson S108-2 Voyager aeroplane, with an
obsolete Franklin 6A4 powerplant. The oil cooler developed multiple leaks and was
beyond repair; replacement certified parts were not available, nor was clear
20 1 What Is Airworthiness?
documentation of the basis upon which the original powerplant installation was
approved (Fig. 1.13).
In discussion with the appropriate authority (the UK CAA) the author agreed
with them that the present form of FAR-23 was an acceptable basis for the approval
of any modifications, as the present version of the standard to which aeroplanes in
that class would now be approved. Refining that discussion, a set of paragraphs
which should apply to the modification were also agreed.
An uncertified part was identified that met the requirements for “form fit and
function” within the powerplant, and an inspection regime that would establish its
fitness for purpose as a one-off part, in lieu of formal release documentation. A set
of required inspections and reports were agreed to show compliance with the
standard—this included physical load tests (Fig. 1.14), inspection points counter-
signed by an independent engineer, a revised weight and balance report for the
aeroplane, and an extended and witnessed engine running test.
These were submitted to the authority, who used this an a basis for approving the
modified aircraft, which then returned to normal use.
1.7 Re-evaluation 21
1.7 Re-evaluation
In most cases and most regimes the initial airworthiness approval of an aircraft or
system remains in force throughout its life. However, there are circumstances where
it may be re-evaluated, either in part, or in whole. The most common such cir-
cumstance is where an accident or incident causes the quality of the original air-
worthiness approval to be brought into question. That this should happen
occasionally is sadly inevitable—airworthiness evaluation, particularly as it often
means evaluating new technologies and designs whose safety implications are not
yet fully understood—will be imperfect. This of course is provides no excuse for
initial airworthiness teams not to perform their duties as completely as possible.
Typically in such a case, the accident investigating body (such as the UK’s
AAIB or the USA’s NTSB) will make recommendations for re-evaluation of
aspects of an aircraft’s initial airworthiness (a wholesale recommendation, whilst
not unknown, is extremely unusual). The overseeing airworthiness authority (such
as CAA, EASA or FAA) will then decide whether to act upon this recommendation,
and if so require and oversee re-evaluation by the Type Certificate Holder—that is
either the manufacturer or their organisational successor. This will re-evaluate the
aircraft, most likely to some agreed combination of original and present standards,
supplemented by a military-esque consideration of fitness for purpose. In extreme
cases the aircraft type may be grounded in the interim, creating significant pressure
of-course to resolve any issues as quickly as posssible—although of-course that
need for speed must never be permitted to reduce standards.
On 25th July 2000 at 1643 local time, an Air France Concorde (AF Flight 4590,
registration F-BTSC—similar to that in Fig. 1.15) took off from Paris Charles de
Gaulle airport. During the take-off run one of the tyres struck a metal “foreign
object” later identified as having fallen from a preceding Continental Airlines
DC10. A tyre failed, throwing 4.5 kg/10 lb lump of debris up into the wing, which
penetrated the fuel tank starting a fire and fuel leak.
The resulting events caused the aircraft to crash into a hotel 2 min later killing all
109 people on board, as well as 4 people on the ground. The severity of the event
caused, as the events started to become understood, grounding of the entire
Concorde fleet. The particular reason given for the grounding at the time was that a
single initiating event—the runway debris—had caused the loss of the aircraft, and
that within the thinking behind modern airworthiness that is unacceptable.
Over the next year, the design authority—Airbus—re-evaluated the airworthi-
ness of the type, and concluded that there was a need for several modifications:
these included stronger burst-resistant tyres, Kevlar armoured fuel tanks, and some
modifications to the aircraft wiring. With these incorporated across the fleet, and the
22 1 What Is Airworthiness?
Fig. 1.15 A surviving British Airways Concorde, on display at Aerospace Bristol in the UK
Type Certificate restored the type was then re-introduced into service in September
2001, remaining in service for another 2 years before being retired on economic
grounds.
Chapter 2
The Atmosphere
Abstract The atmosphere is the medium through which all terrestrial aircraft
travel. Thus, it is vital that any discussion of airworthiness begins with a discussion
of the properties of the atmosphere, and how, in calculation and testing, we char-
acterise the medium. The core components of the atmosphere, its different layers
and how they are characterised are described. The International Standard
Atmosphere (ISA)—its standard definitions and reference latitudes, and how ISA
influences and informs airworthiness practice are also described. An integral part of
airworthiness is human survivability in the airborne environment and as the final
part of this chapter, how atmospheric conditions (pressure, oxygen content and
temperature) affect human survivability are discussed.
The atmosphere is the sphere of gas that surrounds the Earth, it is made up of the
gas mixture which we call air. Air itself is primarily a nitrogen/oxygen mix, with a
lot of small components, made up as shown in Table 2.1.
In practice this is not quite right, since there is often a proportion of water vapour
also present in the air (typically somewhere in the range 0–4%), however this tends
not to be considered for airworthiness purposes (except when considering the
chemical effects of that moisture, such as corrosion or mass gain).
At all aircraft operating altitudes, it is safe to assume that this mix remains
consistent. Therefore, virtually all human endeavours, including the building and
operation of flying machines, takes place in an essentially diatomic gas mixture of
78% Nitrogen, 21% Oxygen, and about 1% of “other” mixed gasses which tend not
to affect gas properties much.
1
A note on terminology: Conventionally, “height” implies a distance above the ground or a
pre-determined point on it (such as a building or runway), whilst “altitude” implies a distance
above mean sea level. We may also refer to pressure altitude or density altitude, each of which
becomes important at different times.
2.1 General Principles of the Atmosphere 25
In fact we consider, based primarily upon temperature variation, that the aero-
nautically usable atmosphere exists in three layers:
From the surface to a nominal 11,000 m (or 36,069 ft) there is the troposphere, in
which temperature and pressure decrease. The top of the troposphere is called the
tropopause.
From about 11,000 m to about 20,000 m (or 65,617 ft) we enter the lower
stratosphere, in which pressure continues to reduce, but temperature remains
constant at a nominal value of 216.7 K (−56.5°). The top of this layer is called the
lower stratopause.
Above 20,000 m and up to 32,000 m (104,987 ft) we enter the middle stratosphere.
Again, pressure continues to decrease, but now the temperature increases. This
terminates at the middle stratopause.
There are further layers above this—the upper stratosphere and then the meso-
sphere, these are mentioned later, but in practice we will only consider behaviour up
to the top of the lower stratosphere, at about 65,617 ft when temperature ceases to
be constant—since virtually all aeronautical endeavours take place below this.
Whilst the generalised characteristics of the earth’s atmosphere have been well
understood since the mid nineteenth century, in the post-WW2 expansion of civil
aviation, it became important for several reasons that the world all used a common
atmosphere model (although several, national standard atmosphere models had
existed since the 1920s). There were many good reasons for this, for example
ensuring a common standard in structural calculations, and ensuring that all com-
mercial air-traffic is using altimeters calibrated to a common scale.
In 1952, NACA—the (US) National Advisory Council for Aeronautics, a pre-
decessor organisation to NASA, published a set of tables and data for what was
referred to as the “ICAO Standard Atmosphere” showing data to 65,800 ft
(20,056 m)—the limit of reliably explored atmosphere at that time; it was based
upon atmospheric data at about a latitude of 40°. This was accepted worldwide,
although more commonly known as the “US Standard Atmosphere”.
As atmospheric data was revised, this atmosphere model was revised by NACA
and later NASA in 1958, 1962, 1966 and most recently 1976—whilst no change
occurred to the low level model, much work was done to expand the model
upwards. The 1976 US Standard Atmosphere was (as with previous NACA and
NASA generated atmosphere models) adopted by ICAO and accepted as a
worldwide standard—a status it retains today, although it is more commonly
referred to nowadays as the “International Standard Atmosphere” or “ISA”.
ISA is published for several latitudes—15°, 30°, 45°, 60° and 75°, however
unless stated otherwise (and for virtually all airworthiness work) published values
correspond to 45° latitude—a nominal temperate condition representing
2.2 The International (or US) Standard Atmosphere 27
2.2.1 Troposphere
T ¼ T0 ð1 h=145;542 ftÞ;
or
q ¼ q0 ð1 h=145;442 ftÞ4:255876 ;
or
q ¼ q0 ð1 h=44;329 mÞ4:255876 ;
P ¼ P0 ð1 h=145;442 ftÞ5:255876 ;
P ¼ P0 ð1 h=44;329 mÞ5:255876 ;
2
The “classic paper” describing tropopause structure and folds is Stratospheric-Tropospheric
Exchange Based on Radioactivity, Ozone and Potential Vorticity, by Edwin F. Danielsen, pub-
lished in Journal of Atmospheric Sciences May 1968, (vol 25, pp 502–518). A more recent and
extremely informative paper is Turbulent Mixing within Tropopause Folds as a Mechanism for the
Exchange of Chemical Constituents between the Stratosphere and Troposphere, by MA Shapiro,
also in Journal of Atmospheric Sciences (Dec 1979, vol. 37, pp 994–1004).
2.2 The International (or US) Standard Atmosphere 29
It is also common and useful to define altitude in terms of a single ISA related
parameter; hence it is common to find altitude referred to as “Standard Pressure
Altitude” (sHp) or “Standard Density Altitude” (sHd). Occasionally it is also
convenient to think in terms of “Standard Temperature Altitude” (sHT) [the author
has on one occasion used an aeroplane’s OAT—Outside Air Temperature gauge
successfully as a crude altimeter after the sole barometric altimeter had failed],
however this is only ever of use if remaining constantly below the Tropopause, and
away from any temperature inversion (this is a common atmospheric feature where
for a short period of a climb temperature increases whilst still relatively low—
usually at a few thousand feet, but often higher).
There are effectively four types of altitude then that we have discussed. These are:
1. Pressure
2. Density
3. Geopotential
4. Temperature
There is a fifth, which was historically not readily available, but now is, and that is:
5. Geometric
Geometric altitude is obviously the distance above the earth’s surface, but that they
raises an obvious question—above what level of the earth’s surface?, since it even a
flat sea is affected by tides, the earth is well established not to be a perfect sphere,
and the land is demonstrably seldom flat. Therefore most sources, including most
GPS receivers, show altitude above a standardised modified sphere—the most
common currently being the WGS84 spheroid3—established since 1984 although
built upon earlier models and also revised several times since. WGS84 assumes a
flattened sphere centred on the earth’s centre of mass, and uses a zero meridian
5.3 arc seconds (just over 100 m at Greenwich) east of the Greenwich Meridian.4
Each has a role within airworthiness practice and aircraft operations, and whilst
these will be discussed in greater depth later, it’s useful to understand now what
each is used for
Pressure Altitude is used for altitude reference in most flying, usually by reference
to a pressure altimeter. It is also the critical parameter when determining the ability
of the human body to absorb oxygen. Pressurised aircraft cabins will normally be
set to an equivalent pressure altitude, which will be a design parameter.
Temperature Altitude is of primary interest in aircraft operations in comparison to
the dewpoint (where cloud tends to start forming) and the freezing level (when ice
starts to form, and thus around which airframe icing becomes most critical). In
theory it could also be used as a crude operational measure of altitude within the
tropopause, but it seldom is.
Density Altitude is the fundamental parameter when determining aircraft perfor-
mance; it is also necessary when converting between Equivalent and True
Airspeeds, and thus is essential to both cruise performance determination, and
instrument calibration.
3
ICAO currently uses the February 1998 V2.4 version of the WGS84 Implementation Manual,
authored by Eurocontrol and Institute of Geodesy and Navigation (IfEN) in Germany. Copies of
this are readily found online.
4
The history of the Greenwich Meridian, and the measurement of longitude is fascinating but
outside of the scope of this book. The author would however recommend Dava Sobel’s popular
science history “Longitude” (Fourth Estate, London, 1996).
2.3 Which Altitude Matters? 31
Whilst ISA (based upon 45° latitude) assumes that the tropopause exists at a uni-
form 36,089 ft/11,000 m, in practice it varies a great deal with latitude and with
local conditions. Typically the polar tropopause is much lower and warmer than the
tropical tropopause, and the summer tropopause is higher than the winter
5
GPS is the most commonly use GNSS (Global Navigation Satellite System) technology, but the
Russian GLONASS and European Galileo systems are also, at-least in theory, available.
Combined GPS/GLONASS receivers are becoming increasingly common and have advantages of
duplication, reliability, acquisition time, and usability at very high latitudes—the GPS constella-
tion can become unreliable beyond 80° latitude—depending upon constellation configuration at
the time.
6
OSGB36 is managed by the Ordnance Survey, website www.ordnancesurvey.co.uk.
7
AGD is managed by Geoscience Australia, website http://www.ga.gov.au/.
32 2 The Atmosphere
8
See Letcher T.M., Climate change: observed impacts on planet earth, Elsevier 2009 [Ch7], also
many reports by the IPCC or International Panel on Climate Change via http://www.ipcc.ch/.
9
See Seidel D.J., Randel W.J., Variability and trends in the global tropopause estimated from
radiosonde data, J GeoR, Vol. 111, D21101 (2006).
2.5 The Effects of Atmospheric Conditions Upon Human Survivability 33
Most airworthiness calculations involve consideration only of the aircraft and not of
its occupant(s). However, a brief mention is made here of human beings—who are
generally the operators and occupants of any aircraft (or at least those aircraft for
which airworthiness is the greatest concern).10
The human body relies upon two things primarily in its environment for
short-term survival; those are an acceptable temperature, and an acceptable partial
pressure of oxygen—the nitrogen component of the atmosphere, so far as the aviator
is concerned, is of sole use as a provider of air density, and not of life support.
10
This book only considers some key airworthiness related aviation medicine topics. For those
wishing to explore the topic further, the definitive reference on aviation medicine is Ernsting’s
Aviation and Space Medicine—regularly updated. However most users find Human Performance
and Limitations and Aviation by Ron Campbell and Mike Bagshaw adequately detailed, much
cheaper, and considerably more accessible. Much simplified texts also exist in various forms as
aids to private and professional pilot theoretical studies.
34 2 The Atmosphere
Whilst it is not safe to assume that every human body will behave identically—for
example a 20 year old non-smoking athlete should certainly have far better toler-
ance to low oxygen levels than a 60 year old sedentary person (although, inter-
estingly some research indicates that smokers are more tolerant to high altitude
conditions), there is a clear relationship—the lower the partial pressure of oxygen,
the poorer the performance of the human body. Figure 2.6 is representative of the
time that it is accepted an average adult can expect at various pressure altitudes to
remain usefully conscious—that is before they pass out or at-least become inca-
pable of performing any rational task.
It can be seen that below about 15,000 ft, it is likely that an average pilot or
passenger should remain usefully conscious for a prolonged flight. Above this
however, they will start to enter a state known as hypoxia, or oxygen starvation. It is
unsurprising then that it is illegal in most countries to fly for prolonged periods
above a certain pressure altitude without the use of supplementary oxygen—in
Europe this is 10,000 ft for continuous exposure with a permission to go up to
13,000 ft for up to 30 min or 16,000 ft for up to 10 min,11 whilst in the USA it is
12,500 ft continuously and 12,500 to 14,000 ft for up to 30 min,12 (the variation
reflecting only slight differences in local interpretation of aeromedical evidence).
11
See EASA SPO.OP.195 “Use of supplemental oxygen.”
12
See 14 CFR 91.211 “Supplemental oxygen.”
13
There are numerous web sources about the life of the repulsive Dr. Rascher, and one eBook: The
Fall of the House of Rascher: The bizarre life and death of the SS-doctor Sigmund Rascher, by
Siegfried Bär. Rascher was executed in 1945 by his own Nazi superiors for fabricating the results
of experiments to breed an improved “master race.”
2.5 The Effects of Atmospheric Conditions Upon Human Survivability 35
Oxygen systems (Fig. 2.7) can assist this; for example reference to ISA tables
(plus remembering that air is only 21% oxygen) will show that the partial pressure
of oxygen at 10,000 ft would be about 0.688 0.21 = 0.144. So, 100% oxygen at
this partial pressure should simplistically give an equivalent level of human body
performance—in this case at about 45,000 ft (Fig 2.6).
In practice, determining the human body’s oxygen requirements is somewhat
more complex, and dependent upon the characteristics of the lungs. It can be taken
that a human’s lungs add water vapour and carbon dioxide to the air-mix. More
precisely, as somebody breathes in, the trachea (or windpipe) adds a fixed partial
pressure of water vapour of about 6.26 103 Pa.14 Then when this gas mix enters
14
For consistency with atmospheric calculations, Pa (Pascals, or N/m2) are used here; however, if
referring to medical textbooks for more detail, be aware that the unit most commonly used there
for gas pressure is mm Hg, or millimetres of mercury. 1 mm Hg = 133.3 Pa. Textbooks for pilots
on physiology (often called Human Performance and Limitations, or HPL) also commonly use
mm Hg.
36 2 The Atmosphere
the alveoli, where transfer of gasses into and out of the blood takes place, a further a
fixed partial pressure of CO2 of about 5.33 103 Pa is added.
So, considering the sea-level case, where air pressure is 101.3 103 Pa, and
Oxygen (O2), the critical gas comprises 0.21 (21%) of the air breathed in. Assuming
dry air and trivial CO2 content, we can assume that prior to being breathed in, air
consists of:
Ambient Air at Sea-Level
0.21 O2 [referred to as fO2 (atm)]
0.79 N2 [referred to as fN2 (atm)]
After entering the trachea, and a partial pressure of water vapour being added,
the critical value of pO2, the partial pressure of oxygen changes to the tracheal
oxygen partial pressure of:
Tracheal Conditions
Inserting values:
f O2 ðatm) ¼ 0:21
After this, this gas mix enters the alveoli where homeostasis maintains the
previously mentioned partial pressure of CO2[pCO2] of 5.33 103 Pa. This, and
similar calculations, allow us to determine the partial pressure in the alveoli:
Alveolar Conditions
Inserting values:
from above
Therefore,
This is a useful figure, because it tells us partial pressure of oxygen in the alveoli,
which might, for example, be required in an air-ambulance’s patient oxygen system,
so as to ensure that the patient requires to simulate sea-level conditions. A more
common calculation however would be to use this information, and the existing
knowledge (as indicated by Fig. 2.6) that a human being can rely upon effectively
indefinite useful consciousness up to 10,000 ft.
Based upon standard atmosphere tables (see the Appendices to this book), we
know that relative pressure at 10,000 ft (ISA) is 0.6877, and can assume that
proportionally this remains 21% oxygen; so, the alveolar partial pressure of oxygen
can be calculated in the following manner:
This offers a critical value, above which one should maintain the alveolar partial
pressure of oxygen. This value can then be used in reverse, to determine the
maximum altitude at which one may fly in an ambient pressure/100% oxygen
environment (i.e. in an unpressurised cockpit, with an oxygen mask providing pure
oxygen). So:
15
Where high-g manoeuvring causes loss of vision this progresses through tunnel vision to what is
known as “blackout” which will typically occur at 3–5 g in a untrained person, several more in
acclimatised aircrew with further improvements achievable using “g-suits”, which are specialist
clothing that constrict the lower body thus keeping more blood in the torso. Around 1.5 g beyond
blackout, aircrew are likely to suffer “g induced loss of consciousness”, commonly abbreviated to
“G-LOC”. When young and fit the author could work easily to 5 g without a g-suit but suffered
blackout at about 6.5 g; he also noted that reversing g from e.g. −2 to +4 g quickly could provide
extremely rapid G-LOC. He does not recommend the experience, but this is a routine risk within
the working life of fast jet aircrew or aerobatic pilots.
40 2 The Atmosphere
Temperature effects are less easy to define since short-term exposure to high
temperatures or (much more likely) low temperatures has no significant effect upon
the human body.
Prolonged exposure to extremes of temperature degrades the body’s perfor-
mance (including at low temperatures an increased susceptibility to hypoxia); in
extremis either hypothermia (exposure) or hyperthermia (heatstroke) can occur,
causing permanent damage to the body and potentially death in very severe cases.
However, local exposure to high or low temperatures (that is, touching some-
thing very hot or very cold with unprotected skin) can cause immediate damage to
the body—again restricting that person’s ability to function.
It’s therefore important that, except in short-term emergency conditions (such as
ejection from a military fast jet) aircraft occupants are protected from either general
(atmospheric) or local extremes of temperature and wherever possible a normal
“room temperature” condition around 20 °C is maintained. This is clearly not
always possible—the author is not about to recommend grounding many thousands
of safe and enjoyable vintage, open cockpit or microlight aeroplanes, and similarly
has worked on a research aeroplane forced to tolerate cabin air temperatures
exceeding 35 °C—but it may often be a consideration when designing or approving
such an aircraft to ensure that certain parts are protected from human contact, or that
provision is made for occupants to wear appropriate clothing. (Additionally some
sensors or electronic equipment can also become temperamental at extremes of
temperature.)
2.6 Exercises
Using the following boundary conditions and formulae, derive an atmosphere table
from 2,000 ft below sea-level to a tropopause of FL400 for the following atmo-
spheric conditions
T0 ¼ 30 C
You are being consulted on the design of the oxygen system for an air-ambulance
helicopter. Whilst the helicopter will never fly above 10,000 ft and thus will not
require a crew oxygen system, casualties may often require oxygen delivered at
ambient pressures, and it is important that their alveolar oxygen levels do not drop
below that which they would receive at sea level breathing normally.
Determine the maximum standard pressure altitude below which it would be safe
for a casualty to breath 100% oxygen?
T0 ¼ 30 C
P0 T0;ISA
q0 ¼ q0;ISA
P0;ISA T0
101; 325 288:15
¼ q0;ISA
101; 325 303:15
¼ 1:164 kg/m3
2.8 Sample Answers to Exercises, with Commentary 43
Inserting values:
Therefore,
Inserting values:
p O2 ðtr Þ ¼ 95 103 Pa
from above
Therefore,
Now working backwards from this value, we can calculate the altitude at which this
may be delivered using 100% breathing oxygen:
PATM ¼ p O2 ðtr Þ=f O2 ðpureÞ
þ pH2 OðtrÞ
ð3Þ ¼ 95:3 10 Pa=1:00 þ 6:33 103 Pa
3
¼ 101:6 103 Pa
Much that an engineer does in the practice of airworthiness, and everything that a
pilot does in operating an aircraft, revolves around airspeed. In airworthiness work:
airspeed determines the aerodynamic loads which apply to the structure, and the lift
and drag which determine inertial loads are all functions of airspeed. In flying, a
pilot must be able to accurately determine his or her time of flight, position, and
ensure that they do not exceed safe operating limits. However, there is not a single
term which one may measure and term “airspeed”, there are a number of different
speeds, which are used in different applications. These are:
1. Groundspeed (G/S): The speed which an aircraft is travelling relative to a fixed
point on the ground.
2. True Airspeed (TAS): The speed at which an aircraft is travelling through the
air surrounding it. In level flight this is simply G/S adjusted for wind; in
climbing or descending flight, it is G/S adjusted for wind and slope.
Alternatively, TAS is obtained from EAS (or vice versa) by correcting for
altitude errors. Flutter onset, if that occurs, is generally at a fixed TAS value
also.
3. Indicated Air Speed (IAS): This is, very simply, the readout of an Airspeed
Indicator (ASI). So, all operating limits need to be declared in IAS.
4. Calibrated Air Speed (CAS): This is the IAS, corrected for known position and
instrument errors. CAS is sometimes also called Rectified Air Speed (RAS),
particularly in American publications.
5. Equivalent Air Speed (EAS): This is the CAS, corrected for compressibility
(not generally necessary below about Mach 0.6 or FL100, where it can be
assumed that EAS CAS). EAS is the value most commonly used for struc-
tural calculations.
So, it can be seen that having used EAS in structural calculations, it is necessary
to make corrections to this value before obtaining indicated values that can be used
as operating limitations. A pilot, in turn, must make corrections from what is
indicated in the cockpit to TAS, and then correct for wind to give a groundspeed,
which can then be used for flight planning. All of this makes an understanding of
airspeed, in its several forms, vital to both flying and airworthiness practice.
There are various means which may be used to measure airspeed; these include
laser and ultrasonic Doppler anemometers, hot-wire anemometry, or even the
venerable cup anemometer. However, for aircraft design purposes, virtually all
practical airspeed methods depend (rather than upon measurement of the velocity of
air molecules) upon the direct measurement of dynamic pressure, as defined by the
term q ¼ 1=2qV 2 : In the earlier days of aviation, some devices were used which
seem crude by modern standards, but nonetheless worked well, for example the
airspeed indicator fitted to the de Havilland Tiger moth consisted of a flat plate
pushed against a spring by dynamic pressure.
Some such unusual devices are still in use particularly at the lighter end of
aviation. However, virtually all fixed and rotary winged aircraft now flying measure
airspeed by use of a Pitot-static system.
Most engineers and pilots will be familiar with the incompressible form of
Bernoulli’s equation, which states
1
p p1 ¼ qV 2 : ð3:1Þ
2
Allowing airspeed to then be estimated from two pressures: Pitot (or dynamic)
and static. On an aeroplane (and many other types of machine) this is done by
positioning two small holes—one forward facing (so as to measure p, which is
called the Pitot pressure) and one (or often several) positioned in as near to still-air
as can be achieved (for example behind a “static dam” or on the side of the aircraft)
so as to measure p0, which is called the static pressure. These holes may be
positioned together on a single device (which is often called a combined pressure
head), or may be positioned in separate parts of the aircraft—forming part of an
overall Pitot-static system (Fig. 3.1).
Fig. 3.1 The huge Pitot tube on scaled composites’ “Proteus” research aircraft—hinged upwards
for safety on the ground
48 3 The Pitot-Static System
Pitot and static lines before flight—to avoid the risk of water freezing and blocking
or rupturing the lines in flight.
Similarly, the F-16 fighter aircraft in Fig. 3.4 shows a similar configuration—a
Pitot-head at the front of the aircraft in clear air, whilst the static ports (not easily
visible in this picture) are in front of the wing leading edges on each side.
Whilst these show the “classical” configuration; depending upon aircraft design,
the Pitot and static could however be positioned virtually anywhere on the aircraft.
ICAO compliant civil airworthiness standards do not usually permit more than the
greater of 5 knot/3% airspeed indication error (whichever is greater) and so much
design time can often be devoted to ensuring positioning that ensures this
throughout the flight envelope. Often it may be found that the Pitot and static are
combined in a single pressure head such as that shown from a PA28 light aeroplane
in Fig. 3.5.
Regardless of the physical location and appearance of the Pitot and static sources
on the aircraft, the basic design will almost certainly be identical and similar to that
shown in Fig. 3.6: the “static line” (a narrow piece of pipe or hose1) will be fed by
one or more static sources (matched statics on opposite sides of the aircraft are
common, so as to minimise the effects of any sideslip: an effect which will normally
be checked for during flight testing), whilst a Pitot line is fed by the Pitot tube or
Pitot-head.
1
The words “pipe” and “hose” often get mixed up. A pipe is rigid, whilst a hose is flexible.
Thinking of a drainpipe or a garden hose helps get this right.
50 3 The Pitot-Static System
There will then normally be three instruments attached to these lines. To the
static line alone are attached the altimeter (which is an absolute pressure gauge—or
barometer), and a vertical speed indicator (also known as a VSI, or a variometer)
which measures the rate of change in static pressure. Attached to both the static and
Pitot lines is the airspeed indicator (ASI), which is a differential pressure gauge
calibrated to read an indicated airspeed. [Machmeters will be discussed later, but if
fitted, would also be fitted between the Pitot and static lines, similarly to the ASI.]
Since dynamic pressure will directly affect the loads upon the aircraft, it is
possible to declare operating limitations as Equivalent Airspeed (EAS), which will
be constant irrespective of altitude. This is always based upon a standard sea level
value of q = q0 = 1.225 kg/m3 (0.076 lb/ft3)—so any airspeed indicator will be
calibrated to read (as near as instrument and pressure errors allow), True Air Speed
3.2 Pitot-Static System Design 51
Pitot
Head
Vertical
Airspeed Reservoir
Altimeter Speed
Indicator
Indicator
Static Source
(TAS) only at ISA sea-level conditions. At higher altitudes, where q\q0 ; therefore
TAS > EAS [one of several reasons why airliners typically cruise at very high
altitudes.].
One may see by inspection that the relationship between TAS and EAS will
pffiffiffiffiffiffiffiffiffiffi
therefore be that TAS = EAS q0 =q; although this is more normally expressed as
TAS ¼ EASpffiffi where r represents the relative air density, q=q : This relationship is
r 0
important for many applications, the most common being that pilots must adjust
EAS to TAS, and then further correct for wind so that they can accurately predict an
aircraft’s heading and groundspeed (Fig. 3.7).
3.3.1 IAS
Indicated Airspeed, or IAS is the value read from the airspeed indicator. It may be
predicted (for example when determining operating limitations) by reversing the
calculations shown below.
3.3.2 CAS
As was mentioned earlier, IAS has to be corrected for system errors to give Calibrated
Airspeed, or CAS, which in turn is (where conditions demand it—normally above
0.6 Mach or 10,000 ft) corrected for compressibility effects to EAS.
52 3 The Pitot-Static System
In a perfect world IAS and CAS would be the same—that is, if there were a
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2ðpD pS Þ
perfect system. They would both be defined by VC ¼ q —and in-fact this is
0
the formal definition of CAS. However, in practice there will be errors in the
system, due to leaks, mechanisation inaccuracies, and particularly the pressure field
around the aeroplane meaning that neither the measured Pitot nor static pressures
are absolutely correct. In particular, static readings tend to be particularly
problematic.
The only way to reliably determine the relationship between IAS and CAS is
experimentally. Data is generally presented in the form of a graph of IAS versus
CAS in an aircraft manual, although increasingly it may become also available
electronically.
Most airworthiness standards define a minimum accuracy for an airspeed indi-
cator system of the greater of 3% or 5 knots, from 1.3 VS to the maximum speed
permitted—this usually ensures that for most piloting purposes it can be assumed
that IAS CAS, although for particularly precise navigational tasks (for example a
long sea or desert crossing) the wise pilot would still carry out the necessary
correction, and corrections are always necessary when the primary Pitot-static
system is used for flight testing. For many microlight aeroplanes and amateur built
aeroplanes, standards do not require a minimum level of accuracy and static vents
may be within the cockpit; this can mean large discrepancies between IAS and CAS
which pilots are not always aware of. This will seldom be acceptable however on
larger aeroplanes.
Regardless of aircraft class, for any engineering and flight test purposes, it is
vital to carry out a data reduction. This applies, for-example, to the use of indicated
airspeeds for any form of performance or handling analysis, or when determining
3.3 Calculating Airspeed Values 53
operating limitations which must always be given to operating crew in the form of
IAS. There are numerous means of calibrating an airspeed indicating system, the
most common of which are described in Sect. 3.4 below.
Once certified it is normal to publish the ASI calibration curve in the operators
manual (indeed, most airworthiness standards will require this), which may also be
referred to as the PEC (or Pressure Error Correction) curve—although this is not a
truly accurate description since, for-example, any inbuilt instrument errors, if
known to exist, should also be accounted for in a PEC curve.
Figure 3.8 shows a copy of the PEC curve as published in the operator’s manual
for a part 23 aeroplane. There is no firm requirement as to the presentation of this
information, but this example is typical.
Fig. 3.8 Airspeed indicator calibration graph for a typical part 23 aeroplane
54 3 The Pitot-Static System
3.3.3 EAS
For the majority of airworthiness and flying tasks, the difference between CAS and
Equivalent Airspeed, or EAS can be ignored. However, if an aircraft is operating at
Mach numbers above about 0.6, or density altitudes above 10,000 ft then the
difference between these two airspeed values must normally be accounted for—
especially in stress related work, where structural loads are a function of EAS, not
of CAS.
The reason for the difference is that air can no longer be considered incom-
pressible, and this invalidates the simple form of the Bernoulli equation from which
the Pitot-static equation is derived. So we require a more complex form that allows
for compressibility:
Equation for Mach Number from Pitot and Static Pressures
ðc1Þ !
2 PD c
M ¼2
þ1 1 ð3:3Þ
c1 PS
or,
Equation for Equivalent Air Speed from Pitot and Static Pressures
ðc1Þ !
2a2 PD c
VE2 ¼ þ1 1 ð3:4Þ
c1 PS
This in itself does not allow a ready correction from CAS to EAS, but if we
consider a simple correction formula:
which may also often be written as: VC þ DVC ¼ VE . Figure 3.8 shows a typical
correction chart, this may be re-arranged to:
And since we know the definitions of EAS and CAS, we can write:
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
!ffi sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
u
u 2a2 PD ðc1 Þ
2ð pD pS Þ
DCAS ¼ t
c
þ1 1 ð3:7Þ
c1 PS q0
3.3 Calculating Airspeed Values 55
It may be seen that from this, and inserting values of T and PS which are derived
from either actual data or (more commonly) a standard atmosphere. It is possible to
create a chart or spreadsheet of corrections. It however is not readily amenable to an
elegant algebraic solution, so normal practice is to use a computer generated
standard chart (since this should be independent of aircraft design) such as that
shown in Fig. 3.9.
An alternative method available to both pilots and engineers, is the venerable
Dalton Flight Computer (such as that in Fig. 3.10 and commonly called a
whizz-wheel2), more advanced versions of which will include the ability to make
Scale Altitude Corrections to airspeed, probably to about 3 significant figures (of
CAS/EAS) which, although it may take some time to learn, offers (eventually!) a
simple method of data reduction. The flight computer is simply a specialist circular
slide rule—slide rules being devices which although rarely used now for everyday
calculation still often find useful applications for specialist (and otherwise complex)
calculations such as this.
2
Or increasingly less polite terms by younger pilots who were brought up on all things electronic.
56 3 The Pitot-Static System
3.3.4 TAS
Remembering that below around 0.6 Mach and 10,000 ft it can reasonably be
considered that CAS EAS, the relationship between EAS (and usually CAS) and
True Airspeed, or TAS can be defined by considering the following:
Relationship Between TAS and EAS
VE
V ¼ pffiffiffi ð3:9Þ
r
3.3 Calculating Airspeed Values 57
Which is the standard conversion between TAS (V) and EAS (VE). The relative
air density, r, may be derived from atmospheric data, although at first approxi-
mation and for much work ISA is adequate.
3.3.5 Groundspeed
Fig. 3.11 Pilatus PC12 cockpit showing the FMS controls and display, incorporating GPS, in the
lower centre console
It has never proved possible to accurately predict the PEC for an airspeed indicator
system, and even if such a method were developed, it would still be essential to
check the results experimentally. Numerous techniques have however been
developed to allow flight test teams to calibrate an ASI experimentally. Since the
relationships between CAS, EAS, TAS and Groundspeed are all known and pre-
dictable, this task should in theory be quite simple—in practice it rarely is.
The ASI calibration task can be done in one of two ways: either by finding a
means of accurately measuring wind vector and groundspeed, or by comparing to
an airspeed indicating system of sufficiently known accuracy. It is common to use
more than one method, so that the results of each method can be cross-verified,
3.4 Methods of ASI Calibration 59
The following was developed around 1999 and has been used to good effect on a
number of projects since.4 Required are turbulence-free conditions (an essential for
any ASI calibration task), accurate knowledge of outside air temperature, a GPS
unit, and approximate wind heading data.
The aircraft is pointed as accurately into wind as the forecast will allow. Precise
wind heading is then obtained by varying heading slightly whilst maintaining
constant speed and height. The aircraft is known to be exactly into wind when the
lowest indication is obtained of GPS groundspeed. This heading is noted.
The aircraft is flown at a range of speeds from just above the stall (the FAA
mandates 1.2 Vs in its flight test guide AC23-8; this guide is a de facto world
standard), to at-least VH (often to VNE) with GPS groundspeed being noted against
indicated airspeed at each increment. Where the airspeed exceeds VH, and thus the
aircraft is forced to descend, the time between two heights (normally about 200 ft,
greater altitude changes potentially causing significant changes in the TAS:CAS
relationship) is recorded to allow correction during subsequent analysis.
The aircraft, maintaining a constant altitude, is then turned (using GPS heading
so as to not be affected by any magnetic anomalies) onto a reciprocal heading, and
this exercise repeated. If necessary (limitations of available airspace tend to control
the flightpath) multiple turns are flown in a “racetrack” method as indicated below.
The data is then reduced, using a table such as that shown in Table 3.1 (Fig. 3.12).
3
A library of information on RVSM certification may be found at https://www.faa.gov/air_traffic/
separation_standards/rvsm/documentation/.
4
GB Gratton, Use of Global Positioning System velocity outputs for determining airspeed mea-
surement error, Aeronautical Journal Vol. 111 No. 1120 (June 2007) pp 381–388.
60
Do
wn
win
d le
g
This data is then plotted to produce an ASI calibration chart. It has been found
important to use at-least 5 points down to as near the stall as is reasonably
achievable (because of the likelihood of low-energy discontinuities, and hence the
desire to allow as little extrapolation into this potentially uncertain area as possible).
At higher speeds, discontinuities or significant changes of curve form have not
generally been noted above VH and this area can be treated with less rigour.
This uses a similar means for groundspeed determination to that described for the
racetrack method above, but instead uses three legs, separated by 120°.
Groundspeed must be measured, using a GPS unit, whilst flying the aircraft on three
headings (not tracks—so heading must be measured using an error corrected
compass, not the GPS) that differ by 120° (e.g. 50, 170, and 290°). These speeds
will be termed V1, V2 and V3.
The mean sum of squared speeds, V′2 is calculated as
We now non-dimensionalise the three groundspeeds and term them each a, so that
Vn2
an ¼ 1; ð3:12Þ
V0 2
A variant upon the triangle method above was devised independently and published
by J. T. Lowry5 and referred to as the “Box Pattern” method. This flies three legs at
90° spaced magnetic headings (one being due North), and then by trigonometry
(reproduced below, using Lowry’s terminology but again without proof, which may
be found in the reference) the TAS is determined at each speed.
Three groundspeeds are recorded for each IAS value; these are g1 (flown due
magnetic North), g2 (flown on magnetic heading 90°) and g3 (flown on magnetic
heading 180°). Variables p, q, a are used within the calculation and have no
physical significance.
g21 þ g22
p¼ ð3:16Þ
2
2
2g2 g21 g23
1
a ¼ tan ð3:17Þ
g23 g21
g23 g21
q¼ ð3:18Þ
4 cos a
5
J T Lowry, Performance of Light Aircraft, ISBN 1-56347-33A-5.
3.4 Methods of ASI Calibration 63
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
p þ p2 4q2
TAS ¼ ð3:19Þ
2
This is again a valid method, the box-pattern method uses three rather than two
speeds (giving greater opportunity for error in an individual datum to be reduced by
calculation) and also does not present the risk of inadvertently appearing to declare
an emergency posed by the triangular method, although still requiring more time to
fly than the racetrack method for the following reasons. It also shares with the GPS
triangle method the disadvantage of potentially being prone to errors due to mag-
netic anomalies.
All of these GPS based methods do rely upon flight being within a constant wind
field—any significant changes in wind vector within the test area will introduce
errors that can’t readily be eliminated subsequently. For lower performance aircraft:
most light aeroplanes and helicopters, this is generally not a problem. For higher
performance aeroplanes it can be a problem, so high altitude jet aircraft testing, or
almost any combat aircraft testing, it must be ensured through the use of expert
meteorological advice that there will be no significant change in wind vector in the
test area—generally speaking this means that on a pressure chart isobars must be
parallel and evenly spaced throughout the test area, and that no testing is close to
frontal systems.
Prior to the advent of GPS, various methods were also developed which were
essentially variants and predecessors of the three methods listed above—these used
other methods of measuring speed such as time over a ground course,
kinetheodolite measurement, or ground radar. All of these remain valid techniques,
although in general are less accurate and more expensive in time and resources than
the more modern use of GPS.6 An alternative philosophy, which has also often been
used, is to assume that the Pressure Error Corrections due to the Pitot (PPEC)
are insignificant and that all significant errors are due to the static (SPEC).
6
The earliest known publication describing the use of the ground course method is F L Thompson
(Langley Memorial Laboratory, NACA), The measurement of air speed of airplanes, NACA
Technical Note No. 616, October 1937. Although clearly not a recent paper, the method published
therein is consistent with that used up until the advent of portable GPS units.
64 3 The Pitot-Static System
This assumption has been commonly used and is usually entirely valid so long as
the Pitot design keeps it always in clear flow forward of the aircraft, and at low
values of local angle of attack.
7
Readers with a strong interest in helicopter flight testing are referred to Alastair Cook and Eric
Fitzpatrick’s book “Helicopter Test and Evaluation”, (Wiley 2002) ISBN 978-0632052479.
8
Keil G, Total-head meter with small sensitivity to yaw, NACA Technical Memorandum No. 775,
Washington August 1935.
3.4 Methods of ASI Calibration 65
Fig. 3.13 Kiel Probe (enclosed Pitot) from X-31 aircraft (Courtesy of NASA)
from the X-31 aircraft which used it as the primary airspeed reference due to the
large range of AoA experienced by that aircraft—although unfortunately the
aeroplane itself was lost when icing of the Kiel probe led to a lack of input to the
flight control system and resultant loss of control.9
This is very straightforward, but expensive. The aircraft whose system(s) require(s)
calibration flies in formation with another aircraft whose calibration state is well
established. XT597, a McDonnell-Douglas Phantom II FG1 aircraft was used by
A&AEE Boscombe Down in this role, participating in most UK military aircraft
certification programmes for 20 years or so, although it was retired in the 1990s (its
engines ending up in Thrust SSC, a supersonic car).
The method remains in use however, particularly for supersonic aircraft SPEC
calibration.
9
This and several other accidents worth learning from are described in NASA’s 2012 book
“Breaking the Mishap Chain: Human factors lessons learned from human factors lessons learned
from aerospace accidents and incidents in research, flight test and development” (ISBN
978-1782662464, but also available free from NASA’s website).
66 3 The Pitot-Static System
3.5 Machmeters
There was a demon that lived in the air. They said whoever challenged him would
die. His controls would freeze up, his plane would buffet wildly, and he would
disintegrate. The demon lived at Mach 1 on the meter, seven hundred and fifty miles
an hour, where the air could no longer move out of the way. He lived behind a
barrier through which they said no man would ever pass. They called it the sound
barrier.
Tom Wolfe, ‘The Right Stuff’
For higher performance aeroplanes, the other way in which airspeeds can be
expressed, is relative to the speed of sound. This is expressed as a Mach Number,
defined as
V
M¼ ð3:21Þ
a
Like an airspeed indicator, a Machmeter is reliant upon both Pitot and static
pressures. The device itself physically resembles an ASI, but contains two pressure
capsules so that both absolute Pitot and static pressures can be determined, then a
mechanical gearing mechanism converts these two values into a dial-gauge indi-
cation of Mach Number (as illustrated in Fig. 3.14). Inevitably more modern airline
and high performance military aircraft will now simulate this effect electronically,
but the way in which the system works overall is unchanged.
It is normal, even with glass cockpit (CRT or LCD displays) for airspeed to be
represented in an analog form: either a dial such as Fig. 3.15 or a strip—experience
has shown that whilst it may be a useful addendum to this, a numeric display of
airspeed takes longer to interpret, and also does not give the intuitive sense of rate
of change in airspeed, nor of the margin between current airspeed, and airspeed
limitations, that a more classical and clearly marked up dial or strip does.
The usual convention as can be seen is for a single needle, which rotates
clockwise from a zero which is vertically upwards. The units of indication are
marked on the dial (units most likely to be encountered are knots, mph, or kph—in
theory knots are the worldwide aviation standard, but in practice this is only true in
large commercial aircraft in the western world).
Blue radial line For most multi-engined aeroplanes, this is the best climb
speed with the critical engine failed (or more precisely,
the speed at which the airworthiness performance
requireme nts for single engined climb were met).
Second red radial line For gliders and motorgliders, this is the best rate of climb
speed, VY. The marking may also be used on some other
aircraft, again to indicate VY, but in such cases is usually
discretionary on the part of the certification team, and not
usually mandated by regulations.
For many multi-engined aeroplanes, this is the most
conservative (highest) value of VMC—the minimum
control speed with the most critical engine failed.
Finally, it should be noted that the ASI shown in Fig. 3.15 shows these coloured
markings on the dial face, and not on the covering glass. This is the most common
means of minimising the risk of instrument mis-reading due to parallax error,
although careful positioning of the markings relative to the pilots design-eye
position should be an acceptable alternative, for a single-pilot cockpit. Again, EFIS
3.6 Display of Airspeed Limits 69
negates this problem, but the majority of aeroplanes still use analog dials for
primary reference, and virtually all aeroplanes will have analog dials in the cockpit,
if only as an emergency backup in case of computer failure. These dials however,
are likely to be put in whatever corner of the cockpit is most out of the way, so
parallax can often be a serious problem.
pressure setting, depending upon conditions of the day; in most of the world, this
setting is in hPa (or mb, which are identical), however in some countries either
mmHg (millimetres of mercury, standard value 760 mm) or inHg (inches of mer-
cury, standard value 29.92″) are used. These settings are referred to in terms of “Q
codes” which are a hangover from the earlier days of aviation, when a great deal of
information was passed by Morse code, and in particular navigation information
was described by three-letter codes beginning with “Q”. Those relating to altimeter
setting, are:
QNE Standard ISA altimeter setting (1013.25 hPa,
760 mmHg, 29.92″Hg). This is usually
selected by aircraft operating above a transi-
tion altitude and is also used to define Flight
levels.
QFE It is also usual for flight test results, except for
those related to field performance, to be
expressed using QNE, and when viewing such
results this can normally be assumed unless
stated otherwise.
Airfield QNH This is an airfield altimeter setting used in a
few countries (primarily the United Kingdom)
to show zero at the highest point of the active
runway.
This is a local altimeter setting, such that an
altimeter would correctly read the height above
mean sea-level (amsl) at the airfield to which it
corresponds. This is often referred to simply as
“QNH” or in North America as “Altimeter”
where it will normally be declared in inches of
mercury.
Regional pressure setting (RPS) This is the most conservative (in terms of
terrain clearance) of a number of airfield QNH
values for a geographic area, and is generally
used for traffic separation below the transition
altitude. E.g. an aircraft flying at 2000 ft
between two airfields within the same altimeter
setting region, where the transition altitude is
set to 3000 ft, will set regional QNH whilst in
transitRegional Pressure Setting (RPS).
None of this changes for electronic altimeters, except that sensing, indicating and
subscale setting, becomes electronic rather than mechanical.
The primary airworthiness implications of altimeter setting (apart from accuracy,
which is discussed in Sect. 5.9) are that the subscale window must be clearly visible
from the pilots design eye position, and that the setting knob must be reachable
3.7 Pressure Altimetry 71
quickly and regularly by a pilot strapped into his seat (in addition, it must be rugged
enough to take rapid and occasionally clumsy operation by that pilot). Figure 3.17
illustrates the typical mechanism of an altimeter.
There are various methods which may be used to calibrate an altimeter or static
system (in effect, to determine SPEC—the Static Pressure Error Corrections). These
include:
Ground Calibration This is done either as a single point calibration (against a
known good altimeter) or as a multiple point calibration using a calibrated test set
connected to the instrument’s static port. It can then be determined what errors exist
within the altimeter itself and the device either accepted, adjusted, or rejected.
However, this is only of use for determining aircraft static system overall errors
(SPEC) if the whole-aircraft system errors are already known. In reality it only
provides useful information about the cockpit instrument itself. Nonetheless it’s a
routine action within continued airworthiness for all certified aeroplanes.
Aneroid/Flypast Method A sensitive and calibrated pressure gauge (normally an
aneroid barometer) is placed on a tower (commonly an airfield control tower) at an
absolutely known height.
72 3 The Pitot-Static System
The aircraft is then flown past the tower as near as possible to the tower height,
at a range of speeds and configurations—this clearly requires a high degree of
precision in the flying task, and so may be inappropriate for the early stages of a test
programme where there may be insufficient confidence in the aircraft to conduct
routine low-level flying.
From a known point (or better, two known points) the aircraft is photographed as
it flies past the tower, generally with a high speed cine or video camera. Analysis of
this film (and a 3d trigonometric analysis to identify exactly where the aircraft was)
allow exact determination of the aircraft’s height relative to the barometer on the
tower. By comparing the known altitude of the aircraft (that is, the altitude of the
tower aneroid, adjusted by photographic analysis) to that indicated on the cockpit
instruments (and also, for very accurate SPEC determination, to an accurate aneroid
connected to the aircraft static line). Thus, by simple data reduction, the SPEC can
be determined. It is common also to use this method for TPEC determination, by
assuming that the PPEC are trivial. However, some justification is likely to be
needed to convince a competent certifying authority that PPEC are effectively zero,
and this is unlikely to be taken on trust—at-least not for an aircraft being certified
for flight by reference to instruments alone.
Formation Flyby Method The formation method has previously been described
as a method for PPEC determination. If a chase aircraft is available whose SPEC are
known to a high degree of confidence, then it is possible to use a formation flyby to
determine SPEC.
The method used is that the slower of the two aircraft maintains a steady con-
stant altitude in a still air mass, whilst providing some source that allows the faster
aircraft (initially positioned well behind the first) to formate upon it—for example
smoke, differential GPS, or air-to-air TACAN (a military variant of VOR/DME).
Maintaining (visually via formation references) an identical altitude, but sufficient
lateral separation to ensure that the airflow around neither aircraft is affecting the
static source of the other.
At the point of flypast, the displayed altitude (or preferably an accurate aneroid
barometer reading connected to the static of each) is recorded for each aircraft. So,
by conducting this at various speeds (and if appropriate, configurations) for the test
aircraft, it becomes possible to build up a picture of the differences in SPEC
between the two. Then, the SPEC data for the known aircraft can be adjusted to
provide SPEC data for the aircraft being evaluated.
This method is expensive and time consuming, but is for example almost cer-
tainly the only reliable method of determining SPEC for supersonic aircraft.
The Trailing Static Method This falls within the leading/trailing Pitot-static
method described earlier. As with other comparison methods, data reduction is
based upon comparison to the known characteristics of the trailing static (generally
calibrated using a wind tunnel).
3.8 Methods of Altimeter/Static System Calibration 73
Note: GPS
GPS measures geometric altitude, an altimeter measures pressure altitude.
Except for small corrections, it is unsafe to assume that the relationship
between GPS altitude and pressure altitude is consistent or predictable.
Therefore, never attempt to use GPS to directly calibrate an altimeter, or as a
reserve altimeter for traffic avoidance purposes. Combined with good map-
ping data however, GPS is extremely valuable for terrain avoidance.
Most airworthiness standards (certainly all those for aeroplanes which may ever be
flown by sole reference to instruments: which includes parts 25 and 23) require a
minimum accuracy to be demonstrated for the combination of static system and
altimeter(s). This will be defined in the airworthiness standard being used for
approval of the aircraft type in question10; however, the following should be a
reasonable guide to the requirements of most airworthiness requirements, and all
are likely to require testing as part of the airworthiness approval process:
• The static source should not be significantly affected by opening of doors or
windows (that may be opened in flight) or the operation of devices such as
airbrakes, flaps, cabin pressurisation, anti-icing equipment or retractable
undercarriages.
• Above 1.3VS1 any primary altimetry system must be accurate to within ±30 ft
per 100 kts EAS, or ±30 ft, whichever is greater.
• If a backup altimeter or static system is provided, this must be within ±50 ft of
the primary system unless a correction card is generated and available.
• If an aircraft is authorised for flight in Instrument Meteorological Conditions
(IMC) then the above must also apply with ice accretion on the airframe.
• There should always be a means, at the lowest point of the static line, to drain
any accumulated moisture.
The descriptions above concerning both airspeed and altitude measurement describe
mechanical devices, most of which have remained unchanged for at-least seventy
years. Slowly, the electronics revolution is having an effect—modern airliners such
10
In civil airworthiness standards, probably about paragraph 1325.
