VEHICLE STIFFNESS ANALYSIS

with a focus on Sports Car Structures and a detailed study of the Westfield
Sports Car Spaceframe Chassis

Wayne Prangnell
November 1992
 SUMMARY
The purpose of a vehicle chassis, the different type of vehicle structures and the analysis of vehicle structures
is discussed by way of introduction to a detailed investigation of a Westfield Sports Car space frame chassis.

The bending and torsional stiffness of a spaceframe chassis was tested in the laboratory and was modelled
using finite element analysis software. Laboratory testing was carried out to establish the validity of the finite
element model. The model was then used to investigate methods of improving the torsional stiffness of the
chassis without altering the layout of the car. A number of recommendations were made to improve the
torsional stiffness of the chassis with some simple modifications.

Page 1
 TABLE OF CONTENTS
List of Figures ..................................................................................................................................................4
List of Tables ....................................................................................................................................................6
1 Introduction ..................................................................................................................................................8
1.1 Outline of Project .........................................................................................................................8
2 Background ..................................................................................................................................................9
2.1 Introduction ..................................................................................................................................9
2.2 Purpose of the Chassis .................................................................................................................9
2.3 Basics of Vehicle Structural Actions ............................................................................................10
2.4 Requirements of a Chassis ...........................................................................................................15
2.4.1 Strength Requirements ................................................................................................15
2.4.2 Chassis Stiffness Requirements ..................................................................................17
2.4.3 Determining Torsional Stiffness ................................................................................19
2.5 Relationship of Suspension and Chassis Stiffness .......................................................................20
2.6 Vehicle Structure Analysis ...........................................................................................................20
2.7 Development of the Structure of Sports and Racing Cars ..........................................................22
2.8 Background of Clubman Cars ......................................................................................................28
3 Analysis of the Westfield Sports Car ............................................................................................................30
3.1 Introduction ..................................................................................................................................30
3.2 Determination of Chassis Geometry ............................................................................................31
3.2 Chassis Bending Stiffness ............................................................................................................32
3.2.1 Laboratory Test Description and Procedure................................................................32
3.1.2 Theoretical Analysis Description ................................................................................34
3.3 Chassis Torsional Stiffness ..........................................................................................................36
3.3.1 Laboratory Test Description and Procedure................................................................36
3.3.2 Theoretical Analysis Description ................................................................................38
4 Results and Discussion .................................................................................................................................39
4.1 Bending Test and Bending Analysis ............................................................................................39
4.2 Torsional Test and Torsional Analysis..........................................................................................41
4.3 Torsional Stiffness - Chassis Variations .......................................................................................45
5 Conclusions ..................................................................................................................................................54
5.1 Recommendations ........................................................................................................................57
5.2 Further Study................................................................................................................................58
6 Acknowledgments ........................................................................................................................................59
7 References ....................................................................................................................................................60
8 Appendices ...................................................................................................................................................61
Appendix A - Westfield Sports Car Data............................................................................................61
Appendix B - Westfield Sports Car Chassis Drawing........................................................................63
Appendix C - Computer Model Data File..........................................................................................64
Page 2
Appendix D - Laboratory Testing Observations ................................................................................76
Appendix E - Diagrams and Information for Chassis Modifications ................................................77
Appendix F - Components of the Westfield Sports Cars ...................................................................91
Appendix G - Calculations ................................................................................................................92

Page 3
 LIST OF FIGURES
Figure 2.1 - Chassis Design Considerations ..........................................10
Figure 2.2 - Exploded View of Girder Chassis ......................................12
Figure 2.3 - Twin Tube Chassis of Triumph TR4 with Cross Bracing ..12
Figure 2.4 - Spaceframe Chassis ...........................................................13
Figure 2.5 - Multitubular Chassis ..........................................................13
Figure 2.6 - Torsion Box Stressed Skin Construction, Ford GT40........14
Figure 2.7 - Monocoque Stressed Skin Construction ............................14
Figure 2.8 - Composite Structure of Mass Produced Renault 16 ..........15
Figure 2.9 - 1966 McLaren Grand Prix Car..........................................20
Figure 2.10 - Lola T92/10 Rollbox Model ............................................24
Figure 2.11 - Live Axle, Swing Axle and Independent Suspension ......26
Figure 2.12 - Independent Suspension Attached to Vehicle Structure...27
Figure 2.13 - Lotus Mark Six ................................................................27
Figure 2.14 - Mercédès-Benz 300SL .....................................................28
Figure 2.15 - Mercédès-Benz W196 ......................................................28
Figure 2.16 - Structure of the Lotus 25 Grand Prix Car ........................29
Figure 2.17 - 1989 Ferrari Grand Prix Car (bodywork cut away) .........29
Figure 2.18 - 1988 McLaren MP4/4 GP Car, Bodywork Removed ......30
Figure 2.19 - Modern Cars with Space Frames .....................................31
Figure 2.20 - Monocoque Chassis Road Cars .......................................32
Figure 2.21 - Westfield Sports Car ........................................................33
Figure 2.22 - Ginnetta G2 ......................................................................33
Figure 2.23 - Lotus Seven Body ............................................................34
Figure 2.24 - Elfin Clubman Car ...........................................................35

Figure 3.1 - Layout of Chassis Survey ..................................................38

Figure 3.2 - Axes System ......................................................................39
Figure 3.3 - Chassis Bending Test .........................................................39
Figure 3.4 - PAFEC 34000 Beam Element ............................................41
Figure 3.5 - Standard Chassis Model Member Groups .........................42
Figure 3.6 - Chassis Bending Model .....................................................43
Figure 3.7 - Chassis Torsional Test ........................................................44
Figure 3.8 - Pattern of Loading for Torsional Test ................................45
Figure 3.9 - Chassis Torsional Test Model ............................................46

Figure 4.1 - Load Deflection Response of Chassis Bending .................47

Figure 4.2 - Shape of Chassis for Calculated Bending Test ..................49
Figure 4.3 Torsional Load Deflection Response....................................51
Page 4
Figure 4.4 - Scatter of Measured Torsional Stiffness ............................52
Figure 4.5 - Torsional Deflections Along Chassis .................................53
Figure 4.6 - Torsional Stiffness Plots of Changes to Member Sizes .....55
Figure 4.7 - Torsional Stiffness Plots of Engine Bay Changes ..............56
Figure 4.8 - Torsional Stiffness Plots for Chassis with Extra Bracing ..58
Figure 4.9 - Torsional Stiffness Plots of Centre Tunnel Changes ..........59
Figure 4.10 - Torsional Stiffness Plots for Changes Using Plates .........60
Figure 4.11 - Torsional Stiffness Plots for Other Changes ....................61

Page 5
 LIST OF TABLES
Table 2.1 - Measured Peak Accelerations of Vehicles ...........................18
Table 2.2 - Chassis Torsional Stiffness ..................................................22

Table 4.1 - Standard Chassis Models ....................................................54

Table 4.2 - Category I, Changes to Member Sizes ................................55
Table 4.3 - Category II, Changes to the Engine Bay .............................56
Table 4.4 - Category III, Addition of Bracing Chassis Nose .................58
Table 4.5 - Category IV, Changes to the Centre Tunnel ........................59
Table 4.6 - Category V, Use of Plates ....................................................60
Table 4.7 - Category VI, Other Changes ...............................................61

Page 6
"You have to have passion to go deep inside, where you can then experience
special feelings, very special moments which trigger some of the unique
sensations, unique touch and feelings that give you something extra when you are
right on the limit."

Ayrton Senna, December 1991

The analysis of a vehicle structure takes the designer deep inside, looking for
something extra to give the driver when he is right on the limit.

Page 7
 1 INTRODUCTION
The motor car is an important part of our lives that most of us use every day. Usually considered as a
mechanical product because of all the mechanisms attached to it, the car is also an important structure. The
motor car is subject to such a variation of loads and a severe fatigue life. For efficiency and performance
reasons the car must weigh as little as possible, thus the design and analysis of this structure is essential. The
analysis of vehicle structures is also very important because the public who use cars will tolerate the
occasional mechanical breakdown, but they expect never to have any problems with the structure of the
vehicle regardless of the severity of conditions the vehicle has been subject through its life.

The analysis of vehicle structures is an area where Civil Engineers, or more specifically Structural Engineers
are well equipped to tackle.

The analysis of vehicle structures was researched and a fairly broad overview provided. A detailed stiffness
analysis of the Westfield Sports Car chassis was then carried out using a finite element computer model and
validated with laboratory testing. Potential modifications to the chassis and their effect of vehicle stiffness
using the computer model.

1.1 Outline of Project

Many ideas were pursued with this project and the aims often shifted with new information that was learned
and new ideas, but the basic goal of this project has remained the same: To learn about the structure of
vehicles. The subject of this project was narrowed by the authors interest in motor sport and sports cars which
led to acquaintances with one of two vehicle manufacturers in Western Australia, Westfield Sports Cars
Australia. Stephen Fox from Westfield Sports Cars showed enthusiasm at learning more about the structure of
the sports car that his company produces and he agreed to lend a completed chassis for testing.

Early plans for the project were ambitious and some of the activities planned were: Track testing of the
Westfield Sports car to determine the loads on a vehicle, analysis of the chassis for stiffness, analysis of the
vehicle for stresses and laboratory testing of chassis stiffness and stresses.

Unfortunately track testing was not viable due to the cost of the equipment that would be required such as
strain gauges and high speed multi channel data loggers. With improvements in data logger technology and
availability, measuring the loads on a vehicle may make an interesting project in the future.

On a simpler level, it was attempted to measure vehicle loads with brake meters. Brake meters work on a
principal of lateral accelerations causing an angular deflection of a pendulum in a damping fluid. These
devices which were used by British authorities for testing the brakes of commercial vehicles were found to be
inaccurate for measuring car accelerations as the pitch and roll of a car about its horizontal axes and the slope
of a road visibly affected the angle of the pendulum.

The laboratory program was limited to stiffness testing because strain gauges for measuring stresses were not
able to be supplied and fitted at the University for financial reasons. In hindsight it was sensible to carry out
laboratory testing for stiffness only for reasons of simplicity and the limited time available to the project.
Page 8
To establish computer models for stress and stiffness analysis of the chassis, various data was collected. The
geometry of the chassis was measured using optical surveying techniques in the first instance and then using a
tape measure. Around thirty of the major components of the Westfield Sports car were weighed, measured and
drawn for use with a lumped mass finite element stress model. However the stress analysis did not proceed
because laboratory stress measurements would not have been available to confirm any model results and the
time available would not have allowed the use of a detailed stress model. A model for stiffness analysis was
created and analysed for bending and torsion load cases. A number of variations to the standard Westfield
Sports Car were also investigated.

All models were created for and analysed using the PAFEC finite element software, running on an Apollo
workstation at Curtin University of Technology.

Some of the results of testing and analysis have been interesting, others were what was expected, but the
overall result was learning a lot about vehicle structures and learning of the potential of computer analysis as a
tool for the development of motor vehicle structures.

The author has found this subject very interesting and hopes that this report conveys its information in a way
that will pass on this interest to the reader.

2 BACKGROUND

2.1 Introduction
Information is presented here as background on vehicle structures. The purpose of a vehicle chassis, its effect
on the performance of a vehicle, the different types of vehicle structures and how analysis of the vehicle
structure is approached is explained. The importance of stiffness of a vehicle structure is also discussed in this
section.

The chassis of a vehicle is frequently referred to throughout this project. The intended meaning is the main
structural parts of the vehicle. This does not include suspension components or non structural bodywork, eg
fibreglass cladding.

A background on the structural developments of racing and sports cars is given as racing and sports cars are
usually at the forefront of chassis development. Background on Clubman cars has been included to help
understanding of the analysis of the Westfield Sports Car chassis which is a Clubman car. Clubman is the
name given to a particular style car and this is explained in the background on Clubman cars.

2.2 Purpose of the Chassis

A car chassis may be thought of as a large bracket. This bracket must keep all the parts of the car rigidly in
place for the normal loads to which a car is subjected. Additionally this bracket must protect the driver in
situations of abnormal loading such as crash loading. A summary of considerations for chassis design is given
in Figure 2.1.

Page 9
 Figure 2.1 - Chassis Design Considerations

Strength is required for safety and long life. Rigidity or stiffness is required for servicability reasons to
eliminate low frequency shaking, fatigue problems, door closure problems on uneven ground. For
performance reasons adequate chassis stiffness ensures that the full road holding and handling potential of the
suspension system and tyres is reached.

2.3 Basics of Vehicle Structural Actions

The vehicle structure is required to be strong and stiff in bending between the front and rear wheels and strong
and stiff in torsion between the front and rear wheels. In addition the vehicle structure must have sufficient
strength and stiffness in local areas where loads are applied by components mounted to the structure. These
include loads from the pedals, steering wheel, seats, engine, fuel tank, differential, aerodynamic devices etc.

