.

Landing gear design in

an automated design
environment

Master of Science Thesis

N.C. Heerens

Landing gear design in an automated

design environment
Master of Science Thesis
by

N.C. Heerens

in partial fulfillment of the requirements for the degree of

Master of Science
in Aerospace Engineering

at the Delft University of Technology,

to be defended publicly on Friday March 7, 2014 at 13:00.

An electronic version of this thesis is available at http://repository.tudelft.nl/.

Faculty of Aerospace Engineering - Delft University of Technology

D ELFT U NIVERSITY O F T ECHNOLOGY

D EPARTMENT O F
F LIGHT P ERFORMANCE AND P ROPULSION

The undersigned hereby certify that they have read and recommend to the Faculty of Aerospace Engineering for acceptance a thesis entitled Landing gear design in an automated design environment by N.C.
Heerens in partial fulfillment of the requirements for the degree of Master of Science Aerospace Engineering.

Dated: February 24, 2014

Head of department:
Prof.dr.ir. L.L.M. Veldhuis

Supervisor:
Dr.ir. M. Voskuijl

Supervisor:
Dr.ir. R. Vos

Reader:
Dr.ir. R. de Breuker

Abstract
The design of the landing gear is one of the prime aspects of aircraft design. Literature describes the design
process thoroughly, however the integration of these design methods within an automated design framework
has had little focus in literature.
Landing gear design includes different engineering disciplines including structures, weights, kinematics, economics and runway design. Interaction between these different disciplines makes the landing gear a
complex system. Automating the design process has shown to have the advantage of increased productivity,
better support for design decisions and can provide the capability of collaborative and distributed design.
The automation tools improve performance of current designs and simplify the development of new aircraft
configurations. In this thesis a systematic and automated landing gear design procedure is proposed.
Positioning the landing gear on the aircraft is limited by several requirements. Requirements include
take-off stability, touchdown stability, wing-tip and engine clearance, ground handling and stability while
taxiing. Evaluation of all these limits results in a feasible design space from which the shortest possible landing gear is found. From the resulting landing gear position, loads on the landing gear struts are calculated.
Tyres and wheels are selected and brakes and shock absorbers are designed. The assembly of landing gear
components can then be used to make an analytical weight estimation. This analytical weight estimation is
based on maximum stresses occurring within the structure due to extreme load cases prescribed in certification specifications. Preventing yielding and buckling within the structure then results in required component
thicknesses.
A multi-body model is then made, where structural parts are seen as rigid bodies. The multi-body model
evaluates and visualises the system dynamics. The oleo-pneumatic shock absorber forces and motion are
modelled using an analytical relation. An empirical tyre model models tyre motion and forces at the contact
point. These two models can accurately describe forces within the tyres and shocks due to externally applied
forces, which then allows for the evaluation of extreme landing load cases. This is done to verify empirically
estimated dynamic landing loads used in the weight estimation. And this is done to identify loading peaks
that could occur during a landing. In addition to the landing simulation a simulation of the landing gear retraction mechanism is done to check the kinematic feasibility and compliance to certification requirements.
Verification of results of the implemented landing gear with reference aircraft shows that landing gear
positions closely match with actual landing gear positions. The analytical weight estimation of the landing
gear assembly estimates the total gear weight with an error of 15 percent. This is comparable to the result of
an empirical weight estimation that has an error of 17 percent.
Multi-body simulation results show that dynamic loads during an extreme landing are similar to empirically estimated dynamic loads. For landing gears with multiple rows of tyres it is especially important to
look at landing loads, since a hard landing then creates peaks at high frequency in the shock loads. These
peaks originate from the interaction between front and rear axle tyres hitting the ground at different times. A
kinematic simulation of retraction and extension then verifies the kinematic feasibility. This simulation also
shows that the retraction and locking mechanisms work and that it can be stowed within the available space.
The resulting landing gear design and analysis tools complete the existing aircraft design tools, which then
forms the basis for the future improvement of automated transport aircraft design.

Acknowledgements
I would like to thank my supervisors dr.ir. Mark Voskuijl and dr.ir. Roelof Vos for their advise and contributions while making this Master thesis. I would also like to thank the other members of my examination
committee prof.dr.ir. Leo Veldhuis and dr.ir. Roeland de Breuker for taking the time to asses my thesis work.
Thanks also to TNO, who provided me with the MF-Tyre tyre model and the aircraft tyre model data. Appreciation is also due to Arne, Kristian and Reno for their valuable input and cooperation while trying to solve
similar programming and thesis problems.
I am grateful for all students and people at the department of Flight Performance and Propulsion who
have made time spend on the faculty very pleasant. Finally many thanks go to my parents and sister, who
always supported me in my efforts.
Niels Heerens
Delft, February 2014

vii

List of symbols
symbol

unit

description

(-) / (m2 )

aspect ratio / area

(m)

wing span / wheelbase

(N s / m)

damping coefficient

Cd

(-)

discharge coefficient

CL

(-)

lift coefficient

C L

(-)

lift slope gradient

(-)

load distribution factor

DI A

(m)

diameter

(Pa)

Youngs modulus

es

(m)

static tyre and shock deflection

(N)

force
2

(m/s )

gravitational acceleration

(m)

height

(m )

moment of inertia / pitch and roll moment of inertia

(m4 )

polar moment of inertia

(N / m)

spring stiffness

KE

(J)

kinetic energy

(N)

lift

(m)

length

(kg) / (Nm)

mass / moment

(-)

landing load factor

(-)

polytropic efficiency

(Pa)

pressure

(-)

scaling coefficient

(m3 )

first moment of area

(m)

radius

(m) / (m) / (N)

distance to edge taxiway / shock absorber stroke / shear force

St

(m)

tyre deflection

(Nm)

torque / forward component of inertia

(s) / (m)

time / track / thickness

tt

(m)

bogie track

(m/s) / (m3 )

velocity / volume

(N)

weight

ix

List of symbols

(m)

width

wb

(m)

bogie wheelbase

X
X

(m)

distance

(m/s)

velocity

(m)

longitudinal position

(m)

lateral position / distance

(m)

vertical distance

(rad)

angle of attack

(rad)

nose gear steering angle

(rad)

dihedral

(rad)

inclination or camber angle

(rad)

angle needed for minimum wheelbase

(rad)

angle needed for turnover angle

(-)

shock efficiency

(-)

tyre efficiency

(rad)

pitch angle / rotation angle

(rad)

longitudinal slip angle

0.25

(rad)

quarter chord sweep

(-)

friction coefficient

(kg / m)

density

(Pa)

normal stress

(Pa)

principal stress in 1 direction

(Pa)

principal stress in 2 direction

(Pa)

shear stress

(rad)

roll angle

(rad)

turn slip angle

(rad)

yaw angle / sideways turnover angle

(rad/s)

radial velocity

List of subscripts
subscript

definition

0.25

quarter chord

at static condition

at extended condition

at compressed condition

180 deg turn

180 degree turn

circle

centreline

taxiway centre line

cg

centre of gravity

drag

drag strut

friction

fillet

taxiway fillet

fwd

forward

ground

hydraulic

inner

length between nose gear and forward cg

LOF

lift-off

main gear

max

maximum

min

minimum

nose gear

outer

pneumatic

side

side strut

TD

touchdown

yield

xi

Glossary
AAC

Airplane Approach Categories

ADG

Airplane Design Groups

AIP

Aeronautical Information Publications

AC

Advisory Circular

ACN

Aircraft Classification Number

CBR

California Bearing Ratio

cg

centre of gravity

EASA

European Aviation Safety Agency

FAA

Federal Aviation Administration

ICAO

International Civil Aviation Organization

LOF

lift-off

MAC

mean aerodynamic chord

MTOM

maximum take-off mass

PCN

Pavement Classification Number

RTO

rejected take-off

TD

touchdown

xiii

List of Figures
1.1 The tricycle, bicycle and tailwheel landing gear layouts (Roskam, 1989a, p. 9) . . . . . . . . . . .

1.2 Activities during landing gear preliminary design (Currey, 1988) . . . . . . . . . . . . . . . . . . .

1.3 The landing gear model created in ADAMS for the simulation of retraction and extension kinematics (Zhang et al., 2000, p. 2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.4 The modelling of landing gear kinematics within a CATIA MDO procedure (Hrlimann et al.,
2011, p. 328) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.5 Research model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.6 Overall structure of the thesis work. In blue the added capabilities are shown. . . . . . . . . . . .

2.1 Typical main gear structure (Roskam, 1989a, p. 6) . . . . . . . . . . . . . . . . . . . . . . . . . . . .

12

2.2 Side view showing limits used in the positioning of the gears . . . . . . . . . . . . . . . . . . . . .

14

2.3 Dimensions used for the sideways turnover requirement . . . . . . . . . . . . . . . . . . . . . . .

15

2.4 Top view showing stability limits used in the positioning of the gears . . . . . . . . . . . . . . . .

16

2.5 Front view showing limits used in the positioning of the gears . . . . . . . . . . . . . . . . . . . .

16

2.6 Definition of dimensions used in calculating landing gear loads . . . . . . . . . . . . . . . . . . .

17

2.7 Nose and main gear longitudinal positions limited by nose gear loading limits and stowage limits. Green indicates feasible gear longitudinal locations. . . . . . . . . . . . . . . . . . . . . . . . .

18

2.8 Dimensions for determining the radius of a 180 degree turn . . . . . . . . . . . . . . . . . . . . . .

19

2.9 Centreline taxiing dimensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

20

2.10 A schematic representation of the structure of a bias ply tyre on the left and of a radial tyre on
the right (Goodyear, 2002) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

22

2.11 An A-frame type wheel cross-section. This wheel can be installed on the Boeing 737. (Honeywell, 2008) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

22

2.12 The electrically actuated carbon brakes developed by Goodrich and currently in operation on
the Boeing 787 (Goodrich, 2012) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

24

2.13 A typical retraction scheme of a wing mounted main landing gear on the left (Currey, 1988) and
a fuselage mounted main gear on the right (Torenbeek, 1982) . . . . . . . . . . . . . . . . . . . . .

24

2.14 Example of a good and poor actuator travel versus actuator load diagram (Roskam, 1989a) . . .

25

2.15 Working principle of an oleo-pneumatic shock absorber (Currey, 1988) . . . . . . . . . . . . . . .

26

2.16 Load stroke curve for a Boeing 707-321 aircraft. From 0 to static is isothermal compression and
from static to the right polytropic compression. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

27

2.17 Model of the oleo-pneumatic shock absorber showing parameters used in equations. . . . . . .

28

2.18 Average fuel price, from 1980 to 2008, paid by US airlines using current dollar cents and using
1987 dollar cents (Doganis, 2009) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

29

2.19 Structural model of a 4 wheel landing gear bogie used in the weight estimation. All externally
applied loads are shown. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

30

2.20 Tube cross-section showing the parameters used in the derivation of the stresses . . . . . . . . .

31

2.21 Comparison Von Mises yield criterion with the maximum shear stress theory. The difference is
largest at pure shear. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
xv

33

xvi

List of Figures
2.22 Level load case on the left and the tail down load case on the right (European Aviation Safety
Agency, 2012, p. 1-App A-2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

35

2.23 One wheel landing load case (European Aviation Safety Agency, 2012, p. 1-App A-3) . . . . . . .

35

2.24 The lateral drift landing load case (European Aviation Safety Agency, 2012, p. 1-App A-3) . . . .

35

2.25 Braked roll load case on the left and the ground turning load case on the right (European Aviation Safety Agency, 2012, p. 1-App A-4) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

36

2.26 The pivoting load case (European Aviation Safety Agency, 2012, p. 1-App A-5) . . . . . . . . . . .

36

2.27 SimMechanics displaying a portion of the model of a landing gear (Mathworks, 2012) . . . . . .

36

2.28 Tyre model description (Besselink, 2000) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

37

2.29 Tyre model coordinate system used . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

38

2.30 Measurement results showing the time histories of a typical loading during landing for an Airbus A300B2 main and nose landing gear (Ladda and Struck, 1991). The y-axis in the graphs
represents the force in x-, y- and z-direction from top to bottom. . . . . . . . . . . . . . . . . . . .

38

2.31 Tyre and shock absorber modelling. The tyre is modelled as a linear spring damper and the
oleo-pneumatic shock as a nonlinear spring damper. . . . . . . . . . . . . . . . . . . . . . . . . . .

39

3.1 Overview of the workflow to make an aircraft class 2.5 weight estimation . . . . . . . . . . . . . .

41

3.2 Class 2 landing gear weight estimation workflow . . . . . . . . . . . . . . . . . . . . . . . . . . . .

42

3.3 Side view showing take-off rotation limit and definition of main gear height. . . . . . . . . . . . .

43

3.4 B707 loading diagram produced by the class 2 weight estimation module. Aft and forward operational cg bounds are shown as vertical lines. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

45

3.5 Components of the landing gear design module . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

46

3.6 Graph showing the feasible design space of rotation angle versus height. . . . . . . . . . . . . . .

47

3.7 Workflow diagram of the gear positioning part of the PositionLandingGear module . . . . . . . .

47

3.8 The bogie layouts that are being evaluated by the PostionLandingGear module . . . . . . . . . .

48

3.9 The bogie design part of the PositionLandingGear module . . . . . . . . . . . . . . . . . . . . . .

49

3.10 The structural components of a bogie with 4 tyres modeled with tubes . . . . . . . . . . . . . . .

50

3.11 The weight estimation part of the PositionLandingGear module . . . . . . . . . . . . . . . . . . .

51

3.12 Class diagram of the PositionLandingGear, Class2WeightEstimation module and a LandingGear

part. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

52

3.13 Landing gear analysis workflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

53

3.14 Landing gear analysis retraction model used for a retraction/extension simulation. The initial
condition is shown. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

53

3.15 Landing gear analysis model used for performing a drop test simulation . . . . . . . . . . . . . .

54

3.16 Landing gear SimMechanics model of a main landing gear. Components include a oleo-pneumatic
shock absorber, side struts, axles, tyres and wheels. . . . . . . . . . . . . . . . . . . . . . . . . . . .

55

3.17 Tyre model block contents, which is part of the landing gear SimMechanics model. . . . . . . . .

55

3.18 The input parameters that define the main gear and nose gear available stowage . . . . . . . . .

57

3.19 The input parameters that define the main gear and nose gear available stowage . . . . . . . . .

57

3.20 Connection between side strut and shock absorber . . . . . . . . . . . . . . . . . . . . . . . . . . .

60

3.21 Locking mechanism between side strut and shock absorber . . . . . . . . . . . . . . . . . . . . . .

60

3.22 Oval blended wing body CWE2 result . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

63

3.23 Boeing 777ER CWE2 result . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

63

3.24 Comparison of class 2 weight estimation results of an oval BWB right and a Boeing 777 aircraft
left. Both aircraft have the same mission requirements. . . . . . . . . . . . . . . . . . . . . . . . .

63

List of Figures

xvii

4.1 Comparison between calculated landing gear positions and actual positions as published by
aircraft manufacturers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

66

4.2 Structural deflections and internal moments for the static load case. These results are for the
right main gear of a Boeing 707. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

68

4.3 Structural deflections and internal moments for the tail down landing load case. These results
are for the right main gear of a Boeing 707. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

69

4.4 Structural deflections and internal moments for the lateral drift landing load case. The top 2
graphs are for the left main gear and the bottom 2 for the right main gear of a Boeing 707. . . . .

69

4.5 Comparison between weight estimation results of the class 2 weight estimation module, the
class 2.5 weight estimation module and actual weight percentages as published by Roskam. The
x-axis is the maximum take-off mass and has a logarithmic scale. . . . . . . . . . . . . . . . . . .

71

4.6 Tyre model verification outputs using the hypothetical model parameters of Pacejka. Tyre forces
F x , F y and self aligning moment M z are plotted against slip angle , slip ratio and path curvature at = a/R. The camber angle is . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

72

4.7 Drop test simulation of the multi-body model of an Airbus A320 main landing gear . . . . . . . .

73

4.8 Shock absorber force, stroke and velocity for a two wheel A320-200 main landing gear landing
at 10 ft/s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

74

4.9 Tyre vertical force, deflection and velocity for a two wheel A320-200 main landing gear landing
at 10 ft/s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

74

4.10 Drop test simulation of the multi-body model a Boeing 777-300Er main landing gear . . . . . . .

75

4.11 Shock absorber force, stroke and velocity for a six wheel B777-300ER main landing gear landing
at 10 ft/s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

76

4.12 Rear axle tyre vertical force, position and velocity for a six wheel B777-300ER main landing gear
landing at 10 ft/s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

76

4.13 Landing simulation with tyre spin-up. Tyre spin velocity, shock absorber force, stroke and velocity for a two wheel A320-200 main landing gear landing at 10 ft/s . . . . . . . . . . . . . . . . .

78

4.14 Retraction motion of the multi-body model an Airbus A320 main landing gear . . . . . . . . . . .

78

4.15 Retraction actuation stroke versus force of an Airbus A320 main landing gear. . . . . . . . . . . .

79

4.16 Retraction/extension angle of an Airbus A320 main landing gear. The gear is initially in a extended position. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

79

A.1 Free body diagram of a general 4 wheel main landing gear bogie . . . . . . . . . . . . . . . . . . .

86

D.1 Airbus A320-200 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

93

D.2 Airbus A340-500 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

93

D.3 Airbus A380-800 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

94

D.4 Boeing 707-321 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

94

D.5 Boeing 727-200 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

94

D.6 Boeing 737-200 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

94

D.7 Boeing 777-300ER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

95

D.8 Boeing 787-800 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

95

D.9 McDonnell Douglas DC10-10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

95

List of Tables
2.1 Airplane Design Group (ADG) classification (Federal Aviation Administration, 2012b, p.13) . . .

19

2.2 Airplane Approach Category (AAC) classification (Federal Aviation Administration, 2012b, p.13)

19

2.3 Required runway width (ft) for a given Airport Approach Category and Airplane Design Group
(Federal Aviation Administration, 2012b, p.263) . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

19

2.4 Taxiway turn dimensions (Federal Aviation Administration, 2012b) (Chai and Mason, 1997, p.26) 19
2.5 Heatsink dimensions (Currey, 1988, p.143) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

23

3.1 List of possible inputs for the PositionLandingGear module . . . . . . . . . . . . . . . . . . . . . .

56

3.2 List of properties of a LandingGear part . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

58

4.1 Calculated tyre results compared with actual aircraft tyres used (Goodyear, 2002), (Michelin
Aircraft Tire, 2001) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

65

4.2 Comparison of flotation calculations of several reference aircraft for rigid and flexible pavements. 67
4.3 Comparison of class 2 wing and nose bogie assembly weight with actual weights as given by
Roskam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

68

4.4 Boeing 707 main landing gear structural component weight . . . . . . . . . . . . . . . . . . . . . .

70

4.5 Comparison of wing and nose bogie structural weight with actual weights as given by Currey
and Chai and Mason . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

70

4.6 Comparison of wing and nose bogie assembly weight with actual weights as given by Roskam .

71

B.1 List of airports with flexible pavements and their accompanying Pavement Classification Number (PCN) and associated subgrade category . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

88

B.2 List of airports with rigid pavements and their accompanying Pavement Classification Number
(PCN) and associated subgrade category . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

88

C.1 List of all output variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

91

xix

Contents
List of Figures

xv

List of Tables

xix

1 Introduction

1.1 Preliminary landing gear design. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.1.1 Textbooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.1.2 Landing gear types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.1.3 General landing gear design process . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.2 Aircraft design initiator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.3 Available research on automated landing gear design . . . . . . . . . . . . . . . . . . . . . . .

1.3.1 Landing gear design in an MDO procedure . . . . . . . . . . . . . . . . . . . . . . . .

1.3.2 Landing gear dynamics analysis and simulation . . . . . . . . . . . . . . . . . . . . . .

1.3.3 Gear dynamics analysis and simulation in combination with MDO . . . . . . . . . . . .

1.4 Thesis objective, aims and questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.5 Thesis contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2 Theory: landing gear design aspects

11

2.1 Landing gear components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

2.1.1 Runway and taxiway surface compatibility. . . . . . . . . . . . . . . . . . . . . . . . . 12
2.1.2 General positioning requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
2.1.3 Tyres . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.1.4 Wheels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.1.5 Brakes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.1.6 Brake actuation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.1.7 Kinematics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.1.8 Shock absorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.1.9 Economics of landing gear design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.2 Landing gear assembly weight estimation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.2.1 Tube stresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.2.2 Side struts sizing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
2.2.3 Landing gear dynamics analysis load cases . . . . . . . . . . . . . . . . . . . . . . . . 34
2.3 Landing gear analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
2.3.1 Tyre model description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.3.2 Drop test load case. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3 Implementation and use cases

41

3.1 Implementing the landing gear design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

3.1.1 Class 2 weight estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.1.2 Landing gear design module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
3.1.3 PositionLandingGear class 2.5 weight estimation . . . . . . . . . . . . . . . . . . . . . 50
xxi

xxii

Contents

3.2 Implementing the landing gear analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

3.3 User manual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.3.1 PositionLandingGear user manual . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
3.3.2 How to run the landing gear design module . . . . . . . . . . . . . . . . . . . . . . . . 57
3.3.3 Module messages description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.3.4 Landing gear analysis user manual. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.3.5 How to run the landing gear analysis module . . . . . . . . . . . . . . . . . . . . . . . 61
3.4 Definition of use cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3.4.1 Class 2 weight estimation example use case . . . . . . . . . . . . . . . . . . . . . . . . 61
3.4.2 Use case kinematic and multi-body analysis . . . . . . . . . . . . . . . . . . . . . . . . 62
4 Results and verification

65

4.1 Positioning results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

4.2 Runway flotation analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
4.3 Gear weight estimation accuracy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
4.3.1 Class 2 weight estimation results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
4.3.2 Structural parts weights . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
4.3.3 Comparison weight estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
4.3.4 Estimating weights for different aircraft types . . . . . . . . . . . . . . . . . . . . . . . 70
4.4 Tyre model verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4.5 Landing gear analysis results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.5.1 Drop test results for an Airbus A320 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
4.5.2 Drop test results for a Boeing 777 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
4.5.3 Tyre spin-up influence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.5.4 Landing gear retraction simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
4.6 Calculation run time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5 Conclusions

81

6 Recommendation

83

A Landing gear structure free body diagrams

85

B Reference airport pavement classification numbers

87

C PositionLandingGear output variables

89

D List of tested reference aircraft

93

Bibliography

97

1
Introduction
The design of an aircraft landing gear is one of the fundamental aspects of aircraft design. Landing gear design has become highly sophisticated, because it includes many different engineering disciplines: structures,
weights, economics and runway design. The process of the design of a landing gear is extensively documented in the books of Conway, Currey, Roskam and Torenbeek. Integration of the design methodology that
can be used in an automated design environment however has had little attention in literature.
In complex engineering systems (such as an aircraft) many different disciplines interact together. An automation framework of the design has as aim to get a better design by using these different interactions (Cumnuantip et al., 2005, p. 2). Automation of the design process has shown to have the advantage of increased
productivity, provide better support for design decisions and has the possibility of distributed and collaborative design. These new tools will improve the performance of current designs and ease the development of
completely new aircraft configurations (La Rocca et al., 2012, p. 1).
The group of FPP is currently developing and extending an automated aircraft design framework. It supports multidisciplinary design, analysis and optimisation of aircraft. It consists of a number of interconnected
multidisciplinary design and analysis tools, preventing that engineers need to do non-creative and repetitive
design work. Within this framework the design of the landing gear has not been taken into account. However
as the framework is continuously extended and becoming more detailed, the need for including the design of
the landing gear has become apparent.
The aim of the master thesis is to expand and enhance the existing design tool by creating and integrating
an automated landing gear design.
In this chapter first the landing gear design literature available is identified. Then the overall landing
gear design process is explained, followed by an overview of the existing aircraft design framework. Then the
current state-of-the-art in the field automated landing gear design and analysis is described. The last part of
this chapter includes the overall structure of the master thesis research.