74 3 The Pitot-Static System
as the 767 or A320, or even the latest model Cessna 172 are using primarily
electronic sensors to emulate a Pitot-static system, whilst the most recent version of
F-16, the block 60, has also finally replaced the classical Pitot-static system of
earlier models with a totally electronic system. Similarly, many light and microlight
aircraft are also increasingly using low cost (and often more importantly,
low-weight) electronic flight instruments.
However, the classical concepts of altitude, IAS, CAS, EAS and TAS as well as
altimeter subscale setting remain equally valid, and these systems are emulating
mechanical instruments—they are not doing anything new with regard to the source
and analysis of data. So, when assessing such instruments, they should be treated
similarly.
That said, new airworthiness issues will also come into play with such devices.
For example: visibility of electronic displays in bright sunlight, means to provide
colour marking of airspeed limits, reliability of both the device itself and its power
supplies. Whilst the advantages of weight and flexibility that electronic devices
offer are appealing, one must bear in mind that the increasing complexity and
inter-relation of electronic system mean that the airworthiness approval task
becomes considerably greater, not less.
Figure 3.18 shows a design outline for the Farnborough F-1 aeroplane, a projected
6-seat turbo-prop commuter aeroplane.
3.11 Sample Problems 75
Estimated Performance
Max operating speed (Vmo) 285 kts
High cruising speed at FL300 324 kts
Long-range cruising speed at FL300 225 kts
Design manoeuvring speed (Va) 150 kts
Stalling speed, flaps and landing gear down 59 kts
Time to FL250 10 min
Service ceiling 35,000 ft
Source Janes’ All the World’s Aircraft
OATðKÞ
r¼d
OATStandard ðK Þ
where OATStandard is the value for OAT at the test standard pressure altitude as
given in ISA tables.
This should then be plotted: column (a) as IAS, and column(h) as EAS to
produce an ASI correction plot.
Etc…
3.12
Sample Solutions
Determination of SPEC
Students should identify the main issues affecting this procedure, which are:
– That the aircraft has retractable gear, which may affect static pressure errors.
– That the aircraft has a speed range from 0 to 120 kts
– That it is the static system under investigation.
They should then identify the special equipment required, which is:
Aircraft: Pitot-static system confirmed to be fully serviceable and representative
of type.
An accurate aneroid attached to the static line.
Tower: An accurate aneroid at a known height amsl.
Externally: Camera (or cameras) mounted so as to give clear distinction and
scaling of vertical position of helicopter relative to tower.
Site survey information, to allow camera data to be accurately used to
determine relative vertical positions of camera and helicopter.
Students should then lay down procedures similar to those listed below (the
exact order and spacing of tests is obviously not critical, but the data requirements
are).
1. Initially hover aircraft adjacent to tower (stationary to wind) with gear down,
record…
– Helicopter aneroid reading (at aircraft)
– Tower aneroid reading (at tower)
– Temperature (normally at tower, to allow density to be determined)
– Photograph of tower and helicopter using camera, or
– GPS geopotential height for both tower and helicopter.
2. Fly away from the tower, then position for and execute a flypast at 10 knots IAS,
transmitting [for synchronisation with the tower and camera who must also
record data as (1)] at the point where data (aneroid reading) is taken.
3. Repeat at 20, 30, 40, 50, 60 knots IAS.
4. Raise the undercarriage, re-establish hover adjacent to the tower, repeat (1).
5. Repeat (2) at 20, 40, 60, 80, 100, 120 knots IAS.
80 3 The Pitot-Static System
Students should then correctly identify the procedure for data reduction (teachers
may prefer either to ask for a schematic, or a full data-reduction table. Below is
shown a schematic:
For each IAS/undercarriage configuration, the following data should be
available:
– Tower aneroid standard altitude (sHpT)—obtained by reversing the ISA formula
for pressure altitude.
– Aircraft aneroid standard altitude (sHpA)—ditto.
– Camera angular correction between tower and aircraft (or difference between
GPS geopotential heights)—the trigonometry behind this is trivial and not given
here.
To calculate SPEC, firstly for each datum:
– Add/subtract the height correction between tower and aircraft to SHpA to give a
nominal indicated pressure altitude, SHpA′
– Calculate dHIND = sHpT − SHpA′
Now plot two graphs, one for each configuration (gear up and gear down),
relating IAS to dHIND , stating “for actual pressure altitude at any given airspeed”,
add dHIND to indicated altitude.
If the two graphs are co-incident, they may be combined.
a. General Pitot-Static system problem
See Figs. 3.19 and 3.20.
Note: Machmeter. VMO = 285 kts = 146.5 m/s CAS
Service ceiling = 35,000 ft
Where from ISA tables, T * 219 K
pffiffiffiffiffiffiffiffiffi
) c ¼ cRT
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
¼ 1:4 287 219
¼ 297 m s1
146:5
) MMAX ¼ 0:49
297
For each IAS condition, the two groundspeeds should be averaged, giving true
airspeed (TAS). This should then be multiplied by the square root of relative air
pffiffiffi
density ( r) to give CAS. If the aircraft exceeded VH, a correction should be made
using rate of descent to give flightpath CAS. IAS is then plotted against CAS to
give a curve for TPEC.
Determination of Static Pressure Error Corrections
The tower flyby method. The aeroplane static system is to be connected to an
accurate aneroid, which has the ability for a reading to be taken at a time determined
by the pilot. A similar aneroid is to be positioned at a known height in a tower,
positioned that it may be flown past. The aeroplane is then to be flown past the
tower at a range of speeds (the widest commensurate with flight test safety), in a
manner which allows the aeroplane’s static and the tower static to be as close as
possible to the same height. At each flypast, a remote camera must be used to
measure the angular difference in height between the tower and aeroplane—which
may then be converted to give a height difference using trigonometry. For each test
condition, the difference indicated altitude at the tower-aneroid and the
aeroplane-aneroid is to be calculated. From this, the SPEC may be determined in
terms of feet error, as a function of airspeed.
Chapter 4
The Flight Envelope
4.1 Introduction
The Flight Envelope, or V-N diagram defines the range of speeds and normal
accelerations to which an aircraft may be subjected in flight without a risk of either
handling problems, or overstress. Bracketed by the general term “flight envelope”,
there are in fact two main envelopes that have to be considered—these are the
“manoeuvre envelope” and the “gust envelope”.
The manoeuvre envelope firstly is the range of conditions through which, in still
air, an aircraft may be manoeuvred, without overstress. In practice, the design team
may need to define many such envelopes, depending upon the number of available
service (flaps, slats, etc.) combinations there are which can be expected to affect the
lifting characteristics of the aircraft; however, it is adequate here to assume only the
one envelope but to discuss the effects of high lift devices such as flaps.
The full flight envelope as shown in Fig. 4.1 is a graph of normal acceleration on
the vertical axis and Equivalent Airspeed (EAS) on the horizontal axis. The
envelope is normally defined at MAUW, but if desired a series of envelopes can be
defined for different weights (some military aircraft manufacturers have attempted
to automate this, with variable success).
In defining the envelope, the first task is to define the normal acceleration limits
and VD (the design maximum speed).
Except in aerobatic or combat aircraft where a designer may specifically be
seeking particularly high normal acceleration limits, the normal acceleration
(nz or g) limits will usually be set to the minimum magnitudes permitted by the
certification standard (see Table 4.1). There are four normal acceleration limits,
which are:
n1 Maximum positive normal acceleration at minimum speed.
n2 Maximum positive normal acceleration at maximum speed.
n3 Maximum negative normal acceleration at maximum speed.
n4 Maximum negative normal acceleration at minimum speed.
Fig. 4.1 Typical basic flight envelope diagram (from CS–23). (Apart from numeric limits, these
diagrams do not vary significantly between airworthiness standards and this may be considered to
apply equally to large (part 25) aeroplanes, or smaller (parts 22 and VLA) aeroplanes, as well
military aeroplanes. Microlights and ultralights, unless relatively high performance, will not
consider gust loadings
1
Aileron reversal—that is, the tendency of a wing to deflect with the torque loading due to aileron
deflection so as to first became an issue in WWII, where it was a problem with the early Griffon
engined Spitfires. However, it is still potentially an issue and aircraft still exist which suffer from it.
86
For simplicity, virtually all designers using BCAR Section S assume that n3 = −2.0
Notes on Table
(1) CS.23 & 25 specifically permit lower load factors if it can be shown that a lower value cannot be exceeded
pffiffiffiffiffiffiffiffiffiffi
(2) Part 23 requires that VC (in knots) must be at least X W=S, where X = 33 (or 36 for aerobatic aircraft) where the wing loading (in lb/ft2) is up to a value of 20, then X may
reduce linearly as W/S increases, to a value of X = 28.6 when W/S = 100 lb/ft2
(3) The multiplying factors when determining the minimum values of VD in part 23 may be reduced from the stated values at a wing loading of 20 lb/ft2 to 1.35 at W/S = 100 lb/ft2
(4) CS.25 and FAR-23 allow the requirements for the minimum value of VD to be disregarded if an alternative set of criteria, based upon the Mach number increase in a 20 s 7.5°
nose-down dive, and 1.5 g pull-up are met. With engines at the lesser of [75% (reciprocating) or MCP (turbine)] or PFLF at VC, the minimum difference between VC and VD must
be at least 0.05Mach (or 0.07Mach for commuter aircraft). (Para 23.335(b)(4) in both cases)
(5) Part 23 allows the requirements for the minimum value of VD to be disregarded if an alternative set of criteria, based upon the Mach number increase in a 20 s 7.5° nose-down
dive, and 1.5 g pull-up are met. The minimum difference between VC and VD must be at least 0.07 Mach (or no less than 0.05 Mach where an analysis including automatic systems
shows that VD wouldn’t be exceeded). (Para 23.335(b)(4) in both cases)
(6) Part 22 has further requirements that VT, the design aerotowing speed shouldn’t be less than 125 km/h and VW, the design winch-launching speed, shouldn’t be less than
110 km/h. No other standard refers to winch or aerotowing so, if for example certifying a glider against FAR-23, these would be useful working values
87
88 4 The Flight Envelope
VD
0
EAS
-2
-4
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2Wmax
VS ¼ ð4:1Þ
q0 SCL:max
where CL.max is that for the whole aircraft, rather than just the mainplane (to what
extent the two differ is inevitably a function of the aeroplane design).
This is simply derived from the equation for lift, and the assumption of a
maximum stall speed (VS, defined in EAS) occurring at MAUW (Wmax) when the
mainplane pitch attitude is such that CL = CL.max. For present purposes we will refer
to this as VS1, meaning the stall speed in the cruise configuration—that being the
most important section of the manoeuvre envelope.
This stall speed is displayed on this graph (referred to as the V-N diagram) as
shown in Fig. 4.3.
This stall speed value must now be extrapolated between the axes, most nor-
mally using the relationship:
1=2
W
VS ¼ VS1MAUW:1g NZ ð4:2Þ
Wmax
2
1 VD
0
VS1 EAS
-2
-4
4.2 Constructing the Manoeuvre Envelope 89
where CAe defines the Aeroelastic Coefficient for the wing. It has a value of 0.5 for
an effectively rigid wing, a value between 0.5 and 1.0 for a wing which decreases
CL.max with increasing loading: typically a Rogallo wing will have a value of CAe
between 0.65 and 0.82. A value of less than 0.5 would imply an aircraft whose wing
increases CL.max as loading is increased: this might apply to a forward swept wing
with moderate unloaded washout such as that of the X-29 experimental jet Fig. 4.4
or the ASK-13 glider (Fig. 4.5). Note that if there is reason to suspect a non
square-law curve in the sector 0-A, then it will be necessary to carry out stall tests at
as wide a range of loadings as reasonably achievable—by using both a range of
Fig. 4.4 X-29 experimental jet aircraft (nose is on the left). (Courtesy of NASA)
90 4 The Flight Envelope
Having established the relationship between Vs and W.Nz, this is plotted on the
diagram as shown in Fig. 4.6. This diagram now is labelled with four critical points:
4.2 Constructing the Manoeuvre Envelope 91
VS VA VD
0
EAS
-2
G E
-4
A, D, E and G—so the stall speed curve is now labelled (and is commonly referred
to as) the O-A curve.
This envelope is generally sufficient information for flight testing to be com-
menced (remembering that if, for example. the aircraft has various flap and slat
configurations—potentially altering both Nz limits and airspeed limits (see the
section on flaps below) several envelopes will be required).
It is likely that some revision will be required during or after the flight test
programme however. The two main reasons for this are firstly that the theoretical
prediction of VS is unlikely to have been totally accurate—so revision to the form of
the O-A curve, and thus the values of VA and/or VF (see later) will be needed.
Secondly it is common that once the flight test programme is well developed, it has
been found and accepted that the maximum safe airspeed is lower than that orig-
inally predicted: this revised value is labelled VDF. The actual reasons for the setting
of this lower VDF value vary, but may include:
• Flutter (generally of insufficient severity, or at high enough speed that an actual
design solution isn’t essential or affordable).
• The aircraft simply cannot be made to fly any faster: this common with exter-
nally strut or cable braced aircraft, but may also occur either with aeroplanes
suffering limited high speed pitch authority, or combat aircraft designed for very
high speeds but which cannot feasibly reach those speeds except by a massive
exchange of potential for kinetic energy.
• Apparent longitudinal static stability starts to become unacceptably poor.
• Unacceptable distortion of the canopy under aerodynamic loads.
• It becomes impossible to prevent a propeller overspeed above this speed even if
the throttle is closed (generally defined as being an engine or propeller rotational
speed above 110% of the rated maximum, but in such cases check the air-
worthiness standard in use and if necessary confirm with the relevant authority
what is considered an appropriate limit).
92 4 The Flight Envelope
VA VNE VDF VD
0
EAS
-2
-4 G E
Finally, limitations are set for the maximum speed to which the aeroplane may be
flown in service. This limit (in all current standards of which the author is aware)
may not be set to greater than 0.9 VDF and in practice usually is set to exactly this
value—this is VNE, the “Never Exceed Speed” (Fig. 4.7).
Concerning the Negative g Stall Curve O-G is the negative g stall line, bound by
zero airspeed-zero load at the top left, and the point at which the aircraft will reach a
negative g stall at n4 at the bottom right. Beware of the assumption, sometimes
made in aerodynamics textbooks, that the shape of the curve will be a mirror image
of the positive g stall curve; this will only be true if the aerofoil and fuselage
aerodynamic shapes are substantially symmetrical. Symmetrical or semi-
symmetrical aerofoils are sometimes used—they may be found on some combat
aircraft, dedicated aerobatic aircraft such as a Su-26 or Pitts S-1, or occasionally on
a very crude amateur designed aircraft with flat wings, but they would not be used
on most aircraft designs because of the gross inefficiency they would give in the
more critical cruise design case.
One of the several manoeuvre envelopes which must usually be determined for an
aeroplane is that with flaps down, which may include multiple envelopes at various
flap settings. For a flapped aeroplane this is likely to at the very least include two
configurations—LAND, which is defined as gear down and full flaps, and PA
(Powered Approach), which is usually defined as gear down and intermediate
flaps—the configurations to be investigated must inevitably be driven by the way in
which the aircraft will be operated, something that is likely to require both piloting
and engineering input. More complex aircraft will have more configurations—
Boeing and Airbus each typically test at-least 5 configurations for any type.
4.2 Constructing the Manoeuvre Envelope 93
Almost invariably the limiting speed for this envelope will be significantly less
than that in the clean configuration, and will be based upon structural analysis and
testing of the flap system based upon a calculation of deflected aerodynamic loads.
This will define VF (flap limiting speed), which may also be referred to as VFE
(limiting speed for flap extension). It is most common to use a single conservative
value for all flap settings—likely to be based upon the maximum flap deflection;
this however is not universal and it may be necessary to use different airspeed limits
for each flap setting—this is likely to be termed VF1, VF2, etc. reflecting first
and second stages of flaps, or might be given as VF15, VF30, etc. for 15° of flap, 30°,
etc.—there is no universal standard. An airliner will most likely use either an
automated system to determine and display limitations, or a series of flip-cards that
define limits by configuration and weight whilst light aeroplanes are likely to
placard either a range of limits, or the most restrictive with greater information
being available in the manual(s).
The justification for these staggered limitations might be that an aircraft will
approach an airfield and use an intermediate flap setting as part of the process of
reducing airspeed and dealing with pitch trim changes caused by flap selection so
that the landing flap setting may finally be selected (this in fact would be a normal
flying practice, to what extent it necessitates variation of flap limits will depend
upon the aircraft).
It is also rare that with flaps selected an aeroplane will require access to the range
of normal acceleration limits which may be required “clean”. So, it is common (and
permitted) to use a reduced set of normal acceleration limits; for example CS.VLA,
the European non-aerobatic light aircraft standard normally uses values of n1 =
n2 = 3.8 whilst with flaps selected n1 = n2 = 2.0 is permitted. This is typical of any
class of aeroplane, but as with many things, the relative simplicity of CS.VLA
provides a useful explanatory environment.
Flaps are clearly not the only systems that an aircraft may have which affect flying
limitations. Other items may include retractable undercarriage, airbrakes, or
openable or removable doors and windows.
Where such devices protrude into the airflow, then almost invariably they will
require a structural assessment, and it is common for this to determine a safe
limiting speed which is below VNE. In such cases, the limits are generally pro-
mulgated—through operating manuals, placards and possibly automated warnings.
The concerns and structural safety factors applicable to these are likely to be similar
to those of flaps, except that it is rare that any device but flaps (or slats) will
significantly effect the stalling speed, allowing a simple limit to be promulgated,
rather than a separate V-N diagram (or related limits).
94 4 The Flight Envelope
All of the limitations determined within the manoeuvre envelope will have been
determined in terms of EAS.
It is important to remember that very few aircraft display EAS directly, most
display IAS, which (via CAS) must be translated to or from EAS. This applies to all
flying speeds which are provided for use by aircraft operating crews. Where an
aircraft will operate over a wide altitude and/or weight range, this may mean that
either several sets of IAS limitations are provided to allow for these changing
conditions, or it may be appropriate simply to provide the most conservative set of
IAS limits. The decision as to what approach should be taken is likely to depend
upon the operating environment, and the number, role and minimum competence of
crews—it is likely that aircrew themselves should be consulted in reaching this
decision.
Limitations should then be promulgated:
• On the ASI and/or Machmeter
• Depending upon aircraft type, usually in placards or checklists
• Invariably in the operators manual.
Conversely, if the aircraft has a canard, then an initial pitch up can be expected
as the canard initially penetrates the canard, increasing angle of attack, and thus
mainplane loading. As the mainplane encounters the gust also, this effect is
amplified, causing a yet greater loading increase.
The “classical” gust considered for airworthiness purposes is a rising column of air
into which an aeroplane flies. The simplest form of this is a “sharp edged gust”,
such as is indicated in part (a) of Fig. 4.8. In calculating the normal acceleration due
to this gust, it is assumed that the aircraft passes instantaneously from (vertically)
still air into the rising column of air. In most cases, this leads to an over-estimation
of gust loads, and so is in the safe-sense, but over-conservative.
A second approach, as indicated in part (b) of the same figure is to assume, more
correctly that the gust intensity (magnitude of the upward velocity of the air col-
umn) increases linearly with the aircraft’s penetration into it, until a steady state
value is achieved.
The revised approach of a graded gust means that it is no longer possible to
assume that the aircraft does not experience any change in flightpath, and so the
pitch attitude response of the aircraft must be calculated.
An approach commonly taken to calculating an aeroplane’s response to a graded
gust of any shape is to treat the increasing gust strength as a series of steps,
calculating aircraft response to each before passing to the next iteration.
A further method also used is to assume a 1-cosine gust, similar to the form
shown in Fig. 4.9. This may again be treated by grading, effectively breaking it up
into discrete time slices treated as sharp-edged gusts.
For most airworthiness purposes, the sharp edged gust remains the standard when
calculating gust loads (whether discretely, or in steps). To calculate the response of
an aircraft to a sharp edged gust, consider an aeroplane flying at a TAS of V, with an
angle of attack a0 ; upon entering a vertical gust of strength u, the angle of attack
increases by tan1 ðu=V Þ u=V; whilst simultaneously the aircraft’s forward speed
increases from V to (V2 + u2)½.
The increase in lift, is therefore:
@CL u @CL
DL ¼ 1=2qV 2 S ¼ 1=2qVS u ð4:4Þ
@a V @a
where @C
@a is the lift-curve slope. Ignoring the change of lift on the tailplane, the gust
L
1=2qVS @C
@a u
L
DnZ ¼ ð4:5Þ
W
1=2qV @C
@a u
L
DnZ ¼ ð4:6Þ
ðW=SÞ
1=2qV @C
@a u
L
nZ ¼ 1 þ ð4:7Þ
ðW=SÞ
4.3 Constructing the Gust Envelope 97
1=2qV @C
@a u
L
nZ ¼ 1 ð4:8Þ
ðW=SÞ
These are however much more useful if modified to give a result in terms of EAS
(VE), and gust strength will also be expressed in EAS, so becomes ue. Similarly, as
with other standardised airspeed equations, we should express density as a standard
(normally ISA sea-level) value, q0 : Thus, these equations are re-written in the
standard form:
Standardised Gust Response for Sharp Upgust in Level Flight (Ignoring
Tailplane)
where ST is the tailplane area, and DCL;T is the change in tailplane lift coefficient,
which itself will be defined by:
@CL;T uE
DCL;T ¼ : ð4:12Þ
@a VE
where @CL;T =@a is the change in tailplane lift coefficient with local angle of attack,
and @e=@a is the change in downwash angle with wing angle of attack. Now,
inserting the definition for DCL;T given above into that for DP, we can write that:
@CL;T
DP ¼ 1=2q0 VE ST uE ð4:14Þ
@a
(Note that this has more than one significance, since it can also be used to estimate
the incremental tailplane loads due to gusts for purposes of evaluating the tailplane
structure.)
4.3 Constructing the Gust Envelope 99
Now, for positive increments of wing lift and tailplane load, we can write that
DnZ W ¼ DL þ DP; so combining the above in this relationship, it can be stated
that:
Normal Acceleration in Sharp Edged Gust with Tailplane Taken Into
Account
1=2q0 VE ð@CL =@aÞuE ST @CL;T =@a
nZ ¼ 1þ : ð4:15Þ
ðW=SÞ S ð@CL =@aÞ
Kg q0 Ude VE a
nZ ¼ 1 þ ; ð4:16Þ
2ðW=SÞ
0:88lg
Kg ¼ ð4:17Þ
5:3 þ lg
Which in turn has been defined in terms of another term, lg which is the
“aeroplane mass ratio”, defined by
2ðW=SÞ
lg ¼ ; ð4:18Þ
q0 Cag
where C defines the aeroplane mean aerodynamic chord, a defines the mainplane
lift curve slope @CL =@a, and g is acceleration due to gravity.
These formulae are universal, and it should be possible to quote them in most
airworthiness reports without any need for proof, referring only to their location in a
relevant airworthiness standard.2 However, the approximations behind gust alle-
viation are appropriate only to a tailplane configured aircraft where it is correct to
consider gust alleviation, whereas a canard configured aeroplane would suffer gust
aggravation—since this is an uncommon case, it has not been explored here,
however as stated above from the author’s personal experience, this is a far from
theoretical problem.
2
For most civil standards, this will be about paragraph 341.
100 4 The Flight Envelope
Using these formulae, gust loadings are usually calculated for four conditions:
• Maximum upgust at design cruising speed, VC.
• Half maximum upgust at maximum speed, VD.
• Half maximum downgust at maximum speed, VD.
• Maximum downgust at design cruising speed, VC.
Maximum gust strengths are normally defined by the airworthiness authority in the
applicable airworthiness standards, but are most commonly 50 ft/s (15.24 m/s) at
VC and 25 ft/s (7.62 m/s) at VD. Assumed gust loadings are then extrapolated
between these points, and these lines plotted onto the manoeuvre envelope as
shown in Fig. 4.11.
Finally the widest envelope that the aeroplane should see in service (assuming
that in severe turbulence the aircraft will be slowed appropriately—usually to a
nominally defined maximum speed in rough air, which is usually referred to either
as VB or VRA) is defined by combining these diagrams into a combined flight
envelope, which encompasses both manoeuvre and gust loads. In faster aircraft it is
common that the gust envelope lies significantly outside the manoeuvre envelope
and increases significantly the loads that must be met, whilst in lower performance
aeroplanes it may not—indeed in most lower performance microlight aeroplanes
gust loadings can be disregarded completely. This results in a combined V-N
diagram as shown in Fig. 4.12 Referring back, the resemblance to Fig. 4.1 should
be clear.
This final envelope describes the limit structural requirements—that is the loads
which the aeroplane may be considered to experience in service, and which must
therefore be accounted for in the structural approval.
VA VC VNE VDF VD
0
Airspeed
-2 -VD gust line
E
-4
G F
VD gust line
g 6
A C
D
4
VA VC VNE VDF VD
0
Airspeed
-2
E
-4 G
F -VD gust line
-VC gust line
Fig. 4.12 Complete V-N diagram combining manoeuvre and gust loadings
Table 4.2 shows the general gust loading requirements at the time of writing for the
main civil airworthiness standards in use worldwide. It should be noted that due to
the inevitability of special conditions or grandfather rights this can only be a general
guide, and shouldn’t be treated as authoritative. Military aircraft requirements may
sometimes be based upon civil standards, but more commonly will be based upon
contract specifications or expert judgement in light of the aircraft’s role, so it’s
imprudent to generalise.
Notes on Table 4.2
VB and VRA VB, the speed for maximum gust intensity, is the speed
(usually only defined for larger passenger carrying
aircraft) above which an aircraft shouldn’t fly in severe
turbulence. It is also usually the maximum permitted
value for VRA, which is the recommended speed for
flight in turbulence (often the two coincide). Both are
semi-arbitrary, and picked following a handling assess-
ment that confirms it is high enough to ensure that a
gust-induced stall is highly unlikely during such condi-
tions. It also needs to provide adequate margins at either
end between low speed stall warning and any high speed
handling effects (sometimes called the low and high
speed buffet boundaries).
Table 4.2 Gust loading requirements for aeroplanes in civil airworthiness standards
102
standard, and any association interpretative material (such as EASA AMCs or FAA ACs)
4.3 Constructing the Gust Envelope 103
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
VB/Part-25 The minimum value for VB is VS1 2
1 þ 498
Kg UREF VC a
ðW=sÞ , where
W/S is mean wing loading over the aircraft, defined in
lb ft−2.
Part-25 other gust cases Part 25 also contains other more complex gust cases
which must be considered, but are outside the scope of
this essentially general volume. Any engineer working
in this area is likely to be working in a company where
considerable expertise and precedent is already in place,
and this should be consulted, along with the actual
standard, and any association interpretative material
(such as EASA AMCs or FAA ACs).
The problem with this is, hopefully, obvious—a neutral resonance is likely to
cause discomfort, control problems and almost certainly material degradation
through some form of fatigue mechanism. A divergent resonance will almost cer-
tainly cause something to fail catastrophically. The most famous example of this on
an engineering artefact is the Tacoma Narrows bridge—not an aeroplane, but
known to every student of high school physics. Tacoma Narrows illustrates how-
ever, the classic mechanism for aeroplane flutter, which is illustrated in Fig. 4.13.
Behind any shape within a field of fluid flow there will be a disruption to the flow,
which will often take the form of an oscillatory vortex field—the Von Karman
vortex street. That vortex street takes energy from the fluid flow to create itself, but
can then put that energy back into the creating structure in the form of cyclic forces
acting perpendicular to the fluid flow.
Vortex shedding is not the only source of forcing resonance—almost all aircraft
powerplants have the potential to create such motion through an airframe, generally
at frequencies that are a linear factor of the engine speed: so typically 1 per rev, 2
per rev, and so-on.
The objective of the designer must be to try and ensure that no piece of structure
with lightly damped resonant characteristics has the ability to be excited by a
forcing resonance at the same frequency. This is not easy as it requires detailed
knowledge of three things, which are variably complex to estimate:
1. Forced resonance frequencies from the powerplant and other items—this is
generally the easiest to estimate, as engine speeds and behaviours are usually
well understood even at the design stage.
2. Resonant frequencies of the structural mass-spring-damper combinations—these
can be predicted after a fashion using modern finite element methods, but there
will be fairly wide tolerances on the results.
3. Predicting the frequency of generation of the Von Karman vortex street behind
any given piece of structure. This frequency, which will primarily be a function
of TAS, Reynolds Number and shape is extremely difficult to estimate with any
accuracy.
So, whilst considerable effort will go into analysing at the higher value end of
aircraft design: particularly part 25 and higher performance military aircraft, there
will ultimately be no substitute for physical testing work aimed at establishing
resonant characteristics. This will have to cover the full flight envelope, however
experience has shown that particular issues are of greatest potential “interest”: these
include potential elevator and/or pitch trim resonance where this can co-incide with
the longitudinal SPO (see Chap. 10), the trailing edge of any extended surfaces
(such as fixed undercarriages and their fairings, struts, and all moveable control
surfaces), and the mounting of engine ancillaries, and for helicopters take-off and
landing phases as the undercarriage loading is changed with rotor pitch. In both
theory and practice resonance can occur at any engine powers or any airspeeds—in
practice however high engine powers and high airspeeds are most likely to prove
problematic because there is the most excess energy available to create problems.
Prior to any flying, typically any new or modified airframe/engine combination
will be subject to several types of tests, the rigour of which will depend upon the
3
Aerodynamics. Selected topics in the light of their historical development. Theodore von Kármán.
Cornell University Press, Ithaca, New York. Oxford University Press, Oxford, 1954.
106 4 The Flight Envelope
value and novelty of the aircraft. During engine ground running (see Chap. 9)
airframe resonance will be monitored and in most aircraft instrumented then sub-
jected to analysis through an FFT (Fast Fourier Transform) analyser or similar to
identify areas of the airframe where potential for resonance is threatened. A further
common test is to subject protuberances, control surfaces and similar to a “bonk
test” where an automated impulse is imparted to areas of the structure, following
which the subsequent motion is also analysed using an FFT analyser to identify
areas of potential concern.
Subsequently the aircraft must of course be flown. Generally only a small
proportion of a flight test programme will be explicitly towards flutter and vibration
analysis—most likely the aircraft’s behaviour will be observed either by crew or
(from mid-range part 23 aircraft upwards) actively instrumented throughout the
programme—although particular interest and care will be taken during any tests
where the aircraft is exploring into previously unflown regions of the V-N diagram:
in other words as speeds and g loadings are increased—but also as greater altitudes
are reached as for any given EAS value, TAS will be greater with increasing
altitude (see Chap. 4). Commonly aircraft controls will be deliberately subjected to
sharp sinusoidal inputs as the flight envelope is expanded, to try (and hopefully fail)
to excite flutter. This is generally regarded as high risk flight testing.
If flutter or resonance is experienced at any point during the test programme, the
most immediate priority is always to safeguard the aircraft by changing the pow-
erplant and/or flight conditions away from those which appeared to be creating the
issue. Generally, good flight test practice is then to terminate the flight as soon as
possible so as to allow ground investigation and analysis since it is generally
impossible in flight to know what damage has been done by even a single flutter
event. Subsequently there are several options available to the project team by which
resonance can be eliminated; these may include:
• For minor powerplant induced resonances, creating engine “avoid” speed bands.
• For inert structure which has suffered resonance, almost certainly mechanical
redesign to tune-out the resonance or to change the vortex street characteristics.
• Where aircraft control surfaces are affected: changing shapes, internal mass
distribution or moving friction.
• For very high speed characteristics, restricting the aircraft’s maximum operating
speed.
The author’s own experience of resonance illustrate these points. The following are
purely anecdotal, but all from direct experiences and illustrate well the range of
problems and fixes:
• A “beaver tail” baffle between two podded engines on the Vickers VC-10, a
rear-engine/t-tail transport aeroplane that was moving sideways a considerable
amount. This was resolved by adding mass to the trailing edge of the baffle, thus
changing the resonant frequency (see Fig. 4.14).
4.4 Flutter and Resonance 107
• Elevator flutter which co-incided with the longitudinal SPO of the CFM
Shadow, a microlight aeroplane with reversible controls (Fig. 4.15). This was
resolved by introducing additional friction at the elevator, this both increased
damping and moved the resonant frequency.
• Trim tab flutter in a single cable trim tab that worked against a spring. The
spring strength was increased.
• Resonance of cabin skin structure between the wing and fuselage of a prototype
light aeroplane. This was fixed by stiffening the structure internally.
• Resonance around a recess in an external instrument on a large research aircraft.
This was resolved by changing the shape of the recess but only after returning
the instrument to the wind tunnel for re-analysis and significant delays in being
able to use the instruments.
• A carbon fibre propeller which suffered high speed resonances within the CFRP
structure, that caused heating and eventually spontaneous combustion. This was
modified, following extensive ground testing, by changing engine speed and
fuel type limits so as to avoid the conditions which could cause this.
108 4 The Flight Envelope
Use the least conservative permitted requirements of CS.VLA and the data below:
define the manoeuvre envelopes for a conventional tailplane-monoplane configu-
ration fixed wing aeroplane, which has a fixed undercarriage and two-stage flaps
(up and down), down being used for landing only; the aircraft structure can be
considered to be rigid. Ignore gust loadings, and you may assume that stalling
speeds will be identical for equivalent positive and negative loadings. A TPEC
chart for the aeroplane is shown below.
4.5 Sample Problems 109
VD ¼ 130 kn EAS:
Stall speeds: flaps up 50 kn IAS, flaps down 45 kn IAS, tested at 700 kgf.
The aircraft started to display pitch instability, regarded by the flight test team as
unacceptable, but not easily solvable above 120 kn IAS.
Present your results graphically, with the cardinal points labelled and supporting
calculations shown (65%).
Also present a proposed cockpit placard of operating limitations, do not use
abbreviations or acronyms for airspeed limits (20%).
Consider whether this aircraft is appropriate for certification, as currently pre-
sented, against the airworthiness requirements of CS.VLA. Describe any difficulties
(15%) (Fig. 4.17).
110 4 The Flight Envelope
These results are shown on the V-N (flight envelope) Fig. 4.18.
Chapter 5
First Principles of Structural Approval
At that time [1909] the chief engineer was almost always the
chief test pilot as well. That had the fortunate result of
eliminating poor engineering early in aviation.
Igor Sikorsky
5.1 Introduction
Any engineering product will be subjected to loads during its service life, and it is
part of the role of the engineer to predict those loads as accurately (or at-least,
conservatively) as possible, and then to ensure that the structure of that product is
capable of withstanding all of those design loads, throughout the service life,
without failure—unless of course failure under certain conditions was part of that
design requirement.
To do this, the engineer must determine three things:
1. What the maximum loads are that a component should ever see during its
service life.
2. The strength of the item when loaded in the same manner.
3. What is the minimum acceptable margin of safety between the maximum load
that may be applied, and the strength of the item.
Whilst the complexity of this task will depend upon the complexity of the machine
being evaluated—and this task is far from the only one in determine the safety of a
machine, this is without doubt one of the skills which sets the engineer apart from
all other professions. Even if one were to consider perhaps the simplest engineering
structure imaginable—a single plank bridge across a ditch, intuition alone will not
guarantee that the right solution is achieved, and it is necessary to conduct either
calculation, or test (or both) to ensure that the plank is strong enough.
When one approaches the rather more complex problem of a flying machine, the
task becomes extremely complex, and if even the simplest aircraft was analysed to
the extent that would theoretically be possible—the resources to support this would
be nearly infinite. Therefore of necessity practices have built up that allow aircraft
structures to be analysed in a manner which is containable, whilst simultaneously
ensuring an adequate level of structural assurance.
A further complication is that it is rarely acceptable in aircraft work to grossly
over-engineer any structure. This is primarily for reasons of weight—it is essential
that any aircraft weighs no more than absolutely necessary, since every ounce of
excess structure either reduces available payload, or degrades aircraft performance
thus reducing the machine’s ability to do the job for which it was designed.
This chapter will describe the main principles of structural approval of aircraft
structures, the intent being to teach the reader sufficient that they should (if already
a competent engineer) understand all of the main principles in approval of any
aircraft structure. Clearly however, the sheer detail that would go into the approval
of, for example, a modern airliner could not be covered—in practice however very
few engineers have this level of understanding (and if they have, it is from many
decades of experience). This is not usually a problem, because most engineers
working on the design and approval of large complex aeroplanes (whether civil or
military) are only required to understand in detail that small part of the aircraft
which is their own responsibility. However, they must still understand something of
the overall process, and also the mechanisms by which approval is demonstrated.
There are several ways in which structural approval of an aircraft can be carried out,
varying from one based entirely upon test, via one based entirely upon algebraic
analysis, to one based entirely upon computer analysis. In practice none of these are
wholly satisfactory, and in order to present a full picture of the integrity of an
aircraft (or any part of it) almost always at-least two are needed in combination,
usually all three.
The role of the structural airworthiness engineer (or more-often, engineering
team) is thus threefold. They must,
1. Determine the most appropriate means by which the structure can be determined
to be satisfactory.
2. Provide the evidence of this.
3. Provide reports, which can be verified independently, proving both that the
proposed method of proof is satisfactory, and that compliance of that structure
with the requirements is satisfactorily proven.
5.2 The Role of the Structural Airworthiness Engineer 115
Therefore the skills required of an engineer, even if only working in this very
narrow field are also threefold, specifically:
1. Loads prediction
2. Structural test and/or analysis (often referred to as “stressing”)
3. Report writing.
The purpose of structural analysis is to ensure that the aircraft is strong enough, light
enough, stiff enough, for long enough.
John Tempest
The first thing that must be determined concerning a piece of aircraft structure, is
the maximum loads which it is expected to ever see in service. At a “whole-aircraft”
scale this for example includes the standard flight envelope as described in the last
chapter. These loads are referred to as limit loads.
However, the only time when it is acceptable to design (or prove) something
purely to limit loads, is when it is actually designed to fail at that load (for example
part of a load-absorbing “crush structure”); this is extremely rare. In most cases, a
design is intended never to fail. Since it must be accepted that there are uncertainties
in both the prediction of loads, and in the calculated or measured strength of any
component, it is normal to apply a safety factor to the limit load—intended to
ensure this desired non-failure.
The determination of safety factors will in the earliest days of aircraft design
have been arbitrary and subject to the whim and opinion of an aircraft’s Chief
Designer (for those with a historical bent, the inquiry report into the loss of the
R101 airship provides many insights into this).1 Nowadays the minimum values of
safety factor will be defined in a combination of airworthiness standards or com-
pany design handbooks, developed and checked over many years. However,
Table 5.1 shows the values that are most commonly used in civil airworthiness
work (showing the paragraph number where the detail is found in most UK,
European and US standards).
These factors are generally multiplied by each other, so that for example in
considering a composite wing mainspar for a + 3.8 g aeroplane (the Beech Starship
shown in Fig. 5.1 is such an example), the total safety factor was 1.5 1.5 = 2.25.
1
Report of the R101 Inquiry, HM Stationery Office 1931. This is available from various sources,
both free online and purchasable in book form.
116 5 First Principles of Structural Approval
Fig. 5.1 Beechcraft 2000A “Starship” a large primarily composite part 23 aeroplane
So, the total load that the mainplane would be expected to withstand is proportional
to 3.8 2.25 = 8.55 g.
This value, of limit load multiplied by total safety factor, is referred to as the
ultimate load. This does not mean that it is expected to fail at this higher load, but
that it is not expected to fail below this load. The load at which something has been
proven or predicted to fail is the Failure Load.
A further concept is that of Proof Load, which is limit load multiplied by the
proof factor; for most aircraft (and the purposes of this book) the proof factor is 1,
so proof load = limit load, however engineers should be aware that this may not
always be true.
5.3 Concepts and Terminology in Structural Approvals 117
In American airworthiness practice, this will often instead be stated as the Margin
of Safety, where margin of safety = reserve factor—1. Both are only ever quoted to
three significant figures, since greater precision would be misleading, and materials
data is rarely, if ever, accurately available to a greater level of accuracy in any case.
So, the structural reports into any aircraft structure must have the objective of
showing throughout that RF 1. Depending upon the nature of evidence, to what
extent RF must exceed 1, can be subjective and must be agreed between the proposing
and approving engineers. For example, if it is intended to justify the strength of a
complex component by use of a finite element (FE) model alone, then the inevitable
uncertainty over the validity of an FE model (which is usually intended as a design,
rather than a certification tool) must necessitate a larger value—in extreme cases this
might exceed 2.0. This is the reason why in practice FE (or other analytic methods)
are almost invariably supported by test—if the computer model can be verified at a
selection of points (or at the very least, at the most safety critical) then this allows a
value of RF closer to 1.0 to be employed with good confidence. It is possible to
interpret this as subjective, and it often is—except where it is possible to claim clear
precedent (the engineering equivalent of case law), then it is invariably necessary for
all parties to agree to the minimum values of RF that will be accepted for each part of a
structure, approved to what method, before too great an expense has been incurred.
Of course, the search (at-least on the part of the designers) will always be for a
value of RF that is as close to 1.0 as possible, since any value greater than that (or
greater than can be agreed with the relevant authority) inevitably represents excess
weight, which almost certainly degrades the aircraft’s performance in its designed
role.
Note
The terminology used above is that common in civil aircraft approvals in the
United Kingdom. Inevitably other environments may use slightly different
terminology, and readers should always ensure that they understand the ter-
minology used in their own place of study or work—there is no absolutely
correct terminology.
118 5 First Principles of Structural Approval
Figure 5.5 shows the arrangement of part of the shoulder harness attachments for a
proposed military reconnaissance aeroplane, similar in configuration to the OV1C
Mohawk aircraft shown in Fig. 5.4. This consists of three elements (plus joints)
(Fig. 5.3):
• A-B is a 2.5 m long unsupported tube, constructed of 6061-T6 aluminium alloy,
with a diameter of 100 mm, and a wall thickness of 1 mm.
• At B is a bolted joint at the centre of the tube, the failure load of the joint is
known to be identical to that of any attached cable.
120 5 First Principles of Structural Approval
Determine for each component (tube A-B, joint B, cable B-C, and webbing C-D)
the following;
The structural reserve factor
Whether any redesign is required?
What testing is likely to be required to provide adequate structural assurance?
Chapter 6
Approving an Aircraft’s Main Flight
Structure
Abstract This chapter describes the methods by which an aeroplane’s main flight
structure is proven to be fit for purpose, through a combination of analysis and test,
explaining that the use of numerical (finite element) methods is usually unsuitable
as an approval tool. Special cases where approval by test only are described.
Materials fatigue is also considered, describing the philosophy behind achieving
fatigue resistant structures.
We might add several other hints to inventors who desire to enter on this enticing
field, but we will conclude with only one more. The newly discovered metal alu-
minum, from its extraordinary combination of lightness and strength, is the proper
material for flying machines.
Scientific American, 8 September, 1860
The V-N diagram, as repeated in Fig. 6.1, represents the limit loads upon an
aeroplane—that is, the maximum load which (assuming that it is operated within
limits) the aeroplane is expected to see throughout it’s service life.
As has been discussed previously however, safety factors must be applied to the
limit loads upon the flight structure, in order to determine the maximum loads that
the structure must be shown able to withstand without catastrophic failure. These
safety factors will not be constant throughout the structure: no engineering structure
(except perhaps the earlier example of the single-plank bridge on page) has
throughout the same material, same type of structure, and no joints.
It is usual practice to list all of the materials, joint types, etc. through the entire
affected structure, with against them the safety factors applicable: consider
Table 6.1 and the corresponding Fig. 6.2.
VD gust line
6
g A C D
4
VA VC VNE VDF VD
0
Airspeed
-2
E
-4 G
F -VD gust line
-VC gust line
Following this, a loads analysis must be carried out, based upon the worst case
limit loads that will be applied to the structure. In this case (the wing strut), there are
likely to be two separate loads—that associated with the maximum positive normal
acceleration combined with most adverse airspeed, the that associated with the
maximum negative normal acceleration and most adverse airspeed. There might
of-course be further cases (for example, hard-landing or VD) and it is important that
the airworthiness team consider all such possibilities—if only so as to discount
them. So the first part of the statement of loads in this strut assembly might look
something like Table 6.2.
6.2 Approval by Test 125
Fig. 6.2 Illustration of a wing strut assembly (Auster). Note the example and photograph do not
fully correspond, in that the inter-strut on the Auster is a rod rather than a cable. Strutted wings
primarily exist now only on light and vintage aeroplanes, but not exclusively—for example the
dHC-6 Twin Otter, and even the C-130 Hercules modified by Marshalls Aerospace in the late
2000 s to test the TP400-D6 engine for the Airbus 400 M use struts
1. First part, the main strut and jury strut assemblies—with a maximum safety
factor of 1.725
2. Second part, the cast base fitting, safety factor of 3.0
3. Third part, the inter-strut cable, safety factor of 2.0
So, it would be an appropriate practice to test the overall assembly to a common
safety factor (say 1.725, if there is confidence that the struts themselves are to an
extent over-engineered), by analysis determine the higher ultimate loads in the base
fitting and inter-strut cable, then to test those separately.
Thus it is possible to test the aircraft structure as a series of large assemblies—
minimising the number of expensive tests required. This method would be based
upon analysis of loads, but the actual demonstration of acceptability is by test only.
In terms of approval cost, this method can often be the most efficient for some light
and microlight aeroplanes; whilst it relies upon the availability of a sacrificial
airframe (since it is unlikely to be sensible to contemplate subsequently flying
airframe components which have been subjected to ultimate loads), the actual time
to conduct such testing, and to produce the reports, can be relatively low and
present a substantial cost saving over an analytic method.
For example, Fig. 6.3 shows the 6 g (ultimate positive Nz) load test upon the
wing of strutted wing structure for a conventional aeroplane. The part being tested
is the metal wing + strut framework, mounted inverted on a testing jig simulating
the fuselage (on the right hand side of the picture). Via load spreaders (in this case a
mesh of nylon rope) the wing is loaded using sandbags distributed to represent the
aerodynamic load distribution in flight. For some standards a simplified load dis-
tribution will be published as part of the interpretative material for the standard.
However, it may often be preferred to carry out a more meaningful analysis, par-
ticularly when trying to ensure the lightest possible structure.
Apart from the requirements for sacrificial hardware, the main disadvantage of
this method is that there is little incentive to load the aircraft structure to more than a
very small percentage above ultimate load. Therefore the proven RF will normally
be near unity (albeit with a very high degree of confidence in the result)—this gives
the engineering team little idea of the actual structural reserves available, which can
present problems later in an aircraft’s life when it may be desired to increase loads
as part of an envelope expansion exercise.
Returning to the example in Table 6.2 testing is clearly not the only, or necessarily
the best means by which approval may be demonstrated. Engineers have at their
disposal a great many analytic tools by which the strength of any component may
be estimated. Whilst it is important to appreciate that such results can only be
estimates, the best analytic tools will produce estimates of sufficient accuracy that
one may be confident that a properly constructed test would yield results within a
very few percent of the predicted failure load, and almost invariably the predicted
failure mode. This degree of accuracy most certainly applies to the classical
methods of framework stress analysis that are clearly described in many textbooks.
However, many structural analysis methods—particularly those which are
computer intensive or deal with particularly complex shapes, do not fall into this
category. Finite element methods are primarily intended as a design tool, and are
rarely a complete replacement for some combination of traditional “longhand”
analysis, nor for at-least a certain amount of testing.
However, because it is usually cheaper than testing, plus allows airworthiness
approval and structural design work to be integrated, analysis has always been, and
will remain, at the core of structural approval work, so repeating the previous
example, consider how this can be taken forward:
If demonstrating acceptability by analysis, it would be normal next to produce a
series of discrete reports for each component. So, taking the example of part #1 the
forward wing strut, this might look something like the example in Table 6.3
Normal practice would be for a series of such reports to be produced, demon-
strating the acceptability (or sometimes, the unacceptability—in which case iden-
tifying the problem to be solved) for each component/load case combination (in
whatever order suits the structural airworthiness team). The hierarchy of reporting
will then be something like that shown in Fig. 6.4.
6.3 Approval by Analysis 129
Materials data
Analysis
10.3 ×106 N
So, reserve factor RF = = 1,140
9000 N
Further The high reserve factors in this piece of structure may allow
observation significant potential for redesign and resultant weight
reduction if required
Overall aircraft
airworthiness
reports
Standard
Flight test reports,
Structural reports applicability
etc.
reports
Subsidiary structural
Reports for discrete
components
ultimately there will be a large degree of professional judgment required on the part
of the designers, the airworthiness engineering team, and the regulators responsible
for ultimately approving the aircraft. This does make one thing absolutely para-
mount—which is regular clear communication between all of these parties, and
agreement as early as possible within a programme as to where the boundaries
between test-only/analysis-only and analysis + test approval are required. Whilst
regulators will seldom commit absolutely, preferring to reserve the right to change
their mind with results and circumstances, firm statements of intent on all sides
provide assurance and a large reduction in the risk of escalating technical costs.
Fig. 6.5 Typical kitplanes. Clockwise from top-left, Supermarine Mk26 Spitfire Replica, Pitts S1,
Dyn-Aero MCR-01 “Ban-bi”, Air Creation KISS-400, Reality Easy Raider
normally be re-enforced by some limited stress calculations (for example for the
mainspar and wing-attachment points, control cables, and so-on: components which
are both highly safety critical, and relatively easy to determine the dimensions and
materials for) where it is particularly important to have a high degree of confidence
in the results, and knowledge of RF values. Typically the materials cost of such an
exercise is likely to be around one-third that of a flightworthy aeroplane.
Particularly for low cost aircraft such as the X’Air shown in Fig. 6.6 this is often far
more cost-effective than a detailed analysis project. Figure 6.3 was taken from the
approval reports for this aeroplane, and is typical of the level of proof (a combi-
nation of photographic and independent oversight) that is usually provided to the
relevant authority1.
The same process has often found favour also with amateur designers of one-off
aeroplanes (who often do not have the capability to engage in any form of math-
ematical analysis), or of the commercial designers and manufacturers of lower cost
1
The world’s most mature microlight airworthiness system is that of the British Microlight Aircraft
Association. Numerous guidance and explanatory documents and aircraft type data sheets can be
freely downloaded from the BMAA’s website at www.bmaa.org.
134 6 Approving an Aircraft’s Main Flight Structure
2
If the reader is unfamiliar with these concepts, a good introduction is in JE Gordon’s The New
Science of Strong Materials, ISBN: 978-0140135978
6.5 Materials Fatigue 135
(d) (e)
(d) (e)
By comparison, Fig. 6.8 shows better practices: at (a) both internal and external
corners have been chamfered off, at (b) the five small fastener holes have been
enlarged and reduced in number reducing peak stresses around the holes as well as
within the now larger fasteners, (c) the V-shaped notch has been curved, (d) the
bolted joint is redesigned so that it is now in double shear (halving shear stresses),
with the bolt replaced with one where only the nut is on the threaded section and an
unthreaded plainshank carries all the shear loads, and at (e) the rectangular cutout
has had all of its corners rounded off, reducing peak stresses at the corners. (This
last practice has become near universal in aircraft windows, particularly since the
advent of pressurised hulls that create cyclic hoop stresses—this contributed to the
losses of the de Havilland Comet in 1954 before this issue was adequately
understood.)
136 6 Approving an Aircraft’s Main Flight Structure
Fig. 6.9 An expert operator carrying out an internal boroscope inspection of a turbofan turbine
section
Assuming that best practices are followed in this manner, the next design feature
which must be considered is inspectability—every component of the aircraft which
is potentially vulnerable to metal fatigue must be accessible so that it can be
inspected on a cycle which permits capture and rectification of fatigue crack growth
before it can endanger the aircraft. Inspection, depending upon the nature of the
structure may be visual, or use technologies such as X-ray inspection, boroscoping
(Fig. 6.9) die-penetrant inspection, or magnetic eddy inspection. Particularly with
complex or cast components cracking may be internal and not findable visually,
even with disassembly, requiring these more advanced inspection technologies. It is
important that the design enables this—or example avoiding large assemblies that
don’t permit either internal access or disassembly for inspection, and a particularly
common fault in smaller aeroplanes, avoiding flexible surface coatings which
would conceal cracks. Similarly, it’s also very important that maintenance and
inspection schedules ensure that every critical part of an aircraft is inspected with
sufficient regularity that cracks will be captured before they become problematic.