In dealing with vehicle loads there are a number of structural systems employed by the different types of
chassis. Looking at the predominant structural action, the four main types of structural actions for vehicle
structures are discussed in the following order: i) Beam structures, ii) Framed structures, iii) Stressed
skin construction and iv) Compound structures.

i) Beam structures

Bending and torsional, are carried by relatively thick walled beams. There are usually two beams
longitudinally along the base of the car. Essentially there have been two types of beam structures used
for vehicles. Historically the first type was the conventional girder chassis which consisted of two
longitudinal steel girders of channel section spaced by transverse members of similar construction. A
girder chassis is shown in Figure 2.2.

Vehicles which commonly employ this structural system are trucks. It is unusual to find this structural
system in a new car today.

Page 10
 Figure 2.2 - Exploded View of Girder Chassis
The second type of beam structure is the twin tube or ladder chassis. This chassis has two large section
hollow members joined by lateral or diagonal bracing or a combination of both which increases the
torsional stiffness of the structure. The torsional stiffness of a twin tube chassis is far superior to a
girder chassis of similar weight. A twin tube chassis with diagonal bracing is shown in Figure 2.3.

Figure 2.3 - Twin Tube Chassis of Triumph TR4 with Cross Bracing

Page 11
ii) Framed Structures

Loads are carried by either a partially or nearly fully braced frame system. Because of the complexity of
loading and the number of components which must be accommodated within a car, a fully braced frame
would be impractical and almost impossible. A vehicle with a frame as its main structure is called a
spaceframe when the frame is well triangulated. It is called a multitubular chassis when the frame is
only poorly braced and the loads are carried partly by the bending of the members and joints and partly
by tension and compression in the members. A spaceframe chassis is shown in Figure 2.4 and a
multitubular chassis in Figure 2.5.

Figure 2.4 - Spaceframe Chassis

Figure 2.5 - Multitubular Chassis

Page 12
iii) Stressed skin construction

With stressed skin construction loads are carried by a series of thin walled panels. The panels are
usually stabilised with stiffeners and reinforced locally in regions of high stress such as near suspension
mountings. The panels are most commonly sheet steel or aluminium, moulded glass fibre composites,
carbon fibre composites.

Stressed skin construction can be categorised into two main forms. Firstly those chassis consisting of
two closed boxed sections down either side of the car, essentially a very large diameter, twin tube
chassis. Figure 2.6 illustrates this type of construction.

Figure 2.6 - Torsion Box Stressed Skin Construction, Ford GT40

Secondly chassis which are like a closed top bath tub; a nearly closed single shell with apertures for
driver and engine. This is illustrated in Figure 2.7 by the Lotus 25 structure.

Figure 2.7 - Monocoque Stressed Skin Construction

These forms of chassis have both been called monocoque, unitary, bath tub, torsion box and stressed
skin bodies. Torsion box probably best describes the former, while the latter fits the definition of a
monocoque. Monocoque comes from French: mono- + coque, shell, from Latin.

Page 13
iv) Compound structures

Many vehicles employ a combination of these principal structural actions. Almost all modern
production cars are a composite structure of frame members such as the roof and door pillars and
stressed skins such as the roof, floor and other panels in the engine bay and boot. Figure 2.8 illustrates
with an exploded view of a Renault production car. Commercial vehicles such as buses and coaches
often use a basic frame, very flexible on its own which is stiffened by the addition of exterior body
panels. The structures of many light buses and four wheel drive wagons are similar in principal to this.
Tray backed vehicles usually have two longitudinal beams along their length and a stressed skin cabin,
often with a frame inside the skin. Some of the more limited volume production sports cars are
composite structures with a braced frame, further stiffened and strengthened with stressed panels.

Figure 2.8 - Composite Structure of Mass Produced Renault 16

The loads in a beam structure are carried by flexure of the main beams, in a braced frame system loads are
carried primarily by tension and compression in the members as in a truss. A partially braced frame carries
load by bending moment and tension and compression in the members and with stressed skin construction,
loads are carried by in plane stresses in the skin.

Page 14
2.4 Requirements of a Chassis

2.4.1 Strength Requirements

The strength requirements of any vehicle structure depend upon :

i) The magnitude of the loads to which it is subject

ii) Whether the loads are dynamic or static.

iii) The method of transmission of the loads into the structure.

iv) The variability of the loading.

v) The factor of safety which is required against failure.

Vehicle structure loadings are generally specified in terms of peak accelerations to which the vehicle is subject.
This is independent of the weight of the vehicle thus allowing uniform comparison between loads on cars of
different weight.

The magnitude of peak accelerations to which a vehicle is subject and the use of these accelerations to
determine the loads on a vehicle structure is described by Garrett (1953). He suggests that the worst
combinations of loading which could affect a vehicle structure are represented by four design cases. These
four load cases do not include any consideration for crash loading, a separate area of vehicle design which is
not considered in this report.

The four load cases are based on peak accelerations of 1g for forward acceleration or braking, 1g for lateral
acceleration due to cornering and 3g for vertical acceleration. These accelerations should be multiplied by 1.5
as a safety factor in arriving at maximum loads for design. A safety factor of 1.5 which is relatively high for
steel structures is used due to uncertainty as to the actual magnitude of the loads.

The four loading cases which should be considered are:

i) Hitting a bump / kerbing while braking in a straight line.

ii) Cornering.

iii) Hitting a bump while accelerating straight ahead.

iv) Hitting a bump while cornering.

Costin and Phipps (1965) used a similar approach in an example of the design of a racing sports car chassis.
The peak accelerations and the safety factor used in arriving at loads they used were identical to those
suggested by Garrett.

Other methods of analysing loads include determining serviceability loads from spring and damper actions and
and analysis of the tyre / road interface. Vertical loading for normal operation would be through determining
the relationship between compression of the spring damper unit and loading which is a combination of simple
elastic deformation of the spring plus dynamic force from the damper unit, whereas strength or ultimate load
cases are invariably outside of the normal spring and suspension movement range and may be more
dependent on other factors including tyre, bump stop and bushing deformation. Analysing the tyre contact
Page 15
patch and tyre deflection may provide an envelope of the longitudnal, lateral and vertical forces potentially
transmitted from the road into the vehicle.

With steady improvements in suspension design, tyre properties, aerodynamics and downforce, vehicle weight,
torsional stiffness and engine outputs, the peak accelerations that a modern production car, sports car or racing
car can generate has increased. Examples of the peak accelerations that can be generated by several different
cars are given in Table 2.1.

Table 2.1 - Measured Peak Accelerations of Vehicles

Peak Accelerations, g
Longitudinal Lateral .
Wheels, October 1992 - Road and race Nissan GTR's
Standard Nissan GTR .87 1.17
Australian Group A Nissan GTR 1.22 1.46

Wheels, May 1992

Ferrari 1992 Formula 1 Car 3.4 4.31
Ferrari F40 1.17 1.29
Ferrari 348 1.14 1.0

Wheels, October 1991 - Australian Group A Racing Cars

Holden Commodore 1.0 1.4
Nissan GTR 1.0 1.7

Motor, July 1991 - Tyre Testing Feature

Nissan 300zx, Dunlop D40 M2 225/50 ZR16 Tyres
Dry, low friction, smooth concrete track 1.09 1.01

Wheels, May 1991 - Handling Test

Tested with Valentine Research Inc. G-analyst
BMW M5 N/A 1.05
Ferrari Mondial t348 N/A 1.03
Honda NSX N/A 1.07
Nissan GTR N/A 1.10
Porsche Carrera 4 N/A 1.06

L J K Setright1968
1953 Grand Prix Car N/A 0.7
1965 Grand Prix Car N/A 1.4

Note that comparisons between these results will not be accurate as factors such as coefficient of friction
of road, temperature, test circuit and driver are variables.

Page 16
Modern cars, in particular sports and racing cars commonly achieve higher peak accelerations than those given
in references by Garrett (1953) and Costin and Phipps (1965). Consequently the loads on a modern car will be
larger, however may be slightly offset by reduction of the safety factor due to:

i) More accurate determination of loads using modern measuring equipment

ii) Better understanding of the vehicle structure through computer structural analysis techniques

iii) Better quality control of materials and manufacturing processes used in vehicle construction

This project is concerned with chassis stiffness rather than strength. This discussion about vehicle loads is
intended to provide background for vehicle strength. Strength analysis of vehicle structures requires further
more detailed information of the loads that apply to a particular vehicle and a range of different load cases
apply ranging from serviceability where fatigue stress is a primary consideration to impact loads where
controlled failure and permanent plastic deformation occurs.

Further discussion focuses on chassis stiffness which can be analysed independently of loads and stresses,
thus detailed load cases are not further developed.

2.4.2 Chassis Stiffness Requirements

Chassis stiffness is important in any vehicle for reasons such as door aperture tolerance, durability of fitments,
occupant comfort and impression of safety, but most importantly for a performance car, chassis stiffness is
fundamental to cornering performance.

Bending stiffness of a chassis is typically expressed as a maximum vertical deflection of a chassis resulting
from a certain mid span load. Fenton (1980) suggests that the maximum deflection for a 680 kg mid span load
should be 1.27 mm. Fenton has not discussed the type or weight range of the vehicles that this would be
applicable to and this would be necessary where deflection is the design criteria. For instance if two cars meet
the requirement of a maximum mid span deflection of 1.27 mm for a 680 kg load, yet one of these cars is very
heavy and the other is much lighter, the in service deflections of the heavier car will be larger approximately in
proportion to the difference in weight.

Chassis torsional stiffness is expressed in vehicle publications and by automotive engineers as the amount of
torque required to twist the chassis one degree over the length of its wheelbase. Metric units are Nm/degree
and imperial units are ft.lb/degree. This expression of chassis stiffness is independent of the wheelbase of the
car, allowing direct comparison between cars of different length.

Page 17
To obtain good road holding and handling, the suspension geometry of a car is carefully designed and often
refined to sub millimetre accuracy. For the suspension system to be most effective, the mounting points for
the suspension on the chassis must be held rigidly in place by a stiff chassis. For this reason torsional stiffness
is most important. Torsional stiffness is almost always more important than bending stiffness for performance
reasons because in bending there is very little deflection at the supports, which in the case of a vehicle are the
suspension mountings. However with torsional deformation the maximum deflections are likely to occur at
the suspension mounting points.

A good illustration of the effect of chassis torsional stiffness is documented by Setright (1968). The 1966
McLaren Grand Prix car was uncompetitive with the leading teams of that season because the engine was
large, heavy and underpowered. However the chassis which was designed by young Aerospace Engineer,
Robin Herd, which had an aluminium skin over a balsa wood core was of exceptional torsional stiffness. The
McLaren had a torsional stiffness of about 13500 Nm/deg compared to about 3300 Nm/deg for a competitive
Lotus 33. Setright timed the McLaren of Bruce McLaren shown in Figure 2.9 through Hunzberg Corner at the
Zandervoort Grand Prix circuit faster than that achieved by Jim Clark in his Lotus although Clark's lap times
were significantly faster than McLaren's.

Figure 2.9 - 1966 McLaren Grand Prix Car

Another illustration of the importance of chassis torsional stiffness was the Porsche 904 Bergspyder developed
for the 1965 European Hillclimb Championship. Its structure was very poor for torsional loads and as a result
the handling was erratic and the car was called 'Kangaroo'. Heavy modifications were needed to make the car
competitive (Cotton 1988).

Page 18
2.4.3 Determining Torsional Stiffness
There are two common strategies for measuring the torsional stiffness of a chassis which are:

i) Fit solid bars in place of spring and shock absorber units and mount either the front or rear suspension
uprights to a rigid datum. Measure the torque required to twist the unrestrained end of the car one
degree, or a similar measured amount. Thus the chassis stiffness is deduced from the rotation of the
unrestrained suspension uprights for the particular torque applied.

ii) Reasonably restrain one end of the chassis from rotation about the centreline at its suspension mounting
points at that end. Apply a known torque to the unrestrained end of the chassis through the chassis
mounting points, measuring the rotations on the chassis at the front and the rear. The chassis stiffness
may be deduced from the relative rotation of the unrestrained end of the chassis to the restrained end for
the torque applied.

The measured torsional stiffness of the chassis may vary if the torsional stiffness is calculated from rotations at
the ends of suspension members as in i) or if the stiffness is calculated from rotations measured at the front and
rear of the actual chassis as in ii).

Table 2.2 summarises literature review of recommendations and observations for chassis torsional stiffness.

Table 2.2 - Chassis Torsional Stiffness

Source Vehicle Recommendation or

Observation
Setright 1968 1962 Lotus 25/33 GP car 3300 Nm/deg
(basic structure weighed 32kg complete)

Setright 1968 1966 Brabham GP car about 1400 Nm/deg

multitubular chassis

Setright 1968 1966 McLaren GP car over 13 500 Nm/deg

chassis of thin aluminium alloy, chemically bonded
to end grain balsa core.