1.1. Preliminary landing gear design

During preliminary landing gear design there are a number of decisions that are made. It for example needs
to be known what type will be used and if it needs to be retractable. There are a number of textbooks available
that provide different methods for finding a good solution.
1

1. Introduction

Figure 1.1: The tricycle, bicycle and tailwheel landing gear layouts (Roskam, 1989a, p. 9)

1.1.1. Textbooks
The book of Conway, Landing gear design is the first book that has been written on landing gear design
(Currey, 1988, p. ix). It was published in 1958 and therefore lacks some information needed for the design of
a new landing gear today. The book of Currey, Aircraft landing gear design: principles and practices is more
recent and is one of the most valuable books that are available for landing gear design (Niu, 1999, p. 431). It
includes all aspects of preliminary landing gear design.
The book of Roskam, Airplane design part 4: Layout design of landing gear and systems just as the book
of Torenbeek, Synthesis of subsonic airplane design includes landing gear design procedures for initial design, without going into details of the structural design.

1.1.2. Landing gear types

All current high speed transport aircraft use the tricycle layout, but there are different general wheel arrangements possible. The tailwheel and the bicycle layout are other design solutions that could be considered
(figure 1.1). For the tricycle layout, the main landing gear is behind the centre of gravity and there is a landing
gear positioned at the nose. This tricycle layout is superior to other layouts due to its improved stability, braking and steering ability, good over-nose visibility, horizontal cabin floor attitude and good take-off rotation
angle. Only specialised aircraft designs might require a different landing gear layout (Torenbeek, 1982). An
example is the British Aerospace Harrier II with a bicycle landing gear. It has a vertical engine in the middle
of the aircraft to provide lift and thus not enough room for a main landing gear in the middle.
Then there is also the choice if the landing gear is going to be fixed or retractable. The advantage of fixed
gears compared to retractable gears is the low weight, low complexity and low cost. The main disadvantage
is the high aerodynamic drag. Generally when cruise speeds go beyond 150 knots the drag penalty becomes
too large and retractable landing gears are more beneficial (Roskam, 1989a).

1.1.3. General landing gear design process

Figure 1.2 shows in general the aspects that need to be considered during the preliminary design phase. Here
flotation is the capability of the ground surface to support the aircraft. Preliminary design starts with a statement of the requirements and a concept formulation phase. During the concept formulation phase there are
a number of completely different aircraft concepts formulated and analysed briefly. At this point, the minimum that needs to be known is the weight of the aircraft and the cg range. The number and the size of the

Figure 2.2 illustrates the preliminary design activity and the factors to be
recognized. Note that, in the early phases, the landing gear designer may be
called upon to influence the requirements in the RFP. For instance, in one
project,
the flotation
requirement was established after an analysis had been
1.1. Preliminary
landing
gear design

STATEMENT'"'i~OF

,,

REQU,!REMENTSI . . . . .
~
-

STUO,E,

MARKETING

CUSTOMER

ANALYSIS

LOCATION

FORMULATION

TYPE

STRUCT. LAYOUT

FLOTATION REQ.

REQUEST
J
FOR

INFLUENCING/ ~

DEFIN TION

;'-

LAYOUT

PRELIM LOADS
PRELIMSTRESS

TIRES,
] A/C WT r., SPEEDS
WHEELS,
BRAKE ENERGY
BRAKES
LOADS

FLOTATION I
ANALYSIS

A/C WT 8 CG

BASIC
AIRFRAME STRUCT
KINEMATI CSJ

I
CONCEPT
i

ii
KNEELING
SPECIAL ~-1 CROSSWIND
FEATURESJ SELF-JACKING
WATER/GRAVEL
DEFLECTION

CONCEPT

FREEZE

1
I

"

PROPOSAL

TRADEOFF
STUDIES

i
Fig. 2.2 Preliminarydesign activity.

Figure 1.2: Activities during landing gear preliminary design (Currey, 1988)

wheels can then be determined and a review of runway flotation, compatibility with the airframe structure,
cost, weight, availability and overall complexity can be done (Currey, 1988). The most cost effective solution will be determined from this review and will be presented in the commercial equivalent of the military
Request for Proposal (RFP).
Following the concept formulation and the concept proposal is the more detailed project definition phase.
In this phase aircraft weight and cg range estimates are improved. As a consequence loads on the landing
gear can be better determined and the landing gear position can be further optimised. Then the stroke of the
gear during landing, the gear dimensions, clearances, brake sizing, materials, weight and cost are evaluated.
This is followed by a reiteration of runway pavement requirements, kinematics evaluation, steering concept
selection and the design of special features (as a kneeling mechanism, a self-jacking capability or a capability
to land on extremely rough surfaces).
In the project definition phase different trade-off studies are applicable and several trade-offs can be
made:
size and number of tyres against cost, weight and flotation;
gear location against cost, weight and performance;
different brake heat-sink materials;

1. Introduction
electric against hydraulic actuation systems.
The project definition phase is ended by a freeze of the landing gear design (Currey, 1988).

1.2. Aircraft design initiator

Computer aided aircraft design is becoming increasingly important, especially for multidisciplinary design
optimisation for nonconventional aircraft concepts. The automated design tools consist of a number of multidisciplinary design and analysis tools and prevent that engineers need to do non-creative and repetitive
design work.
Since 2002, the TU Delft department of Flight Performance and Propulsion has started with a Design
and Engineering Engine (DEE) (La Rocca et al., 2012). This DEE supports multidisciplinary aircraft design,
analysis and optimisation. Recently the DEE has been transformed into a new modular framework called the
Initiator with new capabilities. It now has the capability to make initial aircraft designs for both conventional
and nonconventional concepts as a blended wing body, a box-wing aircraft (Prandtl plane), canard or a three
surface aircraft.
The Initiator is composed of a set of initial sizing tools. It uses a limited set of top level requirements, such
as payload size, range, cruise speed, take-off field length and landing field length, to make a baseline aircraft
design.
The Initiator has been build up out of several independent modules. Based on the need of the user only
the required modules are executed. This is beneficial for reducing computation time and complexity. Different Initiator modules include aircraft component sizing, weight estimation and aerodynamic analysis modules. Classical synthesis methods from Raymer and Torenbeek are also implemented.
Another added capability of the Initiator is the possibility to manually introduce specific aircraft masses
(such as the maximum ramp mass and landing mass) as input. When this is done initial design weight estimates are bypassed and only the input reference weight values are used. This makes verification of module
results with reference data better, since inaccuracies of the initial sizing are factored out. A good information
source for aircraft reference data are airport operations manuals published by aircraft manufactures.

1.3. Available research on automated landing gear design

In the field of automated landing gear design within a multidisciplinary design optimisation (MDO) environment not much research has been done. Most research in the field of MDO focus on other aspects such as
wing geometry optimisation. The MDO environment described by Chai and Mason is one of the few that includes a detailed (similarly detailed as the thesis work) analysis of the landing gear design (Chai and Mason,
1996).

1.3.1. Landing gear design in an MDO procedure

Chai and Mason have integrated the landing gear design in an MDO procedure for the conceptual design of
large transport aircraft. An automated model of the landing gear concept has been developed that can assess
typical EASA requirements automatically. Airfield compatibility issues were automated and the results could
be used in a complete MDO design program.
The main purpose of this implementation was to study the effects of variations of the landing gear design
parameters on the configuration, system integration, airfield compatibility and weight. The work of Chai and
Mason also includes methods for estimating cg range travel, analytical gear weight estimation, selection of
tyres, wheels and brakes, shock absorber sizing and feasibility of retraction/extension. Kraus also describes
an analytical method of weight estimation (Kraus, 1970) that could possibly be used in an automated analysis. This procedure however lacks essential information on load calculations and structural design criteria,

1.3. Available research on automated landing gear design

making implementation of the method very difficult. Other methods of Currey, Roskam and Torenbeek are
all based on statistics.

1.3.2. Landing gear dynamics analysis and simulation

To analyse a landing gear design for the previously mentioned load cases the equations of motion need to be
derived. These were normally determined in the past by hand and converted for use in a computer program.
These programs were mainly developed in-house and written in FORTRAN code.
Now there is the option to use programs like Matlab/Simulink as developing platform that are more flexible than in-house created programs. Tools for eigenvalue analysis, plotting and much more can be easily
used in the analysis (Besselink, 2000, p. 161).
Finite element analysis on landing gears is generally done for analysis of stresses and stiffness. Finite
element packages as NASTRAN and ABAQUS do not efficiently address problems with mechanisms as the
retraction system of the gear. This is because for each position of the shock absorber, actuator and other
parts a separate analysis will need to be performed.
Multi-body analysis software is the preferred tool for dynamical analysis of mechanisms and groundbased vehicles. This type of software has shown to perform the analysis very efficiently (Spieck, 2004, p. 9).
The advantage of using multi-body analysis software is that the equations of motion are automatically derived. With this software it is thus possible to focus on the engineering aspects instead on focussing on the
derivation of the equations. Multi-body simulation packages provide important information and allow for
the virtual testing of new designs and concepts.
Multi-body analysis software can be used for different aspects of ground dynamics in different phases of
the aircraft design process, such as (Spieck, 2004, p. 25):
landing loads, high speed roll;
airframe and landing gear ground loads;
aircraft ground manoeuvres, such as push-back, sharp taxi turns;
behaviour on rough pavements;
shimmy analysis;
certification;
unconventional configuration analysis, or analysis that are difficult or dangerous to perform physically.
Different modes of analysis exist for the analysis using multi-body tools. The analysis can be:
static, when no motion occurs;
kinematic, for closed loop systems such as the extension or retraction of the gear;
linear, the equations of motion are linearized to represent dynamical behaviour;
nonlinear, numerical solutions are used to determine the full non-linear behaviour on events such as
touchdown.
Commercial multi-body simulation tools are nowadays being used by almost all major aircraft and landing gear manufacturers for the analysis of ground dynamics (Spieck, 2004, p. 17). Literature in the field of
landing gear dynamics and simulation that makes use of multi-body dynamics tools is also numerous. This
illustrates the importance that these tools have gained over the years.

loading situation. As this is an integrated system analysis, pressure losses across various components are
calculated based on their location on the flow path. A series of force balance equations and hydraulic
pressure loss equations are embedded in the model to calculate the force at retraction actuators and the
pressure drops across various components of the system during the gear retraction/extension process
based on the instantaneous fluid flow into and out of the retraction actuators. The time history of the gear
1. Introduction
movement is then obtained. The simulation results can then be presented and motion of the landing gears
can be graphically animated. The graphic representation of an integrated landing gear system model is

shown in Figure 1. (with nose gear aft door wire-framed for clarity).
Figure 1.3: The landing gear model created in ADAMS for the simulation of retraction and extension kinematics (Zhang et al., 2000, p. 2)

Figure 1

Integrated Landing Gear ADAMS Model

LANDING GEAR SYSTEM RETRACTION/EXTENSION ANALYSIS

ADAMS is a multi-body dynamics analysis tool also used by Airbus for the dynamical analysis of the landSystem
Description
ing gear (Coetzee
et al.,
2006). Messier-Dowty, the worlds largest manufacturer of aircraft landing gears, uses

ADAMS in theThe
simulation
ofthe
retraction
and
extension
of the gear.
Thearemodel
of thegear
gear structure
objectives of
landing gear
system
retraction/extension
analysis
to verifyconsists
that the landing
system retraction/extension meet the performance requirements under various gear operation conditions
and to provide load information on actuators, door links, and major joints. The landing gear system has to
meetcorrelation
the retraction/extension
requirement
based onfrom
the available
pump
flow.
The tests
load of(Zhang
the dooret al., 2000).
have a very good
with testtime
results
originating
test rigs
and
flight
mechanism during the retraction and extension process has to be determined by simulation to size the
The simulationlinks.
can be used for the sizing of the actuator and the hydraulic system.

and door mechanisms, including a hydraulic system as shown in figure 1.3. The simulation has shown to

Boschetto et al. has used ADAMS for the analysis of landing gear dynamics on the ground of a trainer airThe landing gear system referred herein consists of a nose landing gear (including two hydro-

craft. First the mechanically

behaviour of
theforward
gear during
drop-tests
simulated,
bytwo
implementing
different
driven
doors and
an aft doorwas
attached
to the gear),
main gears (with
outboard tyre models
doors attached to them), gear locking mechanism, hydraulic subsystem (including an engine driven
and a shock absorber
model. Then the gear model was built into a complete aircraft rigid body to simulate
_____________________________________________________________________________________________
ground manoeuvring
load cases. The results of the simulation can be used for the analysis of load path into
2000 International ADAMS User Conference

Page 2 of 10

the fuselage structure during the design phase. The use of ADAMS in the preliminary design phase can be
complex and thus less suitable (Boschetto et al., 2000, p. 8). For later stages of the design it can however prove
to be very valuable.
Cessna for example uses the multi-body simulation package LMS Virtual.Lab to model the components
of the landing gear of a new type of aircraft, including tyres, struts and trunnions. A separate model for the
flexible tyres is thus also included. A parametric analysis can be performed with this software by running
the simulations several times automatically for different aircraft cg positions. The simulation is used to verify
compliance with FAA requirements for static and dynamic flight conditions, such as taxiing, take-off, retraction and landing (LMS International, 2008).
Then there is LMS Imagine.Lab that is capable of doing analysis for different disciplines, such as electrics
and hydraulics simultaneously. Analysis can be done for landing, retraction, braking and steering systems.
All ground loads can be analysed, since the software includes multi-body dynamics, structural dynamics and
optimisation. Also validation of anti-skid systems or steering systems can be done within LMS Imagine.Lab.
Spieck uses the multi-body simulation package SIMPACK to accurately model the ground dynamics of the
landing gear. Aeroelastic and aerodynamic influences have been included in the analysis, which is normally
left out of the analysis. The aerodynamic and aeroelastic effects have shown to be of large importance on the
total dynamic behaviour of the aircraft on the ground (Spieck, 2004, p. 110).
Multi-body dynamics packages allow the use of models for flexible bodies, such as tyres. But if vibrational

1.3. Available research on automated landing gear design

Fig. 8. Slat geometry generation.

Fig. 10. CAD/CFD gateway.

out the CFD calculation. Most importantly, these inpu

ters include the aircraft geometry in the form of cros
for the fuselage, the wings and the empennage (Fig.
Knowledgeware tools, such as Knowledge Patterns and
are used for automatic retrieval of these cross sections
secting the aircraft geometry with a moving cutting plan
3.7. Fuel distribution (CAD)

Fuel loads are a significant contributor towards the

loads. The calculation of the fuel loads requires precise
of the
fuel distribution within the fuel tanks. With CA
Figure 1.4: The modelling of landing gear kinematics within a CATIA MDO procedure (Hrlimann et al., 2011,
p. 328)
automatic method for the calculation of the fuel distrib
interpolation elements (RBE2, RBE3). Spring elements (CLAS2) in
implemented: During static loadcases, the fuel surface
combination with non-stiffening interpolation elements are used
perpendicular to the acceleration vector. In an iterative p
to create realistic wing root restraint conditions that allow for a
fuel height in every fuel tank is adjusted until a target fu
certain compliance in the areas where the wingbox is attached to
is reached. The upper half of Fig. 11 shows the center fue
the fuselage.
The mesh
is variable and
defined by acharacteristics
separate
analysis, such as shimmy analysis,
needs
to besize
performed,
theis flexibility
cannot be changed
the corresponding
fill curve for this fuel tank during cr
input parameter. The standard mesh size of 200 mm results in an
In the lower half, two fuel distributions for acceleration
much in these packages. SAMSEF
MECANO
would
be
a
better
option
for
this
4
4type of analysis (Besselink, 2000,
FEM model with approximately 1e FEM elements and 4e DOFs
vertical direction (top) and horizontal direction (botto
(degrees of freedom).
p. 161).
seen. Also shown is the center of gravity for every fue
CATIA V5 lacks provisions for the generation of hydros
3.6. Aerodynamic model (CFD)
sure loads, a custom MATLAB script is used for the cal
Coetzee et al. used Matlab SimMechanics for the analysis of dynamical loads during
ground handling.
the hydrostatic fuel pressure loads acting on the fuel ta
The advantage of using SimMechanics
is that
a very
toolset
that
can even
be In
used
forV5real-time
The automated
useitofisthe
CFD efficient
solver BLWF
for the
calculaaries.
CATIA
these pressure loads are applied to
tion of the aerodynamic loads requires both export and import
tural model with the help of the built-in surface load
applications (Coetzee et al., 2006, p. 3). The Initiator is largely made in Matlab and since Matlab and Siminterfaces to be implemented in CATIA V5 [2]. The export interinterface. The same pressure mapping functionality is als
the
application
of surface-distributed aerodynamic loads
face provides
BLWF with
necessary
input parameters
carry
Mechanics can be used seamlessly
together
thisthemakes
SimMechanics
antoexcellent
choice
for performing
Fig. 9. Landing gear kinematics.

kinematic and possibly other analysis.

1.3.3. Gear dynamics analysis and simulation in combination with MDO

In general multi-body dynamics software is used for separate case studies for a specific type of landing gear.
When looking in literature for the combination of MDO procedures with more detailed landing gear design
other than the work of Chai and Mason described in section 1.3.1, very little can be found. Hrlimann et al.
shows a MDO implementation where landing gear kinematics are included for the estimation of the mass
of a wing box structure. This implementation is made within CATIA and includes only a simplified model
of the gear structure where the structure is represented by one-dimensional bar elements (figure 1.4). The
landing gear model is used for assessing the position and clearances of the landing gear. When looking at the
computational performance of CATIA within a MDO framework, it has proven to be insufficient for multidisciplinary optimisation (Hrlimann et al., 2011). The creation of a MDO framework within CATIA is therefore
questionable.
Then Cumnuantip et al. shows a multi-body simulation integrated into a MDO process. The multi-body
simulation is used for the selection of a landing gear layout for a Blended Wing Body aircraft. Cumnuantip
et al. also states that this problem has had little attention in literature, confirming the low number of search
results found on this topic.
The multi-body simulation program SIMPACK is used by Cumnuantip et al. for evaluating different landing gear layout concepts. First the gears overall dimensions, the shock strut length and oleo properties required are determined. The gear mass is then calculated analytically based on different load cases from regulations. These details are transferred to SIMPACK and simulated for the 3-point level, the one-wheel and
the tail down landing load case. Then the maximum load on the gear support structure and the total mass
of the aircraft concept is then calculated and used in the optimisation loop as the objective. Results of the
optimisation were found satisfactory (Cumnuantip et al., 2005).

1. Introduction

1.4. Thesis objective, aims and questions

The previous sections described that the focus of current research on automated multidisciplinary aircraft
design is not on landing gear design. As a result a research void on a primary aircraft design aspect has been
left to be investigated.
This thesis research will fill a part of the research void and add new understanding into this subject. The
research objective of the master thesis is to expand the knowledge base of the existing automated design environment by creating and integrating an automated landing gear design tool. The automated landing gear design encompasses disciplinary analysis of structures, kinematics, runway pavement design and weight. The
knowledge that is needed to achieve this research objective is formulated in the form of research questions.
The main research questions are formulated as follows:
Which subjects are relevant for the development of an initial design sizing and selection tool, a parametric landing gear model and an analysis environment for landing gear kinematics and dynamic
loads?
How do we combine the sizing and selection tool, the parametric model and the analysis environment
into a single landing gear design tool?
How to integrate the landing gear design tool into the existing automated design environments?

landing gear design theory

initial design sizing and selection tool

KBE theory

parametric landing gear model

landing gear analysis theory

analysis environment for landing gear

kinematics and dynamic loads

landing gear design

options tool

extended automated
design environment

Figure 1.5: Research model

The research questions can be divided into several research subquestions:

Which parts of landing gear design theory need to be included into an initial design?
Based on which criteria can a initial landing gear design be sized and selected?
Which parameters are required for a parametric landing gear model?
Which landing gear analysis tests need to be performed in order to meet requirements?
Based on which criteria are landing gear design options compared and selected?
How to structure and combine landing gear design with the existing automated design environment?
The answers to these research subquestions together are needed for the main research questions. The
main research questions provide a framework to achieve the research objective. The final result is a complete
landing gear design tool that is integrated into the existing automated design framework.
Figure 1.6 considers the overall picture of the thesis work. New landing gear design capabilities are added
to the Initiator and a new landing gear analysis tool is added. The overall structure is shown as a design
optimisation loop, which is the final goal of the greater aircraft design framework. Design optimisation is
however not the aim of the this thesis. It will be added in a future research project.

1.5. Thesis contents

top level
requirements
add landing gear
design capability

Initiator
aircraft initial
sizing and design

start

landing gear
analysis tool

no
yes
end

converged and complies

with requirements?

converger and
evaluator

Figure 1.6: Overall structure of the thesis work. In blue the added capabilities are shown.

1.5. Thesis contents

The thesis is setup parallel to the research questions. In chapter 2 all landing gear design aspects and theory
required to make an automated landing gear design are discussed. Then in chapter 3 the implementation of
the landing gear design tool, how to use it and different use cases are explained. Results and a verification of
the results is done in chapter 4. Finally a conclusion is drawn in chapter 5 and recommendations are made
in chapter 6 based on the previous chapters.

2
Theory: landing gear design aspects
The landing gear provides several essential functions. The gear absorbs landing and taxiing loads and transmits these loads to the rest of the airframe. Manoeuvring and braking of the aircraft on the ground during
taxiing, take-off roll and landing roll is done by the landing gear. The landing gear also provides the ability of
aircraft towing and protection of damage to aircraft and ground surfaces.

2.1. Landing gear components

A typical main landing gear is built up out of several components, which fulfil different functions (figure 2.1).
The shock absorber and tyres take up most of the shocks during landing and taxiing over rough surfaces.
While the side and drag strut take up side loads on the gear due to engine thrust, braking or a sideways
landing.
Requirements such as minimum weight, component maximum strength, maximum reliability, low cost,
airfield compatibility and others can be conflicting (Currey, 1988, p.5). To prevent that a single requirements
such as minimum cost, gets too much emphasis, aviation authorities prescribe safety requirements. Requirements for large aircraft powered by turbine engines have to follow the certification specifications from
European Aviation Safety Agency (CS25) and have to follow the Federal Aviation Regulations from the Federal
Aviation Administration (FAR25).
The regulations that will be looked at include:
take-off and landing clearance;
touchdown/take-off stability (CS 25.231);
proper clearance between aircraft parts;
taxi stability (CS 25.231 and CS 25.233);
compatibility with runway pavement;
landing load cases (CS 25.723 and CS 25.473);
ground handling load cases (CS 25.235);
landing gear emergency systems (CS 25.729 and CS 25.499);
11

12

2. Theory: landing gear design aspects

Figure 2.1: Typical main gear structure (Roskam, 1989a, p. 6)

The Acceptable Means of Compliance, AMC 25.723, AMC 25.729, AMC 25.735 and AMC 25.745 give information about shock absorption tests, the retraction system, the braking systems certification and nose wheel
steering. The AMCs provide guidance that can be used as a means for showing compliance to the requirements. The relevant requirements and complementary documents are also listed in the AMCs (European
Aviation Safety Agency, 2012).