Or part 25 aircraft in particular, it is common to build a structural test airframe
and test this through a long series of simulated structural cycles—leading
demonstrating and solving any early-life fatigue problems before an aircraft is first
flown, then continuing to several times the life of any in-service airframes with
6.5 Materials Fatigue 137
fatigue being continuously monitored so that where problems arise (and they
usually will) design solutions can be found and incorporated well before they arise
on a flying aircraft. Where this approach has identified discrete components una-
menable to redesign but where fatigue failure is highly likely at a known life, then
typically component replacement will be mandated at the predicted failure life3
divided by a pre-determined safety factor, typically in the order of 2–3. In some
components there may also be a determined calendar life—most likely where
components are manufactured from composite materials that are known to degrade
with time rather than cycles of use. The same principle applies—but if, for example
the minimum failure life of the most critical parts of an undercarriage might be
determined as 12,000 landings or 12 years, then with a factor of 3 applied: removal
and replacement or overhaul might be mandated at 4,000 landings or 4 years
whichever occurs first.
This section has concentrated upon metal fatigue—and this does remain the most
significant fatigue life issue in aviation. However, the potential for composite
material fatigue should not be discounted. The mechanism however is quite dif-
ferent the crack growth mechanism which applies to metals—typically it involves
one or more of microcracking of the composite material’s resin component, partial
disbanding of the matrix from the resin, or changes in material properties due to
moisture or other chemical exposure, or even biological intrusion into the material
following moisture ingress. Non-visible impact damage can precipitate this, as may
in-service loads, or exposure to solar radiation. It is therefore much harder to
generalise about composite fatigue than it is about metal fatigue, and the engineer
determining the strategies for continued airworthiness monitoring of composite
structure, especially primary structure must first familiarise themselves well with
the characteristics of the specific materials. Specialist equipment may need to be
procured or devised to allow monitoring of composite material fatigue, and also
dedicated testing of material samples to accelerate ageing may also be necessary as
part of the design process, so that its fatigue characteristics can be adequately
understood.
3
A good source of guidance on fatigue lifeing of components may be found in NASA Technical
Memorandum 86812 “Prediction of service life of aircraft structural components using the
half-cycle method”, published in May 1987.
138 6 Approving an Aircraft’s Main Flight Structure
Particularly with the recent primacy of composite aircraft, the concept of damage
tolerance is increasingly important. Related somewhat to fatigue mechanisms,
although also in composites to for example barely visible impact damage, this
makes assumptions about the level of structural damage which may have occurred
without being detected and corrected—and then re-assesses the structural strength
of the aircraft.
This is clearly difficult to do, and likely to require both a review of maintenance
records, and significant engagement with experienced maintenance engineers—in
both cases related to similar designs and structures. Done properly, this is likely to
lead to both analysis, and testing work being repeated with damage pre-inserted. It
is likely to be a requirement that the same structural requirements are met with this
damage pre-inserted, although in some instances, slightly reduced safety factors
may be acceptable by mutual agreement between the design, airworthiness and
authority teams. Clearly the specific damage being introduced must be subject to
expert judgment, and must be an informed guess—this can’t be an exact science.
Erring on the side of pessimism in this regard is obviously good practice, but
excessive pessimism will obviously result in the design of an excessively heavy
structure.
This practice is not normally required for the lightest classes of aeroplane—
microlights and part VLA aeroplanes, but starts to become necessary to achieve the
required safety standards at part 23, and is essential at part 25. Military procurers
may not consider this necessary for training aircraft, but it will be essential for all
other classes of aircraft, given their inevitable potential to attract battle damage—in
all likelihood a more critical design requirement than that of fatigue or “hangar
rash” damage to any equivalent civil aircraft. Defence standards4 will contain
guidance on best practice here, but expert opinions in Official Test Centres are
likely to provide the most important guidance.
4
For example AGARD (NATO Advisory Group for Aerospace Research and Development)
AGARDograph 238 “Design Manual for Impact Damage Tolerant Aircraft Structure”
Chapter 7
Undercarriage Structural Approvals
Abstract The undercarriage (or landing gear) of an aircraft is the most important
structure in ensuring safe landing (and take off) of an aircraft. As such, it is subject
to intense stresses at take-off, landing, and indeed, in-flight, and in order to perform
correctly, it must be able to survive these stresses repeatedly. This chapter discusses
in detail the experimental procedures used to determine the response of an aircraft
undercarriage to the various stresses it must endure, in order to determine the initial
(and continued) airworthiness of any given design of undercarriage. Analytical
methods are also briefly discussed, but these tend to be simulations of the experi-
mental tests that this chapter covers in detail.
The undercarriage (or landing gear) is a structure which has three functions. Firstly
it must withstand the loads that are imposed upon it by an aeroplane which is
parked, taxiing, or subject to other ground maneuvering and environmental loads.
Secondly, it must withstand any aerodynamic and inertial loads imposed upon it
during flight. Thirdly, it must withstand the loads that are imposed upon it by
take-offs and landings.
The third of these cases is almost invariably the critical one, and during take-off
or landing, the worst case loads that are experienced are high transitory loads. This
means that ultimately, the undercarriage is primarily a shock-absorber—although
these shock loads can be applied in several orientations and are usually not purely
vertical.
The most common undercarriage configuration that is likely to be met is the
“tricycle”, once known in Cessna’s advertising literature as “Land-O-Matic”
(Fig. 7.1).
Fig. 7.2 Boeing 737-7H4 transport aeroplane, showing a typical tricycle landing gear
arrangement
Fig. 7.3 Airbus A340—a very long aeroplane for which landing and take-off attitudes are very
critical
and the loads consequently much higher; plus, in most aeroplanes with piston-prop
or turboprop installations, the risk of engine damage becomes substantially greater).
A typical example of such an aeroplane is shown in Fig. 7.4; this is the famous
Douglas Dakota, still used worldwide in some specialist applications—although in
reality virtually all new taildraggers now are small single or 2-seat sports aircraft.
Virtually all aeroplanes that may be met will fall into one of these two cate-
gories, but not necessarily all. Many gliders and a few specialist military aircraft
such as the Harrier family (Fig. 7.5) use a monowheel or bicycle undercarriage—
with one or two wheels on the centreline, plus small skids on the wingtips. This is
an interesting case in itself, because although structural considerations for such a
main undercarriage are very straightforward, the loads (in particular drag loads) on
142 7 Undercarriage Structural Approvals
Fig. 7.5 Hawker-Siddeley Sea Harrier, a specialist military aircraft with an unusual undercarriage
configuration
skids at the tips of very long wings can be extremely high and significantly exceed
the loads experienced in flight. This is too specific and specialised design case to
cover in detail in this book, but detailed and helpful guidance can be found in
CS.22, the European certification specification for gliders and motorgliders.
Other configurations may include skis (for operation from ice and snow), floats
(for operation from water), skids (particularly on older aeroplanes—especially for
the tail).
7.2 Determining Undercarriage Energy Absorption 143
The first and oldest method for evaluating the energy absorption characteristics of
an undercarriage is to carry out a drop test: this is still done to some extent on
almost all certification programmes at any scale. In simplistic terms, this means to
take an aeroplane and drop it from a height that has been determined to represent
the limit case within normal landing conditions—that is the worst case hard landing
that the aeroplane should experience during its normally handled service life. In
practice, this means:
• Inertially the aeroplane (or more sensibly and cheaply, a test rig to which the
undercarriage is attached) must be representative of the flying aeroplane. This
includes wheel geometry, tyre type and inflation, and either mass and flying
surfaces being fully representative, or if flying surfaces are omitted—mass must
be adjusted to reflect this.
• The aeroplane must be suspended at a height (of the bottom of the uncom-
pressed undercarriage) above the impact surface which has been calculated
based upon mass and wing area.
• The impact surface must be such that it conservatively represents a realistic
landing surface. Experience has shown that an appropriate test surface is greased
steel plates—which eliminate the “scrubbing” effect (or sidewards friction) that
can otherwise cause results that are unrepresentative of an actual landing.
• The aircraft (or test rig) must be fitted with one or more accelerometers whose
motion compared to the point where the undercarriage is attached to the fuselage
is extremely well damped. Accelerometers may be electronic or mechanical (a
“g-meter”), although in practice high rate electronic instrumentation generally
144 7 Undercarriage Structural Approvals
their high speed and wing-loading), transport aircraft designed for landing on
unprepared strips, and carrier borne aircraft however are inevitably special cases
and it is almost certain that any engineer dealing with such an aircraft will need to
prepare a special case, with considerable analysis (supporting by experimental or
historical results) in order to determine appropriate drop heights.]
The drop height, once determined by the above formula is not generally subject
to change. However, there may be good reasons not to test the undercarriage on a
complete aircraft—the aircraft may not yet exist (or not be available to the
undercarriage team), there may be concerns about the wing structure meeting these
requirements; or, more prosaically, there may simply be insufficient space in the test
facility. In this case, it is permissible to reduce the mass of the test item, by making
assumptions as to the amount of lift which would have applied during the test.
Again, a standard relationship is quoted in most commonly used civil airworthiness
standards, which is given in Eq. 7.2.
Standard Relationship for Reduced Drop Mass
½h þ ð1 LÞd
Me ¼ M : ð7:2Þ
hþd
where
Fig. 7.6 Prepared drop test of a light aircraft undercarriage at reduced weight onto greased steel
plates using a test rig. Courtesy of John Tempest, de Havilland Support Ltd.
Because of the cost, difficulty, and sheer risk to equipment and personnel that
comes with a drop test, it is usually desirable to minimise the amount of drop
testing. This is only possible where it can be shown that no components in the
undercarriage will have failure or deflection characteristics that are significantly
affected by the rate of load onset. This mostly applies to simple composite or metal
spring undercarriages on light aeroplanes, such as that of the PA38 aeroplane
shown in Fig. 7.7, but would almost certainly not apply to an undercarriage with a
fluid-filled oleo such as the very large aircraft shown in Fig. 7.8.
However, if drop testing can be avoided (and often even if it is required and as
well), an alternative method is required by which in particular PZ.max can be
determined—and that is achieved by conducting load versus deflection testing of
either components of or, more likely, the entire undercarriage.
In order to do this, the undercarriage must be arranged—either on the aeroplane
or a test rig, generally on greased steel plates similar to those described for drop
testing, and with the total load upon the undercarriage (including its own weight)
being known. The undercarriage is then loaded, generally up to about three times
(and certainly to at-least twice) the MAUW of the aeroplane for which the
undercarriage is designed. As the aircraft is loaded, the deflection is measured
between the impact surface (the ground), and the point where the undercarriage is
attached to the aircraft. So, a table of values may be built up of load versus
deflection for the whole undercarriage.
7.2 Determining Undercarriage Energy Absorption 147
Fig. 7.7 Piper PA38 showing steel spring main undercarriage structure
Fig. 7.8 Front of Boeing 747-400 and Airbus A380-800, showing compressible oleo noselegs
148 7 Undercarriage Structural Approvals
This data may then be plotted as a polar of load versus vertical deflection, such
as that shown in Fig. 7.9.
Using this information, we know from standard results that work done ¼
energy ¼ force * distance, hence:
Z
E ¼ F ds: ð7:3Þ
Fig. 7.9 Load versus deflection polar for a typical light aircraft undercarriage (single maingear leg)
7.2 Determining Undercarriage Energy Absorption 149
Fig. 7.10 Energy versus force for half of the main undercarriage described in Fig. 7.9
Re-plotting this against the original relationship for force versus displacement,
one may obtain a plot of Energy stored versus applied vertical force. This is shown
in Fig. 7.10.
Returning to the previous result, quoted from various airworthiness standards in
Eq. 7.1, which was that:
12
Mg
h ¼ 0:0132
S
This last relationship may be cross-referenced for the whole aircraft (in this case,
to a curve of FTOTAL versus 2E based upon Fig. 7.10 allowing for two maingear
legs but disregarding the effect of the nose or tailwheel), giving eventually a force
PZ.max which corresponds to the equivalent drop case, if the undercarriage’s
deflection characteristics were not substantially a function of load onset rate.
Where both the above methods were used (this is usually best practice), it is then
normal to compare the PZ.max values derived from each method. If there is not good
(perhaps 10%) correlation between the two then this must be explained—it is
most likely to be either due to rate-dependent characteristics, or to poor quality
experimental methods.
Finally the most conservative (i.e. highest) value of PZ.max from these tests is
taken as applicable to the aircraft under investigation. If this is low (e.g. less than
2.67 W) then it should be confirmed that this is acceptable to the authority and/or
airworthiness standard being used, since generally this is the minimum value that
may be used. This value: PZ.max(a/c) is now known as the Aircraft Load Reaction.
However, it is possible now to assume that some lift (usually up to 2/3 W) may
be considered to act during a landing. This may be subtracted from PZ.max(a/c) to
give a new value PZ.max(u/c) which is called the Undercarriage Load Reaction. This
may also be written as NZ.max, defined by PZ.max(u/c)/W, which is the Undercarriage
load reaction factor and typically has a value in the range 2.0–3.0. Some airwor-
thiness standards may list maximum or minimum values it may take.
Once a working value of PZ.max has been determined, this is then the basis for a
series of load cases which the undercarriage must be demonstrated by either test or
analysis to be able to withstand. Typical load cases differ for mainwheels, nose-
wheels and tailwheels, but are described below.
greater than 1.0) acting vertically, combined with a proportion of PZ.max acting
either forwards or rearwards. The most common version of the maximum vertical
reaction case is that shown in Fig. 7.11.
As is illustrated here, these two forces may be combined (for either test or analysis
purposes) into a single combined force, represented above by a red arrow. Applying
Pythagoras’ theorem, this will have the total force value of (vertical compo-
nent2 þ horizontal component2)½. For the case shown, that equates to 1.044PZ.max.
This force will act (variously before and after) from the vertical at an angle of
tan−1 (horizontal component/vertical component). For the case shown, that equates
to 16.7° either side of the vertical.
Spin Up The spin-up case represents an aircraft’s initial contact with the ground
after landing, during which very high initial drag loads are reacted at the tyres and
axle, in addition to which, there is inevitably a substantial vertical load (Fig. 7.12).
By a similar analysis to that shown above for the maximum vertical reaction
case:
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Total Force ¼ PZ:max 0:52 þ 0:62 ¼ 0:781PZ:max : ð7:7Þ
Sideload The sideload case refers to the crosswind landing condition—rarely does
an aeroplane experience winds that are exactly aligned with the runway during
landing or take-off, and on occasion the crosswind component can be extremely
high. Whilst all pilots prefer to avoid crosswind conditions where possible (and
when experienced, to compensate for them with the careful use of crosswind
landing techniques), it is inevitable that any aeroplane will experience, on occasion,
large sideloads at the undercarriages—either in spite of, or sometimes as a direct
result of, the handling methods used to ensure a safe landing. This is particularly
true when flying a “wing-down” crosswind landing, which requires the pilot to land
with the aircraft banked into the crosswind, held straight with rudder. This means
that the full landing load will initially be reacted at a single mainwheel, with often a
substantial lateral component, as seen in aircraft axes (Fig. 7.14).
A typical load case is illustrated in Fig. 7.15; unlike other cases which are based
upon PZ.max, this case is conventionally based upon W, the aircraft’s [maximum]
landing weight.
This case is combined: 0.5 W inboard combined with 0.67 W vertically on one
side, plus 0.33 W outboard with 0.67 W vertically on the other. Resolving into a
single force vector as previously, this would be:
Side 1: 0.836 W acting at an angle of 36.7° inboard of the airframe vertical,
combined with:
7.3 Typical Undercarriage Load Cases 153
Fig. 7.14 The pilot of this Aeronca K is landing left wing down to compensate for a slight
crosswind from that side—leading to a significant asymmetry in maingear loads during the
landing. A quick look on the internet will show similar examples at every scale of aeroplane
A large number of load cases have been described above. For a tailwheel aircraft,
these may generally be regarded as the sole set of cases to be considered—generally
with the aeroplane assumed to be in the flying attitude, rather than the 3-point
attitude. This may be considered to be because severe landing conditions in normal
taildragger operations generally correspond to the cases where either through
mishandling the aircraft is significantly more nose-down than the preferred 3-point
attitude (the normal objective when landing such an aeroplane is to land with all
three wheels touching down simultaneously), or where the pilot has decided (most
likely due to severe turbulence or crosswinds) to land with increased flying speed
closer, initially, to a flying attitude (this is known to pilots as a “2-wheel” or
“wheeler” landing).
It is possible and normal however to land a nosewheel aeroplane at a wide range
of attitudes, varying from the most nose-down possible where the nosewheel is just
clear of the ground (Fig. 7.16a) to the most nose-up where the tail is just clear of the
ground (Fig. 7.16b).
Therefore for the majority of tricycle undercarriage configured aeroplanes, it
becomes necessary to consider at least two cases for each of the maingear load
conditions described—the most nose-up and the most nose-down—with the actual
load cases being defined in ground axes, then alignment modified into aircraft axes
by adding the aircraft’s pitch attitude in each condition. It is possible that more
cases than these two extremes might require assessment, but in practice this is
highly unlikely; this would be probably considered excessive for the approval of a
light aircraft: which might experience such a range of attitudes, but unnecessary for
most airliners: whose length and field performance requirements dictate a very
narrow range of landing attitudes.
Fig. 7.16 Illustrating the range of pitch attitudes for a landing tricycle configuration aeroplane
7.3 Typical Undercarriage Load Cases 155
the aeroplane, and to assume that the aeroplane is free to rotate in all three axes as a
result of any initial load applied at the maingear. However, such an analysis is
complex, expensive of time and (most likely) computing power, as well as being
very difficult to emulate during a test.
A more conventional (as well as generally conservative and simple to consider)
approach is to pick a part of the aircraft which is sufficiently distant from the
undercarriage (for example, the wing attachments) and to assume that all loads are
ultimately reacted at that point or points. So long as a point is picked which allows
all important internal aircraft structure to be considered, then this method is con-
servative and straightforward to analyse. Further, it should be straightforward to
duplicate test and analytic conditions, should this be considered necessary.
In considering these cases, the conditions described above, must be remembered
to be limit load cases.
There are three load cases which are normally considered for noselegs (this assumes
that, as is generally the case for all aeroplanes, the noseleg is not fitted with brakes).
These are:
Aft Load This is the normal landing load, and comprises a large vertical load
combined with a smaller drag load—both determined normally by reference to the
maximum static load, WN. The most common load case is shown in Fig. 7.17.
Applying Pythagoras’ theorem and basic trigonometry again, as with the
mainwheel cases, this typical case would therefore equate to a total force of 2.39 WT
at an angle of 19.6° behind the vertical.
156 7 Undercarriage Structural Approvals
There are two load cases which are normally considered for tailwheels or tailskids;
these must be met but it is also normal practice to design the structure between the
tailwheel and the main fuselage so that it is at some point weaker than the main
aircraft structure. This is in order that, should a particularly severe load be met, the
tailwheel assembly removes itself from the aircraft rather than potentially removes a
substantial part of the rear of the aeroplane. Clearly neither is a desirable outcome,
but in general it has been found that the consequences of the rear of an aircraft
“skidding” along a runway following a tailwheel loss is not life threatening, should
also allow some directional control via the remaining rudder structure, and is
usually easily repairable.
Obstruction Load This is the normal landing and taxiing load, and is based upon
the static load at the most adverse weight and CG combination for the aeroplane—
which may be termed WT. This is illustrated in Fig. 7.20, with typical load values.
This load case is far simpler than those previously described—the static vertical
reaction load WT is considered to act at an angle of 45° behind the vertical.
Side Load This represents two possibilities. The first is that an aeroplane lands
with the tailwheel not aligned with the direction of travel—either because of rudder
deflection (and a rudder linked to the tailwheel), or because an aeroplane is landed
with some residual “crab”, or angle between the aircraft heading, and aircraft’s
direction of travel. The second (rarer, but more severe case) is that known as a
ground loop which occurs when an aeroplane (almost invariably a taildragger1)
loses directional control only on the ground, usually shortly after landing, resulting
in a yawing departure often causing the aircraft to yaw 180° or more before coming
to rest. This puts very high combined vertical and sideloads on the tailwheel which
must be reacted without damaging the aircraft.
This is illustrated in Fig. 7.21 with typical load values. Please note that although
this diagram shows the tailwheel aligned with the fuselage—as might be the case if
(for example) the aircraft has a tailwheel lock (common on larger taildraggers).
However, if the tailwheel is castoring (as is often the case with smaller, or more
1
Much is written on the handling of tailwheel configured aeroplanes. Two well regarded sources
for those wishing to understand the topic are Conventional Gear: Flying a Taildragger by David
Robson (ASA 2001) and Stick and Rudder: An Explanation of the Art of Flying by Wolfgang
Langewiesche (McGraw-Hill 1944). The latter is particularly widely read by pilots as a seminal
text on the fundamentals of aeroplane handling.
158 7 Undercarriage Structural Approvals
recent designs) then the load case will assume that the wheel is perpendicular to the
aircraft’s longitudinal axis.
As with nosewheel sideload cases, whether this becomes a single case (as shown
above) or a pair of opposing load cases will depend again upon whether the tail-
wheel assembly is symmetric, or asymmetric.
In any case, this (typical) tailwheel sideload case resolves, similarly to previous
examples, to a total force of 1.41WT acting at 45° to either side of the vertical.
determining PZ.max from this drop test (and probably also from vertical load versus
deflection tests). This would then be proven (by some combination of test and
analysis) to meet the airworthiness requirements at both limit and at ultimate loads.
So, if we take the overall safety factor to be applied to the undercarriage to be S.F.
then the actual load cases would be as shown in Fig. 7.22.
It is however a reasonable approximation that the force applied to the under-
carriage during the landing is proportional to the amount of energy to be dissipated,
which in turn is proportional to the height from which the drop test is conducted.
Therefore an alternative is to disregard the static load case entirely, and consider
demonstrating the acceptability of an undercarriage entirely by drop testing—this is
a purely experimental method, largely disregarding any analytical requirements.
In order to do this, a surface (probably greased flat heavy gauge steel as with
previous tests) must be constructed under each wheel assembly whose surface will be
perpendicular to the angle at which the resultant force would normally be considered
to act during the static load test. Then for each test case two drop tests should be
carried out upon a representative undercarriage and associated structure. The first is
from the limit drop height (as shown in Eq. 7.1), after which inspection should show
no permanent deformation should have resulted. The second is from a new ultimate
drop height, being defined as h′ ¼ h.f(S.F.); although unless agreed with an over-
seeing authority otherwise, it’s likely that f(S.F.) ¼ S.F.—this is a very severe case
likely to be highly expensive of equipment to demonstrate, but does at-least require a
minimum of calculation and can, with good test facilities, be done quickly. It may be
particularly favourable for demonstrating the acceptability of lightweight but com-
plex undercarriage designs whose structures are difficult to analyse.
However, it is important to emphasise that this approach is not universally
favoured by airworthiness authorities, and also that the potential for damage to
persons conducting tests is significant (dropping several tonnes of aeroplane, from
potentially over half a meter is a very serious undertaking). This approach therefore,
The majority of aeroplanes are fitted with wheel brakes, usually fitted only to the
mainwheels. This creates an additional load case, whose characteristics will be
based upon a combination of maximum static loads upon the mainwheels (plus, in
the rare event of a design that incorporates nosewheel brakes) with a rolling
resistance determined either by detailed test or analysis of the braking system (plus
the surface friction characteristics acting between tyres and the worst-case, or (more
commonly) by use of a conservative default value for braking efficiency.
Conventionally, braking resistance is represent by a braking coefficient, l
(sometimes written lB ), where the drag due to braking is represented by FB ¼ lWN ,
(WN representing the proportion of aircraft weight, acting normal to the runway
surface, at the portion of the undercarriage under consideration). Realistic values of
l are represented by below, however the majority of airworthiness standards use an
extremely conservative single value around l ¼ 0:8 (Table 7.1).
One may question why it is considered justified to use a value for structural
investigations of braked undercarriages which is double that which would appear
likely to ever be met during actual conditions. In reality it may be seen, comparing
even the conservative braked roll case, to a spin-up case, that this is usually a less
conservative value than that of the spin-up. The use of a high value therefore will
Table 7.1 Typical actual braking coefficient values for different runway surfaces (from ICAO
Annex 14, Volume 1, 2004)
Type of surface Minimum value Maximum value
of l of l
Hard dry runway 0.4 0.4
Damp runway (max. 0.25 mm/0.01″ water) 0.347 0.374
Very light snow 0.334 0.347
Wet concrete (up to 0.75 mm/0.03″ water) 0.294 0.334
Wet tarmac (ashphalt) (up to 0.75 mm/0.03″ water) 0.268 0.347
Gritted compacted snow or ice 0.294 0.321
Compacted snow below −15 °C (5 °F) 0.262 0.268
Heavy rain (0.75–2.5 mm/0.03″–0.1″ water) 0.254 0.268
Snow covered compacted snow below −15 °C (5 °F) 0.241 0.254
Cold ice below −10 °C (14 °F) 0.215 0.241
Wet ice above 0 °C (32 °F) 0.201 0.215
Hydroplaning on standing water above 2.5 mm 0.201 0.201
(0.1″) deep
7.5 Braking Systems 161
not have any significant design impact upon the main undercarriage structure itself,
and will only have an impact upon the design of the internal brake mechanisms and
provides for peak loads or jamming.
Note: Hydroplaning
Hydroplaning is the term used to describe what happens if viscosity causes
water to form a form of laminar boundary layer between a moving tyre and
the runway surface. The phenomenon is most likely to occur where there is an
otherwise clean runway surface contaminated with small quantities of fine
particles (such as pieces of rubber from tyre wear) and recent or standing
water typically at-least 2.5 mm (0.1″) deep. When conditions are suitable, the
mechanism of hydroplaning occurs at or above the hydroplaning speed, a
pffiffiffiffiffiffi
reasonable approximation to which can be given by VHP ¼ 9 PT where VHP
is the hydroplaning speed, expressed as groundspeed in knots, and PT is the
tyre pressure in bar. By inspection this indicates that the hydroplaning speed
(reducing the risk of hydroplaning) is increased by increasing aircraft weight,
that increased density altitude will increase the true airspeed for a given
landing speed (which is always defined in CAS) and thus increase the like-
lihood of hydroplaning, whilst increased headwind will decrease the
groundspeed for a given CAS groundspeed at touchdown, and thus reduce the
likelihood of hydroplaning.
Note that the friction values in Table 7.1 are indicative only, and in particular
specific local measurements are often needed to determine braking efficiency for
aircraft performance purposes. Most authorities will define braking coefficients of 0.4
or above as “good” braking, and 0.25 or below as “poor” braking for aircraft per-
formance and controllability purposes with various intermediate stages. A catch-all
term for runway surfaces with degraded braking performance is “contaminated”.
Note: access to the AMC section of CS.VLA, as well as the main body of
subparts C and D will be needed in solving this problem.
(Note: this was set as an exam question for level III Aerospace Engineering at
Brunel University in May 2006)
You have been asked to investigate the acceptability of a new composite leafspring
maingear for a light aeroplane, with MAUM 1325 lb, and a wing area of 185 ft2.
The undercarriage comprises a single part, attached to the underside of the aero-
plane fuselage. You have available to you drop test data which showed the peak
normal acceleration values given in Fig. 7.23 in a full scale drop test, at MAUW,
with all aerodynamic surfaces fitted:
a. What are the (level landing condition) limit maximum vertical reaction and
spring back load cases to be proven for this maingear assembly? Show all
working. [60%]
164 7 Undercarriage Structural Approvals
b. This undercarriage consists mostly of the composite leaf-spring, all testing for
which has been (and will be) done at room-temperate conditions. However the
attachments to the fuselage are machined chrome-molybdenum steel compo-
nents. What are the ultimate load cases for each part (when considered as part of
the whole undercarriage)? For how long must these components withstand
ultimate loads without catastrophic failure? [20%]
c. The wheels on this maingear will be fitted with hydraulic drum brakes which
have been shown to display a peak braking torque that equates to a friction
coefficient of 0.7 with the ground. What additional load case(s) must be proven
to demonstrate that the main undercarriage structure will withstand this braking
load? [20%]
Explanation of test conditions, etc. are shown in the book and can be derived in
accordance with CS.VLA, but please note the following:
• Requirement for composite superfactor dues to all-composite aircraft.
• The presence of a hydraulic shock absorber means that drop testing is essential,
load-deflection testing along would be insufficient.
• The 1.8 g peak normal acceleration needs to be increased to provide Nz.max in
accordance with the minima of the airworthiness standard.
• Moment calculations should be done about the mainwheel axle (which
co-incides with the to determine the worst case load at the nosewheel, for
purposes of determining the nosegear test loads.
• Whilst the airworthiness standard is phrased in SI units, and thus units need
converting to SI for purposes of determining drop height, all other calculations
can reasonably be done in imperial units. It recommended that students are
encouraged to do so, simply to give them some practice.
lb/2.202 kg/lb = 602 kg. So, Mg/S = (602 * 9.81)/17.2 = 343 N/m2. This results in a
drop height of 0.244 m.
Referring to the graph shown in Fig. 7.23, this relates to a limit value of
N.Z.max = 2.6 g.
This value is below the minimum permitted by the standard, therefore
N.Z.max = 2.67 g.
The most conservative value of ground reaction will be used; hence a wing lift of
0.67 W will be used—making the ground reaction load of N.Z.max = 2.0 g.
This value may now be multiplied by the aircraft weight to give a load of
Pzmax = 1203.4 kgf, (or 2650 lbf, or 11,801 N) distributed across maingear.
Based upon this value, and combining with the constituted load guidance in
AMC.VLA.479(b).c, the maximum vertical reaction load case may be resolved to
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1.044 Pz.max (that is 12 þ 0:32 ) at 16.7° (that is tan1 0:3
1 ) either side of vertical.
This equates to a load of 12,320 N at 16.7° either side of the vertical.
Further considering the spring back load, this is defined in AMC.VLA.479(b).a,
and may be similarly resolved to 0.89 Pz.max at 32° behind the vertical. This equates
to a load of 10,503 N at 32° behind the vertical.
b. The composite leafspring undercarriage is subject to the composite superfactor
as defined in AMC.VLA.619, which is an additional factor of 1.5. In addition,
the normal safety factor of 1.5 as defined in CS.303 must be applied. Therefore a
total factor of 2.25 must be applied to above limit load values to determine
ultimate values.
– The maximum vertical ultimate load is therefore 27,720 N at 16.7° either
side of the vertical, when considering the composite leafspring.
– The spring-back vertical ultimate load is therefore 23,632 N at 32° behind
the vertical, when considering the composite leafspring.
The metal fittings however are only subject to the normal CS.303 load factor
only.
• The maximum vertical ultimate load is therefore 18,480 N at 16.7° either side of
the vertical, when considering the metal fittings.
• The spring-back vertical ultimate load is therefore 15,755 N at 32° behind the
vertical, when considering the metal fittings.
The ultimate loads must withstood for 3 s without catastrophic failure, as
required by CS.VLA.305(b).
c. The loads applied to the undercarriage structure as a result of braked roll, equate
to 1:33 lW rearwards, combined with 1.33 W upwards ¼ 1.5 W at 27.7° behind
the vertical. [CS.VLA.493 refers.]
166 7 Undercarriage Structural Approvals
Drawing this out (see sketch in Fig. 7.24), it can be simply seen that this case
lies within the requirements of the spring back case. Therefore there should be no
requirement to re-prove the maingear structure, although there will clearly still be a
requirement to prove the brake structures themselves (Fig. 7.25).
Chapter 8
Control Surfaces and Circuits
Abstract Control systems and circuits are the mechanisms through which aircraft
are controlled. In order for a system to be considered airworthy, control inceptors
generally follow a standardized layout and must follow defined structural standards.
Circuits must not be prone to jamming under “normal” operational conditions, and
should, generally, be “intuitive” to a pilot. This chapter discusses the recommended
standards and layouts, together with common problems and pitfalls.
8.1 Introduction
Any aircraft must be controllable, in order to fulfil its mission objectives of taking
off, landing, and in between maintaining such airspeed/altitude/manoeuvre condi-
tions as are needed to meet its mission objectives. This control is done by one or
more of three mechanisms:
1. Moveable control surfaces or changing airframe shape within the surrounding
airflow.
2. Changing thrust by control of the powerplant (fuel flow, propeller pitch, mix-
ture, nozzle angle, etc.)
3. Moving the aeroplane’s centre of gravity.
All aeroplanes use at-least one of these mechanisms, most two and a few all three.
The most common however, that applies to the majority of aeroplanes is the use of
moveable control surfaces. The most common configuration of these control sur-
faces is the combination of ailerons, a rudder, and an elevator (with or without
Figure 8.5 shows a typical aeroplane cockpit; as it happens this is the cockpit of a
business jet built in the 1960s, but its layout would be instantly recognisable to any
aeroplane pilot of 1930, 1990 or now. The primary inceptors (cockpit controls)
have a layout which has been (mostly) standardised since powered flight was less
than 25 years old. It is clearly designed around a human body (or here, as is very
often the case, two similar human bodies, seated side by side).
Table 8.1 shows values for cockpit loads which have been shown through expe-
rience to give sensible maximum (i.e. limit) values for the loads applied by a pilot to
cockpit controls. These values are found in many civil airworthiness standards as
either maximum or default values. Note: control forces are commonly given in
either decaNewtons (1 daN = 10 N), or pounds force (lbf). Both are shown here.
Where an aeroplane is fitted with mechanically linked dual controls (which is the
case for virtually all lighter aeroplanes, and a significant number of larger ones)
then a further case that is always considered is that of opposition of controls, e.g.
the captain operating the left hand controls attempts to roll to starboard, whilst the
co-pilot operating the right hand controls attempts to roll to port. The usual case that
is considered there is of 75% of the maximum input (as, for example, defined in
Table 8.1) to be applied at each control, but in opposing directions.
For all controls however, it is possible however that alternate values may be
considered more appropriate; for example if considering an aeroplane such as an
Airbus with a short sidestick (Fig. 8.6, Co-pilot’s seat of an airbus A320), then this
could and should reasonably be designed to much lower stick forces since it would
be almost impossible for any pilot to apply e.g. 75 daN in pitch to such a control.
Alternately non-standard controls may be required: for-example an aeroplane
cockpit which has been modified for disabled use may have a hand operated rudder
control, similar in form to a control stick, in which case it would be more appro-
priate to treat this (structurally) as if it were a pitch control, rather than a yaw
control (Fig. 8.7).
172 8 Control Surfaces and Circuits
So, do not take these values as set in stone. The wise design or airworthiness
team faced with any non-standard cockpit will conduct ergonomic tests to establish
the realistic maximum forces which can be applied by to their controls (a typical
approach to this would be to conduct tests using a reasonably large sample of adult
males, assume that this population is normally distributed, with a true mean and
standard deviation, then determine the 99.9th percentile load applied and use that).
Cockpit controls are normally fitted with “stops”: that is, physical limiting
devices somewhere between the inceptor itself and the control surface (or com-
monly at or before a force/position sensor, or a hydraulic master cylinder)—this
may be a bracket, short cable, shaped hole in a bulkhead, simply any device that
physically limits control travel. It is usually a legally acceptable practice then to
consider these particular loads only as far as the stop. This practice however should
be pursued with caution—if that stop is well within the aircraft and clear of the
cockpit area, then this practice is arguably acceptable. If however the stop is in the
cockpit then it may not be; the issue is one of control jamming. The reason that
these forces are high (generally much higher than the maximum loads that a pilot
should need to apply for any likely manoeuvre) is due to the risk of a foreign body
(often called “FOD” or Foreign Object Debris—encompassing pencils, items of
8.2 Control Inceptors 173
1
The author has encountered everything on this list.
174 8 Control Surfaces and Circuits
Fig. 8.7 Modified Flylight Skyranger cockpit with hand rudder control
then the pilot should be able to use the maximum available force to overcome that
obstruction without fear of instead causing serious damage to the flight control
circuit itself. Therefore, it is unwise to allow any lower loads to be considered when
designing or analysing a control circuit, for any part which is not very thoroughly
protected from FOD ingress. In this regard, FOD may not only enter from the
cockpit or cabin—stray tools have in the past been left within aeroplanes and
caused, sometimes fatal, accidents, and airport debris has sometimes also entered an
aeroplane.
Most control circuits on larger aeroplanes will use hydraulic or electrical actu-
ators at points within the control circuits. In such a case, the actuators themselves
will be rated to a maximum achievable load, and this load should usually be used
when analysing structure between or connected to such actuators.
A further issue that must be considered is that of control system distortion or
“stretch”. It is inevitable that any loads to any part of an aeroplane will result in
some deformation, and if this occurs excessively it results in an aeroplane whose
degree of control is poor or unpredictable. Generally this is evaluated experimen-
tally, although analytical prediction (to ensure that test results should be satisfac-
tory) is normal at the design phase. Tests are carried out by locking the control
surfaces (with any hydraulic or electrical systems in that circuit powered as if
during flight), and applying representative maximum normal flight loads to the
controls. Table 8.2 shows values typically used for this evaluation.
8.2 Control Inceptors 175
Table 8.2 Cockpit inceptor loads used for evaluating system stretch
Control Maximum applied load (SI) (daN) Maximum applied load (Imperial) (lbf)
Pitch 40 90
inceptor
Roll inceptor 20 45
Yaw inceptor 45 100
The total stretch (as measured at the control inceptor) is then evaluated, and
defined by:
Definition of control stretch
a
De ¼ 100 ð8:1Þ
A
where:
a control movement with control surface locked in the neutral position, and
A total available movement.
De is considered in terms of each individual part, where normally it is not
acceptable for more than 25% deformation to occur at any individual part of the
control circuit. Clearly, this may potentially allow a total deformation of 25% over
the entire control circuit—in practice such a degree of movement is likely to result
in a dangerously poor aeroplane and any aircraft showing a total control stretch of
greater than 10% in any axis or direction of control input should be regarded with
considerable caution—this is one of many cases where an aeroplane which in
theory should be perfectly acceptable (most, if not all, airworthiness standards
would allow a total of 25% stretch so long as no individual component stretches by
more than 25%), would in practice be extremely poor. Airworthiness practitioners
should consider this—even if required only to demonstrate bare compliance with an
airworthiness standard, they are as reliant as any other employee upon the con-
tinued health of aircraft sales and manufacture and this relies upon a good aero-
plane, not just a barely safe one. This is the civil view—the military airworthiness
community are likely to be less compromising in this, with their inevitable con-
centration upon mission fitness.
Clearly it is also important that the 25% stretch in any one component (or lower
values), should this occur, does not take the material in any physical component
past the elastic (Hooke’s law) limit. For this reason, analytical investigation of
control stretch must consider whether the elastic limit is exceeded at this value,
whilst experimental evaluation must measure load versus displacement as load is
applied, and then again as the load is removed so that any permanent deformation
can be clearly identified (or more importantly, shown not to be significant). It is
important however to realise that this test or analysis is different to that which is
176 8 Control Surfaces and Circuits
Whilst a pilot clearly requires controls that are sufficiently robust that they will
withstand both normal and occasionally extreme use, without failure there are
several other criteria which are equally essential to the airworthiness of a flight
control system’s inceptors. No list of criteria can be exhaustive, since no two
aircraft types (and often no two aeroplanes) will have identical roles and systems,
thus the airworthiness team should always be aware of the risk of non-standard
“issues”, and the need to find solutions to them. However, the following shows the
most common airworthiness issues that must be addressed by a set of cockpit
controls.
There are many reasons that a control may jam, all of them however are potentially
disastrous. Engineers assessing flying controls need to apply common sense, and
perhaps review accident reports (particularly those reported by the British AAIB, or
the American NTSB which tend to be the most thorough and accessible) for similar
classes or designs of aeroplane in order to determine the most likely causes of
jamming, and thus how to avoid them.
However, there are some aspects of design which particularly should be con-
sidered in order to ensure jam-resistant flying controls. There should be protection
from dropped items (pens, tools, items of clothing, etc.) entering the area of control
runs—most commonly this is achieved by ensuring that as much as possible of this
is enclosed and inaccessible (with flexible “boots” around the inceptors them-
selves), however an alternative sometimes used is to ensure that the controls
8.2 Control Inceptors 177
themselves run through sufficient open space within the aircraft that the risk of
jamming is minimal since there is little for any foreign object to jam against.
To cause jamming it is not essential that the ingress be of solid objects; liquids:
particularly water based such as rain or toilet fluids, present a particular risk because
at the low temperatures commonly experienced in flight freezing is possible. Water
not only solidifies as it freezes, but expands, so is capable of locking solid any
structure that it touches at the time. This type of failure has most commonly been
found with sheathed “Bowden” cables, which are often (although not exclusively)
used particularly on relatively lightly loaded control cables such as engine controls,
undercarriage blow-down, or trim systems.
An additional consideration that relates to jamming is that of geometry. It is
important to ensure that joints cannot lock “over-centre”, or moving parts at any
point rub against other structure within the aircraft (which could also, eventually,
lead to failure). A particular risk with cable systems is of the failure of cable-strands
leading to jamming within cables or runs, or even eventual failure. The most
common cause of this is the failure to ensure that cable has a sufficiently large
radius of turn where it passes around a pulley or through a fairlead. A few air-
worthiness standards, particularly the remaining British Civil Airworthiness
Requirements (BCARs), contain a requirement that the minimum radius of cur-
vature of any pulley or fairlead must not exceed 300 times the radius of the smallest
elemental strand of the affected cable, and that fairleads (curved cable guides)
should wherever possible avoid turns of more than 3°. Whilst this is rarely a
mandatory requirement for larger aeroplanes or those certified to non-UK standards,
the author’s experience is that this simple approach eliminates the failure of cable
elements, and if applied similarly to the routing curvature of Bowden cables (in-
cluding sealed units such as Teleflex cables) then again the rate of failure of such
cables is reduced to near zero.
An example of a generally well designed control run area is shown in Fig. 8.8
which shows the controls under the floor of a BAe-146. In this case the cables and
joints are not protected apart from the floor above them, but are positioned well
away from any other structure so that risk of jamming is avoided; additionally, the
cables used have been selected for extremely small elemental strand diameters, so
that relatively small pulley diameters may be used without significant risk of strand
failure.
Fig. 8.8 Control runs in under-floor area of A BAe-146-300-typical of any aeroplane, but seldom
visible without (as in this case) floors or other trim panels removed
It is unlikely that any pilot will exclusively fly one aeroplane type for most of their
career; even an inexperienced light aircraft pilot is likely to have flown 5+ aero-
plane types within the first few years of their flying career, whilst experienced
professional pilots will probably number the types they have flown in tens, and test
pilots in hundreds2. All of these pilots when faced with difficult circumstances of
2
The world record for most types ever flown is held by the late Captain Eric “Winkle” Brown, a
former Royal Navy Test Pilot, at 487 types (Captain Brown recorded all marks of Spitfire as one
type!). Widely regarded as the most accomplished test pilot in aviation history, his record is
unlikely to ever be broken. His autobiography, “Wings on my sleeve” is highly recommended for
any interested reader.
8.2 Control Inceptors 179
Table 8.4 Control sense of operation for some specialist airsports aeroplanes
Function Sense of the control
Cable release (sailplanes and powered Pull to operate (should normally position for
sailplanes) left hand operation)
Bar (weightshift microlights—see Push for pitch-up, pull for pitch-down Push
Fig. 8.10) right for roll-left, push left for roll-right
Nosewheel steering (weightshift microlights In common with the bar (primary flight
and some very simple ultralights) control), nosewheel steering of a flexwing is
conventionally the reverse of that on a
conventional light aeroplane, i.e. push
right = turn left. The mechanism is generally
similar to that of bicycle handlebars
Hand throttle (weightshift microlights) In UK and Europe usually located on left seat
frame In Australia usually located on right seat
frame. Other countries vary in their
conventions. In all cases, forward = more
power
Foot throttle (weightshift microlights) Right toe forwards for more power, sprung.
Configured so that minimum power is that set
by hand throttle, but this can always be
exceeded by the foot throttle
(Ground) Brakes (weightshift microlights) Left toe forwards, sprung closed
Choke (aeroplanes whose engine requires Usually a lever mounted somewhere close to
one—typically Rotax, Jabiru and Fuji the seat. The only universal requirement is that
engines) it must be clearly distinguishable from the hand
throttle
There are also conventions for the shape and markings of cockpit controls, again
these are not universally mandatory, but divergence from them should never be by
default—only with careful thought, and appropriate information to operating crews.
The first and most important point that should always be remembered when
considering the airworthiness of a cockpit design is that, although it may appear
8.2 Control Inceptors 183
Fig. 8.9 Typical light aeroplane carb-icing avoidance chart (this version is available from CASA
—the Civil Aviation Safety Authority of Australia)
Fig. 8.10 P&M PulsR—the current state of the art in flexwing microlights
very neat, it is very poor practice to position flight controls close together, or to use
similar shaped or coloured controls. In the cockpit, aircrew must be able to con-
sistently go to the various controls, identifying them by a very brief glance, or even
by touch (a cockpit may be poorly lit at night, or circumstances may make it unwise
184 8 Control Surfaces and Circuits
for a pilot to look inside the cockpit—for example when making a flap or under-
carriage selection on final approach).
There are, perhaps confusingly, two conventions for emergency controls. On
military aeroplanes worldwide, these are marked with black and yellow stripes;
conversely on civil aeroplanes these are marked in red. All emergency controls
clearly should also be clearly marked with their purpose unless (such as in the case
of an ejector seat handle) their use is clear and unambiguous.
For most aeroplanes, the following conventions apply:
Flaps
Electrical flap controls should be as shown in Fig. 8.11. With mechanical flaps
it is more common to us a mechanism similar in form to a car handbrake—
either pulled upwards from the cockpit floor, or downwards from the cabin
roof; this however is not a universal convention and generally subject to the
pragmatism of design requirements. What is much more important is to
determine that the control can be reached and operated from any seat from
which this may be necessary. Most commonly flap controls are coloured black,
grey or white.
Undercarriage
Undercarriage controls (whether electrical, mechanical or hydraulic) are
normally shaped as shown in Fig. 8.12; although this normally only applies
to primary controls and not to any emergency “blow-down” controls (which
also, unsurprisingly, are likely to be coloured red on civil aeroplanes, or black
and yellow on military aeroplanes). This control is often referred to as the
“gear lever”, but has no relation to the similarly named control in a car.
Engine Controls
Figure 8.13 is an extract from CS.23, however the conventions is contains are
universal to all modern aeroplanes—both civil and military. Powerplant
controls in aeroplanes may be either panel mounted (push-pull plungers), or
quadrant mounted (usually levers).
186 8 Control Surfaces and Circuits
There are also particular conventions which are accepted for gliders and
motorgliders, which are shown in Table 8.5.
The only convention common to microlight aeroplanes (so far as shape and
colour of controls is concerned) is again that emergency controls, such as they exist,
are normally coloured red.
Control surfaces are an inevitable part of almost any control system, since without
that mechanism to alter the airflow over the flying surfaces of the aeroplane, no
change is likely. In order to determine the loads on a control surface, it is normal to
determine (using whatever is considered appropriate of the many experimental or
analytical methods available) the pressure distribution over the control surface, then
to take moments about the hinge line. The force acting at the control horn (or
whatever mechanism is used for transmitting torque or force to the control) can then
be determined by straightforward mechanics.
It is normal on any aeroplane for the engineers evaluating the loads at the control
surfaces (that is, those used to control the flightpath, which in this context does not
include lift modifying devices such as flaps, or drag modifying devices such as
airbrakes) to consider two sets of values; those are for 1/3rd deflection of the
control at VD, and full deflection at VA. These values will define the limit loads.
Considering in the simple case of an upgoing elevator with a single elevator horn.
Figure 8.14 shows three deflection states. Case (a) is disregarded with regard to the
control surface loads since the elevator since it is not considered to be deflected;
case (b) shows the control deflected one-third of its full scale deflection, whilst case
(c) shows the control deflected fully.
Considering case (b), it must be determined what the lift distribution is as a
function of distance from the hinge, x, from the hinge (x = 0) to the trailing edge
(x = xte). If the distance from the hingeline to the control horn is ych, then it may be
seen for a stationary control (which is the case generally considered) that:
8.3 Control Surfaces and Linkages 187
Zxte
CP dx ych Fch ; ð8:2Þ
0
where Fch is the force acting at the control horn. This basic relationship as
portrayed is almost certainly over-simplistic for any actual case. However all
cases will follow these basic principles and, starting from this basic rela-
tionship it is possible to derive a formula which may be used to determine the
loads at a control horn, and thus through the control circuit for the critical
load cases.
The load cases that must now be considered are the following:
1. That of the control surface structure itself.
2. If the control system is irreversible, that of the structure between the control
horn and the hydraulic (or other type of) slave actuator. If the control system is
reversible, then at-least the control circuit as far as the stop, but if the forces are
greater than those predicted due to inceptor loads, as far as the cockpit control.
3. That of the hinges due to linear loads reacting moments about the control horn
attachment point.
4. That of the main fixed surface (wing, tailplane, fin, etc.) to which the hinges are
attached.
188 8 Control Surfaces and Circuits
There is nothing unusual about any of this analysis, which in outline is identical to
that required for any other part of the aeroplane. However; it is clear that the
complexity of such analysis may become considerable and it is important not to
underestimate the resources that may be required.
During the middle part of the second world war, one of the Royal Air Force’s most
important fighter aeroplanes was the Supermarine Spitfire (Fig. 8.15), an aircraft
which was being constantly modified and updated in the endeavour to achieve and
maintain technical superiority over adversary aircraft operated by axis forces such
as the Bf109 or FW190. As greater power was added, it was noted that at the higher
speeds which could now be sustained, aileron effectiveness became reduced and in
extreme cases (i.e. when flying above the aileron reversal speed) the ailerons
started working the wrong way around—so pushing the stick left would result in a
roll to the right. The reason for this was that the twisting moment acting upon the
wing due to aileron deflection, caused a twist in the wing whose effect was initially
to reduce aileron effectiveness, and above that critical speed, to actually create a
reversal. It is believed that the Luftwaffe was suffering similar problems with its
own front line aircraft at about the same time although historical records are poorer.
Whilst mostly known as a historical problem, aileron reversal still exists and the
author has experienced aileron reversal on the CFM Shadow, a light/microlight
aeroplane type designed in the 1980s—particularly on the higher powered later
variants and this led to the refusal to certify the prototype “Shadow E series”
8.3 Control Surfaces and Linkages 189
Fig. 8.15 A late model Spitfire wing—prone in some variants to aileron reversal at high speed
aircraft. In the 1960s it was also known to occur on the prototype C141 Starlifter.
The main design factor which tends to lead to aileron reversal is low torsional
stiffness of the wing, and the problem seems to have usually appeared when
upgrading the power output of existing aeroplanes with no, or only moderate wing
sweep.
Detailed explanations and derivation of the formulae for aileron reversal may be
found in standard texts on aeroelasticity or aircraft structures, but are too specialised
for inclusion here. Mention is made primarily to warn airworthiness practitioners
that aileron reversal is not just a historical peculiarity, has occurred in recent years,
and may appear again. On a new design, or one which has increased power or
weight, the possibility should always be considered and explored, particularly with
a straight winged planform.
Most airworthiness standards or design specifications will include a requirement
that a minimum steady state roll-rate can be achieved with “normal use of the
controls”. The applicable standard may act as a backstop if one is required, to
ensure that a torsional flexibility problem gets dealt with—although simple
demonstration of the amount of wing torsion and the likely degree of fatigue life
reduction that results, is likely to be sufficient.
Chapter 9
Powerplant Airworthiness
Abstract The powerplant is the part of an aircraft that converts fuel into thrust. As
such, it is also the part of the aircraft that has the potential most hazardous in event
of a failure, and is generally isolated from the main aircraft body by a firewall.