Fenton 1980 typical family saloon, Minimum: 6100 Nm/deg

Recommended: 6500 - 7500 Nm/deg

Webb 1984 family size saloon most cars range 4000 -

9000 Nm/deg

Gard 1992 Ford Falcon EBII 1992 8200 Nm/deg

Campbell 1978 Lotus Elan (about 1963) backbone chassis only 6870 Nm/deg

Fenton 1980 Ford GT40 13560 Nm/deg

Page 19
 Fothergill 1984 Open sports car 4000 Nm/deg is design
aim

light road racing car 4070 Nm/deg suitable

Gard 1992 any performance car ideally 10 times the

suspension roll stiffness

Gard 1992 early 1990's F1 estimated 35 000 to 45

000 Nm/deg

Qld Govt. Low volume cars: 4 cyl 4000 Nm/deg

(Road registration) 6 cyl 6000 Nm/deg
8 cyl 8000 Nm/deg

Gard 1992 English kit Cobra 300 Nm/deg

RMC racing Cobra 8300 Nm/deg
Jaguar 2013 2014 F Type aluminium body Jaguar

2.5 Relationship of Suspension and Chassis Stiffness

An improvement in roadholding and cornering performance may be possible by increasing the stiffness of the
suspension, but often increasing the spring stiffness gives no improvement or even worse overall performance.
The reason for this may be that the torsional stiffness of the chassis has not been considered. For instance
where springs are already quite stiff, or the chassis is quite flexible much of the suspension movement may be
as a result of flexure of the chassis and in such a case stiffer springs are unlikely to increase cornering capacity.

The other problem of fitting stiffer springs, also associated with the torsional stiffness of the car is that stiffer
springs transfer bigger loads into the chassis resulting in larger chassis deflections. When these deflections
become large enough to affect a carefully designed suspension geometry, cornering performance will be lost.
In order to achieve the full potential of the suspension system and the tyres, the torsional stiffness of the
chassis should be ten times the roll stiffness of the suspension. The roll stiffness of the suspension is the
torsional stiffness of the car minus the flexibility of the chassis, measured at the wheel positions with springs
in place and the suspension movement unrestricted.

2.6 Vehicle Structure Analysis

Traditional engineering statics and mechanics formula can be applied relatively easily to early beam and
tubular chassis however manual methods become more difficult with complex three dimensional geometry of

Page 20
space frames, stressed skin and composite construction. Model or prototype testing provided an early solution
but testing numerous iterations of loads and design options is time consuming and expensive.

With modern vehicle design and construction, the structural actions of vehicle structures have become quite
complex. Sports and racing car structures are usually either thin walled (shell) structures or of spaceframe
construction. With many members, joints, load paths and a high degree of indeterminacy, manual analysis
becomes difficult, often impractical and more susceptible to errors. This is further aggravated by the
variability of loading imposed on a vehicle. Thus analysis of these structures has tended toward approximate
analysis techniques, design by rules of thumb, prototype or model analysis (not usually thought of as an
analysis technique). More detailed calculations for localised areas which are important parts of the structure
are usually made. Approximate methods of analysis for thin wall vehicle structures, based on predicting onset
of shear instability are described by Bruhn (1958).

With developments in computer hardware in the 80's and 90's, computers with adequate power for structural
analysis have become more affordable and most major car companies are now using computer finite element
techniques to improve and refine their structural designs. Finite element analysis is also becoming a powerful
design tool for a number of specialist vehicle constructors. For instance Lola Cars who are leading English
constructors of sports and racing car chassis have used finite element analysis in the design of the composite
structure for the rollbox for crash loads of the T92/10 sports car (Baxter-Smallwood J. , Advanced Composite
Engineering, 1992). Figure 2.10 is a diagram of the model used for the analysis of the Lola rollbox. Stress
contours are superimposed on the model.

Figure 2.10 - Lola T92/10 Rollbox Model

The finite element method has become one of the most important tools for vehicle structural analysis. It deals
with structural analysis problems by breaking down a structure which consists of a few complex elements into
many simple elements. Mathematical solution of the systems of simple elements by recognised stiffness
matrix techniques is readily performed using commercial software and a workstation or powerful personal
computer. Current developments in personal computers is leading software companies to produce affordable
finite element software for the PC, effectively bringing this tool within the reach of most engineers.

A finite element software package generally consists of modules which:

i) Assist with creating a model of the structure to be analysed. Interactive graphics is preferable, and
graphics essential for checking input.

Page 21
 ii) Perform the solution

iii) Communicate the results of the analysis. Interactive graphics could be considered essential for this
purpose.

Factors which make the analysis of vehicle structures complex have been discussed, but a factor which can
greatly simplify the structural design and analysis of many vehicles is that the design of a vehicle structure is
usually controlled by deflections rather than stresses. That is, if the car is designed to achieve a suitable
stiffness, the stresses will be below safe limits. This simplifies design because interpreting the results of a
stiffness analysis is generally much simpler and less time consuming than with the analysis of stresses. Hence
most analysis work during the design stage is for stiffness and a stress analysis is carried to check the final
structural details.

2.7 Development of the Structure of Sports and Racing Cars

The historical course of the development of the car chassis has been led in the past by racing and sports car
designers who have either failed or achieved glory in applying new technology and new ways of thinking to
their car designs. In the past it was thought that the car engine embodied the main technology in the car, but
in this era of motor car development and with the benefit of hindsight, the importance of the role that the
vehicle chassis has played in successful cars can be seen. The following brief history pays particular attention
to developments in Grand Prix racing, as this is seen as the show case for automotive technology.

The earliest cars were built on a steel girder frame which supported a timber body. It didn't matter whether the
car was a Grand Prix racing car or a family saloon, the structural action of chassis was the same. This
technology had come straight from the coach building industry and it was generally believed that a degree of
flexing of the chassis was a necessary part of the suspension. If built along substantial lines, the girder
chassis possessed adequate bending stiffness, but its torsional stiffness was very poor. The conventional girder
chassis consisted of two longitudinal steel girders of channel section spaced by transverse members of similar
construction. This was used almost exclusively in sports and racing cars up until the 1930's.

Even racing cars are subject to Newton's laws of motion, and so it is that a heavy racing car requires more
power to accelerate and brake and has a greater desire to continue in a straight line when the driver is trying to
turn a corner. In pursuit of better performance from their racing cars, designers recognised the need to reduce
weight. As these early chassis were particularly heavy for their strength and stiffness, the chassis was an ideal
place to reduce weight.

The move to tubular ladder chassis was led by racing car designers when in 1934 the German Auto Union
team introduced a Grand Prix racing car with a twin tube chassis, Mercédès-Benz also introducing a chassis of
similar layout that year. This considerably increased the torsional stiffness of the chassis with minimal change
in the bending stiffness.

The types of suspension in use at the time, namely live axles and later swing axles, were not dependent on a
stiff chassis to preserve the suspension geometry. These suspension types have the wheels connected to axles

Page 22
and the wheel and axle assembly moves a single unit. Torsional deformation of the car structure has little
effect on the wheel angles whereas the mechanism of wishbone independent suspensions rely on the relative
positions of suspension member pivots to determine the angular positions of the wheels. Figure 2.11 shows
typical independent wishbone, swing axle and live axle suspension systems.

Figure 2.11 - Live Axle, Swing Axle and Independent Suspension

Around 1934 came the application of independent suspension to racing cars. Whereas before this the angular
relationship of the wheels was determined by a live axle acting as a beam joining the wheels, now the car itself
was part of the structure required to preserve the angular relationship of the wheels. Figure 2.12 is a simplified
diagram of the connection of independent suspension to the vehicle structure.

Page 23
 Figure 2.12 - Independent Suspension Attached to Vehicle Structure

The need for increased torsional stiffness was recognised by designers and attempts were made to improve the
basic ladder chassis with extra tubular super structure, however this was generally not effective. It was not
until 1952 that two new sports cars that were to be very successful, designed on space frame principals
appeared; the Lotus Mark Six, see Figure 2.13 and the Mercédès-Benz 300SL, see Figure 2.14.

Figure 2.13 - Lotus Mark Six

Figure 2.14 - Mercédès-Benz 300SL

Spaceframe chassis construction was introduced to Grand Prix racing in 1954 by the Mercédès-Benz W196
which had a significant weight and stiffness advantage over rivals whose car structure were based on the
ladder chassis.

Page 24
 Figure 2.15 - Mercédès-Benz W196

The spaceframe chassis and multitubular chassis were used exclusively in Grand Prix racing until in 1962.

Fuelled by the desire to win races, the search for further chassis stiffness and light weight brought about the
introduction of the stressed skin construction Lotus 25 Grand Prix car. The Lotus 25, later becoming the Lotus
33 with its stressed skin structure achieved a torsional stiffness of around 2 to 2½ times that of the a
conventional Grand Prix. It also achieved a typical weight saving of around 10 kg. The benefits of weight
saving, excellent torsional stiffness and improved driver safety offered by this form of construction were soon
recognised and followed by the majority of Grand Prix teams. The basic structure of the Lotus 25 Grand Prix
car is shown in Figure 2.16.

Figure 2.16 - Structure of the Lotus 25 Grand Prix Car

The excellent stiffness and strength to weight ratio achievable with stressed skin construction currently sees all
Grand Prix teams building their racing cars this way. It has also proven ideal for construction with new
materials that have since become available such as aluminium honeycomb and currently carbon fibre. Figures
2.17 and 2.18 show modern Grand Prix car chassis.

Page 25
 Figure 2.17 - 1989 Ferrari Grand Prix Car (bodywork cut away on top)

Figure 2.18 - 1988 McLaren MP4/4 Grand Prix Car, Bodywork Removed

In Australia today, ladder chassis for cars are common only in go-karts, vintage cars and some drag racing
cars. Space frame chassis are popular for many types of racing cars, for example; Formula Ford and Formula
Vee are restricted to tubular steel construction, Clubman racing cars must be of the "space frame" type and
many original sports cars and sports sedans use space frame chassis. Many of the kit cars that are available in
Australia are of space frame type construction such as Westfield Sports Car and the PRB Clubman and the AT
Riciardi. In contrast to these budget sports cars is the Lamborghini Diablo, currently one of the fastest road
cars it employs a space frame chassis (Sports Car World, 1990/91). Two cars with aluminium space frame
chassis currently undergoing development for mass production are the Pininfarina Ethos (Motor, 1992) and the
Audi Avus (Chiton's Automotive Industries, 1992).

Audi Avus

Page 26
 Pininfarina Ethos
Figure 2.19 - Modern Cars with Space Frames

Aluminium monocoques are required in Australia's premier open wheeler racing category, Formula Holden,
and carbon fibre monocoques are found in Formula 2, Formula Libre and Sports Sedan cars in Australia.
Carbon fibre monocoques provide the basic structure for many of the latest breed of supercars such as Ferrari
F40, Jaguar XJ220, Bugatti EB110, McLaren F1 and the Yamaha OX99-11.

Yamaha OX99-11

McLaren F1

Jaguar XJ220
Figure 2.20 - Monocoque Chassis Road Cars

Page 27
2.8 Background of Clubman Cars
As part of this project theoretical and laboratory analysis of the Westfield Sports Car chassis was carried out.
As this is a clubman style car, background on clubman cars has been included in this report.

The basic formula that defines a clubman car would be: A front longitudinally mounted engine, two seats in
an open cockpit with no doors, live rear axle, multitubular space frame chassis and front wheels separate from
the main body of the car. A typical car is shown in Figure 2.21.

Figure 2.21 - Westfield Sports Car

Two of the earliest clubman cars were the Lotus Mark 6, which was being produced in 1954 and the Ginnetta
G2 which was put into production in 1958. Based around multitubular space frames with aluminium body
panels, these cars were designed to provide an unprecedented level of performance at a price affordable to the
average motoring enthusiast. Their appointments were sparse, with little concession to comfort. They were
suitable for transport during the week and could perform well on the racing track or in trials at the weekend.

Figure 2.22 - Ginnetta G2

Many specials' constructors and limited production manufacturers have since produced similar clubman cars,
some copies of the more recognised designs, others of more original design, but the principals of the clubman
have led to these cars often looking similar and usually performing well.

The structural design of these cars is often very similar, many being based on a Lotus design for the Lotus
Seven which first appeared in 1957. Since this time engine power outputs have risen, the price of steel has
dropped, spring rates of the suspension have risen and there have been significant advances in tyre technology.
Hence there is the desire to improve the chassis to gain the most advantage from these changes.

Page 28
 Figure 2.23 - Lotus Seven Body

Elfin Sports Cars first produced the clubman car shown in Figure 2.20 in Australia in 1962. Currently
clubman cars are available in Australia in kit form from Westfield Sports Cars (WA), PRB Motors (NSW),
Tilke Engineering (NSW) and Fraser Cars Ltd (New Zealand). Specifications and general information
concerning the Westfield Sports Car is included in Appendix A.

Figure 2.24 - Elfin Clubman Car

Page 29
 3 ANALYSIS OF THE WESTFIELD SPORTS CAR

3.1 Introduction
To investigate vehicle chassis analysis, a Westfield SEi chassis supplied by Westfield Sports Cars Australia
was used. The stiffness of this chassis was investigated using a finite element computer model. The computer
model was validated using the results of laboratory testing. Two tests were carried out for evaluation of the
computer model.

i) Bending stiffness of the chassis.

ii) Torsional stiffness of the chassis.