2.1.1. Runway and taxiway surface compatibility

Airfield compatibility is an important issue. Landing gears of new aircraft need to be compatible with existing
runways due to the high cost of runway modifications. The runway pavement bearing strength determines
the arrangement and the number of tyres needed to comply with flotation requirements. Runway and taxiway
geometry also constrain the location of the gears.
Loads on each landing gear strut and also the load on each tyre may not cause (European Aviation Safety
Agency, 2012):
structural damage to the landing gear or to the airplane;
damage to the tyres;
damage to runway or large ground surface deformations.
Different kind of runway surfaces could be considered in the design. For large commercial aircraft 2 types
of pavements are considered: a flexible runway surface pavement with an asphalt top surface and a rigid
pavement with a concrete top surface. About two-third of the major airports have flexible pavement runways.
To make a landing gear design compatible with all airport runways it is intended to operate on, one can
compare aircraft loads with pavement ratings of runways. The pavement ratings are derived from engineering
tests and traffic history or from aircraft currently using the airport. The standard rating is established by
ICAO and uses the Aircraft Classification Number (ACN) and Pavement Classification Number (PCN). The
ACN represents the relative load intensity of the main landing gear of the aircraft. Different methods are used
to calculate the ACN for rigid and flexible pavements (Christy, 2009). ACN values are calculated by aircraft
manufacturers and are published in so called aircraft characteristics for airport planning documents and in
ICAO Annex 14.
The PCN describes the actual load carrying capability of the runway, taxiway or ramp of an airport. All
major airports in the world have been assigned a PCN. PCN values are published in Aeronautical Information

2.1. Landing gear components

13

Publications (AIP) and in the Jeppesen Airport Directory. When the ACN is lower than or equal to the PCN
this means that the maximum take off weight is unrestricted.
The ACN value for a new aircraft cannot be calculated easily. The procedure is described in Annex 14 of the
ICAO airport pavement design manual and requires a computer program. The computer program listed in
Annex 14 is implemented in COMFAA, a computer program made by the FAA. COMFAA is able to calculate the
ACN for different pavement sub-grade categories and it can calculate PCN values and pavement thicknesses.
As input is required the aircraft gross weight, the percentage of GW on the main gear, the number of gear
struts, the number of wheels per main gear and the tyre pressure (Federal Aviation Administration, 2011).
The COMFAA program is a stand-alone program and thus cannot be directly used in an automatic landing
gear design procedure. The FAA though provides the source code of the COMFAA program and also provides
a detailed documentation of how the ACN calculation procedures are implemented. It is thus possible to alter
the source code into a new program that can be used in the flotation analysis of the aircraft design.
The COMFAA program has more capabilities than necessary for the calculation of ACN values. Therefore
only parts of the COMFAA program are used for implementation in the flotation analysis.

2.1.2. General positioning requirements

In the following sections the basic design issues are discussed that need to be taken into account when positioning the landing gear. The issues include the take off/landing requirements, ground handling, ground
clearance and stability requirements.
It is important to choose a landing gear position that does not need to be modified with respect to strut
length and position under the wing when considering future aircraft growth options. An increase in fuselage
length should be taken into account, because this reduces the maximum take-off rotation angle (Chai and
Mason, 1997, p. 22). This is assuming the aircraft horizontal floor attitude is kept the same. Changing the
floor attitude will make the aircraft less convenient for aircraft operations.
The result of positioning the landing gear is a minimum main gear length. From the main gear length and
the desired attitude of the fuselage the nose gear length is found.
Take-off stability
On the ground, during landing and during take-off the aircraft should be able to safely pitch up or down
without hitting the ground or having a tendency to turn on its side. To check whether or not this could occur
the coupling of the gear position with the take-off/landing performance and the aircraft centre of gravity
range needs to be considered.
The required pitch angle for taking off (when the landing gear is fully extended) is estimated using the
equation (Torenbeek, 1982, p.350):

(
)
l 2C L LOF
d 2l 1
+
LOF = LOF +
d t VLOF
gC L

(2.1)

This equation is based on the result of Pinsker (Pinsker, 1969) that takes into account the dynamics of
the aircraft motion at take-off. When more accurate information is not available (typical in an initial design
stage) the lift curve slope is estimated for high aspect ratio wings with (Torenbeek, 1982, p.351):
dC L 2 cos 0.25
=
d
1 + 2/A

(2.2)

(
[
)
]
1
(C L max )to
C L cr
2 (C L max )f=0
1 180
p

(1 + )

2
A
cos 0.25
cos 0.25
cos 0.25
A

(2.3)

C L =
and the angle at lift-off:
LOF =

14

2. Theory: landing gear design aspects

hcg

LOF
takeoff pitch
static ground line

es
touchdown stability

Figure 2.2: Side view showing limits used in the positioning of the gears

Here p is within the range of 0.15-0.20, dependent on aerodynamics. Equation 2.3 takes into account the
ground effect, but tailplane trim is left out, flap deflection at take-off is assumed to have little effect on the
critical lift coefficient and it is assumed that the fuselage is horizontal during cruise. If the lift coefficient with
zero flap deflection is unknown it may be estimated as (C L max )f=0 = 2.10 cos 0.25 when slats are present on
the aircraft.
Landing stability
The previous equations are applicable to the take-off rotation. For landing the maximum pitch angle that
will be achieved is lower than for take-off. This is due to the flap deflection angle, which is higher at landing.
Therefore and because there is limited information available the pitch angle at touchdown, T D can assumed
to be equal to the pitch angle at take-off, LOF (Torenbeek, 1982, p.351).
During landing touchdown the aircraft should have a pitch down tendency. This sets a limit on the minimum longitudinal position for the main gear. The worst-case landing is a landing with the cg aft and the cg
at the highest position. When there are no other loads influencing the pitch down tendency the main gear
should be minimally behind the aft cg by a distance of (Torenbeek, 1982, p.352):
l m (h cg + e s ) tan TD

(2.4)

where e s is the static deflection of the shock and tyre at the static condition and h cg is the height from the
ground of the cg while taxiing.
Sideways turnover and ground stability limits
A crosswind landing or a turn during taxiing at high speed can cause the aircraft to tip on its side. The sideways turnover angle , which should be lower than 63 (Currey, 1988, p.352), is equal to:
tan =

h cg
l n sin

(2.5)

where is equal to:

tan =

t
2(l m + l n )

(2.6)

Here the nose wheel track is assumed to be of minimal importance. Using equations 2.5 and 2.6 the
minimum required track (t ) can be determined to prevent turnover.
To ensure that the aircraft remains stable during taxiing and touchdown another limit should be taken
into account. When assuming that the nose wheel location is fixed, a circle is drawn with a radius of 0.54h cg
placed at the forward cg position as in figure 2.4. This radius is based on the recommendations described

2.1. Landing gear components

15

cg

groun

d line

static

lm

ln

hcg

Figure 2.3: Dimensions used for the sideways turnover requirement

by Torenbeek (Torenbeek, 1982, p.354). Then the line drawn from the nose wheel position to the tangent of
the circle determines the minimum lateral position of the main gear. This can also be calculated analytically
using:
(
(
))
0.54h cg
y mlg min stability = tan sin1
(l n + l m )
ln

(2.7)

When assuming the main gear location is fixed, a limit on the nose gear position can be found by drawing
a line from the fixed main gear position to the tangent of a circle with radius 0.54h cg . This is also shown in
figure 2.4 and this results in:
= tan1

lm
t /2

(
(
)
)
0.54h cg
l n mi n = tan sin1
+ t /2 l m
l m sin()

(2.8)

(2.9)

Wing and engine clearances

During a crosswind landing the aircraft could roll on its side. This rolling should be possible to a certain
degree, without aircraft parts hitting the ground. For the engine nacelles a roll angle of 5 degrees with an
additional 6 inch clearance is given by Raymer (Raymer, 1999, p.62). This requirement is also visible in figure
2.5.
For transport aircraft the wings are usually swept aft. When the aircraft nose pitches up there is a risk that
the wing touches the ground. This can be taken into account by calculating the limit roll angle when the wing
tips just touch the ground (Torenbeek, 1982, p.350):
tan = tan +

2h g

tan tan
(2.10)
bt
Here and are the angles defined in figure 2.4 and 2.5, b is the wing span. In this equation the wings
are assumed to stay in its rigid position.

16

2. Theory: landing gear design aspects

touchdown stability

sideways stability mlg fixed

ln

lm

0.54 hcg
min nlg load

ymlg min stability

sideways stability nlg fixed

max nlg load

Figure 2.4: Top view showing stability limits used in the positioning of the gears

tip
height
loss
wingtip clearance

clearance

engine clearance
5 deg

Figure 2.5: Front view showing limits used in the positioning of the gears

sideways turnover

2.1. Landing gear components

17

Centre of gravity and gear position

The aircrafts cg location is one of the most critical in the design and positioning of the landing gear. An
improper cg position could result in tip back, turn over or tipping on the side of the aircraft. Also during the
landing rollout the weight on the main gears should be high enough to make sure that the brakes can provide
enough braking power.
ln

lm

ll

fwd cg

aft cg

nose gear

main gear

Figure 2.6: Definition of dimensions used in calculating landing gear loads

The load on the nose and main landing gear can be calculated using figure 2.6. The maximum main gear
load for all main gear struts is subsequently equal to (Currey, 1988):
ln
W
lm + ln

(2.11)

lm + ln ll
W
lm + ln

(2.12)

lm
W
lm + ln

(2.13)

F m max =
The maximum nose gear load is:
F n max =
and the minimum nose gear load:
F n min =

Here W is the maximum ramp weight. The maximum and minimum nose load is a design parameter that
has a large influence on the gear positioning. If a first estimate has to be made, Currey advises a maximum
and minimum nose load of 15 and 8 per cent of the maximum ramp weight.
Solving for l n in equation 2.13, results in a maximum nose gear limit that is a function of the main gear
longitudinal position. This limit is the maximum nose gear load limit in figure 2.7. Similarly solving for l l in
equation 2.12 gives the minimum nose gear limit, which is also shown in figure 2.7.
The longitudinal position of the main gear and nose gear is also limited by stowage constraints and by
wing spar constraints. These constraints together leave open a small design space for feasible nose and main
gear longitudinal positions.
Ground operations requirements
Airport taxiways are constructed based on standards defined by the FAA. The Federal Aviation Administration
provides standards and requirements for the design of airports (Federal Aviation Administration, 2012b). In
order to operate on most airports it is advisable to comply with these requirements. The Advisory Circular
gives requirements on the turning radius and centreline guidance taxiing.
The FAA categorises aircraft in different Airplane Design Groups (ADG) and Airplane Approach Categories
(AAC). In table 2.1 and 2.2 the different categories are listed. The aircraft is placed in the highest group based
on its tail height or wingspan. From the AAC category and ADG group a required runway width can be obtained from table 2.3.

18

2. Theory: landing gear design aspects

rear spar location

nn
e
os

30

ar
ge
d,

loa

auxiliary spar location

40

mi
20

ma
x. n

aft
cg

x position nose landing gear [m]

50

os
eg

ea

10

r lo
ad
,

fw
dc

max. nose gear stowage

min. nose gear stowage

20

21

22

23

24

25

26

27

x position main landing gear [m]

Figure 2.7: Nose and main gear longitudinal positions limited by nose gear loading limits and stowage limits. Green indicates feasible
gear longitudinal locations.

The aircraft should be able to make a turn within this runway width. The turn radius of the aircraft is
a function of the nose gear steering angle, , assuming the aircraft only turns by pivoting the nose gear. A
maximum steering angle for transport aircraft is about 60 degrees (Currey, 1988, p.198). Then the turning
radius can be obtained graphically in figure 2.8 or by making use of the maximum steering angle and the
following equation:
t
(2.14)
2
If the turning radius poses a problem due to a too high gear track or wheelbase the turn radius can be
r 180 deg turn = b tan(90 ) +

decreased by installing a main gear steering system. This is for example done for the Boeing 747, Boeing 777
(Chai and Mason, 1997, p.24) and the Airbus A380 main gear bogies (Hebborn, 2008).
Chai and Mason provide a way to determine if the aircraft can turn on airport taxiways. This poses limits
on the aircraft wheelbase and gear track. The angle between the tangent of the taxi turn centreline (with
radius, R centreline ) and the aircraft centre line is called the castor angle, castor . This angle is equal to:
sin castor =

b
R centreline

(2.15)

The castor angle should be smaller than the maximum turn angle of the nose gear. This thus puts a
constraint on the aircraft wheelbase b.
To avoid that the main gear goes off the taxiway while turning, the main gear should be kept a distance
S away from the taxiway edge (see figure 2.8). Using taxiway dimensions and a safety margin in table 2.4 the
maximum aircraft gear track can be found with:
R fillet =

2
R centreline
+ b 2 2R centreline b sin castor t /2 S

(2.16)

2.1. Landing gear components

19

ADG

Tail height (m)

Wingspan (m)

AAC

Approach speed (kts)

<6

< 15

< 91

II

6-<9

15 - < 24

91 - < 121

III

9 - < 13.5

24 - < 36

121 - < 141

IV

13.5 - < 18.5

36 - < 52

141 - < 166

18.5 - < 20

52 - < 65

> 166

VI

20 - < 24.5

65 - < 80

Table 2.1: Airplane Design Group (ADG) classification (Fed-

Table 2.2: Airplane Approach Category (AAC) classification

eral Aviation Administration, 2012b, p.13)

(Federal Aviation Administration, 2012b, p.13)

AAC \ ADG

II

III

IV

VI

60

75

100

150

60

75

100

150

100

100

150

150

150

200

100

100

150

150

150

200

100

100

150

150

150

200

Table 2.3: Required runway width (ft) for a given Airport Approach Category and Airplane Design Group (Federal Aviation Administration,
2012b, p.263)

b
t

r 180 deg turn

Figure 2.8: Dimensions for determining the radius of a 180 degree turn

ADG III

ADG IV

ADG V

ADG VI

R centreline

100

150

150

170

R fillet

55

80

85

85

10

15

15

20

Table 2.4: Taxiway turn dimensions (Federal Aviation Administration, 2012b) (Chai and Mason, 1997, p.26)

20

2. Theory: landing gear design aspects

castor

Rfillet
Rcentreline
S

castor

Figure 2.9: Centreline taxiing dimensions

2.1.3. Tyres
Tyres are exposed to quite severe dynamic and static loads during landing, take-off roll and taxiing. During
touchdown the tyres provide a significant amount of the shock absorption capability of the landing gear. Tyres
are designed around a maximum allowable static load. For any centre of gravity position in combination with
the highest weight pressing on the landing gear this allowable load may not be exceeded.
When selecting tyres one can choose for a conventional bias-ply tyres or radial tyres. Radial tyres are
more recently developed and have as advantage a lower weight and a longer lifespan. Wearing of the tread
is reduced with 40 to 60 per cent (Currey, 1988). Radial tyres are therefore preferred for new aircraft types.
Radial tyres are up to 20 per cent lighter due to their construction, which minimises the shear stresses in the
rubber and efficiently distributes the loads. Radial tyres are constructed with additional steel belts that run in
the radial direction and have an additional advantage that the footprint area is larger (about 10 per cent). A
larger footprint area improves flotation characteristics and reduces hydroplaning. Radial tyres can withstand
higher overload bearing stresses and can withstand under-inflation better. When the tyres do fail they do this
less sudden than bias ply tyres and indications of damaged tyres can be more easily spotted.
For a conventional bias-ply tyre the belts run in varying angles, usually between 30 to 40 degrees. Bias-ply
tyres are however still widely used on current commercial aircraft (Goodyear, 2002). Both bias-ply and radial
tyres can be ordered with different options. An example option are tyres with so called chines. A chine is
a circumferential bulge shaped to deflect water sideways from the engines. Tyres with chines were initially
developed for aircraft with rear-mounted engines.
Aircraft tyre manufactures include Bridgestone, Goodyear, Dunlop and Michelin. All these tyre manufactures provide tyre rated loads, pressures and dimensions of all currently available aircraft tyres. From the
available data the radius of the tyres at static load can be determined for both nose and main gear. This determines the vertical position of the aircraft with respect to the ground. This position is needed for determining
the position of the shock strut at compressed and extended position and for determining the maximum rotation angle of the aircraft during take-off.

2.1. Landing gear components

21

2.1.4. Wheels
Wheels will have to be dimensioned such that there is enough room to house the brakes. Also the selected
tyre needs to fit on it. This should be done while keeping in mind that the weight should be minimal and the
life should be maximal.
Two types of designs for the wheel are available at the moment, namely the A-frame type and the bowltype wheels. The A-frame type can be made lighter, but has as disadvantage that there is less space available
to accommodate the brakes (figure 2.11). Thus if braking requirements are too high, bowl type wheels are the
only option.
Wheels are mainly constructed with forged aluminium. Trends to other materials such as magnesium
are not seen due to serious problems with corrosion. Steel has the problem of increased weight and forged
titanium the high cost.

2.1.5. Brakes
Brakes are used for stopping, turning, speed control and keeping the aircraft in parked position. Recently
new materials are introduced that have a lower weight and better material properties. One of these relatively
new brake materials is carbon.
Carbon has a high thermal conductivity and high specific heat giving a better more uniform distribution
of the heat. At high temperatures carbon keeps most of its specific strength contrary to steel. Additional
advantages are low maintenance, long service life (up to 5-6 times more landings than steel) and low weight.
Disadvantage is mainly the larger required volume to achieve the same amount of energy absorption.
Minor problems that have largely been resolved are the sudden loss of strength due to oxidation of the carbon,
temporarily loss of braking due to moisture and high initial cost. The economic advantages of carbon brakes
have been the reason why they have been used on the most recent large transport aircraft (Chai and Mason,
1997).
To approximate the brake size and get a weight of the brakes, the following conditions are being considered (Currey, 1988, p.140). 250 stops at design landing weight (with 10 ft/s2 deceleration), 5 stops at maximum
landing weight (with 19 ft/s2 deceleration) and a single rejected take-off stop (with 6 ft/s2 deceleration) at the
maximum take-off weight. To determine the total kinetic energy and the associated weight at landing the
power-off stall speed is needed. This power-off stall speed is 1.2 times the stall speed, which depends on the
reference wing area, the weight and maximum wing lift coefficient.
The kinetic energy is equal to:
KE =

MV 2
2

(2.17)

To find the kinetic energy in lbf ft the mass on each tyre M is in lbf and the velocity before applying the
brakes V is in ft/s. Currey provides a figure (Currey, 1988, p.142) from which a relation is contracted that
estimates the brake assembly weight (lbs) for a given kinetic energy at RTO:
2
Wbrake RTO = 9.90e 3K E to
+ 5.41K E to + 9.97e 1

(2.18)

for 5 stops at maximum landing weight:

2
Wbrake 5 stops = 2.99e 2K E max
landing + 8.46K E max landing 2.10

(2.19)

and for 250 stops at design landing weight:

2
Wbrake 250 stops = 1.12e 1K E landing
+ 16.7K E landing + 13.7

(2.20)

Weight
Maximum Static
Load Rating
22
Maximum Limit
Radial Load
Ultimate Loads

wheels

heat
s

41,780 pounds

Applied Ve
(Knots

2. Theory: landing gear design aspects

105,350 pounds

Radial

158,020

100

124.4

139.2

189.6

Combined
Radial

ervice

Structural Torque

Side

Condition No. 1

114,410

57,210

Condition No. 2

158,020

39,500

Burst Pressure

864 psi

41,780 radial load for 2,000 miles

41,780 radial load & 6,270 outboard
Roll
Conditions
side load for 100 miles
Figure 2.10: A schematic representation of the structure of a bias ply tyre on the left and of a radial tyre on the right (Goodyear, 2002)
41,780 radial load & 6,270 inboard
side load for 100 miles

nk

No. of
Stops

41,900 radial load at a speed of not

less than 10 mph for 2.83 miles

Roll On Rim

10.50
1.25

800

900

900ER

cle

1.25

5.25

20.92
DIA

4.75
DIA

2.705
4.3304
DIA
8.850

24.20
DIA

Figure 2.11: An A-frame type wheel cross-section. This wheel can be installed on the Boeing 737. (Honeywell, 2008)

555

2180

C61-0810-000-000
April 2008
2008 Honeywell International Inc.

M
O

2.1. Landing gear components

23

Rim diameter (in)

Outer diameter (in)

Inner diameter (in)

Volume per in width (in3 )

7.25

4.75

23.8

5.1

29.25

10

8.625

5.501

34.7

11

9.375

5.9

40

12

10

6.25

47.9

13

11

6.8

59

14

12

7.376

70.4

15

13

8.126

80.9

16

13.75

8.75

88.4

17

14.75

9.5

100.0

18

15.75

10.126

114.3

19

16.5

10.75

123.1

20

17.5

11.5

136.7

21

18.5

12.25

150.9

22

19.5

12.876

168.5

23

20.375

13.751

176.3

24

21.375

14.375

195.2

25

22.375

15.125

212.1

Table 2.5: Heatsink dimensions (Currey, 1988, p.143)

By averaging these weights the brake weight is found and the brake volume is then estimated with (Currey,
1988, p.142):
V = 3.3Wbrake 84.2

(2.21)

Dimensions of the heatsink can then be found by finding the closest match to the selected rim diameter
within table 2.5:
Carbon brakes sizing can be derived from steel sizing procedures using scaling factors of 1.28 and 0.40 for
the volume and weight respectively (Chai and Mason, 1997).

2.1.6. Brake actuation

Brake actuation systems are currently are mainly hydraulic brake systems. A recent development on the brake
actuation system is the development of electric brake systems. Currently electrically actuated brakes are
already in service on the Boeing 787 aircraft (see figure 2.12. Advantages of electric brakes includes (Goodrich,
2012):
reduced maintenance cost and brake system can be easily replaced;
higher reliability due to redundancy with multiple independent actuators installed on a single wheel;
system health and brake wear reported automatically.
The hydraulic or electric brake system is however not analysed by the landing gear analysis tool, since it
is part of a more detailed design and only the preliminary design is considered.

2.1.7. Kinematics
The design and analysis of landing gear parts relating to the retraction and extension of the gears is called
kinematics. Stowage of the landing gear has to be possible within the available space while the increased

24

2. Theory: landing gear design aspects

Figure 2.12: The electrically actuated carbon brakes developed by Goodrich and currently in operation on the Boeing 787 (Goodrich,
2012)

Figure 2.13: A typical retraction scheme of a wing mounted main landing gear on the left (Currey, 1988) and a fuselage mounted main
gear on the right (Torenbeek, 1982)

weight due to structural reinforcements is minimal. Goal is to make the retraction scheme as simple as possible (based on economic considerations) (Currey, 1988, p.175). The reduced number of parts and the maintenance cost will increase the total cost more than the increase in weight when more complexity is considered
(Chai and Mason, 1997). A requirement that may increase complexity and cost is the requirement to limit
the interference between the gear and the surrounding structure as much as possible. Also the gear must be
properly supported against side forces.
A retraction mechanism generally consists of a retraction actuator, a folding brace and a locking mechanism. The retraction of the gears positioned on the fuselage is most preferably done in the forward direction.
This is to make sure that the gear can lock manually by gravity and air drag in the event of a hydraulic failure. The wing-mounted gears are mainly retracted inboard, because of the limited space available behind the
wing spar to stow the gear. When retracting in the inboard direction the largest part of the gear, the bogie and
wheels can be stowed in the fuselage. Doors for the main gear wheels may be left out of the design if the drag
penalty is of less influence than the increased weight and retraction volume. This has for example been done
on the Boeing 737 main landing gear.