Therefore, airworthiness codes are designed to minimise this risk, and ensure that in
the event of failure, there is no unacceptable damage to the rest of the aircraft. For
the purposes of this book, the powerplant will be one of three main possibilities: a
gas turbine engine, a propeller driven by a gas turbine, or a piston engine driving
one or more propellers. This chapter concerns itself with discussion of airworthi-
ness codes, principles and best practices to ensure airworthiness of everything
forward of the firewall.
The powerplant is that portion of the aircraft which turns fuel into thrust. The
student of the history of aircraft design will note that this covers a very wide
multitude of technologies, but for the majority of aircraft it usually is one (or
several) of the following:
• A gas turbine engine (turbojet, turbofan, etc.),
• A gas turbine engine driving a propeller or rotor via a gearing system (turboprop,
turboshaft),
• A piston engine driving one or more propellers, either directly or via a gearbox
(piston/piston prop).
Fig. 9.1 The range of modern aircraft engines—a Verner 133 M on a Raj Hamsa X’Air and
podded CFM56–5C2 engines on an Airbus A340–300
The first two of these requirements will vary greatly with the applicable air-
worthiness standard, and with the design and role of the aircraft being assessed. The
third is a universal requirement, but the aircraft role will again heavily define the
length and rigour of testing, from perhaps 25–40 h of endurance testing with a few
hours of specific engine tests for a homebuilt or microlight, to some thousands for a
new airliner/engine configuration, and considerable deliberate engine stressing in a
combat aircraft.
If we accept that there is always some risk of a failure of all or part of the engine,
then clearly it is important that the rest of the aircraft is protected so far as possible
from the consequences of that failure—which may include either high energy
debris, or some combination of fumes and heat.
For those aircraft where the engine is built into the fuselage (such as the Pilatus
PC12 in Fig. 9.2) the simplest and most obvious part of this protection is the
firewall. This should, in theory, provide a barrier to particularly fumes, heat and
flames, in the event of uncontained combustion within the engine compartment (an
“engine fire”). Whilst sometimes other materials are used, the normal design
approach is to manufacture the firewall from suitable gauge sheet metal, with
structural members incorporating the firewall, and the design sealing and min-
imising so far as possible the number of holes in it (although some holes—for
engine controls, fuel lines, sensor cables etc. are inevitable). Figure 9.3 shows the
firewall (with a tank behind it—on the right of the photograph, and the engine
mounting points on the left) from an AVRO York, which whilst not a current
Fig. 9.3 Engine and mounting frame from an AVRO York (a 1940s piston engined passenger
airliner)
aircraft, still shows the typical configuration of a piston engined aircraft firewall. On
a podded gas turbine such as shown on the right of Fig. 9.1 or in Fig. 9.7—the
modern airliner configuration—this is less of a requirement because of physical
separation from the wing and fuselage. But in that case, the use of ballistic pro-
tection in case of a high energy failure is instead the main design criterion.
Having created an isolated engine compartment forwards of the firewall (or in a
few aircraft behind it) it is then possible to isolate the fire risk area and to minimise
risk in that area, by measures such as ensuring the use of fire resistant wiring and
hose (although fireproof is preferable), ensuring adequate draining and ventilation
in case of fuel spills or leakages, and ensuring that all metal components are
electrically bonded to each other.
Sheet metal is most commonly used in firewalls and for light aircraft at least it is
accepted that stainless steel sheet at least 0.38 mm thick, mild steel sheet at least
0.5 mm thick, or aluminium alloy sheet at least 2.5 mm thick: combined with steel
or copper alloy fittings provide an acceptable firewall, although other materials
(such as composites or wood plus some form of fireproofing treatment) are also
regularly used, subject to demonstration by analysis or test that they (along with
surrounding structure, such as cowlings) are fireproof. Part 25 aircraft requirements
will be more complex, but following a similar spirit.
9.2 Protecting the Aircraft from Its Engine 195
Engine mounts are the structure that carry the loads imposed by the powerplant
installation upon the airframe. These are:
• Vibration (often necessitating anti-vibration devices within the structure)
• Thrust loads
• Torque loads: Turboprop engines need typically be assessed for torque at
maximum power and speed multiplied by a safety factor typically around 1.6
depending upon analysis. Turbofan and turbojet engine mountings are generally
assessed for engine retention in the event of a sudden seizure or stoppage, or
“slam” accelerations. Piston engines suffer cyclic torque loadings potentially up
to 8 times the mean torque output, depending upon configuration—see
Table 9.1 particularly for piston engines.
• Inertial and static loads (due to the mass of the powerplant)
• Aerodynamic loads upon the engine and its cowlings/nacelle
• Crash case inertial loads
Usually engine mounts comprise of a framework structure with a small number
of attachment points to both the engine and airframe similar to that shown in
Table 9.1 Typical torque multipliers to account for piston engine torsional load variation in
engine mount analysis
No. cylinders 2-stroke engine safety factor 4-stroke engine safety factor
1 6 8
2 3 4
3 2 3
4 2 2
5+ 2 1.33
Note More modern aircraft piston engines—particularly the popular Rotax models are very likely
to be fitted with a gearbox. If calculating engine mount limit loads, then engine maximum torque
output should be multiplied both by these safety factors, and by the reduction ratio of the gearbox
196 9 Powerplant Airworthiness
Fig. 9.3 above since this tends to be light, easily fabricated, easily analysed, and
permits ready removal and refitting of the engine when required.
Typically loads are defined as limit loads, based upon a combination of design
rules of thumb and analysis of the engine characteristics, and then analysis and
approval is carried out and reported on a similar basis to that for other aircraft
structure as described previously. This done primarily by analysis since the loads
(particularly crash loads) can be large and expensive to test, whilst the reasonably
simple design of most engine mounts makes them relatively easy to analyse.
Vibration must also be considered. For simpler aircraft and smaller piston
engines, it is usually possible simply to specify engine anti vibration (A-V) mounts
which are matched to the known characteristics of the engine. However, for larger
aircraft and engines (and most gas turbine engine installations), a more complex
analysis and in particular testing regime is required, which is likely to extend to
flight testing with a heavily vibration-instrumented engine and mounting system.
(The Boeing 737-400 crash at Kegworth, UK on 8 January 1989 was for example
found to be caused in part by inadequate in-flight vibration analysis of the engine
during certification of the then new CFM56-3C engine variant, leading to normal
practice afterwards having substantially greater emphasis on this area.)
Even considering the improvements possible … the gas turbine [engine] could hardly
be considered a feasible application to airplanes mainly because of the difficulty with
the stringent weight requirements.
Gas Turbine Committee of the U. S. National Academy of Sciences, 1940
Once the detail design of an engine has been assessed, then it is necessary
to demonstrate the integrity of an engine. This can be done to some extent by
analysis—and for cost reasons such analysis is particularly inevitable on a larger
engine. However, ultimately the only real way in which it is possible to demonstrate
the integrity of an engine is by testing it.
For medium to large civil aeroplanes, civil helicopters, and most modern military
aircraft, it is usual to conduct the integrity testing of the engine in a manner that
treats it as a separate entity, leading to type certification of the engine, before it is
formally integrated into the aircraft. For lighter civil aeroplanes (those in the
microlight, VLA and motorglider categories), and the majority of older military
aeroplanes this has been done with the engine being treated as a component of the
aeroplane, rather than as a separate entity.
The testing will normally be performed on a test rig, which so far as cost and
facilities reasonably permit, will be designed to represent the interface between the
engine and the external environment—both structurally (with the aircraft), aero-
dynamically (in terms of, so far as is practicable: airflow, density and temperature),
9.4 Integrity of the Engine 197
and if a propeller is fitted loading will also be represented—for piston engines this
is most commonly a dynamometer arrangement, whilst for turbine engines it is most
commonly a test cell. The main purpose of the testing is to run the engine through a
representative set of conditions for a period of time which exceeds the likely
inspection intervals, at the end of which it will be fully disassembled and inspected
(known as a “teardown”). If any major repairs are needed during or immediately
after the his ground test programme, then it is likely that some of the testing will
need repeating. Typically for a small aeroplane engine, this testing will run to about
50 h, or for a large aeroplane engine around 150 h of running, following a repeated
cycle simulating typical operating condition—the extremes of ground testing
regimes are shown below from CS-E (Fig. 9.4—for large gas turbine engines) and
CS.22 (Fig. 9.5—for very light piston engines).
Where an engine has particular features such as pitch locking, reversal or
feathering for propeller driven aircraft, or short periods of high power (typically
called a “combat” or “emergency” setting on military aircraft, or OEI—One Engine
Inoperative—Emergency Power on civil aircraft) or thrust reverse for turbine
engines, or a parking brake for turboprop aircraft then the profile of ground running
will also either include, or be preceded by repetitive representation of these actions,
and it will also of-course include representative start-up and shut-down routines
(with the engine both hot and cold for starting, often also including airborne
restarts). In all cases, it is unlikely that any authority will permit the majority of
testing to simply be at unloaded sea-level conditions: depending upon the engine
and its future role, a proportion of tests at simulated flight altitude conditions, high
power drain or high air bleed are likely to be required, as will engine starts at a
Fig. 9.4 Typical operating profile for transport aircraft engine ground testing
198 9 Powerplant Airworthiness
Fig. 9.5 Operating profile for light aircraft engine testing (typically repeated 25 times)
Fig. 9.6 Typical engine ingestion water and ice densities (based upon EASA CS.E)
weather (most particularly rain and hail). This would be too random and hazardous
to attempt to deliberately investigate in flight, so testing is invariably done under
ground test conditions. Typical tests will include:
• Ingestion of hailstones up to 25 mm in diameter up to 35,000 ft, decreasing to
diameters of 6 mm at 59,000 ft, at a density appropriate to the aircraft’s oper-
ating environment.
• Ingestion of rain, of a density appropriate to the operating environment of the
aircraft (For both rain and hail densities, see Fig. 9.6).
• Ingestion of a single large bird, normally impacting no slower than 200 knots
(103 m/s, 337 fps) and into the most sensitive part of the intake. The size of bird
is usually dictated by the engine intake diameter as shown in Table 9.2, and
after the test the engine is required to continue running without a substantial
(typically more than 25%) loss of thrust.1
• Ingestion of multiple, “flocking” birds, the number, speed and mass of which
will depend upon the size, speed and role of the aircraft.
1
These tests are normally conducted with chickens or turkeys, since they are readily available.
There is an urban legend that somebody, somewhere, once tried to conduct this with a frozen
chicken, without first thawing it, and as a result destroyed a prototype jet engine. The author has
never seen any evidence that this actually happened—but just in case, please take this as a warning
to always fully thaw frozen poultry before either cooking it, or firing it into a gas turbine engine
intake.
For completeness however, the author has learned since the first edition, from retired FAA Test
Pilot James Plackis, that this did indeed happen in 1958 during certification testing of the
Gulfstream 1. However, that was on testing of the windscreen, not the engines, and the frozen
chickens did indeed penetrate the windscreen somewhat destructively.
200 9 Powerplant Airworthiness
Table 9.2 Typical test bird Intake area (m2) Minimum mass of bird
sizes for single bird ingestion
tests <1.35 1.85 kg (4 lb)
1.35–3.9 2.75 kg (6 lb)
>3.9 3.65 kg (8 lb)
Virtually never will any aircraft be allowed to fly without some form of engine
instrumentation. At the simplest level (for example some simpler microlights and
self-sustaining motorgliders) this will probably be no more than an RPM
9.5 Engine Instrumentation 201
(Revolutions Per Minute = engine speed) gauge, and an ignition switch. For the
majority of aircraft however, engine instrumentation is rather more elaborate and
necessarily so—whilst most aircraft can fly without an engine, it tends to limit the
landing options, and thus particularly in a larger aircraft will almost certainly result
in substantial aircraft damage. So it is important that crews can continuously
monitor engine condition and operation.
It is normally necessary for the overseeing airworthiness team to agree minimum
engine instrumentation as part of the MMEL (Master Minimum Equipment list),
which an aircraft operator then can use as the basis for determining an MEL
(Minimum Equipment List—a local document which cannot be less restrictive than
the MMEL, but may at the operators discretion be more restrictive).
The specific instrumentation will vary with application but Table 9.3 illustrates
what is typically installed.
Unsurprisingly, it is likely that most of the instruments fitted will require dupli-
cation in a multi-engined aircraft, and even on piston engines it is not uncommon on
higher performance aeroplanes to see multiple CHT and EGT gauges—representing
each cylinder separately. Similarly this list is not exhaustive—but describes those
aircraft engine instruments most likely to be met.
As has been discussed previously, any engine has potential to fail in flight—for this
reason it has been accepted since early in the history of commercial air transport
(CAT) that most passenger transport should require a multi-engined aircraft. This
often meant more than 3 engines, giving double redundancy—that is that following
9.6 A Special Case—ETOPS 203
a single engine failure an aircraft could continue en-route, and following a second
and subsequent engine failure, sufficient thrust should still be available to make a
diversion and landing.
However, an increased number of engines escalate purchase and operating costs,
and usually also increase fuel consumption—so economic and environmental
pressures created an incentive to minimise the number of engines on a passenger
aircraft. As a result, the concept of ETOPS—Extended range Twin engine
OPerationS originated in the early 1980s. ETOPS approvals are given to specific
aircraft type/operator combinations, and allow flight up to a maximum distance
(defined as a flying time on a single engine in still air) from a suitable diversion
airfield. Issue of these approvals is based upon demonstrated minimum powerplant
reliability criteria for the aircraft. The original ETOPS clearance, issued by the FAA
in 1985 permitted Boeing 767 aircraft operated by Trans World Airlines (TWA) to
operate up to 90 min from a suitable airfield (the earlier rule having been that twin
engined aircraft might not fly more than 60 min from a suitable diversion).
Subsequently, as levels of reliability have improved, ETOPS clearances have
increased in time, with ETOPS-180 (180 min) now being a common clearance, and
very occasional 15% extensions to ETOPS-207 being awarded in special circum-
stances. As a result, about 95% of the earth’s surface may be overflown by
twin-engined passenger airliners, providing a large commercial boost to CAT—the
small proportion of unavailable space mostly being at the centre of large unin-
habitable areas such as the Atlantic, Pacific, or Antarctica (Fig. 9.7).
The level of ETOPS clearance that will be awarded depends upon local regu-
lations (in particular there are substantial philosophical differences between the US
and Europe). However, in general the approach will be:
Fig. 9.7 Boeing 777—an aeroplane which probably wouldnʼt be economically viable without
ETOPS approval
204 9 Powerplant Airworthiness
1. During the certification programme for the aeroplane, the manufacturer must
apply for “ETOPS Type Approval”, which will mandate minimum levels of
equipment and redundancy, and also demonstrated minimum levels of single
engined performance.
2. Initial ETOPS clearance will usually be limited, e.g. to ETOPS-90 or
ETOPS-120.
3. The operator must prove to its overseeing national authority that it has adequate
training, operational and engineering procedures in place to handle ETOPS
operations; this will include maintenance standards, engine condition monitor-
ing diversion planning, loading, single engine cruise, appropriate minimum
equipment list (MEL), communications with alternate airports, and retention of
records.
4. The aircraft is then operated with its limited ETOPS approval for a period of
time (typically a minimum of 100,000 h and 12 months for a new powerplant
type, or 50,000 h and 12 months for a derivative powerplant type). During this
period, engineering oversight of the reliability of the aircraft and in particular
their powerplant is substantial—typically a maximum risk of double engine
failure not exceeding 0.3 10−8 per flying hour needs to be demonstrated.
5. Subject to adequately demonstrated reliability, procedures are modified and an
extended ETOPS approval is awarded.
6. Engineering and operational oversight will continue throughout the life of the
ETOPS approval, with the level of ETOPS approval given remaining condi-
tional upon a continued minimum level of demonstrated engine and associated
systems reliability (see Fig. 9.8).
At the time of writing, the state of ETOPS approvals is not static and will almost
certainly change in the future. In 2014 the Airbus A350XWB was approved by
Fig. 9.8 Typical relationship between demonstrated single engine In-Flight Shut Down (IFSD)
rate and permitted ETOPS diversion time
9.6 A Special Case—ETOPS 205
EASA to ETOPS-370 and the Boeing 787–800 was approved by the FAA to
ETOPS-330. After dallying with the term LROPS—Long Range OPerationS,
ICAO has now labelled this concept EDTO—Extended Diversion Time Operation.
Whilst it has been suggested that this may extend to 3 and 4 engined aircraft, at
present it still only applies to twins.
EDTO certification is becoming a joint process between the manufacturer and
the operator, where the manufacturer is responsible for establishing compliance
with design and reliability criteria, and the operator with operations and mainte-
nance requirements, although the primary responsibility rests with the manufac-
turer. ETOPS and EDTO criteria are now also extending beyond pure powerplant
considerations, through fuel and electrical systems reliability, into less obvious
areas such as cargo area fire suppression, hydraulic system redundancy, and aux-
illiary power unit (APU) start cycle reliability, communications system redundancy
(with SATCOM capability beyond ETOPS 180) and—extremely importantly—
aircrew workload whilst managing the failure and diversion. The fire suppression
criteria are increasingly being applied to 4 engined aircraft such as the A380 for
long distance approvals, but not full EDTO approvals.
9.7 Propellers
The propeller is just a big fan in the front of the aircraft to keep the pilot cool. If you
want proof of this, turn the engine off, and watch the pilot break out into a sweat.
Cliché, remarkably accurate
propeller blades and their retaining structure at the propeller hub will remain intact
during rotation at the maximum rated rotational speed (defining the limit load)
multiplied by safety factors appropriate to the materials and methods of construc-
tion. Calculation of that is a simple matter of centrifugal force determination
(generally dividing the blade into sections, treating each as a point mass, estimating
F = mv2/r for each section, then summing the forces outboard of each section as
being the total centrifugal force at that section), and the propeller blades themselves
usually lend themselves to longhand or finite-element analysis of failure load—
although they are less amenable to load testing that doesn’t modify the propeller
structure through creation of the test method. Thrust loads must also be considered,
but in the majority of cases, those are approximately an order of magnitude less
than centrifugal loads, so will be of less significance. So, usually, evaluation is a
relatively simple stress analysis—unless facilities exist to spin a propeller to failure/
ultimate load. Such facilities are rare, and such tests highly hazardous.
Most propellers will have separate blades set into a central hub, and it is unlikely
that the hub mechanism will be as amenable to simple analysis. So, proof by test is
normally required, where a “dummy blade” is manufactured and within a laboratory
test rig “pulled out” of the hub to the required total centrifugal load plus safety
factors, also for most instances with the addition of an appropriate thrust component
adding bending at the blade root. Tests will be carried out similarly to any other
structural component, followed similarly by disassembly and inspection for dam-
age. Similar evaluation must be made of the torque and thrust carry-through from
the engine to the propeller.
Propellers fitted to engines above about 200 hp/150 kW power—which is larger
piston engines and virtually all turboprop engines, will normally be variable pitch,
also commonly referred to as constant-speed (or colloquially “wobbly props”).
9.7 Propellers 207
A variable pitch (VP) propeller mechanism acts similarly to the gearbox of a car, in
that it allows propeller pitch to be varied so that as airspeed varies the engine can
operate about either a best economy or best power/torque rotational speed. So for
take-off and climb, a VP prop will typically have the cockpit control (normally a
lever next to the throttle) set forwards to fine (equivalent to “low gear”) whilst for
an economical or high speed cruise it will typically be set rearwards to coarse
(equivalent to “high gear”). Most, although not all, such systems are mechanical
and infinitely variable within a set range; the exceptions are some aircraft such as
the Grob G109b motorglider which has a 2-position fine/coarse prop, and some
very modern composite propellers for light and microlight aeroplanes which use an
electrical control system. It is essential that the failure modes of VP props are
considered as part of the powerplant airworthiness evaluation—there must be
means for the pilots to recognise that such a failure has occurred, and failures
should always be into a fixed pitch mode. In the most common systems centrifugal
force drives the mechanism towards fine pitch, whilst a cockpit control uses oil
pressure to work against that towards a coarse pitch, thus ensuring that in the event
of an oil pressure failure (far more likely than a failure of a centrifugal rotating
weight mechanism) the system defaults to a fully fine pitch setting—limiting for-
ward flight speed, but producing best take-off and climb performance, which is the
safest failure mode.
Once these structural and control considerations have been covered, the instal-
lation of the propeller on the aircraft may then be evaluated. Geometry must have
been confirmed at the design stage, but also confirmed by inspection on a com-
pleted airframe—adequate airframe and ground clearance of all parts, which in
particular allows for aeroelastic and landing load deformations so that no rotating
parts can suffer an impact and consequent damage. Airworthiness standards will
define minimum airframe clearances for propellers (typically *25 mm radially and
13 mm axially), usually in a worst case of deformation and for ground clearance
probably also with deflated tyres and compressed oleos. However, particularly for
modern composite propellers flexibility may be rather greater, than was assumed in
drafting these standards and based upon experience the author would at propeller
tips increase those values by a factor of 4. To 100 and 50 mm. Ground clearances
are typically required to exceed a worst case minimum of *180 mm for nosewheel
aeroplanes and 230 mm for tailwheel aeroplanes2 taking into account least
favourable attitudes gear deflections during taxi, take-off or landing (Fig. 9.10).
Once a prototype configuration is established, ground testing of the engine/
propeller/engine combination is essential. The requirement to ensure adequate
engine control, lubrication and ground cooling is obvious and inevitable. Fixed
pitch propellers should be set to deliver around 92–96% of the maximum permitted
engine RPM at full throttle with the aircraft stationary on the ground (secured to an
2
It is tempting when seeing apparently very carefully calculated values listed in SI units within
airworthiness standards to assume considerable research behind those figures. In this case, the
values correspond to semi-arbitrary values of 6 and 9 in. set many years ago, but subsequently
found to be satisfactory or conservative, and left unchanged.
208 9 Powerplant Airworthiness
Fig. 9.10 1934 de Havilland dH88 Comet—showing off two early variable pitch propellers—an
aircraft manufactured today would have substantially the same arrangement, and identical
airworthiness considerations. Note here the importance of fuselage clearance, and when the tail is
lifted for take-off, of ground clearance
unmoveable object behind the aircraft). Variable pitch propellers should be set so
that they do the same at the intended take-off pitch setting, which is usually the fully
fine pitch setting. Flight testing will include evaluation of propeller pitch, as well as
of engine cooling (particularly for piston engines where it is particularly critical,
and heavily related to flow over or through the engine from the propeller(s)).
Airworthiness standards will normally require that at the recommended settings the
maximum engine speed isn’t exceeded in take-off and climb, and 110% of that
maximum can’t be exceeded with the throttle closed in a dive to VNE. In practice,
these are underly conservative requirements for a good aeroplane—better, but not
legally mandatory, requirements would be that an aircraft cannot exceed maximum
engine speed at full throttle in level flight with the cruise pitch setting (coarse for a
variable pitch propeller, or as the fixed pitch prop comes) and that maximum engine
speed cannot be exceeded at-all with the throttle at idle, any pitch setting, and VNE.
Attempting to demonstrate before flight test the compliance of propeller pitch
with airworthiness or best practice requirements is difficult—there are few reliable
analytic tools or generic data. Most propeller manufacturers and some other
organisations will have empirical data that they can share, but this is one of the
major reasons for the popularity in recent years of ground adjustable fixed-pitch
propellers, that allow adjustment between test flights, and even alteration in
9.7 Propellers 209
individual airframes to suit operators who prefer an aircraft optimised for either
take-off and climb performance, or for cruise efficiency. The last can be problematic
from the point of view of noise certification—for those aircraft that require it, as
such adjustments will change the noise profile of an aeroplane. Noise certification is
discussed further in Chap. 17.
On January 17 2008, British Airways flight 38, a Boeing 777-236ER (similar to the
aircraft in Fig. 9.7) was flying an approach to London Heathrow Airport after a
4400 nm high altitude sector from Beijing. As is normal practice, the aircraft
intercepted the ILS for Heathrow’s runway 27L from below. Shortly after inter-
cepting the ILS’ 3° glideslope, the aircraft started to drop below the required
flightpath which the autothrottles failed to increase power to correct. The Captain
manually attempted to increase power, but this also failed. Flaps were partially
de-selected to extend the glide, but the aircraft still impacted the ground before the
runway, causing substantial airframe damage. Of the 152 persons on board, 47 were
injured, one seriously—but nobody was killed.
This incident was clearly a testament to the controllability and crashworthiness
of the type, as well as the management of the flight by its crew, but raised serious
questions about the airworthiness of the powerplant systems. This resulted inevi-
tably in a lengthy investigation, led by the UK’s Air Accidents Investigations
Branch (AAIB), supported by Boeing, British Airways, the USA’s National
Transportation Safety Board (NTSB) and Rolls Royce.
This was an extremely high profile investigation, and unsurprisingly several
interim bulletins were issued in coming months, which permitted other operators of
similar aircraft to consider the developing understanding of cause so that potential
causes could be avoided in their own operations. A final report was issued in
February 2010.3
The process by which the conclusions were determined involved a large
investigation, creation of mock-ups of parts of the fuel system, and simulation of
the conditions of the transit from China to the UK, which turned out to be
abnormally long, slow, and cold—creating relatively unique conditions that
allowed a “jet fuel ice cream” (author’s firmly unofficial term) to form along the
inside wall of fuel hoses. When power was increased during the glideslope inter-
cept, this unstuck, partially blocking a fuel filter, restricting fuel flow to the engine,
and thus only enabling partial power to be generated.
3
Air Accidents Investigations Branch, Formal Report AAR 1/2010. Report on the accident to
Boeing 777-236ER, G-YMMM, at London Heathrow Airport on 17 January 2008. This report is
available from the AAIB’s website at www.aaib.gov.uk.
210 9 Powerplant Airworthiness
By the time of this report, revised fuel system design and recommendations for
the use of anti-icing fuel additives were already in place, and the report itself lists
115 conclusions and 18 recommendations, going far beyond just the fuel system.
Whilst this report highlights some failures in the system which had approved the
aircraft and its operating procedures—it also shows that these failures were due to
circumstances not being foreseen because they had not previously been docu-
mented. Otherwise however this case shows the airworthiness system at its best.
Specifically:
(i) The initial accident was rendered entirely survivable.
(ii) All organisations with involvement in the background to the accident, par-
ticipated immediately and competently in the investigation.
(iii) Immediately possible causes were identified, these were published so that
organisations with an interest and ability to prevent a recurrence were able to
commence early their own investigations and actions.
(iv) Multiple recommendations were made, including for revisions to design
codes, to prevent a recurrence. These were developed in conjunction between
investigators and stakeholders, so that they would be accepted and adopted
as quickly as possible.
At the same time, whilst these laudable features of the conduct of the investi-
gation are worthy of note, the investigation was also built upon teams of engineers
building an extremely thorough understanding of the aircraft and its engines and fuel
systems, then using that understanding to work the problem of how to prevent future
recurrences of this or similar accident, through combinations of analysis and test.
Note—the future desire to be able to run aeroplanes on alternative fuels with
reduced environmental impact (such as described in Chap. 17), may cause cases
and issues such as are described here to be revisited and become increasingly
critical in the future.
In 2002 the author was part of the team working on certification of the UK version
of the MXP740 Savannah (Fig. 9.11), an Italian kitplane based upon the American
Zenair CH701 STOL. The aircraft was fitted with a relatively new 4-stroke
air-cooled engine direct driving (i.e. not through a gearbox) a fixed pitch propeller.
During a handling test flight, the test pilot experienced a strong smell of burning
in the cockpit, causing him to shut down the engine and make a glide approach and
landing. On investigation the central components of the carbon-fibre composite
propeller (blade roots, spinner and clamshell hub) had partially combusted. This
appeared to be due to a slippage in the propeller hub causing a shift of the propeller
disc centre of gravity and a consequent high frequency vibration in flight that had
led to heating.
Flight testing was temporarily suspended for investigations, which highlighted
that there had been a number of other—albeit less dramatic—failures of
high-rotational inertia composite propellers on other aircraft fitted with the same
engine—these failures had not occurred with wooden propellers (which tend to have
lower rotational inertia, and also are generally better at absorbing vibration).
In an attempt to investigate this further, a suitable aeroplane forward fuselage
with the right engine fitted was obtained, and the engine fitted with high rate
accelerometers connected to an automatic recording system. Then a series of tests
were carried out with different propellers and permissible grades of fuel. Data from
accelerometers were analysed using a fast Fourier transform (FFT) software
package.
Analysis showed that there was a marked increase in resonant peak amplitude of
the 2, 4 and 6 per-rev cycles, particularly the 2 per rev cycle at certain conditions. In
particular the use of low octane unleaded fuel [MOGAS] (compared to higher
octane leaded fuel [AVGAS]), propeller tip Mach numbers calculated to be in
excess of M = 0.82. Conversely resonance was, contrary to suspicions not worse
with composite propellers than wooden—but some propellers at the same operating
conditions did distinctly show worse characteristics than others.
The conclusion from this was to prohibit propeller installations on all aircraft
with the engine fitted that would in any predictable flight regime allow tip Mach
numbers to exceed M = 0.8. In practice this brought maximum RPM for some
aircraft installations down by about 10% from the manufacturer’s manual limit,
something relatively easily done for most installations by setting ground adjustable
propellers slightly coarser than might otherwise have been the case—in effect
configuring for slightly poorer take-off and climb performance, and slightly
improved climb performance. Whilst not prohibited, advice was also offered to
favour the use of certain propeller models over others, and to favour the use of 100
octane 100LL AVGAS over 85 octane unleaded MOGAS.
Following this approach appeared to eliminate all such future incidences, in a
manner which was happily relatively (apart from the higher cost of AVGAS
compared to MOGAS) unproblematic for aircraft owners and operators.
Chapter 10
Crashworthiness and Escape
It has become normal and legally required practice to require demonstration that for
any medium to large sized aeroplane (typically with more than 44 passengers) it is
possible to evacuate any aircraft within 90 s, with half of the emergency exits
closed. This is normally demonstrated at night or in simulated night conditions
using untrained personnel, with typical obstructions (bags, pillows, blankets, etc.)
distributed around the cabin although obviously includes the use of emergency
egress devices, particularly slides, emergency lighting. A typical make-up of the
test passenger load, is defined in part 25, and would be that they are:
• 30 + percent female
• 5 + percent aged over 60
• 5–10% are children aged under 12
• representative cabin crew
For smaller aeroplanes, the requirements are less prescribed, but nonetheless it is
important that egress from the aircraft can be reasonably rapidly achieved—
10.2 Escaping from an Aircraft 215
including with the aeroplane inverted on the ground (which can cause difficulties
e.g. with sliding canopies). Assessing this will often require the design and air-
worthiness teams to consider carefully the nature of both the design and intended
operation of the aircraft.
Helicopters can provide some significant additional challenges, given that larger
helicopters particularly will often require certification for ditching—a necessity if
they are to regularly serve in an offshore role such as rig resupply, search and rescue
or inter-island ferrying. In such a case, whilst the assumption of an untrained
“population representative” passenger load can often be relaxed (those flying on
board offshore helicopters, even only as passengers, normally receive special
training for this) new requirements emerge for submerged abandonment. This is
likely to include allowances for various floating attitudes, partial blockage of exits,
partial failure of floatation devices, various sea-states, and the quite probable
inversion of the helicopter underwater1.
In all cases, it must be easy and intuitive to open emergency exits, without use of
excessive strength (or language-specific signage). And of-course, these exits must
also be large enough for use, and positioned so that (even if the engines are still
running) the passenger or crew leaving the aircraft will not suffer subsequent injury.
Unsurprisingly, for large aeroplanes FAR-25 and CS.25, the civil airliner require-
ments contain clear requirements for the minimum size, position, visibility and
number of exits (as a function of number of passengers)2, and also of minimum
aisle width and numbers of seats abreast3.
There are aircraft for which a means of airborne abandonment is essential to the
role. This mostly includes aeroplanes that are designed for some form of aerobatic
manoeuvering—for example combat aircraft and light aeroplanes used for display
aerobatics; however the use of some form of escape system is not constrained to
those: most early flight test prototypes will use something of this nature, regardless
of long term intentions, and most gliders as the nature of glider operations (many
aircraft chasing the same thermal in a small area) causes enhanced risks of airborne
collision. Airborne abandonment can be broken into three categories:
1
Whilst not yet amended into CS.27 and CS.29 at time of writing, Europe has recently completed a
major exercise in researching and enhancing helicopter ditching and abandonment requirements.
These are found in EASA Notice of Proposed Amendment (NPA) 2016-01, “Helicopter ditching
and water impact occupant survivability”.
2
Paragraphs 803–812 of the main standards contain these requirements.
3
Paragraphs 815–817.
216 10 Crashworthiness and Escape
Some national regimes may require (or enable) approval of parachutes, and some
may mandate the use of parachutes in some operations—particularly aerobatics.
There is however no universal standard in this case.
Practical points for abandonment system approval are that any systems should be
evaluated using an actual or mocked up aircraft, and the widest possible range of
pilot anthopometric characteristics and acceptable parachute types. This should be
assessed using multiple witnessed tests (video evidence is helpful both for vali-
dation purposes and analysis), typically on the ground with an aeroplane sur-
rounded by crashmats or mattresses onto which test subjects can fall. Prior to flight
trials alone this can be done only for the test aircrew with their own equipment, but
clearly for a more general release to operate with parachutes, much more extensive
work is needed.
For the special case of parachuting aeroplanes—that is aeroplanes designed for
skydiving, there are few formal requirements: the best advice here would be to
consult with experienced skydiving operators4 and design both modifications (if
required) and approval regimes in the knowledge of established best practices. The
major sport parachuting associations worldwide tend to be very competent and are
likely to have considerable available expertise. However, aircraft modification and
approval of those modifications are likely to include:
– In-flight operable door function, and airflow around those doors.
– Attachment points for parachute static lines.
– Steps and handrails.
– Indicator lights.
– Signage stating what may, or may not, be held onto whilst exiting the aircraft.
– Training and operating documentation.
Automated abandonment
The use of ejection seats or similar systems was pioneered in the 1940s as
some aeroplanes became too high performance for manual abandonment to be safely
possible. The first systems were German or Swedish and used compressed air to eject
the pilot (wearing a personal parachute) from the aeroplane. Post WW2 both of these
lines of development ceased, but a new line of development was commenced by
British company Martin-Baker: instigated by company owner James Martin, fol-
lowing the death in a flight test by his business partner, test pilot Valentine Baker.
Whilst other companies have developed systems since (British company Folland,
Russian company NPP Zvezda and various American companies including Stanley,
Weber, AMI, Douglas and North American), the supply market has shrunk sub-
stantially to mainly Martin Baker supplying most of the world outside of the Former
Soviet Union, and NPP Zvezda supplying most companies in that part of the world.
4
The British Parachute Association’s “Jump Pilot’s Manual” is an excellent starting point for such
investigations; additionally UK CAA’s CAP660 “Parachuting” and FAA’s AC 105-2 “Sport
Parachuting”.
218 10 Crashworthiness and Escape
Fig. 10.1 A Martin Baker Mk.9 ejection seat as used in the Jaguar and Harrier. Not the latest
technology but illustrative of the basic principles of current ejection seat design
Fig. 10.2 Cirrus SR22 G3—incorporating a whole aircraft recovery parachute behind the cockpit
10.2 Escaping from an Aircraft 221
The reality is more complex. These systems have substantial potential to save
life, in extreme situations but experience has shown that they cannot be simply
incorporated into a conventional aeroplane and flown by pilots without special
training, without further considerations. Those considerations should in particular
involve specialist initial and recurrent training in decision making and safety sys-
tems, in a similar manner to that which a military pilot in an ejection seat equipped
aeroplane might require.
Ballistic parachutes have also been used well as safety devices in flight test
programmes—the author has required them for several programmes (particularly
stalling, spinning and high speed testing), and a well known incident occurred
where Cessna used one to save the test pilot’s life during development testing of the
C162 Skycatcher5. The two most mature standards available for approval of such
systems are in the UK: BCAR Section S sub-section K (intended for microlight
aeroplanes), and in the USA: FAA Special Conditions for Cirrus’ SF50 CAPS.
The FAA special conditions are the more mature of the two, but they have similar
requirements, which may be summarised as:
– System installation and use is optional.
– Non-use must not endanger the aeroplane.
– There must be safety mechanisms to prevent inadvertent operation.
– All parts of the system must be secure on the aircraft.
– The operation and maintenance of the system must be fully covered by the
aircraft’s operating documentation.
– Similar reserve factors to most other primary structure should be applied (in the
case of FAA rules, this explicitly applies this to failure energy states, rather than
forces).
That these rules have been in use in these two countries for some years, makes them
valuable precedents likely to be useable for a wide range of other aircraft. At the
same time—not unlike ejection seats in military aeroplanes—technology substan-
tially leads regulation, and the expertise built up by manufacturers in particular
should be extensively consulted as part of design and certification work for these
installations.
Historically a small number of military aeroplanes such as the F-111 in Fig. 10.3
have used cockpit “capsules” with explosive bolts to release the cockpit section and
multiple parachutes to bring it to the ground. These are not now widely used as a
solution, and can be regarded as a historical anachronism—although the basic
problems of their use are substantially the same as modern general aviation ballistic
parachute systems. Outside of the province of this book, space recovery capsules
such as the Apollo and Soyuz use similar technologies.
5
Aircraft registration N162CE on 19 March 2009 near Wichita, Kansas. This has been widely
written upon.
222 10 Crashworthiness and Escape
Statistics do not exist in a form that allows an easy breakdown by numbers of the
causes of injuries in accidents—analysis generally concentrates upon the causes of
accidents and their prevention. However, it is possible from experience, analysis,
and the narrative of accident reports to identify the main issues in crashworthiness
that must be addressed—these are listed below.
If occupants have survived the initial ground impact of a crash, the most common
subsequent cause of death is fire related. A substantial part of this has historically
been because occupants were in some way trapped within the aircraft, or were
injured in a manner which prevented them readily leaving the aircraft (in particular,
lower leg entrapment/injuries due to poor adoption of the brace position, or dis-
tortion of structure, have commonly been cited in this regard). Subsequently they
were killed by smoke, noxious fumes or fire.
Whilst this clearly emphasises the importance of ensuring that cabin structure,
seats, harnesses, etc. are designed to restrain and protect occupants, and to restrain
other objects, in the event of a crash—it is also important that the materials from
which the cockpit and cabin, seats, floor, etc. are made are designed and manu-
factured in a manner which minimizes the risk of fire—and should a fire still occur,
minimises the likelihood of severe fire and fume emissions.
Whilst strict definitions and test requirements (very rarely will any authority
accept proof entirely by analysis in this regard) will vary with application and
standard, materials used within cabins then, will normally be evaluated in terms of
10.3 Common Causes of Post Crash Injuries and Means of Prevention 223
Fig. 10.4 Air Liquide smoke hood in use during a cabin firefighting exercise
that there is sufficient ventilation to allow the majority of fumes to escape rather
than build into the cockpit. For a larger aircraft however, cockpit/cabin ventilation
is not possible and in the short term occupants are likely to have to “suffer” the
smoke or fume filled environment; eventually evacuation is provided through the
use of illuminated markings, particularly at floor level, powered by a reserve battery
system separate from the aircraft’s main power supply.
In a larger aircraft, flight crew will have access to smoke goggles and an
emergency oxygen system capable of supplying 100% (and thus fume free) oxygen;
it is unlikely that any such provision will normally be provided for passengers.
Airliner passenger oxygen systems will normally mix oxygen with cabin air, which
is life-saving in the event of cabin depressurization, but unlikely to provide much
10.3 Common Causes of Post Crash Injuries and Means of Prevention 225
Fig. 10.6 Piper PA34–200: a typical light twin with a fuel burning cabin heater
assistance if there are smoke or fumes in the cabin. This may seem harsh, but it
should be remembered that a 100% oxygen system carries a substantial weight
penalty, and also that survival of all on board is dependent upon the functional
consciousness of the flight crew—temporary incapacitation of passengers and cabin
crew, whilst undesirable, is less likely to be life threatening.
Finally it is an unsurprising requirement for most aircraft to carry some form of
smoke or fume detection system on board, the only real exception being piston
engined aeroplanes whose cabin heaters do not require fuel to be burned for
heating. At the simplest level (for example an aircraft such as the PA34 in
Fig. 10.6) this would probably comprise of a simple “black spot” carbon monoxide
detector, costing a few pounds and which changes colour in the presence of CO and
is replaced every few months. In an airliner this is likely to comprise a much more
sophisticated distributed electronic monitoring system—in particular the toilets
(where passengers who cannot cope with a long flight without a cigarette are known
to retreat: hence that regulations also normally require toilet bins to contain fire
suppression systems) and cargo compartments (since it is not unknown for
potentially incendiary items to be inadvertently or maliciously dispatched by air
transport) require reliable fume monitoring systems which will quickly alert crews
as to the risk of fire or fume hazards on board.
One of the most likely failures in the event of an emergency landing, is failure in
whole or part of the undercarriage. So, some effort must be made to understand the
nature of failure (an inspection of the reserve factors for various undercarriage
components is likely to give a reasonable clue, although for larger aeroplanes a
much more detailed analysis, and possibly test, is likely to be required).
An additional concern, and the main consideration for smaller aircraft, is that
undercarriage collapse does not cause high shock loads to be carried to the
10.3 Common Causes of Post Crash Injuries and Means of Prevention 227
occupants. Since about 1985, a great deal of research has been done into this,
following the realization that a lot of accidents which should have been very
survivable in smaller aircraft were actually resulting in severe injuries to occupants.
So, newer designs over this period have increasingly made use of energy absorbing
“sacrificial” structure and conformal foam materials. The latter in particular, using
materials such as Dynafoam (a conformal energy absorbing foam)—all of this is
primarily aimed at reducing the compressive load on the spine—which is the most
likely failure mode of the human body in an aircraft crash, and is particularly a
function of the seat occupants age (as is shown in Fig. 10.7).
The objective behind using energy absorbing structures and materials then is
clearly to minimise the loads upon the spine, which is done partly by ensuring that
the energy of impact is absorbed by any means other than via the occupant’s body,
and secondly by avoiding any form of rebound or “bounce”, which can often act to
increase shocks upon the spine. Design features which can tend to increase the
likelihood of this (and thus of injury) are trapped air, soft seat cushions, and loose
fitting harnesses. Regrettably, this does mean that the safer seat designs (with
conformal/vibration damping foams, lack of give in the seat cushions, and tight
fitting harnesses) are often less comfortable, and thus less appealing to the occu-
pants for whose benefit these safety features are provided. This results in some
difficult design contradictions6.
For larger aeroplanes the main concern is instead that of damage to the fuel
system or electrical system—potentially causing a fire. A normal assumption is that
failures will be symmetric, since the results of an asymmetric undercarriage
6
The pre-eminent researcher on spinal injuries and their prevention in light aircraft is retired
physician and glider pilot Dr. Tony Segal. He published the bulk of his work in OSTIV’s obscure
but valuable journal Technical Soaring. Those interested in his work need a copy of special
edition: Volume 32, Number 1/2, January/April, 2008 which collated and republished many of his
most significant papers.
228 10 Crashworthiness and Escape
overload are too variable and unpredictable to analyse rigorously, as well as being
very rare events.
Fig. 10.8 Robinson R44 Helicopter—designed and certified for the “fuller bodied” pilot
in safety margins for a great many crew and passengers. In the light aircraft world,
this practice is starting to change—110 kg (243 lb) is now commonly being used in
British light aircraft designs, and different values globally by smaller helicopters
such as the Robinson R22 and R44 (109 kg/240 lb and 136 kg/300 lb respectively)
are also deliberately using higher seat weights (Fig. 10.8). Whilst it may not yet be
legally required, the author would recommend to any design or airworthiness
engineer using a value of at-least 110 kg. This is likely to future proof designs, and
probably avoid future litigation when a more full-bodied passenger suffers injuries
during a hard landing: this will happen!
The geometry of the harnesses is also important—poor geometry, particularly of
upper torso restraints can cause death or serious injury in the event of crash loads; a
good example of this was a Whittaker MW6-S aircraft which suffered a loss of
control on landing at Newnham in Hertfordshire (UK) in March 1999, leading to
death of the pilot7. The cause of death was found to be that the harness, subjected to
an unauthorised modification (presumably for comfort) broke the pilots neck.
Figure 10.9 shows the geometry which has been found from experience,
although the science behind it is difficult to trace, to offer the best retention for seat
occupants—whether wearing a 2, 3 or 4 point harness (a 2-point harness is com-
monly referred to as a lapstrap, a 3-point is similar to that normally found in a car);
it is unlikely that any aircraft will be successful in achieving certification if it does
not conform to these dimensions. Also, whilst lapstraps for passenger seats remain
the norm (since passengers can assume a “brace” position whilst aircrew do not
have this luxury), the only class of aircraft likely now to receive certification with
only a lapstrap for the pilot’s seat(s) is a flexwing microlight—which normally only
7
UK Air Accidents Investigation Branch Bulletin No: 8/99 Ref: EW/C99/3/1.
230 10 Crashworthiness and Escape
Fig. 10.9 Preferred geometry for aircraft occupant restraints. (from CS-23)
has a full harness in the rear seat, if at-all, to allow use of full body movement in
control.
Up until the 1980s, virtually all aircraft were manufactured from a combination of
aluminium, magnesium and steel alloys, and wood, with small proportions of
natural fabrics and rubber or processed oil-based products. Following an impact and
possibly subsequent fire, these would break up or burn in a reasonably predictable
manner, with the major risk to occupants being the consequences of the fuel fire,
but relatively little overall environmental impact, post-crash.
Glass Reinforced Plastic (GRP) was first developed in the 1940s, and its initial
use in aircraft was for the major structure of gliders and smaller parts of some larger
aeroplanes (most notably Concorde) in the 1960s. As confidence was gained in this
new material, and it was realised that complex and aerodynamically efficient shapes
that weren’t previously possible could be readily achieved, increasingly whole light
aeroplanes, and substantial parts of larger aeroplanes became manufactured pri-
marily of these new composite materials which developed from the original GRP—
the primarily composite airliner is an inevitability—indeed the new Airbus A350
promises to be that, whilst the primarily composite fighter already exists in the
Eurofighter Typhoon (Fig. 10.10), Lockheed F-117 and others.
However, the composite matrix used in these materials, and also increasingly the
fibre-reinforcing materials can present very real problems in a post-impact envi-
ronment. The fumes of any fire can be extremely toxic, whilst the small particles
emitted from the breakup of the material can be at-best irritants, and potentially
present very serious risk to health—in addition these materials often store a great
10.5 The Challenge of New Materials 231
Fig. 10.10 The primarily composite Eurofighter Typhoon: 70% of the surface area is carbon fibre
composite
deal of energy which can be released during burning. The range of materials in use
now under the common label “composites” is of-course very wide, but the majority
of them offer some form of elevated post-crash risk. This is not a theoretical risk:
particularly in the military environment there have been a number of documented
cases of rescue crews suffering post-task medical problems (including histamine
reaction, respiratory problems and headaches) after attending an accident to an
aeroplane with substantial composite structure, this has included accidents to
F/A-18, Harrier II and F-117 aeroplanes. Experiments have shown that the main
volatile compounds emitted are likely to be benzene-based, whilst the main airborne
particulates are likely to be primarily carbon-compound particles in the range
2–7 lm across. The actual post-burn composite structure tends to be only
mechanically hazardous. It may not be possible to protect aircraft occupants from
such emissions (although the use of personal smoke hoods is a possibility as
onboard safety equipment, this hasn’t as yet found great favour—mainly for eco-
nomic reasons since as has been previously discussed the evidence in favour of their
availability is strong). It is nonetheless essential that rescue crews are protected
from these emissions not least because short term incapacitation will prevent them
from rescuing aircraft occupants—this has received considerable attention in recent
years, and all airport rescue crews (and most general fire and rescue crews) should
now be provided with adequate body and respiratory protection.
A related point of growing concern is electrical energy storage on board aircraft.
Historically this was a trivial issue of continuous generation and a small number of
liquid containing (e.g. lead-acid) batteries used for startup buffering in flight, and
emergency requirements. Increasingly aircraft both require, and carry large quan-
tities of energy storage: the present most popular technology being lithium (lithium
ion or lithium polymer) based, and some small aeroplanes now are entirely electric
232 10 Crashworthiness and Escape
such as the Slovenian Pipistrel Alpha Electro currently entering service. The search
for more efficient electrical energy storage has created its own problem—a high
energy density battery is (being simplistic) a bomb. The problems of thermal
runaway in lithium based batteries are well documented and have created under-
standable growing restrictions on their carriage, and mandatory training for cabin
crews in how to fight fire in them. Aggressive cooling is usually the best way to
solve a fire once it has broken out, but the most important work must be on safe
storage and the prevention of overload or damage which might cause such thermal
runways. Clearly a crash is the most severe circumstance and at present, this issue is
poorly understood—in particular being essentially ignored by electric light aircraft
designers and certifying teams. That cannot be a sustainable position, and the next
few years are likely to result in significant learning in this area.
Chapter 11
An Introduction to Flying Qualities
Evaluation
Abstract The flying qualities of an aircraft can be split into two components—
handling and performance. Handling qualities determine how responsive the air-
craft is to perturbations from regular flying conditions, and how easily, and safely, a
pilot can recover from such a perturbation. This chapter will concentrate mainly on
handling qualities, and in particular the civil codes that define minimum handling
standards and how these relate to preferable handling qualities.
should be, since high workload makes human errors more likely and, perhaps,
inevitable.
The term flying qualities breaks down into two categories: performance, and
handling. Largely they can be treated separately (although there are exceptions
where the single term P&HQ: Performance and Handling Qualities becomes a
single subject—in particular stalling characteristics, and the asymmetric behaviour
of multi-engined aeroplanes). This section will concentrate primarily upon handling
qualities—partly because of the vital importance of that subject, and partly because
there are already many excellent textbooks on aircraft performance.
The term “stability” is arguably misleading, and has several possible meanings in an
aeronautical context—the word should never be used on its own, always requiring
qualifiers.2
Static stability describes a parameter’s tendency, or not, to move initially
towards the original neutral state after being disturbed. Consider an aircraft trimmed
to a speed x kn which is disturbed without re-trimming to x + 10 kn. If the air-
craft’s immediate response is to reduce speed back towards x again, then it has
positive static (speed or pitch) stability. If the aircraft remains at x + 10 kn then it is
has neutral static speed stability; whilst if the aircraft continues to increase its speed,
it has negative speed (or apparent longitudinal) static stability.
Secondly dynamic stability describes a parameter’s long term behaviour after
disturbance. Considering the case above, if the long term behaviour of the aircraft’s
speed is to eventually stabilise back at speed x, then it has positive (or convergent)
dynamic speed stability. If it oscillates indefinitely within the range x ± 10 kn, then
it has neutral dynamic speed stability; if the oscillations increase in amplitude then
the dynamic stability mode is negative (or divergent).
Consider a fictional scenario of a ball bearing, which is released at the side of a
curved channel, friction being disregarded (Fig. 11.1).
The ball bearing’s position has positive static stability about a neutral state at the
bottom of the channel but, because of the lack of friction, the dynamic stability is
neutral.
Now add a fluid, introducing friction, in the bottom of the channel (Fig. 11.2)—
this will slow the ball bearing each time it passes through the bottom of the channel,
reducing the amplitude of oscillations and eventually causing the ball bearing to
stop at the bottom. Whilst static stability is unchanged, dynamic stability has
changed from neutral to positive (or convergent).
2
As a rule of thumb: any engineer or pilot using terms such as “the aeroplane was very stable”, or
“stability is high” (or poor/weak) does not understand stability and control and their opinions
should not be relied upon in this context.
11.2 The Essential Terminology of Aeroplane Stability 235
Now invert the channel and place the ball bearing on top (Fig. 11.3). If there is
any disturbance whatsoever, the ball bearing will fall off, and will not return to the
top—it therefore has negative static stability, and also negative dynamic stability.