The effect of variations on the torsional stiffness of the chassis was investigated using a computer model. The
model was created for, and analysed with PAFEC finite element software on an Apollo workstation. Variations
that were tested were aimed at either improving the torsional stiffness of the chassis or reducing construction
costs.

Only the stiffness of the chassis was investigated because of the following reasons:

i) The strength of the Westfield Sports Car has been well proven

ii) Measurement of stresses is expensive and was beyond the finances available to this project

iii) Computer stress analysis of a vehicle requires a much more complicated model than does stiffness
analysis. The number of load cases that must be considered for stress analysis also extends the time
required to set up and analyse a model.

iv) Stresses predicted by a model can only be as accurate as the loads that are used. To determine loads with
reasonable accuracy would require special measuring equipment, unavailable to this project. Alternatively
loads may be used as determined from other peoples work, however it appeared that the references that
were available (Garrett 1953 and Costin and Phipps 1965) were somewhat dated as are the analysis
methods that were used when these load cases were first suggested.

v) Developments in suspension and tyre technology mean that the cornering performance of the car is likely
to benefit from improved chassis stiffness.

Page 30
A quick overview of the activities that were involved in testing and analysis follows:

Laboratory Testing
Construction of sub frames for attachment of chassis to testing frame.
Setting up the chassis, loading devices and measuring equipment for testing.
Carrying out the test.
Recording observations and the results of the test for later analysis and scrutiny.

Theoretical Model Analysis

Determining the geometry of the model
Creating a data file that describes the geometry of the chassis, member and section properties, loads and
restraints.
Checking the data for errors
Analysis of the model
Interpretation of the results of analysis.

3.2 Determination of Chassis Geometry

The physical characteristics of the chassis were required for a theoretical model of the chassis to be generated.
Information such as section types and plate thicknesses were available from the management of Westfield's,
however no plans or drawings of the chassis were available. Two methods of determining chassis geometry
were considered:

i) Survey using optical surveying instruments.

ii) Tape measure, measuring from reference beams.

At the time it was thought that an optical survey would provide the most accurate measurement of the chassis
geometry, so with the assistance of Associate Professor L. A. White a survey of the chassis was commenced.

Two theodolites, two subtense bars and the chassis were layed out as shown in Figure 3.1. A subtense bar is a
bar with markings accurately calibrated to two metres.

Page 31
 Figure 3.1 - Layout of Chassis Survey

Vertical and horizontal angles to each end of the subtense bars and several of the nodes were observed and the
distance between stations A and B was measured. This data enabled calculation of positions of the nodes in
three dimensional space. However the chassis survey technique of measurement was found unsuitable for the
following reasons:

i) Making the observations was very time consuming.

ii) A check on measurements determined by the survey with a tape showed errors of 4 to 7mm.

The time consuming nature of the theodolite observations and the large errors were partly to the level of skill
of the operator.

The geometry of the chassis was subsequently measured using a tape. Beams were clamped to the chassis to
act as a reference for measurements. A one fifth scale orthogonal drawing was produced as a reference for
further work. A copy of this drawing has been included in Appendix B.

The geometry of the theoretical model was compiled into standard file format for the PAFEC finite element
software by typing the node coordinates, member connectivities and other information defining loads,
restraints and member properties. A graphics interface was used for checking that information was correct.
The following diagram, Figure 3.2 shows the global axes of the model. This is the axes system used
consistently in this report.

Figure 3.2 - Axes System

3.2 Chassis Bending Stiffness

3.2.1 Laboratory Test Description and Procedure

The chassis bending test of the Westfield Sports Car involved simply supporting the chassis on its front and
rear extremities as shown in Figure 3.3 and applying loads near the middle of the chassis while the deflections
at known positions were measured.

Page 32
 Figure 3.3 - Chassis Bending Test

The chassis was placed on timber blocks in the four restraint positions shown in Figure 3.3. The blocks were
supported on a smooth concrete floor. Dial gauges were set up to measure deflections at nodes 12 and 24
relative to the concrete floor. A load hanger was placed midway between nodes 151 and 154 on which dead
weights were placed.

The following steps were carried out during testing:

i) The chassis was first proof loaded to with 50 kg to bed in the chassis at the supports and to ensure the
chassis was sitting evenly on its supports.

ii) The proof load was removed and dial gauge readings were observed at nodes 12 and 24.

iii) A load of 10 kg was applied and dial gauge readings at nodes 12 and 24 were observed. This was
repeated for loads of 10 kg, 20 kg, 40 kg, 50 kg and 60 kg. Observations were made as the load was
increased to 60 kg and then reduced in the same increments back to zero.

iv) Dial gauge deflections were then observed for a loading pattern of 0 kg, 50 kg, 0 kg, 50 kg, 0 kg.

v) The average deflections of the gauges at nodes 12 and 24 were calculated for the load increments.
These are plotted in Figure 4.1 in the results section.

Page 33
3.1.2 Theoretical Analysis Description
The chassis model was created using primarily PAFEC type 34000 beam elements. These elements are
straight uniform beams with two nodes. They cater for bending in two principal directions, axial forces and
twisting about the shear centre. The beam section is described by second moments of area Iyy and Izz, area A
and torsional constant C (same as J). The flexural and shear centres of this element must coincide. Six
degrees of freedom are modelled at each of the two nodes of this element; Ux, Uy, Uz, φx, φy and φz. Figure
3.4 shows the degrees of freedom of the 34000 element.

Figure 3.4 - PAFEC 34000 Beam Element

The theoretical basis of this model is that bending displacements in each direction vary as a cubic along the
length, giving linearly varying bending moments. Axial force and twisting moment are constant along the
length. Results produced by this element are exact in statics. The beam members in the model are represented
by the member centre lines.

Other elements used in the model were the suspension members which were only axial force elements and
plate elements for the engine and gearbox mounts. The plate elements were capable of accepting both in plane
and out of plane forces.

The elements of the model were separated into eleven groups of similar members with the same sectional
properties. This was done to enable quick and simple specification and changing of member properties. A
diagram showing the node numbering and the data file which defined the chassis model is included in
Appendix C. Figure 3.5 on the following page shows the member groups and their colours with brief
explanations of each of the groups used in the model.

Page 34
 Figure 3.5 - Standard Chassis Model Member Groups

The restraints and loading used for the chassis bending analysis as a model of the chassis bending test are
shown in Figure 3.6.

Figure 3.6 - Chassis Bending Model

A load of 1000 N was applied mid way between nodes 151 and 154. The chosen value of the load was not
important, just that the value was known because the theory used to analyse the model assumed linear elastic
response.

Page 35
3.3 Chassis Torsional Stiffness

3.3.1 Laboratory Test Description and Procedure

The torsional stiffness test was carried out in the Curtin University of Technology Civil Engineering concrete
laboratory, using a substantial reaction frame designed for testing concrete beams. The chassis was supported
on the frame at three points; one front and two rear simulated wishbones. The connections to the wishbones
were retained by loosely fitting bolts that allowed the connections to act as joints pinned in three dimensions.
The simulated wishbones along with solid links replacing the spring and shock absorber units distribute test
loads into the chassis similarly to how loads in a car are transmitted through the suspension.

At the rear of the chassis, restraint from rotation about the longitudinal X axis was provided by the wishbone
connections. At the front of the chassis the one wishbone connection prevented mechanistic rotation of the
chassis and facilitated the application of a torque loading to the front of the chassis.

The load was applied at the unsupported front wishbone and deflections of the chassis were measured by dial
gauges supported from the reaction frame. Loads were applied as steel dead weights in 20 kg and 18.1 kg
increments with deflections observed at selected node points for each increment. Figure 3.7 shows the chassis
with simulated wishbones and the loading and restraint conditions.

Figure 3.7 - Chassis Torsional Test

The test was carried out on two occasions; after the first test was carried out, inspection of the results indicated
inconsistent load - deflection behaviour. To rectify this two steps were taken before and during the second
test:

i) The removal of a pulley through which the load was initially being applied. The pulley allowed the
vertical load to be applied upwards, however it appeared that the pulley was jamming when load was
applied.

Page 36
ii) After incrementing the load, the chassis was tapped with a small block of wood until the dial gauge
readings became steady, before deflections were observed.

The testing procedure was carried out as follows:

i) The chassis and subframes were set up in the test frame as shown in Figure 3.7.

ii) Dial gauges were set up at selected nodes on the chassis. Two dial gauges at the front measured rotation
at the front while two dial gauges at the rear measured rotation of the rear of the chassis. Deflections
were measured at the rear of the chassis because although the rear was prevented from mechanistic
rotation, elastic deformations of the sub frame and wishbones allowed some rotation.

iii) Distances between dial gauges and the load lever arm were measured with a tape.

iv) A 60 kg proof load was applied to settle in the chassis at its supports. Loads were applied at the front
wishbone as shown in Figure 3.7.

v) Dial gauge readings were observed before loading was commenced. Loads were then applied and
removed in the pattern shown in Figure 3.8. Before observing dial gauge readings and after the load
was applied, the chassis was tapped with a small block of wood until dial gauge readings stabilised.

vi) Measurements of distance between dial gauges and load lever arm distance were checked.

A full set of observations from the laboratory tests is given in Appendix D.

Figure 3.8 - Pattern of Loading for Torsional Test

Page 37
3.3.2 Theoretical Analysis Description
The chassis model was the same as that used for the bending analysis but the loads and supports were changed
to model the torsion test conditions. The restraints and loading used for the chassis torsional analysis are
shown in Figure 3.9.

Figure 3.9 - Chassis Torsional Test Model

The load applied to the front wishbone was 1000 N. Where deflections for other loads are required the
deflections obtained from analysis with the 1000 N load may be scaled directly proportionally to the change of
load because the computer model was linear.

Page 38
 4 RESULTS AND DISCUSSION

4.1 Bending Test and Bending Analysis

The chassis bending test was a simple test which was carried out to help establish the accuracy of the
theoretical model of the chassis.

The chassis was loaded between nodes 151 and 154 with deflections at nodes 12 and 24 being observed.
(These nodes have been identified previously in Figure 3.6). The average vertical deflections of nodes 12 and
24 are plotted for different loads in Figure 4.1.

Figure 4.1 - Load Deflection Response of Chassis Bending

Page 39
The relationship between load and deflection observed in the test followed a linear pattern with a coefficient of
correlation, r2 = 0.995. The measured stiffness of the chassis in bending is less than the calculated stiffness by
about 11%. The linear relationship between load and deflection from Figure 4.1 being:

Observed stiffness: Δ = load ∗ 0.70

Theoretical stiffness:
Δ = load ∗ 0.62

where Δ = deflection in mm, load = load in kN

The linear relationship for observed stiffness ignores the permanent set which is labelled in Figure 4.1. The
permanent set was probably the result of local crushing of the timber supports where high spots of welds were
in contact with the supports.

The difference in stiffness between the model and the measured stiffness of 11% may be due at least partly to
the following reason. The chassis was supported on timber blocks into which stresses were transmitted across
the grain. These timber blocks were quite thick and supported the chassis on a relatively small bearing area.
As any deflections at the supports will be reflected in the measured deflections, compression of the supports
will result in measured stiffness apparently less than the actual stiffness of the chassis. A quick check for
elastic compression of the supports indicates that up to 1 - 3% of the difference may be as a result of
deformation in the supports. This does not include local deformation of the supports which may occur where
there are local high spots in the chassis above the supports.

Other reasons for the model stiffness differing from the observed stiffness were: The supports used in the test
may not have been properly level. Some member eccentricities were difficult to include in the model and
where they were not large they were excluded. This would increase the effectiveness of the bracing and
slightly increase the stiffness of the model.

Angular deformation of the members to which the dial gauges were attached may have caused increases to the
observed deflections.

The shape of the deflected chassis undergoing bending was determined from the model. Deflections for the
bottom plane outer edge members are plotted in Figure 4.2 for a load of 1000 N between nodes 151 and 154.

Page 40
 Figure 4.2 - Shape of Chassis for Calculated Bending Test

The average deflections of nodes 12 and 24 are shown in Figure 4.2 by a dashed line. The maximum vertical
deflection of the chassis between front and rear wheel centres of 0.26mm is shown on Figure 4.2.

The observed vertical deflection of 0.26 mm per 1000 N can be linearly extrapolated to 1.73 mm per 680 kg
mid span load. This compares to a recommended bending stiffness of not more than 1.27mm deflection for a
mid span load of 680 kg by Fenton (1980). The bending stiffness achieved by the Westfield Sports Car chassis
is obviously less than that recommended by Fenton, however Fenton's recommendations include no discussion
on the weight of the vehicle for which his recommendation is made. It would be logical to include the weight
of a car in a recommendation for bending stiffness as the bending deflections are likely to increase
proportionally to the weight of the car.