2.1. Landing gear components

25

Figure 2.14: Example of a good and poor actuator travel versus actuator load diagram (Roskam, 1989a)

There are many different retraction schemes possible. To keep the retraction scheme as simple as possible
each gear rotates about a single axis. To store the gear more efficiently it might also be necessary to rotate the
gear bogie about the bogie pivot point (see figure 2.13 right). On most forward retracting gears the shock strut
is shortened during retraction to minimise the stowed space. Also drag struts and side struts fold away during
retraction with a more complicated scheme than for the main gear. The drag/side strut mechanism and the
strut shortening mechanism are part of a more detailed analysis than the preliminary design. It is therefore
only included in the multi-body dynamics computer simulation.
For the automatic design of the landing gear an algorithm needs to be developed that can determine the
stowed position of the gear such that interferences with the wing/fuselage structure can be found. At the
same time the pivot axis of the gear has to be aligned such that the extension or retraction of the gear can be
done most effectively.
The actuator needs to be positioned such that the retracted actuator length is not smaller than one half
of the extended length. The efficiency of the actuator needs to be checked as well. The forces acting on the
gear during retraction are normally the aerodynamic drag and the gear weight. The aerodynamic drag can be
determined from the gear drag estimation method from Roskam. The gear weight can be determined from a
Class II weight estimation or a more sophisticated method.
Currey mentions that the geometric layout should be replaced with a mathematical analysis as soon as
possible and that the moment arms should be checked to be satisfactory throughout the retraction motion.
To check the retraction efficiency, actuator travel should be plotted versus actuator load. The efficiency is
then the area underneath the curve (which is the energy absorbed) divided by the product of the maximum
actuator load and the total actuator travel. An efficiency of 70 per cent is considered high. The efficiency
should be at least about 50 per cent (Torenbeek, 1982) and extreme variations in force should be avoided
during retraction as shown in figure 2.14. Generally a low efficiency is used to obtain simplicity or stow the
gear in a certain space. Only drawbacks of low efficiency are a longer retraction time or higher weight (Currey,
1988).
The time for extending the gear is limited to 15 seconds at temperatures higher than -29 C and to 30
seconds at temperatures between -54 and -29 C . Retraction time of the gear is limited to 10 seconds at all
temperatures.
All these checks could be performed by doing a handbook analysis as presented by Roskam, Currey and
Chai and Mason. But a simulation with SimMechanics can produce a more accurate result. SimMechanics is
therefore used to check the kinematics of the landing gear design.

2.1.8. Shock absorption

Shocks during landing and taxiing need to be absorbed by the landing gear and loads need to be reduced to
an acceptable level. Both the tyres and the shock absorber(s) take up most of the loads.

Fig. 5.7

26

Dowty sprung wheel.2. Theory: landing gear design aspects

AIR VALVE
UPPER CHA MBER
(AIR)
OIL
ORIFICE

ORIFICE SUPPOR
TUBE

UPPER BEARING
RECOIL VALVE
METERING PIN

RECOIL CHAMBE

SEAL
LOWER BEARIN G

y//P

I STON

..--AXLE

Fig. 5.8 Oleo-pneumatic shock absorber.

Figure 2.15: Working principle of an oleo-pneumatic shock absorber (Currey, 1988)

76
Shock absorbers can be constructed differently. They can be made as solid steel springs, rubber springs
or a fluid spring with gas and/or oil. An oleo-pneumatic fluid spring is the only type of shock absorber that is
considered. This is because of their widespread use on commercial transport aircraft and due to its relative
low weight and high gear efficiency (Currey, 1988).
Oleo pneumatic shock absorbers absorb the loads by forcing oil through an orifice to a chamber with dry
air or nitrogen. The area of the orifice is often controlled by a metering pin that has a varying radius as shown
in figure 2.15. In this way the strut load is kept relatively constant at dynamic loading (Currey, 1988). When
shock loads decrease the air pressure will press the oil back to the other chamber at a controlled rate.
The stroke of the absorber is an important design parameter. BothRoskam and Currey provide a method
to size the shock stroke. The method of Currey is more detailed and is therefore used and explained here.
In this method first the landing load factor is selected from a range of 0.7 to 1.5. The value of 1.2 is however
mostly used. Then using the required sink speed at landing (from FAA/EASA regulations) the energy absorbed
during touchdown is approximated. In this approximation the tyres absorb a part of the energy and the other
part is absorbed by the shock absorber. When the lift is assumed to be equal to the weight the shock stroke S
is equal to (Currey, 1988, p.84):

S=

V 2 /2g
N

S t t

(2.22)

Here V is the sink speed, N the landing load factor, S t the tyre deflection due to load N . The influence of
the tyre efficiency is not very high when the stroke is large and the shock efficiency is high. The tyre efficiency
t can therefore be fixed at a value of 0.47. The shock absorber efficiency n s is estimated between 80 and 90
per cent (Currey, 1988, p.35,77).
In the design of the shock absorber the stroke from static to fully compressed can be adjusted to the needs
of the designer. For an initial design the ratio between the static and extended pressure is set to P 1 /P 2 = 4/1.
The ratio between static and compressed pressure is set to P 1 /P 3 = 1/3. This is a typical value for transport

2.1. Landing gear components

27

aircraft. The pressure from fully compressed to static varies within the shock absorber according to Boyles
law:
P 1 V1 = P 3 V3 = const ant

(2.23)

The piston area is the static load divided by the static pressure (assumed 1500 psi). The change in volume
is the piston area times the total stroke length. Using these all pressures and volumes can be found at the
static, compressed and extended state. Pressures should be between 60 and 6000 psi (Currey, 1988, p.102).

shock force (lbs)

105

polytropic compression
from static position

2
isothermal compression
static
0.
0

10

12

14

16

18

stroke (in)
Figure 2.16: Load stroke curve for a Boeing 707-321 aircraft. From 0 to static is isothermal compression and from static to the right
polytropic compression.

In figure 2.16 the shock absorber load is plotted as a function of the stroke. This is a combined plot of an
isothermal part based on the equation
Px =

P 1 V1
Vx

(2.24)

and an polytropic part using the equation

P x = P1

V1n

(2.25)

Vxn

Assuming n = 1.35 for a shock with oil and gas separated during compression. Polytropic compression
only has a significant effect on the shock force when the shock stroke is large. Therefore polytropic compression is used from the static stroke position onwards.
Figure 2.16 also shows the load at the fully compressed stroke position. The g-force is then 3.9 for this
particular aircraft. For transport aircraft a g-force of 4 when it is fully compressed is suitable (Currey, 1988,
p.102).
The minimum allowable overlap between the shock piston and cylinder should be 2.75 times the piston
outer diameter. The minimum length of the cylinder should then be the stroke plus the minimum overlap.
The shock absorber orifice that restricts the flow of hydraulic fluid is sized with the following equation
A orifice =

0.3A
r

AS
W

(2.26)

Here A is the piston area, r the compressed load g-force and W the static load (Currey, 1988, p.119).

28

2. Theory: landing gear design aspects

air pressure

kp

cylinder

Aorifice

fluid pressure
cp

ffriction

piston

Figure 2.17: Model of the oleo-pneumatic shock absorber showing parameters used in equations.

A model of the oleo-pneumatic shock is a spring damper system. The pneumatic force of the shock can
be modelled as a nonlinear spring which produces a force of (Milwitzky and Cook, 1953):
(
Fp = P2 A

V2
V2 AX

)n
(2.27)

Where P 2 and V2 are the pneumatic pressure and volume at the fully extended state. This equation is derived using equation 2.25. The flow of oil through an orifice of the oleo-pneumatic shock produces a damping
force that depends on the velocity squared and is given by (Milwitzky and Cook, 1953):

Fh =

A 3h
2(C d A orifice )2

X X

(2.28)

is the density of the hydraulic fluid, A h = A A orifice is the hydraulic area and C d is the discharge coefficient.
The discharge coefficient can change during compression and is a function of fluid properties and the
orifice shape. The discharge coefficient can range between 0.6 and 1.0. A different discharge coefficient can
result in a difference of maximum displacements of 20 per cent (Milwitzky and Cook, 1953, p21). The orifice
and fluid properties are part of a more detailed design of the shock absorber. Since there is only an initial
design considered the discharge coefficient is set at a fixed value of 0.8.
The friction within the shock absorber will result in an additional force:
F f = (F p + F h )

(2.29)

A pessimistic value of the coefficient of friction, is 0.1 (Currey, 1988, p.99). Then the total shock absorber
force is equal to
F = F p + Fh + F f

(2.30)

2.1.9. Economics of landing gear design

As mentioned in the previous sections, cost are an important factor in the landing gear design decisions that
are made. Costs are to be considered for the design process, manufacturing, maintenance and disposal at the
end of life.

2.1. Landing gear components

29

Figure 2.18: Average fuel price, from 1980 to 2008, paid by US airlines using current dollar cents and using 1987 dollar cents (Doganis,
2009)

The total cost of the gear as part of the aircraft when the aircraft will be delivered to the customer has to
take into account many aspects. Some of these aspects include:
development;
materials;
production processes;
certification;
marketing;
overhaul;
refurbishment;
spares.
A study by Chai and Mason estimates the total gear program cost to be $10 to $12 million dollars. Relatively fixed costs are the cost of tyres, wheels and brakes. A bias ply main gear tyre for the Boeing 747 is priced
at $2100, while a comparable radial tyre for a Boeing 777 main gear is $2900. The more expensive radial tyre
is however still mainly chosen by airlines for new types of aircraft, because of a longer service life and lower
weight. Wheels and tyres are replaced after 300 landings, making the cost for the wheel and tyre $5 per landing. Carbon brake replacement on the 747 is done after 1200 to 1500 landings (results in bake cost of $10 per
landing).
Development of new larger tyres is costly, due to the cost of new manufacturing and testing equipment
that will be necessary. The maximum size of tyres currently used is therefore limiting. For bias-ply tyres the
maximum diameter is 56 inch and for radial tyres 58 inch (Chai and Mason, 1997).
Maintenance cost form an important part of the total operating cost: about 10% in the recent years (Doganis, 2009). The time between overhaul of the landing gear varies between 33000 and 42000 flight hours.
Overhaul of the landing gear is preferably done on the complete gear to minimise downtime. Cost of this
overhaul for a Boeing 747 type aircraft gear is estimated at $400000.
Weight is an important factor in cost calculations. When the aircraft airframe weight can be reduced with
1 per cent, the fuel consumption reduces with 0.25% for small aircraft (e.g. B737) to 0.50% for large aircraft
(e.g. B747) (Greene, 1992).

2. Theory: landing gear design aspects

c

z
x

Fside

Ckx

Fdrag
Fky
Chx

Fkx
k

Clx
Flx

Flz
m

kb

ea

Cix

tru
c

Fhx

Fly

Fkz
f

Fhy

piston-cylinder

tr
un

ni
on

30

Fix

Fiy

wb

h
g

axle

Fiz

Fhz
tt

Figure 2.19: Structural model of a 4 wheel landing gear bogie used in the weight estimation. All externally applied loads are shown.

The part that fuel prices take up of the total operating cost has increased considerably the last 2 decades.
In 1994 the fuel and oil cost were 11.4 per cent of the total operating cost, while this has increased to 25.4
per cent in 2007 (Doganis, 2009). Figure 2.18 shows the recent steep fuel price increase. Prices are not expected to decrease, due to factors as larger demand, lower reserves and the use of oil production quotas by
oil producing countries. Further reductions in the weight of the landing gear is thus becoming increasingly
important.
A further more detailed analysis of the impact of cost on the landing gear design is difficult to make. This
is because there is very limited cost information available for aircraft systems, including the landing gear.
Manufactures are not willing to present this information due to competitive concerns.

2.2. Landing gear assembly weight estimation

Estimating the landing gear total weight based on empirical relations cannot always produce accurate estimates. Also they do not respond to changes in landing gear design variables, making optimisation to minimum weight impossible (Chai and Mason, 1996, p. 72). A landing gear weight can be estimated analytically
by modelling the landing gears as simple geometric shapes. Maximum stresses within the structure, due
to dynamic and static loads, are calculated with simplified analytical equations and assuming an idealised
structure. The dynamic and static loads, described in section 2.2.3, are the loads that are considered.
Figure 2.19 shows the three dimensional structural model used for the analytical weight estimation. In
appendix A the complete landing gear structural model is shown, including applied and internal forces and
moments for each component separately.

2.2.1. Tube stresses

The landing gear structure is constructed out of circular tubes. The structure is sized with fixed radii (scaled
to the piston area) and by varying the thicknesses of the tubes. A beam can be accurately modelled as a
thin walled structure if the maximum diameter divided by the thickness is larger than 20 (Hibbeler). The
thicknesses required are expected to be larger than one tenth times the mean radius of the tubes. A thin-

2.2. Landing gear assembly weight estimation

31

Mz
Sz
z

ri

ro

Sy

My

Figure 2.20: Tube cross-section showing the parameters used in the derivation of the stresses

walled approximation is therefore not valid and not used.

Figure 2.20 shows the used parameters in the following derivation graphically. The geometric properties
of the tubes are:
A = (r o2 r i2 )
J = /2(r o4 r i4 )
I y y = I zz = J /2 = /4(r o4 r i4 )
The normal stress in a tube due to the combination of an axial force and bending moments is
xx = N /A + M y /I y y z M z /I zz y

(2.31)

The maximum normal stress occurs at the outer edges

xxmax = N /A +

M y r o sin() M z r o cos()
/4(r o4 r i4 )

Differentiate with respect to and find the maximum at d xx /d = 0

xxmax
dx

M y r o cos() + M z r o sin()
/4(r o4 r i4 )

solving for results in

tan(max ) =

=0

M y
Mz

(2.32)

which can be rewritten using sin()2 = (1 cos()2 )

cos(max ) =

Mz

(2.33)

(M y )2 + M z2 )

and using cos()2 = (1 sin()2 )

sin(max ) =

M y

(2.34)

M z2 + (M y )2

Using these equations the maximum normal stress in the tube is at = 0 and =

xxmax =

N
(r o2 r i2 )

r o

M z2 + (M y )2

/4(r o4 r i4 )

(2.35)

32

2. Theory: landing gear design aspects

Then the shear stress in the tube due to an axial torque
Tr o
Tr o
=
J
/2(r o4 r i4 )

xsmax =

(2.36)

The first moment of area which is needed to get the shear stresses due to shear forces is

Qy =

z dA =

ro

/2
/2+

ri

z r ddr =

ro

/2
/2+

ri

r sin( + /2) r ddr

2
cos()(r o3 r i3 )
3
and the first moment of area needed for the bending about the y-axis is
=

Qz =

y dA =

ro

y r ddr =

ri

ro

ri

r cos() r ddr

2
sin()(r o3 r i3 )
3
Shear stress due to shear forces in the y and z direction is
=

xs =

S y Q z
I zz t

SzQy

(2.37)

Iyy t

Filling in the equations for the first moment of area results in

xs =

2
(S y (sin()(r o3 r i3 )) + S z (cos()(r o3 r i3 )))
3I zz t
=

2(r o3 r i3 )
3/4(r o4 r i4 )t

The maximum shear stress occurs when

dxs
d

(S z cos() S y sin())

=0

2(r o3 r i3 )
dxs
=
(S z sin() S y cos()) = 0
d
3/4(r o4 r i4 )t
which results in (similarly as for xxmax )
tan(max ) =

sin(max ) =

Sy
Sz
Sy

S 2y + S 2z

cos(max ) =

Sz
S 2y + S 2z

Using these equations in the equation for shear stress

xsmax =

8(r o3 r i3 )
3t (r o4 r i4 )

S 2y + S 2z

(2.38)

The total maximum shear stress due to torque and shear forces is then using equation 2.36 and 2.38

xsmax =

2
(r o4 r i4 )

(Tr o

4(r o3 r i3 )( S 2y + S 2z )
3t

(2.39)

2.2. Landing gear assembly weight estimation

2

33
von Mises yield
maximum shear stress theory

pure shear

Figure 2.21: Comparison Von Mises yield criterion with the maximum shear stress theory. The difference is largest at pure shear.

The maximum shear stress theory predicts when yielding of the material begins. The maximum absolute
shear stress to prevent yielding should be (Vable, 2012, p.486):
max =

Y
2

(2.40)

The maximum-distortion-energy theory sometimes also called the Von Mises yield criterion gives the
maximum stress (Vable, 2012, p.487):
21 1 2 + 22 = 2Y

(2.41)

+ R and 2 = 2xx R are the two principal stresses in Mohrs circle of radius R. Rewriting

xx y y 2
this equation and using R = (
) + 2xs and equation 2.41 results in
2
Where 1 =

xx
2

2xx + 32xs = Y

(2.42)

The Von Mises yield criterion is used to determine limit loads, because it gives a more accurate prediction
of the yield load than the maximum shear stress criterion. Actual load tests have shown that a maximum
improvement in accuracy of 15 percent can be achieved by using the Von Mises yield criterion (Hibbeler,
p.527). This happens when the structure is loaded in pure shear. This is also made visible in figure 2.21. The
eliptical curve represents the Von Mises Yield criterian. If the stress is outside the boundary of the curve the
material will fail.
Since the structure of the landing gear is based on the safe live principle an addition safety factor is applied. By adding a safety factor the structure is capable to deal with fatigue during the entire life of the aircraft
structure.

2.2.2. Side struts sizing

Both the drag strut and the side strut can be simplified as a truss member. A truss is a two force member that
is only loaded axially. When the truss is loaded axially in tension the required thickness to prevent yielding
can be found from the yield criterion: Y = F /A. The side struts cross-section can be modelled as a thinwalled I-beam with a constant thickness making the area, A equal to t (2w + h) and moment of inertia I =
t (h/12 + w/2)h 2 .
When the truss is loaded in compression yielding is not longer the limit. The truss will buckle before
yielding. The critical buckling load is equal to (Vable, 2012, p.503):
F cr =

2 E I
l2

(2.43)

34

2. Theory: landing gear design aspects

2.2.3. Landing gear dynamics analysis load cases

Several different limit load cases have been specified in the Certification Specifications Part 25. Loads are
applied externally and the effect of the centre of gravity position has to be taken into account. The most
critical position of the cg has to be considered. The maximum descent velocity to be considered is 10 feet per
second. Then the analysis of the loads on the landing gear also has to take into account (European Aviation
Safety Agency, 2012, p. 1-C-16):
dynamic characteristics of the gear
spring back and spin-up
the rigid body response
if significant: the airframe structural dynamic response
There are 7 different load cases that need to be considered. These are shown graphically in figure 2.22 to
2.26. In these figures T is the forward component of the inertia force and I is the pitch and roll moment of
inertia that are needed for equilibrium. The subscript N and M stand for the nose gear and for the main gear.
1. The first load case is a landing at level attitude at the limit descent velocity of 10 fps at design landing
weight. For a coefficient of friction between the ground and tyres a value of 0.8 is sufficient. This means
that the horizontal force D M on the main gear is 0.8 times the vertical force VM and a similar horizontal
load is applied on the nose gear.
2. The second load case is the tail down landing case, which is the same as the first only now the attitude
of the aircraft is at the maximum possible angle of attack. The tail structure then hits the ground or the
stall angle will be attained at this position.
3. Then the third load case is a landing on a single wheel. The ground forces on the gear are the same as
for the first load case as shown in figure 2.23.
4. For the fourth load case, the lateral drift landing, only the main gear is in contact with the ground. A
side load in the inward direction has a magnitude of 0.8 times the vertical reaction on one side (VM )
and on the other side a load of 0.6 times VM acts outwards. These forces are assumed to be resisted by
inertial forces and moments as is visible in figure 2.24.
5. A braked roll at design take-off weight (and at 1.2 times the design landing weight) is the fifth load case.
A drag force is applied to each wheel of 0.8 times the vertical ground load on the wheel.
6. Then the sixth load case is a static loading condition: the ground turning load case. The side load on
each wheel is 0.5 times the vertical ground load on each wheel. These forces are counteracted by a side
load of 0.5 times the weight acting at the location of the cg (figure 2.25).
7. The final load case is a pivoting load case. On one side of the gear the brakes are applied with a coefficient of friction of 0.8. The aircraft is for this case at static loading conditions as can be seen in figure
2.26.
There are more load cases mentioned in CS 25 (European Aviation Safety Agency, 2012, p. 1-C-16), but the
previously mentioned ones are the most significant. For example a rebound landing, a towing load case and
a reversed braking load case are described.

2.2. Landing gear assembly weight estimation

35

Annex to ED Decision 2012/008/R

Annex to ED Decision 2012/008/R

CS25
Figure 2.22: Level load case on the left and the tail down load
caseBOOK
on the 1right (European Aviation Safety Agency, 2012, p. 1-App A-2)
CS25 BOOK 1
Appendix A (continued)
Appendix A (continued)

Figure 2.23: One wheel landing load case (European Aviation Safety Agency, 2012, p. 1-App A-3)

Figure 2.24: The lateral drift landing load case (European Aviation Safety Agency, 2012, p. 1-App A-3)

1App A3

1App A3

Amendment 12

Amendment 12

CS25 BOOK 1

36

Appendix A (continued)

2. Theory: landing gear design aspects

Figure 2.25: Braked roll load case on the left and the ground turning load case on the right (European Aviation Safety Agency, 2012,
p. 1-App A-4)

Figure 2.26: The pivoting load case (European Aviation Safety Agency, 2012, p. 1-App A-5)

2.3. Landing gear analysis

The landing gear can be seen as a system of interconnected rigid and deformable components, a multi-body
system. The dynamics of this system is a complex problem, because its highly nonlinear. In most cases the
problem can only be solved with computer-based techniques (Shabana, 2005).
Matlab SimMechanics can model the multi-body dynamics of the landing gear model. The multi-body
system is modelled with blocks that represent bodies, joints, constraints and force elements. Using provided
input parameters SimMechanics then evaluates the equations of motion of the complete mechanical system
and tests the landing gear for different load cases.

Figure 2.27: SimMechanics displaying a portion of the model of a landing gear (Mathworks, 2012)

1App A5

Amendment 12

2.3. Landing gear analysis

37

l
ee
wh

ne
pla

Vx

Ro

nwhee l
Fz Re
Mz
.

Vsy
road p
lane

nroad

Vcy
Fy, My

wheel
spin axis

Fx, Mx
Vsx

Figure 2.28: Tyre model description (Besselink, 2000)

2.3.1. Tyre model description

Simulating the dynamics of the landing gear can only be done accurately when using an accurate tyre model.
The widely used tyre model of Pacejka, an empirical model called Magic Formula Tyre Model, can provide
this accuracy.
MF-Tyre, made by TNO Automotive, is a software implementation of the tyre model of Pacejka. This tool
is mainly focussed on automotive tyres, but new tyre data can be inserted such that aircraft tyres can also be
modelled.
A description of the equations used in the MF-tyre implementation have been made available (Pacejka,
2006) (TNO Automotive, 2010), such that these can be used in the implementation. Output forces are forces
in x, y direction F x , F y and moments M x , M y , M z at the tyre contact point.
Input variables of the tyre model are the tyre slip angle , the turn slip angle t , the forward velocity Vcx ,
the tyre inclination angle and the longitudinal slip . The tyre slip angle is equal to tan = Vc y /Vx . A turn
slip angle occurs when the wheel is on a circular path with radius R c and is equal to t = /Vx = 1/R c 2 (as

e
shown in figure 2.29). The tyre longitudinal slip is defined as = Vx R
, where is the radial velocity of
Vx

the tyre.
In vertical direction the tyre vertical is modelled as a linear spring damper with stiffness K tyre and damping
coefficient C tyre (see figure 2.31). The vertical force the tyre produces is then equal to:
F z = K tyre X C tyre X

(2.44)

TNO Automotive provides basic tyre model parameters for an aircraft tyre (H40x14), that is similar to an
Airbus A320 tyre. The tyre data of this tyre is of the reduced format, because there is limited data available.
Because the input data is of reduced format, estimation of the tyre response is less accurate. The unknown
parameters are given the default value. The parameters included are the tyre dimension, nominal load, vertical stiffness, vertical damping, friction coefficients, slip stiffness coefficients and relaxation length coeffi-

38

2. Theory: landing gear design aspects

Fy7

Fx

Fy
F F

F.

RO.8

Vx

Fy

F
F, A-

FF 1

SMCONTGN

N (POSmV DU1ECTlONS)

Mz
2.3
Presentation
of the results on main and nose landing gear
Mz,
Rc
Fy
-
2.3.1 General

The landing gear loads have been measured by st- ., gages calibrated and combined to produce
pure
Fy
vertical, lateral and longitudinal loads using Skopinski's method J 13, 14, 19 ] cn noseFand
L/H main gear.
z
A typical example of the measured ",ad time histories during landing and taxiing is shown on Fig. 9. The
top view
rear viewby means of the
maximum values during landing 1.ave
been marked. These maximum loads have been analyzed
extreme value distribution ' 4 1, the loads during taxiing by means of the statistical counting method level
Figure 2.29: Tyre
coordinate system used
crossings[
25model
1.