This shows that to have positive or neutral dynamic stability, a parameter must
have positive static stability. Of course—positive static stability can still occur in
combination with negative dynamic stability, it simply does not lend itself to this
analogy.
Every parameter in flight mechanics can be described in this way; however,
further terms are required. Firstly consider whether we are using “actual” condi-
tions, or “apparent” conditions. Actual conditions are those which are experienced
236 11 An Introduction to Flying Qualities Evaluation
A fundamental problem in flight testing is, and always has been, how to quantify
pilot opinion. If you took an aircraft and asked three different test pilots to fly the
same manoeuvre at the same conditions in it, their narrative descriptions of the
merits of it will inevitably be different as a result of their different writing styles and
experience levels. For somebody who is not another test pilot, and even if they are,
to analyse the semantics of those reports and reach a meaningful conclusion is
extremely difficult. However test pilot opinion is nonetheless vitally important in
analysing the flying characteristics of any aircraft and needs to be effectively used.
Clearly the engineers and key decision makers in any project will be much better
at their jobs if they do have a good understanding of how the aeroplane will be
operated, but this does not remove that problem. In the 1950s George Cooper at
NASA developed and published in 1957 the “Cooper pilot compensation rating
scale” which made a major step forward in providing a mechanism for putting
numbers on pilot opinion of the qualities of an aircraft. Over the 1960s Cooper
continued to develop this method in collaboration with researcher Bob Harper of
Cornell Aeronautical Laboratory, and they refined the method which was published
in 1969 and became known as the Cooper-Harper scale.3 That scale, shown in
Fig. 11.5 rapidly became a worldwide standard for the purpose, and remains
unmodified and in routine use today. It places a score of 1 ! 10 on the pilot’s
opinion about various aspects of his or her ability to perform a flying task.
Other scales have come into being since Cooper-Harper, for example
NASA-TLX and the Cranfield Pilot Workload rating scale; however all of these are
basically special-purpose derivatives of Cooper-Harper, which remains the baseline
method.
At the root of use of the Cooper-Harper scale is the construction of flying tasks
which can be assessed—such tasks need to be very precisely defined. So for
example, tasks such as “fly straight and level”, “land”, or “maintain 250 kn IAS”
are not useful.
Consider however the task of flying straight and level. A typical hand-flying task
might be to fly an aeroplane in the cruise configuration at 8000 ft, straight and level,
within ±100 ft, ±2° heading and whilst maintaining an airspeed of 250 kn ± 10
kn. We can even refine that task a bit more finely by making ±50 ft the desirable
altitude tolerance with ±100 ft essential; similarly ±10 kn as desirable airspeed
tolerance, with ±20 kn essential and finally heading as ±2° desirable and ±5°
essential.
So, the test pilot then takes the aeroplane to 8,000 ft, trims it out, and then
proceeds to try and handfly accurately at 8,000 ft and 250 kn on a convenient
3
Their work is explained in: G. Cooper and R. Harper. The use of pilot rating in the evaluation of
aircraft handling qualities. Technical Report TN D-5153, NASA, April 1969, which is essential
knowledge for anybody working in manned aircraft handling qualities.
238 11 An Introduction to Flying Qualities Evaluation
heading. He or she is then flying three separate Cooper-Harper tasks, all of which
are relevant to the reporting. So the test pilot’s report may then read something like:
It was reasonably straightforward to maintain an altitude of 8,000 ft within ±50 ft once
trimmed and only occasional corrections were required to the throttle (HQR2), airspeed also
was reasonably constant although over several minutes it tended to vary through about
±15 kn unless it was monitored and small corrections occasionally made (HQR3);
however, the aeroplane despite being trimmed suffered a slow, neutral but moderate
amplitude Dutch Roll through about ±4°of heading, and ±2° could not be achieved of the
target heading of 250° (HQR4).
The assessing airworthiness team then will treat this as three separate tasks
within the basic straight and level flying manoeuvre: height hold, speed hold, and
heading hold, but it is appropriate to try and achieve all three at once as this is a
realistic flying task. The first two are in different ways allied to aLSS, whilst the
third is related to directional stability.
The next stage of analysis is normally then to also examine the background of
the pilot(s) making these comments. So in this case for example we might treat
these scores differently depending upon whether they are for a test pilot new to the
aircraft type, and one with 2,000 h on the aeroplane. It is almost never appropriate
to manipulate HQR (Handling Quality Rating) scores obtained using
11.3 The Use of the Cooper Harper Pilot Compensation Rating Scale 239
Fig. 11.6 Gratuitous name dropping. Left to right, Bob Harper, the author, George Cooper. At a
test pilot’s dinner in 2012
Abstract Longitudinal stability and control links control in speed, altitude and
normal acceleration. This chapter relates to the stability seen by the pilot, as
opposed to the aerodynamic characteristics of the aircraft. This Chapter will discuss
static and dynamic stability, and minimum airworthiness codes that define
acceptable ranges and responses. Definitions of stability and control terms are
discussed, as are the usual tests for determining whether an aircraft’s response is
airworthy.
Longstab (as it is commonly known) or LSS is very important but often poorly
understood. This is perpetuated by the use of the term “stability” which is tech-
nically misleading, and by many aerodynamic textbooks, which favour theoretical
analysis over practical understanding.
This section will refer to “apparent” LSS—that is the stability as seen by the
pilot on the controls, not necessarily the aerodynamic characteristics of the aero-
plane itself. This is valid since aircraft are more likely to be damaged by handling
problems via cockpit inceptors, than by aircraft misbehaving without pilot input.
Most modern aircraft use stability augmentation to alter the relationship between
cockpit feel and actual aerodynamics. The design and approval of such systems is a
complex and highly specialist subject, for advice on which the reader must largely
look elsewhere.
Apparent LSS is about the relationship between control input and aircraft air-
speed response: impacting upon height change, and thus flightpath stability.
However, the classical textbook longstab graph of input vs. response which shows a
continuous curve through the origin is misleading. Before explaining the mecha-
nism by which apparent LSS works, it is helpful to explain some standard terms, as
shown in Table 12.1.
To understand longstab from an airworthiness perspective, we ust quantify
FCMC (Flight Control Mechanical Characteristics); specifically BO + F, Freeplay
and TSB. The tests for them are shown in Table 12.2 (all test pilots will be familiar
with these, although where possible they are all best executed with the aid of an
observer—who may well be an engineer, and/or flight test instrumentation).
Table 12.2 Tests for longitudinal flight control mechanical characteristics (LFCMC)
Quantity Test
Breakout and Establish trimmed flight (if sortie time is limited this can be done in the
friction climb) and apply force to the stick until the aircraft nose moves relative to
the horizon. To be carried out in both senses, recording the force to make the
aircraft respond
With irreversible controls, this test can usually be performed on the ground
(although airborne confirmation is wise), however with reversible flight
controls this MUST be carried out airborne, because of likely characteristic
changes as the airframe deforms under flight loads
Freeplay A similar test to BO + F, except that the displacement of the sticktop or grip
position is measured
Trim speed band The aircraft is trimmed, and the trim speed recorded. The aircraft is then
accelerated by about 10 knots using the stick before gently releasing the
control. The aircraft should slow before stabilising at an airspeed, which must
be recorded. Repeat in the opposite sense. The difference in results is TSB
12.1 Apparent Longitudinal Static Stability, CG Range … 243
Having found longitudinal FCMC, aLSS can be measured. There are two ways
in which this is measured at the cockpit, and both are significant. These are
apparent stick fixed LSS, and apparent stick free LSS.
The aircraft must be trimmed at constant speed in the configuration and climb/
descent condition required: airspeed must be constant. Airspeed is varied in suitable
increments (for most aeroplanes, *10 knots: perhaps 20 knots for a fast jet or
airliner or 5 knot for very low performance aircraft) away from trim, with in each
case the stick (or yoke or bar) force and displacement being measured. For a “quick
check” measurement, this can be done just moving away from trim (such as con-
firmation of correct characteristics on a new production aircraft). But, for a prototype
programme at-least ±15% of trim speed, paused, and then returned through trim to
the other side before returning back to the trimmed condition again. Power, propeller
pitch, or trimmer settings must not be adjusted.
Major test programmes will have force and displacement instrumented on the
aircraft, but “line” aircraft, production aircraft, or a low budget projects (including
almost any general aviation programme), require handheld instrumentation making
it necessary to carry out the test twice, because of the workload of controlling the
aircraft, measuring stick forces, measuring stick displacements, and maintaining
reasonable airmanship at the same time. Some marking up of the cockpit to provide
reference data is advisable.
Once the data has been obtained, it must be plotted on two graphs, demon-
strating apparent stick-free and stick fixed LSS. Illustrated in Figs. 12.2 and 12.3
are sample graphs although in reality, they are not this neat, as shown by Fig. 12.4.
Several points should be noted with regard to these graphs:
• The measurement of gradient either at the middle or end points of the curve can
be unreliable. If a precise measurement of gradient is required, testing must be
continued beyond the range for which the gradient is needed.
• Non-linearities are not necessarily a bad thing, in-fact, if the gradient steepens at
the high and low speed ranges, this protects the aircraft from mishandling. They
must however be considered and understood if they occur.
• If the aircraft in question has non-linear PEC (ASI Errors)—which is quite
common, particularly with smaller aeroplanes, helicopters at any speed, or any
aeroplane close to the stall, then graphs, as with Fig. 12.4, must be plotted
against CAS, not against IAS. Otherwise, non-linearities can be seen which are
actually ASI non-linearities (or alternatively, LSS non-linearities could con-
ceivably be masked by PEC). If in any doubt, plot against CAS, not IAS.
We plot both stick-free and stick-fixed LSS because pilots respond to stick force
and stick displacement in different ways. In most circumstances, most pilots will fly
primarily by reference to pitch feel—so control forces are the most important factor.
However, control displacements remain essential and must be of an appropriate
12.1 Apparent Longitudinal Static Stability, CG Range … 245
Fig. 12.4 Apparent stick fixed LSS graph obtained by the author whilst testing a Reims-Cessna
F152 (illustration of a similar aircraft in Fig. 12.5. This was part of a research project whose results
are in the public domain. “A study of fatal stall or spin accidents to UK registered light aeroplanes
1980–2008”, published by the General Aviation Safety Council (GASCo) in 2010 provides an
opening to it, with more detail in various academic papers, mainly by the author and Dr. Mike
Bromfield, now of Coventry University)
order, otherwise the pilot will either become distracted by the need for huge dis-
placements, or will find themselves unable to accurately control the aircraft. The
F-16 (Fig. 12.6), was designed originally to fly by stick force alone with no dis-
placements, the prototype was very nearly lost on its first flight because the test
pilot could not make sufficiently small corrections to his attitude. The author met a
similar problem with the prototype X’Air UK (Fig. 12.7), which at aft CG had an
acceptable stick-fixed gradient of about 0.1 daN/knot, but a stick-free gradient of
about 0.1 mm/knot, combined with an approach speed of 50 knots, made flying an
Fig. 12.7 X’Air Mk.2 Falcon a slightly later development of the originally problematic Mk.1
12.1 Apparent Longitudinal Static Stability, CG Range … 247
accurate approach virtually impossible. The solution was to reflex the ailerons by
7.5°, and move the aft CG limit forwards by some 1.5”, providing a gradient of
1 mm/knot at the new aft CG limit, and thus good speed control over the aircraft.
Civil standards rarely define minimum or maximum aLSS values, and the assessing
team must decide what are acceptable characteristics for the role of the aircraft.
A very low stick force gradient is entirely appropriate for an aircraft designed for
air-combat or for display aerobatics, but inappropriate for most training or transport
aircraft. (Military standards do define values, and these may be useful for civil
projects also.)
For training aeroplanes, setting the aft CG limit to give a worst gradient of
1 mm/knot (stick free) or 0.1 daN/knot (stick fixed) is usually in the right order.
Differing levels of aLSS will be appropriate at aft CG depending upon role and
control geometry; for example a transport aeroplane is likely to require a steeper
gradient to be safely controllable, whilst a fighter aeroplane wants a more shallow
gradient for the required level of manoeuvrability (plus a fighter pilot’s flying skills
are good to exceptional: not necessarily true of all other pilots). A further and
perhaps even more important consideration is the total pull force to the point of stall
—research suggests that aircraft with less than around 10 daN (2 lbf) of pull to stall
in the approach or go-around configurations are prone to inadvertent stall, poten-
tially leading to fatal accidents. This is an area where Test Pilots opinion: developed
in understanding of the role of the aircraft, is more important than hard numbers—
although numbers are essential to support experts’ opinions.
The forward CG limit is easier to define, but again will be a function of the role
and geometry of the aircraft; here the aircraft must possess adequate pitch control to
rapidly unstall the wing, to execute an efficient roundout flare and landing, to
recover from a Vdf dive safely (or a dive created by 18 daN (40 lbf) push in many
standards) or to carry out any other manoeuvres specifically required by the type
(for example, a tailwheel aeroplane must be able to execute a 3-point landing).1
Formal requirements are vague, leaving the assessing team to make their own
decisions of acceptability. A good practice, in case of limited misloading, is to find
the acceptable fwd and aft CG limits and then to impose operational limits 5–10%
of total CG range towards mid CG at either end.
This makes no mention of the standard aerodynamic term “static margin”, or of
any form of theoretical analysis of CG. This is a necessary part of the aircraft design
process, and details can be found in any good textbook on aircraft design or fixed
1
And not overload the tailplane at any permitted speed or weight—but that is a structural
consideration.
248 12 Longitudinal Stability and Control
wing aerodynamics. But, when an aircraft is presented to the airworthiness team for
certification, these theoretical CG limits will contain approximations and inaccu-
racies in their calculation. Best practice is to start at estimated mid-CG condition,
and then to cautiously work outwards through the test programme, until practical
forward and aft CG limits are proven. The same caution should be applied to testing
of kitplanes or ultralights from countries with little or no regulatory supervision of
such aircraft, such as the USA or France. Even if the CG limits have been declared,
there is seldom much available justification.
Standards do give limits for maximum transient and permanent control forces,
and usually a maximum Trim Speed Band; certification test conditions will also be
stated, although development programmes will need a wider range. Transient forces
are those due to thrust, speed or service changes, but which will then be trimmed
out (or a trimmer failure). Permanent control forces will be those which have to be
held continuously. These limits are shown in Table 12.3, but should be regarded as
absolute limits and any aircraft close to them probably has very poor handling.
There is no standard answer as to how an LSS problem can be fixed, and technically
this should be the province of design, rather than airworthiness. However,
Table 12.4 is intended to give the reader an indication of where to look for solu-
tions, once a problem has been identified.
When any service of the aircraft is selected or changed, then the change in pitch
must be identified, and proven not to have an adverse effect upon the aircraft.
Typical services are:
• Power or thrust changes
• Changes in flap setting
• Gear extension or retraction
• Airbrake selection or retraction.
The aircraft must be trimmed, the service selected, then constant airspeed main-
tained, by making any necessary adjustment at the pitch inceptor/stick. Stick-forces
must be measured, and not exceed the temporary pitch force requirement of the
standard, then to be trimmable out, to a value below the continuous pitch force limit
of the standard.
Some control forces are acceptable or even good—one example of an aircraft
which should have had such a force and didn’t was the SEPECAT Jaguar
Table 12.3 Guidance on aLSS minima found in the main civil airworthiness requirements
12.3
Minimum LSS slope Trim speed if no Where are test Maximum TSB Temporary Continuous
trimmera conditions found? pitch force pitch force
Old Part-23 (143, 145, Gradient not defined, positive within ranges 1.4 VbS1 with gear up 145, 161, 175 ±10% 60 lbf (stick) 10 lbf
161, 171, 173, 175) and in configurations given in 175 or down, flaps up ±7.5% (cruise conditions, 75 lbf (two
In general, within 15% of trim at all 1.3 VS0 with gear and commuter category only) hands on
conditions flaps down wheel)
Climb speed in climb 50 lbf (one
config hand on wheel)
VNO, clean, no power
3° approach, 1.4 VS1
and VREF
New part 23 (2145) Aeroplanes not certified for aerobatics must have [positivee] static longitudinal stability in normal operations and provide stable control feedback throughout
the operating envelope. No aeroplane may exhibit any divergent longitudinal stability characteristic so unstable as to increase the pilot’s workloadf or create a
Fixing Apparent LSS Problems
dangerous condition
Part-25 (143, 145, 161, As old part 23, with greater detail in definition of test conditions
171, 173, 175)
CS-VLA (143, 145, Not defined Lower of 0.9VH, VcC 161, 175 ±10% trim CAS 20 daN 2 daN
161, 173, 175)
CS-22 (143, 145, 161, 1 N/10 km/h 1.2 ! 2.0 VdS1 145, 175 Greater of ±15% or 20 daN 2 daN
171, 173, 175) ±15 km/h
British civil airworthiness standards
Section S (143, 161 Not defined 1.3 ! 2.0 VS1 175 ±10% trim CAS 20 daN 2 daN
173, 175)
Section K (2–8, 2–9, 2– Not defined 2–8, 2–9 Not defined No limit 45 N/10 lbf
10)
a
Must lie somewhere within this range
b
Part 23 assumes that a trimmer is fitted. The conditions given are those which must be trimmable to. The table here is necessarily simplified, and readers should read the small print in part
23 before making anything but general conclusions or plans
c
VLA also assumes that a trimmer is fitted and requires that the aircraft can be trimmed at any condition between this and 1.4 VS1
d
Part 22.173 also has limitations upon minimum and maximum possible trimmed speeds, as a function of VS1 and Vne
e
The author has inserted “positive” here, which is not stated in the standard, but is the only reasonable interpretation. It seems most likely also that this implies apparent stick-fixed static
stability, but this is also not defined in the standard
f
This reference to not increasing the workload is unhelpful since it is again vague, and it is likely the use of role-relatable Cooper-Harper type tasks will be needed to demonstrate
acceptability
249
Table 12.4 Possible fixes for LSS problems
250
Problem ! Good Unable Unable to Neutral Neutral Good stick Lack of nose-up Excessive
Possible Fix# stick-fixed, to flare maintain LSS at LSS at free, weak or nose-down control
weak stick approach speed high speed low speed stick fixed authority forces
free control
Elevator centring X X X
springs
Move fwd CG X
limit aft
Change pitch X X
gearing
Move aft CG X
limit Fwd
Reflex ailerons or X X
trailing edge
Look for X
excessive
freeplay
12
Fit a larger X X X
tailplane
Change shape of X X
cockpit control
Change rigging X X X X
angle of tailplane
Alter washout X X X X
(delta or swept
wings only)
Longitudinal Stability and Control
12.3 Fixing Apparent LSS Problems 251
(Fig. 12.8), which showed only minimal handling effect of airbrake selection—
arguably a poor characteristic in an aeroplane where a pilot will have his head out
the cockpit and may not notice an inadvertent selection or deselection immediately.
Considering power: an aircraft that pitches nose-up with a reduction in power
may be deficient if the stall will be approached following a rapid (e.g. unplanned)
power reduction. In the Eurowing Goldwing (Fig. 12.9) closing the throttle requires
a large forward stick movement to prevent an inadvertent stall and if the aeroplane
does stall, it will not recover without thrust due to poor elevator authority combined
with high nose-up trim, at those low speeds.
There are two main longitudinal dynamic modes, known as the short and long
period modes. Both influence an aeroplane’s controllability and fitness for purpose.
Airworthiness standards, particularly civil standards, often consider them in only
limited depth. So, airworthiness evaluators must understand the role of the aircraft
and evaluate LDS (Longitudinal Dynamic Stability) in that context.
The short period mode is an oscillation in angle of attack (AoA or a) about the short
term flight path. It is essential to the successful control of an aeroplane—influencing
the ability to adjust pitch attitude. This affects in all aeroplanes the stall and/or stall
recovery, in combat aircraft the ability to accurately track either a stationary or
moving target, and in most aeroplanes pitch control in landing or go-around.
SPO is predicted through analysis described in numerous textbooks and papers;
however, acceptability must be evaluated in flight test, at a range of conditions and
manoeuvres. Two test techniques—the pitch doublet and the sinusoidal stick pump
(or SSSP) the latter also referred to as a frequency sweep. Either may be automated
or flown manually.
The pitch doublet is flown by making an approximately sinusoidal pitch control
input of small or moderate magnitude (if in doubt starting small then increasing the
input magnitude) whilst in a relevant test condition—for example a gear down/full
flap (called “configuration land”, or just LAND) simulated approach at VREF, a
gear-up flapless configuration (called “configuration cruise”, or just CR) at a
cruising or combat speed and power setting. Test pilot opinion is valuable in
defining test conditions, as well as executing and analysing the tests. Where
automatic recording of pitch attitude and/or a is available then this is ideal, but
much information can still be observed by recording the obvious characteristics—
typically by timing the cycles (thus giving the frequency) and if dynamically stable,
counting the number of cycles for the oscillation to become imperceptible to the
pilot. A neutrally or negatively damped SPO is harder to quantify, apart from
frequency, but it is usually unacceptable, so a simple statement of frequency and
unacceptability is likely to trigger rectifying investigations.
The SSSP requires care as incorrect application can cause aircraft damage.
Sinusoidal pitch inputs are made at a very low rate—perhaps 0.1 Hz; then slowly
frequency increased whilst observing aircraft response. The aircraft will slowly
respond at a similar rate to the inputs at very low frequency, whilst at very high
frequency response will be minimal. At the SPO frequency, the aeroplane should
respond in antiphase to the pitch inputs: as the stick is pushed forwards the aeroplane
pitches nose-up, and vice versa. This identifies the SPO frequency, and the mag-
nitude of aircraft response gives empirical indication of damping. An aeroplane with
12.4 Longitudinal Dynamic Stability 253
a poorly damped SPO at around the frequency used during high gain handling tasks
such as fine course adjustments when following an ILS, or correcting to remain in
the basket during air-to-air refuelling (AAR) for military aircraft can be extremely
problematic—indeed most modern military aircraft will have specific AAR control
laws to avoid such issues, but the test and evaluation load for evaluating these
conditions is substantial. Such deficiencies cannot be properly evaluated without
high rate automated recording of aircraft motion and control inputs as well,
probably, as external cameras. Role-relatable specialist test manoeuvres (such as
high gain tracking or condition acquisition tasks) are always needed and automated
precise calibrated automated control inputs, will usually be required for fly by wire
aeroplanes.
The author has experienced flutter in either the elevator or the pitch trimmer at a
frequency co-incident with a poorly damped SPO. This is highly hazardous and an
aircraft can destroy itself in a few seconds from flutter onset. Fortunately, it is also
relatively easily solvable by modification of control surface spring or friction
characteristics.
24,000 300
22,000 290
18,000 270
16,000 260
14,000 250
12,000 240
10,000 230
0 50 100 150 200 250 300 350 400
Time, seconds
Fig. 12.10 Illustration of the phugoid mode in a poorly damped jet transport aeroplane in the
cruise
The phugoid, whilst almost always present, is a pure nuisance mode, not useful
in any way to the aeroplane. So, ideally the phugoid damps to no significant
perturbation in half a cycle—called “critically damped” or “deadbeat”.
Figure 12.11 is from a well adjusted training aeroplane, showing initial deliberate
perturbation, then two phugoid cycles as airspeed and pitch attitude return back
towards the original trimmed condition.2
The Case of the Sioux City Crash
On July 19 1989, a Douglas DC10 en—route from Chicago to Denver was en-route
from Denver to Chicago suffered an uncontained failure of the No. 2 engine, which
caused failure of multiple on-board systems, rendering the aircraft uncontrollable.
Through impressive teamwork, the crew managed to maintain adequate control
over the aircraft and navigate to Sioux City where a landing was effected.3 In the
final part of the flight, in the words of Captain Al Haynes,
2
This apparently shows an aeroplane trimmed to a very slow 55 knot—in practice it was flying
about 85 knot, but this is GPS data for an aeroplane flying into a strong wind—a useful illustration
of the importance of understanding all aircraft instrumentation and what it is really saying.
3
The Sioux City air disaster, also referred to as United Airlines Flight 232 is widely written upon,
as well as there being a TV movie “A Thousand Heroes” starring Charleton Heston and James
Coburn that celebrates the story. This case study has only referred to one very narrow aspect of a
much more complex history that is widely studied and published upon.
12.4 Longitudinal Dynamic Stability 255
that picture in the air is very deceiving. It looks like we have everything pretty much in
control. We were starting a down phugoid, and starting a right bank, 300’ in the air. And we
just, that’s where our luck ran out. We just ran out of altitude, trying to correct it. We had
the time in the air, trying to correct it. But that close to the ground, we didn’t have time. In
an attempt to stop the phugoid and the turn, Dennis added power, and unfortunately the left
engine spooled up faster than the right, the first time in the day we noticed that it happened,
and the bank increased. And in four seconds, we went from four degrees of right bank, to
twenty degrees of right bank, and hit the ground.
This portrays a more complex picture than “just” a phugoid, but does illustrate that
control and ability to damp out the phugoid mode is a significant safety issue—
failure to maintain an entirely predictable flightpath during the final approach and
landing, creates a high probability of aircraft damage and associated loss of life—as
occurred at Sioux City.
If the pilot of an aeroplane pulls back onto the control column, normal acceleration
(“g”) will increase at-least briefly. In a turn, a pilot must pitch up as well as bank to
maintain a level turn. If only a very small pull force was required to pull significant
“gs”, then it would be easy through mishandling to overstress the aeroplane. So,
airworthiness standards require a significant pull force to achieve significant normal
acceleration.
256 12 Longitudinal Stability and Control
Table 12.5 What do the standards say about manoeuvre stability? (Note varying units here, to
reflect what is used in each standard)
Standard Requirements
BCAR Section S (3-axis • 15 lbf (7 daN) of pull to achieve N1
aeroplanes)
CS.22 • 0.5 daN (1 lbf) of pull to achieve 45° balanced turn at 1.4 VS1
from a trimmed condition (flaps and airbrakes at most critical
condition, gear retracted)
CS.VLA • 15 lbf (7 daN) of pull to achieve N1 in the “clean” configuration
(implying gear and flaps or airbrakes retracted)
CS.23 or FAR-23—yoke • (MTOW in Newtons/10) or (MTOW in lbf/100) of pull to
achieve N1. But, not less than 89 N (20 lbf) and need not exceed
222 N (50 lbf)
Tested at 75% MCP for piston engines, or MCP for turbine
engines
CS.23 or FAR-23—stick • (MTOW in Newtons/14) or (MTOW in lbf/140) of pull to
achieve N1. But, not less than 67 N (15 lbf) and need not exceed
156 N (35 lbf)
Tested at 75% MCP for piston engines, or MCP for turbine
engines
CS.25 No requirements exist. The author would recommend using the
FAR-25 part 23 requirements
needed, requiring considerable time and planning, and possibly use of dedicated
“range” airspace.
An alternative is the “pull-up” where the aircraft is dived, then pitched up using
the stick. This creates a transient condition where control force and normal accel-
eration should be measured when the aeroplane is passing through the nominal
level pitch attitude. This is difficult to fly accurately, but particularly for display
aerobatic or ground attack aeroplanes valid and role-relatable.
Data recording can be manual, using a handheld stick-force gauge and a panel
mounted g-meter, but where possible use automated recording of control forces,
pitch attitude, airspeed and normal acceleration to reduce test pilot workload and
improve data analysis.
Ultimately what is required is a graph of normal acceleration against stick force
at each of the critical conditions of weight, CG and configuration. This will usually
be linear. Figure 12.12 shows typical lines which might be obtained from a part 23
aeroplane with a 3.8 g normal acceleration limit—indicating “good” (it will be very
hard to overstress), “acceptable” (compliant with the standard but relatively easy to
overstress), and “unacceptable” (relatively easy to overstress, and also
non-compliant) polars.
A significant role relatable issue concerning manoeuvre stability is the “manikin
effect”. This reference to a manikin implies that a pilot’s body will move in direct
response to changes in normal acceleration, and thus that where the pilot is holding
the controls, particularly the pitch control, that body movement may in-turn create a
control input. This generates an aircraft response which may be undesirable, but is
certainly uncommanded and in extremis can induce a problematic Pilot Induced
Oscillation, or PIO that can lead to loss of the aircraft. This effect was noted on a
number of 1980s–1990s era gyroplanes where the pitch control movement had a
258 12 Longitudinal Stability and Control
significant vertical component of motion,4 but the specific term was first met by the
author when used by Test Pilot Keith Dennison to describe the characteristics of the
prototype e-Go aeroplane.5
4
These were known as “pump action sticks”. CAA Paper 2009/02: The Aerodynamics of
Gyroplanes is a useful resource, as are the AAIB reports into accidents to gyroplanes G-BIGU and
G-BXEM, all freely available online.
5
Dennison’s lecture was delivered on 17 February 2015 and is available to download from the
Royal Aeronautical Society’s website.
Chapter 13
Lateral and Directional Stability
and Control
…an extra 360° because it seemed like a good idea at the time.
I can still remember the excitement in the eyes of the company
photographer when he came up to me afterwards…. Most of all
I can remember raising the nose at the end of the roll and how
wide the runway looked at that point.
John Farley, A view from the hover
Abstract While the theoretical basis of lateral and directional stability and control
is highly mathematical and often taught in universities, the practical application is
less so. This chapter, like the previous one, again focuses on the “apparent” lateral
and directional stability, i.e. on what the pilot sees and feels. This is the method-
ology used in most flight test schools and companies when assessing lateral and
directional stability from an airworthiness perspective. Thus, an understanding of
the methodology is vital for anyone intending to work in the field.
Lateral and directional stability and control are a combined subject essential to the
safe controllability of any aircraft. It is often inconsistently understood: no two
airworthiness standards have identical sets of static lat-dir requirements. So, the
airworthiness team must decide what characteristics are desirable for a particular
aeroplane, in the light of “Role Relation”.
Descriptions of the various modes will again be limited to the apparent char-
acteristics: that is what is seen by the pilot. Many textbooks exist which describe the
theoretical basis behind the control characteristics and oscillatory modes of fixed
wing aircraft if this knowledge is required. Readers trained already in the theoretical
basis to stability and control which is often taught in universities may find the
approach used in this chapter unusual, but it is broadly the approach taught by the
major test pilots schools and used in most company flight test departments.
Lateral Stability is rolling moment due to sideslip, and Directional Stability is
yawing moment due to sideslip. Virtually all aspects of lat-dir are driven by
sideslip, so are the most significant modes and require understanding.
If a rudder (or other yaw control) input is made in isolation, this will cause a
sideforce at that part of the aircraft, generating yaw, and thus sideslip. If the aircraft
possesses positive directional stability, it will then yaw back towards balanced
flight, whilst neutral stability would mean that the sideslip angle remains constant
after the rudder pedals are released. Negative stability would mean that the aircraft
would continue to yaw away from balanced flight—a potentially dangerous situa-
tion that would require constant pilot input to correct it. The last case can occur,
although seldom deliberately.
A second effect of this sideslip is usually a rolling moment. If in response to
nose-right sideslip, the aircraft rolls right wing down, the aircraft is said to have
positive lateral stability. If however no roll occurs, the aircraft has neutral lateral
stability and finally a roll away from the sideslip (so that yaw and roll oppose each
other) implies negative lateral stability. All of these characteristics might be seen,
and none need necessarily be unacceptable, depending upon the role and design of
the aircraft.
Low or neutral lateral and directional static stability are usually poor charac-
teristics, since it is usually desirable that an aircraft, when disturbed, will return to
level, balanced flight; this is particularly true in a training aircraft where it should be
a reasonable expectation that a student who has made some handling errors can let
go of the controls and allow the aircraft to restore itself to balanced flight, or in an
aircraft likely to be used in prolonged cruising flight since a constant need for the
pilot to correct small perturbations would increase his or her workload consider-
ably. For aeroplanes which may be flown routinely in IMC (Instrument
Meteorological Conditions) ideally the pilot should not be making constant cor-
rections to maintain course when flying without visual references, most likely inside
cloud, which demands low positive lateral stability, although neutral stability may
be acceptable. However, the converse—very high lateral or directional stability—is
equally problematic: during any flight in turbulence, the constant variations in
sideslip would cause many large rolling and yawing excursions, creating a high
workload, uncomfortable flying environment. The class of aircraft where the latter
is most noticeable is tailless delta winged aeroplanes, which display considerably
increased lateral stability at high AoA/low speed—requiring pilots to maintain
relatively high speeds during flight in turbulence. There has been at least one fatal
accident to the author’s knowledge where loss of control was caused by an
ill-advised low-speed, steep climb-out in turbulence in an aircraft with high lateral
stability.
Initial testing for static lateral and directional stability is typically commenced
using a test known as a Steady Heading Sideslip (SHSS). The aircraft is established
on a safe heading, in balance flight, then progressively more rudder is introduced,
creating sideslip, which is balanced by use of the ailerons to maintain a steady
heading flight. Rudder and aileron deflections can be measured at the cockpit
control, or by use of instrumentation, and bank angle can be measured either by use
of an artificial horizon (AH, also called attitude indicator, or AI) such as shown in
Fig. 13.1, or by reference to the visual horizon. Ideally sideslip would be measured
using a calibrated vane on a boom ahead of the aircraft, however a test pilot can
13.1 Lateral and Directional Static Stability and Control 261
Fig. 13.1 Three typical aeroplane artificial horizons. Far left within a modern electronic standby
unit, main part of left image integrated into an EFIS system, right a standalone more traditional
instrument
estimate sideslip to within a few degrees by noting the aircraft heading using an HSI
(or if none is available, a compass), allowing the aircraft to swing back to balanced
flight quickly, and then noting the new heading. The difference between the two
was the sideslip angle. This method obviously would mean that the test must be
done in iterations, returning to balanced flight in between, rather than a smooth
progression to full rudder and aileron deflection.
There are some safety points to note when planning and conducting SHSS tests:
1. If carrying out SHSS at low speeds, crews must be aware of the risk of inad-
vertently entering a spin should the aircraft stall whilst a large rudder input is
applied.
2. If planning SHSS at high speeds, remember that to remain within structural
limits normally no more than 1/3 rudder or aileron should be applied above VA.
This however is an operational limit, and not necessarily absolute as a flight test
limit—nonetheless before conducting such tests, the team must prepare a plot of
maximum permitted rudder deflection versus airspeed. Structural specialist
inputs are likely to be required, and potentially specialist instrumentation—
particularly in transport category aircraft where high inertias can make limitation
avoidance particularly difficult.
3. The pitot-static system may suffer additional errors as an aircraft is sideslipped.
This problem should be assumed in all new designs until proven otherwise.
Listed in Table 13.1 are the points which should be noted when reviewing the
results from an SHSS test, all potentially having airworthiness implications which
must be considered when determining the acceptability of an aircraft.
262 13 Lateral and Directional Stability and Control
So, this one test can show much useful information about the aeroplane.
However, it is important, particularly when evaluating a new aircraft type, not to get
hung up on an academic test such as this alone. Role relatable (or Cooper-Harper,
see Chap. 11) tests, related to the intended purpose of the aeroplane are also
important. For example:
13.1 Lateral and Directional Static Stability and Control 263
1. Most aircraft will experience flight during service in moderate to severe tur-
bulence. Deliberate flight in such conditions, noting the extent to which the
aircraft tends to suffer undemanded rolling or yawing excursions, can indicate
how appropriate the lateral and directional stability values are to normal flight.
2. During a crosswind landing, lateral and directional static stabilities will affect
the handling techniques used. For example, the Lockheed L1011 Tristar
(Fig. 13.2) has reasonably low side force due to sideslip, and moderate lateral
stability—as a consequence a sideslipped “wing down” approach can be readily
flown down to the ground. Alternatively some aeroplanes may have much
higher sideforce due to sideslip, making a prolonged wing down approach
extremely uncomfortable—which by increasing the pilot’s workload inevitably
degrades safety. In those aircraft, a crabbed approach, only using the wing down
method during the round out is more appropriate.
3. In many aeroplanes it is useful to prove that the aircraft can be flown by rudder
alone—requiring a combination of good rudder power, and significant positive
lateral and directional stabilities. Any aircraft which can be safely flown and
landed without use of the primary roll control has significant safety benefits, but
also at high angles of attack some aeroplanes (particularly those with much
higher yawing than rolling inertia) can depart controlled flight if large roll
control inputs are used; the McDonnell Douglas F-4 Phantom II (Fig. 13.3),
typical of aircraft in that class is routinely rolled at high angles of attack (such as
during a combat turn) using the rudder instead of the ailerons. So far as civil
standards are concerned, directional controllability by rudder alone is a
requirement of only the part 23 standards, but good practice is that it should be
demonstrated in any aircraft.
4. Although significant rudder co-ordination is normal in most light aeroplanes (or
heavier aeroplanes in the landing configuration) during an aileron turn, it is
worth ascertaining whether any aircraft can be safely rolled using the ailerons
alone in the event of a rudder failure. If an aircraft is not controllable in this case,
Conditions to demonstrate positive Bank angles Rudder forces Roll control forces Conditions to Steerable
lateral stability in sideslips demonstrate by rudder
positive alone?
directional
stability in
sideslips
Joint airworthiness requirements/EASA certification specifications
Part 25 1.13 VSR1 to 1.23 VSR1 with flaps Sufficient to To be roughly To be roughly 1.13 VSR1 to the Not
(147(e), in most extended take-off setting show proportional to b for small proportional to b for small greater of VFE, required
177) 1.13 VSR1 to 1.23 VSR1 with flaps sideslip sideslips sideslips VLE, VFC, or
in any setting, unless divergence is unless a yaw May not reverse below MFC
gradual and easily recognised and indicatora is lesser of 180 lbf/full
controlled fitted deflection
Legacy All configurations, power to 75% Not defined Must continuously Must continuously All Required
Dynamic Lateral and Directional Stability
Part 23 MCP, speeds above 1.2 VS1 increase with b—need not increase with b—need not configurations
(147(c), (take-off config)/1.3 VS1 (other be linear but may not be linear but may not and power to
177b) configs) reverse in landing reverse in landing MCP, speeds
In landing configuration, use power configurations up to 50% configurations up to 50% above 1.2 VS1
for 3% descent MCP MCP b may be
b may be appropriate to aircraft appropriate to
role, but at least equivalent to lesser aircraft role, but
of 10° of bank./667 N (150lbf) at least
rudder force equivalent to
10° of bank
Part 22 Not defined Not defined Any reversal must not Must continuously Not defined Not
(177) require exceptional skill to increase with b—need not required
control be linear but may not
reverse
Part VLA All gear and flap positions, power Must be Must recover from SHSS Not defined 1.2 VS1 to VDF, Not
(177)c up to 75% MCP, all permitted sufficient in with rudder released in all idle to MCP required
265
Conditions to demonstrate positive Bank angles Rudder forces Roll control forces Conditions to Steerable
lateral stability in sideslips demonstrate by rudder
positive alone?
directional
stability in
sideslips
appropriate to aircraft role, but at constant May not reverse below
least equivalent to 10° of bank heading lesser of 40 daN lbf/full
deflection between
1.2 VS1 and VA
A specific test is required
at 1.2 VS1, landing
configuration, 50% MCP
Federation aviation requirements
FAR-25 1.13 VSR1 to 1.23 VSR1 with flaps Not defined To be roughly To be roughly 1.2 VS1 to the Not
13
in most extended take-off setting proportional to b for small proportional to b for small greater of VFE, required
1.13 VSR1 to 1.23 VSR1with flaps sideslips sideslips VLE, VFC, or
in any setting, unless divergence is May not reverse below MFC
gradual and easily recognised and lesser of 180 lbf/full
controlled deflection
FAR-23 All configurations, power to Not defined Must continuously Must continuously All Required
(147(c), 75% MCP, speeds above 1.2 VS1 increase with b—need not increase with b—need not configurations
177d) (take-off config)/1.3 VS1 (other be linear but may not be linear but may not and power to
configs) reverse in landing reverse in landing MCP, speeds
In landing configuration, use power configurations up to 50% configurations up to 50% above 1.2 VS1
for 3% descent MCP MCP b may be
b may be appropriate to aircraft appropriate to
role, but at least equivalent to lesser aircraft role, but
of 10° of bank./667 N (150 lbf) at least
rudder force equivalent to
10° of bank
Lateral and Directional Stability and Control
(continued)
Table 13.2 (continued)
13.2
Conditions to demonstrate positive Bank angles Rudder forces Roll control forces Conditions to Steerable
lateral stability in sideslips demonstrate by rudder
positive alone?
directional
stability in
sideslips
British civil airworthiness requirements
Section S Not defined Not defined Must continuously Must continuously Not defined Not
(177) increase with b—need not increase with b—need not required
be linear but may not be linear but may not
reverse reverse
Section Ke All speeds above 1.2 VS1, all flap Not defined Must continuously Must continuously As lateral Not
(2–8 6.2 positions (extremes and those used increase with b—need not increase with b—need not stability, up to required
2–10 4) for steady flight), gear up and be linear but may not be linear but may not MCP
Dynamic Lateral and Directional Stability
So there are three main modes—the first two (spiral and roll) being essential to the
good handling of the aircraft, whilst the last (Dutch roll) is a pure nuisance in an
aeroplane (although some research suggests that many bird species use it for rapid
airborne direction changes).
sideslip, heading and bank angle versus time can give all the data required, however
such instrumentation is not vital and the data required can be obtainable visually
(Figs. 13.6 and 13.7).
When the DR is excited, the most obvious clue is that the aircraft nose will
describe a “figure of 8” motion about the aircraft heading, as illustrated in Fig. 13.6.
This is the pilot’s best clue to the excitation of Dutch Roll, and when examining
its real world effects upon the aeroplane (“role relation”), it is what a pilot will be
considering most—can the aircraft track a target sufficiently accurately—that target
potentially being an enemy aircraft, a visual or instrument approach, a runway
centreline, or some other appropriate mission task. If there is a problem with the
DR, then one must analyse it either using automatic data recording, or visually
acquired data. To do so visually, the best sighting point is out the side, where a
Fig. 13.7 llustration of side view from aircraft during Dutch roll motion
270 13 Lateral and Directional Stability and Control
reference point should be found on the horizon which can be lined up against a
wingtip, or other fixed point on the aircraft (if all else fails, a chinagraph mark on
the side of the canopy). This will be seen to describe an oval (Fig. 13.7).
If a Dutch Roll problem is identified, then this test, with the DR being excited by
a rudder doublet is most useful since it gives the ratio of roll:yaw in the oscillation
(which is a reasonable approximation to the ratio of lateral to directional stability,
since it is these stabilities that drive the mode); by timing and counting cycles until
the DR apparently damps out, the frequency and damping ratio can readily be
estimated:
2p:No:Cycles
xN ¼ ð13:2Þ
Time
Thus the damping ratio may then be estimated by feeding this value of the fre-
quency into the Dutch Roll characteristic equation:
S2 þ 2fxn S þ x2n ¼ 0; ð13:3Þ
and solving it. (Where automatic data recording is available, such as shown in
Fig. 13.8 this should be used, but the eventual analysis is substantially the same.)
However, such analysis is of limited usefulness in airworthiness evaluation.
Engineers should always be guided by the opinions of the test pilot concerning
whether DR is a problem or not. Terminology that may be used, particularly
although not exclusively by a less technically qualified pilot, might include
“snaking” (referring to DR primarily in yaw) or “wing rock” (referring to DR
primarily in roll). Less often, Dutch Roll may be excited by a roll input, leading to
undesirable variations in roll rate through a rolling manoeuvre. Generally, a good
aeroplane’s DR mode will appear to the pilot to damp out in 1–4 cycles after being
excited by a rudder doublet.
Table 13.3 What civil airworthiness standards say about Dutch roll
Joint airworthiness requirements/EASA certification specifications
Part 25 Must be controllable, using primary controls and without requiring exceptional
(181(b)) pilot skill, at any speed between 1.13 VSR and maximum permitted
Part 23 Must be heavily damped with controls free or in any fixeda position. At any
(181(b)) speed, DR must damp to 1/10 amplitude within 7 cycles
Part 22 Must be heavily damped, at any speed between stalling speed and VDF, with the
(181(b)) primary controls fixed or free. For powered sailplanes this must be met at any
allowable power setting
Part VLA At any speed permitted above the stalling speed, Dutch Roll must damp to 1/10
(181(b)) amplitude within 7 cycles with the primary controls (1) Free; and (2) In a fixed
position
Federation aviation requirements
FAR-25 Must be positively damped with controls free, and controllable with normal use
(181(b)) of primary controls and without exceptional pilot skill, at any speed between
1.2 VS and maximum permitted speed in any configuration
FAR-23 Must be heavily damped with controls free or in any fixed7 position. At any
(181(b)) speed, DR must damp to 1/10 amplitude in 7 cycles
British civil airworthiness requirements
Section S Any combined lateral-directional oscillations occurring between the stalling
(181(b)) speed and VDF must be damped with the primary controls (a) free, (b) fixed.
These requirements must be met with the engine running at all allowable powers
Section K Must be heavily damped with controls fixed or free
(2–10 5)
a
The controls fixed section of part 23 need not be met if a Stability Augmentation System is being
used to maintain acceptable handling qualities. Para 672 contains the SAS approval requirements
272 13 Lateral and Directional Stability and Control
LV NR NV LR : ð13:4Þ
If this quantity is negative, an aircraft is likely to be spirally unstable (that is, once
disturbed the bank angle is likely to continue to increase), whilst if it is positive the
aircraft is likely to be spirally stable (that is the aircraft will naturally tend to roll
back to wings level). An aircraft with stronger directional than lateral stability is
likely to be spirally unstable, and visa versa. A highly negative spiral stability can
be dangerous, since it tends to create an aircraft reluctant to recover from a spiral
dive, which is an accelerating descent with increasing bank angle (see Chap. 15).
Weak but positive spiral stability is usually preferable. In the former case, a low
ability pilot will benefit from the ability to release the aircraft’s controls after a
handling difficulty and being able to expect the aircraft to naturally roll wings level;
the workload of an experienced pilot flying a precise flightpath will be reduced if he
or she is not constantly having to make bank angle corrections against instrument
references. Propeller driven aircraft, particularly those with relatively powerful
engines, often display asymmetric spiral stability—tending to roll away in one
direction and towards wings level in the other—the Pilatus PC12 in Fig. 13.9 is
such an aircraft. Aerobatic aircraft may prefer neutral spiral stability to aid
manoeuvrability. There is rarely a good reason to have negative spiral stability,
although it may not necessarily be unacceptable—always consider the role of the
aircraft before drawing conclusions (Table 13.4).
Notes Conditions for roll reversal test(s) Maximum time for Maximum time for 30
45 to 45° roll to 30° Roll reversal
reversal
Joint airworthiness requirements/EASA certification specifications
Part 25 CS25 and JAR-25 do not give specific requirements for the roll mode
Part 23 Take-off configuration, aeroplanes N/A 5s
(157(a)& under 2730 kg (6000 lb) MTOW
(b))
Part 23 Take-off configuration, aeroplanes Flaps for take-off, gear up, MTOP (twin N/A Lesser of 10
s
(157(a)& over 2730 kg (6000 lb) MTOW þ 200
engined aeroplanes, CEIO), 1.2 VS1, aircraft in or W 590 s
(b)) trim (W = MTOW in kg)
Part 23 Approach configuration, Flaps for landing, gear down, all engines as for N/A 4s
(157(c)& aeroplanes under 2730 kg 3° approach, VREF, aircraft in trim
(d)) (6000 lb) MTOW
Dynamic Lateral and Directional Stability
Part 23 Approach configuration, Flaps for landing, gear down, all engines as for N/A Lesser of 7 s
(157(c)& aeroplanes over 2730 kg (6000 lb) þ 1300
3° approach, VREF, aircraft in trim. or W 1000 s
(d)) MTOW (W = MTOW in kg)
Part VLA Take-off configuration Flaps for take-off, gear up, MTOP, 1.2 VS1 N/A 5s
(157(a))
Part VLA Approach configuration Flaps extended, gear down, power (idle and N/A 4s
(157(b)) PFLF), 1.3 VS1
Part 22 No significant sideslip to be Flaps in greatest en-route position, airbrakes b/3 s, where N/A
observed and gear retracted b = wingspan
in metres
(continued)
273
Table 13.4 (continued)
274
Notes Conditions for roll reversal test(s) Maximum time for Maximum time for 30
45 to 45° roll to 30° Roll reversal
reversal
Federation aviation requirements
FAR-25 FAR-25 does not give specific requirements for the roll mode
W þ 500
FAR-23 Identical requirements to JAR-23, except that weight based calculations use lb (not kg), giving formulae 1300 for take-off configuration,
þ 2800
and W 2200 for approach
British civil airworthiness requirements
Section K Take-off configuration, aeroplanes N/A 5s
(2–8 under 2730 kg (6000 lb) MTOW
6.4.1 +
Appendix)
Section K Take-off configuration, aeroplanes Flaps for take-off, gear up, MTOP (twin N/A Lesser of 10
s
þ 200
13
(2–8 over 2730 kg (6000 lb) MTOW engined aeroplanes, CEIO), 1.2VS1, aircraft in or W 590 s
6.4.2 + trim (W = MTOW in kg)
Appendix)
Section S Take-off configuration, aeroplanes N/A 5s
(147) under 2730 kg (6000 lb) MTOW
a
The special requirement in Section S for what is in effect a maximum value of sR as a function of PSS was introduced after some (early 1980s) flexwing aircraft
with comparatively low sail tension displayed very high sR values leading to inadvertent (and dangerous) excedences of Section S’s 60° bank angle limit
Lateral and Directional Stability and Control
13.2 Dynamic Lateral and Directional Stability 275
We evaluate the spiral mode (in still air) by introducing a small bank angle
(usually around 15°), usually using the yaw rather than roll control (so as to prevent
residual roll control input upsetting the test). The controls will then be released and
the aircraft response noted and timed. This will be reported as t2 or t½: that is the
time for the bank angle to double or halve. As a rule, if either value is more than
10 s, the spiral stability should be regarded as Neutral. A good training aircraft or
one designed for instrument flying should probably display 3 < t½ < 6 s, but this
should not be regarded as a firm rule.
Airworthiness standards rarely discuss spiral stability, so it will usually be
entirely up to the airworthiness team to decide upon the acceptability criteria for the
specific aircraft being examined, in light of its intended role.
Ln
PSS n: ð13:5Þ
LP
A
sR ¼ ð13:6Þ
LP
More usefully, from experimental data, sR is the time taken for the roll rate to reach
0.623PSS from initial response. The pilot will perceive sR as being the time from
initial control input—which will be a slightly longer time—perhaps 1/3 of a second
longer depending upon control mechanisation—which can be a significant time
interval when particularly crisp roll control is required such as in an aerobatic
aeroplane—although certainly trivial in an airliner. Both actual and perceived
values of sR can be extracted from a good high frequency trace, and this instru-
mented testing can be necessary exercise if pilot opinion and the theoretical analysis
do not coincide.
It is important that sR is well matched to the roll rate, and to the intended role of
the aircraft. For example, a training aeroplane with sR < 0.2 s is likely to have such
“snappy” roll response as to intimidate most students, whilst similarly an aerobatic
or combat aeroplane with sR > 0.5 s is likely to be sluggish and hard to control with
the desired accuracy. (Even then however, too low a sR can lead to “roll ratcheting”,
requiring the pilot to make short “burst” roll inputs—this author hasn’t met any
certified civil aircraft exhibiting this problem but the F-14 and F-16 fighters do
exhibit it—requiring a degree of skill and hand-to-eye co-ordination of the pilot that
could not reasonably be expected of all civil pilots).