The sharp change in stiffness graph of Figure 4.2 at point A is as a consequence of the presentation of the data
for this graph. Point A is an external node on the bottom plane of the chassis, point B is a node on the same
member as point A, but it is closer to the longitudinal centreline of the chassis and directly under the rear
support. Also there is no reason for concern over a sharp decrease in the stiffness of the chassis in this position
because this part of the structure is outside of the wheelbase of the car and only subject to small loads.

4.2 Torsional Test and Torsional Analysis

The chassis torsional stiffness test was carried out to establish the accuracy of the theoretical model of the
chassis.

As mentioned earlier the torsional stiffness test was carried out on two occasions. In the first instance there
was large scatter of the results and virtually no consistency. This was thought to be the result of applying the
load through a pulley in the first test. The pulley was not of good quality and although it appeared to run
Page 41
smoothly and freely when unloaded, it was likely that the pulley was binding against the pulley shaft when
load was applied. Thus the pulley was discarded for the second test.

Another precaution that was taken for the second test was to tap several times checking that dial gauge
positions did not fluctuate before dial gauge readings were recorded.

The results of the first torsional test are not included in this report because due to their inconsistent nature,
they are of little use.

The torsional deflection response for the second torsion test of the Westfield Sports Car chassis is shown in
Figure 4.3. The load deflection response calculated from the chassis model is also shown on this graph for
comparison with the measured response.

Figure 4.3 Torsional Load Deflection Response

The results of the second torsional stiffness test show very little scatter. A response which is clearly linear may
be observed.

For the torsional stiffness test, deflections were measured at the front and rear of the chassis on each side of
the chassis at nodes 102 and 123 at the front and nodes 105 and 111 at the rear. Figure 3.9 previously defined
these node numbers.

To calculate the torsional stiffness, the rotation at the rear of the chassis was subtracted from the rotation at the
front of the chassis. The torque applied at the front of the chassis was calculated from the magnitude and lever
arm of the load. Thus the torsional stiffness was the Torque applied divided by the rotation between the front
and rear of the chassis.

Page 42
The torsional stiffness of the chassis in the test calculated in this way was 1134 Nm/deg on average (see Figure
4.4) and the stiffness of the model was 1121 Nm/deg.

Figure 4.4 - Scatter of Measured Torsional Stiffness

For the torsional test errors such as error of measurement of the chassis geometry, approximations in the model
by ignoring some eccentricities and error in reading dial gauges should be the same as those for the bending
stiffness test.

Page 43
The difference between the measured torsional stiffness of 1134 Nm/deg and the model torsional stiffness of
1121 Nm/deg of 1.2% was extremely good and suggests that the model was a good representation of the
chassis. The range of torsional stiffness observed during testing was from +5.4% to -10.4% of the model
stiffness. The biggest difference of 10.4% between model and test is still within the average difference
observed for the bending test..

When the torsional stiffness of the model was calculated from the rotation of the front suspension wishbones
with the rear wishbone restrained from movement by the supports, the torsional stiffness of the chassis was
found to be 1050 Nm/degree. The difference between the two calculated stiffnesses is due to the position of
the load relative to where the stiffness was measured. Measuring the torsional stiffness from the wishbones
resulted in an apparently more flexible structure because the loads and supports were attached directly to the
wishbones. Nodes 102, 123, 111 and 115 were away from the loads and supports which were the most highly
stressed regions, thus the measured stiffness was higher.

The graph of angular deflections along the chassis in Figure 4.5 highlights the most flexible areas of the
chassis. The most flexible areas, which are where the curve is steepest are the first 70mm from the front of the
chassis and 200mm to 500mm from the front of the chassis which is in the engine bay area. The stiffest parts
of the chassis is the 250mm directly behind the hoop on which the steering wheel is mounted and the very
front of the chassis, after the first 70mm and where there is corner bracing in the front, top plane of the engine
bay.

If the entire chassis was able to be increased to the same stiffness as directly behind the steering hoop, the
chassis would have a torsional stiffness of over 2000 Nm/deg.

Figure 4.5 - Torsional Deflections Along Chassis

Page 44
4.3 Torsional Stiffness - Chassis Variations
A number of variations to the standard chassis have been considered. Mostly these variations are intended to
be suitable for production at some time in the future, however some less practical variations have been
considered on the basis that they may help understanding of the structural actions of the chassis. Table 4.1 to
Table 4.7 describe the various changes made and the effect of these changes on the torsional stiffness and
weight of the chassis. Following each table is a graph with the torsional stiffness along the length of the
chassis plotted. In each case Car1, the standard chassis configuration is included as a reference. The
variations to the chassis are shown graphically in Appendix E along with information about masses, centres of
mass and moments of inertia of the chassis variations.

The types of variations to the basic chassis structure are grouped together according to the type of change
which was made. In general terms the changes which resulted in a worthwhile increase in chassis stiffness
were extra centre tunnel bracing, increased member section sizes with same or even reduced wall thicknesses,
extra engine bay bracing and extra bracing in the nose. The changes which were least desirable were the
removal of the existing main engine bay brace and attaching steel plates to various areas such as the front of
the drive train tunnel and the sides of the engine bay.

Table 4.1 - Standard Chassis Models

File Description Torsional Weight % Change from Stiffness
Stiffness (kg) Car1 to Weight
(Nm/deg) Stiffness Weight Ratio
Car1 Standard chassis with minimum three
point restraint
Torsional stiffness calculated as per 1121 63.3
laboratory test
Torsional stiffness calculated from 1050 63.3 0.0 0.0 16.6
wishbones deflections

Car20 As Car1, but using PIGS generated 1050 63.3 0.0 0.0 16.6
data file (as a check)
Car23 This file models the bending stiffness 63.3 0.0
test

Hereafter all files are the same as the standard Westfield Sports Car chassis except for those variations
specified. Minimum three point restraint and torsional stiffness calculated from deflections at the wishbones is
used consistently.

Page 45
 Table 4.2 - Category I, Changes to Member Sizes
File Description Torsional Weight % Change from Stiffness
Stiffness (kg) Car1 to Weight
(Nm/deg) Stiffness Weight Ratio
Car2 Top plane members changed to 31.8 x 1152 64.4 9.7 1.7 17.9
1.6 SHS
Car3 Bottom plane members changed to 1112 65.2 5.9 3.0 17.1
31.8 x 1.6 SHS
Car 16 Top and bottom plane members 2051 71.0 95.3 12.2 28.9
changed to 40 x 1.2 SHS
Car26 Bottom side members changed to 40 x 1185 67.0 12.9 5.8 17.7
1.6 SHS
Car27 Bottom and top side members changed 1412 69.4 34.5 9.6 20.3
to 40 x 1.6 SHS
Car29 All member changed to 40x40x1.0 2845 74.6 171 17.9 38.1
SHS

Figure 4.6 - Torsional Stiffness Plots of Changes to Member Sizes

Changes to member sizes produced the biggest increases in torsional stiffness to weight ratio when the section
sizes of the members were increased significantly and the wall thicknesses of the hollow members decreased.
Comparing changes of the top longitudinal members to changes to the bottom longitudinal members showed
that changes to the top longitudinal members produced a more pronounced effect on torsional stiffness.

Page 46
 Table 4.3 - Category II, Changes to the Engine Bay
File Description Torsional Weight % Change from Stiffness
Stiffness (kg) Car1 to Weight
(Nm/deg) Stiffness Weight Ratio
Car4 Extra brace in top plane of engine bay 1261 63.8 20.1 0.8 19.8
Car5 Side bracing in engine bay changed 1040 63.5 -1.0 0.3 16.4
Car6 Extra lateral member across top of 1105 64.0 5.2 1.1 17.3
engine bay
Car9 Extra top, right hand engine bay brace 1060 63.8 1.0 0.8 16.6
Car10 Normal engine bay brace removed 683 62.5 -35.0 -1.3 10.9
Car11 Normal engine bay brace replaced by 1148 63.5 9.3 0.3 18.1
LH and RH braces
Car15 Engine bay side braces replaced by 1061 66.9 1.0 5.7 15.9
1mm steel panels
Car 18 Extra cross members in engine bay 1475 64.9 40.5 2.5 22.7

Figure 4.7 - Torsional Stiffness Plots of Engine Bay Changes

Page 47
The changes around the engine bay consisted of changes to the top plane bracing, changes to the bracing in the
side of the engine bay and use of plates instead of bracing in the sides.

Additional bracing in the top plane, correctly positioned achieved an excellent increase in torsional stiffness
for a simple change. When the added bracing was not well positioned only insignificant increases to torsional
stiffness were observed. The removal of the main top plane engine bay brace caused a dramatic reduction in
the torsional stiffness of the chassis.

Changes to the the bracing in the side of the engine bay was carried out so that the degree of triangulation was
not reduced. Consequently there was no large changes to the torsional stiffness for the variations analysed.

Using plates instead of bracing was a solution which increased the weight of the chassis with no significant
gain in torsional stiffness.

Page 48
 Table 4.4 - Category III, Addition of bracing to Chassis Nose
File Description Torsional Weight % Change from Stiffness
Stiffness (kg) Car1 to Weight
(Nm/deg) Stiffness Weight Ratio
Car8 Nose fully braced (some of this may 1453 65.8 38.4 3.9 22.1
not be practical)
Car17 Extra nose bracing 1334 65.0 27.0 2.7 20.5
Car24 Some extra nose bracing 1149 64.0 9.4 1.1 18.0
Car25 Some extra nose bracing 1215 64.6 15.7 2.1 18.8

Figure 4.8 - Torsional Stiffness Plots for Chassis with Extra Bracing

Bracing added to the nose of the chassis produced some worthwile increases in torsional stiffness. The fully
braced nosed produced a 38% increse in torsional stiffness, and while this may not be practical for production,
each of the other changes that were more suitable for production resulted in an increased stiffness to weight
ratio.

The nose is a relatively short part of the overall length of the chassis so the magnitude of these increases in
stiffness was unexpected.

Page 49
 Table 4.5 - Category IV, Changes to the Centre Tunnel
File Description Torsional Weight % Change from Stiffness
Stiffness (kg) Car1 to Weight
(Nm/deg) Stiffness Weight Ratio
Car12 1mm steel plates added to front of 1157 71.1 10.2 12.3 16.3
tunnel
Car19 Extra centre tunnel bracing, except on 1450 67.6 38.1 6.8 21.4
bottom plane
Car21 Extra centre tunnel bracing all around 1504 68.4 43.2 8.1 22.0
Car28 Tunnel members changed to 25 x 1.6 1183 65.9 12.7 4.1 18.0
SHS

Figure 4.9 - Torsional Stiffness Plots of Centre Tunnel Changes

Fully bracing the centre tunnel section, which had external dimensions of 220mm x 140mm over the main part
of its length produced a 43% increase in torsional stiffness. This shows that the centre tunnel section plays an
important part in the overall torsional stiffness of the chassis. Fully bracing the centre tunnel may not be
practical in that it would restrict access to the driveshaft which passes through this tunnel. Bracing the tunnel
on all but the bottom plane also resulted in a large increase in torsional stiffness of 38%. Welding plates to the
tunnel section was a solution which was not analysed but could be worthwhile considering in the future.

Table 4.6 - Category V, Use of Plates

File Description Torsional Weight % Change from Stiffness
Stiffness (kg) Car1 to Weight
(Nm/deg) Stiffness Weight Ratio
Car12 1mm steel plates added to front of 1157 71.1 10.2 12.3 16.3
tunnel
Altered engine support beams webbed 1179 64.4 12.3 1.7 18.3
with 1mm plate
Car14
Page 50
Car15 Engine bay side braces replaced by 1061 66.9 1.0 5.7 15.9
1mm steel panels

Figure 4.10 - Torsional Stiffness Plots for Changes Using Plates

The increases in torsional stiffness that were achieved by adding plates to the chassis were generally offset by
the increases in weight and reduction in accessibility that resulted from using plates. The plated engine
support beams gave a torsional stiffness increase of 12% but the engine support beams were also modified in
this case. The modified engine support beams increased torsional stiffness by 11% without the use of plates.
These results should not be construed to suggest that plate solutions will not be viable, rather that the
variations which were tested were not particularly viable.

Page 51
 Table 4.7 - Category VI, Other Changes
File Description Torsional Weight % Change from Stiffness
Stiffness (kg) Car1 to Weight
(Nm/deg) Stiffness Weight Ratio
Car7 Member added under gearbox (Node 1066 63.6 1.5 0.5 16.8
61 to 63)
Car13 Geometry of engine support beams 1163 63.3 10.8 0.0 18.4
altered
Car14 Altered engine support beams webbed 1179 64.4 12.3 1.7 18.3
with 1mm plate

Figure 4.11 - Torsional Stiffness Plots for Other Changes

The modified engine support beams gave an 11% increase of torsional stiffness for a relatively simple change
in geometry and no increase in weight. The member under the gearbox was a change that would be difficult to
put into production. The effect of this member was minimal.

