This action might not be possible to undo. Are you sure you want to continue?
https://www.scribd.com/doc/38003231/2009REVSAESuspensionCorke
07/28/2013
text
original
Analysis of the 2009 REV Race Car Suspension
Peter Corke
20139918
School of Mechanical Engineering, University of Western Australia
Supervisors
A/Prof Thomas Braunl
School of Electrical, Electronic and Computer Engineering, University of
Western Australia
Dr Kamy Cheng
School of Mechanical Engineering, University of Western Australia
Final Year Project Thesis
School of Mechanical Engineering
University of Western Australia
Submitted: October 23
rd
, 2009
2
Project Summary
The purpose of this project is to analyse and optimise the suspension system of the 2009
REV race car. An effective suspension system acts as a vibration filter, and protects the
vehicle from excitation caused by irregularities in the road.
Suitable coil springs are to be selected for the front and rear of the car, to ensure that both
ends do not oscillate violently out of phase after going over a bump. The Half Car Model,
which has four degrees of freedom, can be used to simulate the vehicle driving in a straight
line. By providing excitation to the wheels, the model can be used to show the vertical
accelerations of the sprung mass and the front and rear unsprung masses at different
frequencies. The selected spring combination should give the least possible vertical
accelerations of the sprung and unsprung masses. A common method of determining the
extent of driver discomfort is to measure the vertical acceleration of the sprung mass.
It is also important that the suspension is able to damp the unsprung masses adequately to
prevent wheel hop. Wheel hop is the vertical vibrating motion of the wheel between the
sprung mass and the road, and has the capacity to impede handling and performance.
The Half Car Model will also be used to determine the effects of changing the amplitude of
the assumed road, and changing the damping rate of the wheel.
3
Letter of transmittal
Peter Corke
6/57 Lockhart Street
Como, WA, 6152
23
rd
October, 20091020
Professor David Smith
Dean
Faculty of Engineering, Computing and Mathematics
University of Western Australia
35 Stirling Highway
Crawley, WA, 6009
Dear Professor Smith,
I am pleased to submit this thesis, entitled “Analysis of the 2009 REV Race Car
Suspension”, as part of the requirement for the degree of Bachelor of Engineering.
Yours sincerely
Peter Corke
20139918
4
Acknowledgements
I would like to thank the following people who have helped make this project possible.
All of the REV team, for being great people to work with.
Lachlan Tomlin, for the hours that he put into the race car.
Chris Rowles and Grant Keady, for giving me good advice with my computer modelling.
My friends and family, for the support that they have given me this year.
5
Table of Contents
1. Introduction........................................................................................................................ 7
Project objectives ............................................................................................................... 7
Definitions.......................................................................................................................... 7
a...................................................................................................................................... 7
b...................................................................................................................................... 7
CG.................................................................................................................................. 8
Sprung mass ................................................................................................................... 8
Unsprung mass............................................................................................................... 8
Wheelbase, l ................................................................................................................... 8
Axes ................................................................................................................................... 8
Camber ............................................................................................................................... 9
Toe.................................................................................................................................... 10
Importance of suspension................................................................................................. 10
Wishbone independent suspension .................................................................................. 11
Equal length and parallel wishbones........................................................................... 12
Unequal length wishbones ........................................................................................... 13
Unsprung mass oscillations.............................................................................................. 14
Shimmy ......................................................................................................................... 15
Wheelfight .................................................................................................................... 17
Wheel hop..................................................................................................................... 18
2. Model Formulation .......................................................................................................... 20
Car information................................................................................................................ 20
Coil springs ...................................................................................................................... 21
Fox Vanilla RC Shox ....................................................................................................... 22
Quarter Car Model ........................................................................................................... 24
Half Car Model ................................................................................................................ 26
3. Results.............................................................................................................................. 31
Available springs.............................................................................................................. 31
Motion ratio...................................................................................................................... 31
Wheel rates....................................................................................................................... 32
Tyre rate ........................................................................................................................... 33
Ride rates.......................................................................................................................... 33
Undamped natural frequencies of the front and rear sprung masses ............................... 34
Spring selection................................................................................................................ 35
Criteria1....................................................................................................................... 35
Criteria 2...................................................................................................................... 36
Spring combination candidates.................................................................................... 37
Wheel hop: rate and mass ratios....................................................................................... 38
Vertical accelerations produced by the spring combination candidates .......................... 41
Most appropriate springs to be used on the REV car .................................................. 45
The problem associated with using springs that are too stiff ........................................... 46
Changing amplitude, A
1
................................................................................................... 48
Changing damping ratio, ζ ............................................................................................... 49
4. Safety................................................................................................................................ 51
Emergency phone numbers.............................................................................................. 51
Fire procedure .................................................................................................................. 51
6
Smoking ........................................................................................................................... 52
Electrical safety................................................................................................................ 52
Working in labs................................................................................................................ 52
Driving the REV car......................................................................................................... 52
5. Conclusion and Future Work ........................................................................................... 54
6. References ........................................................................................................................ 56
7. Appendices....................................................................................................................... 58
Appendix A: Negative swing arm effect.......................................................................... 58
Appendix B: Determining a and b ................................................................................... 59
Appendix C: Quarter Car Model Vertical Acceleration Calculations ............................. 60
Appendix D: Road Excitation Calculations ..................................................................... 62
Appendix E: Half Car Model Vertical Acceleration Calculations................................... 64
Appendix F: Damping calculations for the Half Car Model............................................ 66
Appendix G: MATLAB Code for Half Car Model Vertical Accelerations..................... 67
Appendix H: Damping calculations for the Quarter Car Model ...................................... 70
Appendix I: MATLAB Code for Quarter Car Model Vertical Accelerations ................. 71
7
1. Introduction
Project objectives
The original plan for the REV team was to receive the 2001 Motorsport SAE car in ‘rolling
chassis’ form at the beginning of the year, and to transform it into a batterypowered race
car with hub motors. The aim of this thesis was going to be to redesign the suspension
system of the vehicle, making it more suitable to the new weight distribution caused by
replacing the petrol motor with a battery cage, batteries and hub motors. Unfortunately,
due to significant time delays, the ‘rolling chassis’ was not received until the mid year
holidays. Therefore, in order to get a drivable car constructed by the end of the year, as few
alterations as possible were made to the car, and, consequently, this thesis became
concerned with analysing and optimising the existing suspension system of the car instead
of designing a new one.
About midway through second semester, it was decided that trying to design and install a
hub motor setup complete with pulleys into each of the existing rear uprights of the car was
too difficult. Therefore, it was decided that inboard motors would be used as this approach
allowed for much more space and easier installation.
Definitions
a
‘a’ is the distance between the front track and the centre of gravity (CG)
b
‘b’ is the distance between the rear track and the centre of gravity (CG)
8
CG
CG is an acronym for ‘centre of gravity’.
Sprung mass
The sprung mass is comprised of parts of the vehicle that are supported by the suspension,
such as the chassis, driver, battery cage, batteries and motors. The sprung mass moves
when the springs are deflected (Waldron & Kinzel 2004).
Unsprung mass
The unsprung mass is comprised of parts of the vehicle that are not supported by the
suspension, such as the wheels, tyres and uprights (Waldron & Kinzel 2004).
Wheelbase, l
The wheelbase (l) is the distance between the front track and the rear track
Axes
Below is a diagram showing the pitch, roll and yaw axes of a vehicle. The Half Car Model
later in this paper is concerned with the pitch axis.
9
Figure 1.1: Pitch, roll and yaw axes (Milliken & Milliken 1995)
Camber
Below is a diagram showing the camber angle.
Figure 1.2: Camber angle (Milliken & Milliken 1995)
Below is a diagram showing a negative camber angle of both wheels.
10
Figure 1.3: Negative camber (Milliken & Milliken 1995)
Toe
Below is a diagram showing both toe in and toe out. Please note the oval shape represents
the driver at the front of the car.
Figure 1.4: Toe in and toe out (Desert Rides 2007)
Importance of suspension
The key requirement of a suspension system is to ensure that all tyres, especially the two
attached to the driving wheels, maintain the greatest possible contact with the road. The
11
suspension should protect the vehicle from jolts in the road, and provide stability and
roadholding throughout the entire power range of the motors. While cornering, the wheels,
especially the outside wheels, should remain as upright as possible. An effective
suspension system is very much a compromise between roadholding and ride. The stability
of a racing car is influenced by many factors, including:
• Vehicle weight
• Vehicle weight distribution
• Sprung to unsprung weight ratio
• Location of the centre of gravity
• Roll centre heights on the front and rear suspensions
• Wheelbase and track
• Suspension layout
• Steering geometry
• Loading condition
• Characteristics of the tyres, including pressure and tread depth
• Wheel and tyre balance
A successful suspension system generally has nonlinear camber change, which means that
camber is not directly dependent on vehicle roll (Costin & Phipps 1962).
Wishbone independent suspension
Wishbone independent suspension systems do not encounter as many problems as beam
axles.
1. The front end does not experience shimmy, which means that the steering requires no
damping. This allows the driver to have a better “feel” for the handling of the car,
which is important for race car driving. Also, bump steer can be entirely eradicated at
any particular steering angle.
12
2. Unsprung mass is heavily reduced. Since there is much less mass moving up and down
with the wheels, it is easier to maintain contact between the tyres and the road,
especially on uneven surfaces.
3. The rear wheels can be cambered, which can increase the cornering power of the tyres.
Also, any amount of toein or toeout can be provided.
(Aird 1997)
The advantage with wishbone suspension systems is that they are normally built with a
steering linkage that is symmetrical about the centreline of the vehicle. This means that any
“inaccuracies” in the linkage will not cause the vehicle to veer off unnecessarily (Milliken
& Milliken 2002).
Equal length and parallel wishbones
The main disadvantage of having equal length and parallel wishbones is that the wheels
become cambered the wrong way when the chassis rolls in cornering. To ensure that the
treads of the tyres remain flat on the road while cornering, a substantial amount of negative
camber can be incorporated into the wheels. This, however, means that when the vehicle is
traveling forward, the tyres will be running on their edges. This causes the tyre edges to
wear and overheat, and does not allow for much traction when accelerating or braking
(Aird 1997).
Figure 1.5: Equal length and parallel wishbones in roll (Aird 1997)
13
Using equal length and parallel wishbones also means that as the suspension moves
vertically in bump and rebound, the wheels will rise and fall without cambering. This
completely eliminates the problem of gyroscopic “kick”, which occurs when a rotating
wheel is tilted, although it introduces a new problem. The vertical motion of the wheel is
accompanied by some sideways motion, referred to as scrub, and this generates “kickback”
in the steering. This is due to the line of the axis the wheel steers around crossing the
ground ahead of the tyre contact patch’s centre (Aird 1997).
Figure 1.6: Equal length and parallel wishbones in bump (Aird 1997)
Unequal length wishbones
The REV car has independent suspension with unequal length wishbones. By using
unequal length wishbones, that is, by making the top wishbone shorter than the bottom one,
it is possible to combine small amounts of camber and scrub. Having small amounts of
camber and scrub prevents the detrimental effects of having large quantities of either by
itself (Aird 1997).
14
Figure 1.7: Independent suspension with unequal length wishbones (Smith 1978)
Unsprung mass oscillations
The oscillations of the unsprung mass can cause potential problems. These oscillations
include:
1. Shimmy
2. Wheelfight
3. Wheel hop
Wheels that are mounted to independent suspension linkages have an advantage over
wheels that are attached to beam axles, because, in theory, they have only one degree of
freedom, as opposed to six. However, in racing cars, it is essential to avoid too rigid a
limitation of the wheel motion, which therefore establishes additional degrees of freedom.
Vibrations in the various modes can sometimes supply energy to each other, if they are
coupled in a particular way. Of the listed varieties of oscillation, only shimmy is
indefinitely selfsustaining on a smooth road (Milliken & Milliken 2002).
15
Shimmy
Shimmy is considered the most hazardous type of unsprung mass oscillation, and is
prevalent in cars where the front wheels are connected via a beam axle. The shimmy cycle
begins by one of the front wheels hitting a chance bump in the road, or by an imbalance in
one of the front wheels. This causes one front wheel to be flattened into the road, while the
other wheel is in the air. The positions of the wheels are then reversed. This cycle is
infinitely selfsustaining on a smooth road, and generally occurs at higher speeds.
Shimmy can cause many problems, such as the hands of the driver being shaken off the
steering wheels, a tyre blowing or coming off its rim, and damage of steering and
suspension linkages. If the driver applies the front brakes during the shimmy cycle, the
frequency of the shimmy decreases while its amplitude increases. This can cause the car to
become out of control.
16
Figure 1.8: A severe shimmy cycle (Milliken & Milliken 2002)
The most effective approach in preventing shimmy is to use independent front suspension,
as well as having a torsionally stiff car frame. By placing the sprung mass of the car
between the front wheel mountings, the gyroscopic torque of the shimmy cycle is unable to
generate the high frequency required for shimmy (Milliken & Milliken 2002).
It is highly unlikely that a shimmy cycle will develop on the REV car, as it has independent
suspension at both ends of the car. The chassis, which is a part of the sprung mass, is
torsionally stiff, and is manufactured from mild alloy steel 4140. It is fair to state that the
chassis is “overengineered”.
17
Wheelfight
SAE Vehicle Dynamics Terminology defines wheelfight as “a rotary disturbance of the
steering wheel produced by forces acting on the steerable wheels” (Milliken & Milliken
2002). Wheelfight is unlike shimmy as it is not selfsustaining. Instead, it relies on
constant excitation supplied by irregularities in the road.
Any complex analysis of wheelfight requires a detailed investigation of the steering gear’s
elastic resonance. Pictured below is a steering wheel test rig.
Figure 1.9: Steering gear test rig (Milliken & Milliken 2002)
Wheelfight generally takes place when the frequency of the steering gear is close to the
frequency at which wheel hop occurs.
18
This means that preventative measures against wheelfight can be implemented in either of
two locations: at the steering gear or at the steerable wheels (front wheels).
To assist in preventing wheelfight at the steering gear:
i. increase the steering wheel’s moment of inertia
ii. reduce the steering wheel shaft diameter
To assist in preventing wheelfight at the steerable wheels,
i. attach rubber mountings to the lower control arms
ii. have suspension linkages designed so that as the front wheels rise they toe out
iii. make the wheel centres move backwards as the wheel rises by sloping the wheel
arm pivots
iv. induce a negative swing arm effect (please refer to Appendix A)
It is unlikely that wheelfight will pose a threat to the performance of the REV car, as the
steering gear will probably not reach high enough frequencies (wheel hop frequency). To
verify this, the natural frequency of the steering gear would have to be determined,
however, this is beyond the scope of this project.
Wheel hop
Vehicle ride has two modes: they are primary ride (often simply referred to as ride) and
secondary ride. Primary ride is the lower frequency oscillation of the sprung mass, while
secondary ride is the higher frequency oscillations of the unsprung masses. Secondary ride
is sometimes referred to as wheel hop, which is defined as “the vertical oscillatory motion
of a wheel between the road surface and the sprung mass” (Milliken & Milliken 2002).
Too much wheel hop has adverse effects on the handling and stability of the car. Therefore
it is important to appropriately damp the unsprung masses, and also the sprung mass, so
that the driver is comfortable.
19
Wheel transmissibility is defined as the ratio of the vertical wheel motion to road input
amplitude. Generally transmissibility under 2.5 will not permit enough wheel hop to hinder
handling (Milliken & Milliken 1995).
20
2. Model Formulation
Car information
The total mass of the car when it loaded with a 75kg driver is 274 kg. There are two
inboard motors that are mounted to the chassis above the rear track that have a combined
mass of 25kg, while the battery cage and batteries are located just behind the driver’s seat
and have a combined mass of 55kg. The car has an approximate mass distribution of 48%
at the front and 52% at the rear.
Below is a screen shot from a computer model of the car that was created using a software
package called Solid Works.
Figure 2.1: Solid Works model of the car (Ivanescu & Tomlin 2009)
The Solid Works software was used to calculate the mass moment of inertia of the sprung
mass about its pitch axis through its centre of gravity. This value was found to be 616.363
kg m
2
.
21
The important information about the car is expressed in the table below.
Information Symbol
Wheelbase l 1.85m
Distance from CG to front track a 0.959m
Distance from CG to rear track b 0.891m
Sprung mass M 202kg
Front sprung mass M
f
97.29kg
Rear sprung mass M
r
104.71kg
Unsprung mass (same at front and rear) m 18kg
Mass moment of inertia (about pitch axis
through CG) I 616.363kg m
2
Table 2.1: Important information about the car
Please refer to Appendix B to see the method used for calculating a and b.
Coil springs
Coil springs are the most commonly used type of springs in racing cars with independent
suspensions. They can be used in both compression and extension (tension). The helix is
the most widespread shape of coil spring, in which the mean diameter of the coil remains
constant (Milliken & Milliken 1995).
Below is a diagram of a helixshaped coil spring.
Figure 2.2: Helixshaped coil spring (Milliken & Milliken 1995)
22
Fox Vanilla RC Shox
The shock absorbers used in the REV car suspension are of the Fox Vanilla RC Shox
variety, which are commonly used on mountain bikes.
The Fox Vanilla RC is a telescopic type monotube damper, which means that its damping
force is due to oil being forced through an orifice at high pressures. It has an external tank
of pressurised nitrogen gas, which is increasingly compressed as the piston rod moves
further up the cylinder (Guzzomi 2004). Below is a diagram of the Fox Vanilla RC
damper.
Figure 2.3:Fox Vanilla RC damper (Finlayson 2003)
The Fox Vanilla RC damper has a casing that is manufactured out of hard, anodised
aluminium, which increases wear resistance. It has a 25mm bore and has a mass of just
over 315 g.
23
The damper is adjustable in both bump and rebound. To modify the bump settings, the
spring preload on the relief valve can be altered (see Figure 2.3). A high spring preload
will result in higher damping rates in bump. To modify the rebound settings, there is an
adjustable needle valve that controls piston bypass in both bump and rebound, although it
has only a small effect on bump damping (Guzzomi 2004).
Adjustments in bump and rebound are done in increments of ¼ turns, or “clicks”.
The graphs below show force (N) vs. velocity (m/s) for the Fox Vanilla RC damper in both
bump and rebound. The numbers in the legend refer to the number of “clicks” open.
Bump
0
500
1000
1500
2000
2500
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Velocity (m/s)
F
o
r
c
e
(
N
)
0
1
3
5
7
9
11
Graph 2.1: Fox Vanilla RC damper in bump
24
Rebound
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
0 0.05 0.1 0.15 0.2 0.25 0.3
Velocity (m/s)
F
o
r
c
e
(
N
)
4
5
6
7
8
9
10
11
12
13
Graph 2.2: Fox Vanilla RC damper in rebound
It is noticeable in the “bump” graph that there is a distinct kink in each of the curves, which
occurs at around the same velocity, regardless of the number of “clicks”. This is due to the
presence of a “blowoff” valve, which causes the bypass of a proportion of the oil at high
pressures. The reason for this is to avoid very high damping forces at high velocities
(Milliken & Milliken 1995). From the graph, it can be seen that “blowoff” occurs at
approximately 0.2 m/s.
Quarter Car Model
The Quarter Car Model is used to simulate one corner of the car. It has two degrees of
freedom: the vertical movement of the unsprung mass, and the vertical movement of the
sprung mass.
Below is a diagram of the Quarter Car Model.
25
Figure 2.4: Quarter Car Model
Where
m=unsprung mass
M
1
=sprung mass at one wheel
X
1
=vertical displacement of the unsprung mass
X
3
=vertical displacement of the sprung mass
K
W
=wheel rate
K
T
=tyre rate
c=wheel damping rate
The Quarter Car Model assumes that the road is a sine wave with amplitude A
1
. Therefore,
the excitation caused by the road is
F = A
1
e
jωt
(2.1)
26
Assumptions of the Quarter Car Model
Listed below are the assumptions incorporated into the quarter car model
• The CG of the sprung mass is located directly above the CG of the unsprung mass
• The quarter of the vehicle being analysed acts independently to the rest of the
vehicle
• There is no tyre damping
• The damping of the wheel is theoretical linear fluid damping
• The excitation from the road is sinusoidal excitation with amplitude A
1
The Quarter Car Model will be used to determine the vertical accelerations of the sprung
and unsprung masses. Please refer to Appendix C for the Quarter Car Model vertical
acceleration calculations.
Half Car Model
The Half Car Model, sometimes called the bicycle model, is used to simulate one half of
the car. It is more realistic than the quarter car model, because the front and rear are
coupled so that their motions affect one another. It has four degrees of freedom: the
vertical movement of the front unsprung mass, the vertical movement of the rear unsprung
mass, the vertical movement of the centre of gravity of the sprung mass, and the pitching of
the sprung mass about its CG.
Below is a diagram of the Half Car Model.
27
Figure 2.5: Half Car Model
Where
m
f
=front unsprung mass
m
r
=rear unsprung mass
M=sprung mass of half of the car
X
1
=vertical displacement of the front unsprung mass
X
2
=vertical displacement of the rear unsprung mass
X
3
=vertical displacement of the CG of the sprung mass
K
Wf
=front wheel rate
K
Wr
=rear wheel rate
K
T
=tyre rate
28
c
f
=front wheel damping rate
c
r
=rear wheel damping rate
l=wheelbase
a=distance from CG to front track
b=distance from CG to rear track
The Half Car Model assumes that the road is a sine wave with amplitude A
1
, and with a
length of one cycle of road profile of s. Below is a diagram of the assumed road.
Figure 2.6: Assumed road
Therefore, the excitations caused by the road are:
t j
e A F
ω
1 1
=
(2.2a)
t j
e
s
l j
j A F
ω
π

