You are on page 1of 15

HANDLING ANALYSIS

Vehicle Ride and Handling

Everest Chiboli

Contents
Introduction ............................................................................................................................................. 2
Background ............................................................................................................................................. 2
Aims and Objectives ............................................................................................................................... 3
Methodology ........................................................................................................................................... 3
Equations ................................................................................................................................................ 4
Results ..................................................................................................................................................... 4
Model 1: Simulink .............................................................................................................................. 4
Model 1: Vehicle 1 Results ................................................................................................................. 4
Model 1: Vehicle 2 Results ................................................................................................................. 4
Model 2: Simulink .............................................................................................................................. 5
Model 2: Vehicle 1 Results ................................................................................................................. 5
Model 2: Vehicle 2 Results ................................................................................................................. 5
Model 3 Simulink ............................................................................................................................... 6
Model 3: Vehicle 1 Results ................................................................................................................. 6
Model 3: Vehicle 2 Results ................................................................................................................. 6
Validation................................................................................................................................................ 7
Conclusion .............................................................................................................................................. 7
Further Work........................................................................................................................................... 8

Introduction
The study of vehicle dynamics is important to automotive engineers, manufacturers and the
public. As powertrains become increasingly complex, the stability and handling
characteristics of the chassis must meet these increasing performance demands.
From an engineering aspect, the capabilities of a vehicle is limited simply by the tyres and
being able to linearise the handling capabilities of a vehicle ensures greater safety for the
occupants, as well as better ride and handling characteristics.
To understand the dynamics of a tyre and all the forces subjected upon it, software can be
used to simulate tyre behaviour to predict vehicle handling from driver input. The softwares
used in this report are Matlab and Simulink.
Matlab stands for Matrix Laboratory and put simply is a complex mathematical language
based software used to compute equations.
Simulink on the other hand is a software that resides within Matlab that provides an
interactive and graphical interface for modelling, simulating and analysing of dynamic
systems as block diagrams.
This report looks at vehicle dynamics by creating tyre models using model-simulating
software to analyse tyre behaviour, to further determine whether if over steer or under steer
characteristics are demonstrated and to analyse overall handling performance.

Background
As tires are the only point of contact with the road a vehicle has, tyres are regarded as the
most important safety feature of any vehicle.
History teaches us In 1895, Andr Michelin was the first person to use pneumatic tires on an
automobilei although deemed unsuccessful, the foundations for a tyre were born and was
applied to an automobile.
The purpose of a tyre is to support the vehicles weight, allow the vehicle to steer in the
desired direction, absorb road bumps and reduce road noise, although amongst this numerous
forces are subjected onto a tyre when in motion and especially when cornering. It is important
to understand the forces a tyre is subjected to.
Some of the main forces are;
Lateral force: Also known as side or cornering force, the lateral tyre force is the force
required to maintain a vehicle through a corner.
Longitudinal force: Similar to the lateral force but in a longitudinal direction.
It is considered that the limits of handling is usually defined by the maximum available
lateral force. Tyres usually have an optimum operating range and can generate up to 0.4G.
2

This is the linear range of the tyre and allows the handling to become and remain more
predicable to the driver. Anything after this point renders the tyre to behave in a nonlinear
fashion where conditions of under or over steer reside.

Aims and Objectives


The aim of this report is to build and analyse the handling model of a vehicle using software
called Matlab and Simulink.
The objectives are to build three Simulink models:

Linear 2 degree of Freedom bicycle model


2 Degree of freedom bicycle model with a non-linear tyres
2 Degree of freedom bicycle model with a non-linear tyres and load transfer

On completion of the above models, tests will be carried out to show the response of each
model. The tests include:

Model 1 and both vehicle parameter sets to conduct repeated step steer tests for 10
different speeds using the same steer angle.

Models 2 and 3 to conduct sweep steering tests at 20m/s and 50m/s using both sets of
vehicle parameters with a reasonable steer angle, frequency range and simulation
time.