Page 52
The changes that have been considered are only a sample of the changes that could be considered in a serious
effort to improve the torsional stiffness of this chassis. For instance no additional bracing was considered for
the rear part of the chassis. The changes that were considered targeted the more flexible areas of the chassis,
as indicated by the torsional stiffness diagram, Figure 4.5 where changes could be made without disrupting the
layout of the chassis. Where positive improvements to the chassis have been determined, these changes could
be refined by further analysis with the computer model.

Page 53
 5 CONCLUSIONS
The first objective of this project that was carried out was to investigate the use of a computer model for
predicting the stiffness of a vehicle chassis. The space frame type chassis of the Westfield Sports Car was
modelled using finite element software. The model consisted of mainly beam elements and it was found that a
beam model predicted the stiffness of the chassis with good accuracy.

The accuracy of the computer model was established from carrying out the same tests both in the laboratory
and with the computer model and comparing the results. Two distinct tests were carried out that were bending
stiffness and torsional stiffness.

The bending stiffness test showed that the computer model was within 11% of the laboratory observations.
This difference may be partly accounted for by a problem with the way the laboratory test was carried out.
The chassis was supported on compressible timber supports and because this was not foreseen as a problem,
no measurement of the deflections at the supports were taken during the test. Consequently the results of the
test indicated the chassis was more flexible than it actually was. The computer model was stiffer than the
results of the bending test so at least some of the 11% difference was due to compression at the supports in the
laboratory test.

The torsional stiffness test did not have the problems of the bending test and comparison between model and
average measured stiffnesses was of the order of 1%. The torsional stiffness was measured a number of times
for different loads and the range of all measurements was about 10% if the extremes are discarded. The 1%
difference between the model and average observed stiffness is well within this 10% range. Hence the
accuracy to which the model can be confirmed is limited by the accuracy of test measurements.

The load deflection response of the chassis was consistent and linear for both laboratory tests which further
confirms the use of a simple beam model based on linear elastic theory.

The second objective that was carried out was to make use of the computer model for testing changes to the
chassis. The torsional stiffness only was tested for the chassis modifications because it was desirable for the
torsional stiffness of the chassis to be improved, where as the bending stiffness was already adequate.

Page 54
There were a number of changes that Westfield Sports Cars may test further on a full scale chassis and there
were a number of other changes that may be beneficial. The changes fell into five broad categories:

i) Changes to member sizes

ii) Addition or removal of bracing in the engine bay

iii) Addition of bracing to the nose of the chassis

iv) Addition of members to the centre tunnel

v) Use of plates instead of bracing

vi) Other changes

The first category which was changes to member sizes showed excellent improvements to torsional stiffness
for a minimal weight penalty with increased section sizes. Reducing the wall thicknesses of the hollow
members when the section sizes were increased minimised increases of weight in the chassis. The most
pronounced effects of changing member sizes were observed where members in areas with a lack of bracing
were changed such as the top and bottom plane members in the engine bay and cockpit. The largest increase
of stiffness of all the changes analysed, changing all members to 40x40x1.0 SHS, was in this category.
Although this change could not be directly incorporated into manufacture of new Westfield’s because there is
physically not enough space in some places for these larger members, it demonstrates the efficiency of larger
section sizes and smaller wall thicknesses for this type of chassis.

The second category which was the addition and removal of bracing in the engine bay showed that the bracing
in the top plane of the engine bay was very significant. The removal of the existing main engine bay brace in
the top plane reduced torsional stiffness by 35% while adding a second main engine bay brace in the top plane
increased the torsional stiffness by 20%. No major changes were made to the side bracing of the engine bay
but presumably there is little potential for increased torsional stiffness by adding bracing to the sides because
the sides are already well braced. Significant decreases in torsional stiffness would be likely if the side
bracing of the engine bay is partly or wholly removed.

Page 55
Bracing in the nose of the chassis was the third category and some worthwhile increases in torsional stiffness
were obtained with nose bracing. However at the nose of the car there are other requirements which limit the
use of bracing. Steering arms protrude through the side of the nose and the engine cooling system limits
bracing in the front plane. Thus the benefits of nose cone bracing indicated by the model would not achievable
on a finished car.

Extra bracing of the centre tunnel was very effective in increasing the torsional stiffness of the model. The
largest increase in torsional stiffness was observed when the centre tunnel was braced on the sides and top and
bottom planes. The torsional stiffness increased by 43% for only an 8% increase in weight for this case.
Changing the member section sizes of the tunnel increased torsional stiffness by 13% and while this increase is
much less than that achieved with extra bracing, it would require less labour for construction than would the
extra bracing. If additional bracing and increased section size is applied to the centre tunnel, the individual
increases are very unlikely to be cumulative. The reason why increased member sizes gave such an increase in
torsional stiffness was because the centre tunnel was poorly braced. With a well braced tunnel, section sizes
will be much less significant.

The use of plates in the place of hollow section bracing was generally not structurally advantageous. Flat
plates 1mm thick were analysed with the model. Where plates were used instead of hollow section bracing,
the plated solution was heavier and the model suggested only a small increase in torsional stiffness. Plates of
1mm thickness were used because thinner plates may have been difficult to weld. It may be beneficial to
investigate the use of thinner plates and also the use of profiled plates. Plate solutions have the advantage that
they are simpler to fabricate than tubular members when no services are required to pass through the plate and
disadvantages of restricting access through the chassis and difficulty of fabrication where services are required
to pass through the plate.

The final category considered two unrelated changes; addition of a member under the gearbox and alteration
of the engine support beams. The member under the gearbox would be a difficult member to add in practice
and had little effect on torsional stiffness. Alteration of the engine support beams did not affect the weight of
the chassis but produced a worthwhile increase in torsional stiffness of 11%. This change would be relatively
simple and it should be considered for incorporation into production. The further addition of 1mm steel plates
to the altered engine support beams had little effect.

Only a small number of the possible changes to the chassis have been analysed. With the work in creating and
verifying the model already done, it would not require much more work to investigate numerous other
changes. Testing a large number of changes to the chassis would have been a relatively simple task but in the
context of this project it was seen as more advantageous to concentrate on establishing the accuracy of the
model and showing how this was done. The results of a smaller number of variations could also be presented
in more detail, making the use of a computer model for testing vehicle structures more clear.

With the importance of light weight and good stiffness for a car chassis and the high cost of building a testing
prototype vehicles computer model analysis is likely to be cost effective method of determining upgrades to an
existing chassis. Computer model analysis can be effective for large production and special production
vehicles alike.
Page 56
5.1 Recommendations
From analysis of the computer model, there are a number of changes to the chassis that should be investigated
for immediate inclusion into the production of the Westfield Sports Car chassis. The most practical and
effective changes were:

i) Additional top plane engine bay brace

ii) Additional bracing of the centre tunnel

iii) Increased top plane member section size with same or reduced wall thickness.

iv) Geometry of the engine support beams altered

v) Extra nose bracing. Suitable bracing geometry may be determined by investigation of complete car
and further analysis with the computer model.

In general for a structure of this type the stiffness will be increased for any given weight when section sizes are
increased and wall thicknesses decreased. Such changes should be subject to further investigation to
determine if welding thinner walls will cause a problem and if particular wall thicknesses are required for
withstanding rust, abrasion and local stresses around mounting brackets such as suspension mounting brackets.

A recommendation not associated with the analysis of the Westfield Sports Car, comes from applying an
engineering knowledge to the background information given in this report. It very beneficial to consider the
weight of vehicle as well as the vehicle's purpose or engine size when recommending or legislating for
stiffness of the vehicle. Whether torsional stiffness or bending stiffness is considered, the reason stiffness is
required is to limit deflections. The deflections of a structure are just as dependent on the applied loads as the
stiffness of the structure. In the case of a vehicle, the loads are as a direct result of the weight of the vehicle,
thus any sensible recommendations or legislation for vehicle stiffness should include consideration for vehicle
weight.

Page 57
5.2 Further Study
With the limited amount of time and resources available for a final year project, there remains much on the
analysis of vehicle structures that could be investigated of the structure of the Westfield Sports Car. This
project should provide a good background for any further work investigating vehicle loads, structures and
dynamic response. The original intention to carry out a detailed stress analysis was not carried out, however
diagrams were drawn and masses measured for around thirty components, covering the major components of
the Westfield Sports Car. A lumped mass model from this data was partially completed for stress analysis
work and could be made available. Information about the components of the Westfield Sports Car has been
included in Appendix F.

Page 58
 6 ACKNOWLEDGMENTS
Atkinson, Scott Mr Atkinson provided a great deal of assistance with numerous
computing problems, writing programs for manipulating data, fixing up
sick computers, processing information for plotting and printing and
transferring information between different computers.

Chandler Dr I Dr Chandler was always prepared to talk about ideas and

problems encountered in the course of the project. I have enjoyed these
discussions, often being infected with his positive attitude. His time spent
sifting through computer print outs or laboratory results to try a find where
I had gone wrong was invaluable. Dr Chandler has also helped to improve
the standard of this report by his efforts in proof reading several draft
copies, making valuable comments.

Fox, Stephen I wish to thank Mr Fox for his interest in my project and specially
thank him for supplying the Westfield Sports Car chassis on which my
project was based. It has also been a privilege talking to Mr Fox and
learning from his experiences building road and racing cars.

Gard, Jaime I would like to thank Mr Gard for the interest he has shown in my
project and for the time he has spent discussing vehicles and vehicle
structures with me, in particular his work chassis torsional testing and race
car suspensions. Mr Gard has also spent time checking the accuracy of
information in my report and I greatly appreciate this and express my
sincere thanks.

Kong, Paul Mr Kong has helped me by checking computer models for errors.
I would like to thank him for this.

Curtin University of Technology Civil Engineering laboratory technical staff

The laboratory staff have provided assistance in carrying out laboratory
tests, different to the testing normally carried out in the Curtin Civil
laboratories.

Sceresini, Robert I would like to thank Mr Sceresini for helping with physical
problems such as moving chassis and lifting engines.

Associate Professor L.A. White I extend my thanks for assistance in carrying out the survey of the chassis.

Page 59
 7 REFERENCES
Baxter-Smallwood, J (1992) "FEA gets Lola rolling", Advanced Composites Engineering June

Beermann, H J (1989) The Analysis of Commercial Vehicle Structures, Mechanical Engineering Publications
Ltd, London.

Bruhn, E F (1958) Analysis and Design of Flight Vehicle Structures

Bureau of Transport and Communications Economics (1990) Cost of Road Crashes in Australia - 1988

Campbell, Colin (1973) Design of Racing Sports Cars, Robert Bentley Inc., Cambridge.

Campbell, Colin (1978) The Sports Car,: Its Design and Performance, Robert Bentley Inc., Cambridge.

Carey J (1991) "The G Force", Wheels Magazine, May

Carey, J (1992) "Max Factor", Wheels Magazine, May

Costin, M and Phipps, D (1965) Racing and Sports Car Chassis Design, B. T. Batsford Ltd, London

Cotton, M (1988) Classic Porsche Racing Cars, Patrick Stephens Ltd, England.

Coulter, J (1986) The Lotus and Caterham Sevens, Motor Racing Publications Ltd., England.

Crombac, G (1986) Colin Chapman. The Man and His Cars, Patrick Stephens Ltd, England.

Dubensky, R G (1986) What Every Engineer Should Know About Finite Element Analysis Methods, Chrysler
Motors Corp.

Federal Office of Road Safety (1989) Australian Design Rules for Motor Vehicles and Trailers, Third Edition,
Federal Department of Transport and Communications.

Fenton, J (1980) Vehicle Layout and Analysis, Mechanical Engineering Publications, London.

Fothergill, D J, Southall, R, Osmond, E, (1984) "Computer Aided Concept Design of a Sports Car Chassis
System", Proceedings of Institution of Mechanical Engineers

Gard, J (1992) Oral Communication

Garrett, K (1953) "Automobile Dynamic Loads", Automobile Engineer, February

Garrett, T K (1953) "Structure Design", Automobile Engineer, March/April

General Motors Holden's (1990) FEM of Motor Body Structures, ACADS Seminar, 25 June 1990

Greenway, W R, (1969/70) "Automobile Body Testing Techniques", Proceedings of Institution of Mechanical

Engineers

Lake, B (1992) "Budget Barnstormers", Motor Magazine, September

McCarthy, M, (1987) Great Australian Sports Cars and Specials, Australian Consolidated Press, Sydney.

National Council of CAMS, (19992) CAMS 1992 Manual of Motor Sport 1992, Confederation of Australian
Motor Sport.

Page 60
Niemi, E, Makelainen, P (1990) Tubular Structures, Third International Symposium,

PAFEC Limited (1984) Data Preparation User Manual Level 6.1

PAFEC Limited (1984) PAFEC Theory

Page, E (1991) "Two Way Stretch", Wheels Magazine, October

Palmer Tube Mills Australia Pty Ltd (1991) Catalogue 1991, Palmer Tube Mills, Australia.