¹

\

− =
2
exp
1 2
(2.2b)
Please refer to Appendix D to see how the above excitations were calculated.
Below is a diagram of the sprung mass when it is at a pitch angle of θ. Not that the
situation in the diagram occurs when the force from the front unsprung mass is greater than
the force from the rear unsprung mass.
i.e.
(2.3)
29
Figure 2.7: Sprung mass from the Half Car Model pitched at an angle θ.
Where, for small angles of θ,
θ a X Y + =
3 1
(2.4a)
θ b X Y − =
3 2
(2.4b)
Assumptions of the Half Car Model
Listed below are the assumptions incorporated into the half car model
• The car is travelling in a straight line
• The half of the vehicle that is being analysed acts independently to the rest of the
vehicle
• There is no tyre damping
• The damping of the wheels is theoretical linear fluid damping
• The road is a sine wave with amplitude A
1
and with a length of one cycle of road
profile of s
• The pitch angle, θ, is small
• The mass moment of inertia, I, is constant. This can be assumed as the mass
moment of inertia will vary very little for small pitch angles.
30
The Half Car Model will be used to determine the vertical accelerations of the front
unsprung, rear unsprung and sprung masses. Please refer to Appendix E for the Half Car
Model vertical acceleration calculations.
31
3. Results
Available springs
The springs available to the REV team are an assortment of Fox coil springs, varying in
spring rate (K
S
) and travel distance. The spring rate and travel distance are written on the
side of the spring. For example, 600 x 2.00 indicates a spring rate of 600 lbs/in and a travel
distance of 2.00 in.
The table below lists the available Fox coil springs, as well as the accompanying spring
rates in SI units.
Spring Spring Rate (N/m)
600 x 2.00 105076
550 x 2.00 96320
500 x 2.30 87563
450 x 2.39 78807
300 x 2.57 52538
Table 3.1: Spring rates of the available Fox coil springs.
Motion ratio
If springs were mounted directly above the tyres, as pictured below, then the wheel rate
would be equal to the spring rate.
32
Figure 3.1: Spring mounted directly above wheel (Smith 1978)
However, due to limitations of space, it is necessary for the springs to be mounted inboard
of the tyre centerline. This means that the wheel travel will not be equal spring travel. The
motion ratio is defined as
Travel Spring
Travel Wheel
Ratio Motion =
(3.1)
For example, in the diagram below, the motion ratio is 2.2 (Smith 1978).
Figure 3.2: Spring mounted inboard of the tyre centerline (Smith 1978)
Using experimental data from the 2001 Motorsport team, the motion ratio for the front and
rear wheels was calculated to be 1.764.
Wheel rates
The wheel rate (K
W
) is defined as
( )
2
S
W
Ratio Motion
K
K =
(3.2)
The table below shows the available springs and the corresponding wheel rates.
33
Spring Wheel Rate (N/m)
600 x 2.00 33768
550 x 2.00 30954
500 x 2.30 28140
450 x 2.39 25326
300 x 2.57 16884
Table 3.2: Available springs and the corresponding wheel rates.
Tyre rate
A standard Goodyear racing tyre suitable for the application of the REV car with an
inflation pressure of 83kPa has a tyre rate (K
T
) of 180190N/m (Winzer 2002). This value
was used for further calculations involving both the front and rear tyre rates.
Ride rates
The ride rate (K
R
) is the combined rate of the wheel rate and the tyre rate. The deflections
of the wheel and the tyre are additive, so therefore the formula for two springs in series is
used (Milliken & Milliken 2002). The formula for ride rate is
W T R
K
1
K
1
K
1
+ =
(3.3a)
or
W T
W T
R
K K
K K
K
+
=
(3.3b)
The table below shows the available springs and the corresponding ride rates.
Spring Ride Rate (N/m)
600 x 2.00 28439
550 x 2.00 26416
500 x 2.30 24339
34
450 x 2.39 22205
300 x 2.57 15437
Table 3.3: Available springs and the corresponding ride rates
Undamped natural frequencies of the front and rear sprung masses
The natural frequency of the front sprung mass is
f
Nf
=
1
2π
2K
R
M
f
(3.4a)
The natural frequency of the rear sprung mass is
r
R
Nr
M
2K
2
1
f
π
= (3.4b)
At each end of the car there are two wheels, meaning that at each end there are two “ride
rate” springs. These springs share a load, and are said to be in parallel. Therefore to find
their composite rate, they are added. This explains the “2K
R
” term in both of the equations
above (Milliken & Milliken 1995).
The front and rear sprung masses were calculated to be 97.29kg and 104.71kg respectively.
The table below shows the available springs and the corresponding undamped natural
frequencies of the front and rear sprung masses.
Spring
Natural Frequency of
Front Sprung Mass (Hz)
Natural Frequency of
Rear Sprung Mass (Hz)
600 x 2.00 3.848 3.709
550 x 2.00 3.709 3.575
500 x 2.30 3.560 3.432
450 x 2.39 3.400 3.278
300 x 2.57 2.835 2.733
35
Table 3.4: Available springs and the corresponding natural frequencies of the front and rear
sprung masses
Spring selection
To select which combination of springs to use at the front and the rear of the car, two
different criteria had to be satisfied. Any combination that satisfied both criteria was then
considered a possible candidate for use in the REV car suspension.
Criteria1
Staniforth recommends that, as a general guideline, the frequency of the front sprung mass
should be about 10% lower than the frequency of the rear sprung mass, in order to avoid the
front and rear of the vehicle oscillating violently out of phase after going over a bump
(Staniforth 1999).
This means that
90 . 0 ≈
Nr
Nf
f
f
(3.5)
However, he states that his race experience has shown that this guideline is more of a “rule
of thumb”, and is not always valid. Therefore, the author has been rather lenient and
considered any spring combination which produces frequency ratios in the range of
95 . 0 75 . 0 ≤ ≤
Nr
Nf
f
f
(3.6)
to have satisfied Criteria 1.
36
Below is a table showing the values of f
Nf
/ f
Nr
for each spring combination. The
combinations satisfying Criteria 1 are highlighted in bold.
Rear Springs f
Nr
(Hz) Front Springs f
Nf
(Hz) f
Nf
/ f
Nr
600 x 2.00 3.709 600 x 2.00 3.848 1.037
600 x 2.00 3.709 550 x 2.00 3.709 1
600 x 2.00 3.709 500 x 2.30 3.56 0.960
600 x 2.00 3.709 450 x 2.39 3.4 0.917
600 x 2.00 3.709 300 x 2.57 2.835 0.764
550 x 2.00 3.575 550 x 2.00 3.709 1.037
550 x 2.00 3.575 500 x 2.30 3.56 0.996
550 x 2.00 3.575 450 x 2.39 3.4 0.951
550 x 2.00 3.575 300 x 2.57 2.835 0.793
500 x 2.30 3.432 500 x 2.30 3.56 1.037
500 x 2.30 3.432 450 x 2.39 3.4 0.991
500 x 2.30 3.432 300 x 2.57 2.835 0.826
450 x 2.39 3.278 450 x 2.39 3.4 1.037
450 x 2.39 3.278 300 x 2.57 2.835 0.865
300 x 2.57 2.733 300 x 2.57 2.835 1.037
Table 3.5: Criteria 1
Criteria 2
In his notes, Maurice Olley suggests that the natural frequencies of the front and rear
sprung masses should be close, however, the natural frequency of the rear sprung mass
should be no greater than 1.2 times the natural frequency of the front sprung mass (Milliken
& Milliken 2002).
This means that
Nf Nr
f f 2 . 1 ≤
(3.7)
As in Criteria 1, this is to avoid the situation where the front and rear of the vehicle
oscillate violently out of phase after going over a bump.
37
Below is a table showing the values of 1.2f
Nf
for each spring combination. The
combinations satisfying Criteria 2 are highlighted in bold.
Rear Springs f
Nr
(Hz) Front Springs f
Nf
(Hz) 1.2 f
Nf
600 x 2.00 3.709 600 x 2.00 3.848 4.6176
600 x 2.00 3.709 550 x 2.00 3.709 4.4508
600 x 2.00 3.709 500 x 2.30 3.56 4.272
600 x 2.00 3.709 450 x 2.39 3.4 4.08
600 x 2.00 3.709 300 x 2.57 2.835 3.402
550 x 2.00 3.575 550 x 2.00 3.709 4.4508
550 x 2.00 3.575 500 x 2.30 3.56 4.272
550 x 2.00 3.575 450 x 2.39 3.4 4.08
550 x 2.00 3.575 300 x 2.57 2.835 3.402
500 x 2.30 3.432 500 x 2.30 3.56 4.272
500 x 2.30 3.432 450 x 2.39 3.4 4.08
500 x 2.30 3.432 300 x 2.57 2.835 3.402
450 x 2.39 3.278 450 x 2.39 3.4 4.08
450 x 2.39 3.278 300 x 2.57 2.835 3.402
300 x 2.57 2.733 300 x 2.57 2.835 3.402
Table 3.6: Criteria 2
Spring combination candidates
Therefore, the combinations of springs satisfying both criteria are
Combination Front Springs Rear Springs
1 450 x 2.39 600 x 2.00
2 300 x 2.57 450 x 2.39
Table 3.7: Candidates for the spring combination of the car
38
Wheel hop: rate and mass ratios
Assuming that the oscillation centres of the front and rear sprung masses are directly above
the front and rear tracks respectively,
Critical damping of the sprung mass is
c
crit,M
= 2 M
1
K
T
K
W
K
T
+ K
W
(3.8)
Critical damping of the unsprung mass is
c
crit,m
= 2 m K
T
+ K
W
( )
(3.9)
Where,
M
1
=sprung mass at one wheel
m=unsprung mass
For wheel hop and ride to be equally damped,
c
crit,M
= c
crit,m
(3.10a)
This can be rearranged to give
m
M
1
=
K
W
K
T
1+
K
W
K
T