MAIN GEAR

__

01F,

F..

F,

Figure 2.30: Measurement results showing

of a typical
during
FIG,9the time
LOADhistories
TIE HISTORIES
OFloading
THE LOADS
ONlanding
THE for an Airbus A300B2 main and nose
MIN
NOSE
LANDMGGEAR
landing gear (Ladda and Struck, 1991). The y-axis
in theAND
graphs
represents
the force [121
in x-, y- and z-direction from top to bottom.

cients both longitudinal and lateral. Rolling resistance and aligning coefficients are also included. The highly
recommended
2.3.2
Landing

parameters (TNO Automotive, 2012) are thus available within the aircraft tyre data file and it

can thus be used to model the combination of lateral and longitudinal slip, tyre relaxation length effects and

The acceleration of the landing gear wheels to a circumferential speed which corresponds to the aircraft
response due
to short
wavelength
obstacles. drag load, the so-called spin-up load. Caused by the energy
horizontal
landing
speed
causes a longitudinal
built up in the main gear this leads to a so-called spring back load in form of an attenuated oscillation. In
reason of the elasticity of tl' landing gears side loads occures during landing. Due to the vertical sinking
speed of the A/C at the moment of touch down a landing impact in vertical direction comes off. Distributions
2.3.2.
Drop speed
test load
case types of A/C are shown on Fig. 10, from which the vertical load can be
of
the sinking
for different
derived.
In figure 2.30 it can be seen that a typical landing impact load on the main gear is only about 40% of the static

(Ladda
and Struck,operational
1991). At adata
normal
touchdown
procedure
the aircraft
load
(Christy,
The
hatched2009),
scatter
band represents
for transport
aircraft
1 26 1only
and part
is in of
agreement
witl. ( 27 1. The distribution marked by 1 depicts the spectra predicted for A320 as well as DC1O [ 26, 28 1.
weight
is
put
on
the
ground.
This
is
because
the
wings
still
provide
a
large
lift
force
at
touchdown.
When
Curve 2 results from an A320 landing gear investigation performed during flight tests, it lies in the upper
region
of
the
scatter
band
but
under
the
predicted
one.
Curves
3
to
5
show
landing
sinking
spectra
speed is decreased during landing rollout, the lift is decreased gradually until all of the totalspeed
aircraft
weight
for different military aircraft f 6 J.
presses on the landing gear.

The CS-25 certification specifications specify that during the landing the lift provided by the wings is equal
to the landing weight. The dynamic landing load cases are at a descent velocity of 10 ft/s. For the dynamic
touchdown the vertical ground load factor per shock strut N v can be estimated with (Kraus, 1970):
Nv =

)
( 2
Vs
Fv
1
=
+ S cos
cW
s S cos g

(2.45)

Here c is the load distribution factor, s the shock efficiency, S the shock stroke (in), Vs the descent velocity
(ft/s), g the gravitational acceleration (32.17 ft/s2 ), the pitch angle at touchdown. This vertical ground load
factor is applicable locally at a single landing gear strut (see figure 2.31). At the aircraft centre of gravity a
different load factor applies. Assuming the lift is equal to weight the load factor at the aircraft cg is equal to
(Currey, 1988, p.34):

Ncg =

F ext F v + L
=
= Nv + 1
W
W

(2.46)

2.3. Landing gear analysis

39

aircraft with weight W

Ncg
Kshock

Ktyre

Cshock

Ctyre

c W Nv
Figure 2.31: Tyre and shock absorber modelling. The tyre is modelled as a linear spring damper and the oleo-pneumatic shock as a
nonlinear spring damper.

Tyre spin-up forces should also be taken into account for the drop test load case (European Aviation Safety
Agency, 2012, p. 1-C-16). The maximum spin-up force that will occur is when the tyres are at rest before
touchdown. The difference in speed between the ground and the tyre is the landing velocity. The tyre will
spin-up from zero to landing velocity in a very short amount of time, putting a high load on the tyres.

3
Implementation and use cases
The implementation is split up in a landing gear design part and a landing gear analysis part. Implementation
starts with designing the landing gear.

3.1. Implementing the landing gear design

The landing gear design is completely integrated into the TU Delft Aircraft Initiator, which is built using Matlab (Elmendorp, 2014). Using specified top level requirements, the Initiator can generate an aircraft parametric model. Which includes geometry and performance characteristics of a conventional or unconventional
aircraft. The initiator is built-up out of several separate modules. In this way the user can generate an aircraft
using only the required modules. Landing gear design is done in a new module called PositionLandingGear
and in the class 2 weight estimation module (which is extended with basic landing gear design methods).

aircraft
input file

settings
file

Initiator
start

preliminary sizing
(class 1
methods)

database

class 2 weight
estimation

aerdynamic
analysis

Position landing
gear

class 2.5 weight

estimation

no
end
yes

weight and range

estimation converged?

performance
estimation

Figure 3.1: Overview of the workflow to make an aircraft class 2.5 weight estimation

The workflow needed to make a class 2.5 weight estimation of the aircraft is shown in figure 3.1. It starts
with an aircraft input file and settings file that defines all requirements set by the user. Using these, similar
reference aircraft are selected from a database. These are used to make a first estimate of the weight, performance and general geometry. Then a closer look is taken at the design of the cabin and a more detailed class
41

42

3. Implementation and use cases

2 weight estimate of aircraft components can be made.

Now there is enough information available to make a class 2.5 component weight estimation which runs
several modules iteratively. These modules include control allocation, aerodynamics estimation (AVLVLM),
wing weight estimation (EMWETWeight), drag estimation, fuselage weight estimation (Schmidt, 2013) and
finally landing gear weight estimation (position landing gear). The iteration stops when the estimation of the
aircraft weight and range has converged.

3.1.1. Class 2 weight estimation

A class 2 weight estimation (CWE2) is a method that uses empirical relations to improve the class 1 weight estimation. The CWE 2 module implemented in the Initiator uses the empirical relations of Raymer to estimate
aircraft component weights. One of these components is the landing gear weight.
The class 2 weight estimation methods require little computation time and are applicable to both conventional and unconventional aircraft. Raymers empirical gear weight estimation relation only requires 10
different parameters for both the nose and main gear. Therefore simplified methods can be used to find the
location of the gears. The results of the class 2 weight estimation is thus not a detailed analysis of the landing
gear layout and position. For this purpose a separate module has been developed.
Class 2 landing gear weight estimation process
The flow of work performed by the CWE2 landing gear weight estimation is shown in figure 3.2. The gear
component weight estimation is done last in the class 2 weight estimation. This is done because then the most
accurate position of the aircraft centre of gravity is known by combining all aircraft component cg positions.

preliminary sizing

class 2 weight estimation

class 2 landing gear weight estimation

aircraft
geometry
position nose
gear

statistical
estimate of #
tyres and size

main gear
longitudinal
positioning

wing and engine

clearance reqs

get gear height

using scrape
angle

determine
tailbumper
position

sideways
turnover limit

gear length
estimate

main gear lateral

positioning

settings
file

create gear parts

Figure 3.2: Class 2 landing gear weight estimation workflow

The nose gear is assumed to consist of 2 tyres and it is placed at a fixed position based on nose gear
loading settings. It is also checked if nose gear loading will add limitations to aircraft operations. This will be
explained in section 3.1.1.

3.1. Implementing the landing gear design

43

The number of wheels required for the main landing gear are determined from a statistical relationship
between the maximum take-off weight and the number of main gear wheels. All gear tyre dimensions are
found using a relation from Raymer (Raymer, 1999, p.234). When the number of wheels is larger than 12, the
number of main gear struts is set to 4. The track of the main gear bogie is fixed to 2.5 times the tyre width.
This is based on reference data from reference aircraft shown listed in appendix D.
By checking the fuselage width to span ratio it can be determined if all main gears can be placed on the
fuselage. For gears that can be placed on the fuselage, the longitudinal position of the main gears is only
related to the aft centre of gravity position (independent of the wings). This also holds for aircraft that have a
high wing and aircraft that have their main wing positioned far forward. Placement of the main landing gears
under the nacelle of a turboprop is currently not taken into account.
If the wing is positioned far aft of the fuselage (canard), the gear can be placed on the wings. The longitudinal position of the main gear is set near the front spar position. This is done by setting the gear to a certain
percentage of the mean aerodynamic chord depending on the forward or aft sweep of the wing. In this way
the exact spar locations do not need to be calculated and complexity is reduced. For a conventional aircraft
all these exceptions do not hold and the main gear can be positioned on the wing near the auxiliary spar. The
MAC percentages used for the longitudinal position is 0.55 for an aft swept wing (Currey, 1988) or 0.88 for a
forward swept wing. For the forward swept wing the MAC percentage is higher, because then a point inboard
the MAC on the wing is further aft.
The point where the fuselage would hit the ground at take-off (a too large rotation) is called the tail
bumper position (see figure 3.3). From the geometry of the fuselage xz-plane cross-section a third order
polynomial is constructed. The location of the tail bumper is found when the derivative of the polynomial
equation is equal to the fixed scrape angle of 12 degrees (to keep dependency on other modules and calculation time minimal). Only when the aft fuselage is not steep enough the tail bumper is positioned at the end
of the fuselage.

0.55 MAC
tail bumper position
LOF
gear height

takeoff pitch
static ground line
maximum tyre and shock displacement

Figure 3.3: Side view showing take-off rotation limit and definition of main gear height.

The main gear height (as defined in figure 3.3) is the vertical distance of the line from the tail bumper
forward towards the longitudinal position of the main gear. The clearance between the tyres and fuselage
or other aircraft parts should be minimally 6 inch. Otherwise the main gear height is increased (Torenbeek,
1982).
The minimum lateral position of the main landing gear is found from engine clearance, wing clearance
and the sideways turnover limit requirements. The engine clearance is found by drawing a line at 5 degrees
(plus 6 inch clearance) from the lowest point of the engines to the main gear position. Similarly for wing
clearance a line of 5 degrees plus clearance is drawn from the wing tip. Turnover of the aircraft can occur
along the line from the most outboard wheel towards nose wheel, due to wind shear or a high speed turn
during taxi. This turnover angle should not be larger than 63 degrees (Currey, 1988) as explained in section

44

3. Implementation and use cases

2.1.2. The maximum of the engine, wing and turnover limits is the minimum lateral gear position.
Another check is done to see whether or not the gears collide when retracted. This will put an additional
constraint on the lateral gear position. To do this, the complete length of the main gear needs to be known.
In the case of a wing mounted main gear this is the length from the ground to the wing connection point. In
the case of a fuselage mounted gear the gear length is the distance from the ground to the passenger floor.
In case of a high wing with fuselage mounted main gears it is assumed these are placed in pods at the most
outboard part of the fuselage. The gear length for this case is estimated as the length from the ground to the
fuselage belly.
At this point the location of the main gear and the general dimensions are known. The resulting landing
gear is then added to the aircraft model. A three dimensional plot can now be made where the gear position
and dimensions are represented as tyres.
Finally the class 2 weight estimate equation for the nose gear and main gear is evaluated. This results in a
mass for the main gear and the nose gear groups. The centre of gravity position is set at the centre of the gear
at a height one third of the gear length from the ground.
Class 2 landing gear estimation assumptions
Some of the assumptions made in the CWE2 gear weight estimation have been mentioned in the previous
section. The most important CWE2 landing gear sizing assumptions are listed below. When a specific value
or percentage is used it is based on different conventional reference transport aircraft, listed in appendix D.
Only the tricycle gear layout is supported;
If the number of wheels is larger than 12 it is assumed that the number of main gear struts is 4 (3 main
gears are not considered);
The nose gear is assumed to consist of 2 wheels and placed at 25 percent of the prescribed nose gear
stowage space (from settings or the input file);
The aft cg position is estimated to be at the mean cg position plus 10 percent of the mean aerodynamic
chord (MAC);
Specific rear or forward spar locations are not taken into account;
The maximum take-off scrape angle is assumed 12 degrees;
In case of wing mounted gears the longitudinal position is set at 55 percent of the MAC for aft swept
wings and at 85 percent for forward swept wings;
In case of a wing mounted very far aft of the fuselage (aft of 60% of the fuselage length) the longitudinal
position is assumed on 0% of the MAC for aft swept wings and 40% for forward swept wings;
In case of a high wing or a wing far forward (forward of 30% of the fuselage) the main gears are assumed
to be placed laterally on the fuselage most outboard position;
If the main gears are positioned on the fuselage the longitudinal position is estimated as the aft cg
position plus 4% of the fuselage length;
Propeller clearances are not taken into account;
Kneeling gears are not considered, because these are rarely applied to aircraft

3.1. Implementing the landing gear design

.
.
.
.
.
.
105

45

.loading maximum fuel mass

.loading according to mission requirements
.forward ops bound, max nose gear load
.aft ops bound, min nose gear load
.actual forward cg position
.actual aft cg position

fuel

0.8
cargo

mass (kg)

0.9

0.7
pax

0.6
.
0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

xpos w.r.t. MAC /MAC

Figure 3.4: B707 loading diagram produced by the class 2 weight estimation module. Aft and forward operational cg bounds are shown
as vertical lines.

Aircraft centre of gravity operational bounds

A class 2 weight estimation is usually done on an aircraft concept from which very little is known about the
mass distribution of the aircraft. The CWE2 module calculates the range of centre of gravities that could occur
during aircraft operations. A sample result for the Boeing 707-321 aircraft is displayed in figure 3.4.
The centre of gravity range that is produced by the class 2 weight estimation module seems too large. This
is also visible by the operational bounds that need to be applied to the cg range. Operational bounds are
calculated using the method described in section 2.1.2 of chapter 2. Operational bounds on the cg are not
desirable, because they limit airplane operations. The forward operational bound prevents that the nose gear
load will be too high. The aft operational bound will prevent that the nose gear load becomes too low to steer
the aircraft adequately.
The Boeing 707 airport manual (Boe, 2011) provides the actual cg positions of the aircraft: the most forward cg is at 14% and the most aft cg at 23% of the mean aerodynamic chord (MAC). Clearly the centre of
gravity distribution of the aircraft needs to be changed in order to improve the stability and control of the
aircraft and to improve landing gear positioning.
This problem occurs on preliminary sizing of some aircraft. Adjustments on the centre of gravity range of
the aircrafts will solve this, but is currently not done. Future work on the TU Delft Initiator will address this
problem.

3.1.2. Landing gear design module

The landing gear design module called PositionLandingGear has the capability to position, size and estimate
the weight of each landing gear. The module is composed of 3 parts: a gear positioning, a bogie design and

46

3. Implementation and use cases

weight estimation part as is shown in figure 3.5. Each part of the module runs sequentially.

aircraft
geometry

settings
file

PositionLandingGear
find feasible gear
locations

find feasible
bogie layouts

tyre +
wheel
data

estimate landing
gear weight

Figure 3.5: Components of the landing gear design module

Gear positioning
The gear positioning part of the PositionLandingGear module evaluates a design space of all possible landing
gear heights versus take-off scrape angles. Figure 3.6 shows this design space bounded by positioning limits.
Positioning starts with the parameters that do not need to be evaluated for every design point. This includes the available stowage space and the main gear longitudinal position limits that are imposed by maximum and minimum nose gear loads.
The positioning steps in the PositionLandingGear module are displayed in figure 3.7. For each design
point the take-off and landing angle is calculated. This also requires the calculation of the fuselage tail
bumper location. The wing spar limits are estimated as well as main gear longitudinal limits from the maximum and minimum nose gear loads.
Currently the take-off angle is estimated empirically. If this angle can be calculated more accurately the
module can be easily changed to use this more precise angle. When calculating gear positions, shock absorbers are at the extended position and tyres are at the undeformed state.
The nose gear lateral position is now also set from nose gear loading requirements (set in the input or
settings file). The lateral position of the main landing gear is found from the wing and engine clearance, the
sideways turnover limit, the stowage requirement and the limit preventing colliding gears when retracted.
When the actual lateral gear position is known the position of the spars is also known.
The ability to turn on a runway and taxi requirements are also calculated. From all feasible landing gear
design points the shortest possible is chosen. All methods used for the positioning of the landing gear are
further explained in section 2.1.2 of chapter 2.
Bogie design
When the position of all landing gear bogies are known the static load is known for a forward and aft cg
position. The aft cg position is most critical for the main gear and the forward for the nose gear. Tyres transfer
the loads of the struts to the ground. There are many different bogie layouts with different number of tyres
per bogie that can be made. But only the 8 different bogie types of figure 3.8 are considered.
For each of these bogie types the load per tyre is calculated. A database of 342 different tyres with accompanying wheels has been made using data from Goodyear (Goodyear, 2002). The tyre with a suitable rated

3.1. Implementing the landing gear design

47

Lateral positioning limits

2.5

M
ain
3

ge
ar

sto
wa
g

3.5

ake
own/T

4.5

Feasible design space

off limit

5
5

10

11

im

it

im
Re um
n
ar
/ a ose
ge
ux
ar
ili
ar
lo
y
ad
sp
lim
ar
it
lim
its

Touchd

main landing gear height

el

M
ax

12

13

14

15

rotation angle

Figure 3.6: Graph showing the feasible design space of rotation angle versus height.

class 2 weight
estimation

gear positioning
calculate
available
stowage space

main gear lon.

limits due to req.
nose gear loads
for each design point evaluate:

wing and engine

clearance reqs

set nose gear

position

set main gear lon

position

takeoff and
landing reqs

turnover limit

general stowage
reqs

prevent colliding
gears

set main gear lat

position

find lowest
feasible

bogie design

Figure 3.7: Workflow diagram of the gear positioning part of the PositionLandingGear module

48

3. Implementation and use cases

1 single

2 dual

4 triple

5 dual tandem

3 tandem

6 dual twin

8 dual twin tandem

7 tri-twin tandem

Figure 3.8: The bogie layouts that are being evaluated by the PostionLandingGear module

load is selected. This rated load includes a tyre safety factor (1.25 by default) such that for future growth of the
aircraft design the tyres do not need to be changed (Currey, 1988). If the bogie type is capable of sustaining
the static loads, the bogie design is refined by estimating tyre positions and tyre clearances. This is done using
the dimensions and relations of Torenbeek (Torenbeek, 1982, p.382).
Pavement requirements for airports with flexible and rigid pavements are then evaluated by running an
external module. This module is called FlotationAnalysis and is derived from the source code of the Federal
Aviation Administration COMFAA 3.0 analysis tool (Federal Aviation Administration, 2011). The code has
been rewritten from Fortran to Matlab to make sure that flotation analysis is completely integrated into the
Initiator. Flotation analysis results in a so called Aircraft Classification Number (ACN) as explained in chapter
2, section 2.1.1. Calculation of the ACN requires many iterations and thus has a large impact on calculation
time. The Matlab JIT-accelerator speeds up the code by converting it into a more efficient programming
language. Therefore the flotation module has been optimised for the Matlab JIT-accelerator, such that the
greatest performance benefit can be achieved (Mathworks, 2002).
Then the bogie bounding box that includes clearances is determined and the stowed position and retraction angle is calculated. It is then checked if the stowed bogie is still within the stowage boundaries predefined
by the user.
The previous steps are done for every feasible bogie type as is shown graphically in figure 3.9. From these
the least complex bogie type (the bogie with the least number of wheels) is selected. Reduced complexity
keeps cost and weight minimal.
The nose gear type is of the dual bogie type, which is used on virtually all transport aircraft. Clearances,
tyres and shock absorber sizing for the nose gear is done in a similar way as for the main gear. Finally the
main gear brakes and the oleo pneumatic shock absorbers can be sized.
In the landing gear positioning and the bogie design part of the PositionLandingGear module the following most important assumptions are made:
Only the tricycle gear layout is supported;
Only conventional transport aircraft;
The number of main gear struts is 2, 3 or 4. The third and fourth are assumed to be positioned on the
fuselage;
The nose gear has a fixed layout: 2 wheels placed symmetrically along a nose gear strut with a fixed
track of 0.5m;

3.1. Implementing the landing gear design

49

gear
positioning

bogie design
calculate bogie
loads

for each bogie layout evaluate:

size bogie
dimensions

tyre and wheel

selection

tyre +
wheel
data

calculate aircraft
flotation ACNs
calculate stowed
position and reqs

design brakes
and shock
absorbers

find best bogie

solution

design nose
wheel bogie

weight
estimation

Figure 3.9: The bogie design part of the PositionLandingGear module

The aircraft is symmetric in the xz-plane;

The maximum touch down angle is assumed to be equal to the maximum take-off angle. And the
maximum take-off angle is calculated empirically;
In calculating the turnover angle, the nose gear track is not taken into account;
Wing bogies are assumed to be connected to the wing between or on the rear and auxiliary spar. The
gear-spar connection point is assumed located halfway along the spar height;
A clearance angle of 5 degrees for nacelle with an additional 6 inch clearance is used. The wing tip has
a set clearance angle and also an additional 6 inch clearance (Torenbeek, 1982);
The maximum turnover angle allowed is 63 degrees;
The aircraft will need to be able to turn on a runway and manoeuvre on a taxiway according to FAA
advisory circular No. 150/5300-13A. Main gear steering is not taken into account;
The drag strut on the wing main gear is set at a fixed angle of 10 degrees w.r.t the vertical and 0 degree
w.r.t. the aircraft xz-plane. The side strut is assumed at a fixed angle of 45 degrees w.r.t. the yz-plane.
The fuselage and nose gear bogies are assumed to have a side and drag strut at the same angle and are
positioned symmetrically;
The drag strut and side strut for each bogie is assumed to be attached at the same point on the main
strut. This point is fixed halfway along the gear length;

50

3. Implementation and use cases

z
x
y
V

F side strut
F drag strut

wheel base

tyre track
Figure 3.10: The structural components of a bogie with 4 tyres modeled with tubes

All main gear tyres are assumed to be braked. The brake weight is derived from the requirement to
perform 1 RTO, 5 stops at maximum landing mass with 10 ft/s deceleration and 250 stops at design
landing mass with 10 ft/s deceleration (Currey, 1988);
For the shock absorber the design methods explained in chapter 2, section 2.1.8, are used. The design
parameters used are as follows. The maximum load factor is 1.2 (2.2 at the aircraft cg), the maximum
sink speed is 10 ft/s, the tyre efficiency is 47% and the shock efficiency is 85%. The compression ratio
is 4/1 static to extended, 3/1 compressed to static and the static pressure is 1500 psi (Currey, 1988).
Polytropic compression is assumed where the oil and gas are separated. Only oleo pneumatic shock
absorbers are considered.

3.1.3. PositionLandingGear class 2.5 weight estimation

The general dimensions of the landing gear components were calculated in the previous sections. To make a
weight estimation it is convenient to model the trunnion, shock strut, truck and axles as tubes as is shown in
figure 3.10.
At this point only the radii and thicknesses of the structural components need to be known to make a
weight estimation. Only the required shock absorber inner radius is know from the piston area. The radii
of the other structural components are set to a fixed value related to the piston area. Thicknesses of the
structural members can be found by considering the maximum stresses within the structure that could occur
due to the extreme load cases defined in chapter 2, section 2.2.3. The maximum stresses are found analytically
using structural analysis theory (Megson, 1999). The complete analytical derivation of the maximum stresses
can be found in section 2.2.1.
Figure 3.11 shows all steps of the class 2.5 weight estimation. First all dimensions of the structural members that are needed for the analysis are obtained. Then the maximum normal stress and axial force due
to bending are calculated. The shear stress due to torsion and shear forces are also calculated. All stresses
are combined to find the maximum stresses within the cross-section at different critical points within the
structure. Stresses are then multiplied with a safety factor of 1.5. Thicknesses of all structural members are

3.1. Implementing the landing gear design

51

bogie design

weight estimation
get size of
structural parts

for each bogie and load case evaluate:

size side struts to

yield or buckling
criterion

size tubes
thicknesses to
yield criterion

calculate max
stresses

determine mass
and c.g.