Rose, J (1988) Ginetta The Illustrated History, Haynes Publications Inc., California.

Setright, LJK (1968) The Grand Prix Car 1954-1966, George Allen and Unwin Ltd, London.

Setright, LJK (1976) The Designers, Weidenfeld and Nicolson, London.

Sturz, W D, (1990/91) "Hell Fire", Sports Car World, Summer pp 14-21

Timishenko and Gere (1968) Elements of Strength of Materials,

Webb, G G (1984) "Torsional Stiffness of Passenger Cars", Proceedings of Institution of Mechanical

Engineers

Westerman, A (1991) "Tyre Supertest", Motor Magazine, July

Westfield Sports Cars Ltd, Westfield SE - SEi Instruction Manual, Westfield Sports Cars Ltd

Williams, G (1991) McLaren. A Racing History, The Crowood Press Ltd, Wiltshire.

8 APPENDICES

Appendix A - Westfield Sports Car Data

Motor magazine recently conducted tests of four clubman cars available and able to be licensed in Australia
(September 1992). The following information about the Westfield Sports Car is sourced from tests conducted
by Motor.

Kits sold in Australia 60

Cars registered in Australia 14
Engine Front, longitudinally mounted
1.6L, 88kW (Toyota Corolla)
Suspension front - independent double wishbones
rear - double wishbones or live axle
Tyres Yokohama A-008R, 205/60 R13 85H
Wheelbase 2270mm
Front Track 1310mm
Rear Track 1330mm
Overall Length 3515mm
Overall Width 1580mm
Height 1040mm
Ground Clearance 105mm
Page 61
Kerb Weight 580kg
Weight/Power 6.6 kg/kW
Acceleration 0 - 100m 6.53s
Standing 400m 14.85s

Member Properties of Westfield Sports Car Chassis

Tubemakers B.T.M. Square Hollow Sections

Section Size mm Wall Area kg/m Ixx mm4 J mm4

Thickness mm²
mm
2020 20 1.6 111 0.873 6080 10300
2525 25 1.6 143 1.12 12800 21200

Page 62
Appendix B - Westfield Sports Car Chassis Drawing

Page 63
Appendix C - Computer Model Data File

(diagrams showing nodes and elements at end of data file listing)

C
C Standard Chassis
C Generally units are in Newtons, N and metres, m
CONTROL
CONTROL.END
C
C .......1.........2.........3.........4.........5.........6.........7.........8
BEAMS
MATERIAL=1
C (NOTE THAT SECTION.NUM IS THAT REFERRED TO BY PROPERTY NO. IN ELEMENTS
MODULE)
SECTI IYY IZZ TORSION AREA KY KZ ZY ZZ
1 12.8E-9 12.8E-9 21.2E-9 143E-6 .9 .9 816E-9 816E-9 C BOTTOM
PLANE
2 6.08E-9 6.08E-9 10.3E-9 111E-6 .9 .9 608E-9 608E-9 C CENTRE
TUNNEL
3 12.8E-9 12.8E-9 21.2E-9 143E-6 .9 .9 816E-9 816E-9 C TOP PLANE
4 5.36E-9 5.36E-9 10.7E-9 103E-6 .9 .9 487E-9 487E-9 C
5 6.08E-9 6.08E-9 10.3E-9 111E-6 .9 .9 608E-9 608E-9 C
6 5.36E-9 5.36E-9 10.7E-9 103E-6 .9 .9 487E-9 487E-9 C
7 6.08E-9 6.08E-9 10.3E-9 111E-6 .9 .9 608E-9 608E-9 C SUSP'N
MEMBERS
8 12.8E-9 12.8E-9 21.2E-9 143E-6 .9 .9 816E-9 816E-9 C UPRIGHTS
9 16.1E-9 16.1E-9 1.0E-9 111E-6 .9 .9 800E-9 800E-9 C BRACKETS
10 16E-9 810E-9 20E-9 492E-6 .3 .7 100E-9 12.1E-6 C FLOORPANS
C
C
MATERIAL
MATE.NUM E NU RO
1 200E9 0.3 7850
C
C
C *** THE NODES MODULE IS PRINTED
C *** USING GLOBAL CARTESIAN AXES
C
NODES
NODE X Y Z
C
C NODES ON PERIMETER OF BOTTOM PLANE
C
1 0 0.000 0.000
3 0.000 -.123 0.000
4 0.000 -.145 0.000
5 .196 -.182 0.000
6 .257 -.199 0.000
7 .341 -.224 0.000
8 .499 -.272 0.000
9 .585 -.297 0.000
10 .962 -.411 0.000
12 1.553 -.517 0.000
13 2.307 -.517 0.000
14 2.293 -.517 0.000
16 2.930 -.442 .127
17 2.930 -.126 .127
Page 64
 18 2.930 0.000 .127
19 2.930 -.456 .127
20 2.930 .126 .127
21 2.930 .442 .127
22 2.930 .456 .127
23 2.307 .517 0
36 2.293 .517 0
24 1.553 .517 0
26 .962 .411 0
27 .585 .297 0
28 .499 .272 0
29 .341 .224 0
30 .257 .199 0
31 .196 .182 0
32 0 .145 0
33 0 .123 0
34 1.809 -.517 0
35 1.809 .517 0
38 .542 -.2845 0
37 .542 .2845 0
C
C OTHER NODES ON BOTTOM PLANE
C
50 .257 -.180 .003
51 .341 -.203 .003
52 .499 -.213 .003
53 .585 -.218 .003
54 .962 -.243 .003
55 .962 .253 .003
56 .585 .218 .003
57 .499 .213 .003
58 .341 .203 .003
59 .257 .180 .003
60 .962 -.162 0
61 .962 -.156 -.003
62 .962 .107 0
63 .962 .097 -.003
64 2.306 -.432 -.003
65 2.306 -.140 -.003
66 2.306 -.124 -.003
67 2.306 -.088 -.003
68 2.306 .088 -.003
69 2.306 .124 -.003
70 2.306 .140 -.003
71 2.306 .432 -.003
72 .542 -.2155 .003
73 .542 .2155 .003
C
C NODES FOR CENTRE TUNNEL
C
75 1.558 -.067 .191
76 1.558 .055 .191
77 1.580 -.064 .191
78 1.580 .054 .191
79 2.052 -.059 .196
80 2.052 .055 .196
81 2.122 -.059 .197
82 2.122 .055 .197
Page 65
 83 1.553 -.065 -.003
84 1.553 .055 -.003
85 1.553 -.143 -.003
86 1.553 .135 -.003
87 1.553 -.434 -.003
88 1.553 .432 -.003
89 1.286 -.100 -.003
90 1.286 .073 -.003
91 1.386 -.087 -.003
92 1.386 .065 -.003
93 2.282 -.060 -.003
94 2.282 .060 -.003
95 2.307 -.060 -.003
96 2.307 .060 -.003
97 1.336 -.0935 -.003
98 1.336 .069 -.003
C
C PERIMETER NODES ON TOP PLANE
C
100 .069 0 .324
101 .069 -.046 .324
102 .069 -.250 .324
103 .069 -.281 .324
104 .167 -.267 .322
105 .277 -.288 .320
106 .282 -.289 .320
107 .967 -.411 .308
108 1.222 -.457 .303
109 1.553 -.517 .297
110 2.197 -.517 .297
111 2.489 -.517 .515
112 2.489 -.299 .515
113 2.489 0 .515
114 2.489 .299 .515
115 2.489 .517 .515
116 2.197 .517 .297
117 1.553 .517 .297
118 1.222 .457 .303
119 .967 .411 .308
120 .290 .289 .319
121 .167 .267 .322
122 .069 .281 .324
123 .069 .250 .324
124 .069 .043 .324
125 .276 .285 .319
126 2.489 -.325 .515
127 2.489 .325 .515
128 1.537 -.5 .392
129 1.537 .5 .392
C
C OTHER NODES ON TOP PLANE
C
150 .967 -.143 .310
151 1.222 -.107 .305
152 .967 .087 .310
153 .967 .095 .310
154 1.222 .075 .305
155 .967 -.161 .310
Page 66
 156 .967 .108 .310
C
C EXTRA NODES FOR THE BRACING
C
200 .275 -.280 .287
201 .275 .280 .287
202 .959 -.411 .029
203 .959 .411 .029
204 1.551 -.517 .266
205 1.551 .517 .266
206 1.551 -.517 .280
207 1.551 .517 .280
208 2.183 -.517 .288
209 2.183 .517 .288
210 2.312 -.517 .020
211 2.312 .517 .020
C
C NODES FOR REAR INTERNAL FRAME (UP TO BEHIND SEAT BACK)
C
220 2.372 -.203 .208
225 2.372 .203 .208
221 2.687 -.124 .077
224 2.687 .124 .077
222 2.697 -.124 .0795
223 2.697 .124 .0795
226 2.387 -.135 .250
227 2.387 .135 .250
228 2.387 -.120 .250
229 2.387 .120 .250
230 2.387 -.088 .250
231 2.387 .088 .250
232 2.387 -.055 .250
233 2.387 .055 .250
234 2.697 -.203 .208
235 2.697 .203 .208
236 2.412 -.203 .208
237 2.412 .203 .208
C
C NODES FOR SUSPENSION MOUNTINGS AND SUSPENSION
C
250 .005 -.195 .033
251 .005 .195 .033
252 .074 -.293 .261
253 .074 .293 .261
254 .167 -.265 .288
255 .167 .265 .288
256 .267 -.293 .265
257 .267 .293 .265
258 .315 -.192 .032
259 .315 .192 .032
260 .008 -.163 .046
261 .008 .163 .046
262 .057 -.262 .278
263 .057 .262 .278
264 .315 -.195 .003
265 .315 .195 .003
266 .273 -.275 .270
267 .273 .275 .270
Page 67
 268 2.327 -.120 .0425
269 2.327 .120 .0425
270 2.694 -.120 .045
271 2.694 .120 .045
272 2.414 -.231 .225
273 2.414 .231 .225
274 2.693 -.231 .225
275 2.693 .231 .225
276 .170 -.492 .275
277 .170 .492 .275
278 .170 -.487 .070
279 .170 .487 .070
280 2.550 -.55 .23
281 2.550 .55 .23
282 2.550 -.55 .04
283 2.550 .55 .04
284 2.497 -.325 .478
285 2.497 .325 .478
286 .170 -.49 .17
287 .170 .49 .17
288 2.550 -.55 .17
289 2.550 .55 .17
290 .17 -.55 -.1
291 .17 .55 -.1
292 2.550 -.60 -.1
293 2.550 .60 -.1
294 .17 0 -.1
C
RIGID.LINK
N1 N2
7 51
29 58
C
C
ELEMENTS
ELEM= 34100
NUMB GROUP PROP TOPO
C
C BOTTOM PERIMETER PLANE ELEMENTS
C
1 1 1 1 3
2 1 1 3 4
3 1 1 3 5
4 1 1 5 6
5 1 1 6 7
6 1 1 7 8
7 1 1 8 38
33 1 1 38 9
8 1 1 9 10
9 1 1 10 12
10 1 1 12 34
11 1 1 34 14
12 1 1 14 13
14 1 1 16 19
15 1 1 16 17
16 1 1 17 18
17 1 1 18 20
18 1 1 20 21

Page 68
 19 1 1 21 22
21 1 1 23 36
22 1 1 36 35
23 1 1 35 24
24 1 1 24 26
25 1 1 26 27
26 1 1 27 37
34 1 1 37 28
27 1 1 28 29
28 1 1 29 30
29 1 1 30 31
30 1 1 31 33
31 1 1 33 32
32 1 1 33 1
C
C BOTTOM PLANE INTERNAL ELEMENTS
C
50 1 1 5 31
51 1 1 6 50
52 1 1 50 59
53 1 1 59 30
54 2 2 50 264
93 2 2 264 51
55 2 2 51 52
56 2 2 52 72
452 2 2 72 53
57 2 2 53 54
58 1 1 10 54
59 1 1 55 26
60 2 2 55 56
61 2 2 56 73
453 2 2 73 57
62 2 2 57 58
63 2 2 58 265
94 2 2 265 59
64 1 1 12 87
70 1 1 87 85
71 1 1 85 83
72 1 1 83 84
73 1 1 84 86
99 1 1 86 88
74 1 1 88 24
66 1 1 54 60
67 1 1 60 61
68 1 1 55 62
69 1 1 62 63
75 2 2 63 90
76 2 2 90 98
450 2 2 98 92
77 2 2 92 84
78 2 2 84 94
96 2 2 94 96
79 2 2 61 89
80 2 2 89 97
451 2 2 97 91
81 2 2 91 83
82 2 2 83 93
97 2 2 93 95