\

¹

2 (3.10b)
(Milliken & Milliken 2002)
Below is a graph of K
W
/K
T
vs. m/M
1
39
Graph 3.1: Wheel Hop: Rate and Mass Ratios
Both spring combination candidates have been plotted onto the graph. For both spring
combinations, the mass ratios are the same.
For the front wheels,
m
M
1
≈ 0.27
For the rear wheels,
m
M
1
≈ 0.25
The rate ratios differ between the spring combinations.
Combination 1
40
The rate ratio for the front wheels is
K
W
K
T
≈ 0.14
The rate ratio for the rear wheels is
K
W
K
T
≈ 0.19
Combination 2
The rate ratio for the front wheels is
K
W
K
T
≈ 0.09
The rate ratio for the rear wheels is
K
W
K
T
≈ 0.14
Anywhere to the right of the “equal damping” line means that wheel hop damping is less
than ride damping. This means that in order to damp out wheel hop, the ride will be
overdamped. As can been seen from the graph, the plotted points (especially the front
wheel from Combination 2) are closer to “50% damping on unsprung” than to “equal
damping”.
To get closer the “equal damping” line, and therefore not to have to overdampen ride to
prevent wheel hop, there are a variety of changes that can be made.
I. The unsprung mass could be reduced. This could prove difficult, but even
manufacturing the wishbones out of carbon fibre instead of mild steel would help
improve the situation.
II. The sprung mass could be increased. An acceptable way of doing this would be to
modify the battery cage to house a larger number of lithiumion batteries. Although
this would be expensive, it would not only increase the sprung mass, but it would
also allow the car to travel a greater distance before having to recharge the batteries.
Another way would be to put some panels and a body kit on the car, as the chassis is
41
currently just a “skeleton” frame, as can be seen on the screen shot of the Solid
Works model of the car in Figure 2.1.
III. Stiffer springs could be used. This would be a simple change to make, however, if
the springs are too stiff, the ride will be too uncomfortable for the driver. For
further information, please refer to “The problem associated with using springs that
are too stiff”.
Vertical accelerations produced by the spring combination candidates
Both candidates for the spring combination were entered into the Half Car Model.
The “road” parameters were arbitrarily chosen as follows.
A
1
= 0.1m
s = 10m
The wheel damping rates were arbitrarily chosen to be 15% of the critical damping of the
unsprung masses (damping ratio, ζ=0.15). Please refer to Appendix F for the calculation of
the damping rates.
Below are the MATLAB generated graphs of magnitude of vertical acceleration vs.
frequency of the two unsprung masses and the sprung mass from the Half Car Model using
both combinations of springs. Please refer Appendix G for the MATLAB code for the Half
Car Model vertical accelerations.
42
Graph 3.2: Vertical accelerations of the three masses at frequencies between 035Hz.
Below is a MATLAB generated graph showing the magnitude of vertical acceleration vs.
frequency of the front unsprung mass from the Half Car Model using both combinations of
springs.
43
Graph 3.3: Vertical acceleration of the front unsprung mass at frequencies between 0
35Hz
Below is a MATLAB generated graph showing the magnitude of vertical acceleration vs.
frequency of the rear unsprung mass from the Half Car Model using both combinations of
springs.
44
Graph 3.4: Vertical acceleration of the rear unsprung mass at frequencies between 035Hz
As can be seen from the accelerationfrequency graphs of the front and rear unsprung
masses, Combination 2 produces slightly larger vertical acceleration of the unsprung
masses compared to Combination 1, although this difference is so small that it can be
considered negligible.
Vehicle suspension is used to minimise discomfort cause by irregularities in the road. A
common method of determining the extent of this discomfort is to measure the vertical
acceleration experienced by the driver, i.e. the vertical acceleration of the sprung mass
(Milliken & Milliken 1995).
Below is a MATLAB generated graph showing a “close up” of the magnitude of vertical
acceleration vs. frequency of the sprung mass from the Half Car Model using both
combinations of springs.
45
Graph 3.5: Vertical acceleration of the sprung mass at frequencies between 025Hz
On closer inspection of the accelerationfrequency graph of the sprung mass, it can be seen
that Combination 2 produces lower vertical accelerations compared to Combination 1. This
means that Combination 2 would provide a more comfortable ride for the driver, and,
therefore would be the most appropriate combination of springs to be used on the REV car.
Most appropriate springs to be used on the REV car
So the most appropriate springs to be used on the REV car are:
Rear Springs Front Springs
Combination 2 450 x 2.39 300 x 2.57
Table 3.8: Most appropriate springs to be used on the REV car
46
The problem associated with using springs that are too stiff
Referring to the “Wheel Hop: Rate and Mass Ratios” graph, one of the suggestions to get
closer to the “equal damping” line was to use stiffer springs. The problem associated with
using springs that are too stiff is that the ride can become too uncomfortable for the driver.
This can be shown by the following extreme example.
On the “Wheel Hop: Rate and Mass Ratios” graph, the plotted points have an m/M
1
ratio of
either 0.25 or close to 0.25. On the “equal damping” line, an m/M
1
ratio of 0.25
corresponds to a K
W
/K
T
ratio of 1.
This means that in order to be on the “equal damping” line, the wheel rate must be the same
as the tyre rate, which is 180 190 N/m. Taking into account the motion ratio, this
corresponds to a spring rate of just over 3200lbs/in.
Both the new “stiff spring” (where the wheel rate is equal to the tyre rate) and the springs
from Combination 2 were entered into the Quarter Car Model.
Arbitrarily, the “road” amplitude A
1
was chosen to be 0.1m, and the wheel damping rates
were chosen to be 15% of the critical damping of the unsprung masses (damping ratio,
ζ=0.15). Please refer to Appendix H for the calculation of the damping rates.
Below are the MATLAB generated graphs of magnitude of vertical acceleration vs.
frequency of the unsprung and the sprung masses of the front wheel from the Quarter Car
Model using both the “stiff spring” and the front spring from Combination 2. Please refer
Appendix I for the MATLAB code for the Quarter Car Model vertical accelerations.
47
Graph 3.6: Vertical accelerations of the sprung and unsprung masses of the front wheel at
frequencies between 035Hz
Below are the MATLAB generated graphs of magnitude of vertical acceleration vs.
frequency of the unsprung and the sprung masses of the rear wheel from the Quarter Car
Model using both the “stiff spring” and the rear spring from Combination 2.
48
Graph 3.7: Vertical accelerations of the sprung and unsprung masses of the rear wheel at
frequencies between 035Hz
As can be observed from the two previous graphs, the “stiff spring” produces very large
“spikes” in the vertical accelerations of the sprung masses of the front and rear wheels.
This is a clear indication that using springs this stiff would cause the driver to be very
uncomfortable.
Changing amplitude, A
1
Below are the MATLAB generated graphs of magnitude of vertical acceleration vs.
frequency of the two unsprung masses and the sprung mass from the Half Car Model using
the Combination 2 springs. These graphs show the effect of changing the “road” amplitude
A
1
.
49
Graph 3.8: Changing the amplitude
The MATLAB graphs shows curves for A
1
=0.03m, A
1
=0.1m and A
1
=0.17m.
It is clear from observing the graphs that increasing the amplitude of the “road” causes the
vertical accelerations of each of the masses to increase.
Changing damping ratio, ζ
Below are the MATLAB generated graphs of magnitude of vertical acceleration vs.
frequency of the two unsprung masses and the sprung mass from the Half Car Model using
the Combination 2 springs. These graphs show the effect of changing the damping ratio ζ.
50
Graph 3.9: Changing the damping ratio
The MATLAB graphs shows curves for ζ=0.05, ζ =0.15 and ζ =0.25.
In the case of the front and rear unsprung masses, increasing the damping ratio reduces the
vertical accelerations at all frequencies.
However, in the case of the sprung mass, increasing the damping ratio reduces the
acceleration peaks at around 3Hz and 17Hz, but causes the acceleration level over the
remainder of the frequency range to be increased.
Using a damping ratio of 0.15 provides a reasonable compromise of reducing the unsprung
mass acceleration peak, while reducing both sprung mass acceleration peaks, and producing
sprung mass acceleration levels that are not too high for the remainder of the frequency
range.
51
4. Safety
It was vitally important that high levels of safety were adhered to during this project, to
ensure the wellbeing of the author and other members of the University, including other
students, lecturers, tutors and lab technicians. It is responsibility of everybody associated
with the REV project to assist in creating and maintaining a culture of safety, encouraging
all participants to remain vigilant against possible health and safety hazards.
Emergency phone numbers
In the case of any emergency, call 2222 (from any internal university phone) or 6488 2222
(from a mobile phone), and give the following information:
• type of emergency
• location
• name
• mobile phone number or the telephone extension of the room
Below is a list of some other important phone numbers
Hospital (QE 2 ‘G Block’ Emergency) (0) 9346 3380
Doctor (UWA Medical Centre) Extension 2118
Poisons Information Centre (All hours) 0 13 11 26
Fire procedure
In the event of a fire, smoke detectors should sound an alarm. If the fire is too large to
extinguish, or the person or persons are not comfortable with fighting the fire, they should
immediately evacuate the area, and assemble at the designated evacuation area. If the fire
is small and the person or persons are comfortable with fighting the fire, they should select
an appropriate fire extinguisher and extinguish the fire from a safe distance. Most fire
52
extinguishers are the carbon dioxide or dry chemical type, and they are suitable for most
fire. Beware that foam and water extinguishers are not to be used on electrical fires, and
water extinguishers are not to be used on flammable liquid fires.
Smoking
To protect people from tobacco smoke, smoking is strictly forbidden within buildings or
within 5 metres from the entrance to a building. Also, smoking is not permitted in the
vicinity of gas storage bottles.
Electrical safety
It is important that all electrical equipment is properly used and well maintained. All
equipment should be regularly inspected, and any equipment believed to be unsafe or faulty
should be clearly tagged, and not used until it has been fixed and approved by a qualified
electrician.
Working in labs
When working in the labs, personal protective equipment (PPE), such as fully enclosed
footwear and approved safety glasses, must be worn at all times. When working alone
outside of the normal operating hours of 9am to 5pm, people are not permitted to use tools.
Driving the REV car
When driving the REV car, fully enclosed footwear and a helmet must be worn. The driver
must also be strapped in the seat with a fivepoint safety harness.
If when the car is in motion, one or more of the wheels starts to oscillate violently due to
wheel hop or some other form of unsprung mass oscillation, it is important not to
53
immediately brake. This will cause the frequency of vibration to decrease, while the
amplitude will increase significantly, which can lead to loss of control of the car. Instead, it
is recommended that the driver simply takes their foot off the accelerator and allows the car
to slow down until the violent oscillation of the wheel(s) stops. The car suspension should
then be immediately serviced, meaning that the dampers should be either retuned or
replaced (if damaged) so that they are able to damp out any violent oscillations of the
wheels.
54
5. Conclusion and Future Work
After considering both spring selection criteria and observing the accelerationfrequency
graphs generated by MATLAB, the most suitable spring combination to use on the REV
car is the 300 x 2.57 springs at the front and the 450 x 2.39 springs at the rear (Combination
2). As in Combination 1, these springs satisfied both criteria and produced very similar
vertical accelerations for both the front and rear unsprung masses, however, they produced
much smaller vertical accelerations of the sprung mass compared to Combination 1,
meaning that they would provide a much more comfortable ride for the driver.
After observing the “Wheel Hop: Rate and Mass Ratios” graph, it became apparent that
there are three methods in order get closer to the “equal damping” line, and to avoid
overdamping the ride to control wheel hop. The first method is to reduce the unsprung
mass. This could be achieved by replacing the current mild steel wishbones with ones that
are manufactured out of carbon fibre. Another way to reduce unsprung mass would be to
redesign the uprights so that they are lighter, although doing this would possibly be a final
year project by itself. The second method is to increase the sprung mass. As can be seen
by the screen shot Solid Works model in Figure 2.1, the car chassis consists only of a
frame, and is in “skeleton” form. By adding panels and a body it, the sprung mass would
be increased, and the car would be more aesthetically pleasing and may even perform better
aerodynamically. The third and last method was to use stiffer springs. This, however,
should be done with caution, as using springs that are too stiff can have a detrimental effect
on ride.
There is plenty of future research that can be done on the REV car. This project has only
simulated situations where the car is pitching. It would be desirable if future REV students
could model transient situations where the car is turning, and the associated rolling, yawing
and pitching of the car.
Another interesting study would be to design and implement a springdamper arrangement
that could be attached to the base of the seat, which could damp out any violent oscillations
55
of the sprung mass. This would allow stiffer springs to be used, in order to control wheel
hop, and would ensure that the ride was comfortable for the driver.
If the REV car is going to compete in future competitions, it would be advantageous for a
kinetic suspension system to be investigated. The UWA Motorsport team has been using
this technology since 2004, and it seems as though using a kinetic suspension system will
be the way of the future
56
6. References
Aird, F 1997, Race Car Chassis: Design and Construction, MBI Publishing Company,
Osceola, WI
Costin, M & Phipps, D 1962, Racing and Sports Car Chassis Design, Robert Bentley, Inc,
Cambridge, Massachusetts
Desert Rides 2007, Desert Term Glossary, Desert Rides, Available from:
<http://www.desertrides.com/reference/images/terms/toe.gif>, [21 October 2009]
Finlayson, D 2003, Design, Optimisation & Development of a FSAE Suspension System,
University of Western Australia, Perth
Guzzomi, F 2004, Modification, Testing and Analysis of FormulaSAE Dampers,
University of Western Australia, Perth
Milliken, W & Milliken, D 1995, Race Car Vehicle Dynamics, Society of Automotive
Engineers, Inc., Warrendale, PA
Milliken, W & Milliken, D 2002, Chassis Design: Principles and Analysis, Society of
Automotive Engineers, Inc., Warrendale, PA
Smith, C 1978, Tune to Win, Aero Publishers, Inc., Fallbrook, CA
Staniforth, A 1999, Competition Car Suspension 3
rd
Edition, Haynes Publishing, Sparkford,
UK
Waldron, K & Kinzel, G 2004, Kinematics, Dynamics, and Design of Machinery, John
Wiley & Sons, Inc., Hoboken, NJ
57
Winzer, N 2002, Design of a Formula SAE Racing Car Suspension System, University of
Western Australia, Perth
58
7. Appendices
Appendix A: Negative swing arm effect
The negative swing arm effect causes camber to become increasingly positive as the wheel
rises, as shown below.
Figure A: Negative swing arm effect (Milliken & Milliken 2002)
59
Appendix B: Determining a and b
A set of scales was placed underneath each wheel of the car, which was “loaded” with a 75
kg driver.
The average scale readings of the front and rear wheels was 66kg and 71kg respectively.
This information can be represented in the diagram below.
Figure B: Mass representation diagram
m b a l 85 . 1 = + =
( )
m b
m a
a a l
a b
M
CG
891 . 0
959 . 0
66 71
66 71
0
= ∴
= ∴
= −
= ∴
=
∑
60
Appendix C: Quarter Car Model Vertical Acceleration Calculations
[ ]
[ ]
[ ]
−
− +
=
−
−
=
=
W W
W W T
K K
K K K
ix SpringMatr
c c
c c
rix DampingMat
M
m
MassMatrix
0
0