None of the standard specify either roll rate, or sR —the roll response is defined
by minimum standards for rolling reversals from a steady bank angle, or times to
change bank angle—obviously this has to be done in both directions. This is
practical since, although instrumentation is always helpful, it means that it is never
absolute requirement and compliance can be demonstrated without an instrumented
aircraft. All of the standards (and common sense) obviously also require that there
should be no undemanded pitch departures or uncommanded high roll rates—the
characteristics of civil aircraft are such that such rolling departures are unlikely, and
are usually the preserve of military “fast-jet” aircraft whose long, thin, metal tubes
lend themselves to undesirable inertia coupling problems—the SEPECAT Jaguar
was particularly notorious for this, with the Military C of A (formerly known as the
Controller Aircraft Release) being so full of caveats and warnings about rapid
rolling during high-g manoeuvres (or as near as can be achieved in a Jaguar),
maximum numbers of continuous rolls, etc., that few Jaguar pilots were ever
entirely sure what all the operating limitations were in any given configuration.
13.2 Dynamic Lateral and Directional Stability 277
When a second engine is introduced into an aircraft, it is rarely possible to put both
onto the centreline, and if 3 or more engines are fitted, it is impossible for all
engines to be on the aircraft centreline. Because engines can never be 100% reli-
able, the consequence of this is that the handling effects of asymmetric thrust must
be considered as an airworthiness issue.
A further issue is that of asymmetric services. If an aircraft is fitted with twin
rudders, flaps, or other more esoteric devices, then unless there is a direct
mechanical interlink between them, the possibility of an asymmetric failure may
have to be considered. These cases are included here because the issues are natu-
rally related to those of asymmetric thrust. Described will be only the case of an
asymmetric flap failure, which is the most common problem likely to have to be
considered, and the only one specifically listed in any of the civil design codes
(military codes include, for example, detailed evaluation of asymmetric stores).
Consider a conventional twin engined aircraft (the diagrams below show the view
from above). The aircraft suffers a failure of the starboard1 engine (Fig. 14.1).
The natural behaviour of the aircraft is to yaw right, towards the failed engine.
The pilot (or an advanced autopilot) will attempt to counter this by applying left
rudder, reducing the sideslip to zero. This, illustrated in Fig. 14.2 tends to create a
sideforce towards the good engine, naturally tending to create sideslip.
In order to try and prevent this sideslipping towards the good engine, the pilot
must then use the ailerons to roll the aircraft towards that engine. The consequence
is a permanent rudder input and roll input towards good engine and a small resultant
sideforce (which will be visible in the cockpit as a small displacement of the slip
ball) but if the aircraft is flown correctly, no sideslip (Fig. 14.3 viewed from
behind).
The aircraft can then be flown, either for purposes of a diversion and landing, or
until the failed engine can be restarted, with a small continuous bank angle, and
(unless the aircraft is fitted with sufficiently powerful trimmers) some continuous
pilot input, in both roll and yaw.
In flight, the two minimum control speeds, VMCA and VMCG2 (minimum control
speed in the air and minimum control speed on the ground respectively) now come
into play. These are the speeds above which (a) the rudder power is adequate to
overpower the yawing moment due to the failed engine, and (b) the combination of
aileron power and lateral stability is such that the resultant bank angle is small
enough to be deemed acceptable. These values, once determined, will then become
part of the operating advice for the aircraft. The bank angles (and possibly yaw
angles following initial engine failure) are given in the airworthiness standards.
VMC values will be determined using the “Most Critical Engine”, that is, the
engine which when failed, results in the greatest yawing moment to be counter-
acted. This will normally be estimated by calculation, then confirmed by flight
testing. In jet aircraft, the difference is normally minimal, but in propeller driven
aircraft whose propellers all rotate the same way, the difference between engines
can be quite marked, depending upon a combination of propeller positions,
thrustlines (for example, the thrustlines in the SAAB 2000 are about 2½° set to one
side to counteract the torque effects of the engines during normal flying), and the
resultant yawing and rolling moment effects can be complex. An associated systems
assessment may be required alongside this, given that the failure of this engine may
also degrade power to other services on board the aircraft.
1
Starboard = Right (and carries a green navigation light); Port = Left (and carries a red conspicuity
light). This is easily remembered in English because the three longer words, and three shorter
words are grouped.
2
CAUTION: minimum control speed terminology varies a great deal, ensure that the terminology
used on any particular aircraft or project is adequately explained and understood.
14.3 Testing for Control Speeds in the Air 281
VMC tests come under the general heading of “potentially hazardous” and assess-
ment teams should regard their planning and execution with great care—particu-
larly VMCG tests which almost by definition will end up running an aircraft off the
side of the runway at some point. Also, it is likely that simply closing a throttle on
an engine and feathering a propeller will not be acceptable to a certification
authority; shutting off the fuel to an engine creates a much more rapid reduction in
thrust, and thus more pronounced characteristics which are closer to the case of a
genuine sudden engine failure. A further caution is that when carrying out tests in
piston engined aircraft, their engines can be particularly reluctant to restart fol-
lowing a sudden shutdown. Always plan any airborne engine shutdown tests
overhead a useable runway!
When testing a new aircraft type, or a substantially modified existing aircraft
type (most likely fitted with either modified flying controls, or a different power-
plant) tests will start at airspeeds well (50% +) above the predicted minimum
282 14 Aeroplane Asymmetry
control speed. At a safe height, and with all the usual precautions taken (minimum
crew, briefed recovery procedures, good visibility with clear horizon, etc.) the
critical engine will be failed, usually by shutting off the fuel to the engine via a
remotely operated tap as close as possible to the engine (often on turbine engine
aeroplanes the high pressure or HP fuel cock will suffice for this purpose). The
degree of control available will be recorded, possibly over several tests, before
repeating this test at progressively lower airspeed. Testing ceases when either the
threshold of acceptable control is reached, or often when a target speed is reached
(even though control is still comfortably adequate). In general specific flight test
instrumentation is not essential; however, if available: traces of control forces,
aircraft bank angle, engine power settings and bank angle against time are useful in
subsequent analysis. Calibrated airspeed and altitude instrumentation is essential.
If the critical engine cannot be determined by analysis, or the analysis is suffi-
ciently close that there is some uncertainty over the critical engine, then the test may
need to be repeated for more than one engine—although it is a reasonable
assumption that the critical engine will be at an outboard station. The
post-evaluation declared critical engine is then that which results in the highest
minimum control speed.
When carrying out the post manufacture air test on a newly built aircraft of a
known type, normally tests will only be carried out at the required minimum control
speeds, and then the results confirmed to match (a) the minima of the airworthiness
standard, and (b) the flight test results for the original prototype.
Another issue to be considered during the work up to asymmetric testing is the
effect of sideslip on airspeed pressure errors. Whilst a well designed aircraft
shouldn’t suffer varying PECs as the aircraft is sideslipped, many will. This should
be established so that the test team knows the relationship between IAS and CAS
throughout the testing regime.
Following an engine failure, an aircraft will naturally tend to yaw into the dead
engine. The standards will require that a given set of conditions at engine failure,
followed by a pre-determined minimum reaction time, mustn’t cause an unacceptable
degree of undemanded motion of the aircraft—defined in terms of bank and yaw.
Phrases along the lines of “without exceptional piloting skill” tend to be introduced at
this point, which in the author’s opinion is far too vague for a serious engineer or test
pilot, but is the general approach in civil standards (Table 14.1). However, this does
become an area where the Cooper Harper rating scale (Chap. 11) comes into its own:
without exceptional piloting skill” is normally CHR 4.
Following failure of the critical engine(s), the aircraft must be able to carry out
normal manoeuvres of climbing, turning, descent, approach and landing at
Table 14.1 Requirements of the standards immediately following an engine failure in flight
14.4
Notes on test conditions Test CAS Test power Reaction Maximum bank Maximum
time delay heading
change
Part 25 Most critical configuration VMC (which may not MTOP Not 5° 20°
(149) available after take-off (with exceed 1.13 VSR) specified
gear up)
Maximum 150 lbf rudder force
to maintain control
Legacy Part 23 Flaps and gear up Normal climb speed MCP, with 2s 45° Not
(147(b)) Control must also be possible with all engines, also propellers set specified
with partial or total loss of the demonstrate at VMC for normal
primary roll control system climb
New part 23 (2135) At all conditions, including “During all phases of flight” Undefined Undefined
[Legacy material is retained as an configuration changes and “It is possible to maintain control of
“Approved means of compliance” reasonably predictable failures the airplane
but is no longer mandatory]
British civil airworthiness requirements
Section K No loss of height beyond that VMCA MTOP, Not No attitude change 20°
Behaviour Following an Engine Failure, and Control …
whatever speeds will be used. It would be unrealistic to expect that “full” control
over the aircraft will be available, but it does need to be adequate. This is another
area where the minima of the standards to which aircraft will usually be tested are
particularly minimal, and the author would advise any conscientious airworthiness
team to try and produce an aircraft which can provide considerably more control at
the required speeds and conditions than those listed in Table 14.2.
3
Regarding forward/aft CG and MTOW: the configuration of many aeroplanes, particularly
transport types makes achieving any combination of forward CG limit and MTOW impossible—
the aeroplane may be loaded aft of the CG range to achieve MTOW but therefore a rearwards CG,
or lightly loaded to achieve a forward CG. This makes identifying worst case conditions difficult,
and the inevitable result is probably an increased number of test conditions.
14.5
feathered
For 4+ engined aircraft, also
perform yaw test only with
2 critical engines
inoperative
Legacy Power at lesser of PLF for 1.3 Retracted Retracted 15° yaw both ways without using more than 5° bank. 150 lbf
Part 23 1.3 VSR1 and MCP VSR1 and rudder force need not be exceededa
(147(a)) extended
British civil airworthiness requirements
Section K Propeller for final approach VMCL-1 As Extended 20° away from inoperative Not specified
(K2-8 (good engine), feathered required engine, within 5 s
4.2) (failed engine) for CEIO
Control forces “shall not be approach
excessive”
Trim as for CEIO approach
at VAT1
a
The meaning of the rather ambiguous requirements concerning rudder force in JAR & FAR is that if the rudder forces are low enough, yaw the aircraft to 15°
each way. If however they exceed 150 lbf, then it is only necessary to yaw to whatever angle equates to that rudder force
285
286 14 Aeroplane Asymmetry
Whilst VMCL is defined as the airspeed at which adequate control can be maintained
when, with all engines initially operating normally, the critical engine fails; VMCL-1
is defined as the speed at which, subsequently adequate control can be maintained
with the critical engine inoperative (CEIO). The term VMCL-1 is in-fact only defined
by BCARs and not used by parts 23 and 25, which assume that the two quantities
are the same (and it is generally convenient to assume that is the case).
Testing for VMCL-1 is usually done by trimming the aircraft at a safe speed
(around 1.5 times the estimated value for VMCL-1) and failing the critical engine.
Airspeed will then be gradually be reduced, until the appropriate limit is reached of
bank angle or control force (for JAR and FAR tests use the VMCL limits in this case).
This airspeed is noted, the failed engine restored and finally, a test for VMCL, as
described above is carried out.
BCARs give maximum values of the control forces which should be experi-
enced. Although (just) acceptable, an aircraft which just meets these requirements is
going to impose very high workload upon the pilot(s). Therefore, it should always
be a design objective regardless of regulatory regime that either through the use of
powerful trim devices, or powered controls, these forces are no more than transient
and can be mostly eliminated during sustained flight with asymmetric power.
Whilst the concept of VMCL-2, the airspeed at which a double engine failure can be
safely controlled following an existing CEIO case, is universal—it is only specif-
ically defined in part 25 (at paras 149(g) and 149(h)). The concept is only used in
aircraft with 3 or more engines—so the second most critical engine in a 4-engined
(or more engined) aircraft would almost certainly be the next engine inboard from
the already failed outboard engine. For a 3 engined aircraft (such as a B-N
Trislander (Fig. 14.4, Britten-Norman Trislander three engined transport aeroplane)
or Boeing 727 it is likely to be the centreline engine.
VMCG, the minimum control speed on the ground, is defined by the ability to
maintain the runway centreline during take-off, within a given tolerance, in the
event of failure of the most critical engine—which is again usually performed by
shutting off fuel to an engine during a take-off run. Above that speed, an aircraft
may safely continue a take-off (assuming that sufficient thrust is available to do so)
whilst below that speed, in order to remain on the runway, braking must be used
and a take-off aborted. It is taken into account when determining V1– the take-off
decision speed, which can never be less than VMCG: the other bound for V1 is
obviously VR—the rotation speed. Runway length, wind, slope and surface are
additional considerations, as is density altitude.
Testing VMCG involves a series of take-offs, ideally into wind, with the critical
engine deliberately being failed at a series of airspeeds, starting just below VR and
reducing in small increments until the aircraft is unable to remain within the
required distance of the runway centreline. The take-offs should be filmed from
either behind or ahead of the aircraft, and subsequent analysis of the take-off films
should determine at which engine failure speed speed adequate directional control
on the runway could no longer be maintained. Inevitably, the tests will be carried
out at whatever is determined to be the most critical take-off configuration, with the
worst case weight and CG conditions. Take-off trim is universally used.
This again is a high risk exercise, normally carried out after all airborne mini-
mum control speeds have been determined. The use of a very wide runway, with no
significant obstructions to either side is essential, as is a very competent and well
briefed team on every aspect of the trial. It is important in more complex aircraft to
be particularly aware in VMCG tests of the flight control system—spoilers in par-
ticular can often lead to non-linear rolling moment with stick input characteristics,
and may also lead to increased directional stability (weathercocking tendency) just
when it is least wanted.
During a VMCG test, use of the nosewheel steering is usually not permitted.
Almost certainly this is to allow use of the nosewheel to counter crosswinds
290 14 Aeroplane Asymmetry
Table 14.6 Requirements for VMCG in the main civil airworthiness standards
Notes Power Maximum Roll Maximum Nosewheel
drift from control rudder steering
centreline force
EASA certification specifications (and previously JARs)
Part 25 Any propeller MTOP 30 ft Only to be 150 lbf Not permitted
to take its (9.1 m) used to to be used in
natural keep demonstration
position, with wings
no pilot input level
to pitch setting whilst on
the ground
JAR-23 The quantity VMCG is not defined or used in JAR-23. If VMCG is required, the
author recommends use of part 25
Federation aviation requirements
FAR-25 Any propeller MTOP 30 ft Only to be 150 lbf Not permitted
to take its (9.1 m) used to to be used in
natural keep demonstration
position, with wings
no pilot input level
to pitch setting whilst on
the ground
FAR-23 Any propeller MTOP 30 ft Only to be 150 lbf Not permitted
to take its (9.1 m) used to to be used in
natural keep demonstration
position, with wings
no pilot input level
to pitch setting whilst on
the ground
British civil airworthiness standards
Section K The quantity VMCG is not defined or used in BCAR section K. If VMCG is required,
the author recommends use of JAR-25
(which are not taken account of in these tests), but the additional workload of
using the tiller in a large aeroplane may also be deemed unreasonable (see
Table 14.6).
Civil design codes only define one flight control asymmetry, and that is an
asymmetric flap failure, although military standards may also consider asymmetric
weapons carriage cases and design review may require consideration of other
asymmetries which will be discussed later. The requirement (which is at paragraph
14.9 Asymmetric Controls and Services 291
4
A Royal Air Force Tucano T1 was once lost due to a nose-down pitch trimmer runaway. The
control was so powerful that it overwhelmed the pilot’s ability to control the aircraft (admittedly
the situation was not helped by the fact that he had just recovered from a spin and therefore
presumably partly disoriented), leading to a voluntary ejection. Do not underestimate the ability of
trimmers to cause problems!
Chapter 15
Departures from Controlled Flight
Abstract All aircraft are potentially prone to departures from controlled flight:
most are recoverable, but are nonetheless likely to be undesirable. The stall mode is
considered in depth as this must be evaluated on all aeroplanes, and the spin mode
is considered with regard to different depths of evaluation depending upon whether
an aircraft will be certified for deliberate spinning or not, and whether it is to be
certified for as spin resistant. Two further modes, poorly catered for in airworthiness
standards, are also described—the spiral dive, and the tumble.
Potential exists in any aircraft to lose control. Given the nature of flight, any loss
of control potentially will cause damage, and may well cause complete destruction
of an aircraft. This might make it appear that all loss of control should be avoided at
all reasonable costs; in practice this is not necessarily so. Many loss of control
modes have potential to be recoverable and in some circumstances can be desirable
—for example display aerobatics, or basic flying training.
Loss of control modes exist in all classes of aircraft but this chapter will only
concern itself with aeroplanes, and in particular four modes: the stall, the spin, the
spiral dive, and the tumble. The former will be concentrated upon primarily because
it is, of the four modes, the one most concentrated upon in certification pro-
grammes, and also because it defines part of the flight envelope (Chap. 4). The other
three will be dealt with in more brief form as they are of more specialised interest
than the stall.
There are other modes of yet greater specialist interest, not discussed here, for
departures in helicopters, including loss of tail rotor effectiveness (LTRE), vortex
ring state, and dynamic static rollover.1 Those are not explored here, nor is the
poorly understood rollover mode in weightshift controlled microlight aeroplanes, or
some of the more esoteric departure modes in aeroplanes such as Mach tuck or pitch
divergence.2
15.2 Stalling
“Son, be careful; fly low and slow.” Low and slow is where a pilot doesn’t
want to be.
Don Moore
1
A good source here is Shawn Coyle’s book “Cyclic and Collective: the art and science of flying
helicopters” (Iowa State University Press 1996, or Lulu 2009).
2
The reader wishing to explore these more specialist departure modes might start at the Pilots
Handbook for Critical and Exploratory Flight Testing, published by the Society of Experimental
Test Pilots in 2003.
15.2 Stalling 295
A more simple and universally applicable definition is that a stall is the point as
the angle of attack is increased at which the pilot ceases to have full control over
the aeroplane. This is compatible with the definition above, since an uncontrolled
motion or the longitudinal control being on the stop are clear indicators that the
pilot does not have full control over the aircraft in all axes; however, wing rocking
(undemanded rolling oscillations, initially of low amplitude but potentially enough
to roll an aircraft inverted if not controlled), or other low-speed departures from
controlled flight may also be included.
Aerodynamicists would normally either define the stall when considering lift
versus AoA characteristics as the point at which lift ceases to increase with
increasing AoA, by reference to a flow visualisation as the point where a given
degree of flow detachment occurs from the lifting surface, or as the point at which
there is a marked increase in the gradient of @@aCM
: However, this definition can be
misleading in airworthiness work as, for example, the wing may still be at an
aerodynamically unstalled condition, but the limit of elevator authority reached.
During a test programme, the test team must define the stall for a specific
aircraft. Notwithstanding that other definitions may be useful in certain circum-
stances, the three most common definitions are:
1. The longitudinal control being on the nose-up control stop (often termed “mush”
by pilots). This is most common at forward CG where insufficient nose-up
control authority exists to fully aerodynamically stall the wing.
2. A downward pitching motion (often termed a “pitch break”). This is caused by a
loss of lift at the mainplane (or canard) altering the balance of forces and
moments on the aircraft and causing a net nose-down pitching moment. This is
most common at aft CG/hangpoint states, where there is sufficient nose-up
control authority to fully aerodynamically stall the wing.
3. A wing drop, sometimes accompanying a pitch break. This occurs where the
two sides of the wing do not stall simultaneously and may be caused by a small
amount of uncorrected sideslip, a rigging asymmetry in the wings and airframe,
or by an inadvertent control input.
The term stall warning describes those characteristics of the aircraft which indicate to
a pilot that he or she is flying at conditions close to the stall and caution may be needed.
Stall warning characteristics will vary between aircraft and must be identified for a
given aeroplane and documented in the Pilots Operating Handbook. Stall warning
may typically include one or more of the following:
• Airframe buffet, as localised airflow starts to detach.
• Stick buffet, as localised airflow, usually over the wing root in a conventional
3-axis/tailplane aircraft, detaches and strikes the tail control surfaces.
• Artificial stall warning devices, normally either based upon an AoA sensor or a
localised airflow pressure sensor. [Note: not based upon upon airspeed as the
stalling airspeed will vary with configuration, weight, pitch rate and normal
acceleration.]
296 15 Departures from Controlled Flight
• An aircraft pitch attitude which is perceptibly more nose-up than that normally
seen in level flight.
• The aircraft’s primary pitch control being noticeably displaced in the nose-up
sense compared to its position in level flight.
• Lack of control responsiveness.
During the airworthiness evaluation process for any aircraft, the following ques-
tions need to be addressed:
• What are the stalling characteristics at representative deceleration rates? Are
these characteristics acceptable?
• What are the stall warning cues? Are they adequate?
• Is the aircraft fully controllable during deceleration down to the point of stall?
• Can the aircraft, post-stall, be returned to controlled flight without the use of
exceptional piloting skill, or whilst suffering an unacceptable degree of height
loss or uncommanded (or undemanded, the terms can be used interchangeably
here) manoeuvre?
Finally, operating data must be confirmed to accurately and safely address the
stalling characteristics and stall speeds of the aeroplane.
5. Stalling characteristics will vary with CG position and power setting. Although
little is universal where stalling characteristics are concerned, expect to most
often see higher stall speeds at fwd CG, and more dramatic post-stall behaviour
at aft CG. With increased power settings, stalling speeds are likely to be lower
with more nose-up stalling attitudes. Increased power again, is likely to generate
more dramatic post-stall behaviour; some aircraft, particularly those with high
powered piston or turboprop powerplants, which may be extremely benign
when stalled at idle power will tend to enter an uncommanded severe rolling
manoeuvre or even enter immediate incipient spin (sometimes with no stall
warning) when stalled at high power.
6. The stall is not necessarily marked by the classical pitch-break beloved of flying
instructors and pilot textbooks. It may be marked by a loss of nose-up authority
(many training aeroplanes, for example the PA28), an AoA triggered Klaxon
(the Sepecat Jaguar), an air horn supplied by air pressure from the wing
under-surface (Cessna 150), severe wing drop (the Britten-Norman Islander)
unacceptable vibration (Lockheed L1011), operation of a stick pusher (BAe
Jetstream), or other cues. The bottom line is that the stall occurs when the pilot
ceases to have absolute control over the aircraft—but teams must learn and
understand the range of definitions accepted by the standard in use. Test pilots,
flight test engineers, and airworthiness engineers must all be in complete
agreement as to the stall definition in use for a test programme and subsequent
documentation.
7. Some degree of wing-drop at the stall is common regardless of aeroplane class;
it is the magnitude of wing drop that is critical to airworthiness, along with any
subsequent tendency to further loss of control. The wing drop limits of all
standards assume that correct recovery action is being taken.
In Table 15.1 are listed the main points of the various standards with regard to
the 1 knot/s unaccelerated stall at flight idle. It is normal to carry out the tests for
compliance with the below standards in all normally used combinations of flaps/
slats/undercarriage/airbrake. If dealing with a simple aircraft with fixed undercar-
riage and no or perhaps 3 position (cruise, take-off, landing) flaps, this is very
straightforward. If dealing with a more complex aircraft with a wide range of
configurations possible: such as any transport aeroplane, then consult with both the
standard and experienced aircrew, before proposing a test grid to the approving
authority.
Before attempting to manage stalling tests or approvals, the flight-test or air-
worthiness engineer is well advised to try and gain some personal exposure to
stalling in a similar class or type of aeroplane. It is far easier for an engineer to draw
sensible conclusions and write a meaningful report if they have experienced the
phenomena themselves.
Table 15.1 General and wings level stalling requirements from the main civil airworthiness requirements
298
Maximum Maximum Artificial Maximum Maximum Maximum stall speed Stall warning
pitch change height loss Stall undemanded wing drop interval
warning? yaw
EASA certification specifications/joint airworthiness requirements
Part 25 ‘controllable’ No limit Permitted, No limit 20° Formula Greater of 5 knots/
(103, 201, but not visual 5%
207) alone
Legacy ‘controllable’ No limit Permitted, 15° 15° 61 knots Vso (single engine 5 knots
Part 23 but not visual and sub-2730 kg twins unable
(49, 201, alone to climb at 1.5% SEI)
207)
Part 22 No limit No limit Permitted, No limit 30° 80 km/h (Vso—no ballast) 5–10%, less or none
(49, 201, but not visual 90 km/h (Vso—full ballast) if a/c totally
207) alone controllable down
to stall
Part VLA No limit No limit Permitted, 15° 15° 45 knots (Vso) 5–10 knots
(2, 201, but not visual
207) alone
15
Maximum Maximum Artificial Maximum Maximum Maximum stall speed Stall warning
pitch change height loss Stall undemanded wing drop interval
warning? yaw
British Civil airworthiness requirements
Section K 30° 100 ft, Permitted No limit 15° 60 knots Vso (single engine 5 knots, less at
(2–11) unless and miti-engine twins unable assessors discretion
documented to climb at 1% gradient SEI)
Section S No limit No limit, Discouragedb No limit but 20° 35 knots (Vso)c At assessors
(2, 201, but must be mustn’t tend discretion
207) documented to spin
a
The FAR-23 requirement that single engine aircraft, or multi-engine aircraft must have a Vso and VS1 not exceeding 61 knots need not be complied with if
either a single-engine-inoperative climb can be maintained at a 1.5% gradient (FAR 23.67), or extra safety requirements for emergency landings in FAR
23.562 are met
b
Section S is unusual in that a range of acceptable handling characteristics are listed if no stall warning is provided. This is appropriate to the class of aircraft,
where stall warning is often very poor, but low speed handling excellent. An artificial stall warning is absolutely a last resort but may be used
The Unaccelerated and Turning Flight Stalls
c
Section S permits a higher Vso if the wing loading is no greater than 25 kg/m2. In reality, that would be an extremely inefficient wing design and is a largely
irrelevant requirement
299
300 15 Departures from Controlled Flight
For civil certification, the turning flight stall is usually carried out at 30° of bank in a
co-ordinated turn, although requirements for power and flap settings may vary. For
military aeroplanes the test conditions will vary far more and be necessarily dictated
by role. Since this is an unaccelerated stall, again the entry rate should be no greater
than 1 knot/s.
Most aeroplanes will suffer from some undemanded rolling at the point of a
turning flight stall, and this is usually more marked following a turning flight stall.
The terminology used for this undemanded rolling motion is “into the turn” and
“out of the turn”. If an aircraft is banked to the right, and at the point of stall it banks
more to the right, then it is said to have rolled “into the turn” (a characteristic most
commonly associated with low wing aircraft), whereas if it tends to roll towards
wings level (a characteristic most often found in high wing aircraft) then it is said to
have rolled “out of the turn”.
It is always important that stalling from a co-ordinated turn cannot cause a spin.
The author has seen several aircraft which did display this tendency (i.e. never
assume that it won’t be there!), the best known of which is the North American
Harvard (Fig. 15.1) which routinely will enter an incipient spin from a stall with
power above idle—a characteristic which is believed to have killed many student
pilots during WW2. Similarly any marked pitch-up or longitudinal control force
lightening at or near to the stall should be regarded as unacceptable (Table 15.2).
Trim speed(s) Maximum Required power Gear Flaps Max into Max out
height loss settings turn of turn
rolling rolling
EASA certification specifications/joint airworthiness requirements
Part 25 1.5 VS1 No limit Idle “Any likely combination… approved for 30° 60°
(103, PLF for 1.5 VS1 operation”
201, 203)
Legacy 1.5 VS1 No limit Idle with fine pitch, Retracted Retracted and fully extended + each 60° 90°
The Turning Flight Stall
Trim speed(s) Maximum Required power Gear Flaps Max into Max out
height loss settings turn of turn
rolling rolling
British civil airworthiness requirements
Section K 1.5 VS1 No limit Idle, and Retracted Retracted and fully 60° 30°
(2–11) PLFd and extended + other settings “likely to
extended be used”
Section S Not specified No limit, but Idle Retracted In all available settings 30° 60°
(203) must be MCP and
documented extended
a
The “PLF” setting dictated by JAR-23 is normally 75% MCP unless this results in unacceptable nose-up pitch attitudes, in which case the lesser of PLF at
Maximum Landing Weight and 1.4 VS1, or 50% MCP should be used. Cowl flaps should be as required for the configuration
b
Part 22 is unique in requiring turning flight stalls to be carried out at 45°, not 30°. JAR-22 also requires the test grid to include airbrakes in any position,
asymmetric water ballast (unless this is shown unlikely to occur). Obviously if the standard is being applied to an unpowered sailplane, then all the powerplant
items should be disregarded
c
The “PLF” setting dictated by FAR-23 is normally 75% MCP unless this results in unacceptable nose-up pitch attitudes, in which case the lesser of PLF at
Maximum Landing Weight and 1.4 VS1, or 50% MCP should be used. Cowl flaps should be as required for the configuration
d
The “PLF” setting dictated by Section K is that for aeroplanes not above 2730 kg MTWA, 75% MCP shall be used; for heavier aircraft the power shall be the
15
greater of 50% MCP and the power required for level flight at MTWA and 1.4 Vso (but need not exceed 75% MCP). Cowl flaps (referred to in Section K as
“Cooling Gills” should be as required for the configuration)
Departures from Controlled Flight
15.5 Defining Test Conditions for Large … 303
Part 25 does not specify the test conditions for test stalls in turning flight. So, the
airworthiness team must examine the proposed and likely range of operating envi-
ronments for their aircraft and use this to define the test grid. Defining test conditions
in this way is part of the training of a test pilot, and engineers should usually defer to
them in this part of test planning. Before then embarking upon testing however, the
company must agree the test-grid which will comprise the means of compliance with
the relevant airworthiness authority (and get their agreement in writing!).3
Recently Boeing and Airbus have worked together to publish joint recommen-
dations on best practice in large aeroplane stall testing.4 This is developing science—
in particular because the FAA especially is increasingly favouring stall avoidance as
the paramount aspiration in both part 23 and part 25 aeroplanes, however at present
this join Boeing/Airbus work is the best advice available. Broadly it indicates that,
apart from the need for greater levels of safety precaution and instrumentation,
stalling and stall testing of large aeroplanes is broadly similar to the same activity for
smaller aeroplanes. Altitude loss of large aeroplanes massing 100 tonnes plus is
likely to be in the order of 5–7,000 ft in the stall and recovery, compared to perhaps 1/
20th of that for a light single. Large aeroplanes, particularly when operating at high
altitude are however particularly prone to the risk of stalling at high Mach numbers—
a condition hard to both predict, and achieve at well defined conditions. It also raises
concerns about the difficulties presented by the increasingly highly automated nature
of large aeroplane flight controls, and their response, in combination with that of
aircrew, to stalls or stall-related events.5
Most aircraft will (and should) pitch nose-down, at the point of stall. This naturally
puts the aircraft into a dive from which the existing trim setting should allow it to
recover without pilot input—although for an efficient (and comfortable) recovery
correct handling by the pilot—initially to allow the aircraft to pitch nose-down to
3
The alternative, of having to explain to the financial director or chief engineer very late in the day
why several hours of additional expensive flight testing will be required is not worth contemplating
4
The present most mature publication on this work is: “Stalling Transport Aircraft” by Paul
Boles-Moorehead et al., Aeronautical Journal, Volume 117, Issue 1198 , pp. 1183–1206
(December 2013). This is essential reading for anybody with any interest in stall qualification of
large aeroplanes.
5
The reports and commentary into the loss of Air France flight 447, an Airbus A330-203 off Brazil
on June 1st 2009 provide particularly strong evidence as to why this is a matter of concern.
Extensive material is available in the public domain.
304 15 Departures from Controlled Flight
reduce AoA, use of thrust to reduce initial altitude loss, then to pull out of the dive
and re-establish straight and level (or climbing) flight is normal.
It is important to ensure that the acceleration (a secondary effect of the nose-down
pitching motion) immediately following the stall cannot lead to an inadvertent
exceedence of VNE/VMO/MMO. Also, the pullout (especially that pull-out naturally
caused by the pitch trimmer setting) must not cause NZ or a limits to be exceeded.
Most research shows that the optimal stall recovery in most aeroplanes is a
simultaneous application of a nose-down pitch control input, and increase in power.6
In practice this means that the power should invariably slightly lag pitch, as pitch input
response in most aeroplanes will be more immediate than thrust response.
However, for airworthiness purposes, what is paramount is that the aeroplane
has, through extensive testing, been shown to absolutely and consistently recover
using the set of stall recovery actions that have been determined and laid out in the
aircraft’s operating documentation. In some airworthiness regimes it is also a
secondary requirement to identify the amount of height loss—which is straight-
forward, although of marginal genuine usefulness since this value is extremely
dependent upon handling technique and circumstance. Whilst in most aircraft actual
stalls and recoveries are regularly flown as part of either pilot training, and/or
periodic airworthiness flight tests—these will always be flown at a safe height
above the ground and with adequate preparation and briefing; the reality is that
inadvertent stalls are most likely to happen close to the ground, and completely
unexpectedly. It is unreasonable and unsafe to be completely representative in
testing such scenarios. It is preferable that, so far as is safely achievable, the
recommended stall recovery actions are identical to that used on other similar
aeroplane types, and/or the types that pilots will have trained on—so that pilots
“instinctive” actions in the event of an inadvertent stall are most likely to be correct.
6
Self citing unfortunately—but: Gratton et al., Finding an optimal set of stall recovery actions for
light aeroplanes, RAeS Aeronautical Journal. Vol. 118, No. 1202, pp. 461–484 (May 2014).
15.7 Other Stall Cases—The Accelerated and Dynamic Stalls 305
Table 15.3 Dynamic and accelerated stall requirements of the main civil airworthiness standards
Maximum Maximum Maximum Maximum Turning flight Wings level
height loss pitch-down into-turn out-turn case(s) case(s)
wing drop wing drop
Joint airworthiness requirements/EASA certification specifications
Part 25 Not Not 60° 90° Up to 3 knots/s No requirement
(201, 203) specified specified at 30° bank
Part 23 Not Not 60° 90° 3–5 knots/s at No requirement
(201, 203) specified specified 30° bank
Part 22 (Wings level special case only)
(201) With the sailplane in straight flight at 12 VS1 in the configuration appropriate to
winch-launching by pulling rapidly on the control stick, a pitch attitude approximately 30°
above the horizon must be achieved and the resulting stall must not be severe and such as to
make prompt recovery difficult
Part VLA “Not No limit 60° 60° 3–5 knots/s at No requirement
(203) excessive” 30° bank
Federation aviation requirements
FAR-25 Not Not 60° 90° Up to 3 knots/s No requirement
(201, 203) specified specified at 30° bank
FAR-23 “Not Not 60° 90° 3–5 knots/s at No requirement
excessive” specified 30° bank
British civil airworthiness requirements
Section K No limit No limit Total not Total not 30° plus, No requirement
(2–11 2.4) exceeding exceeding 2 g (utility)a
90° 60° 4 g (aerobatic)
Section S No limit No limit No requirement to consider turning dynamic Stall following
(143) stallsb engine failure
in steepest
permitted
climb
a
BCAR Section K’s 2 and 4 g stall cases for the utility and aerobatic cases respectively, are to be with flaps and
gear UP, trim and power as required for the manoeuvre (although engine power can’t be increased during the
manoeuvre or recovery), and CG most adverse for recovery (likely to be a heavy-forward case)
b
Notwithstanding the requirements of Section S, the BMAA’s own test schedules BMAA/AW/010 (as well as
in-house schedules used by P&M Aviation, the UK’s largest manufacturer) do include dynamic turning stalls
306 15 Departures from Controlled Flight
which reflect the way in which the aircraft will actually be flown. Most organisa-
tions dealing routinely with flight testing (particularly those dealing with the lighter
end of aircraft certification) will have type or class specific schedules and guidance
which can be used—and may well be much more useful than the main certification
standard. Unless stated otherwise, test conditions are the same as for 30° turning
flight stalls discussed above.
Almost universally the 1 knot/s stalling speed can be considered to scale as:
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 Actual Weightp ffiffiffiffi
VSACTUAL ¼ VSMTOW 2
N; ð15:1Þ
MTOW
where N is the normal acceleration factor. However, due to the delayed onset of the
stall during a rapid increase in angle of attack, the stall speed in the dynamic stall is
likely to be less than that predicted (or similarly, the stalling AoA likely to be
higher, but AoA is seldom instrumented). It would take considerable aerodynamic
study in any individual case to accurately predict this—a study which in most cases
is of little value (the exception is likely to be modern combat aircraft, where this
information becomes fundamental to the control laws of the fly-by-wire system);
however, typically expect the dynamic value to be 5–10% below that predicted by
the equation above.
Formal dynamic stall warning requirements are generally the same as for the
unaccelerated stall. Whilst at first glance this is entirely sensible, in practice it is not
and the sensible airworthiness team should aim to go beyond the minima of the
standards.
Consider for an example, CS.23, which has a requirement for a minimum 5 knot
stall warning margin, and a test deceleration speed for the dynamic stall of 3–
5 knots/s. This potentially would give a pilot as little as 1 s warning of the
impending stall—only enough for the most alert pilot (and then in an aircraft with
high elevator power and a very short SPO period) to take appropriate action. For
any aircraft that is likely to be routinely flown in violent turning manoeuvres (e.g.
aerobatic aircraft, utility aircraft used for short field or agricultural operations, any
military aircraft or most training aircraft) it is important that the pilot is given clear
and unambiguous warning of the impending dynamic stall. Apart from the obvious
artificial stall warning devices or natural buffet, this might be achieved through high
stick forces, very high nose-up pitch attitudes, or alarms triggered by AoA sensors.
15.7 Other Stall Cases—The Accelerated and Dynamic Stalls 307
Going beyond the minima of the airworthiness standard in this way can require
some political and financial courage, particularly if potentially expensive modifi-
cations to the aircraft may be required. However, any recent set of fatal accident
reports should furnish one or two cases of aircraft loss due to a pilots failure to
recognise the dynamic stall—this and the risk of litigation is generally sufficient to
convince the most obstructive company accountant.
The spin is a post stall mode that can occur if an aeroplane is stalled with significant
sideslip present. It then comprises a combined rolling/yawing/pitching motion and
high rate of descent that is often self sustaining (termed “autorotation”) unless
deliberate action is taken to recover the aeroplane.7 In theory all aeroplanes can spin
—in practice there are some exceptions: stall protection in most large aeroplanes is
sufficient to effectively prevent a spin ever developing, and whilst rarely applied,
“spin resistant” criteria have been developed and scope exists to declare an aero-
plane as spin resistant.
Predicting spinning characteristics is highly problematic—trustworthy analytical
tools do not exist for either spin characteristics, or the reliability of recovery from
the spin.8 The use of models either radio-controlled and free-flying, or in a vertical
wind tunnel may be more trustworthy, but should also be treated with considerable
caution as the scale laws between a model and full scale aeroplane, where they
affect spinning characteristics, are also poorly understood.9
So, any airworthiness programme for an aeroplane is likely to some extent, to
have to explore the spin. In virtually all cases this will be regarded as high risk
testing, and thus will require substantial safety oversight.10 Nonetheless, it is
essential, because if an aeroplane is capable of spinning, pilots must know how to
identify that spin, and how to recover—from an airworthiness viewpoint also the
potential for a spin entry must be commensurate with the role of an aeroplane. In a
passenger aircraft that must be microscopically small, whilst in an aeroplane
designed for display aerobatics the spin entry and recovery should simply be
7
A good introduction to the spin may be found in Chap. 4 “Maintaining Aircraft Control: Upset
Prevention and Recovery Training” of the FAA’s “Airplane Flying Handbook” (reference
FAA-H-8083-3B, available online).
8
There is a concept of the “Tail Damping Power Factor” or TDPF as a predictor of good spin
recovery characteristics. This concept was originally developed at the Royal Aircraft
Establishment in the 1930s then further by NASA in the 1940s–50s. However, later 1970s NASA
work concluded that the method was untrustworthy and should not be used.
9
See Spin-Tunnel Investigation of the Spinning Characteristics of Typical Single-Engine General
Aviation Airplane Designs, NASA Technical Paper 1076 dated November 1977.
10
See “High Risk Flight Testing” by Wilson “Rusty” Lowry, a chapter in Encyclopedia of
Aerospace Engineering, https://doi.org/10.1002/9780470686652.eae602.
308 15 Departures from Controlled Flight
consistent and well understood. In training aeroplanes the potential to spin may be
more controversial—the flying training world is divided over whether spinning
should be taught at all; one argument is that as aeroplanes can spin pilots should be
familiar with it and able to recognize and respond appropriately, the opposing
argument is that historically more fatalities were caused by the conduct of spin
training, than were prevented by that training having taken place. At present, except
in gliding and military training where spin training is still the norm, most training
regimes do not include spinning and most modern training aeroplanes are not
qualified for deliberate spinning. However, whilst it is essential of course to conduct
extensive spinning evaluation—including mishandled and multi-turn spins, during
the flight testing of an aerobatic aeroplane, there will almost always still be a need
to conduct spinning (or spin resistance) evaluation of almost all aeroplanes.
Typically a simple aeroplane not intended for a deliberate spinning clearance is
likely to need 50–100 spins to demonstrate compliance with requirements
(Table 15.4), an aeroplane being certified to be spin-resistant around 300–400 test
points (Table 15.5), and an aeroplane being certified for deliberate spinning in the
order of 500–1000 flight test spins, depending upon the nature of the clearance
being sought.
The concept of a “Standard Spin Recovery” exists, which is typically summa-
rized as:
(i) Close throttle
(ii) Full back stick
(iii) Full opposite rudder
(iv) Stick progressively forwards until the spin has stopped
(v) Ease out of the ensuing dive.
There is a common assumption that this is, or at-least should be, the sole way to
recover any aeroplane from a spin. This is untrue for several reasons that it is
helpful to understand. These include that:
– Some aeroplanes do not respond best to these actions; for example most 3-axis
microlights have potential to be kicked into a spin in the opposite direction by
the “full opposite rudder” actions and are better recovered from a spin using
centralised controls, due to their extremely powerful fin and rudder at low
speeds.
– Some aeroplanes require modification to these actions. For example the Auster
(Fig. 15.2) requires a short delay between rudder and pitch inputs, otherwise the
aeroplane enters a period of significant and problematic negative g.
– Some aeroplanes, whilst they may recover adequately with these actions, will
respond better to variations. In particular aerobatic pilots may often increase
power again to create accurately timed spin recoveries during recovery. In such
aircraft, it is appropriate to test and document both recoveries, so that good
operating data is available to support both non-aerobatic and aerobatic
operations.
Table 15.4 Civil design code minima for spin testing
15.8
Part 25 There are no published standards. This is likely to be subject to agreements between companies and their authorities, and developed
company manuals
Legacy part 23 [no With and without flaps (flaps may be retracted Longer of 1 turn or 3 s Max 1 additional turn
spinning clearance during recovery)
required] (221a)
Legacy part 23 With and without flaps (flaps may only be retracted At least 6 turns (more as required for Max 1½ additional turns
[spinning clearance during recovery if a deliberate clearance to spin with clearances), or to commencement of There must also be means to
required] (221b-c, flaps is not sought) natural spiral dive if occurring after 3 abandon the aircraft
807d) turns
CS.VLA (221) With and without flaps (flaps may be retracted Longer of 1 turn or 3 s Max 1 additional turn
(No desire for a during recovery)
deliberate spinning
clearance is
assumed)
CS.22 (221) Sailplanes, and powered sailplanes with engine at At least 5 turns or commencement of a Seeking deliberate spinning
idle natural spiral dive, whichever is first clearance: at any point in the
Including: water ballast at critical asymmetry, spin in max 1 additional turn
neutral and in-spin ailerons, opposite rudder and Not seeking spinning
aileron, flaps and airbrakes both neutral and most clearance: max 1½ additional
critical condition turns
BCAR Section S Flaps, airbrake, gear in any permitted position. Flaps Longer of 1 turn or 3 s Maximum 1 additional turn
(221) and airbrakes may be retracted during recovery Para 143 control force limits
(No desire for a not to be exceeded
deliberate spinning
clearance is
assumed)
All standards and best practices will also insist on testing over a full range of weight, power and CG conditions
309
310 15 Departures from Controlled Flight
Less often considered in airworthiness is the spiral dive, which is to some extent
interlinked with the spiral mode (Chap. 13). Some aeroplanes, particularly those
where directional stability is significantly stronger than laterial stability, have
potential to enter a mode where the aircraft has a significant bank angle, nose-down
pitch attitude, and is increasing airspeed whilst reducing altitude. This can be
induced deliberately by pitching and rolling the aeroplane together, and pilots are
familiar with this as part of a phase of pilot training known as “unusual attitudes”.
Less deliberately a spiral dive may occur naturally as a result one of a number of
circumstances: an aeroplane self recovering from a spin, a pilot failing to maintain
15.9 The Spiral Dive 311
the aeroplane in balance during a steeply banked turn, or failure to monitor aircraft
attitude for prolonged periods of time—particularly in turbulence (this combination
is easily achieved in cloud with an aeroplane being hand flown and/or with an
autopilot failure).
Recovering from the spiral dive is, theoretically straightforward—the aeroplane
is at high speed and thus should have good control authority, therefore both pitch
and roll control should be readily available. The problem here lies however in the
stresses an unwise manoeuvre will place on the airframe—the combination of
aileron and elevator input, with pitch and roll rates, and high speed does not exist
within the structural design envelope for most aeroplanes, and can cause catas-
trophic airframe damage. Therefore it is essential that a sequence of actions are
designed and tested that are assured to provide a consistent recovery—ideally with
reasonably minimal height loss. The actions are most likely to involve reducing
power to reduce the accelerating effect of thrust, rolling the aeroplane wings level
with co-ordinated use of the roll and yaw controls whilst keeping the nose low (and
thus avoiding increasing normal acceleration); once the aeroplane is wings level,
then pitching the aeroplane up to the level flight attitude, re-applying power as that
is reached to restore altitude. These actions are common to most aircraft and cir-
cumstances—but the cues used by the pilots may not be. In a light aeroplane in
visual flight conditions it’s likely that this will be flown by reference to wind noise
and external cues. In the same aeroplane in convective cloud this is not feasible,
therefore recovery must be by reference to instruments alone—the use of particu-
larly the artificial horizon then airspeed indicator. An airliner is likely to use sub-
stantially the same actions. In either case however, there must also be consideration
of partial instrumentation failures: for example of the primary attitude indication
source, requiring airspeed and heading alone to be used as guidance. All of these
issues are normal and routinely solved, but must be understood when certifying an
aeroplane, but especially when designing operating documentation.
A further consideration that must be allowed for is that whilst aerodynamically
they may be very different, visually the spiral dive and the spin look extremely
similar to a pilot—the major differences being visible from airspeed (wind noise, or
the airspeed indicator will both provide this information, but both must be
specifically observed). The differences and their identification must also be included
in operating documentation for any aeroplane. It is observable, and supported to
some extent by flight mechanics theory, that aeroplanes prone to spin are highly
resistant to the spiral dive, and vice versa; this however should not be considered a
universal rule.
Most airworthiness standards provide little or no guidance on the spiral dive,
although rather more advice can be found in texts on advanced flying training,
particularly under the heading of “upset recovery training”. In terms of airworthi-
ness assessment however, this is a necessary collaboration between structural and
handling qualities specialists—albeit hopefully one unlikely to require substantial
effort.
312 15 Departures from Controlled Flight
Fig. 15.4 Illustration from wind tunnel testing of upper surface vortex in tumbling flying wing.
Wing illustrated is transitioning downwards on page and rotating anticlockwise as seen about a
centre of rotation below the wing
15.10 The Tumble 313
The earliest well documented instance of the tumble was in the YB49 flying
wing experimental bomber (Fig. 15.5) in 1948 by test pilot Robert Cardenas, and
then subsequently the aircraft was destroyed in an accident believed to be related to
the same mode. Earlier but less well document aircraft losses of other experimental
aircraft are (unprovably) believed by the author to also be related to tumble
departures—these include the Northrop XP79P in 1945 de Havilland dh108
Swallow in 1946.
Later this has become a known mode of failure in Rogallo-winged hang-gliders
and microlight aeroplanes—one type which suffered a number of tumbling
departures was the Mainair Gemini flash 2 alpha (similar to the Flash 2 in
Fig. 15.6), although this was resolved with a mixture of mandatory rigging mod-
ifications and pilot education. The author has discussed this also with colleagues
who worked on flight testing of the later Northrop Grumman B2 Spirit “Stealth
Bomber” (Fig. 15.7)—in that case the team were aware of the potential for such a
mode, but addressed it by flight protections within the aircraft’s control laws that
prevented extremes of either pitch attitude or pitch rate and so it has never been
encountered in that aeroplane in flight—any aerodynamic design analysis of the
potential remains classified and inaccessible.
The tumble mode has seldom found to be recoverable, and in most cases the
inertial forces have been so great as to cause aircraft to break up before ground
collision, and often pilot incapacitation. Therefore for highly swept tailless aero-
planes, the emphasis must always be on ensuring that a mixture of handling advice
and design prevent manoeuvres being flown that allow a sufficiently high pitch rate,
that the pitch-autorotative tumble can initiate.
There are no airworthiness standards for any class of aeroplane but hang-gliders
defining the tumble mode, and the airworthiness standards for hang-gliders are only
advisory. Therefore any team investigating the airworthiness of such aircraft, whilst
they should make use of all available experience and knowledge, are essentially on
their own concerning what best practices should be applied.
Chapter 16
Systems Assessment
The future airliner will need in the cockpit a pilot and a dog.
The pilot is there to talk to the passengers, whilst the dog is
there to bite him if he tries to touch anything.
Old joke, still not true
All aircraft use systems—this might encompass navigation systems, cockpit dis-
plays, control linkages, hydraulic power storage and/or transfer, electrical power
storage and/or transfer, fuel storage and supply—the potential list is virtually
endless. One can of-course simply regard the entire aircraft as a single complex
system, but in practice this is unhelpful as the sheer complexity—even of a light
aeroplane, let alone of a modern transport or combat aircraft, prevents any mean-
ingful analysis of the whole aircraft at once.
So, the airworthiness team must make decisions about the breakdown of systems
for assessment purposes. This is a difficult, and often subjective, task—for example
should the whole hydraulic system be treated as a single system, or should it be
broken down into the undercarriage, primary flight controls, distribution system,
Fig. 16.1 Example partial systems hierarchy tree for a fictional light combat aircraft
and so-on? Similarly, how should a complex integrated avionics system be asses-
sed?—as a single unit, or split to the multiple GPS, autopilot, VOR/DME, EGPWS
components. There is no straightforward answer to this, and decisions must be
made which are informed by the role and mode of usage of the aircraft—so getting
operator (pilot!) input into the early stages of the assessment is particularly helpful
as engineers who are not pilots will often make incorrect assumptions about how
systems are used, and the consequences of their failure.1
A useful way to visualise the systems arrangement is to create a systems hier-
archy tree such as that shown Fig. 16.1, albeit that this will inevitably become
ambiguous and suffer overlap before it is finished. This does show that for example
the pitch trimmer can be evaluated in isolation, as a single mechanical system, but
that some consideration is needed about the overlap between how the emergency
blowdown system should be considered between the mechanical secondary systems
and the undercarriage. For most real aircraft of course the diagram will be far more
complex than shown here—for the student trying to understand the subject matter,
the author would recommend starting much simpler, for example representing a
light aeroplane or a guided missile. Light aeroplanes have the advantage that many
operators or maintenance manuals are readily available to provide study informa-
tion and are the author’s preferred teaching method, even for students destined to
work on much larger and more complex aeroplanes.
1
There are many useful textbooks on aircraft systems, aimed at professional pilots and can be
found in the catalogues of most “pilot shops”. A good source for engineers is Aircraft Safety
Systems Safety by Duane Kritzinger: ISBN 978-0-08-10088908 (Elsevier 2017).
16.2 System Failure Numeric Analysis 317
Once the discrete systems that must be analysed have been identified, a second and
also slightly subjective process must be followed. For each of these systems,
repetitively three questions need to be asked, which are:
1. What can fail?
2. What are the consequences of that failure?
3. What is the probability of that failure?
All of these questions require significant experience and expert judgment to
answer—and the third in particular can difficult given that systems are inevitably
designed to be as reliable as reasonably achievable, and thus an accurate prediction—
or experimental demonstration—of the failure rate can be very difficult to achieve.