Methodology
It is important to note two sets of pre-determined parameters are used to compute the three
models. The parameters are as follows:
Parameter
Mass
Yaw inertia
CG to front axle
CG to rear axle
Front axle
cornering stiffness
Rear axle cornering
stiffness
CG height from
ground
Track width

Symbol
m
I
a
b
Cf

Units
kg
kgm2
m
m
kN/rad

Vehicle 1
2000
3650
1.35
1.85
85

Vehicle 2
1150
1200
1.05
1.15
120

Cr

kN/rad

75

140

0.5

1.8

2.0

Equations
In order to create the Simulink tyre models the following equations were used and calculated:
=(1)
= Fy(max) =
= /()
=(1)(1)
= /()

(1)
(2)
(3)
(4)
(5)

Results
Model 1: Simulink

Lateral Velocity Vs Time


1

16
0

14
-1

Lateral Velocity (m/s)

Yaw (deg/sec)

12
10
8
6

-2

-3

-4

4
-5

2
0

10
12
Time (sec)

14

16

18

-6
0

20

10
12
Time (sec)

14

16

18

20

Mo
del
1:
Ve
hic
le
1

Lateral Acceleration Vs Time


12

10

Lateral Acceleration (m/s^2)

Yaw Vs Time
18

10
12
Time (sec)

14

16

18

20

Results

Model 1: Vehicle 2 Results


Lateral Acceleration Vs Time

Yaw Vs Time

Lateral Velocity Vs Time

30

20

0.5

18
0

25

16

Lateral Velocity (m/s)

Yaw (Deg/Sec)

15

10

Lateral Acceleration (m/s^2)

-0.5

20

-1
-1.5
-2
-2.5

14
12
10
8
6
4

5
-3

2
0

-3.5

10
12
Time (sec)

14

16

18

20

10
12
Time (sec)

14

16

18

20

10
12
Time (sec)

14

16

18

20

Simulink Model 1 shows a linear two degree of freedom bicycle model. The graphs for this
model display results for Yaw, Lateral Velocity and Lateral Acceleration against time ran at
speeds ranging from 5 50 m/s in increments of five. Two sets of graphs are produced as
there are two vehicles taking part in each test. On reflection, it is clear to see that these graphs
reach a steady state in less than two seconds generating a linear response.
To further compare both vehicles graphs vehicle two reaches steady state slightly faster than
vehicle one. The magnitude for acceleration and yaw rate is greater for vehicle two. The
velocity reached in model two is of lower magnitude than model one, the combination of this
and the increased magnitude for acceleration means steady state is achieved faster.

Model 2: Simulink

Model 2: Vehicle 1 Results


Yaw Vs Time

Lateral Acceleration Vs Time

Lateral Velocity Vs Time

15

5
4

10

Lateral Velocity (m/s)

Yaw (Deg/Sec)

Lateral Acceleration (m/s^2)

-5

1
0
-1
-2
-3

-10

10
12
Time (sec)

14

16

18

20

-5

0
-2
-4

10
12
Time (sec)

14

16

18

-8

20

10
12
Time (sec)

14

16

18

20

16

18

20

Lateral Velocity Vs Time

Yaw Vs Time
20

15

10

Lateral Velocity (m/s)

Yaw (Deg/Sec)

-6

-4

-15

5
0
-5

-1

-10
-2

-15
-20

10
12
Time (sec)

14

16

18

20

-3
0

10
12
Time (sec)

14

16

18

20

Model 2: Vehicle 2 Results


Lateral Acceleration Vs Time
15

Lateral Acceleration (m/s^2)

10

Model two constructs on from model one with extension of the


tyre dynamics. The outcome for this model displays a non-linear
two degree of freedom bicycle tyre model. The equations used

-5

-10

-15

10
12
Time (sec)

14

to complete this model are equations (1) and (2) listed above. The test speeds carried out
were at 20 m/s and 50 m/s.
The graphs displayed for both vehicles show Yaw, lateral velocity and lateral acceleration
against time. It is important to note to produce the above graphs a chirps signal was used
using a value of 1Hz. A chirps signal is an angle module signal used to simulate a driver
turning left and right which is what each wave on the above graphs represents.
By increasing the Hz for the chirps signal, the outcome would show the graphs with a higher
frequency. It is advised to choose a value less than 3Hz as in real world conditions it would
be impossible to achieve anything higher then 3Hz.