¹

\

=
3
1
X
X
X


¹

\

=
0
0
3
1
0
X
X
X
t j
e X X
ω
0
=


¹

\

=
0
1
A
A
[ ] [ ] [ ]
t j
e A X ix SpringMatr X ix DamperMatr X MassMatrix
ω
= + +
& & &
[ ] [ ] [ ] A X ix SpringMatr X rix DampingMat j X MassMatrix = + + −
0 0 0
2
ω ω
61
[ ] [ ] [ ] ( ) A X ix SpringMatr rix DampingMat j MassMatrix = + + −
0
2
ω ω
[ ] [ ] [ ] ix SpringMatr rix DampingMat j MassMatrix B + + − = ω ω
2
A B X
1
0
−
=
t j
t j
t j
e X X
e X j X
e X X
ω
ω
ω
ω
ω
0
2
0
0
− =
=
=
& &
&
62
Appendix D: Road Excitation Calculations
Assuming that road is a sine wave with amplitude A
1
and a length of one cycle of road
profile of s, pictured below:
Figure D: Assumed road
Period of oscillation, T:
Where, s= length of one cycle of road profile
v= velocity
Frequency:
Excitation at front wheel:
Converting to complex form:
t j
e A
ω
1
Rear wheel will be excited some time after the front wheel, depending on the velocity of
the vehicle.
Therefore, excitation at rear wheel:
T =
s
v
ω =
2π
T
=
2π
s
v
=
2πv
s
A
1
sinωt
63
Converting to complex form:
t j
e
s
l j
j A
ω
π

¹

\

−
2
exp
1
A
1
sinω t + ∆t
( )
= A
1
sinω t +
l
v

\

¹

= A
1
sinω t +
2πl
ωs

\

¹

= A
1
sin ωt +
2πl
s

\

¹

l = wheelbase
64
Appendix E: Half Car Model Vertical Acceleration Calculations
+ − −
+ − −
− +
− +
=
+ − −
+ − −
−
−
=
=
r f r f r f
r f r f
r r r
f f f
W W W W W W
W W W W
W W T
W W T
r f r f r f
r f r f
r r
f f
r
f
K b K a bK aK bK aK
K K K K
K K K
K K K
ix SpringMatr
c b c a bc ac bc ac
c c c c
c c
c c
ix DamperMatr
I
M
m
m
MassMatrix
2 2
2 2
0
0 0
0 0
0
0 0
0 0
0 0 0
0 0 0
0 0 0
0 0 0