Figure 3.11: The weight estimation part of the PositionLandingGear module

calculated such that yield does not occur. If the side struts are loaded under compression the buckling criteria
is applied.
Structural analysis is done for each main gear bogie separately, because lateral side loads put different
loads on the side struts of the left and right main gear bogies. The nose gear is also evaluated separately.
Finally when the thicknesses are known, masses and centre of gravity position are quantified. An example
result for different load cases is presented in the next chapter (section 4.3.2).
The following assumptions have been made in the weight estimation part of the PositionLandingGear
module.
The landing gear structure is assumed to be made out of LESCALLOY 300M VAC ARC ultra high strength
steel. This material is commonly used for landing gears (ASM International, 2013);
A safety factor of 1.5 on maximum stresses;
The structure is not allowed to yield at maximum load;
All structural parts are assumed to be made out of tubes except for the side struts. The side struts are
modelled as two force members with an I-beam cross-section. The side struts are thus only loaded
axially;
The tubes are analysed as thick-walled tubes. The side struts are considered thin-walled. The width
and height of the strut cross-section are fixed at 0.15 m, only the thickness is varied;
8 different load cases determine the maximum loads. These include: static load, 3 point landing, one
wheel landing and tail down landing at 10 fps, the lateral drift landing, braked roll turning and pivoting;
Side loads on the wheels are assumed to only cause a moment about the x-axis. No other external
moments are applied at the axle ends;
The deflection of the side struts in z-direction is negligible;

52

3. Implementation and use cases

All radii of the tubular structural members are derived from the piston area;
All properties of the PositionLandingGear, Class2WeightEstimation module and LandingGear part are

shown in a class diagram in figure 3.12.

InitiatorController

DesignModules

Part

Part

PositionLandingGear

Class2WeightEstimation

LandingGear

(Access = protected)
xRef
zRef
WeightResults
debug
displayWarnings
CLmax_l
CLmax_to
CLcruise
wingTotalArea
wingTotalSpan
wingPosition
wingRootTE
wingTipTE
wingDihedral
wingGearTE
MAC
xMAC
GearHeight
NoseGearLoadCGfwd
MainGearLoadCGaft
MainGearPosition
MainGearLength
NoseGearPosition
NoseGearLength
NoseGearNWheels
NoseBogieSStrutCylCxx
NoseBogieSStrutSparCxx
NoseBogieDStrutSparCxx
SparX
SparY
SparZ

FuelScalingFactor
PaxScalingFactor
CargoScalingFactor
(Access = protected)
WeightResults
UsingClassI
RemainingPayloadMass
HasCargo
calculateTotals
createLandingGear
getCargoWeight
getEngineWeight
getFuelWeight
getFurnishingWeight
getFuselageWeight
getLandingGearWeight
getLoadingCGRange
getMainWingWeight
getOperationalItems
getOpsBoundsCG
getPaxWeight
getSystemsWeight
getWingWeight
plotCGRange

Type
TyreDiameter
TyreThickness
Length
NRows
NWheelsPerRow
ExtPosition
Position
XPositions
YPositions
(Access = private)
TyreGeometry
getTyreGeometry
(Access = protected)
generate

(Static)
sumStruct

findLowestLandingGear
designBogies
getLandingGearWeight

Figure 3.12: Class diagram of the PositionLandingGear, Class2WeightEstimation module and a LandingGear part.

3.2. Implementing the landing gear analysis

53

3.2. Implementing the landing gear analysis

The landing gear analysis can verify loads within the structure for different European Aviation Safety Agency
CS 25 load cases. The kinematics of the landing gear retraction mechanism of nose and main gear bogie is
also verified.
Analysis of the landing gear has been made separate from the modules in the TU Delft Initiator. The
analysis however does use the XML output of the Initiator as input. Dynamic analysis of the landing gear
model is made as a SimMechanics application in Matlab. The multi-body system is modelled with blocks
that represent rigid bodies, joints, constraints and force elements. The blocks of the implemented multibody model can be seen in figure 3.16. Using provided input parameters, SimMechanics then evaluates the
equations of motion of the complete system and tests the main landing gear for different load cases.

Initiator output

get size and

mass of parts

for each bogie:

drop test
including spin-up

drop test

kinematic
analysis

setup
SimMechanics
model

loading results

Figure 3.13: Landing gear analysis workflow

Steps taken in the landing gear analysis are displayed in figure 3.13. For each bogie the SimMechanics
model is setup by automatically adding blocks from a library to the model. Added blocks include blocks for
axles, wheels, tyres and a truck beam.
Then a kinematic analysis of the retraction mechanism is performed. The initial shock absorber stroke
and the retraction angle is set to the extended position by default. The side strut locking mechanism is set
initially to the locked position. The bogie rotates about a single axis as shown in figure 3.14. It is retracted and
extended using actuators that control the retraction angle, the truck beam angle and the locking mechanism.
retraction actuator force
fixed
g

extended position

lock actuator
force

retraction angle

Figure 3.14: Landing gear analysis retraction model used for a retraction/extension simulation. The initial condition is shown.

54

3. Implementation and use cases

fixed

fixed
V

extended position
spinup

spinup

initial truck angle

drop height

W/2

W/2

W/2

W/2

Figure 3.15: Landing gear analysis model used for performing a drop test simulation

The retraction analysis is followed by a drop test simulation, where the bogie is oriented in a landing state
and dropped. The degree of freedom of the landing gear strut is limited only the vertical motion. The bogie
initial height is set such that it will hit the ground at 10 feet per second due to gravity. This speed is prescribed
by the three point landing, regular landing and one wheel landing load case in FAR 25 (European Aviation
Safety Agency, 2012, p. 1-App A-2)). For the drop test simulation gravity is reversed in the upward direction,
the ground will thus move upward. Tyre spin-up is not included for the first test case.
The drop test is then done a second time, but now a tyre spin-up initial condition is included in the touchdown analysis. Tyre spin-up is applied to all tyres with an equal angular velocity.
When the simulations complete, all results are combined and verified with results obtained from the class
2.5 weight estimation in the PositionLandingGear module.
The most important assumptions in the kinematic analysis are:
The simplest form of gear retraction is assumed, which is a rotation only about the trunnion beam
x-axis (figure 3.10);
Only the number of wheels, tyres, axles, tube radii and thicknesses are varied;
The shock absorber is oleo-pneumatic (non-linear);
The empirical model of Pacejka is used to model the tyres. This model is called the magic formula tyre
model;
The inputs and outputs of the tyre model are shown in figure 3.17. The actual tyre model is implemented
as a Matlab function that is evaluated at each time step. The Matlab function implements the equations of the
MF-tyre implementation (TNO Automotive, 2010) (Pacejka, 2006). In section 2.3.1 details of the tyre model
coordinate system and input and output values are further explained.
The non-linear oleo-pneumatic shock absorber forces are modelled as a Matlab function block. The equations of chapter 2, section 2.1.8 are implemented in this function block.

3.3. User manual

Since the landing gear design and the landing gear analysis are separate modules, the user manual has also
been split up in two parts. First the use of the landing gear design module is explained.

3.3. User manual

55

Figure 3.16: Landing gear SimMechanics model of a main landing gear. Components include a oleo-pneumatic shock absorber, side
struts, axles, tyres and wheels.

Figure 3.17: Tyre model block contents, which is part of the landing gear SimMechanics model.

56

3. Implementation and use cases

Description

Input name

Default

number of struts

NumStruts

min selected airport flexible pavement

MinSelectedAirportFlex

Rotterdam Zestienhoven

min selected airport rigid pavement

MinSelectedAirportRigid

London Luton

heat sink material

HeatsinkMaterial

Carbon

forward and aft main gear stowage (fuselage length fraction)

MainGearStowage

0.35;0.45

main gear stowage width (fuselage width fraction)

MainGearStowageWidth

1.0

fuselage fairing height (meter)

FuselageFairingHeight

forward and aft nose gear stowage (fuselage length fraction)

NoseGearStowage

0.05;0.2

maximum nose gear load (maximum ramp mass fraction)

MaxNoseGearLoad

0.15

minimum nose gear load (maximum ramp mass fraction)

MinNoseGearLoad

0.6

tyre load safety factor

TyreLoadSF

1.25

aft centre of gravity x-position

xCGaft

forward centre of gravity x-position

xCGfwd

Table 3.1: List of possible inputs for the PositionLandingGear module

3.3.1. PositionLandingGear user manual

For the PositionLandingGear module the following is required to run the module.
The take-off and approach speeds need to be defined in the input file;
The aircraft needs to have a feasible aft and fwd centre of gravity position;
The aft section of the fuselage needs to be properly shaped, such that the take-off scrape angle can be
achieved;
The wing needs to consist of either 1 or 2 parts;
Results of the class 2 weight estimation module
One of the most important requirements is that the cg positions are feasible. This can be checked within
the class2weight estimation. Results of the class 2 weight estimation are required, because at this stage the
aircraft component geometries, weights and aircraft performance data is available.
When creating an input file one can add specific inputs listed in table 3.1. The inputs are all optional and
if no input is given the default values are used.
The available stowage space within the aircraft is controlled by the NoseGearStowage, MainGearStowage,
FuselageFairingHeight and MainStowageWidth inputs. The definition of these parameters is also made visible in figures 3.18 and 3.19.
The minimum selected airport for flexible and rigid pavements represent the airports where the aircraft
should minimally be allowed to land. These parameters can be given as an airport name listed in appendix
B. It can also be given as a PCN number, this has to be the PCN for subgrade category 4 (CBR 3).
When setting the aft and forward centre of gravity x-position in the input file, these values will be used in
the analysis. This is usually done when results for a reference aircraft are needed. Centre of gravity positions
are then already known. If xCGaft and xCGfwd are not entered in the input file, a centre of gravity position
calculated by the class 2 weight estimation is used.
All default values can be set in the settings.xml file. These settings are currently set such that large number
of aircraft by default can successfully run. If in future landing gear positions pose a problem (in a optimisation
loop) settings can be set even less strict.

3.3. User manual

Nose
G

earSt

57

owag
e

Nose
G

(1)

earSt

owag
e

(2)

Main
G

earSt
owag
e

(1)

Main
G

earSt
owag
e

(2)

Figure 3.18: The input parameters that define the main gear and nose gear available stowage

available stowage space

MainGearStowageWidth

FuselageFairingHeight

Figure 3.19: The input parameters that define the main gear and nose gear available stowage

There is also an addition setting called PlotRetractedGear which will put the right wing and right fuselage
bogie in retracted position in the 3-dimensional plot.
Properties of the landing gear part are listed in table 3.2. A landing gear part is created by the class 2 weight
estimation. When the PositionLandingGear module is run the class 2 gear parts are replaced by the results of
this module.
All outputs of the landing gear sizing and selection tool are listed in table C.1 in appendix C. In this table
some properties have the prefix WingBogie. The same properties names are applied to a fuselage bogie (if it
exists) with a prefix FuseBogie or they are applied to a nose gear bogie without a prefix.

3.3.2. How to run the landing gear design module

The landing gear design module is run by first starting the Initiator:

C = InitiatorController(A380.xml);
Then the PositionLandingGear module can be run:

C.runModule(PositionLandingGear)
When the module finished results are written to the A380.xml file. Results are also available in the Matlab
workspace as a C InitiatorController object. Another way to display the results is by making a report. The
report will include positioning and bogie design details. If class 2.5 landing gear weights need to be included
in the report the Class25WeightEstimation also needs to be run. This can be done by:

58

3. Implementation and use cases

Description

Property

Position vector of the bogie centre on the ground (meter)

Position

Rotation about x, y and z axis (origin at the wing gear connection) (deg)

Orientation

Extended position vector of the bogie centre on the ground (meter)

ExtPosition

NoseGear or MainGear

Type

Maximum tyre diameter (meter)

TyreDiameter

Maximum tyre section width (meter)

TyreThickness

Gear length from ground to wing connection point (meter)

Length

Number of wheel rows

NRows

Number of wheels per row

NWheelPerRow

Option to override default position to predefined tyre x-positions (meter)

XPositions

Option to override default position to predefined tyre y-positions (meter)

YPositions

Table 3.2: List of properties of a LandingGear part

C.runModule(Class25WeightEstimation)
C.runModule(ReportWriter)

3.3.3. Module messages description

The Class2WeightEstimation and PositionLandingGear module can produces a number of messages. An description of each possible message is listed in the following sections.
The following warning messages could be displayed by the Class 2 weight and centre of gravity range
estimation:
CG range limited by fwd ops bound due to nose gear load limits
The forward operational bound limits the centre of gravity range too much. The forward operational
bound is imposed due to the maximum nose gear load (MaxNoseGearLoad in the settings or input file).
CG range limited by aft ops bound due to nose gear load limits
The aft operational bound limits the centre of gravity range too much. The aft operational bound is
imposed due to the minimum nose gear load (MinNoseGearLoad in the settings or input file).
The warning messages of the positioning module are only shown if GearPositioningWarnings is set to true
in the settings.xml file.
Unconventional aircraft types not supported
The aircraft that is being analysed is of unconventional type. Only the conventional aircraft type is
supported by PositionLandingGear
Gear height needed seems large, check ac aftfinenessratio or nosegear loading setting
The lowest possible feasible landing gear layout is larger than 3 meters heigh from the belly of the
fuselage.
Check wing position w.r.t. fuselage
The centre of gravity range is such that the nose gear cannot be placed properly. This is probably due
to the lateral wing fuselage position.
wheelbase too large for taxiing
The landing gear wheelbase does not comply with the FAA Advisory Circular No. 150/5300-13A.

3.3. User manual

59

track too large for taxiing

The landing gear track does not comply with the FAA Advisory Circular No. 150/5300-13A.
turn radius too high
The landing gear turn radius does not comply with the FAA Advisory Circular No. 150/5300-13A. This
problem can be solved by applying main gear steering.
Main gear cannot be positioned forward of auxiliary spar. Change the aft, fwd cg or aux spar position.
Nose gear loading limits and the current auxiliary spar position prevent the result of a feasible main
gear position.
nacelle goes through ground
The lowest possible gear height selected places the nacelles through the ground. This is a validation
check of the engine clearance calculations.
No feasible landing gear positions found, adjust cg range, aux spar, nosegearloading setting or stowage
This error message occurs when there can be no feasible landing gear position found in the design
space. This only can be solved by adjusting the centre of gravity range, the nose gear loading setting
and/or stowage settings. This error message stops further results of the PositionLandingGear module.
Number of struts higher than 4 not supported
The number of struts in the input file is set higher than 4. Only 2, 3 or 4 main gear struts are supported.
No feasible bogies found
Of the 8 different landing gear layouts none can be applied. This is usually due to improper stowage
requirement settings or wing rear/auxiliary spar positions.
small wheels not supported for brake sizing
The given aircraft requires tyres that have a relatively small radius. Only conventional transport aircraft
are supported by the PositionLandingGear module.
shock absorber pressure too low
The required shock absorber pressure is lower than 60 psi which will result in sticking due to friction.
This is a validation check of the shock absorber design.
shock absorber pressure too high
The required shock absorber pressure is higher than 6000 psi, seal leakage could occur. This is a validation check of the shock absorber design.
trunnion radius increased
Loads on the bogie are high; a large tube wall thickness is required. To reduce the trunnion weight and
manufacturability the trunnion radius is increased by 10 percent.

3.3.4. Landing gear analysis user manual

The landing gear analysis module performs several kinematic and multi-body simulations. The SimMechanics kinematic and multi-body analysis will give as output:
a SimMechanics model of the nose gear, the wing landing gears and if they exist the fuselage landing
gears;
loads within the gears structural members for several load cases;
duration of retraction/extension and the retraction actuator efficiency;

60

3. Implementation and use cases

response behaviour to dynamic loading
Results of the landing gear design module in the Initiator are used as input of the analysis. The Position-

LandingGear module gives several feasible landing gear layouts. Only one can be used in the multi-body
analysis. By default the solution with the least number of wheels is used. If one would like to analyse a different combination, this can be set in the runsim.m Matlab run file.
There are a few extra settings that can be altered if the default kinematic solution is not sufficient. These
include the position of the side strut bracket that forms the connection between the shock absorber and the
side strut, shown in figure 3.20. Also the angle between of the locking mechanism needs to be set properly
(figure 3.21).

Figure 3.20: Connection between side strut and shock absorber

Figure 3.21: Locking mechanism between side strut and shock absorber

After finalising all simulations, results will be displayed in figures and as parameters in the Matlab workspace.
The results are grouped together in a struct called Results.
The explanation of the messages that could be displayed during the simulation are listed below.

3.4. Definition of use cases

61

Gear needs to be connected above spar centre

The shock absorber stroke length required is longer than the piston cylinder length. Therefore the gear
strut wing spar connection point is moved upwards.
Gear stroke too long to be placed on wing
The shock absorber stroke length required is longer than the piston cylinder length. The gear strut
wing spar connection point required is even higher than the highest point on the wing spar. This is not
feasible.
Structural thickness results incomplete
The structural thicknesses calculated by the weight estimation of the PositionLandingGear module
contains a NaN value. This is a validation check of the weight estimation result.
Side strut top bracket goes through cylinder, position brackets differently
The bracket that connects the strut with the side struts needs to be moved.

3.3.5. How to run the landing gear analysis module

First the filename and path of the Initiator output xml file needs to be set in the runsim.m file. In this file also
the initial conditions of the different analyses can be changed. The landing gear analysis module is then run
by starting the simulation in Matlab with the following command.

runsim

3.4. Definition of use cases

The landing gear design modules in the Initiator can be used for many different use cases. Two different use
cases will be explained in the following sections. The first example use case is a class 2 component weight
analysis of 2 different aircraft concepts with comparable missions.

3.4.1. Class 2 weight estimation example use case

For this example use case the weight of the landing gear of a blended wing body aircraft can be compared
with the gear weight of a conventional aircraft. It needs to be possible to do this rapidly, since the design is in
a preliminary design phase.
This use case starts by defining an input file for a blended wing body. An input file called ovalbwb.xml
for a blended wing body aircraft already exists in the CleanInputFiles folder. In this input file the mission
requirements are set. Mission requirements include the required passenger and cargo payload, cruise speed,
altitude and range. For a Class 2 weight estimation also other inputs are required, namely:
performance parameters, such as lift coefficients at take off and landing;
configuration parameters, as wing aspect ratio and location;
fuselage shape definition, because of the nonconventional shape of a blended wing body this shape
needs to be defined more precisely than would be necessary for a conventional aircraft. It is important
to set the nose and aft shape to get the desired wide shape of the fuselage (Schmidt, 2013);
a detailed right wing shape definition, which includes span, airfoils, thicknesses, taper, sweep, twist
and dihedral at different wing sections. A similar separate left wing is also defined;
engine positions;
This input file can then be analysed by the Initiator by running the following commands in Matlab

62

3. Implementation and use cases

Initiator ovalbwb.xml --interactive

run Class2WeightEstimation
run PlotTool
plot geometry
A comparable aircraft to the oval blended wing body, in terms of mission requirements, is the Boeing
777-300ER aircraft. Class 2 weight estimation results are generated similarly as for the oval BWB aircraft. The
B777-300ER.xml file in the CleanInputFiles folder is used here as input.
Figure 3.22 shows the graphical results of the oval BWB and figure 3.23 the results of the Boeing 777. Both
aircraft have the same 6 wheel main gear bogies. For a blended wing body the fuselage is wide enough to
place all main gear bogies on the fuselage. The 777 has the main gears placed on the main wing and therefore
requires longer main gear struts. The weight of the 777 gear is therefore expected to be higher. Components
weight results of both aircraft are compared in figure 3.24. The percentages of gear group weights are both
comparable: 5.9 percent for the 777 and 5.5 percent for the BWB. When comparing the weights in kg, the total
weight of the oval BWB aircraft is found 14 percent lower than the Boeing 777 gear weight.

3.4.2. Use case kinematic and multi-body analysis

Another use case is a kinematic and multi-body analysis on an existing conventional reference aircraft. The
PositionLandingGear module and the kinematic module have been specifically designed for this use case.
First create an input file of the reference aircraft. Also define reference values for the weights and centre
of gravity ranges in the input file. By defining these reference values, results of the preliminary sizing will not
be used. Instead the more accurate reference values are used to create a more realistic model.
To perform the landing gear analysis the following steps need to be done:
Run the PositionLandingGear module;
Define the resulting output xml file in the input for the kinematic analysis;
Select the bogie combination to be analysed;
And select the load case to be analysed and change the initial conditions in the runsim.m file
Analyse results
This is done by editing the runsim.m file and running the following commands in Matlab the PositionLandingGear module:

Initiator B777-300ER.xml --interactive

run PositionLandingGear
run ReportWriter
quit
runsim
Additionally by running the report writer module a pdf report is generated with landing gear characteristics. In this report also figures are included that show the wing, engine clearance angles, the take off scrape
angles and the landing gear footprint.
The kinematic feasibility of the kinematic model and loads due the different load cases are the result of
the analysis. These can be evaluated by inspecting the results displayed in figures and in the Results struct in
the Matlab workspace.

3.4. Definition of use cases

63

Figure 3.22: Oval blended wing body CWE2 result

Figure 3.23: Boeing 777ER CWE2 result

cargo
fuel
fuel

cargo
15.7%
30.0%

17.5%

23.7%

pax
11.8%

.
4.9%

13.0%

1.2%

pax

8.3%
2.9%
0.3%
2.6%
2.6%
1.4%
4.3%
nose gear
main gear 1
main gear 2
systems

wing
tail

2.7%
0.5%
2.7%
nose gear
2.7%
main gear 1
1.3%
main gear 2
5.1%

fuselage

11.0%

12.0%
wing
2.6%
10.6%

engines

8.6%
tail

systems
engines

fuselage

Figure 3.24: Comparison of class 2 weight estimation results of an oval BWB right and a Boeing 777 aircraft left. Both aircraft have the
same mission requirements.