The answers to these questions need to be expressed both as a broad concept,
and numerically—a classical example of the numerical approach would rotate
around Table 16.1, which is unsurprisingly very similar to the approach used in
many other forms of risk assessment. Expert analysis is used for each discrete
system to predict the ways in which it might fail, and the worst potential conse-
quence of that failure; these consequences are allocated scores corresponding to the
worst outcome. For example failure of a VOR/ILS receiver or CDI, which in almost
all cockpits will be duplicated by a similar or alternate system would be classed risk
level 2 (maintenance rectification required), whilst failure of ground brakes, which
might cause a runway overrun may be classified level 5 (potential loss of life). More
difficult, the probability of the failure must also be estimated—this may be esti-
mated based upon experience from similar systems on earlier aircraft, extensive
ground testing, or analysis—although generating a theoretical prediction of system
reliability is difficult and again an element of expert judgement will usually be
needed as well: this probability will also be typically given a score from 1 to 5,
where 5 is most likely and 1 least.
The two values will normally be multiplied by each other to give a value
between 1 and 25, as indicated in the last part of Table 16.1, allowing classification
between “low” and very high of the total system risk. Generally for civil aeroplanes,
all results by the time an aircraft is fit for flight, and certainly for final certification
will be in the low bracket—although some “medium” assessments may be accepted
subject to contingency procedures and authority acceptance. Any “high” or “very
high” result is likely to be unacceptable and require rectification—most likely
through system redesign or duplication. Clearly if the probability of failure of two
or more duplicated systems can be made essentially independent, then results can
be improved radically: for example if the probability of failure of a primarily flight
controls hydraulic system is assessed at 1 10−3 per flying hour with a potentially
catastrophic outcome, then this gives a score of 25, which is unacceptable.
Duplicating this with a second system would bring the probability to (1 10–3)2 =
1 10−6, and thus the total outcome to 15—a large improvement, but still unac-
ceptable. Introduction of a third (triplicated) hydraulic system brings the probability
318 16 Systems Assessment
2 The term “time unit” used here by the author is deliberately ambiguous.
Depending upon the nature of the aircraft, system, and assessment, this might
potentially be: flying hours, undercarriage cycles, flights, engine cycles, weapon
firing cycles, or some other unit appropriate to the task.
16.2 System Failure Numeric Analysis 319
of complete failure to 1 10−9, and the full outcome score to 5 or 10: a medium
but acceptable outcome, subject to agreement of the airworthiness team and
authorities.
For part 23 aeroplanes (or part 27 helicopters) and second line military aircraft,
it’s normal to usually only consider single failure cases (and for smaller than part 23
aircraft only a very small number of basic cases such as engine failure and control
cable jamming). For part 25, 29, or first line military aircraft double failure cases are
likely to have to be considered. This again requires significant expert judgement,
primarily to determine the combinations of failures, which might be problematic. So
for example, it can reasonably be stated that the combination of radio communi-
cations failure and nosewheel steering can be disregarded as they do not increase
the severity of the consequences of each other; alternately a combined ground
braking and nosewheel steering failure could produce a completely uncontrollable
aircraft on the ground, and that this should be considered as a failure case.
All such failure modes will require analysis, and one consequence of that
analysis will be aircrew procedures: promulgated as checklists (with usually
expanded checklists within manuals that are used for training purposes); this is not
it should be appreciated a new concept as illustrated by Fig. 16.2, which is an
extract from the manual of a WW2 era fighter aircraft. The changes since then has
been that military and transport aeroplanes in particular have become far more
complex, but also that the manuals themselves have become far more complex,
assuming much less personal decision making on the part of aircrew. Whether this
is appropriate is a common topic for debate amongst experienced aircrew, but it is
certainly an established fact. The reader will learn much about this from reviewing
some aircraft manuals, which are readily obtainable.
An instance available readily for study of this being got wrong, was Concorde,
nominally a part 25 aircraft (in practice certified to a one-off and dedicated air-
worthiness standard written for the project in the 1960s); specifically, the aircraft
lost on 25 July 2000 at Paris [Air France Flight 4590]. This was lost when a piece of
debris on the runway ruptured a tyre, which in turn threw up a large piece of
material puncturing a fuel tank and causing the escaping fuel to ignite. The result
killed 113 people and led to the grounding of the type for over a year; the reason for
the grounding was that it was determined that a single failure (runway FOD2 being
thrown into the tank) led to a catastrophic outcome, which should be an unac-
ceptable event in any part 25 aircraft, but also that there were no procedures in place
for dealing with the specific combination of failures that occurred. The fleet were
grounded for over a year and substantial system redesign and recertification was
required.
2
FOD is a dual purpose acronym, meaning both “Foreign Object Debris”, and “Foreign Object
Damage”. Most aircraft hangars will have readily available rubbish or “gash” bins marked with
“FOD”, and most airports will have regular “FOD sweeps” indicating the importance of always
ensuring such debris is retrieved and disposed of before it becomes a safety hazard.
320 16 Systems Assessment
Fig. 16.2 Extract from the emergencies section of the Spitfire Mk.IX Pilots Notes showing
aspects of system failure management
Clearly, the lists of events to be considered will be extremely long and complex—
particularly for modern systems-reliant aircraft such as any modern airliner or combat
aircraft, and the scale of this task shouldn’t be underestimated. It is also important to
understand that failure modes are not necessarily constrained to those systems sud-
denly stopping working—often more challenging both for the airworthiness engi-
neers at the design stage to understand it, and the aircrew who may then have to deal
with the consequences, are undemanded variations in performance—for example
differences in performance between port and starboard mainwheel brakes, which
could cause a partial loss of directional control during a braked landing roll, or a
navigation system which suffers a drift in indication whilst still producing an
apparently normal display causing reduced situational awareness.
16.2 System Failure Numeric Analysis 321
The complexity of this array of “problems” creates a need for three sets of
concepts:
1. The identification of realistic “partial failure” modes of systems for inclusion in
the systems failure analysis described above.
2. Means by which these failures can be identified, ideally before flight, but par-
ticularly during flight.
3. How those failures can be managed and mitigated.
It is easy to obsess with systems failure analysis, but of course most of the time,
most aircraft systems, will be behaving normally and must be managed and
understood as such. What “normal” is, needs to be well understood, tested, and both
ground and air crew provided with procedures to manage systems within normal
operating parameters.
For more complex aircraft, both civil and military, it is normal to create an Iron
Bird (more properly, but rarely, called a Systems Integration Test Bed), which is a
simulation of all of the aircraft systems laid out on the ground with representative
structure and loads around them. This will be run on an accelerated series of
simulated flight cycles so that well before a real aircraft flies, a good level of
systems operating experience has been obtained, and what were predicted to be
normal operating parameters and procedures have been provisionally confirmed to
work. Typically the operation of the iron bird will continue well beyond flight
testing and the entry into service of the aircraft, so that at any point in the aircraft’s
service life, there should be ground experience of systems use with significantly
greater ageing than in flight so that potentially catastrophic failures or degradations
can be dealt with long before they are ever seen in flight.3
The latest modern practices link the iron bird with a flight simulator, ensuring
that the flight cycles tested are as accurate as possible representations of real flight
loads. The principle of the iron bird has also been adopted by the space industry
who will often evaluate satellite systems with the mechanical or fluid flow com-
ponents laid out on a very large flat table (so as to minimise gravity gradient
effects), leading to that industry’s term flat-sat. The principles and purposes of the
flat-sat and iron bird are substantially the same, although the pressures are subtly
different: few spacecraft offer the subsequent opportunity for rectification once in
3
A fictional but accurate treatment of this, particularly considering structural failure prediction is in
Nevil Shute’s brilliant novel No Highway (ISBN 978–0099530091); this was filmed as a major
Hollywood movie No Highway in the sky, with the hero: Aeronautical Engineer Theodore Honey
played by James Stewart.
322 16 Systems Assessment
service, whilst spacecraft failures seldom have potential to cause the loss of life that
a large aeroplane accident can.
Once the aircraft systems have been adequately tested on the ground, and this
used to develop maintenance, operation and rectification procedures, a prototype
aircraft or system can be flight tested. For simpler aeroplanes—single engine/fixed
pitch propeller/fixed gear aeroplanes the aircraft systems testing will be minimal
and incorporated in the main handling and performance testing programme,
although the increasing incorporation of complex cockpit electronic systems is
changing that. For larger aeroplanes however, systems testing will now comprise
the major part of the development then certification flight test programme. This will
be a case of identifying the various system mode combinations and the flight
conditions where they will be used, predicting the aircraft’s behaviour, then eval-
uating and confirming that behaviour in flight. Given the complexity of a modern
airliner or combat aircraft, this testing phase can potentially require several test
aircraft, and several years. The basic process is however predictably methodical—a
case of constructing the elaborate grid of conditions which may potentially occur,
and then determining an efficient way to evaluate all of those in as few flying hours
(and months!) as possible. The right balance of thoroughness and efficiency in this
will be difficult to achieve and will be the subject of continuous debate.
Aspects of the philosophy behind this has necessarily changed with the steady
shift from analogue (electrical, mechanical and electronic) systems to a much
greater reliance upon digital electronic systems, driving a reduced mass of (much
heavier) analogue systems. In theory, a digital system should be more predictable
and amenable to ground testing—in practice this has proved not to be the case, and
the systems flight test component of a certification programme is much larger than
previously; this trend shows no tendency to reverse. That most high value modern
aircraft use digital fly-by-wire systems, only exacerbates that, with the handling
qualities flight test programme of an aircraft such as Typhoon needing to be much
greater in length and complexity than the Panavia Tornado—which used a primarily
analogue FBW system, and in turn greater than the F-4 Phantom II which it
replaced, and used a hydraulic flight control system with minimal computing input.
The same is true in civil aircraft if one compares a fully fly-by-wire aircraft such as
the Boeing 787 to the older Boeing 737 with stability augmentation but essentially
conventional hydraulic controls, and then the Boeing 707 with primarily unpowered
reversible controls. In Europe, a similar progression can be seen in the A320, before
it the A300 which had powered and augmented but essentially direct controls, and
before that the de Havilland Comet with mechanical reversible controls and no
augmentation. In each case, the flight test programme was longer than the previous,
and the proportion of systems testing also substantially greater (Fig. 16.3).
16.3 Systems Testing and Performance Identification 323
Fig. 16.3 Progression of systems intensity in airliners: top to bottom, de Havilland Comet, Airbus
A300, Airbus A320
324 16 Systems Assessment
As with any other electrical equipment, all electrical aircraft systems have at-least a
theoretical potential to fail in an undesirable manner—causing potentially over-
heating and even fire. For this reason all systems must be able to be isolated.
A normal part of the aircraft systems assessment therefore is to ensure that every
electrical system can be discretely isolated with a switch, circuit breaker or fuse
(modern practice will most commonly use a switch and circuit breaker in line with
each other: fuses are relatively rarely used now). These need to be accessible to
aircrew so that systems can be shut down should they malfunction—the most
critical systems within the cockpit, but in larger aircraft it may be necessary to
locate banks of circuit breakers within a separate avionics bay—typically a com-
partment below the cockpit an accessed through a hatch in the floor. The ability to
do so needs to be confirmed as part of the airworthiness evaluation, as must the
utility of manuals, placards and checklists providing aircrew with instructions in
their use. Where circuit breakers or fuses (fuses are highly unusual now, but will
commonly be found in aircraft built pre 1980) are used, they must also be
re-settable from a position that crew can access. Normal pilot training advice is that
if a circuit breaker on a piece of equipment fails it may be reset once, but left alone
for maintenance action if it fails a second time. Obviously all fuses and circuit
breakers that might need to be accessible in flight must also be readily accessible
and clearly labelled. Lighter aeroplanes (microlights, part VLA, part 22, lighter part
23) may often use these circuit breakers also as equipment switches, although on
larger part 23 aeroplanes and part 25 aeroplanes this is prohibited unless the units
have been specifically designed for use in this way.
It is also in the nature of all electrical and electronic systems that they will
generate electromagnetic fields, and this can create mutual interference between
devices. Anybody who has inadvertently left their cellphone next to a loudspeaker
will probably be familiar with these interference effects.4 A modern aircraft is
clearly a far more complex set of systems than this simple example, and so the
potential for mutual electronic interference is significant. Additional complexity is
created by external RF generators such as ground radar, radio transmitters or power
lines. As a matter of course, aircraft systems are tested during approval of new or
modified electrical installations where systems will be started up in turn until all are
running together, with functional checks run on all potentially affected systems to
determine whether any unacceptable levels of mutual interference exists. For mil-
itary aeroplanes this is particularly critical as this testing must be extended to
electrically managed weapons and emergency explosive systems such as ejector
4
Can mobile phones interfere with aircraft systems? Probably—BAE Systems have reported, at the
very least, one definite incident. This was, apparently, a Nokia 3210, at seat 14F, which interfered
with Honeywell GNS-XLS Flight Management System in the cockpit of a BAe-146. That one
incident, for now, justifies the automatic ban of most functioning mobile phones in most airliners,
until the combinations have been explicitly proven safe.
16.4 Electrical and EMC Considerations 325
seats and canopy fracture and military Test and Evaluation (T&E) organisations
will also use a facility called a Radio Electrofrequency Generator, or REG, to
simulate the wide range of RF spectrum power and frequency combinations which
might be experienced in service.5
Duplication of electrical supply is normally required in all but the most simple
VFR only aeroplanes
Virtually all aircraft require some form of backup electrical supply. On smaller
aeroplanes this is likely to simply be provided by a reasonable battery capacity, and
advice in operating documentation about the capacity (and thus time available from
engine or generator failure to full power failure). On larger aeroplanes, auxilliary
power units (APUs), duplicated engine generators and even ram air turbines (RATs)
are likely to be used, combined with system hierarchies and predicted failure rates
that achieve acceptably low overall electrical power loss risks. [APUs and RATs are
usually multi-function, providing electrical, hydraulic and potentially from APUs
conditioned air supplies—which might be provided by separate devices fitted to
5
Also cars! Some models of older British car, particularly the Rover 200 were known to routinely
refuse to start in the car parks at London Gatwick Airport and had to regularly be towed a quarter
mile from the airport at which point they would start straight away.
326 16 Systems Assessment
main engines.] A small number of piston engined aeroplanes use battery driven
electrical ignition systems—such must also be backed up, although they are unusual
—most piston engines use magneto ignition systems that are separate to the elec-
trical generator used for other services. In turn, the majority of these magnetos are
also duplicated as a matter of course.
Demonstration of all electrical supply capabilities must be modelled (a very
simple task for a microlight, considerably less so for an airliner or combat aircraft—
other classes sit in between), tested on the ground, then if absolutely essential,
tested in the air—including shutting down of various generators and bus-bars to
ensure adequate power delivery in degraded states.
Aircraft systems will see a wide range of environmental conditions during service,
which are difficult to readily design for. These issues are generally ignored for light
aircraft, and to a certain extent only limited assessment is carried out for large
aeroplanes—limited typically to contaminated runway testing and evaluation of anti
and de-icing systems. (Anti icing systems are those which prevent ice accretion,
whilst de-icing systems remove ice already formed.)
Military equipment will typically be tested to a much wider range of environ-
mental conditions—obviously whole aircraft will be evaluated against basic
requirements to evaluate controllability and performance degradation on contami-
nated runways, and any anti- and/or de-icing systems will be tested in known icing
conditions. Most usually this is done in natural conditions, although this can be
difficult to achieve reliably and a normal practice is to identify the desired condi-
tions and the earliest point in a flight test programme where exposure to progres-
sively worse known icing conditions can be accepted. Then, on an opportunity
basis, a test aircraft will be diverted from an ongoing flight test programme to
permit evaluation in those known icing conditions. It should be mentioned that the
first flight in known icing conditions is classed as high risk testing.
An alternative approach is particularly used for aircraft that do not possess
anti-icing systems but seek a limited clearance to fly in known icing conditions (or a
contingency clearance—where it is accepted that brief exposure may occur before
deliberate flight out of it). In this case either a specialist ice accretion prediction
modelling program such as NASA’s LEWICE or SmaggIce packages is used to
predict the shape of ice build up on the airframe, or aircraft components are exposed
to artificial icing conditions on the ground to determine experimentally the shape of
ice accretion. Then, artificial ice accretion is created, probably using rapid proto-
typing technology, and attached to a flight test airframe, which (following usually
some prior flight mechanics modelling to predict the likely performance and handling
effects) will then be test flown. This is still high risk testing, given the relatively
unknown changes to performance and handling, but a viable evaluation method
typically used on top-end part 23 aeroplanes such as 9–19 passenger seat turboprops.
16.5 Environmental Testing 327
We had 103 years of flying experience there in the cockpit, trying to get that airplane
on the ground, not one minute of which we had actually practiced, any one of us. So
why would I know more about getting that airplane on the ground under those
conditions than the other three. So if I hadn’t used [CRM],6 if we had not let
everybody put their input in, it’s a cinch we wouldn’t have made it.
Captain Al Haynes, US Airlines Flight 232
6
CRM in this context is Crew Resource Management, a set of concepts of best practice in
communication and decision making in a stressed and safety critical environment.
328 16 Systems Assessment
the interaction of human beings with each other is substantially outside of the
practice of initial airworthiness assessment.7
Whilst none of these four should be ignored, the practice of initial airworthiness
assessment will tend to revolve around the first three of these: software-liveware
and hardware-liveware which will be considered here in reverse order.
16.6.1 Hardware—Liveware
The most obvious part of the hardware-liveware interface is the cockpit. For civil
aircraft, technically, it is entirely at the discretion of the design team whether they
design this solely for midgets or Masai warriors—in practice either will be some-
what sales limiting so it is important that a cockpit has sufficient scope to allow a
wide range of people to occupy it and function effectively. A typical “full popu-
lation” model might for example require a range from 5th percentile adult oriental
female to 95th percentile Caucasian male—although exceptionally 1st to 99th
percentile populations might be explored. Military standards will generally be much
more prescriptive as they both can do, and must do so—a military organisation is
dealing with a small known population who should be (compared to the back-
ground population) relatively fit and lacking overweight individuals—at the same
time the military will value optimal performance more than civil operators who
generally only seek acceptable levels of performance.
The dimensions of the human body can be defined by series of key dimensions.
NASA in NASA-STD-3000 gave (amongst much other valuable data) the dimen-
sions in 2000 for 40 year old Americans males and Japanese females, the core
points of which are shown in Table 16.2.
Modern computer based design tools make it possible to evaluate these and other
dimensions against prototype cockpits or other workstations. However, it is
nonetheless common practice to identify for any assessment programme the largest
and smallest members available within the available team to take part in more active
assessments of workstations, cockpits, or even (particularly for military applica-
tions) clothing, known typically as the AEA or Aircrew Equipment Assembly.
One particularly important aspect of these assessments relate to what is known as
the Design Eye Position, or DEP—that is the location in space within the cockpit
where the pilot’s or pilots’ eyes are assumed to be. It is important to assess that
visibility both of the critical instruments within the cockpit, and externally are fit for
purpose; the classical method of illustrating the external view from a cockpit is an
equal areas modified Mollweide projection called a Hammer diagram—an illus-
tration of which is given in Fig. 16.5 for the SEPECAT Jaguar, a 1980s era ground
7
Readers seeking a good introduction to human factors and their contribution to error in aviation
and other activities might do well to start with “The Human Contribution” by James Reason
(ISBN-13: 978–0754674023) and the UK CAA’s document CAP 737 “Flightcrew human factors
handbook”. Both are very extensive.
330 16 Systems Assessment
Fig. 16.6 Hammer diagram for the Captain’s seat of a Boeing 737
attack aeroplane: combat aeroplanes and all helicopters are the classes of aircraft
where this is most critical, but the approach is valuable for all aircraft. Obviously
the example given here is for an aeroplane where the pilot sits on the aircraft
centreline—for most aircraft of any class, these diagrams will be asymmetric and it
is advisable to produce one for each seat—for example Fig. 16.6 is for a Boeing
737. A relevant aside is that these diagrams do not illustrate one important point
about the view out of a cockpit, which is that obstructions narrower than the
distance between the pupils (ranging from 53 to 70 mm in adults [American male:
5th and 95th percentile 55 and 70 mm respectively, American female: 53 mm and
65 mm]) will not in reality provide any obstruction to seeing objects at a long
distance, whilst obstructions wider than that, will do so. Therefore any obstructions
wider than 52 mm should be of particular interest in cockpit assessments, irre-
spective of angular projection.
Not unrelated to this, a consideration in many aircraft is lighting. For pilots the
lighting in the cockpit is critical during night operations is very important—for the
ability to see all of the instruments clearly, along with operational documents in the
cockpit, whilst also not making the cockpit so bright as to make it difficult for pilots
to see outside the cockpit. Whilst pilots engaged in night flying operations will of
course carry torches, etc. to aid this, using those significantly increases workload
and should not be necessary. For some specialist aircraft (e.g. helicopters in military
or police support roles) this problem exaccerbates if pilot mounted night vision
devices are worn, as these will require extensive matching and evaluation to ensure
that their characteristics are compatible with internal cockpit lighting, whilst still
332 16 Systems Assessment
being able to work as required looking outside of the cockpit.8 Externally, care
should also be taken over stroboscopic anti-collision lights, which if reflected back
into the cockpit (for example during flight in cloud) can create unpleasant, dis-
tracting, and in great extremis incapacitating effects in the cockpit. Frequencies in
the range 4–20 Hz are usually most problematic.9)
Vibration, particularly in the 1–40 Hz range is also problematic, depending upon
magnitude. Standard references suggest 1–4 Hz can affect breathing, 4–10 Hz
cause chest and abdominal pain, 8–12 Hz cause backache, and 10–20 Hz cause
headaches and muscular tension, and 30–40 Hz interfere with vision. The research
basis behind these values is a little tenuous, so these values and claimed effects
should be treated with caution and probably verified with experimental work in a
representative environment.
A further assessment which is prudent for some applications is to formally assess
operator workload. This requires experienced operators, who should be allowed to
familiarise themselves with the cockpit or workstation, and then carry out both
routine and non-standard tasks (such as handling simulated in-flight emergencies),
during which their workload should be assessed. There are various means to do
so—including for example heart-rate monitoring, but the most common is a series
of questionnaires. The most universally used system for this is a US originated
system called NASA-TLX (Task Loading Index), illustrated in Fig. 16.7, although
other systems such as the Bedford and Cranfield workload scales also exist and may
be preferred for some applications.
16.6.2 Software—Liveware
Typically operating data for aircraft and systems will fall into five categories, some
or all of which may exist for any given aircraft or system. These will be:
1. Checklists, for both normal and emergency operations—these are typically used
during actual operations
2. Expanded checklists primarily used in a training environment
3. Operators Manual (also called pilots notes, flight manual, pilot’s operating
handbook, or other terms depending upon aircraft, system and/or environment)
4. Maintenance manuals
5. Placards
8
Night Vision Goggle qualification is a developing body of knowledge, and the practitioner should
conduct an up to date literature review if entering the field. A useful starting point however is
NASA Technical Memorandum 101039 “Helicopter Flights with Night Vision Goggles—Human
Factors Aspects” published in March 1989.
9
Allied ground forces during WW2 used a strobe light weapon, operating about 6 Hz and
intending to distract enemy troops. In British service this was known as the “Canal Defence
Light”, and in American service the “T10 Shop Tractor”; both names deliberately ambiguous.
German searchlights may have on occasion used a similar principle against bombers.
16.6 Ergonomics: The Human in the System 333
Good practice is that such of these as will be required should be written during the
initial development process of any aerospace product—but at an appropriate level
of complexity to the product—whilst a very comprehensive suite of documentation
will be required for a whole aeroplane, for a relatively minor system such as a new
cockpit display considerably less will be needed. Then all should be reviewed for
utility, practicality, accuracy and workload throughout the airworthiness evaluation
process.
16.6.3 Environment–Liveware
To some extent this has already been discussed in Chap. 2, concerning atmospheric
conditions and their impact on the human beings in an aircraft. However there are
numerous other aspects of the surrounding environment which can also impact
upon human performance, and fall within the province of airworthiness. Comfort,
vibration, temperature, light and noise levels all will have significant impact, and
design and airworthiness engineers can do much to alleviate these. Much of this
alleviation may fall within the province of the “aircrew equipment assembly” or
AEA—the set of equipment and specialist clothing worn by crew or some pas-
sengers, which is often an airworthiness issue. For example there may be aspects of
helmet, sunglasses, flying clothing selection where design and airworthiness can
influence and improve best practices. Some other aspects are not—such as stress,
working hours, and the conditions crew enjoy (or endure) outside of the aircraft.
Once the system overall is deemed airworthy and fit for entry into service, higher
value aircraft will then normally go through further stage, which the military call
Operational Test and Evaluation or OT&E (compared to the earlier phase
Development Test and Evaluation, or DT&E), although does not have a defined
term in civil aviation. In this phase, experienced operators will take on the aircraft
and evaluate it in the operational environment—the military would simulate
warfighting, an airline would fly representative route sectors operating as if carrying
paying passengers, and a training organisation would evaluate the aircraft as if it
was being used for flying training. In all cases the objective is that these experi-
enced operators will generate “final use” operating documentation, based upon the
earlier documents which meet and suit their particular operating requirements and
fit their available skillsets.
A modern military approach, at least within NATO countries, is that the DT&E
and OT&E processes are combined such that the company airworthiness and flight
test departments, government employed evaluators, and front line operational
evaluators work in (theoretically) a single combined evaluation team. This is typ-
ically called a Combined Test Force, or Integrated Test Team. The need for this
approach is generally accepted now after initial controversy, but considerable
debate continues about where this process should be led with differing approaches
taken in different places. The most expert flight test evaluators tend to favour
customer leadership being at the government flight test centre (called an OTC, or
16.6 Ergonomics: The Human in the System 335
Official Test Centre), but they do not necessarily hold the greatest political power
when making such decisions. It may be pragmatic that leadership in reality starts
with the government agency, but moves partway to the OTC, and concludes with
the OT&E organisation. The OT&E process goes beyond, but particularly con-
centrates upon operating practices and documentation.
Chapter 17
Environmental Impact
When aircraft were first developed, there was probably no consideration at-all of
their emissions. Subsequently this has become a matter of massive public concern
on four levels. These are:
(1) Noise,
(2) Greenhouse gases,
(3) Particulates and surface pollutants, and
(4) End of life wastage.
For military purposes only, a fifth emission category also exists, which is of heat
and radar signatures, and consequent detectability. This will not be considered in
depth here.
Whilst much of the recent discussion about aircraft emissions in recent years has
concentrated on greenhouse gasses, and in particular Carbon Dioxide (CO2), the
reality for anybody living close to an airport is that these are entirely abstract issues.
For anybody living close to any airport, or underneath either the approach or climb
out paths of any major transport airport, noise emissions are everything. Probably
every major airport in the world has resident protest groups about noise, and
invariably there are regular political concerns, protests and hopefully regular
community engagement meetings about noise between airport and community
stakeholders. This has been reinforced by research that has shown a significant
correlation between noise exposure and human health.
This fundamental issue of noise then, has become an issue of the ability to
operate—authorities globally are setting increasingly stringent rules that must be
met—and aircraft which do not meet these rules are not permitted to operate.
Therefore, whilst not a safety issue, noise has become very firmly an airworthiness
issue.
Noise legislation in aircraft emerged from international discussions in the 1960s,
and for airliners the first major piece of legislation was “Chapter 2”1 implemented
in 1972 and providing the first binding standards for noise emissions. Chapter 2
used measurements of external noise at three points—directly below the take-off
flightpath 3.5 nm (6,500 m) from brake release, the highest reading at any point
450 m off the runway axis, and 1 nm (2,000 m) before the runway threshold under
the approach path, normally on a standard 3° instrument approach. These three
readings were initially added up to provide a nominal “cumulative level” noise
value. Subsequently as volume of air traffic has expanded, and often tolerance to
noise has become less, a series of chapters have been implemented either bringing
classes not previously considered into legislation, or tightening the standards—with
a classic case occurring of technology and regulation driving each other
(Table 17.1).
Chapter 3 superceded Chapter 2, which was only ever an initial standard, and
defined more complex limits based upon each type of measurement and the max-
imum mass of the aeroplane (Table 17.2).
1
Chapter 2 to ICAO Annex 16 is “Noise certification of subsonic jet and large propeller driven
aircraft”.
17.2 The Impact of Noise 339
Additional complexities for airliners have defined precisely how the take-off
reference procedure had to be flown (with a minimum altitude for power cutback on
initial climbout increasing from 210 m (689 ft) for a four engined aeroplane via
260 m (853 ft) for a 3-engined aeroplane to 300 m (984 ft) for singles and twins),
and not permitting power cutback below that required to maintain a 4° climb
gradient or level flight with the critical engine inoperative. Everything is stan-
dardised using approved procedures to ISA sea-level pressure, ISA + 10 °C and
70% relative humidity in still air, with testing carried out between −10 and 35 °C,
20% < RH < 95% and wind not exceeding 12 kts (maximum 7 kts crosswind) at
10 m (32 ft) above the ground.
At time of writing (2017) a new Chapter 14 is incipient, looking to be imple-
mented in a staged manner between 2017 and 2020, introducing a similarly con-
structed, but still more stringent equivalent to Chapter 4. Already ICAO is debating
yet more stringent requirements again through 2030, and the introduction of these
seems inevitable, although the detail still unclear. Figure 17.1 shows the progres-
sion of total noise permissions from the 1970s to 2020s, and as both technologies
improve, and intolerance to noise increases, one may assume that future chapters
will be more stringent again—although also future “super-super-jumbo” and
supersonic classes may see this diagram made more complex by heavier and faster
aeroplanes.
For other classes of aircraft, inevitably procedures have been developed and
mandated, but compared to the large aeroplane case, varied to reflect the nature of
their operations. Smaller aeroplanes under Chapter 6 were initially tested using a
340
simple flypast at Maximum Continuous Power (MCP) 300 m (984 ft) above the
instrument, and this developed under Chapter 10 to a take-off flightpath measure-
ment underneath the take-off path at Maximum Take-Off Power (MTOP). Large
helicopter testing under Chapter 8 has involved a best rate of climb/MTOP climb, a
150 m (500 ft) overflight and a 6° approach; small helicopter testing under Chapter
11 has involved a single point overflight at 150 m (500 ft). Similarly to safety
related airworthiness standards, this has effectively scaled the level of proof to the
level of risk (or in this case, more accurately, nuisance).
Historically Concorde and the Tupolev Tu-144, the only two civil supersonic
aircraft were permitted considerable leniency in noise standards. Presently various
programmes are proposing new, smaller, supersonic business jets such as the
“Boom” supersonic transport being sponsored by Virgin group (Fig. 17.2). New
noise standards will be developed for these reflecting what is possible, but also a
likely high degree of intolerance to sonic booms and the noise of high thrust
turbojet in engines, in regions which have not previously experienced those.
The source and control of noise is a complex issue—and just as an aircraft
should be designed to meet safety standards, increasingly it must also be designed
to meet noise requirements. The aircraft designed disregarding those standards, is
likely to then prove problematic to gain operating approval for. The sources of
noise will be multiple—obviously the majority component being the engines and
the engine installation, particularly on the climbout path. Additionally however
aerodynamic drag (especially due to flaps, undercarriage and creators of interfer-
ence drag) is a substantial player, most especially during the final approach.
342 17 Environmental Impact
Because of the complex nature of noise generation, so also the solutions tend to be
multiple, requiring a large number of design solutions, each making a small but
iterative contribution to noise reduction on an aircraft. Consider for example the
Airbus A319 and A350 in Figs. 17.3 and 17.4. Whilst the A320 family were state
of the art when they entered service, it is visible in the newer A350 that consid-
erable effort has gone into smoothing contours, reducing edges and protuberances,
fairing undercarriage components… any increment that could reduce noise (which,
conveniently, also reduces drag and thus improves operating efficiency).
A further major component of noise reduction is operational. Measures such as
Continuous Descent Approaches (CDA) and Performance Based Navigation
17.2 The Impact of Noise 343
(PBN) initial climb out paths can be used, in conjunction with noise transmission
models and population maps to minimize noise nuisance to populations. Both
climbout and approach procedures can be designed to reduce noise emissions—for
example airbrakes (typically mounted on the top of a wing) will generally create
less ground-perceivable noise than extended undercarriage, and of course drag
induced noise will be greater at higher speeds. So an approach procedure that
extends landing gear as late as possible, and at as low a speed as possible, is likely
to be quieter to ground observers. To manage this however requires procedures,
controls and especially navigation systems, almost certainly coupled to autopilot,
that will ensure sufficiently precise adherence to designed best procedures. At
present some older types—the Boeing 747 for example, are proving to not to have
sufficiently fine navigational control to be able to adhere closely enough to the latest
climb-out noise control flightpaths.
For smaller, and usually propeller driven aeroplanes, design solutions will differ
to those for larger aeroplanes. Control of propeller tip speed can be useful: through
a combination of pitch, diameter, and engine speed—and piston engines and some
turboprops can enjoy reduced noise emissions through intelligent exhaust design—
using mufflers and routing exhausts upwards. Drag induced noise is far less of an
issue for light aeroplanes, because of their lower speed and whilst similar opera-
tional procedures can be applied, the use of autopilot coupling to deliver
noise-optimised flight paths probably cannot.
Other non-airworthiness operational practice used to reduce perceived noise
include scheduled re-routings of flight paths (termed respite), towing aircraft to the
runway before engine start, locating ground power and air conditioning sources
underground, earth bunds used to deflect noise upwards, displaced take-off points,
funding of noiseproofing of homes, and night flying curfews. All of these can have
significant benefits to communities around airports, but are not in themselves air-
worthiness issues.
344 17 Environmental Impact
Fig. 17.5 Airbus A320 underwing aperture showing noise reducing vortex generator (courtesy of
Iain Clark)
17.3 Greenhouse Gas Emissions 345
The fact of climate change, and that the major contributors to it are man-made, is
well known. It is also well known that the major contributing factor is the emission
(both man-made and natural) of substances that have come to be known as
greenhouse gases (GHGs) and their movement into the stratosphere.
Although the science had been known since the early 1970s, the first major effort
to reduce GHG emissions was the 1989 Montreal Protocol on substances that
deplete the ozone layer, more commonly just referred to as the Montreal Protocol.
This had a very immediate impact upon aircraft, because it restricted the use and
implemented steady withdrawal of a set of chemicals known as Chlorofluorocarbons
(CFCs), Hydrochlorofluorocarbons (HCFCs) and Hydrofluorocarbons (HFCs). This
group included 1-1-1 Tricholoroethane, which was widely used as a solvent in
aircraft manufacture and maintenance, and the Halon family, which are an excep-
tionally effective and lightweight firefighting agent that (save asphyxiation) are
harmless to humans. The majority of these chemicals have been withdrawn
worldwide, including within aviation—the exception for now being Halon. Halon
production was ceased worldwide in 1994, but it is still being recycled and used on
aircraft as the safest and most effective firefighting chemical: however replacements
are under development—these are in use for handheld and lavatory firefighting
systems, although alternatives for engine firefighting remain elusive.
Carbon dioxide has subsequently become the most widely discussed GHG, and
its production is well understood. At first approximation, each unit mass of kero-
sene burned in an aircraft engine creates 3.15 units of CO2. This obviously means
that the more financially efficient an aircraft is, the more efficient it also is in terms
of reducing GHG emissions. Aviation emissions of CO2 were estimated in 2015 to
be about 2% of the worldwide total (using about 13% of world transport’s fossil
fuels),2 which might be argued to be small compared to the massive economic and
social importance which aviation has to the world. However, aviation has two
problems; firstly it is relatively unique in being unable to exist as yet without
hydrocarbon fuels due to their extremely high calorific value per unit mass, and
secondly annual passenger air traffic is increasing at over 4% per year—approxi-
mately doubling use every 15 years. This appears unsustainable in a world actively
trying to reduce its collective carbon footprint; some hope may be gained from the
fact that between 1965 and 2015 a per passenger-mile reduction in fuel burn of
about 80% has been achieved, but realistically gains now are much harder to
achieve—reductions in fuel burn due to engine efficiency and operational proce-
dures (not dissimilar to those already discussed to reduce noise) will continue, but
industry growth will probably still significantly outstrip efficiency improvements.
Following agreements at the 2016 ICAO general assembly, ICAO and IATA are
2
There are numerous reports detailing these breakdowns from the IPCC, or Intergovernmental
Panel on Climate Change. See www.ipcc.ch.
346 17 Environmental Impact
3
See https://www.icao.int/environmental-protection for the latest ICAO publications on this topic.
17.3 Greenhouse Gas Emissions 347
longwave (infra red) radiation back to the earth creating a net warming effect on the
earth, although may also create cooling during the daytime by reflecting short wave
(solar) radiation back into space. These effects are poorly understood as yet,
although an interesting hypothesis being widely discussed is that in the future
aircraft cruising altitudes might be set, using meteorological data, to optimize for
planetary cooling. Contrail formation and control will almost certainly be a matter
for future airworthiness evaluation, as this science becomes better understood and
can be exploited.4
A note on contrails
Contrails have been a matter of concern for military purposes for many years.
The Royal Air Force High Altitude Flight (HAF) was established at
Farnborough in 1942 specifically to investigate the formation of contrails—
and throughout that war they were in particular a telltale for high altitude
bombers. During the cold war of the 1950s to 1980s it was a similar telltale
for high altitude reconnaisance aircraft such as the U2 or Canberra PR9.
Research at that time generally failed to adequately predict contrail formation,
although did identify the specific mechanism. Cold war spyplanes mostly just
used a rear view mirror to monitor for their formation, then changed altitude if
this occurred (Fig. 17.7).
4
Readers wishing to explore this topic should look into the publications of Professor Keith Shine
and Dr. Emma Irvine at the University of Reading in the United Kingdom, and Professor Ulrich
Schumann at DLR in Germany.
348 17 Environmental Impact
Fig. 17.7 Cold War era British Canberra PR9 high altitude reconnaissance aircraft
It is difficult to define the scale of near surface pollutants from aircraft because
airports commonly suffer numerous other ground based polluters. This may include
cars, ground power units, trains, combined heat and power units and simply large
numbers of people.
Nonetheless there are certainly chemical emissions from aeroplanes on or near
the surface at airports—in particular unburned hydrocarbons, particulates, NOX
compounds, CO and CO2. There is some evidence also that rubber dust from the
touchdown point of landing aircraft can travel short distances and be harmful.5
Air quality around airports is important, and there are research papers indicating
that close to airports incidences of heart disease (this may be more noise related)
and respiratory disease increase.6 However, the impossibility of generally sepa-
rating aircraft and ground sources, mean that this is primarily a land use and
planning issue, rather than a pure airworthiness issue. Nonetheless, airport and land
use planners may often need detailed information on aircraft emissions to permit
them to predict and resolve air quality issues. Providing that information may
however be very difficult—such information is not at present routinely obtained,
although that may change in the future as regulations develop.
5
A good starting point to investigate this topic generally is the publications of Dr. Michael Bennett,
formerly of the UK’s Manchester Metropolitan University.
6
See Schlenker, Wolfram and Walker, Reed, (2016), Airports, Air Pollution, and
Contemporaneous Health, Review of Economic Studies, 83, issue 2, p. 768–809 which evaluates
this problem for 12 airports in California.
17.5 End of Life Wastage: The Recycling Problem 349
It is usual, both from environmental and economic imperatives, that at the end of
its life, there will be a desire to recycle an aircraft. Very large sums of money
have been sunk into that aircraft through its life, and also it will contain materials
that can be damaging to the environment if not properly removed and recycled or
disposed of.
Managing this process well depends initially upon that aircraft having been
subject to competent continued airworthiness. Records that show the origin, life
remaining and serviceability of parts already fitted to the aircraft allow those to be
removed and returned following inspection to the stores supply chain. Whilst it can
take months or years to realise that value, this can often in itself give an end of life
aeroplane, particularly of a type still widely in service, greater financial value than
its market value immediately prior to retirement. At the same time all fluids must be
removed from the aircraft: typically hydraulic fluids, lubricants and water based
fluids—it is unlikely that any of these will be fit for immediate re-use, but there are
many routes that permit contaminated fluids to be recycled, and they are likely
to be.
Of the remaining materials, steels and aluminium alloys in particular can nor-
mally be readily recycled and return to engineering use as new metal. Still good
through-life documentation and understanding of the aircraft is critical; for exam-
ple, the author is aware of an incident when a Boeing 747 was recycled by the
“lowest compliant bidder” who inadvertently included in aluminium alloy sent for
smelting a significant amount of depleted uranium that had been incorporated into
the aircraft during service as ballast to control centre of gravity position.
A growing problem with modern aircraft is the use of fibre-polymer composites
such as CFRP, which are not easily amenable to recycling. However, research is
ongoing in this area, and increasingly carbon fibre composite material and similar
will be finely shredded and then used as the base for manufacturing tertiary aircraft
structure (such as interior trim) or similar non-aviation applications as a lightweight,
bulky, but not structurally critical material. It is becoming increasingly clear that
aircraft designers must consider the end of life recycling process for such materials,
and it is likely that this will become at the very least a de facto airworthiness/
certification requirement, if not a legally binding one. It is particularly becoming
important to companies to be able to demonstrate that they are giving the maximum
consideration to recyclability of such parts.
Sometimes recycling of aircraft can become more imaginative—and again this
may require the involvement of airworthiness engineers, as they are the specialists
best placed to understand how to ensure aircraft parts are fit for their new role. For
example Fig. 17.8 shows the forward fuselage of a DC10 airliner which at end of
life was recycled at Manchester Airport in the UK as a lecture room and classroom
—including the use of recycled aircraft seating (Fig. 17.9).
Whilst it is unlikely that the initial airworthiness standards detailed earlier in this
book are appropriate for sign-off of a structure of this nature, they are known and
350 17 Environmental Impact
define the minimum strength and integrity of the starting component. It is therefore
relatively straightforward for engineers working cross-discipline to map airwor-
thiness standards to the building and health and safety standards applicable—thus
allowing recycling of large aircraft components into their new role without
wholesale re-analysis.
It is currently estimated that of more recent aircraft (such as for example an end
of life A320 or Boeing 737-700) about 90% of the aircraft can be recycled, whilst
17.5 End of Life Wastage: The Recycling Problem 351
an older type with limited re-use interest in parts such as a Boeing 747-200, about
50% can be recycled. The remainder will end up in landfill, a recycling industry
centred on the Aircraft Fleet Recycling Association (https://afraassociation.org/),
estimating about 600 large aircraft being recycled annually worldwide, is however
steadily managing to improve these values.
Chapter 18
Facilitating Continued Airworthiness
The bulk of this book is about initial airworthiness: that is deciding whether an
aircraft or system is fit for entry into service. That function is vital, but equally vital
is ensuring that it stays both compliant with an approved design standard, and fit for
use. An aircraft, like any engineering product, will deteriorate with use, but also
may be found to contain designed or manufactured-in deficiencies that require
rectification—the initial airworthiness process cannot be assumed to be infallible.
This chapter will introduce the reader to the basic principles of continued airwor-
thiness, which aims to ensure that through life an aircraft is fit for use, and that an
approved design standard, and actual build standard of the aircraft, co-incide.
The importance of this is defined in the (US) FAA’s definition of airworthiness,
which is:
The aircraft conforms to its type design, and; it is in a condition for safe flight.
DI The daily inspection, carried out before each flight, or the first
flight of the day. This may often be carried out by a suitably
qualified pilot, rather than a technician, engineer or mechanic.1
A-check The lowest level of inspection and maintenance, typically carried
out every 500 flying hours or 3 months (whichever is the lower).
On all but some light aeroplanes, this will require a technician,
engineer or mechanic to carry out the inspection.
B-check Inspection and maintenance typically carried out at about 6 month
intervals or after 1,000 h of operation.
C-check Inspection and maintenance typically carried out at 12–24 month
intervals, or after 1,000–3,000 h of operation. This may sometimes
be referred to as an “Annual” and can often require the aircraft to
be grounded for several days or weeks.
D-check Inspection and maintenance typically carried out at 3–5 year
intervals, or after 5,000–10,000 h of operation. This will require
substantial disassembly of the aircraft and inspection or overhaul
of a great many components. In some regimes this will be referred
to as a “Star-Annual”. In other regimes it does not exist as the A to
C checks are considered to cover all that is important.
All of the flying hour figures shown here are representative of a 1990s era
airliner such as the A320 or B737. Most other categories of aircraft are likely to
require inspection and maintenance at much lower numbers of flying hours—for
example in many light aeroplanes the A-check co-incides with the DI, the B-check
is likely to be at about 50 h, and the D-check every 1,000 h or 5 years (see for
example Fig. 18.1). Combat aircraft may have much more sophisticated means of
determining maintenance intervals, based upon aircraft useage, whilst intervals are
significantly greater in modern high value civil aircraft such as the A380 or B787.
1
The terms “engineer”, “mechanic” and “technician” mean different things in different places and
regimes. The implication here is somebody trained and qualified to carry out and certify main-
tenance actions upon an aircraft. In Britain the term “licenced engineer” would be most common,
whilst in the USA “certified mechanic”, and Europe will tend to use variations upon the term
“technician”. Considerable offence amongst highly qualified professionals can result from using
the wrong term in the wrong place; this should be avoided!
18.2 Constructing Maintenance Procedures 355
Fig. 18.1 Extract from a typical generic light aircraft maintenance schedule (UK CAA’s “LAMS”
Light Aircraft Maintenance Schedule, reference CAP 411)
The most rigorous inspections, particularly the B and C checks may well be
carried out staged in small parcels to minimise the amount of time that the aircraft is
withdrawn from operational use. This approach is more expensive of time and
resource overall, but can often make good sense to an operator trying to get the
maximum use of valuable assets. Similarly operators with multiple aircraft will
stagger aircraft usage so as to ensure that the major inspection and maintenance do
not occur at the same time—which would otherwise result in both a lack of aircraft,
and overloading of maintenance capacity.
There are international minimum standards for civil aircraft that define the
training and qualification of staff who may carry out maintenance; these are main-
tained by ICAO. In practice, the most stringent standards will only normally need to
be met by the senior maintenance staff who are in a supervisory and certifying role,
whilst much of the work is carried out by less qualified staff under their direction.
The specific details of these qualifications vary internationally, and are subject to
regular revision. So, latest and local authority publications should always be referred
to for guidance on qualifications and standards. The qualifications required for
maintenance of military aircraft will follow national decisions and priorities, and are
also subject to regular revision, but they will tend to broadly follow the same general
pattern as civil requirements and indeed, many military organisations now will offer
dual qualification opportunities as a recruiting incentive. There are two main
exceptions, the first is that civil aircraft rarely have any requirement for maintenance
staff qualified to maintain weapons systems and ejection seats—these specialist
maintenance staff are commonly referred to as “armourers”; the second is that
military organisations increasingly use narrower specialisms allowing more rapid
training and reflecting fixed length military careers. The main (non-explosive) areas
356 18 Facilitating Continued Airworthiness
test maintenance procedures and practices which were developed in draft form
during the design process, can be pressure tested and revised. Military organisations
may further refine these during the Operational Test and Evaluation (OT&E)
process.
These procedures and any associated guidance and information must always be
controlled documents, permitting the overseeing design organisation throughout the
aircraft or system’s life to revise them whenever issues arise which may require
changes. Historically these were paper documents with amendments sent out on
paper (or electronically transferred in printable form), and this remains the case for
most part 23, part 27 and sub-ICAO aircraft. However, the majority of part 25, part
29 or operational military aircraft will use some form of continuously amended
secure online resource which will be accessed continuously by maintenance staff to
look up required information which will always be up to date. (Although, experi-
ence has shown that using this in any electronic form at the “shop floor” seldom
actually works, so it will still be printed out for use—but then usually discarded,
including doubtless oil stains and handwritten notes, immediately afterwards.)
Additional documents will also need to be generated as part of this process.
These are described below, but the actual titles will again often change with reg-
ulatory regime:
Type Certificate Data Sheet (also known, amongst other titles, as “Military
Aircraft Release”, “Type Approval Data Sheet”). The TCDS defines the approved
configuration standards for an approved aircraft and lists the titles and approved
modification states of derivative documents such as the flight manuals and main-
tenance schedules.
Master Minimum Equipment List The MMEL will normally be defined by the
aircraft design owner for aircraft used in commercial operations, and will show the
designer’s opinion of the absolute minimum of serviceable equipment required for
any given type of operation. For example, an aircraft may be fitted with three
altimeters, but the MMEL will define two as essential, with a maximum time
permitted before an unserviceable third must be repaired or replaced.
Minimum Equipment List The MEL is normally defined by a commercial
operating organisation, and is based upon both the MMEL and their own operating
preferences—for example crew training, their Safety Management System (SMS)
assessments, and the specifics of their own operating environment. Normally this
may not be any less restrictive than the MMEL.
As described here, the precise concepts of an MEL and MMEL are unique to
civil commercial operations; however the broader concepts of a minimum level of
equipment that should be serviceable to commence or continue a flight will be
found somewhere in the operating documentation for most aircraft—subject
perhaps, particularly in the case of military aircraft, to local operational expediency.
358 18 Facilitating Continued Airworthiness
The first line of airworthiness oversight starts with local documentation of main-
tenance and inspection actions. Typically this will be in two documents, the “tech
log” completed by cockpit crew, and for passenger aircraft the “cabin defects log”
completed by senior cabin crew.
When the aircraft is returned to its home operating base, then normally the
content of these logs will be reviewed by home base maintenance staff and trans-
ferred to the main aircraft logbooks. These are increasingly being maintained
electronically, particularly for high value military and transport aeroplanes although
at the root of such systems is still often physical paper based records. Normally an
airframe will have its own logbook, but then additional logbooks will be required
for each engine, and any variable pitch propeller. Military organisations may
require similar documentation for large portable weapons systems (for example an
air launched missile), whilst research organisations may do so for large airborne
scientific instruments. This allows large transferrable components to easily be
transferred between compatible aircraft whilst maintaining continuity of
documentation.
At the most local level, an expert maintenance engineering staff will be routinely
reviewing these logbook records and ensuring that maintenance schedules are being
met, also both advising operations planners and maintenance resource managers of
coming issues so that aircraft usage and positioning, and maintenance staff and
resource deployment, can be scheduled to facilitate efficient maintenance.
They will also be monitoring maintenance records for any events which are of
sufficient significance that they must be highlighted to either the design authority
(typically the owner of the Type Certificate in the civil environment, or the prime
contractor for supply of the aircraft in the military environment). Such events (for
example a failure on test of a piece of safety equipment, or a requirement to replace
before service life for a safety critical lifed component defined by preventative
maintenance practices) will be subject to a document typically called a Mandatory
Occurrence Report, or MOR. An example of such an event is shown in Fig. 18.2—
this was a bracket within the tailplane attachment structure of a part 23 class
aeroplane which failed at considerably less hours than the designed “safe life”
resulting in mandatory modification across all examples of the type (and a serious,
but thankfully not fatal accident).
This provides an illustration of what will typically happen when a fault is
MORd2; design engineering staff will then normally be required to investigate the
reasons for the fault. In the case shown above, it was identified that an inappropriate
manufacturing technique had been used (retention of a bearing housing by use of
sharp edged peening which had potential to initiate fatigue cracks), and the design
2
The term MOR is common in certified civil aeroplanes, but other terms will exist within
sub-ICAO or military aircraft. The concept however, is quite universal.
18.3 Continued Airworthiness Oversight 359
Fig. 18.2 Failed primary structural component in a part 23 class aeroplane which led to
mandatory reporting
3
“Service bulletin” or “SB” is again a term specific to some regulatory regimes, but like MOR, the
concept is universal.
360 18 Facilitating Continued Airworthiness
underline all airworthiness practice, this is an area where complex ethical decision
making is particularly “interesting”.
Authorities will normally oversee and routinely audit this function, bringing
independent thought to this decision making. Some regulatory regimes may,
depending upon severity of actions, have to sanction any mandatory modification
action. This can create occasionally international dichotomies, where an SB is
universally published, but some authorities will make it mandatory, whilst others
make it optional. This is undesirable and should ideally be avoided, given the
inherently international nature of aviation—but avoidance will be heavily depen-
dent upon negotiation as there is no true international authority in most cases.