¹

\

=
θ
3
2
1
X
X
X
X





¹

\

=
0
3
2
1
0
0
0
0
θ
X
X
X
X
65
t j
e X X
ω
0
=






¹

\


¹

\

−
=
0
0
2
exp
1
1
s
l j
j A
A
A
π
[ ] [ ] [ ]
t j
e A X ix SpringMatr X ix DamperMatr X MassMatrix
ω
= + +
& & &
[ ] [ ] [ ] A X ix SpringMatr X rix DampingMat j X MassMatrix = + + −
0 0 0
2
ω ω
[ ] [ ] [ ] ( ) A X ix SpringMatr rix DampingMat j MassMatrix = + + −
0
2
ω ω
[ ] [ ] [ ] ix SpringMatr rix DampingMat j MassMatrix B + + − = ω ω
2
A B X
1
0
−
=
t j
t j
t j
e X X
e X j X
e X X
ω
ω
ω
ω
ω
0
2
0
0
− =
=
=
& &
&
66
Appendix F: Damping calculations for the Half Car Model
crit
c c × = ζ
( )
W T crit
K K m c + = 2
15 . 0 = ζ
Damping of front unsprung mass:
Combination 1
c
crit
= 3846.7 Ns/m
c = 0.15 x 3846.7 = 577.0 Ns/m
Combination 2
c
crit
= 3766.9 Ns/m
c = 0.15 x 3766.9 = 565.0 Ns/m
Damping of rear unsprung mass:
Combination 1
c
crit
= 3924.9 Ns/m
c = 0.15 x 3924.9 = 588.7 Ns/m
Combination 2
c
crit
= 3846.7 Ns/m
c = 0.15 x 3846.7 = 577.0 Ns/m
67
Appendix G: MATLAB Code for Half Car Model Vertical Accelerations
Below is the MATLAB code for determining the vertical accelerations of the front
unsprung, rear unsprung and sprung masses in the Half Car Model.
%Half Car Model  Acceleration
%Peter Corke 20139918
clear;
%Changeable variables
kwf=; % front wheel spring rate [N/m]
kwr=; % rear wheel spring rate [N/m]
cf=; % front wheel damping rate [Ns/m]
cr=; % rear wheel damping rate [Ns/m]
%Constant variables
mf=; % front unsprung mass [kg]
mr=; % rear unsprung mass [kg]
M=; % sprung mass of half of the car[kg]
I=; % mass moment of inertia of sprung mass [kg
m^2]
a=; % distance from front wheel to CG [m]
b=; % distance from rear wheel to CG [m]
l=a+b; % wheelbase [m]
ktf=; % front tyre spring rate [N/m]
ktr=; % rear tyre spring rate [N/m}
A1=; % Amplitude [m]
s=; %length of one cycle of road profile [m]
68
hzmin=; % Frequency range [Hz]
hzmax=;
n=1000;
hzamount=(hzmaxhzmin)/(n1);
Mass_Matrix=[mf,0,0,0;0,mr,0,0;0,0,M,0;0,0,0,I];
Damper_Matrix=[cf,0,cf,0;0,cr,cr,0;cf,cr,cf+cr,0;
a*cf,b*cr,a*cfb*cr,a^2*cf+b^2*cr];
Spring_Matrix=[ktf+kwf,0,kwf,0;0,ktr+kwr,kwr,0;kwf,
kwr,kwf+kwr,0;a*kwf,b*kwr,a*kwfb*kwr,a^2*kwf+b^2*kwr];
A=[A1;A1*j*exp((j*2*pi*l)/s);0;0];
wmin=hzmin*2*pi;
wmax=hzmax*2*pi;
wamount=hzamount*2*pi;
k=1;
for w=wmin:wamount:wmax;
B=w^2*Mass_Matrix+j*w*Damper_Matrix+Spring_Matrix;
X0(:,:,k)=B\A;
Acc(:,:,k)=w^2*X0(:,:,k);
freq(k)=w;
k=k+1;
end;
[x,y,z]=size (Acc);
for u=1:1:z;
X1 (u)=Acc(1,1,u);
X2 (u)=Acc(2,1,u);
X3 (u)=Acc(3,1,u);
end;
freq_hz=freq/(2*pi);
magnitudeX1=abs(X1);
69
magnitudeX2=abs(X2);
magnitudeX3=abs(X3);
figure (1);
subplot (3,1,1);
plot (freq_hz,magnitudeX1,'b');
xlabel ('Frequency (Hz)');
ylabel ('Magnitude of Acceleration (m/s^2)');
title ('Front Unsprung Mass');
subplot (3,1,2);
plot (freq_hz,magnitudeX2,'b');
xlabel ('Frequency (Hz)');
ylabel ('Magnitude of Acceleration (m/s^2)');
title ('Rear Unsprung Mass');
subplot (3,1,3);
plot (freq_hz,magnitudeX3,'b');
xlabel ('Frequency (Hz)');
ylabel ('Magnitude of Accerleration (m/s^2)');
title ('Sprung Mass');
70
Appendix H: Damping calculations for the Quarter Car Model
crit
c c × = ζ
( )
W T crit
K K m c + = 2
15 . 0 = ζ
Damping of front unsprung mass:
Stiff spring
c
crit
= 5093.9 Ns/m
c = 0.15 x 5093.9 = 764.1 Ns/m
Combination 2
c
crit
= 3766.9 Ns/m
c = 0.15 x 3766.9 = 565.0 Ns/m
Damping of rear unsprung mass:
Stiff spring
c
crit
= 5093.9 Ns/m
c = 0.15 x 5093.9 = 764.1 Ns/m
Combination 2
c
crit
= 3846.7 Ns/m
c = 0.15 x 3846.7 = 577.0 Ns/m
71
Appendix I: MATLAB Code for Quarter Car Model Vertical Accelerations
Below is the MATLAB code for determining the vertical accelerations of the unsprung and
sprung masses in the Quarter Car Model.
%Quarter Car Model  Acceleration
%Peter Corke 20139918
clear;
m=; % unsprung mass [kg]
M=; % sprung mass at one wheel[kg]
c=; % wheel damping rate [Ns/m]
kt=; % Tyre spring rate [N/m]
kw=; % Wheel spring rate [N/m]
A1=; % Amplitude [m]
hzmin=; % Frequency range [Hz]
hzmax=;
n=1000;
hzamount=(hzmaxhzmin)/(n1);
Mass_Matrix=[m,0;0,M];
Damper_Matrix=[c,c;c,c];
Spring_Matrix=[kt+kw,kw;kw,kw];
A=[A1;0];
wmin=hzmin*2*pi;
wmax=hzmax*2*pi;
wamount=hzamount*2*pi;
k=1;
for w=wmin:wamount:wmax;
B=w^2*Mass_Matrix+j*w*Damper_Matrix+Spring_Matrix;
X0(:,:,k)=B\A;;
Acc(:,:,k)=w^2*X0 (:,:,k);
freq (k)=w;
k=k+1;
end;
[x,y,z]=size (Acc);
for u=1:1:z;
X1 (u)=Acc (1,1,u);
X3 (u)=Acc (2,1,u);
end;
72
freq_hz=freq/(2*pi);
magnitudeX1=1000*abs (X1);
magnitudeX3=1000*abs (X3);
figure (1);
subplot (2,1,1);
plot (freq_hz,magnitudeX1,'b');
xlabel ('Frequency (Hz)');
ylabel ('Magnitude of Acceleration (mm/s^2)');
title ('Unsprung Mass');
subplot (2,1,2);
plot (freq_hz,magnitudeX3,'b');
xlabel ('Frequency (Hz)');
ylabel ('Magnitude of Acceleration (mm/s^2)');
title ('Sprung Mass');
Project Summary
The purpose of this project is to analyse and optimise the suspension system of the 2009 REV race car. An effective suspension system acts as a vibration filter, and protects the vehicle from excitation caused by irregularities in the road.
Suitable coil springs are to be selected for the front and rear of the car, to ensure that both ends do not oscillate violently out of phase after going over a bump. The Half Car Model, which has four degrees of freedom, can be used to simulate the vehicle driving in a straight line. By providing excitation to the wheels, the model can be used to show the vertical accelerations of the sprung mass and the front and rear unsprung masses at different frequencies. The selected spring combination should give the least possible vertical accelerations of the sprung and unsprung masses. A common method of determining the extent of driver discomfort is to measure the vertical acceleration of the sprung mass.
It is also important that the suspension is able to damp the unsprung masses adequately to prevent wheel hop. Wheel hop is the vertical vibrating motion of the wheel between the sprung mass and the road, and has the capacity to impede handling and performance.
The Half Car Model will also be used to determine the effects of changing the amplitude of the assumed road, and changing the damping rate of the wheel.
2
Letter of transmittal
Peter Corke 6/57 Lockhart Street Como, WA, 6152 23rd October, 20091020
Professor David Smith Dean Faculty of Engineering, Computing and Mathematics University of Western Australia 35 Stirling Highway Crawley, WA, 6009
Dear Professor Smith,
I am pleased to submit this thesis, entitled “Analysis of the 2009 REV Race Car Suspension”, as part of the requirement for the degree of Bachelor of Engineering.
Yours sincerely
Peter Corke 20139918
3
for being great people to work with. All of the REV team. Lachlan Tomlin. My friends and family. 4 . Chris Rowles and Grant Keady.Acknowledgements I would like to thank the following people who have helped make this project possible. for giving me good advice with my computer modelling. for the support that they have given me this year. for the hours that he put into the race car.
.......................................... 35 Criteria1................................................................................................................................................................................................................................................................................................................................................................................................ A1 ......................................................... 37 Wheel hop: rate and mass ratios..................................................................................................................................................................................... 51 Fire procedure ....................................................... 26 3..................................................... 10 Importance of suspension............................ 9 Toe............................... 18 2......................................................................................... Safety................................................................................................................... 8 Camber .................................... 8 Unsprung mass............................................................................................................................................................................................................................................................................................................ l ............ 7 b..................................... ζ .......................................................................................................................................................................................................................................... 22 Quarter Car Model ............................................................................. 7 Project objectives .............. 20 Car information ............... 51 Emergency phone numbers ..................................... 13 Unsprung mass oscillations.......................................................................................................................................................................................................................................................................................... 38 Vertical accelerations produced by the spring combination candidates ................................................................................................... 46 Changing amplitude...... 33 Undamped natural frequencies of the front and rear sprung masses .................................................................................................................................................................................................. 41 Most appropriate springs to be used on the REV car .................. 31 Wheel rates.................................................................................................. 49 4.... 32 Tyre rate ......................... 35 Criteria 2......... 15 Wheelfight ........................ 51 5 ................ 31 Available springs........ Model Formulation .............. 21 Fox Vanilla RC Shox ............................................................................................................................................................................................. 8 Wheelbase....................................................................................................... 36 Spring combination candidates..................................................................... 12 Unequal length wishbones ....................................................................................................................................................................................................... 33 Ride rates...................................................................... 14 Shimmy ................................................................................................. 10 Wishbone independent suspension ............................................................................................ 7 a..................... 20 Coil springs .. Results ...... 8 Sprung mass ........ 34 Spring selection......................................................................... Introduction ............................... 7 Definitions....................................................................... 8 Axes ................................ 31 Motion ratio.......................................................................................Table of Contents 1....................................... 45 The problem associated with using springs that are too stiff ................................................................................................................................................................ 48 Changing damping ratio............................................................................................................................................................................................... 7 CG ................................................................................................................................................................................................................................... 11 Equal length and parallel wishbones.................................................................................................................................................................................................................................... 24 Half Car Model ........................ 17 Wheel hop......................................................................................................................................................................................................................................................
.......................................................................................................................... 54 6................................................................ 70 Appendix I: MATLAB Code for Quarter Car Model Vertical Accelerations ......................... 52 Driving the REV car....................................................... 62 Appendix E: Half Car Model Vertical Acceleration Calculations... 67 Appendix H: Damping calculations for the Quarter Car Model ........... References .......................... 64 Appendix F: Damping calculations for the Half Car Model................................................Smoking ........................... 56 7.............................. 60 Appendix D: Road Excitation Calculations ................................................................ 52 Working in labs ................................................. 59 Appendix C: Quarter Car Model Vertical Acceleration Calculations ................. 58 Appendix A: Negative swing arm effect.................................................................................................................... Appendices......................................................................................................... 71 6 .................................................................. 66 Appendix G: MATLAB Code for Half Car Model Vertical Accelerations....................................................................................................................................................................................... Conclusion and Future Work ...................................... 52 Electrical safety......... 58 Appendix B: Determining a and b .... 52 5......................................................................................................................................
batteries and hub motors. as few alterations as possible were made to the car. The aim of this thesis was going to be to redesign the suspension system of the vehicle. Introduction Project objectives The original plan for the REV team was to receive the 2001 Motorsport SAE car in ‘rolling chassis’ form at the beginning of the year. Therefore. this thesis became concerned with analysing and optimising the existing suspension system of the car instead of designing a new one. making it more suitable to the new weight distribution caused by replacing the petrol motor with a battery cage. Unfortunately. due to significant time delays. consequently. About midway through second semester. it was decided that inboard motors would be used as this approach allowed for much more space and easier installation. it was decided that trying to design and install a hub motor setup complete with pulleys into each of the existing rear uprights of the car was too difficult. and to transform it into a batterypowered race car with hub motors.1. and. in order to get a drivable car constructed by the end of the year. Therefore. Definitions a ‘a’ is the distance between the front track and the centre of gravity (CG) b ‘b’ is the distance between the rear track and the centre of gravity (CG) 7 . the ‘rolling chassis’ was not received until the mid year holidays.