4
Results and verification
4.1. Positioning results
To check whether the output of the positioning module makes sense the result is compared against real aircraft data found in airport planning manuals published by aircraft manufacturers. This is done graphically in
figure 4.1 and tyres are compared in table 4.1. Reference aft cg, forward cg, reference stowage positions and
reference aircraft weights have been used here. Estimations of the class 2 weight estimation have been left
out to make a better comparison.
Description

Nose gear tyre

Rated load (kg)

Main gear tyre

Rated load (kg)

A320-200

39x13

6.80e3

H46x18.0-20

23.2e3

A320-200 actual

30 x 8.8 R15

6.50e3

46 x 17 R20

20.9e3

A340-500

B46x16.0-23.5

24.4e3

1400x530R23

34.0e3

A340-500 actual

45x18.0R17 36PR

22.8e3

1400x530R23

34.0e3

A380-800

54x21.0-23

31.1e3

54x21.0-23

32.7e3

A380-800 actual

1270 x 455R22 32PR

24.9e3

1400 x 530R23 40PR

34.0e3

B787-800

40x14

12.6e3

1400x530R23

23.7e3

B787-800 actual

40 x 16.0 R16 26PR

13.2e3

50 x 20.0 R22 34 PR

25.9e3

B777-300ER

H42x16.0-19

17.1e3

1400x530R23

34.0e3

B777-300ER actual

43X17.5R17, 32PR

20.6e3

52X21R22, 36PR

30.2e3

B707-321

15.50-20

9.30e3

49x17

18.0e3

B707-321 actual

39x13, 16PR

8.80e3

46x16, 30PR

20.3e3

Table 4.1: Calculated tyre results compared with actual aircraft tyres used (Goodyear, 2002), (Michelin Aircraft Tire, 2001)

For the 6 different aircraft in figure 4.1, the lateral tyre positions differ with 0.49 m and longitudinal positions with 0.28 m. Deviation with reference data is thus small. Also the number of wheels and clearances
between the wheels match, indicating that the flotation analysis and Torenbeek clearance equations produce
good results for these type of aircraft.
The calculated A380 landing gear differs from the actual aircraft. The fuselage and wing bogies have been
interchanged. Also the calculated fuselage gear has been placed further aft, since only forward retracting
fuselage gears have been taken into account. This resulted in a nose gear tyre that has a significant higher
rated load (see table 4.1). The calculated gear position is however a good result. Airbus also considered this
65

66

4. Results and verification

lat. position [m]

lat. position [m]

4
2
0
2
4

.
0

10

15

20

. .calculated
. .actual
.
0
10

40

5
.
0

10

20

30

.
0

40

A380 gear lon. position [m]

10

20

30

B787 gear lon. position [m]

4
lat. position [m]

5
lat. position [m]

30

A340 gear lon. position [m]

lat. position [m]

lat. position [m]

A320 gear lon. position [m]

20

5
.
0

2
0
2
4

10

20

30

B777 gear lon. position [m]

40

.
0

10

15

20

B707 gear lon. position [m]

Figure 4.1: Comparison between calculated landing gear positions and actual positions as published by aircraft manufacturers

25

4.2. Runway flotation analysis

67

option in the design of the A380 (Hebborn, 2008). Airbus discarded the 4 wheel fuselage bogie concept in a
trade-off further in the design process. At this design stage Airbus had more details available (more than in
the current analysis) to make a better trade-off.

4.2. Runway flotation analysis

The FAA COMFAA program has an extensive database of commercial and military aircraft (Federal Aviation
Administration, 2011). As a check of the flotation results the calculated ACN values of a number of aircraft
are compared with results of the original FAA COMFAA program. This is done using the same COMFAA input
values. It has to be notes that the results of the COMFAA program might differ from actual ACN values as
published by aircraft manufacturers.
Results are shown in table 4.2 and as can be seen calculated results have a small error. Only for the Airbus
A380-800 aircraft the results deviate more. Maximum deviation found here is 15 percent. A cause for this
deviation has not been found. The deviation is however small enough to not influence the final landing gear
result.
Aircraft

CBR 3

CBR 6

CBR 10

CBR 15

COMFAA: ACN flex A380-800 fuselage bogie

105.2

74.7

61.7

55.9

calculated: ACN flex fuselage bogie A380-800

105.2

74.4

61.3

55.9

COMFAA: ACN rigid A380-800 fuselage bogie

109.6

88.1

67.6

54.8

calculated: ACN rigid fuselage bogie A380-800

109.7

88.1

68.5

63.1

COMFAA: ACN flex wing bogie A380-800

102.1

75.1

63.6

58.2

calculated: ACN flex wing bogie A380-800

110.8

82.2

67.7

62.7

COMFAA: ACN flex rigid wing bogie A380-800

90.2

78.1

65.8

56.3

calculated: ACN flex rigid wing bogie A380-800

90.2

78.1

65.9

57.5

COMFAA: ACN flex B777-300ER

120.3

89.3

71.3

63.8

calculated: ACN flex B777-300ER

120.3

89.3

70.3

63.5

COMFAA: ACN rigid B777-300ER

131.9

109.8

85.7

66.1

calculated: ACN rigid B777-300ER

131.9

109.8

85.8

68.3

COMFAA: ACN flex A320-200

50.2

44.4

40.0

38.5

calculated: ACN flex A320-200

50.2

44.3

39.4

38.4

COMFAA: ACN rigid A320-200

50.4

48.4

46.0

43.5

calculated: ACN rigid A320-200

50.4

48.4

46.0

43.5

COMFAA: ACN flex B737-200

39.3

35.2

31.1

30.0

calculated: ACN flex B737-200

39.3

35.2

30.6

29.7

COMFAA: ACN rigid B737-200

39.3

37.8

35.9

34.0

calculated: ACN rigid B737-200

39.3

37.8

35.9

34.0

Table 4.2: Comparison of flotation calculations of several reference aircraft for rigid and flexible pavements.

4.3. Gear weight estimation accuracy

4.3.1. Class 2 weight estimation results
Table 4.3 lists the actual total landing gear group weights for 4 different aircraft (Roskam, 1989b). The aircraft
size of these aircraft can be categorised from small to large. The actual weights are compared with landing
gear weight results from the class 2 weight estimation module. Class 2 results are of the same order of magnitude and the average error is 17 percent.

68

4. Results and verification

Aircraft

Estimate (kg)

Actual (kg)

Est/Act

Boeing 737-200

2528

1975

1.28

Boeing 727-200

3770

3271

1.15

Boeing 707-321

4643

5799

0.80

DC 10-10

9855

9576

1.03

Table 4.3: Comparison of class 2 wing and nose bogie assembly weight with actual weights as given by Roskam

The class 2 weight estimation gives all Boeing aircraft 2 main gear struts with 4 tyres, but with different
tyre sizes. The McDonnell Douglas DC10 is given 2 main gear struts with 6 tyres. In reality this is different: the
B737 has 2, the B727 2, the B707 4 and the DC10 has 4 tyres per main gear strut. Considering the assumptions
that have been made at this design stage, positioning and weight estimation results are better than expected.

4.3.2. Structural parts weights

The weights of the structural parts are determined from maximum stresses that could occur for 8 different
load cases. The thickness of each part is adjusted such that the maximum stress in the structure does not
exceed the yield stress multiplied by a safety factor of 1.5. The maximum stress is here determined from the
Mises yield criterion as explained in section 2.2.1.
Deflections and internal moments as calculated by the class 2.5 weight estimation part of the landing
gear design module can be plotted for each load case. The reference load case is the static load case when the
maximum ramp weight is put on the main gear. The results of this load case for a Boeing 707-321 is shown in
figure 4.2. The displacement u, v and w is the displacement in x-, y- and z-direction. The coordinate system
of figure 3.10 is used here.
There are only a few load cases that are most critical. For the Boeing 707 aircraft analysed here, the most
critical load cases are the tail down landing (figure 4.3) and the lateral drift landing (figure 4.4). The axles
are sized based on the tail down landing and the other parts are sized on the lateral drift landing load case.
Specifically the lateral drift load case for the left landing gear, since the aircraft gear is loaded asymmetrically.
The loads are significantly higher than the right gear, because of the side struts that will be loaded differently.
The load on the side struts are also highest for the left gear for this load case.
Structural thicknesses and weights as calculated for the Boeing 707 are listed in table 4.4. The longest and
the most heavily loaded parts are given the highest thicknesses as expected. As a result these parts also have
the highest weights.

internal moment (Nm)

displacement (m)

103
6
4
2
0.
0

0.5
.
.

1.5

structural member length (m)

.axle w
.
.truck w
.piston cylinder u .
.trunnion w

105

2
1
0
1 .
0
.
.

0.5

1.5

structural member length (m)

.axle Mx
.
.truck My
.piston cylinder My .
.trunnion My

Figure 4.2: Structural deflections and internal moments for the static load case. These results are for the right main gear of a Boeing 707.

4.3. Gear weight estimation accuracy

10

internal moment (Nm)

displacement (m)

69

0
5
.
0

0.5

0
1
2 .
0

1.5
.
.
.
.

structural member length (m)

.axle w
.
.axle u
.truck w
.
.piston cylinder u
.trunnion w .
.trunnion v

.
.
.

106

0.5

1.5

structural member length (m)

.axle Mx
.
.axle Mz
.truck My
.
.truck Mz
.piston cylinder My .
.trunnion My
.trunnion Mz

Figure 4.3: Structural deflections and internal moments for the tail down landing load case. These results are for the right main gear of a
Boeing 707.

106
internal moment (Nm)

102

2
0
2

.
0

0.5

0.1

0.2 .
0
.
.
.
.

0.5

1.5
internal moment (Nm)

displacement (m)

displacement (m)

1.5

structural member length (m)

.axle w
.
.truck w
.truck v
.
.piston cylinder u
.piston cylinder v .
.trunnion w
.trunnion v

.
0
106

0.5

1.5

0.5

1.5

0
1
2 .
0
.
.
.
.

structural member length (m)

.axle Mx
.
.truck My
.truck Mz
.
.piston cylinder My
.piston cylinder Mx .
.trunnion My
.trunnion Mz

Figure 4.4: Structural deflections and internal moments for the lateral drift landing load case. The top 2 graphs are for the left main gear
and the bottom 2 for the right main gear of a Boeing 707.

70

4. Results and verification

Part

Beam radius (m)

Wall thickness (m)

Weight (kg)

axle

57.1e-3

16.2e-3

106

truck beam

111e-3

12.2e-3

127

piston cylinder

143e-3

22.7e-3

281

trunnion

117e-3

30.5e-3

183

side strut

1.29e-3

8.3

drag strut

2.05e-3

8.3

Table 4.4: Boeing 707 main landing gear structural component weight

Aircraft

Estimate (kg)

Actual (kg)

Est/Act

Boeing 737-200 main gear

300.1

348

0.86

Boeing 737-200 nose gear

75.2

65.7

1.14

Boeing 727-200 main gear

708

751

0.94

Boeing 727-200 nose gear

121

148

0.82

Boeing 707-321 main gear

712

1150

0.62

Boeing 707-321 nose gear

119

101

1.18

Table 4.5: Comparison of wing and nose bogie structural weight with actual weights as given by Currey and Chai and Mason

4.3.3. Comparison weight estimation

To validate that the class 2.5 weight estimation method produces satisfying results, the calculated structural
weight are compared to a landing gear structural weight estimation of Roskam. In this weight estimation the
landing gear structure is modelled similarly. The structural weight here do not include all structural parts of
the gear. Items such as torsion links, fittings and internal components of the shock absorber are not included.
These, including controls, represent about 23 percent of the total main gear and 44 percent of the total nose
gear structural weight (Currey, 1988, p.264). Results are presented in table 4.5 and do not include these items.
Structural weight data of the landing gear cannot be found easily. The weight data of Chai and Mason
and Currey are the only data sources found usable for a comparison. These data sources contain data for
only a few different transport aircraft. For the total landing gear group weight more information is available.
Torenbeek gives landing gear group weights for several commercial aircraft as well as Roskam, Currey and
Chai and Mason.
Results for the main gear structural weights deviate with an average of approximately 19 percent from the
actual value. Nose gear weight deviate with an average of approximately 17 percent. These structural weight
errors match with results obtained by Chai and Mason and Kraus. Kraus found an average error of 13 percent
and Chai and Mason found an error of 10 percent.
Tyres, wheels and brake weights have all been determined before and can now be added. All remaining
parts can also be added. These include fittings, controls and miscellaneous parts. The total assembly weights
of all main and nose gears combined are compared with actual total assembly weights of Roskam in table 4.6.
The average deviation from actual weights for these 4 aircraft is 15 percent.

4.3.4. Estimating weights for different aircraft types

The overall weight estimation capability of both the class 2 weight estimation module and the class 2.5 weight
estimation is compared by generating results for many different aircraft. Figure 4.5 shows the results for all
analysed aircraft. The aircraft size ranges from the Boeing 737 to the largest Airbus A380.
The class 2 module overestimates the gear weight for the smaller aircraft types. This is mainly due to
the number of wheels required per bogie is calculated higher than actual as mentioned before. For the most

4.4. Tyre model verification

Aircraft

71

Estimate (kg)

Actual (kg)

Est/Act

Boeing 737-200

1741

1975

0.88

Boeing 727-200

3200

3271

0.98

Boeing 707-321

4143

5799

0.71

DC 10-10

11337

9576

1.18

.
.
.

.CWE2
.CWE2.5
.actual
A380

B777
A340

DC10
B787

B707

A320

B727

B737

landing gear weight (percent MTOM)

Table 4.6: Comparison of wing and nose bogie assembly weight with actual weights as given by Roskam

3
.
104.7

104.8

104.9

105

105.1

105.2 105.3
MTOM (kg)

105.4

105.5

105.6

105.7

Figure 4.5: Comparison between weight estimation results of the class 2 weight estimation module, the class 2.5 weight estimation
module and actual weight percentages as published by Roskam. The x-axis is the maximum take-off mass and has a logarithmic scale.

heavy aircraft analysed actual gear weights are not available. But for the Boeing 747 the gear weight is gear
weight is 4.2 percent of the MTOM of 3.2e5 kg (Chai and Mason, 1997). Both the class 2 and the class 2.5 lines
in figure 4.5 are close to the B747 values. A further comparison for the heaviest aircraft cannot be made due
to the absence of reference data.
Overall the class 2.5 weight estimation is more close to the actual weight results of Roskam than the class
2 module. The class 2.5 weight estimation is based on a lot more information than the class 2 estimate. It is
therefore anticipated that the class 2.5 results are better.
There are many factors that can be changed in the design of the landing gear. Each of these factors influence the overall weight of the landing gear. If for example safety factors are changed, calculated weights
will change significantly. This is the most important reason why weight estimates shown in figure 4.5 are
deviating quite a bit.

4.4. Tyre model verification

The tyre model equations listed in Pacejka (2006) and TNO Automotive (2010) are implemented in a Matlab
script. The model will give the resulting forces F x , F y , F z and moments M x , M y , M z produced by the tyre due
to a given velocity vector, slip angle , slip ratio , path curvature at = a/R and camber angle .
To verify that the magic formula equations produce correct results reference tyre parameters of Pacejka
are used. Several plots are made as can be seen in figure 4.6 and these are compared with the results as shown
in Pacejka (2006, p. 189)

72

4. Results and verification

4,000
2,000

20

= t = 0

6
4
0

= -4

2,000 .

2,000

20

100
0

= -4

-100
-200

= t = 0

200
Mz [Nm]

Fy [N]

12

2,000

2,000

Fx [N]

Fx [N]

20

= 8
t = 0

4
0

Mz [Nm]

Fy [N]

4,000
2,000
0
= -4

2,000 .

2,000

200

= 8
t = 0

100

= -4

0
-100
-200

20
.

2,000

2,000

Fx [N]
-4
Mz [Nm]

Fy [N]

200

2,000

= 0
4
8

100
0

=0
.
0

=0
100 .
0

at = -a/R

=0
Mz [Nm]

2,000

0.7

2,000
.
20

-a/R = -0.15
10

0
[deg]

at = -a/R
200

Fy [N]

2,000

Fx [N]

8
4
0
= -4

2,000

2,000

10

20

0.7

=0

100
0
100
200 .
20

10

-a/R = -0.15
0

10

20

[deg]

Figure 4.6: Tyre model verification outputs using the hypothetical model parameters of Pacejka. Tyre forces F x , F y and self aligning
moment M z are plotted against slip angle , slip ratio and path curvature at = a/R. The camber angle is .

4.5. Landing gear analysis results

73

The results given here exactly match with the results of Pacejka. The equations of the tyre model thus
produce correct results for the given reference tyre parameters. The results of the tyre model are more detailed
than required for the aircraft tyre model that will be used. This is because the number of parameters available
for an aircraft tyre from TNO Automotive is less than the reference tyre presented here (TNO Automotive,
2012). When more detailed aircraft tyre data is available tyre model results will be more accurate.

4.5. Landing gear analysis results

4.5.1. Drop test results for an Airbus A320
A landing gear analysis has been performed on an Airbus A320-200 main landing gear. This is the simplest
bogie type and does not introduce loads due to rotation of the bogie truck. A schematic representation of the
simulation, including degrees of freedom, is available in section 3.2. The default ode45 solver can efficiently
solve the equations of motion and is therefore used in the drop test simulations. The order of accuracy of this
solver medium (Mathworks, 2012). When more accuracy is required a ode113 solver can be used, but this
requires a lot more time steps to run the complete simulation.
Figure 4.8 and 4.9 displays the results for which it has been assumed that lift is equal to weight. The
landing is performed at maximum landing weight with a touchdown vertical velocity of 10 ft/s (3 m/s). A
maximum load of 4.32e5 N is found on the main gear strut. This is 1.27 times the static maximum ramp
weight strut load (which equal to a load of 3.62e5 N per strut). An estimate of the dynamic load factor of 1.33
for a 10 ft/s landing has been found before using equation 2.45. Estimated dynamic loads were used in the
analytical weight estimation and are thus only slightly over-dimension the landing gear structural parts.

Figure 4.7: Drop test simulation of the multi-body model of an Airbus A320 main landing gear

The drop test performed here is not a regular landing, but an extremely hard landing. This should almost
never occur during the life of the aircraft. The landing gear has been designed for a load factor of 1.2. During a
normal landing the impact load will be much lower: about 40 percent of the static load (as mentioned before
in section 2.3 of chapter 2).
The tyre deflection when a static load is applied is for the A320 tyre 0.21 m. In figure 4.9 the maximum
tyre deflection is only 0.13 m. The vertical force of a hard landing is thus not the critical vertical force on the

4. Results and verification

shock force (N)

74

105
0
2
4

.
0

0.5

1.5

2.5

3.5

2.5

3.5

2.5

3.5

shock stroke (m)

Time (seconds)
0.3
0.2
0.1
0

.
0

0.5

1.5

shock velocity (m/s)

Time (seconds)
2
1
0
1

.
0

0.5

1.5

2
Time (seconds)

Figure 4.8: Shock absorber force, stroke and velocity for a two wheel A320-200 main landing gear landing at 10 ft/s

tyre force (N)

105
2
1
0.
0

0.5

1.5

2.5

3.5

2.5

3.5

2.5

3.5

tyre deflection (m)

Time (seconds)
0.00
0.05
0.10
.
0

0.5

1.5

tyre z-velocity (m/s)

Time (seconds)

0
1
2
3 .
0

0.5

1.5

2
Time (seconds)

Figure 4.9: Tyre vertical force, deflection and velocity for a two wheel A320-200 main landing gear landing at 10 ft/s

4.5. Landing gear analysis results

75

tyre, when tyre spin-up before touchdown and a horizontal friction force is not taken into account.
When comparing the results with actual measurement results of Ladda and Struck the shock force response in z-direction is initially similar. When looking on a longer time span the result of the model does not
increase as the measurement results. This is mainly due to the change in lift force, which is not taken into
account. Therefore an evaluation for a longer time period is not represented accurately by this model. In
reality the lift on the wings reduces rapidly when lift dumpers are deployed during the landing rollout.
The behaviour of the oleo-pneumatic shock absorber and the tyre to the applied load is good. The vibrations in the shock and tyre damp out rapidly during a short time. There are no high frequency loads seen in
the results. This is desirable when considering fatigue in the landing gear material.

4.5.2. Drop test results for a Boeing 777

The Boeing 777-300ER has a large 2 strut main landing gear with 6 wheels per strut. For this simulation of
a drop test the same initial conditions apply as for the A320 landing gear. The gear lands at 10 fps and lift is
assumed equal to weight. Only now there is a bogie truck placed at an initial angle of 15 degrees with respect
to the horizontal. The results for the shock absorber and the rear axle tyre are shown in figures 4.11 and 4.12.
The rear axle tyre will hit the ground first at touchdown.

Figure 4.10: Drop test simulation of the multi-body model a Boeing 777-300Er main landing gear

The interaction with the front axle and the rear axle tyres is important here. The result of this interaction
is that the landing gear bounces several times after the first impact. Also there are high frequency loads on the
gear. At the point when both the front and rear axle tyres simultaneously hit the ground the load on the shock
strut is 2.6 times the strut static load. Because of the high rotational velocity of the truck beam the shock and
tyre is highly loaded briefly. When the truck beam velocity reduces, maximum loads are considerably lower.
The maximum load factor after the first simultaneous impact is 0.7.
After 2.5 seconds the bouncing of the gear stops and the landing loads can be damped more effectively.
The shock force however still fluctuates at high frequency. This is because the truck beam motion is undamped. The peak loads can be reduced by adding a damping force to the truck beam actuator as well as
adjusting the shock absorber design.

4. Results and verification

shock force (N)

76

106
0
2
4 .
0

0.5

1.5

2.5

3.5

2.5

3.5

2.5

3.5

shock stroke (m)

Time (seconds)
0.1
0
0.1
0.2 .
0

0.5

1.5

shock velocity (m/s)

Time (seconds)

2
0
.
0

0.5

1.5

2
Time (seconds)

tyre force (N)

Figure 4.11: Shock absorber force, stroke and velocity for a six wheel B777-300ER main landing gear landing at 10 ft/s

105

2
0.
0

0.5

1.5

2.5

3.5

2.5

3.5

2.5

3.5

tyre z position (m)

Time (seconds)
0.4
0.2
0
0.2 .
0

0.5

1.5

tyre z-velocity (m/s)

Time (seconds)

5
0
5
10 .
0

0.5

1.5

2
Time (seconds)

Figure 4.12: Rear axle tyre vertical force, position and velocity for a six wheel B777-300ER main landing gear landing at 10 ft/s

4.6. Calculation run time

77

4.5.3. Tyre spin-up influence

If the landing gear tyres are not spin-up before touchdown, the difference in speed of the tyres with the ground
is the approach velocity. To simulate such a landing the tyres are given an initial radial velocity of = V /R,
which is for an Airbus A320 68.9 m/s / 0.58 m = 118 rad/s or 19 rotations per second. The simulation is
performed using the same initial conditions as in 4.5.1. Results are shown in figure 4.13.
The tyre stops spinning after the first impact. There is a high force required to stop it from spinning:
the maximum tangential force is 3160 kg. This high force is a consequence of the fast change in rotational
velocity. The tyre and shock dynamics after the first impact are the same as for the impact without spin. The
influence of the tyre spin is only that the maximum force on the shock is increased to 4.38e5 N. This is only
1 percent higher than a landing without spin-up. The influence of the tyre spin on the landing gear model
response is thus minimal.

4.5.4. Landing gear retraction simulation

To make sure that the landing gear design can actually be stowed in the fuselage a retraction and extension
simulation is performed. There is an actuator installed on the trunnion beam, on the lock links and if applicable on the truck beam. Retraction is performed by applying a predetermined sequence. The gear initial
condition is the completely extended position. A schematic representation of the simulation is available in
section 3.2. The default ode45 solver is used to solve the equations of motion in the simulation. This solver is
efficient and has a medium accuracy that is sufficient for the retraction simulation.
First the truck beam is positioned in retracted position (horizontally). Then the lock actuator is enabled,
which then pushes the lock links upward. Finally the largest actuator rotates the complete assembly inboards.
Rotation stops when the maximum rotation angle (calculated in the stowage analysis) is reached.
The actuator forces are controlled by PID controllers. PID controller are widely used for many industrial
control problems, because the controller structure is simple and performance is satisfactory in many applications (Toscano, 2005).
By adjusting the deploy/retract sequence and the PID parameters the retraction system is given the desired properties. The PID parameters are tuned to achieve a low steady state error percentage (fully reaching
the retraction angle) and are tuned to get no overshoot in the response. The stroke-force curve for the main
retraction actuator is shown in figure 4.15. The retraction angle is almost completely achieved and the steady
state error is 5 percent. There is no overshoot and retraction happens without sudden movements. A drawback of the chosen retraction tuning is that the force changes are large during the stroke of the actuator (see
the stroke force curve). Efficiency of the actuator is thus not high and can be improved if actuator efficiency
proves too low later in the design.
Figure 4.14 and 4.16 displays the retraction and extension motion of an Airbus A320 main landing gear.
The retraction signal is given at 2.5 seconds. Then it takes about 3 seconds to fully retract the gear. Extension
starts at 6 seconds and complete extension and locking finishes about 4 seconds later. At the final stage of
extension it takes some time for the lock actuator to fully lock the landing gear. Retraction is faster than the
required 10 seconds and extension is faster than 15 seconds.
The design of the actuation systems is not completed at this stage. Further optimisations and a more
detailed analysis should be done later in the design process. The results presented here show that the landing
gear design is kinematically feasible and the required retraction and extension motion can be achieved by the
system. Also it is shown that the required stowage space is kept minimal.