Chapter 19
Professional Ethics Within
Airworthiness Practice
I pledge:
• To give the utmost of performance;
• To participate in none but honest enterprise;
• To live and work according to the laws of man and the highest standards
of professional conduct;
• To place service before profit, the honor and standing of the profession
before personal advantage, and the public welfare above all other
considerations.
In humility and with need for Divine Guidance, I make this pledge.
Adopted by US National Society of Professional Engineers, June 1954.
19.1 A Caution
This chapter commences with a caution. The author has been a professional
engineer for approaching three decades, and a Chartered Engineer for well over
two. As such he’s been bound throughout by requirements to apply the highest
possible standards of professionalism, and by several applicable codes of conduct.
However he has no formal qualifications in philosophy or ethics—so please treat
this chapter therefore as personal views presented for consideration, and as such
less authoritative than the rest of the book. Nonetheless, they are built on a lot of
thought and experience, and hopefully will prove of value.
Most likely the first profession to publish and adopt a formal code of conduct is
medicine, in the form of the Hippocratic Oath. A translation to English of the
earliest known form of this is shown below (this version was published in 1923 by
Dr. W. H. S. Jones of St. Catherine’s College Cambridge.)
It is interesting to break this apart into sections and translate these into plain
modern language:
Original Simplified
1. I swear by Apollo Physician, by Asclepius, by This is a solemn promise
Health, by Panacea and by all the gods and
goddesses, making them my witnesses, that I will
carry out, according to my ability and judgment,
this oath and this indenture
2. To hold my teacher in this art equal to my own I owe a debt of gratitude and return to those who
parents; to make him partner in my livelihood; taught me my profession, and to their families
when he is in need of money to share mine with
him; to consider his family as my own brothers,
and to teach them this art
3. and to teach them this art, if they want to learn I will share my professional knowledge with
it, without fee or indenture; to impart precept, oral those entitled to that knowledge
instruction, and all other instruction to my own
sons, the sons of my teacher, and to indentured
pupils who have taken the physician’s oath
(continued)
19.2 The Use, and Tyranny of Professional … 363
(continued)
Original Simplified
4. But to nobody else I hold some knowledge that is confidential, and
must not be shared with people not entitled
5. I will use treatment to help the sick according to I will use my professional skills to do good, as
my ability and judgment best as I am able
6. but never with a view to injury and wrong- Even if asked to, I will not deliberately do any
doing. Neither will I administer a poison to harm
anybody when asked to do so, nor will I suggest
such a course. Similarly I will not give to a woman
a pessary to cause abortion
7. But I will keep pure and holy both my life and I will maintain my professional skills
my art
8. I will not use the knife, not even, verily, on I will only do work that I am competent to carry
sufferers from stone, but I will give place to such out, and defer to specialists when I’m not.
as are craftsmen
Put simply in this manner, one might reasonably apply these moral principles to
any other profession, such as professionalism is important—as it most certainly is in
the practice of aeronautical engineering in any form. The USA’s National Society of
Professional Engineers1 publish their own more complex code, which starts with
the following Fundamental Canons, which state that:
Every item here except for (3) clearly has a direct map to the third century
Hippocratic Oath previously shown. That item is actually the direct opposite of
Hippocrates’ clauses (2) and (3). Information should no longer be held in a closed
shop, but instead should (often must) be openly published—but where it is, must be
rigorously confirmed to be truthful.
1
The latest and expanded form of this code, and much else, will be found at https://www.nspe.org/.
364 19 Professional Ethics Within Airworthiness Practice
This code, which is developed from the general guidelines published by the UK
engineering council presents another general, and consistent view—albeit one that
is, given the organisation, perhaps surprisingly generic, with no specific discussion
of anything related to aerospace.
19.2 The Use, and Tyranny of Professional … 365
This Code of Ethics has been developed to provide us, the Members of the
Society, with a guide to the way that we will engage in the profession of flight
test, and advance the aims and objectives of the Society, thereby achieving
excellence in flight test.
Members acting in accordance with this Code will have the support of the
Society. The manner and extent of the support will be determined on the
merits of each case.
Excellence in flight test is judged not solely on the flight test products
themselves, but also by the way that the flight test products are derived. As a
multi-disciplinary activity, excellence in flight test requires inspired leader-
ship and effective teamwork, combined with application of the utmost in
contemporary professional standards.
As Members of the Society, we accept a personal obligation to our profes-
sion, other Society members and the communities we serve. Both individually
and collectively we are committed to upholding the core values of the Society
and to act in accordance with those values and the principles they represent,
as expressed further hereunder.
As Members of the Society, we accept a personal obligation to our profes-
sion, other Society members and the communities we serve. Both individually
and collectively we are committed to upholding the core values of the Society
and to act in accordance with those values and the principles they represent,
as expressed further hereunder.
Safety Consciousness
To guard against the inherent risks associated with flight test and aeronautics
more generally, we shall always hold concern for safety foremost in our
minds; in particular, we will actively assess the risks of individual activities
and reduce them as far as practicable.
Ethical Behaviour
Individually and collectively we shall always place our responsibility for the
welfare, health and safety of the community before any responsibility to
sectional or private interests, or to other members. We shall avoid real or
perceived conflicts of interest
whenever possible, and disclose them to affected parties when they do exist.
We shall not be apathetic to unethical behaviour by our colleagues and shall
ensure that breaches of this Code are drawn to the attention of the Society
and treated equitably but firmly.
366 19 Professional Ethics Within Airworthiness Practice
Competent Performance
Individually, we shall only offer services, advise on, or undertake flight test
work within our respective areas of competence, and we shall practise our
profession in a careful and diligent manner.
It is possible to distill all of these various Codes of Ethics down to a very simple
set of rules, which might be termed thus:
1. Ensure you are sufficiently skilled to perform a task, and if not, either improve
your skills or pass the task on to somebody else who is.
2. Consider the broader impact of your work and behaviour on everybody and
anybody else, and satisfy yourself that these are acceptable.
3. If communicating about your work, ensure that you do so truthfully and when
you should.
4. Stay within the law.
5. Do things only for the right reasons of duty or employment, not concealed
personal gain.
6. Try to support others in applying these same principles.
Specifically for airworthiness practice, the author would however add two fur-
ther points—not explicitly described in any of the previous codes:-
7. Work with sufficient rigour to ensure good outcomes, even if this requires
exceeding the minimum requirements of laws and regulations.
8. Always ensure independent verification of any decisions with an impact on risk
to life.
These might be referred to as a “moral code”—and indeed this is the author’s
preferred terminology, to differentiate it from the following topic of ethical decision
making. They are however without exception, extremely simplistic and generic—
they do not address the real complex specifics of practice in any general or
sub-field, nor do they really address what is really meant by professionalism, which
must surely be significant here.
Ethical Decision Making
All human beings operate to some form of moral code. We obtain that from our
parents, teachers, religious leaders, friends and colleagues and every individual
probably has some personal aspects to their own codes of behaviour. The moral
codes above, whatever termed, will overlay a societal moral code—as would, for
example specific religious codes of conduct such as the Ten Commandments or
legal requirements on somebody’s broader moral standards.
In reality however, most people will regularly encounter situations that are more
complex than can be addressed a single moral code alone. Consider for example the
following simplistic and hypothetical situation:
19.2 The Use, and Tyranny of Professional … 367
You are the engineer in charge of a sizeable industrial process in a poor country. You have
become aware that a major local outbreak of poor health is due to processes under your
control polluting the local water supply. You have the power and authority to shut this
process down, but in doing so, will inevitably put out of work the breadwinners for several
dozen local families, leaving them without sufficient income for food and education costs.
2
This is modified by the author from an original version in Michael Davis, Ethics and the
University (Routledge, London, Dec 1998)
368 19 Professional Ethics Within Airworthiness Practice
The following incidents occurred whilst the author was working as Chief Technical
Officer to the British Microlight Aircraft Association.
The Aviasud Mistral (Fig. 19.2) is a French aircraft that was imported into the
United Kingdom in the early 1990s. The company importing the type failed after
bringing 12 aircraft into the country (of about 200 in built in France within a
deregulated regime) but before obtaining “type approval” certification, that is
approval for factory built aircraft. This left 12 aircraft unable to be issued with
“permits to fly”, yet having been paid for by individual customers of the aircraft.
BMAA investigated and sought to resolve this, and ultimately they were issued
with “amateur built permits to fly”, in a legal deal between the BMAA, CAA and
owners of the aircraft. This form of permit to fly implies that the aircraft have been
shown to comply with UK safety standard for microlight aircraft, (BCAR Section S,
see Chap. 1) but does not imply that they had been manufactured by a supervised
factory.
A significance of this is that the aircraft were considered fit for private use, but
could not legally be used for commercial flying instruction (the only form of aerial
work apart from certification test flying, and later rental allowed on a microlight
aircraft). One significant point of this approval is that the aircraft did not display
during certification handling qualities at the stall that met the safety requirements, in
that it could drop a wing and show signs of an incipient spin. For this reason the
fitting of an artificial stall warning device (a buzzer which sounded in the cockpit,
triggered by an AoA vane on one wing) was mandated.
A qualified but not experienced pilot took his privately owned Mistral for a
pleasure flight in June 2001. About 40 min into the flight the aircraft was seen by a
number of witnesses flying at low level. Witnesses thought that the aircraft was
flying too slowly and two witnesses, one of whom was an experienced glider pilot
thought that the aircraft was flying out of balance. Some seconds later, the aircraft
was seen to yaw and roll rapidly into a steep dive before disappearing from view.
One witness thought they heard the engine stop just before the aircraft apparently
departed from controlled flight. Several people proceeded in the direction that the
aircraft was last seen and found the aircraft wreckage almost vertically nose-down with
extensive damage in the cockpit area. Both occupants were certified dead at the scene.
Subsequent examination of the wreckage by the UK’s Air Accidents
Investigation Branch3 found that:
• Ingress of an unidentified loose object had caused an engine failure, assumed
just prior to the accident.
• There was supporting evidence of low speed flight prior to the accident.
• The removable canopy of the aircraft was not fitted.
It should be understood that an engine failure in an aeroplane in this class should
never inevitably lead to a fatality—it was the stall/spin event, subsequent to the
engine failure which caused that.
Establishing the facts and issues—stage 2 and 3 of the Davis process.
The author spent some time investigating the background to this incident, and
found that:
• French authorities had a record of a similar fatal accident in 1996, they also had
record of two non-fatal French occurrences of loss of control at the stall. No
record could be found of these having been reported to any UK authority.
• Anecdotal evidence and one formal report to the BMAA, came to light of an
inadvertent spin in an Aviasud Mistral. The formally reported incident had been
filed but had not been investigated further.
• The decision to accept an audible stall warning device in lieu of acceptable
stalling characteristics had been recommended by the test pilot who flew the
aircraft for original certification. This alternative was not explicitly permitted by
the safety standard at the time, and there was no historical record of in-depth
discussion of this decision, leading to a conscious decision that this provided
acceptable equivalent safety. Also when carrying out testing of a further rep-
resentative aeroplane after the accident, this stall warner did not provide ade-
quate and consistent stall warning.
• The aircraft was fitted with a removable canopy. This was not fitted, yet on
investigation throughout the UK certification testing of the type the canopy had
been fitted, so there was no formal knowledge of what changes removing the
canopy made to the handling, and thus there should have been explicit prohi-
bition on flight without the canopy fitted.
3
AAIB reports are available online at www.aaib.gov.uk the registration of the accident aeroplane
was G-MYST.
370 19 Professional Ethics Within Airworthiness Practice
A case could be made that any of these four points were failings contributing to
this accident, and such would certainly fit James Reason’s well known Swiss
Cheese Model.4 However, the BMAA endeavoured to maintain the assumption of
good intent wherever possible in airworthiness work, and so concentrated upon
prevention of future recurrence—something that would be distracted from by
seeking to allocate blame for previous errors.
4
Philosopher Professor James Reason, formerly of the University of Manchester is well known for
his seminal work on the causes and avoidance of error. This author would caution however against
use of his earlier writings such as Human Error (1991), as his later writing is much more
developed in its thinking. The Human Contribution: Unsafe Acts, Accidents and Heroic
Recoveries (Routledge 2008) is a good introduction to his work and its use.
5
European Commission Regulation (EU) No 691/2010 dated 29 July 2010.
6
A useful explanation of the intentions behind Just Culture may be found in Global Aviation
Information Network (GAIN)’s September 2004 document “A roadmap to a just culture:
enhancing the safety environment”. Some readers may note differences between the culture as
described in that document, and the implementation in some regimes.
19.3 Case Studies 371
• Grounding all 11 remaining aircraft—this would be safe, but defeat the BMAA’s
mission to deliver safe yet affordable flying by its members.
• Presenting the owning community with an ultimatum to work together to
identify and resolve the problems that existed. Experience had taught that this
approach seldom reached any useful solution, as private aircraft owners seldom
have the skill or resource to effect such solutions, and therefore BMAA’s pri-
mary mission would again be defeated.
• Applying resource to design solutions to any deficiencies with the aeroplane,
and then presenting the owning community with the options of grounded air-
craft, applying the centrally designed solution, or finding their own solution: this
presented the difficulty that the design and approval organisations for the
modifications would become the same, creating unwanted potential for
untrapped errors (as arguably had occurred previously).
Selection Actions—Stage 6 of the Davis Process
Decisions could not be made in isolation—in this case there was lengthy discussion
between specialists in BMAA and AAIB, whilst ensuring that overseeing specialists
at the Civil Aviation were informed and content with the process. The AAIB report
details the interim decisions that were made:
• The BMAA was to carry out flight tests to determine the acceptability of the
Mistral’s handling characteristics with the canopy removed. Further, the BMAA
should not be flown with the canopy removed until these tests have been carried
out.
• The BMAA was to review the adequacy of the current stall warning system fitted
to Mistral aircraft with particular emphasis on the standard of construction and
adequacy of the aural warning.
The first recommendation was pursued initially, and it was found that a pilot
leaning their arm out of the cockpit would influence flow over the lower wingeron
(the lower wing of the biplane is all flying, providing roll control), rolling the
aircraft and concluded that this was highly likely to have contributed to the loss of
control in the accident flight. Reviewing options, it was concluded that the utility of
the aeroplane was not significantly affected by mandating flight with the canopy
fitted, and that as all aeroplanes had one, no cost would be incurred by mandating
this—so this was done. Ethically this was a good and easily justifiable decision: the
safety enhancement is obvious, the cost in terms of money and (less importantly)
reputation minimal.
The second recommendation led to expenditure of resource, and the potential for
poor decisions because the design and associated flight testing work was being done
by the airworthiness office. This was resolved, by ensuring external authority
oversight of the process, and that in addition to the internal test pilot, a second
external test pilot verified the flight testing of the solution. The actual solution was a
new stall warner, fitted in a different location, and calibrated to each individual
aeroplane.
372 19 Professional Ethics Within Airworthiness Practice
On 2nd September 2006, Nimrod MR2 (Fig. 19.3) XV230 was on a routine
reconnaissance flight over Afghanistan as part of the UK’s contribution to an
ongoing war against Taliban forces there. Shortly after aerial refueling, the aircraft
caught fire and crashed killing all 14 servicemen on board, the Royal Air Force’s
worst single loss since the 1982 Falklands Conflict. The reader unfamiliar with the
events is referred to:
Fig. 19.3 Nimrod MR2 (Crown Copyright, reproduced under Open Government Licence v2.0)
19.3 Case Studies 373
The Nimrod Review: An independent review into the broader issues surrounding
the loss of the RAF Nimrod MR2 Aircraft XV230 in 2006, Charles Haddon-Cave,
28 October 2009.
Many thousands of pages have been written on that accident, including the
seminal report listed above, the Haddon-Cave Report explicitly naming a number
of engineers who were considered to have failed in their professional duties. The
root of the loss of XV230 was the inadequacy of a safety case and supporting
management structure that were supposed to ensure a full and rigorous assessment
of the systems safety of the aircraft in service, and was essentially a collaboration
between the manufacturer BAE Systems, the Ministry of Defence and Royal Air
Force, and partial state owned specialist contractor Qinetiq.
Sir Charles Haddon-Cave in his report particularly concluded that BAE Systems
engineering managers had been dishonest in their construction of the safety case—
concentrating upon producing “an impressive looking set of reports on time” rather
than on rigour and completeness—some managers knew that there were significant
gaps in the reports but failed to draw attention to them, these then went to Qinetiq
for independent verification—who failed to assess these with sufficient rigour—in
some cases signing them off without the right specialists or managers having read
them. This process overall was overseen by an Army General followed by an RAF
Group Captain in the Defence Logistics Organisation whose priorities included
implementing 20% organisation cost savings, and were quite probably not fully
competent in the issues themselves—delegating this to a more junior civilian
engineer who was insufficiently skilled in such tasks, but failed to either raise his
skill level or alert managers to his lack of skill.
Comparing this to the eight principles of ethical behavior described earlier, we
can see failings at every level (Table 19.1).
5. Test op ons
6. Choose course of ac on
7. Review
3. If communicating about your work, ensure that you do so truthfully and when
you should.
4. Stay within the law.
5. Do things only for the right reasons of duty or employment, not concealed
personal gain.
6. Try to support others in applying these same principles.
7. Work with sufficient rigour to ensure good outcomes, even if this requires
exceeding the requirements of laws and regulations.
8. Always ensure independent verification of any decisions with an impact on risk
to life.
Many real ethical decisions will prove complex, and require research and a
decision making process. A good decision making process is that proposed in 1998
by Davis, and shown in Fig. 19.4.
20.1 Introduction
Any successful programme to take an aviation product from concept to service must
have a significant component of programme management incorporated into it.
Programme management of course has become a specialist discipline in its own
right since the 1960s and is likely to remain so. This chapter does not aim to
provide any detailed advice on programme management as a topic, but to show
airworthiness practioners how they can usefully integrate with broader programme
management, and conversely indicate to programme managers how they might
constructively engage with airworthiness specialists, with a view to achieving the
best possible outcomes to the project.
The imperative for airworthiness integration into programme management
should be clearcut; a product cannot achieve successful sales and introduction to
service if it is not certified. At the same time, achieving only minimum certification
requirements is likely to mean a poor product on the marketplace—and therefore
early engagement at the design stage to ensure significant exceedence of certifi-
cation minima should not only ensure a smooth and affordable certification process,
but a product that has necessary saleable products in the marketplace, a better
product for the user, or a fighting edge for the warfighter.
The analytical phase commences with engagement with design teams, where the
airworthiness team will support the design process in ensuring designers are
working to the current airworthiness requirements, and aware of any latest points of
interpretation. For civil programmes, and to a large extent military programmes at
the company side, airworthiness teams will also be commencing preparation of the
compliance checklist reports, that detail the extent to which the product is predicted
to exceed standards minima. This of course is a set of predictions based upon initial
analysis and design data and cannot be exact, but it will provide several vital tools
towards eventual certification:
– It provides a framework for the eventual final form airworthiness reports.
– It will identify clearly the tests resources that will be required later on.
– This should identify before prototype manufacture and testing any major
potential noncompliances that slipped through the design stage and require
rectification.
– It should identify the cardinal points of smallest margins compared to certifi-
cation minina that should be the critical end points of test plans (and, particu-
larly for flight test programmes, the start points being where the safety margins
appear to be the greatest).
– All compliance that can be demonstrated by analysis, should have this done at
this stage where (as is usually the case) analysis is the cheapest way to do it.
380 20 Running a Certification Programme
It is likely that the next stage (with the obvious reality that all major “report writing”
phases overlap and interlock with each other and this is not a neat series of suc-
cessive tasks) will be the major ground test stage phase. This is the intermediate
stage between analysis, and flight test—being typically also the intermediate in
terms of cost. Some aspects of this (particularly anything related to product life,
such as full scale fatigue tests) will extend through subsequent flight testing and into
a product’s service life. The nature of the ground testing will depend entirely upon
the nature of the certification task, but may include (but certainly need not be
limited to):
– Ultimate or destructive load tests
– Initial fatigue tests prior to flight test
– Coupon tests of materials with variable properties (this is likely to need to be
commenced well before the end of the analytical phase)
– Evacuation tests (likely from mock-ups rather than any potentially flightworthy
vehicles)
– Ergonomic evaluations
– Engine ground or test rig running
– Opening iron bird or flat-sat testing (again, highly likely to continue through
service for aeroplanes, less likely for spacecraft)
– Electromagnetic Compatibility (EMC) testing
– Initial taxi tests and pre flight ground testing of all new or modified aircraft
systems
– Pre-flight-test inspections.
The objective here is to reach a position where as much as can reasonably have
been achieved without expensive and risky flight testing has been achieved. It is
likely that prior to the commencement of flight testing a further iteration and review
of compliance reports will be carried out. The objective of that review is to satisfy
the programme that as much compliance as can reasonably have been demonstrated
without flight testing, has been demonstrated—and also to identify any incomplete
analysis or ground testing that can be continued reasonably in parallel with the
flight test phase.
Risk assessments, competences and resource availability reviews are clearly
essential throughout this phase. It is generally the case however that ground testing
risk assessments, and consequent processes, are used to ensure safe delivery of
activities that will in any case be carried out—note that this is differs from best
practice in the next phase of flight testing.
20.6 The Development Flight Test Phase 381
Flight testing is the final major phase of certification, and likely to only be com-
menced once all major uncertainties that can be addressed without flight testing
have been resolved. In general the flight test programme will be progressed in the
following order, so far as resources and expediency permit
(i) Demonstration of basic safe function
(ii) Establishment of safe operating limits
(iii) Demonstration of safety critical compliance points
(iv) Obtaining of non-safety critical performance data.
At the end of this phase, which is likely to be complex and require various iterations
to design configuration as issues are determined through the programme, the
demonstration that the product is satisfactory and fit for approval and production
should have been completed. Almost inevitably analysis and ground testing will
have continued in parallel with flight testing, but programme management should
aim to achieve completion of both of these by the time flight test reporting is also
completed.
Not all projects of course will require flight testing, and some of those will not
require the four separate stages indicated above. For example a new cockpit engine
instrument may be subject to analysis, ground test and inspection, but then flight
testing can be entirely limited to a simple functional test, or in-flight entertainment
systems approval may be entirely done by analysis and ground testing. Equally a
new prototype aircraft, or major airframe changes such as wing shape or powerplant
is likely to require the full four phases of flight test development described above.
Of necessity, developmental flight test reverses ground test’s safety priorities. In
virtually all cases, absolute primacy must go to safety of the flight vehicle and its
occupants, and then ways found to obtain the data required.1
1
The FAA’s airworthiness circulars AC90-89 for very small aeroplanes, AC23-8 for part 23
aeroplanes, and AC25-7 for part 25 aeroplanes are the best readily available guidance on flight test
programme execution in the civil domain. Multiple NATO AGARD guides provide similar facility
in the military environment. Of course as with other engineering specialisms, access to these is no
substitute for also having key staff with the right specialist technical educations. The classical, but
not the only route to this is the 8 “Test Pilot Schools” across the world and accredited by the
Society of Experimental Test Pilots. Inevitably many other experts were also trained “on the job”
in company flight test departments, and these should not be discounted.
382 20 Running a Certification Programme
between. In practice this almost never occurs that simply—there will normally be
feedback from the authority and requirements for additional analysis and testing.
Clearly it should be the objective that this prolonging corrective activity is min-
imised: this can be done by active involvement throughout with the approving
authority, and by a calculated degree of exceedence of minimum standards so that
there is little potential to create disputes over whether those minimum standards
have genuinely been exceeded or not.
However, approval should be obtained, in whatever form is appropriate to the
regulatory regime. Inevitably this is always then a cause for significant celebration
on the programme, triggering of progress payments or customer orders, and so-on.
However, it is not yet a point at which the product can enter service. This then
requires the operational test and evaluation (OT&E) phase as it is termed in military
environments.
The final pre-service stage of OT&E is where expert users have access to the
product and can develop training material for both air and ground crew, integrate
the product into operating strategies (which in the modern world are likely to be
highly complex, IT intensive and integrated, whether in a military or civil envi-
ronment), and train initially core senior operators, then all operators as product is
manufactured and introduced into service. Until the 1990s, this may have actually
been what happened.
Now, OT&E will have been deeply integrated into the later stages of the flight
test programme. In the military environment high ability front line users and
maintenance staff will have been closely engaged with the military DT&E function
(which will have included both company and OTC specialists), so that by the time
approval is obtained they have considerable expertise in its use and have prepared
the bulk of their essential material for introduction to service, as well as having built
up a core of product familiar military personnel. Similarly in a civil environment for
major launch products, customer staff: both flight and ground will have been
involved with the latter stages of the approval process—including probably flying
the aeroplane on endurance tests, which will have also gone into eventual desti-
nation airports so that experience in these locations has been obtained. This inte-
gration requires complex management, but is essential given the cost and timescales
of a major programme. To do otherwise would introduce large cost and time
increases that are unacceptable.
This transition from serial to parallel test and evaluation processes has been
largely welcome and uncontroversial in the commercial aviation world and usually
centres still on manufacturers’ sites. In the military world it has been controversial
and tended to be driven by senior front line military commanders, who are not
20.8 Operational Test and Evaluation, and User Integration 383
historically sympathetic to the costs and timescales of DT&E. The need is now
generally accepted, although the question of where this combined T&E activity
should be centred does remain controversial as both operational and developmental
communities tend to believe that they should hold the lead. The author tends to
believe that given the larger bulk of technical expertise that must sit behind the test
enterprise in development activity, this should have primacy—but this view is not
universally held, particularly by senior military customers in the UK and USA
where, for example, Typhoon DT&E activity has been co-located away from
Boscombe Down and Warton where the major technical expertise lies, towards
front line RAF stations.
Once a man has spent his time in messing about with aeroplanes he can never
forget their heartaches and their joys, nor is he likely to find another occu-
pation that will satisfy him so well, even writing.
Nevil Shute, Slide Rule, 19542
2
Nevil Shute Norway 1899–1960 was an aeronautical engineer present at the gestation of many
major aviation products—most famously the R100 airship where he worked under Chief Designer
Barnes Wallis. In later life he became one of the world’s most successful novelists—all of his work
still bears reading now and much is still in print, although often—whilst socially crusading—
nonetheless rooted in very dated social constructs. His autobiography “Slide Rule” published first
by Heinemann in 1954 is perhaps historically unreliable in places but highly readable. For a more
in-depth and accurate understanding of his life, “Parallel Motion” by John Anderson was pub-
lished by Paper Tiger in 2011. The Nevil Shute Norway Foundation has also been active for some
years and is a valuable source of wisdom at www.nevilshute.org.
Appendix A
International Standard Atmosphere
Tables
Note: the values here are typical, and suitable for learning exercises or
conceptual design work: but they are not necessarily accurate for any par-
ticular application or material source and have also been simplified. Always
refer to current and traceable data sheets for actual airworthiness tasks,
although these should be sufficient for quick-estimate or educational pur-
poses. For woods and composites, remember always that properties depend
upon grain/fibre direction, and must always be confirmed in any case by batch
coupon testing.
Material Density Elastic (Young’s) Ultimate tensile strength Yield tensile strength Shear strength
(kg m−3 103) modulus (N m−2 109) (N m−2 106) (N m−2 106) (N m−2 106)
Aluminium alloys
2014-T0 2.80 73.1 186 96.5 28.0
2014-T4 2.80 73.1 427 290 260
2014-T6 2.80 72.4 483 414 290
2024-T0 2.78 73.1 179 75.8 124
2024-T3 2.78 73.1 448 310 276
2024-T6 2.78 72.4 427 345 283
6061-T0 2.70 68.9 117 48.3 75.8
6061-T4 2.70 68.9 228 131 152
6061-T6 2.70 68.9 290 255 186
6082-T6 2.70 290 250
Steels
AISI 4130 normalised 7.85 205 670 435
AISI 4130 tempered 7.85 205 807 689
Mild steel 7.87 205 350 220
T300 annealed stainless 8.00 195 585 240
T300 cold drawn 100% 8.00 195 1550 1450
hardened stainless
Titanium alloys
Ti.CP-1 4.51 105 241 172
Ti.CP-2 4.51 105 345 276
Ti.CP-3 4.51 105 448 379
Ti.CP-4 4.51 105 552 483
(continued)
Appendix B: Typical Properties of Common Aerospace Materials
Table B.1 (continued)
Material Density Elastic (Young’s) Ultimate tensile strength Yield tensile strength Shear strength
(kg m−3 103) modulus (N m−2 109) (N m−2 106) (N m−2 106) (N m−2 106)
Plastics and plastic based composites
Transparent acrylic 1.19 21.0 69.0 40.0
Nylon 66 (not reinforced) 0.70 2.50 60.0 60.0 57.0
CFRP, low strength 1.40 14.0 110
CFRP, high strength 1.70 220 890 55.0
CFRP, very high strength 1.60 170 2100 120
GRP 0.99 2.76 55.0
13 mm Fibrelam 3.00 Fibrelam’s properties are not expressed in these terms
Woods
Red spruce 0.38 8.14 1.52 5.31
White spruce 0.37 6.76 1.38 4.62
Douglas fir 0.40 8.41 2.41 6.07
Balsa (kiln dried) 0.16 4.10 17.5 3.00
Plywood 0.40 8.00 31.0 6.20
Appendix B: Typical Properties of Common Aerospace Materials
389
390 Appendix B: Typical Properties of Common Aerospace Materials
All standards define ZFW as the empty weight of the aircraft, with unusable fuel,
working fluid levels (everything from coolant, through lubricants to toilet fluid),
and minimum essential equipment. Aircraft documentation must clearly define the
particular state and levels, and weight and balance (W&CG) reports must always
give the modification state of the aircraft.
There are manufacturers who start to argue about what is the minimum equip-
ment list so as to minimize the apparent ZFW; arguably if an aircraft reaches the
state that this is important, it is probably time to consider some modification or
lightening. If it starts at such a state, (particularly the case for some higher per-
formance light or microlight aeroplanes) then matters are altogether more serious,
not lease because historically all aeroplanes tend to become heavier with age.
W/S, where used, is the MTOW (in kgf) divided by the total wing area. Total wing
area includes the mainplane(s), and canard, but does not include a tailplane. It is
common, on aircraft where a definition is based upon wing area is important, to
include canard, the area of aerofoil section struts and exceptionally undercarriage
legs.
Max no. Engine type Wing Max stall speed Max. Max ZFW Max. MTOW
seats loading no.
engines
EASA certification specifications/joint airworthiness requirements
CS.25 and JAR-25 No limit No limit No limit Formula No No limit No limit
(“Part 25”) limit
CS.23 and JAR-23 As Piston or No limit 61 knots Vso (single 2 As FAR-23 As FAR-23
(1, 3, 25, 49, 67) FAR-23 turboprop engine and sub-2,730 kg
(“Part 23”) twins unable to climb SEI)
CS.22 and JAR-22 2 Piston 3 kg/m2 80 km/h (Vso—no ballast) 1 MTOW—180 kg (s-seat) or 850 kg
(1, 3, 25, 29, 49) (W/span2) 90 km/h (Vso—full ballast) 110 kg (1-seat)—30 min fuel at
(“Part 22”) max cont power
CS.VLA and 2 piston no limit 45 knots (Vso) 1 MTOW-86 kg per seat—1 h 750 kg
JAR-VLA (1, 3, 25, fuel at max cont. powera
29, 49) (“Part VLA”)
Federal aviation requirements
FAR-25 No limit No limitb No limit No limit No No limit No limitc
limit
FAR-23 (1, 3, 25, 49) 19+ No limit No limit 61 knots (Vs1) (single No MTOW—77 kg (170 lb) per seat 19,000 lbf/8618 kg
(commuter category) pilot(s) engine or under 6,000 lbf limit —30 min fuel at max cont.
seat(s) only) power
FAR-23 (1, 3, 25, 49) 19+ No limit No limit 61 knots (VS1)e (single No MTOW—86 kg (190 lb) per seat 12,500 lb/5,670 kg
(normal, utility and pilot(s) engine or under 6000 lbf limit (77 kg (170 lb) in normal
acrobaticd categories) seat(s) only) category)—30 min fuel at max
cont. power
FAR-103f 1 No limit but No limit 24 knots Vso No 254 lbf powered + weight of No limit
Vh 55 knot limit floats and safety devices or
CAS 155 lbf unpowered
(continued)
Appendix C: The Main Civil Airworthiness Standards
Table C.1 (continued)
Max no. Engine type Wing Max stall speed Max. Max ZFW Max. MTOW
seats loading no.
engines
British civil airworthiness requirements
Section Dg No limit No limit No limit No limit No No limit No limit
limit
Section K (1–1, 5–1) No limit No limit No limit 60 knots Vso (single 2 No limith 5,700 kg
engine and miti-engine
twins unable to climb at 1
% gradient SEI)
Section S (2, 25) 2/1 (see No limit 25 kg/m2 (or) 35 knots (Vso) No MTOW-86 kg per seat—1 h fuel 450 kg/300 kg
MTOW) (W/area)i limit at max cont. powerj (2/1 seat)
a
CS.VLA also states that MTOW must not be exceeded with an 86 kg pilot and full fuel, although in practice this is rarely an issue
b
FAR-25 applies special engine power and capacity limits for DC-3 and L-18 derivatives
c
FAR-25 applies special MTOW limits for DC-3 and L-18 derivates using FAR-25 for STC applications
d
The American term “acrobatic”, and British/European term “aerobatic” are interchangeable. The reasons for the linguistic divergence is unclear, although some British
Appendix C: The Main Civil Airworthiness Standards
The inclusion of a lifting body area is not usual, and if this is considered
necessary, the engineer should reach an independent agreement with their
approving authority.
CS-22 divides the MTOW by the square of the wingspan—however, take care
not to mix this up with classically defined wing loading—they have the same units
and dimensions but very different values. CS.22 is simply ensuring that the standard
is only applied to aeroplanes with long slender wings.
Vso is the stalling speed (or minimum flying speed) in the landing configuration.
That is with gear (if retractable) down, flaps and/or slats in the setting for landing,
and normally with the engine(s) at flight idle. It is the most critical speed from the
perspective of determining whether a standard can be applied to an aircraft,
although multiple stall speeds will be required for determination of performance
and demonstration of more detailed issues of compliance and flying limitations.
Accurately measuring stalling speed is a two part process. Firstly the pitot-static
system must be accurately calibrated. Ideally this should be done at a weight a little,
but not significantly below MTOW.
The second part of the process is that stalling tests must be carried out as near as
possible to MTOW at a deceleration rate not exceeding 1 knot/s. The mean value
from a series (usually at least 6) tests must be cross referenced to the ASI calibration
graph to determine Vso as a calibrated airspeed. For more detail on stalling tests,
see the appropriate chapter of this book.
Both FAR-23 and CS/JAR-23 define the SEI (Single Engine Operative) climb as a
deciding factor for Vso. In both cases the definition used (Para 67 of each standard)
is that the aircraft must be able to maintain a 1.5% climb gradient, with the most
critical engine inoperative and at its lowest drag state, at 5000 ft standard pressure
altitude.
Ideally the conscientious airworthiness engineer will attempt to ensure that this
margin is comfortably exceeded—nonetheless this has often not been achieved for
many lighter twin engined aeroplanes. A common colloquialism for light twin
engined aeroplanes is that if one engine fails, the second is “there to get the aircraft
to the scene of the accident”.
Appendix D
Conversion Factors
Dihedral Effect Alternative term for lateral stability (does not necessarily
rely upon dihedral)
Directional Stability Yaw due to sideslip (mathematically, the partial derivative
of yawing moment with respect to sideslip angle)
DLR Deutsches Zentrum für Luft- und Raumfahrt e.V: the
German national aerospace centre
Dutch Roll Combined rolling and yawing oscillation
EAS Equivalent airspeed
EASA European Aviation Safety Agency
ECU Engine control unit
EPR Engine pressure ratio
ETOPS Extended range twin engine (airline) OPerationS (also less
charitably referred to as “Engines Turn Or Passengers
Swim”.)
ETPS Empire Test Pilots School (based at Boscombe Down,
Wiltshire, UK)
Extremely remote A probability between 10−7 and 10−9 per flying hour
FAA (US) Federal Aviation Administration
FAR (US) Federal Airworthiness Requirement
fN2 Proportion of the air which is gaseous nitrogen (normally a
nominal value of 0.79)
fO2 Proportion of the air which is gaseous oxygen (normally
0.21)
FOD Foreign object debris (often leading to foreign object
damage)
FTO Flight test observer (in aircraft)
g Acceleration due to gravity (9.81 N/kg, or m/s2 or 32.2 ft/s/s)
Note: the value of g will in fact vary around the planet,
however this standard value is universally used for
airworthiness work. The only occasion when a different
value may be required, is the practice of instrument
calibration (or related measurement and navigation tasks)
GPS Global positioning system (satellite navigation)
GS Ground speed
H Altitude
h Gust gradient distance, that is the distance parallel to flight
path for a gust to reach its peak velocity. It is usually
assumed that 2 h = 25C
HQR Handling qualities rating (alternative term for CHR)
HSI Horizontal situation indicator (gyroscopic compass)
IAS Indicated air speed
ICAO International Civil Aviation Organisation (International
treaty based organisation setting international standards for
overflight)
400 Appendix E: Common Acronyms, Terms and Abbreviations …
PS Static pressure
PSS Steady state roll rate
PZ.max Maximum vertical force on aircraft during landing
PZ.max(a/c) Aircraft load reaction (vertically, during landing)
PZ.max(u/c) Undercarriage load reaction (vertically, during landing)
Q or q Dynamic pressure, defined by q ¼ 1=2qV 2
QFE Altimeter setting giving an indication of zero height on the
ground at a destination aerodrome. Given in hPa
(heptoPascals) or mb (millibars) the units being identical,
ISA sea-level value being 1013.25
R2 Coefficient of determination, definingP theP quality
P of a line
n xy x y
fit (defined by R ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P 2 P 2 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P 2 P 2 ),
nð x Þð x Þ nð y Þð y Þ
has value R = 1 for perfect line fit, R = 0 for totally
2 2
random distribution
RAS Rectified air speed, alternative term to CAS mainly used in
the USA
Reasonably probable A probability between 10−3 and 10−5 per flying hour
Remote A probability between 10−5 and 10−7 per flying hour
RF (Structural) reserve factor
Roll Mode Mode concerning roll rate
RPM Revolutions per minute (shorthand for engine speed)
S Reference wing area (including a canard, if fitted, but not
tailplane)
SAS Stability augmentation system
Science Speed A speed value (usually declared in IAS by aircrew but TAS
by scientists) at which instruments are likely to be
optimised on a research aeroplane. This is most likely to
be close to best endurance speed
sHp Standard pressure altitude (altimeter reading with
1013.25 hPa set on subscale)
SHSS Steady heading sideslip
SPEC Static pressure error corrections
Spiral Mode A measure of the tendency of an aircraft to return to, or
bank away, from wings level after a lateral disturbance
ST Tailplane area
Stopway Prepared overrun capable of holding the weight of an
aircraft, but not normally available for take-off
t2 or t½ Time to double or halve a bank angle (when examining
spiral mode)
TAS True air speed
TCDS Type certificate data sheet
TO Configuration for take-off, normally mid flaps gear down
TP Test pilot
Appendix E: Common Acronyms, Terms and Abbreviations … 403
“AN”—a very old abbreviation for “Army and Navy specification” is a standard
still in use worldwide for fasteners and some other hardware commonly used in
aircraft manufacture. The components are always specified in imperial units.
At the root of the system is the AN bolt. An AN bolt is specified according to
diameter and plain shank length, and will always have a UNF (Unified Fine (im-
perial)) screw thread. Bolts are specified by “ANn-nn” where “n” will be the
nominal plain shank (or grip) diameter in sixteenths of an inch, and “nn” will
correspond to a plain shank length as shown in the table below (other larger sizes
also exist, these are only the most common aircraft type bolts). They are normally
drilled through the screw thread to carry a securing pin or wire locking, and if they
do not a “C” is appended after the screw thread designation. So, a 1/4ʺ UNF bolt
with a 11/16ʺ plain shank length and a hole drilled for a clevis ring will be desig-
nated AN4-14, but if the hole is not required it will be designated AN4C-14
(Table F.1).
A variety of corresponding nuts will always be available for each screw thread,
including plain, nylon self-locking, drilled and wing. Both nuts and bolts are
usually available in either cadmium plated steel, or stainless steel.
Loads in an AN bolted joint should preferably be in shear, across the bolt plain
shank—it’s not regarded as good practice to apply loads the nut, which is only there
to ensure that the bolt stays in place (and then the locking pin, ring, or wire, is there
to ensure that the nut stays in place).
Appendix G
All About Units
Darrol Stinton
Worse still, there is a “market metric” in common use worldwide,1 which is not
technically part of either system, so units such as decaNewtons, centimetres and
kilogrammes-weight proliferate and may often be found in aeronautical use also.
This is just something else which we must understand and live with.
This book therefore has used both SI (as is mostly used in Europe and
Australasia) and Imperial (as is mostly used in North America, and prior to about
1970 was most common in the UK and its former empire), and if appropriate
“market metric” depending up the nature of each example, and the most commonly
used units. If this offends any reader who feels that a particular system should be
used exclusively, The author invites that reader to look more broadly at the aero-
space industry within which they work—and for which this book was written.
Between the three systems only one unit remains inviolate—the second of time,
thankfully provides us with some degree of linkage.
What is referred to here as the Imperial system, may also correctly be referred to as
the fps (foot-pound-second) system, or (for mass units) as avoirdupois; some
non-British textbooks refer to it as “British units”.
The basic units of the Imperial system are the pound of mass (lb) and the foot
length (ft), plus the second of time (s).
There are two alternative units for force which may be routinely used. The pound
force (lbf) is the gravitational force experienced by 1 lb of mass at 1 g (mean
acceleration due to gravity at sea level). The poundal (pdl) however is the force
required to accelerate a mass of 1 lb at 1 ft/s/s. It is important to take care, when
values of mass or weight/force are reported in pounds, that it is clear which is
intended, before making calculations upon that value. (Even more care is needed
when using texts and formulae written for lb/lbf if switching to SI units.)
An alternative unit of mass is also in use, which is the slug (also sometimes
called the geepound). 1 slug is the mass which when acted upon by a force of 1 lbf,
accelerates at 1 ft/s/s. 1 slug = 32.17 lb.
1
It is believed that the term “Market Metric” was originally coined by the late Dr. Darrol Stinton, a
well known aviation author and test pilot, so this author claims no originality.
Appendix G: All About Units 409
2
For historical interest: “A mile ſhall contain eight Furlongs, every Furlong forty Poles, and
every Pole ſixteen Foot and a half”.
410 Appendix G: All About Units
One should take care to be familiar with the differences between British and
American Imperial volume measures which can be sufficiently different as to cause
problems. For example—an aeroplane requiring 1000 Imperial gallons of fuel for its
journey, but which is filled with 1000 US gallons may find itself short of fuel before
the destination, US Gallons being 17% smaller. Anybody dealing with Imperial
volume measures should always take great care. [When dealing with fuel or oil, also
note than some organisations and aircraft will use mass or weight, rather than
volume units—there is no universal standard, and fuel density particularly can vary
with storage temperature.] (Tables G.3 and G.4).
It can be seen that the most important difference between the UK and US
Imperial volume systems is that whilst in the UK there are 20 fluid ounces to a pint,
in the US there are 16. The two units of a fluid ounce are very similar, but not quite
identical either, viz:
Table G.3 US-imperial units of fluid volume (sometimes incorrectly called “British Units”)
1 US bbl (US 1 US gal (US 1 US qt (US 1 US pt (US 1 US fl oz (US fluid
barrel) gallon) quart) pint) ounce)
1 31.5 126 252 4032
1 4 8 128
1 2 32
1 16
1
Appendix G: All About Units 411
Table G.4 British-imperial units of fluid volume (sometimes incorrectly called “British Units”)
1 gal (UK gallon) 1 qt (UK quart) 1 pt (UK pint) 1 fl oz (UK fluid ounce)
1 4 8 160
1 2 40
1 20
1
1 US fl oz ¼ 1:0408 UK fl oz:
It will also be noted that there is no UK barrel shown. Whilst the barrel exists in
Britain, it has several versions and is almost exclusively used for foodstuffs such as
wine, beer or salt—therefore if an aeronautical quantity is expressed in barrels, it
should almost certainly refer to US barrels.
Pressure in Imperial units is most usually expressed as “psi” (pounds per square
inch: lbf/in.2), but more properly should for calculations be expressed as lbf/ft2. It is
common in some countries however (particularly the USA) to express atmospheric
pressure in in.Hg (inches of mercury)—where the standard sea-level value of
atmospheric pressure is 29.29 in.Hg. Inches of mercury are also commonly used in
declaring gas partial pressures, particularly in relation to human physiology.
The standard Imperial scale for temperature is the Rankine Scale, which is an
absolute scale—setting absolute zero at 0°R, the freezing point of water at 491.67°
R, and its boiling point at 671.67°R.
The Rankine scale compares linearly to the more widely familiar Fahrenheit
scale by R = °F-459.67.
The basic units of the SI system are the kilogramme (kg) of mass and the metre
(m) of length, plus of-course the second (s) of time.
Force is defined related to mass and time: the Newton (N) is defined that 1 N is
the force required to accelerate a mass of 1 kg at 1 m/s/s.
A further important point that should be understood in the use of SI, is that
strictly all units scale in thousands. So:
The daN is the only regularly used unit within the SI system which is not defined
in thousands. daN are commonly used for measuring control forces, and so are also
listed here.
The kilogramme is not formally a unit of force, but please note its alternative
usage shown under “market metric” below. For weight, in strict use of SI, the
Newton is also the unit that should be applied.
The basic unit of volume used in SI is the cubic metre (m3), however:
414 Appendix G: All About Units
1; 000 L ðlitersÞ ¼ 1 m3
1; 000 mL or ml ðmillilitersÞ ¼ 1 L
Degrees Celcius/Centigrade (both written as °C) is a scale set such that distilled
water at sea level freezes at 0 °C and boils at 100 °C.
Mean acceleration due to gravity at sea level, is defined as 9.80665 m/s/s, which
may also be written as 9.80665 N/kg. For most engineering work however, 9.81 is
sufficiently accurate, whilst for some scientific work, local variation may need to be
allowed for.
Note: Capitalisation
It is the tradition and practice in SI units that they are always written with lower
case letters (kilogramme, second, millimetre), except when the unit was named
after somebody, in which case the unit name is capitalised, e.g. Newton, Pascal,
Joule. The two main exceptions to this are large-value prefixes such as Mega
(106) or Giga (109), and the use of a capital “L” which is to abbreviate litre,
to avoid inadvertent confusion with the numeral “1” (one).
Market metric is a term which does not describe a unified system of scientific units,
but instead a series of derived units which may commonly be used and with which
the engineer should be familiar since, although probably they shouldn’t, they can
regularly creep into engineering work.
The most common of these is the centimetre, where 100 cm=1 m—this is
commonplace in many domestic applications such as DIY, clothes sizes, etc. and
the author has seen it used for weight and balance calculations in some European
light aeroplanes.
416 Appendix G: All About Units
Also regularly used, and which one should take care with, is the
kilogramme-force, usually written kgf. This is often used, particularly for weight
and balance purposes, presumably because so many scales are calibrated in kgf.
Whilst it is safe practice to use kgf for some purposes (in fact, there is one British
organisation which regularly, and safely, uses kgf.in. for aircraft moment arms),
engineers should normally be careful to ensure that it is properly converted to
Newtons (by multiplying by g) before performing most calculations which are
based upon force.
In volume, whilst the litre and millilitre are most common, the centilitre
(100 cL= 1 L) and decilitre (10 dL = 1 L) are also sometimes used, although rarely
for aeronautical or engineering purposes.
Other modified SI units may be encountered from time to time—the importance
to the engineer is to be able to recognise them as such, and convert them to a
consistent (SI or fps) unitary system where necessary to do so before performing
calculations.
G.5 c.g.s.
c.g.s. is an alternative to the SI system which was for many years in regular use in
scientific papers—being introduced around 1873, and rendered technically obsolete
by the international standardisation of SI in 1960. c.g.s. units are still however
regularly used in much electrical and electronic work, via subset systems called esu,
emu and the Gauss system.
Since this book does not consider electrical and electronic systems in great
depth, c.g.s. units will not be discussed further here.
It is possible to combine unitary systems, but this should be done only with great
care. The author cautions against this practice, except where the units being com-
bined have different dimensions—so for example calculating moments by multi-
plying Newtons by inches would be a reasonably safe practice, whilst any
calculation including both pounds and kilogrammes of mass is likely to result in
error.
If in doubt—convert all units to a single scientific system: either SI (kg-m-s) or
Imperial (f-lb-s/fps) before performing any calculations. (Or at the least, this is a
very good practice for check calculations.)
Appendix G: All About Units 417
There are worldwide conventions for what units are used for what, albeit with local
variations. The most common units in use are:
Airspeed knots (air transport and military flying), mph (some light and
vintage aeroplanes), kph (some light and microlight aeroplanes).
Altitude feet (most civil and military aviation), metres (some light
aviation, some operations in China and Russia), flight levels (air
transport above a locally defined transition altitude)—which are
standard pressure altitudes defined in hundreds of feet: so
FL100 = 10,000 ft pressure altitude at 1013.25 hPa.
CG position mm or inches, usually aft of an arbitrarily defined datum. Some
aeroplanes may use %smc (standard mean chord) with a typical
range of *15–30%.
Cloudbase Usually feet, occasionally metres.
Distance (over the surface) nautical miles, (within an airframe), m, mm or
inches.
Engine power (Piston engines) kW or hp: horsepower, where 1 hp = 0.764 kW
[but take care, as numerous variants upon the horsepower exist,
and there is no universal standard for the test conditions at which
engine power is determined], (turboprop engines) kW, (turbojet
and turbofan engines), kN or lb.
Fuel flow Normally gauged fuel quantity per hour.
Oil qty (European aeroplanes and engines) litres, (North American
aircraft engines) quarts. [1 quart = 2 pints, which is very close
to a litre.]
Fuel qty litres, kg, lb, US gal or Imperial gal, depending upon aeroplane.
[Take care that conversions from volume to mass unit will
depend upon storage temperature.]
Mass kg or lb.
Rate of climb (Aeroplanes) ft/min, (gliders) m/s—occasionally knots may also
be used.
Runway length (including take-off and landing distances)—in North America
feet, the rest of the world, metres. Older aeroplane operating data
may often however use feet: requiring regular conversions by
crew.
Temperature Celcius
Torque Nm or ft.lb (commonly used in tooling, but also as part of power
expression for turboprop and turboshaft engines).
Time seconds, except for navigation where hours, or occasionally
minutes may be used.
Velocity (in a wind tunnel): m/s or fps.
418 Appendix G: All About Units
Visibility metres, except in North America, where statute miles are used.
Wind (English speaking world and western Europe) knots, (elsewhere)
probably metres/second.
It is likely that many of these may require conversion to SI or fps for the
purposes of engineering calculations. To not do so is to invite errors, but equally if
there is a standard unit used operationally, that should be changed only with great
care and standard engineering units may be inappropriate and confusing if included
in operating data.
Appendix H
Useful Further (Non Web) Sources
of Reference
These are the most thumbed reference books on the author’s desk, and may be
useful in particular as expansion of the information in this appendix.
(continued)
What Why Reference
The Aeroshell Book Specifications on all of the fluids and Published by Shell Aviation Limited,
lubricants routinely used in aviation London. (The latest edition can also
be downloaded from their website)
The Checklist A brilliant treatise on the use of ISBN: 978-1846683145
Manifesto checklists and their benefits in results
Atul Gawande critical working environments, based
upon aviation principles
The properties of Data on all of the common (and most Published by The Aluminium
aluminium and its of the uncommon) aluminium alloys, Federation
alloys with comparisons between
specifications
Index