Sprung mass The sprung mass is comprised of parts of the vehicle that are supported by the suspension. such as the wheels. The sprung mass moves when the springs are deflected (Waldron & Kinzel 2004). 8 . batteries and motors. Wheelbase. battery cage. roll and yaw axes of a vehicle. driver. tyres and uprights (Waldron & Kinzel 2004).CG CG is an acronym for ‘centre of gravity’. The Half Car Model later in this paper is concerned with the pitch axis. such as the chassis. l The wheelbase (l) is the distance between the front track and the rear track Axes Below is a diagram showing the pitch. Unsprung mass The unsprung mass is comprised of parts of the vehicle that are not supported by the suspension.
roll and yaw axes (Milliken & Milliken 1995) Camber Below is a diagram showing the camber angle.1: Pitch.Figure 1.2: Camber angle (Milliken & Milliken 1995) Below is a diagram showing a negative camber angle of both wheels. 9 . Figure 1.
Please note the oval shape represents the driver at the front of the car.3: Negative camber (Milliken & Milliken 1995) Toe Below is a diagram showing both toe in and toe out.Figure 1. maintain the greatest possible contact with the road. especially the two attached to the driving wheels.4: Toe in and toe out (Desert Rides 2007) Importance of suspension The key requirement of a suspension system is to ensure that all tyres. Figure 1. The 10 .
including pressure and tread depth Wheel and tyre balance A successful suspension system generally has nonlinear camber change. which means that the steering requires no damping. including: • • • • • • • • • • • Vehicle weight Vehicle weight distribution Sprung to unsprung weight ratio Location of the centre of gravity Roll centre heights on the front and rear suspensions Wheelbase and track Suspension layout Steering geometry Loading condition Characteristics of the tyres. 11 . Also. This allows the driver to have a better “feel” for the handling of the car. especially the outside wheels. The front end does not experience shimmy. should remain as upright as possible. which is important for race car driving. 1. The stability of a racing car is influenced by many factors.suspension should protect the vehicle from jolts in the road. An effective suspension system is very much a compromise between roadholding and ride. the wheels. Wishbone independent suspension Wishbone independent suspension systems do not encounter as many problems as beam axles. bump steer can be entirely eradicated at any particular steering angle. which means that camber is not directly dependent on vehicle roll (Costin & Phipps 1962). While cornering. and provide stability and roadholding throughout the entire power range of the motors.
which can increase the cornering power of the tyres. Also. means that when the vehicle is traveling forward. (Aird 1997) The advantage with wishbone suspension systems is that they are normally built with a steering linkage that is symmetrical about the centreline of the vehicle. however. Since there is much less mass moving up and down with the wheels. it is easier to maintain contact between the tyres and the road. This causes the tyre edges to wear and overheat. This. The rear wheels can be cambered. To ensure that the treads of the tyres remain flat on the road while cornering. the tyres will be running on their edges. and does not allow for much traction when accelerating or braking (Aird 1997). especially on uneven surfaces.5: Equal length and parallel wishbones in roll (Aird 1997) 12 . any amount of toein or toeout can be provided. Unsprung mass is heavily reduced. This means that any “inaccuracies” in the linkage will not cause the vehicle to veer off unnecessarily (Milliken & Milliken 2002). Equal length and parallel wishbones The main disadvantage of having equal length and parallel wishbones is that the wheels become cambered the wrong way when the chassis rolls in cornering. 3. Figure 1. a substantial amount of negative camber can be incorporated into the wheels.2.
although it introduces a new problem. The vertical motion of the wheel is accompanied by some sideways motion.6: Equal length and parallel wishbones in bump (Aird 1997) Unequal length wishbones The REV car has independent suspension with unequal length wishbones. which occurs when a rotating wheel is tilted. This completely eliminates the problem of gyroscopic “kick”. 13 . the wheels will rise and fall without cambering. by making the top wishbone shorter than the bottom one. This is due to the line of the axis the wheel steers around crossing the ground ahead of the tyre contact patch’s centre (Aird 1997). it is possible to combine small amounts of camber and scrub. that is. By using unequal length wishbones. referred to as scrub.Using equal length and parallel wishbones also means that as the suspension moves vertically in bump and rebound. Figure 1. and this generates “kickback” in the steering. Having small amounts of camber and scrub prevents the detrimental effects of having large quantities of either by itself (Aird 1997).
it is essential to avoid too rigid a limitation of the wheel motion. because. in racing cars. if they are coupled in a particular way.Figure 1. in theory. Of the listed varieties of oscillation. Wheel hop Wheels that are mounted to independent suspension linkages have an advantage over wheels that are attached to beam axles. only shimmy is indefinitely selfsustaining on a smooth road (Milliken & Milliken 2002). However.7: Independent suspension with unequal length wishbones (Smith 1978) Unsprung mass oscillations The oscillations of the unsprung mass can cause potential problems. they have only one degree of freedom. Vibrations in the various modes can sometimes supply energy to each other. Shimmy 2. 14 . as opposed to six. Wheelfight 3. which therefore establishes additional degrees of freedom. These oscillations include: 1.
while the other wheel is in the air. or by an imbalance in one of the front wheels. This causes one front wheel to be flattened into the road. This can cause the car to become out of control. The shimmy cycle begins by one of the front wheels hitting a chance bump in the road. and damage of steering and suspension linkages. a tyre blowing or coming off its rim. the frequency of the shimmy decreases while its amplitude increases. 15 . This cycle is infinitely selfsustaining on a smooth road. and generally occurs at higher speeds. The positions of the wheels are then reversed. and is prevalent in cars where the front wheels are connected via a beam axle. such as the hands of the driver being shaken off the steering wheels.Shimmy Shimmy is considered the most hazardous type of unsprung mass oscillation. If the driver applies the front brakes during the shimmy cycle. Shimmy can cause many problems.
as well as having a torsionally stiff car frame. as it has independent suspension at both ends of the car.8: A severe shimmy cycle (Milliken & Milliken 2002) The most effective approach in preventing shimmy is to use independent front suspension. It is fair to state that the chassis is “overengineered”. and is manufactured from mild alloy steel 4140. The chassis.Figure 1. It is highly unlikely that a shimmy cycle will develop on the REV car. By placing the sprung mass of the car between the front wheel mountings. which is a part of the sprung mass. the gyroscopic torque of the shimmy cycle is unable to generate the high frequency required for shimmy (Milliken & Milliken 2002). 16 . is torsionally stiff.
Wheelfight SAE Vehicle Dynamics Terminology defines wheelfight as “a rotary disturbance of the steering wheel produced by forces acting on the steerable wheels” (Milliken & Milliken 2002).9: Steering gear test rig (Milliken & Milliken 2002) Wheelfight generally takes place when the frequency of the steering gear is close to the frequency at which wheel hop occurs. it relies on constant excitation supplied by irregularities in the road. Instead. Wheelfight is unlike shimmy as it is not selfsustaining. Any complex analysis of wheelfight requires a detailed investigation of the steering gear’s elastic resonance. 17 . Pictured below is a steering wheel test rig. Figure 1.
Therefore it is important to appropriately damp the unsprung masses. To verify this. Too much wheel hop has adverse effects on the handling and stability of the car. so that the driver is comfortable. iii. Primary ride is the lower frequency oscillation of the sprung mass. while secondary ride is the higher frequency oscillations of the unsprung masses. ii. i. as the steering gear will probably not reach high enough frequencies (wheel hop frequency). which is defined as “the vertical oscillatory motion of a wheel between the road surface and the sprung mass” (Milliken & Milliken 2002). Wheel hop Vehicle ride has two modes: they are primary ride (often simply referred to as ride) and secondary ride. and also the sprung mass. ii. To assist in preventing wheelfight at the steering gear: i.This means that preventative measures against wheelfight can be implemented in either of two locations: at the steering gear or at the steerable wheels (front wheels). increase the steering wheel’s moment of inertia reduce the steering wheel shaft diameter To assist in preventing wheelfight at the steerable wheels. Secondary ride is sometimes referred to as wheel hop. induce a negative swing arm effect (please refer to Appendix A) It is unlikely that wheelfight will pose a threat to the performance of the REV car. 18 . attach rubber mountings to the lower control arms have suspension linkages designed so that as the front wheels rise they toe out make the wheel centres move backwards as the wheel rises by sloping the wheel arm pivots iv. however. the natural frequency of the steering gear would have to be determined. this is beyond the scope of this project.
Generally transmissibility under 2. 19 .5 will not permit enough wheel hop to hinder handling (Milliken & Milliken 1995).Wheel transmissibility is defined as the ratio of the vertical wheel motion to road input amplitude.
1: Solid Works model of the car (Ivanescu & Tomlin 2009) The Solid Works software was used to calculate the mass moment of inertia of the sprung mass about its pitch axis through its centre of gravity. Below is a screen shot from a computer model of the car that was created using a software package called Solid Works.2. 20 . There are two inboard motors that are mounted to the chassis above the rear track that have a combined mass of 25kg. Figure 2. The car has an approximate mass distribution of 48% at the front and 52% at the rear. Model Formulation Car information The total mass of the car when it loaded with a 75kg driver is 274 kg. while the battery cage and batteries are located just behind the driver’s seat and have a combined mass of 55kg. This value was found to be 616.363 kg m2.
71kg 18kg 616. Information Wheelbase Distance from CG to front track Distance from CG to rear track Sprung mass Front sprung mass Rear sprung mass Unsprung mass (same at front and rear) Mass moment of inertia (about pitch axis through CG) Symbol l a b M Mf Mr m I 1.363kg m2 Table 2.959m 0. They can be used in both compression and extension (tension). in which the mean diameter of the coil remains constant (Milliken & Milliken 1995).85m 0.891m 202kg 97. The helix is the most widespread shape of coil spring.The important information about the car is expressed in the table below.2: Helixshaped coil spring (Milliken & Milliken 1995) 21 . Below is a diagram of a helixshaped coil spring. Coil springs Coil springs are the most commonly used type of springs in racing cars with independent suspensions.29kg 104. Figure 2.1: Important information about the car Please refer to Appendix B to see the method used for calculating a and b.
anodised aluminium. The Fox Vanilla RC is a telescopic type monotube damper.3:Fox Vanilla RC damper (Finlayson 2003) The Fox Vanilla RC damper has a casing that is manufactured out of hard. Figure 2.Fox Vanilla RC Shox The shock absorbers used in the REV car suspension are of the Fox Vanilla RC Shox variety. Below is a diagram of the Fox Vanilla RC damper. which increases wear resistance. 22 . It has an external tank of pressurised nitrogen gas. which means that its damping force is due to oil being forced through an orifice at high pressures. which is increasingly compressed as the piston rod moves further up the cylinder (Guzzomi 2004). It has a 25mm bore and has a mass of just over 315 g. which are commonly used on mountain bikes.
The damper is adjustable in both bump and rebound. To modify the bump settings, the spring preload on the relief valve can be altered (see Figure 2.3). A high spring preload will result in higher damping rates in bump. To modify the rebound settings, there is an adjustable needle valve that controls piston bypass in both bump and rebound, although it has only a small effect on bump damping (Guzzomi 2004).
Adjustments in bump and rebound are done in increments of ¼ turns, or “clicks”.
The graphs below show force (N) vs. velocity (m/s) for the Fox Vanilla RC damper in both bump and rebound. The numbers in the legend refer to the number of “clicks” open.
Bump
2500
2000
0 1
Force (N)
1500
3 5 7
1000
9 11
500
0 0 0.2 0.4 0.6 0.8 1 Velocity (m/s) 1.2 1.4 1.6
Graph 2.1: Fox Vanilla RC damper in bump
23
Rebound
5000 4500 4000 3500 Force (N) 3000 2500 2000 1500 1000 500 0 0 0.05 0.1 0.15 Velocity (m/s) 0.2 0.25 0.3
4 5 6 7 8 9 10 11 12 13
Graph 2.2: Fox Vanilla RC damper in rebound
It is noticeable in the “bump” graph that there is a distinct kink in each of the curves, which occurs at around the same velocity, regardless of the number of “clicks”. This is due to the presence of a “blowoff” valve, which causes the bypass of a proportion of the oil at high pressures. The reason for this is to avoid very high damping forces at high velocities (Milliken & Milliken 1995). From the graph, it can be seen that “blowoff” occurs at approximately 0.2 m/s.
Quarter Car Model The Quarter Car Model is used to simulate one corner of the car. It has two degrees of freedom: the vertical movement of the unsprung mass, and the vertical movement of the sprung mass.
Below is a diagram of the Quarter Car Model.
24
Figure 2.4: Quarter Car Model
Where
m=unsprung mass M1=sprung mass at one wheel X1=vertical displacement of the unsprung mass X3=vertical displacement of the sprung mass KW=wheel rate KT=tyre rate c=wheel damping rate
The Quarter Car Model assumes that the road is a sine wave with amplitude A1. Therefore, the excitation caused by the road is
F = A1e jωt
(2.1)
25
Below is a diagram of the Half Car Model. It has four degrees of freedom: the vertical movement of the front unsprung mass. the vertical movement of the centre of gravity of the sprung mass. is used to simulate one half of the car. Please refer to Appendix C for the Quarter Car Model vertical acceleration calculations. It is more realistic than the quarter car model. and the pitching of the sprung mass about its CG. 26 . the vertical movement of the rear unsprung mass.Assumptions of the Quarter Car Model Listed below are the assumptions incorporated into the quarter car model • • • • • The CG of the sprung mass is located directly above the CG of the unsprung mass The quarter of the vehicle being analysed acts independently to the rest of the vehicle There is no tyre damping The damping of the wheel is theoretical linear fluid damping The excitation from the road is sinusoidal excitation with amplitude A1 The Quarter Car Model will be used to determine the vertical accelerations of the sprung and unsprung masses. sometimes called the bicycle model. Half Car Model The Half Car Model. because the front and rear are coupled so that their motions affect one another.
5: Half Car Model Where mf=front unsprung mass mr=rear unsprung mass M=sprung mass of half of the car X1=vertical displacement of the front unsprung mass X2=vertical displacement of the rear unsprung mass X3=vertical displacement of the CG of the sprung mass KWf=front wheel rate KWr=rear wheel rate KT=tyre rate 27 .Figure 2.