4.6. Calculation run time

The calculation time required to run the PositionLandingGear module is in the order of 5 seconds for a 2 strut
main gear and 9 seconds for a 4 strut main gear. Timing tests have been performed on a computer with a

4. Results and verification

tyre z-velocity (m/s)

tyre force (N)

shock force (N)

tyre rotational velocity (rad/s)

78

100
50
0

.
0

0.5

1.5

2.5

3.5

0.5

1.5

2.5

3.5

0.5

1.5

2.5

3.5

0.5

1.5

2.5

3.5

105
0
2
4 .
0
105
2
1
0.
0
0
1
2
3 .
0

Figure 4.13: Landing simulation with tyre spin-up. Tyre spin velocity, shock absorber force, stroke and velocity for a two wheel A320-200
main landing gear landing at 10 ft/s

Figure 4.14: Retraction motion of the multi-body model an Airbus A320 main landing gear

actuator stroke (m)

4.6. Calculation run time

79

0.2
0.4
0.6
0.8 .
0

10

12

14

10

12

14

actuator force (N)

Time (seconds)
4

104

2
0
2 .
0

retraction actuator force (N)

Time (seconds)

104

2
0
2 .
0.8 0.75 0.7 0.65 0.6 0.55 0.5 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1
retraction actuator stroke (m)

Figure 4.15: Retraction actuation stroke versus force of an Airbus A320 main landing gear.

retraction angle (deg)

80
60
40
20
0
.
0

10

12

Time (seconds)
Figure 4.16: Retraction/extension angle of an Airbus A320 main landing gear. The gear is initially in a extended position.

14

80

4. Results and verification

2GHz Intel Core 2 Duo processor. These run times are comparable to other weight estimation modules that
are available within the Initiator. The time required is low enough to be able to use the module within an
optimisation loop.
Profiling of the code shows that most time is spent on the flotation analysis: 55 percent on the rigid ACN
calculation and 9 percent is spent on the flexible ACN calculation. This is due to the large number of iterations
required to get an ACN result. An increase in the number of main gear struts will increase the number of
bogie combinations that need to be analysed for the flotation requirement. This explains the relatively large
increase in calculation time for aircraft with more than 2 struts.
The flotation analysis calculation time has been reduced as much as possible by optimising for the Matlab
JIT accelerator as described in section 3.1.2. A further reduction in calculation time is difficult to achieve.

5
Conclusions
The objective of the master thesis was to expand the knowledge base of the existing automated design environment by creating and integrating an automated landing gear design tool. The automated landing gear
design includes analysis of structures, kinematics, runway flotation and weight.
When finding a feasible position of the nose and main landing gear there are several limiting factors. First
there is the take-off stability requirement that makes sure that the aircraft can safely pitch up or down. During
landing the aircraft should have a pitch down tendency in order to make a safe landing. A crosswind landing
should also be possible without touching the ground and taxi turns need to be possible without the tendency
to tip on its side. Additional ground clearance should be kept between the engines or wing tips for a possible
sideways landing. Loads on the nose gear should be high enough to keep the ability of proper steering. The
main gear also should keep a high enough load to make sure that the brakes can provide enough braking
power. Ground operational requirements dictate that the aircraft should be able to make a 180-degree turn
on a runway. Also airport taxiway turns need to be made by keeping the nose gear on the turn centre line
without going outside the pavement boundary. The combination of all these limits results in a main and nose
landing gear lateral position, a longitudinal position and a landing gear height.
Based on the landing gear position maximum loads on the landing gear struts can be calculated and tyres
can be selected from a catalogue. Accompanying wheels are then also selected. The size of the brakes is
found from the required stopping power to make a rejected take-off stop, several landing stops at design
landing mass and several stops at maximum landing mass. Both steel and carbon brakes are considered.
The design and analysis of the retraction and extension of the gear (kinematics) results in joint and actuator positions. When considering different kinematic schemes, the least complex scheme is preferred, considering safety and maintenance and costs. Retraction of each landing gear is done by rotation about a single
axis while side braces and a locking mechanism are folded in. The geometric kinematic analysis is replaced
by a mathematical analysis or numerical simulation as the landing gear design gets more detailed.
When shocks occur caused by hard landings and by taxiing over rough surfaces they are absorbed efficiently by oleo-pneumatic shock absorbers and tyres. The shock absorbers need to be able to cope with
different load cases provided by certification specifications. To show compliance with these specifications
the piston stroke, the internal pressures and volumes are designed. The oleo-pneumatic shock absorber
forces can then be modelled accordingly and landing load cases can be evaluated. For maintenance, keeping
complexity low is important as maintenance forms an important part of the costs. Reduced complexity will
also keep landing gear weight minimal. And as fuel prices increase, reductions in weight are given higher
priorities.
81

82

5. Conclusions
Empirical relations mainly estimate the landing gear weights. But empirical relations are limitedly de-

pendent on aircraft and landing gear design variables. An analytical weight estimation does not have this
drawback and is therefore better suitable for use in an optimisation loop. Analytically estimating the weight
of the landing gear can be done by generating a structural model and finding maximum stresses when the
structure is under extreme loads. The extreme load cases are prescribed in certification specifications. Yield
stresses should not occur during the entire life of the structure, also buckling should not happen when the
side struts are put under compression. From the maximum stresses, required component thicknesses are
found by including an additional factor of safety.
Based on the information of the previous analyses a multi-body dynamical model can be automatically
made. In this model structural parts are modelled as rigid bodies. The oleo-pneumatic shock absorber forces
and motion are modelled using an analytical relation. And tyre motion and forces at the contact point are
modelled using the empirical relations of the MF-tyre model.
A detailed model of both the shock absorbers and tyres allows the evaluation of extreme landing load
cases and verification of the estimated dynamic landing loads used in the weight estimation. In addition to
the landing simulation a simulation of the landing gear retraction mechanism is done. The simulation forms
a check on kinematic feasibility and a check of compliance to certification requirements.
Implementation of the above-mentioned procedures is done by integrating the landing gear positioning,
bogie design and weight estimation into the TU Delft Initiator. This is done in the form of a landing gear
design module. Additionally a separate empirical module is made that is also applicable to nonconventional
aircraft concepts. And a separate multi-body dynamics simulation program is also implemented that validates the kinematics and performance of the shock absorber and tyres during an extreme landing.
To verify the results produced by the implemented landing gear design modules, landing gear designs
for a number of different aircraft are generated. Calculated landing gear positions closely match with actual
landing gear positions given by aircraft manufactures. The analytical weight estimation of the landing gear
assembly estimates the total gear weight with an error of 15 per cent compared to an empirical weight estimation error of 17 per cent. This result shows that an empirical weight estimation method cannot be fully
replaced by an analytical method. Both methods complement each other, but when using a weight estimation in an optimisation loop the analytical method is preferred due to its highly dependency on landing gear
and aircraft design parameters.
Multi-body simulation results show that dynamic loads during an extreme landing are about 5 per cent
lower than the estimated dynamic loads. A hard landing with a landing gear with multiple rows of tyres creates
high frequency peaks in the shock loads. These peaks originate from the interaction between front and rear
axle tyres hitting the ground at different times. Loading peaks of twice the maximum load for a single axle
gear could occur. Therefore measures need to be taken in the design of multi-axle landing gears to damp
out these high peaks. When adding the effect of tyre spin-up forces, these are found to have a low impact on
overall landing loads.
The kinematic simulation produces bogie retraction/extension time and retraction actuator efficiency,
which can be used to further optimise the landing gear design. The simulation also shows that the retraction
and locking mechanisms work and can be stowed within the available space.
Finally it can be concluded that the implemented landing gear design tools add essential extra information to the overall aircraft design. By not leaving out the landing gear design in the aircraft design, unfeasible
aircraft configurations can be identified early in the design process.

6
Recommendation
There are a number of recommendations for improvement and additional research that can be made. An
improvement would be to increase the level of detail of the landing gear analysis model by adding systems
such as a steering system, a hydraulic actuation system and structural component interfaces.
When more detail is added to the landing gear model, structural component weight estimates can be
further improved. Also adding a finite element analysis of the structure improves confidence in the produced
landing gear result. This also allows for the optimisation of the shape of landing gear components and can
reduce overall weight.
Aerodynamic drag and interference with the flow over the wing is a subject that also is valuable to be
analysed. It can then also be checked if the landing gear is able to withstand aerodynamic loads and can be
extended by gravity in an emergency during the landing.
When the landing gear multi-body model is added to a complete aircraft model further simulations and
verification can be done. A simulation of ground handling or a crosswind landing simulation is a valuable
addition.
For the TU Delft Initiator it can be recommended that the centre of gravity range optimisation needs to be
improved. It can often occur that this range is outside operational bounds, leading to an operational limited
aircraft. Centre of gravity range optimisation is not only important for landing gear positioning, stability and
control of the aircraft during flight is also highly dependent on the cg range.
The analytical weight estimation is currently applicable only to conventional aircraft. Support for nonconventional aircraft can be added by adding additional limitations and checks in the analysis. Added support for evaluation of nonconventional aircraft concepts will then add the capability to perform the design
process within a multidisciplinary optimisation loop.

83

A
Landing gear structure free body
diagrams

85

86

A. Landing gear structure free body diagrams

Raz

May

b
a

Rax

Fside

Ckx

Fky
Chx

Fkx
k

zdrag

Cix

Fhx

Fhy

z'

Flz

Fkz

Rbx

Mby

Fly

Mcy Rcx

Rcz

Rbz

Flx

Rcx

Rbz

Clx

Fdrag

Mcy

x'

May

Raz

Rcz
Mby Rbx

Rax

Fdrag

wb

zside

Fside

Fix

h
g

Fiy

xz-plane

Fiz

Fhz

Mfy
x

Rfx

tt

Rgz

Rfz

Rjz

Rfx
Mfy

Rgx

x'

Rgx

Rjx
Rfz

Rgz

Fhy

Fhx

y
xy-plane: top view
x

Rgx

Rgx

Mgz

Rfx
Rgy

Rfy

Fix

Mbx

Rjy

Rfy

y'

Rjx

Mfz

x'
Rgy

Rbz

Mjz

Mgz

Fside

Rbz
Mbx
z'

phi drag
Mfz

Rbx

Rby

phi side

Fdrag

Fiy

Rby

Fdrag

Rfx

Rby
Fside
Rby
Mbz

Rfy

x'

Mfx
Rbx

y
Rfz

yz-plane

Rfz
Rfy

Mgx + Mjx

Cix

Fgy

Chx

y'

Fiy
Fiz

Figure A.1: Free body diagram of a general 4 wheel main landing gear bogie

Mgx

Fhy
Rgz

Fhz

B
Reference airport pavement classification
numbers

87

88

B. Reference airport pavement classification numbers

Airport

PCN

Subgrade category

Min runway length

Min runway width

London Heathrow

83

3660

50

Amsterdam Schiphol

90

3300

45

New York John F Kennedy Intl

90

3048

45

San Diego Intl

78

2865

61

Rotterdam Zestienhoven

74

2200

45

Trondheim Norway

50

2999

45

Svalbard Norway

58

2483

45

Maastricht Aachen

71

2500

45

Eindhoven

60

3000

45

Groningen Eelde

55

1500

45

Lelystad

19

1250

30

Table B.1: List of airports with flexible pavements and their accompanying Pavement Classification Number (PCN) and associated subgrade category

Airport

PCN

Subgrade category

Min runway length

Min runway width

Atlanta Intl

62

2743

45

Chicago OHare Intl

108

2461

45

Los Angeles Intl

70

3134

45

Kuala Lumpur Intl

90

4000

60

Washington Dulles Intl

81

2865

45

Washington Andrews AFB

54

2973

45

Los Angeles Edwards AFB

64

3658

60

Las Vegas Nellis AFB

43

3085

60

Manchester

79

3050

45

London Stansted

86

3049

46

London Luton

75

2160

46

Prague

62

3250

45

Auckland Intl

120

3535

45

Buffalo Niagara Intl

38

2996

45

Table B.2: List of airports with rigid pavements and their accompanying Pavement Classification Number (PCN) and associated subgrade
category

C
PositionLandingGear output variables

89

90

C. PositionLandingGear output variables

Description

Property

Aircraft Design Group

ADG

auxiliary spar x-position at main gear y-position

AuxSparX

rear spar x-position at main gear y-position

RearSparX

aircraft ACNs flexible pavements

ACNsFlex

aircraft ACNs rigid pavements

ACNsRigid

single brake volume (m3)

BrakeVolume

single brake weight (kg)

BrakeWeight

feasible main gear bogie types

FeasibleBogies

landing gear height w.r.t fuselage belly

Height

landing load factor

LandingLoadFactor

load fraction of MRM at aft cg position

LoadCGaft

load fraction of MRM at forward cg position

LoadCGfwd

wing maximum z-position at main gear y-position

MaxGearSparCxxZ

bogie type with the least number of wheels

MinNWheelsSolution

minimum shock piston length (m)

MinPistonLength

shock piston orifice area (m2)

PistonOrificeA

rim diameter (m)

RimDiameter

rim flange height (m)

RimFlangeHeight

rim weight (kg)

RimWeight

rim width (m)

RimWidth

shock stroke length at static loading (m)

ShockStaticStroke

maximum shock stroke length (m)

ShockStroke

g force when fully compressed

StrutCompressedG

shock absorber p1 pressure

StrutP1

shock absorber p2 pressure

StrutP2

shock piston area (m2)

StrutPistonArea

strut static load (kg)

StrutStaticLoad

shock absorber v1 volume

StrutV1

shock absorber v2 volume

StrutV2

shock absorber v3 volume

StrutV3

total number of wheels on all main gears

TotalNWheelsFeasible

tyre name

Tyre

tyre diameter (m)

TyreDiameter

tyre database row number

TyreId

tyre pressure (kPa)

TyrePressure

tyre rated load (kg)

TyreRatedLoad

tyre diameter at rated load (m)

TyreStaticDiameter

tyre weight (kg)

TyreWeight

tyre width (m)

TyreWidth

wing bogie cg z-position (m)

WingBogieCGz

wing bogie drag strut connection point on wing (m)

WingBogieDStrutSparCxx

structural tubes inner radii (m)

WingBogieInnerR

wing bogie length (m)

WingBogieLength

wing bogie mass (kg)

WingBogieMass

91

Description

Property

wing bogie number of wheel rows

WingBogieNRows

wing bogie number of wheels

WingBogieNWheels

wing bogie number of wheels per row

WingBogieNWheelsPerRow

wing bogie loading MRM fraction

WingBogiePcntLoad

right wing bogie position (m)

WingBogiePositionRight

wing bogie retraction angle (deg)

WingBogieRetractionAngle

wing bogie side strut cylinder connection point (m)

WingBogieSStrutCylCxx

wing bogie side strut wing spar connection point (m)

WingBogieSStrutSparCxx

wing bogie structural tube thicknesses (m)

WingBogieThicknesses

wing bogie weight per structural tube (kg)

WingBogieTotalWeights

wing bogie layout number

WingBogieType

wing bogie tyre track (m)

WingBogieTyreTrack

wing bogie wheelbase (m)

WingBogieWheelBase

shock absorber polytropic N number

polyN

total landing gear mass of least number of wheel solution (kg)

Mass

nacelle height from ground (m)

NacelleHeight

nacelle clearance angle (deg)

Phi_nacelle

wing clearance angle (deg)

Phi_wing

sideways turnover angle (deg)

SidewaysTurnoverAngle4

total tube structure weight per bogie (kg)

TotalStrucWeight

total landing gear mass (kg)

TotalWeight

scrape angle when fuselage hits the ground (deg)

theta

maximum take off scrape angle (deg)

thetaLOF

Table C.1: List of all output variables

D
List of tested reference aircraft
4
2
0
2
4

.
6

10

12

14

16

Figure D.1: Airbus A320-200

10

10

.
10

Figure D.2: Airbus A340-500

93

20

30

18

94

D. List of tested reference aircraft

10

10
.
10

20

30

40

Figure D.3: Airbus A380-800

5
.
5

10

15

20

10

15

20

Figure D.4: Boeing 707-321

5
.
Figure D.5: Boeing 727-200

4
2
0
2
4

.
4

Figure D.6: Boeing 737-200

10

12

14

95

10

10
.
10

20

30

Figure D.7: Boeing 777-300ER

5
0
5
.
5

10

15

20

25

Figure D.8: Boeing 787-800

5
0
5
.
10
Figure D.9: McDonnell Douglas DC10-10

15

20

25

30

Bibliography
The Materials Information Society ASM International.

Asm materials information, 2013.

URL http:

//products.asminternational.org/matinfo/index.jsp.
I.J.M. Besselink. Shimmy of aircraft main landing gears. Masters thesis, Delft University of Technology, Mekelweg 2, Delft, The Netherlands, September 2000.
Boeing 707 Airplane Characteristics, Airport Planning. Boeing, Seattle, WA, revision 3 edition, May 2011.
M. Boschetto, R. Bianco Mengotti, GL Ghiringhelli, and S. Gualdi. Analysis of landing gear behaviour for
trainer aircraft. In Proceedings of the 15th European ADAMS Users Conference, Rome, pages 1516, 2000.
Sonny T. Chai and William H. Mason. Landing gear integration in aircraft conceptual design. Technical
report, Virginia Polytechnic Institute and State University, Blacksburg, Virginia, 24061-0203, March 1997.
URL http://www.dept.aoe.vt.edu/~mason/Mason_f/M96SC.html.
S.T. Chai and W.H. Mason. Landing gear integration in aircraft conceptual design. Technical report, Virginia
Polytechnic Institute and State University, 1996.
John Christy. Flight operations engineering: runway loadings. Technical report, Boeing Commercial Airplanes, 2009.
EB Coetzee, B. Krauskopf, and MH Lowenberg. Nonlinear aircraft ground dynamics. 2006.
H.G. Conway. Landing gear design, volume 3. Chapman & Hall, 1958.
S. Cumnuantip, M. Spieck, and WR Krueger. An approach for sizing and topology optimization integrating
multibody simulation. In 46 th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials
Conference, pages 110, 2005.
N.S. Currey. Aircraft landing gear design: principles and practices. American Institute of Aeronautics and
Astronautics, 370 LEnfant Promenade, S.W., Washington, D.C. 20024, 4th edition, 1988.
R. Doganis. Flying Off Course IV: Airline Economics and Marketing. Taylor & Francis, 2009.
R.J.M. Elmendorp. Synthesis of novel aircraft concepts for future air travel. Masters thesis, Delft University
of Technology Faculty of Aerospace Engineering, January 2014.
European Aviation Safety Agency. Certification specifications and acceptable means of compliance for large
aeroplanes cs-25, July 2012.
Federal Aviation Administration. Comfaa 3.0. software, August 2011. URL http://www.airporttech.tc.

faa.gov/naptf/download/index1.asp.
Federal Aviation Administration.

Federal aviation regulations part 25airworthiness standards: Trans-

port category airplanes, September 2012a. URL http://www.faa.gov/regulations_policies/faa_

regulations/.
97

98

Bibliography

Federal Aviation Administration.

Advisory circular: Airport design ac 150/5300-13a, September 2012b.

URL http://www.faa.gov/airports/resources/advisory_circulars/index.cfm/go/document.

current/documentNumber/150_5300-13/.
Goodrich.

787 electro-mechanical brake system, 2012. URL http://www.goodrich.com/Goodrich/

Businesses/Aircraft-Wheels-and-Brakes/Products/Literature-Listing.

Goodyear. Aircraft tire data book. Technical report, The Goodyear Tire and Rubber Co., 2002.
D.L. Greene. Energy-efficiency improvement potential of commercial aircraft. Annual Review of Energy and
the Environment, 17(1):537573, 1992.
Andy Hebborn. A380 landing gear and systems the feet of the plane. Lecture for RAeS, DGLR, VDI and HAW
Hamburg, June 2008.
RD Hibbeler. Mechanics of materials. 1997.
Honeywell. Boeing 737 NG Wheel and CERAMETALIX Brake. Honeywell, 1944 E, Sky Harbor Circle Phoenix,
AZ, April 2008.
F. Hrlimann, R. Kelm, M. Dugas, K. Oltmann, and G. Kress. Mass estimation of transport aircraft wingbox
structures with a cad/cae-based multidisciplinary process. Aerospace Science and Technology, 15(4):323
333, 2011.
P.R. Kraus. An Analytical Approach to Landing Gear Weight Estimation. SAWE, Incorporated, 1970.
G. La Rocca, T.H.M. Langen, and Y.H.A. Brouwers. The design and engineering engine. towards a modular
system for collaborative aircraft design. Technical report, Delft University of Technology, 2012.
V Ladda and H Struck. Operational loads on landing gear. AD-A239 914, 1991.
LMS

International.

Cessna

uses

lms

virtual.lab

motion

for

the

mak-

ing

of
the
citation
columbus,
2008.
URL
http://www.lmsintl.com/
cessna-citation-columbus-uses-lms-virtual-lab-motion-to-design-biggest-business-jet.
Mathworks. Accelerating matlab: The matlab jit-accelerator. Technical report, The Mathworks Inc., 2002.
Mathworks. Key features - simmechanics for matlab and simulink, 2012. URL http://mathworks.nl/

products/simmechanics/description1.html.
T.H.G. Megson. Aircraft Structures for Engineering Students. Referex Engineering. Edward Arnold, 1999. ISBN
9780340705889. URL http://books.google.nl/books?id=5qTPMgEACAAJ.
Michelin Aircraft Tire. Aircraft tire: engineering data, 2001.
Benjamin Milwitzky and Francis E Cook. Analysis of landing-gear behavior, 1953.
M.C.Y. Niu. Airframe structural design: practical design information and data on aircraft structures, volume 67. Conmilit Press, 1999.
H.B. Pacejka. Tire and vehicle dynamics, volume 372. 2006.
J.J.W. Pinsker. The dynamics of aircraft rotation and lift-off and its implication for tail clearance requirements,
especially with large aircraft. Technical report, Ministry of Technology. Aeronautical Research Council,
1969.

Bibliography

99

D.P. Raymer. Aircraft design: a conceptual approach. American Institute of Aeronautics and Astronautics,
1999.
J. Roskam. Airplane design Part 4: Layout design of landing gear and systems. 1989a.
J. Roskam. Airplane design Part 5: Component weight estimation. 1989b.
R.K. Schmidt. A semi-analytical weight estimation method for oval fuselages in novel aircraft configurations.
Masters thesis, Delft University of Technology Faculty of Aerospace Engineering, November 2013.
A.A. Shabana. Dynamics of multibody systems. Cambridge Univ Pr, 2005.
M. Spieck. Ground dynamics of flexible aircraft in consideration of aerodynamic effects. PhD thesis, Technische
Universitt Mnchen, Universittsbibliothek, 2004.
TNO Automotive. Mf-tyre/mf-swift 6.1.2 equation manual. Technical report, TNO Delft-Tyre, 2010.
TNO Automotive. MF-Tyre/MF-Swift 6.1.2 Help Manual. TNO, March 2012.
E. Torenbeek. Synthesis of subsonic airplane design: an introduction to the preliminary design, of subsonic
general aviation and transport aircraft, with emphasis on layout, aerodynamic design, propulsion, and performance. Springer, 1982.
R Toscano. A simple robust pi/pid controller design via numerical optimization approach. Journal of process
control, 15(1):8188, 2005.
M. Vable. Mechanics of Materials. Oxford University Press, August 2012. ISBN 9780195133370.
H. Zhang, J. Ning, O. Schmelzer, and M.D. Incorporation. Integrated landing gear system retraction/extension
analysis using adams. In North American ADAMS User Conference, June, pages 1921, 2000.