Page 69
 65 1 1 13 64
83 1 1 64 65
84 1 1 65 66
85 1 1 66 67
86 1 1 67 95
87 1 1 95 96
88 1 1 96 68
89 1 1 68 69
90 1 1 69 70
91 1 1 70 71
92 1 1 71 23
C
C TOP PLANE PERIMETER ELEMENTS
C
100 3 3 100 101
101 3 3 101 102
102 3 3 102 103
103 3 3 102 104
104 3 3 104 105
105 3 3 105 106
106 3 3 106 107
107 3 3 107 108
108 3 3 108 109
109 4 4 109 110
110 3 3 110 111
111 3 3 111 126
126 3 3 126 112
112 3 3 112 113
113 3 3 113 114
114 3 3 114 127
127 3 3 127 115
115 3 3 115 116
116 4 4 116 117
117 3 3 117 118
118 3 3 118 119
119 3 3 119 120
120 3 3 120 125
125 3 3 125 121
121 3 3 121 123
122 3 3 123 122
123 3 3 123 124
124 3 3 124 100
C
C TOP PLANE INTERNAL ELEMENTS
C
150 3 3 101 106
151 3 3 124 120
152 3 3 120 153
153 3 3 107 155
154 3 3 155 150
155 3 3 150 152
156 3 3 152 153
157 3 3 153 156
158 3 3 156 119
159 2 2 150 151
160 2 2 152 154
161 3 3 108 151
162 3 3 151 154
Page 70
 163 3 3 154 118
164 5 5 111 16
165 5 5 115 21
166 2 2 151 77
177 2 2 77 78
167 2 2 154 78
168 2 2 77 79
169 2 2 78 80
170 2 2 79 81
171 2 2 80 82
172 2 2 81 232
173 2 2 82 233
174 4 4 109 128
175 4 4 128 129
176 4 4 129 117
C
C UPRIGHTS (WELL THOSE THERE ABOUT)
C
254 8 8 10 202
255 8 8 202 107
256 8 8 26 203
257 8 8 203 119
258 8 8 12 204
259 8 8 204 206
260 8 8 206 109
261 8 8 24 205
262 8 8 205 207
263 8 8 207 117
264 8 8 14 210
265 8 8 36 211
266 8 8 210 111
267 8 8 211 115
268 2 2 62 156
269 2 2 60 155
270 2 2 83 77
271 2 2 84 78
272 2 2 93 79
273 2 2 94 80
250 8 8 4 260
274 8 8 260 262
275 8 8 262 103
252 8 8 32 261
276 8 8 261 263
277 8 8 263 122
251 8 8 6 200
280 8 8 200 266
278 8 8 266 105
253 8 8 30 267
279 8 8 267 201
281 8 8 201 125
C
C MEMBERS REPRESENTING SUSPENSION COMPONENTS
C
300 9 9 104 254
301 9 9 121 255
302 9 9 260 250
303 9 9 261 251
304 9 9 262 252
Page 71
 305 9 9 263 253
306 9 9 256 266
307 9 9 257 267
308 9 9 258 264
309 9 9 259 265
310 9 9 274 234
311 9 9 275 235
312 9 9 270 221
313 9 9 271 224
314 9 9 268 66
315 9 9 269 69
316 9 9 236 272
317 9 9 237 273
C Front Suspension
ELEMENTS
ELEM= 34400
NUMB GROUP PROP TOPO
318 7 7 250 278
319 7 7 251 279
320 7 7 278 258
321 7 7 279 259
326 7 7 254 278
327 7 7 255 279
C Rear Suspension
328 7 7 268 282
329 7 7 269 283
330 7 7 270 282
331 7 7 271 283
336 7 7 282 126
337 7 7 283 127
C Uprights
ELEMENTS
ELEM= 34100
NUMB GROUP PROP TOPO
C
C BRACING MEMBERS
C
200 6 6 200 202
201 6 6 201 203
202 6 6 202 204
203 6 6 203 205
204 5 5 206 34
205 5 5 207 35
206 3 3 34 208
207 3 3 35 209
208 3 3 208 110
209 3 3 209 116
214 3 3 14 208
215 3 3 36 209
216 2 2 154 84 C TUNNEL FRONT BRACE
C
C MEMBERS OF THE REAR INTERNAL FRAME (UP TO BEHIND SEAT
BACK)
C
350 1 1 66 221
351 1 1 221 222
352 1 1 222 17

Page 72
 353 1 1 69 224
354 1 1 224 223
355 1 1 223 20
356 5 5 221 19
357 5 5 224 22
358 3 3 226 228
359 3 3 228 230
360 3 3 230 232
361 3 3 232 233
362 3 3 233 231
363 3 3 231 229
364 3 3 229 227
365 2 2 226 112
366 2 2 227 114
367 1 1 228 220
377 1 1 220 210
368 1 1 229 225
376 1 1 225 211
369 2 2 67 230
370 2 2 68 231
371 1 1 234 235
372 1 1 220 236
379 1 1 236 234
378 1 1 225 237
380 1 1 237 235
373 1 1 234 222
374 1 1 222 223
375 1 1 223 235
C
C FLOOR PAN MEMBERS
C
400 10 10 87 64
401 10 10 85 65
402 10 10 86 70
403 10 10 88 71
C
ELEMENTS
ELEM=46210
NUMB GROUP PROP TOPO
425 11 11 8 9 52 53 38 0 0 72
426 11 11 57 56 28 27 73 0 0 37
427 11 11 89 91 90 92 97 0 0 98
C
C
PLATES.AND.SHELLS
PLATE.OR.SHELL MATER.NUMB THICKNESS
11 1 0.0030
LOADS
CASE NODE DIRE VALUE
1 279 3 1000
C
C
RESTRAINTS
NODE.NUM DIRECTION
278 123
282 23
283 3
C

Page 73
C REACTIONS
C NODE.NUM
C 278
C 279
C 282
C 283
C
C STRESS.ELEMENT
C START FINISH STEP
C 1 9999 1
C
END.OF.DATA

Page 74
Page 75
Appendix D - Laboratory Testing Observations

Page 76
Appendix E - Diagrams and Information for Chassis Modifications

i) Diagrams Showing Changes to Chassis

Page 77
Page 78
Page 79
Page 80
Page 81
Page 82
Page 83
Page 84
Page 85
Page 86
Page 87
ii) Summary of Calculated Masses for Chassis Modifications
CAR1 - Masses and Moments of Inertia About Global Axes
GROUP CENTRE OF MASS MOMENTS OF INERTIA

NUMBER MASS X Y Z Ixx Iyy Izz

ALL 63.3 1.60 0.155E-02 0.130 9.35 206. 210.

CAR2

GROUP CENTRE OF MASS MOMENTS OF INERTIA

NUMBER MASS X Y Z Ixx Iyy Izz

ALL 64.4 1.58 0.152E-02 0.133 9.64 207. 212.

CAR3

GROUP CENTRE OF MASS MOMENTS OF INERTIA

NUMBER MASS X Y Z Ixx Iyy Izz

ALL 65.2 1.58 0.148E-02 0.126 9.67 209. 214.

CAR5

GROUP CENTRE OF MASS MOMENTS OF INERTIA

NUMBER MASS X Y Z Ixx Iyy Izz

ALL 63.5 1.60 0.155E-02 0.130 9.36 206. 211.

CAR6

GROUP CENTRE OF MASS MOMENTS OF INERTIA

NUMBER MASS X Y Z Ixx Iyy Izz

ALL 64.0 1.59 0.154E-02 0.132 9.44 206. 211.

CAR7

GROUP CENTRE OF MASS MOMENTS OF INERTIA

NUMBER MASS X Y Z Ixx Iyy Izz

ALL 63.6 1.60 0.142E-02 0.129 9.36 206. 211.

CAR8

GROUP CENTRE OF MASS MOMENTS OF INERTIA

NUMBER MASS X Y Z Ixx Iyy Izz

ALL 65.8 1.54 0.356E-02 0.131 9.50 206. 211.

CAR9

GROUP CENTRE OF MASS MOMENTS OF INERTIA

Page 88
NUMBER MASS X Y Z Ixx Iyy Izz

ALL 63.8 1.59 0.326E-02 0.131 9.44 206. 211.

CAR10

GROUP CENTRE OF MASS MOMENTS OF INERTIA

NUMBER MASS X Y Z Ixx Iyy Izz

ALL 62.5 1.61 -0.850E-03 0.127 9.24 206. 210.

CAR11

GROUP CENTRE OF MASS MOMENTS OF INERTIA

NUMBER MASS X Y Z Ixx Iyy Izz

ALL 63.5 1.60 -0.959E-03 0.131 9.41 206. 211.

CAR12

GROUP CENTRE OF MASS MOMENTS OF INERTIA

NUMBER MASS X Y Z Ixx Iyy Izz

ALL 71.1 1.53 0.780E-03 0.134 10.1 214. 219.

CAR13

GROUP CENTRE OF MASS MOMENTS OF INERTIA

NUMBER MASS X Y Z Ixx Iyy Izz

ALL 63.3 1.60 0.141E-02 0.130 9.34 206. 210.

CAR14

GROUP CENTRE OF MASS MOMENTS OF INERTIA

NUMBER MASS X Y Z Ixx Iyy Izz

ALL 64.4 1.59 0.159E-02 0.128 9.42 207. 211.

CAR15

GROUP CENTRE OF MASS MOMENTS OF INERTIA

NUMBER MASS X Y Z Ixx Iyy Izz

ALL 66.9 1.55 0.141E-02 0.131 9.86 208. 212.

CAR16

GROUP CENTRE OF MASS MOMENTS OF INERTIA

NUMBER MASS X Y Z Ixx Iyy Izz

ALL 71.0 1.59 0.205E-02 0.133 10.6 231. 236.

CAR17

Page 89
GROUP CENTRE OF MASS MOMENTS OF INERTIA

NUMBER MASS X Y Z Ixx Iyy Izz

ALL 65.0 1.56 0.218E-02 0.132 9.51 206. 211.

CAR18

GROUP CENTRE OF MASS MOMENTS OF INERTIA

NUMBER MASS X Y Z Ixx Iyy Izz

ALL 64.9 1.57 0.499E-03 0.133 9.55 207. 211.

CAR19

GROUP CENTRE OF MASS MOMENTS OF INERTIA

NUMBER MASS X Y Z Ixx Iyy Izz

ALL 67.6 1.58 0.240E-03 0.132 9.55 216. 220.

CAR20

GROUP CENTRE OF MASS MOMENTS OF INERTIA

NUMBER MASS X Y Z Ixx Iyy Izz

ALL 63.3 1.60 0.155E-02 0.130 9.35 206. 210.

CAR21

GROUP CENTRE OF MASS MOMENTS OF INERTIA

NUMBER MASS X Y Z Ixx Iyy Izz

ALL 68.4 1.59 0.209E-03 0.131 9.55 219. 223.

CAR22

GROUP CENTRE OF MASS MOMENTS OF INERTIA

NUMBER MASS X Y Z Ixx Iyy Izz

ALL 67.5 1.54 0.151E-02 0.136 9.86 208. 213.

Page 90
Appendix F - Components of the Westfield Sports Cars

Page 91
Appendix G - Calculations
i) Chassis Torsional Stiffness

The deflections at four dial gauges are as follows:

Node 102 123 111 115

DZ -5.77 -14.01 0.84 -1.15

Load = 94.4 kg

lever arm of load = 974mm

ii) Measure of Effectiveness of WSC Spaceframe

Page 92
As a means to evaluating the effectiveness of the geometry of a space frame in resisting torsional loading, it
would be helpful to consider: How much does the torsional stiffness of the members independently contribute
to the overall torsional stiffness of the chassis.

Note that a low percentage contribution from individual members will suggest an effective geometry of the
overall chassis.

A typical cross section of the chassis would consist of four 25x1.6 SHS and four 20x1.6 SHS.

The chassis alone was calculated as having a stiffness of 1121 Nm/deg over a length of 2420mm. Consider
member stiffnesses alone over this length:

The percentage to which the members alone contribute to torsional stiffness is: 6.5%

Hence the stiffness of the chassis is primarily derived from the members "acting as a team" within the structure
rather than from the individual member stiffnesses, as would be the case with a ladder type chassis.

ii) Equivalent Structures

It may be helpful in understanding the improvements in chassis technology over the years to consider the
typical torsional stiffness of different types of chassis of the same weight.

Westfield Chassis: weight = 63 kg

wheelbase = 2420 mm

torsional stiffness = 1121 Nm/deg

Girder Chassis of Equivalent Weight

Two girders 2.9 m long and three 0.8 m cross members of the same section, weighing 63 kg combined would
require a member of 7.7 kg/m.

Page 93
Using 100 TFB (Taper Flanged Beam); 7.20 kg/m the weight of the chassis will be 59 kg and the torsional
stiffness of the chassis will be:

Page 94
Ladder Chassis of Equivalent Weight (ignoring the effect of lateral bracing)

Use 102 CHS 3.2 7.77kg/m. Two 2.9m beams and three 0.8m laterals weigh 63.7 kg.

Stressed Skin Chassis of Equivalent Weight

Using 0.9mm steel skin, the dimensions for two boxes with a combined weight of around 63 kg are:

430mm x 300mm x 0.9mm thick; two 3m boxes would weigh 62kg

Page 95