and with a length of one cycle of road profile of s. Below is a diagram of the assumed road. the excitations caused by the road are: F1 = A1 e jωt j 2πl jωt F2 = A1 j exp − e s (2. (2. i.2a) (2.e. Figure 2. Below is a diagram of the sprung mass when it is at a pitch angle of θ.cf=front wheel damping rate cr=rear wheel damping rate l=wheelbase a=distance from CG to front track b=distance from CG to rear track The Half Car Model assumes that the road is a sine wave with amplitude A1.6: Assumed road Therefore.3) 28 .2b) Please refer to Appendix D to see how the above excitations were calculated. Not that the situation in the diagram occurs when the force from the front unsprung mass is greater than the force from the rear unsprung mass.
Where. for small angles of θ. θ. This can be assumed as the mass moment of inertia will vary very little for small pitch angles.7: Sprung mass from the Half Car Model pitched at an angle θ.4b) Assumptions of the Half Car Model Listed below are the assumptions incorporated into the half car model • • • • • • • The car is travelling in a straight line The half of the vehicle that is being analysed acts independently to the rest of the vehicle There is no tyre damping The damping of the wheels is theoretical linear fluid damping The road is a sine wave with amplitude A1 and with a length of one cycle of road profile of s The pitch angle. 29 . is small The mass moment of inertia.4a) (2. is constant. I. Y1 = X 3 + aθ Y2 = X 3 − bθ (2.Figure 2.
rear unsprung and sprung masses. Please refer to Appendix E for the Half Car Model vertical acceleration calculations. 30 .The Half Car Model will be used to determine the vertical accelerations of the front unsprung.
then the wheel rate would be equal to the spring rate.00 550 x 2.57 Spring Rate (N/m) 105076 96320 87563 78807 52538 Table 3. as well as the accompanying spring rates in SI units.30 450 x 2. For example. The spring rate and travel distance are written on the side of the spring.00 in. as pictured below. The table below lists the available Fox coil springs. Results Available springs The springs available to the REV team are an assortment of Fox coil springs.00 500 x 2. varying in spring rate (KS) and travel distance. Motion ratio If springs were mounted directly above the tyres.00 indicates a spring rate of 600 lbs/in and a travel distance of 2. 600 x 2.39 300 x 2. Spring 600 x 2.3.1: Spring rates of the available Fox coil springs. 31 .
it is necessary for the springs to be mounted inboard of the tyre centerline. due to limitations of space. This means that the wheel travel will not be equal spring travel. Figure 3.1: Spring mounted directly above wheel (Smith 1978) However.1) For example. Wheel rates The wheel rate (KW) is defined as KW = (Motion Ratio )2 KS (3. the motion ratio for the front and rear wheels was calculated to be 1.Figure 3.2: Spring mounted inboard of the tyre centerline (Smith 1978) Using experimental data from the 2001 Motorsport team.764. in the diagram below. The motion ratio is defined as Motion Ratio = Wheel Travel Spring Travel (3. 32 .2 (Smith 1978). the motion ratio is 2.2) The table below shows the available springs and the corresponding wheel rates.
30 Ride Rate (N/m) 28439 26416 24339 33 .39 300 x 2. The deflections of the wheel and the tyre are additive.57 Wheel Rate (N/m) 33768 30954 28140 25326 16884 Table 3. This value was used for further calculations involving both the front and rear tyre rates.00 550 x 2.00 550 x 2. so therefore the formula for two springs in series is used (Milliken & Milliken 2002).00 500 x 2.Spring 600 x 2.3a) KR = KTK W KT + KW (3.00 500 x 2.2: Available springs and the corresponding wheel rates.3b) The table below shows the available springs and the corresponding ride rates. Spring 600 x 2. The formula for ride rate is 1 1 1 = + KR KT KW or (3. Ride rates The ride rate (KR) is the combined rate of the wheel rate and the tyre rate. Tyre rate A standard Goodyear racing tyre suitable for the application of the REV car with an inflation pressure of 83kPa has a tyre rate (KT) of 180190N/m (Winzer 2002).30 450 x 2.
Therefore to find their composite rate.57 22205 15437 Table 3.39 300 x 2.733 34 . Spring 600 x 2.71kg respectively.848 3. The table below shows the available springs and the corresponding undamped natural frequencies of the front and rear sprung masses.709 3. These springs share a load.57 Natural Frequency of Front Sprung Mass (Hz) 3.39 300 x 2.450 x 2.00 500 x 2. they are added.30 450 x 2. This explains the “2KR” term in both of the equations above (Milliken & Milliken 1995).575 3. The front and rear sprung masses were calculated to be 97.4a) The natural frequency of the rear sprung mass is f Nr = 1 2π 2K R Mr (3.432 3.29kg and 104.278 2.709 3.835 Natural Frequency of Rear Sprung Mass (Hz) 3.00 550 x 2.4b) At each end of the car there are two wheels. meaning that at each end there are two “ride rate” springs.400 2.560 3.3: Available springs and the corresponding ride rates Undamped natural frequencies of the front and rear sprung masses The natural frequency of the front sprung mass is fNf = 1 2π 2K R Mf (3. and are said to be in parallel.
in order to avoid the front and rear of the vehicle oscillating violently out of phase after going over a bump (Staniforth 1999).95 (3. This means that f Nf f Nr ≈ 0. two different criteria had to be satisfied.5) However. Any combination that satisfied both criteria was then considered a possible candidate for use in the REV car suspension.6) to have satisfied Criteria 1. Therefore.75 ≤ f Nf f Nr ≤ 0. he states that his race experience has shown that this guideline is more of a “rule of thumb”.4: Available springs and the corresponding natural frequencies of the front and rear sprung masses Spring selection To select which combination of springs to use at the front and the rear of the car. the frequency of the front sprung mass should be about 10% lower than the frequency of the rear sprung mass. and is not always valid. 35 .Table 3. as a general guideline.90 (3. the author has been rather lenient and considered any spring combination which produces frequency ratios in the range of 0. Criteria1 Staniforth recommends that.
575 3.432 3.037 1 0.56 3.793 1. Rear Springs 600 x 2.00 500 x 2.00 550 x 2.30 450 x 2. Maurice Olley suggests that the natural frequencies of the front and rear sprung masses should be close.00 600 x 2.57 300 x 2.951 0.30 450 x 2.764 1.39 300 x 2. This means that f Nr ≤ 1.39 300 x 2.57 fNr (Hz) 3.Below is a table showing the values of fNf / fNr for each spring combination.57 fNf (Hz) 3.57 550 x 2.4 2.39 450 x 2. the natural frequency of the rear sprung mass should be no greater than 1.00 500 x 2.037 0.733 Front Springs 600 x 2.39 300 x 2.278 2.996 0.00 500 x 2.30 450 x 2.00 550 x 2.39 300 x 2.30 500 x 2.865 1.432 3.917 0.709 3. this is to avoid the situation where the front and rear of the vehicle oscillate violently out of phase after going over a bump.00 600 x 2.709 3.991 0.30 500 x 2.835 3.037 0. however.835 3.30 450 x 2.037 0.835 2. 36 .5: Criteria 1 Criteria 2 In his notes.39 300 x 2.575 3.4 2. The combinations satisfying Criteria 1 are highlighted in bold.00 550 x 2.2 times the natural frequency of the front sprung mass (Milliken & Milliken 2002).278 3.00 600 x 2.56 3.4 2.00 550 x 2.709 3.00 600 x 2.56 3.432 3.037 Table 3.709 3.57 450 x 2.7) As in Criteria 1.960 0.848 3.575 3.709 3.709 3.4 2.00 550 x 2.2 f Nf (3.57 500 x 2.826 1.835 3.709 3.835 fNf / fNr 1.575 3.
00 600 x 2.432 3.402 4.575 3.08 3.402 3.08 3.00 550 x 2.00 600 x 2.6: Criteria 2 Spring combination candidates Therefore.432 3.709 3.30 500 x 2.30 450 x 2.272 4.709 3.6176 4.39 300 x 2.00 450 x 2.56 3.00 500 x 2.278 2.709 3.00 550 x 2.709 3.4 2.402 4.30 450 x 2.848 3. The combinations satisfying Criteria 2 are highlighted in bold.00 500 x 2.733 Front Springs 600 x 2.30 450 x 2.57 3.00 550 x 2.272 4.57 550 x 2.2fNf for each spring combination.835 1.709 3.57 300 x 2.00 600 x 2.4 2.39 300 x 2.402 Table 3.Below is a table showing the values of 1.30 500 x 2.39 300 x 2.835 3.575 3.08 3.39 300 x 2.432 3.709 3.00 500 x 2.575 3.575 3.2 fNf 4.4 2.7: Candidates for the spring combination of the car 37 .00 550 x 2.00 600 x 2.402 4.835 2.4 2.57 fNf (Hz) 3. the combinations of springs satisfying both criteria are Combination Front Springs Rear Springs 600 x 2. Rear Springs fNr (Hz) 600 x 2.39 300 x 2.835 3.39 1 450 x 2.57 450 x 2.57 Table 3.08 3.56 3.39 2 300 x 2.4508 4.00 550 x 2.835 3.709 3.272 4.39 450 x 2.4508 4.278 3.57 500 x 2.30 450 x 2.56 3.
8) Critical damping of the unsprung mass is c crit.m = 2 m(KT + KW ) Where. (3. Critical damping of the sprung mass is c crit.m (3. c crit.10a) This can be rearranged to give m = M1 K 2 W 1+ KT (Milliken & Milliken 2002) KW KT (3.Wheel hop: rate and mass ratios Assuming that the oscillation centres of the front and rear sprung masses are directly above the front and rear tracks respectively. m/M1 38 .M = c crit.9) M1=sprung mass at one wheel m=unsprung mass For wheel hop and ride to be equally damped.10b) Below is a graph of KW/KT vs.M = 2 M1 KT KW KT + KW (3.
m ≈ 0. m ≈ 0. Combination 1 39 .27 M1 For the rear wheels. For the front wheels.1: Wheel Hop: Rate and Mass Ratios Both spring combination candidates have been plotted onto the graph. For both spring combinations. the mass ratios are the same.25 M1 The rate ratios differ between the spring combinations.Graph 3.
the ride will be overdamped.14 KT The rate ratio for the rear wheels is KW ≈ 0. To get closer the “equal damping” line. there are a variety of changes that can be made. Although this would be expensive. as the chassis is 40 . but it would also allow the car to travel a greater distance before having to recharge the batteries. An acceptable way of doing this would be to modify the battery cage to house a larger number of lithiumion batteries. and therefore not to have to overdampen ride to prevent wheel hop. I. but even manufacturing the wishbones out of carbon fibre instead of mild steel would help improve the situation. This could prove difficult. As can been seen from the graph.19 KT Combination 2 The rate ratio for the front wheels is KW ≈ 0.The rate ratio for the front wheels is KW ≈ 0. the plotted points (especially the front wheel from Combination 2) are closer to “50% damping on unsprung” than to “equal damping”. This means that in order to damp out wheel hop. it would not only increase the sprung mass. The unsprung mass could be reduced. II.09 KT The rate ratio for the rear wheels is KW ≈ 0. The sprung mass could be increased. Another way would be to put some panels and a body kit on the car.14 KT Anywhere to the right of the “equal damping” line means that wheel hop damping is less than ride damping.
however. please refer to “The problem associated with using springs that are too stiff”. Vertical accelerations produced by the spring combination candidates Both candidates for the spring combination were entered into the Half Car Model. ζ=0. 41 . For further information.currently just a “skeleton” frame. Below are the MATLAB generated graphs of magnitude of vertical acceleration vs. if the springs are too stiff.1. frequency of the two unsprung masses and the sprung mass from the Half Car Model using both combinations of springs. Please refer to Appendix F for the calculation of the damping rates. The “road” parameters were arbitrarily chosen as follows.1m s = 10m The wheel damping rates were arbitrarily chosen to be 15% of the critical damping of the unsprung masses (damping ratio. Please refer Appendix G for the MATLAB code for the Half Car Model vertical accelerations.15). A1 = 0. the ride will be too uncomfortable for the driver. This would be a simple change to make. III. Stiffer springs could be used. as can be seen on the screen shot of the Solid Works model of the car in Figure 2.
Below is a MATLAB generated graph showing the magnitude of vertical acceleration vs. frequency of the front unsprung mass from the Half Car Model using both combinations of springs.2: Vertical accelerations of the three masses at frequencies between 035Hz. 42 .Graph 3.
3: Vertical acceleration of the front unsprung mass at frequencies between 035Hz Below is a MATLAB generated graph showing the magnitude of vertical acceleration vs. 43 .Graph 3. frequency of the rear unsprung mass from the Half Car Model using both combinations of springs.
4: Vertical acceleration of the rear unsprung mass at frequencies between 035Hz As can be seen from the accelerationfrequency graphs of the front and rear unsprung masses. i. the vertical acceleration of the sprung mass (Milliken & Milliken 1995). Vehicle suspension is used to minimise discomfort cause by irregularities in the road. A common method of determining the extent of this discomfort is to measure the vertical acceleration experienced by the driver.Graph 3. Combination 2 produces slightly larger vertical acceleration of the unsprung masses compared to Combination 1. frequency of the sprung mass from the Half Car Model using both combinations of springs. Below is a MATLAB generated graph showing a “close up” of the magnitude of vertical acceleration vs. 44 .e. although this difference is so small that it can be considered negligible.
it can be seen that Combination 2 produces lower vertical accelerations compared to Combination 1. and.5: Vertical acceleration of the sprung mass at frequencies between 025Hz On closer inspection of the accelerationfrequency graph of the sprung mass. Most appropriate springs to be used on the REV car So the most appropriate springs to be used on the REV car are: Rear Springs Combination 2 450 x 2.Graph 3.57 Table 3.39 Front Springs 300 x 2. This means that Combination 2 would provide a more comfortable ride for the driver.8: Most appropriate springs to be used on the REV car 45 . therefore would be the most appropriate combination of springs to be used on the REV car.
25. On the “Wheel Hop: Rate and Mass Ratios” graph. On the “equal damping” line.1m. Please refer Appendix I for the MATLAB code for the Quarter Car Model vertical accelerations. This can be shown by the following extreme example. the plotted points have an m/M1 ratio of either 0. The problem associated with using springs that are too stiff is that the ride can become too uncomfortable for the driver. Please refer to Appendix H for the calculation of the damping rates.25 or close to 0.25 corresponds to a KW/KT ratio of 1. Taking into account the motion ratio. ζ=0. frequency of the unsprung and the sprung masses of the front wheel from the Quarter Car Model using both the “stiff spring” and the front spring from Combination 2. This means that in order to be on the “equal damping” line.15). Arbitrarily. an m/M1 ratio of 0. this corresponds to a spring rate of just over 3200lbs/in. the wheel rate must be the same as the tyre rate. which is 180 190 N/m. 46 . and the wheel damping rates were chosen to be 15% of the critical damping of the unsprung masses (damping ratio. the “road” amplitude A1 was chosen to be 0. Below are the MATLAB generated graphs of magnitude of vertical acceleration vs. Both the new “stiff spring” (where the wheel rate is equal to the tyre rate) and the springs from Combination 2 were entered into the Quarter Car Model.The problem associated with using springs that are too stiff Referring to the “Wheel Hop: Rate and Mass Ratios” graph. one of the suggestions to get closer to the “equal damping” line was to use stiffer springs.
Graph 3. 47 . frequency of the unsprung and the sprung masses of the rear wheel from the Quarter Car Model using both the “stiff spring” and the rear spring from Combination 2.6: Vertical accelerations of the sprung and unsprung masses of the front wheel at frequencies between 035Hz Below are the MATLAB generated graphs of magnitude of vertical acceleration vs.
A1 Below are the MATLAB generated graphs of magnitude of vertical acceleration vs. the “stiff spring” produces very large “spikes” in the vertical accelerations of the sprung masses of the front and rear wheels. 48 . This is a clear indication that using springs this stiff would cause the driver to be very uncomfortable. frequency of the two unsprung masses and the sprung mass from the Half Car Model using the Combination 2 springs. Changing amplitude.Graph 3.7: Vertical accelerations of the sprung and unsprung masses of the rear wheel at frequencies between 035Hz As can be observed from the two previous graphs. These graphs show the effect of changing the “road” amplitude A1 .
Graph 3.17m.8: Changing the amplitude The MATLAB graphs shows curves for A1=0. frequency of the two unsprung masses and the sprung mass from the Half Car Model using the Combination 2 springs. ζ Below are the MATLAB generated graphs of magnitude of vertical acceleration vs.03m.1m and A1=0. It is clear from observing the graphs that increasing the amplitude of the “road” causes the vertical accelerations of each of the masses to increase. These graphs show the effect of changing the damping ratio ζ. Changing damping ratio. A1=0. 49 .
and producing sprung mass acceleration levels that are not too high for the remainder of the frequency range. 50 . in the case of the sprung mass. increasing the damping ratio reduces the vertical accelerations at all frequencies.9: Changing the damping ratio The MATLAB graphs shows curves for ζ=0. However. In the case of the front and rear unsprung masses. ζ =0. Using a damping ratio of 0.05. but causes the acceleration level over the remainder of the frequency range to be increased.15 provides a reasonable compromise of reducing the unsprung mass acceleration peak. increasing the damping ratio reduces the acceleration peaks at around 3Hz and 17Hz. while reducing both sprung mass acceleration peaks.15 and ζ =0.25.Graph 3.
lecturers. Emergency phone numbers In the case of any emergency. It is responsibility of everybody associated with the REV project to assist in creating and maintaining a culture of safety. smoke detectors should sound an alarm. Safety It was vitally important that high levels of safety were adhered to during this project. tutors and lab technicians. to ensure the wellbeing of the author and other members of the University.4. and give the following information: • • • • type of emergency location name mobile phone number or the telephone extension of the room Below is a list of some other important phone numbers Hospital (QE 2 ‘G Block’ Emergency) Doctor (UWA Medical Centre) Poisons Information Centre (All hours) (0) 9346 3380 Extension 2118 0 13 11 26 Fire procedure In the event of a fire. If the fire is too large to extinguish. they should select an appropriate fire extinguisher and extinguish the fire from a safe distance. encouraging all participants to remain vigilant against possible health and safety hazards. If the fire is small and the person or persons are comfortable with fighting the fire. or the person or persons are not comfortable with fighting the fire. and assemble at the designated evacuation area. including other students. they should immediately evacuate the area. Most fire 51 . call 2222 (from any internal university phone) or 6488 2222 (from a mobile phone).
All equipment should be regularly inspected. one or more of the wheels starts to oscillate violently due to wheel hop or some other form of unsprung mass oscillation. The driver must also be strapped in the seat with a fivepoint safety harness. Driving the REV car When driving the REV car. and not used until it has been fixed and approved by a qualified electrician. and any equipment believed to be unsafe or faulty should be clearly tagged. smoking is not permitted in the vicinity of gas storage bottles. it is important not to 52 . and water extinguishers are not to be used on flammable liquid fires. Electrical safety It is important that all electrical equipment is properly used and well maintained. When working alone outside of the normal operating hours of 9am to 5pm. must be worn at all times. Beware that foam and water extinguishers are not to be used on electrical fires. Working in labs When working in the labs. and they are suitable for most fire. Smoking To protect people from tobacco smoke. personal protective equipment (PPE). Also. such as fully enclosed footwear and approved safety glasses. people are not permitted to use tools. smoking is strictly forbidden within buildings or within 5 metres from the entrance to a building. If when the car is in motion.extinguishers are the carbon dioxide or dry chemical type. fully enclosed footwear and a helmet must be worn.
meaning that the dampers should be either retuned or replaced (if damaged) so that they are able to damp out any violent oscillations of the wheels. it is recommended that the driver simply takes their foot off the accelerator and allows the car to slow down until the violent oscillation of the wheel(s) stops. This will cause the frequency of vibration to decrease. Instead.immediately brake. which can lead to loss of control of the car. 53 . while the amplitude will increase significantly. The car suspension should then be immediately serviced.
The second method is to increase the sprung mass. This could be achieved by replacing the current mild steel wishbones with ones that are manufactured out of carbon fibre. As in Combination 1. There is plenty of future research that can be done on the REV car. Another way to reduce unsprung mass would be to redesign the uprights so that they are lighter. should be done with caution. The first method is to reduce the unsprung mass.1. This project has only simulated situations where the car is pitching. As can be seen by the screen shot Solid Works model in Figure 2. however. although doing this would possibly be a final year project by itself. By adding panels and a body it. Conclusion and Future Work After considering both spring selection criteria and observing the accelerationfrequency graphs generated by MATLAB. and the car would be more aesthetically pleasing and may even perform better aerodynamically. and to avoid overdamping the ride to control wheel hop. as using springs that are too stiff can have a detrimental effect on ride. however. meaning that they would provide a much more comfortable ride for the driver. and is in “skeleton” form. This. Another interesting study would be to design and implement a springdamper arrangement that could be attached to the base of the seat. these springs satisfied both criteria and produced very similar vertical accelerations for both the front and rear unsprung masses.5.39 springs at the rear (Combination 2). The third and last method was to use stiffer springs. After observing the “Wheel Hop: Rate and Mass Ratios” graph. It would be desirable if future REV students could model transient situations where the car is turning. they produced much smaller vertical accelerations of the sprung mass compared to Combination 1.57 springs at the front and the 450 x 2. it became apparent that there are three methods in order get closer to the “equal damping” line. yawing and pitching of the car. the car chassis consists only of a frame. which could damp out any violent oscillations 54 . the sprung mass would be increased. the most suitable spring combination to use on the REV car is the 300 x 2. and the associated rolling.
of the sprung mass. and it seems as though using a kinetic suspension system will be the way of the future 55 . This would allow stiffer springs to be used. If the REV car is going to compete in future competitions. it would be advantageous for a kinetic suspension system to be investigated. The UWA Motorsport team has been using this technology since 2004. and would ensure that the ride was comfortable for the driver. in order to control wheel hop.
Sparkford. NJ 56 .6. PA Smith. Race Car Vehicle Dynamics. References Aird. W & Milliken. Design. Fallbrook. F 2004. Massachusetts Desert Rides 2007. Perth Milliken. Inc.. Racing and Sports Car Chassis Design..desertrides. Kinematics. A 1999. Society of Automotive Engineers. Dynamics. Chassis Design: Principles and Analysis. Robert Bentley. Modification.. F 1997. [21 October 2009] Finlayson. Hoboken. Perth Guzzomi. Competition Car Suspension 3rd Edition. Desert Term Glossary. D 2002. K & Kinzel. C 1978. Tune to Win. Race Car Chassis: Design and Construction. Inc. Osceola. D 2003. Warrendale. UK Waldron. D 1995. Warrendale. Haynes Publishing. PA Milliken. University of Western Australia. Inc. John Wiley & Sons. Available from: <http://www. Optimisation & Development of a FSAE Suspension System. Inc. Cambridge. Society of Automotive Engineers. Inc. Aero Publishers.. D 1962. W & Milliken. Testing and Analysis of FormulaSAE Dampers. CA Staniforth. and Design of Machinery.com/reference/images/terms/toe.gif>. G 2004. University of Western Australia. MBI Publishing Company. Desert Rides. M & Phipps. WI Costin.
N 2002. Design of a Formula SAE Racing Car Suspension System. Perth 57 . University of Western Australia.Winzer.
as shown below. Appendices Appendix A: Negative swing arm effect The negative swing arm effect causes camber to become increasingly positive as the wheel rises.7. Figure A: Negative swing arm effect (Milliken & Milliken 2002) 58 .
This information can be represented in the diagram below.959m ∴ b = 0.Appendix B: Determining a and b A set of scales was placed underneath each wheel of the car.891m 59 . The average scale readings of the front and rear wheels was 66kg and 71kg respectively. which was “loaded” with a 75 kg driver.85m ∑M CG =0 ∴ 71b = 66a 71(l − a ) = 66a ∴ a = 0. Figure B: Mass representation diagram l = a + b = 1.
Appendix C: Quarter Car Model Vertical Acceleration Calculations [MassMatrix] = m 0 0 M − c − c c c − KW KW [DampingMatrix] = [SpringMatrix] = X X = 1 X 3 X1 X0 = 0 X3 0 K T + KW − KW X = X 0e jωt A A= 1 0 & & [MassMatrix]X& + [DamperMatrix]X + [SpringMatrix]X = Ae jωt − ω 2 [MassMatrix]X 0 + jω [DampingMatrix]X 0 + [SpringMatrix]X 0 = A 60 .
(− ω [MassMatrix] + jω[DampingMatrix] + [SpringMatrix])X 2 0 =A B = −ω 2 [MassMatrix] + jω [DampingMatrix] + [SpringMatrix ] X 0 = B −1 A X = X 0 e jωt & X = jω X 0 e jωt && X = −ω 2 X 0 e jωt 61 .
depending on the velocity of the vehicle. excitation at rear wheel: 62 . Therefore. Frequency: ω= 2π 2π 2πv = = s T s v Excitation at front wheel: A1 sin ωt Converting to complex form: A1e jωt Rear wheel will be excited some time after the front wheel. pictured below: Figure D: Assumed road Period of oscillation.Appendix D: Road Excitation Calculations Assuming that road is a sine wave with amplitude A1 and a length of one cycle of road profile of s. T: T= s v s= length of one cycle of road profile v= velocity Where.
A1 sin ω (t + ∆t ) l = A1 sin ω t + v 2πl = A1 sin ω t + ωs 2 πl = A1 sinωt + s Converting to complex form: l = wheelbase j 2πl jωt A1 j exp − e s 63 .
Appendix E: Half Car Model Vertical Acceleration Calculations m f 0 MassMatrix = 0 0 0 mr 0 0 0 0 M 0 0 cr − cr bcr 0 0 0 I − cf − cr c f + cr ac f − bcr 0 a 2 c f + b 2 cr 0 0 − KW f − K Wr K W f + K Wr aKW f − bKWr a 2 KW f 0 + b 2 K Wr 0 0 cf 0 DamperMatrix = − cf − ac f K T f + KW f 0 SpringMatrix = − KW f − aKW f X1 X X = 2 X 3 θ X 10 X 20 X0 = X 30 θ 0 0 K Tr + KWr − KWr bKWr 64 .
X = X 0 e jω t A1 A1 j exp − j 2πl A= s 0 0 & & [MassMatrix]X& + [DamperMatrix ]X + [SpringMatrix]X = Ae jωt − ω 2 [MassMatrix ]X 0 + jω [DampingMatrix ]X 0 + [SpringMatr ix ]X 0 = A (− ω [MassMatrix] + jω[DampingMatrix] + [SpringMatrix])X 2 0 =A B = −ω 2 [MassMatrix] + jω [DampingMatrix ] + [SpringMatrix ] X 0 = B −1 A X = X 0 e jωt & X = jω X 0 e jωt && X = −ω 2 X 0 e jωt 65 .
0 Ns/m Combination 2 ccrit = 3766.0 Ns/m Damping of rear unsprung mass: Combination 1 ccrit = 3924.9 = 565.7 = 577.9 Ns/m c = 0.9 = 588.7 Ns/m c = 0.15 x 3846.15 x 3924.Appendix F: Damping calculations for the Half Car Model c = ζ × ccrit ccrit = 2 m(KT + KW ) ζ = 0.7 Ns/m Combination 2 ccrit = 3846.7 = 577.15 x 3846.9 Ns/m c = 0.15 x 3766.0 Ns/m 66 .7 Ns/m c = 0.15 Damping of front unsprung mass: Combination 1 ccrit = 3846.
% front unsprung mass [kg] % rear unsprung mass [kg] % sprung mass of half of the car[kg] % mass moment of inertia of sprung mass [kg % distance from front wheel to CG [m] % distance from rear wheel to CG [m] % wheelbase [m] % front tyre spring rate [N/m] % rear tyre spring rate [N/m} % Amplitude [m] %length of one cycle of road profile [m] 67 . s=. % front wheel spring rate [N/m] % rear wheel spring rate [N/m] % front wheel damping rate [Ns/m] % rear wheel damping rate [Ns/m] %Constant variables mf=. kwr=. cf=. M=. ktr=. ktf=. m^2] a=. I=. b=. cr=.Acceleration %Peter Corke 20139918 clear.Appendix G: MATLAB Code for Half Car Model Vertical Accelerations Below is the MATLAB code for determining the vertical accelerations of the front unsprung. %Changeable variables kwf=. rear unsprung and sprung masses in the Half Car Model. %Half Car Model . l=a+b. A1=. mr=.
hzmin=.0.0.kwr. k=1. wamount=hzamount*2*pi.0.kwf.:. magnitudeX1=abs(X1).a^2*cf+b^2*cr]. k=k+1.y. end.0.I].1. X3 (u)=Acc(3. hzamount=(hzmaxhzmin)/(n1). for u=1:1:z.0.k)=w^2*X0(:.cf. Acc(:. % Frequency range [Hz] hzmax=.z]=size (Acc). freq_hz=freq/(2*pi).cr.0.0.0. Damper_Matrix=[cf.a*kwf.0.u).0. X1 (u)=Acc(1.0.M.a*cf.0. B=w^2*Mass_Matrix+j*w*Damper_Matrix+Spring_Matrix.:.0. X0(:. freq(k)=w. Mass_Matrix=[mf.b*kwr. n=1000. wmin=hzmin*2*pi.u). X2 (u)=Acc(2.k)=B\A.0.a^2*kwf+b^2*kwr].0.ktr+kwr.A1*j*exp((j*2*pi*l)/s). [x. 68 .cf.cf+cr.kwf.mr.u).1.:.a*cfb*cr.0.0. for w=wmin:wamount:wmax. A=[A1.cr.0. Spring_Matrix=[ktf+kwf.kwf+kwr.0.0.b*cr.0].0. wmax=hzmax*2*pi.1. end.0.a*kwfb*kwr.cr.0.kwr.k).
plot (freq_hz.2).1. plot (freq_hz.1. 69 .3).'b'). figure (1).magnitudeX2. subplot (3.magnitudeX2=abs(X2). title ('Front Unsprung Mass'). ylabel ('Magnitude of Accerleration (m/s^2)'). ylabel ('Magnitude of Acceleration (m/s^2)').magnitudeX1.1). ylabel ('Magnitude of Acceleration (m/s^2)'). title ('Rear Unsprung Mass'). subplot (3. xlabel ('Frequency (Hz)'). xlabel ('Frequency (Hz)'). plot (freq_hz. subplot (3. magnitudeX3=abs(X3).'b').'b'). title ('Sprung Mass').1.magnitudeX3. xlabel ('Frequency (Hz)').
15 x 5093.15 x 5093.1 Ns/m Combination 2 ccrit = 3846.9 Ns/m c = 0.0 Ns/m Damping of rear unsprung mass: Stiff spring ccrit = 5093.15 x 3766.0 Ns/m 70 .1 Ns/m Combination 2 ccrit = 3766.15 x 3846.7 = 577.Appendix H: Damping calculations for the Quarter Car Model c = ζ × ccrit ccrit = 2 m(KT + KW ) ζ = 0.9 = 764.9 Ns/m c = 0.9 = 764.7 Ns/m c = 0.9 = 565.9 Ns/m c = 0.15 Damping of front unsprung mass: Stiff spring ccrit = 5093.
wamount=hzamount*2*pi.k).u). % Tyre spring rate [N/m] kw=.kw. Acc(:. X3 (u)=Acc (2. for w=wmin:wamount:wmax. %Quarter Car Model .k)=B\A. % wheel damping rate [Ns/m] kt=.:. % Frequency range [Hz] hzmax=. wmax=hzmax*2*pi.Acceleration %Peter Corke 20139918 clear.0.0]. % unsprung mass [kg] M=.M]. end. freq (k)=w. % Wheel spring rate [N/m] A1=. k=1.:. Mass_Matrix=[m. k=k+1. A=[A1.Appendix I: MATLAB Code for Quarter Car Model Vertical Accelerations Below is the MATLAB code for determining the vertical accelerations of the unsprung and sprung masses in the Quarter Car Model.. m=.u).1. n=1000. X0(:.kw].c. hzamount=(hzmaxhzmin)/(n1).c].0.:.kw. X1 (u)=Acc (1. wmin=hzmin*2*pi. B=w^2*Mass_Matrix+j*w*Damper_Matrix+Spring_Matrix.y. % Amplitude [m] hzmin=. % sprung mass at one wheel[kg] c=.z]=size (Acc). Damper_Matrix=[c. [x. Spring_Matrix=[kt+kw. 71 .c. for u=1:1:z.1. end.k)=w^2*X0 (:.
ylabel ('Magnitude of Acceleration (mm/s^2)').1.1).freq_hz=freq/(2*pi). subplot (2. magnitudeX3=1000*abs (X3). xlabel ('Frequency (Hz)'). title ('Sprung Mass').'b').'b'). ylabel ('Magnitude of Acceleration (mm/s^2)').magnitudeX1. subplot (2.2). figure (1). plot (freq_hz.magnitudeX3.1. title ('Unsprung Mass'). plot (freq_hz. magnitudeX1=1000*abs (X1). xlabel ('Frequency (Hz)'). 72 .
This action might not be possible to undo. Are you sure you want to continue?