Model 3 Simulink

Model 3: Vehicle 1 Results


Yaw Vs Time

Lateral Velocity (m/s)

Yaw (Deg/Sec)

-5

15

10

-1

-10

-2

-15

-3

10
12
Time (sec)

14

16

18

Lateral Acceleration (m/s^2)

10

Lateral Acceleration Vs Time

Lateral Velocity Vs Time

15

20

-5

-10

Yaw Vs Time

10
12
Time (sec)

14

16

18

20

-15

Lateral Velocity Vs Time

25

2.5

20

15

1.5

10

10
12
Time (sec)

14

16

18

20

Lateral Acceleration Vs Time


20

5
0
-5
-10

0.5
0
-0.5
-1

-15

-1.5

-20
-25

Lateral Acceleration (m/s^2)

Lateral Velocity (m/s)

Yaw (Deg/Sec)

15

10
12
Time (sec)

14

16

18

20

-2.5

5
0
-5
-10
-15

-2
0

10

10
12
Time (sec)

14

16

18

20

-20

10
12
Time (sec)

14

16

18

20

Model 3: Vehicle 2 Results


Model three introduces load transfer to the two degree of freedom nonlinear bicycle model.
Load transfer is an important aspect to consider when analysing handling as this takes into
consideration the vehicle mass and provides a more realistic results.
6

As the vehicle turns from one direction to the other, the response generated for vehicle one
shows that when traveling at 20 m/s over a set period of time, yaw oscillates at the same
frequency, unlike when traveling at 50 m/s yaw increases with time. Vehicle two on the other
hand increases its yaw rate marginally as compared with vehicle one. It is also important to
note that vehicle one has higher lateral velocity with lower later acceleration as compared
with vehicle two that has a higher lateral acceleration but lower lateral velocity. A reason for
this could be due to vehicle 2 being lighter.
Comparing models two and three, they appear very similar although a few differences are
present. Due to considering nonlinear load transfer, a more accurate model is portrayed in
model three.
All equations listed in the equation section of this report were used to generate model three.

Validation
For the validation process, hand calculations were performed for validating model one yaw
rate Vs time. The following calculations were performed and were correlated with model one
vehicle one steady state graph of Yaw vs time. Please note U = 50 m/s. The equations used
to perform this task are:

( )

1

= 0.002

= = 295.495

= (+)+2 = 0.1101036269

= = 0.1926813472 rad/sec

= 3.853626
=

= 9.63

To validate the values from the equations above, Rss was converted from radiant to degrees
to achieve a value of 11.03 degrees. This matches up with the model one vehicle one graph
thus validating model one.

Conclusion
In order to help visualise the handling characteristics, Simulink is a very good method as it
can assist engineers study the responses generated from driver input. Many factors can be
incorporated as have been with this report and the beauty of this software is that it can be
easily expanded to include more factors.
As the models produced in this report progressed form 1 3, it became more realistic as the
models became more complex introducing more factors. Model one for example looks at the
front and rear tyres of a simple non-linear bicycle model, whereas model two adds in linearity
to the same model for the front and rear tyres. Model three on the other hand as can be seen
in this report introduces all four tyres of a vehicle which although makes the model more
complicated, it is the more real in terms of real world application.

The main outcomes of this report highlight two main responses. The first being steady state
and the other transient. Model one clearly shows steady state whilst models two and three
show transient.
From analysing all the data provided in this report, performance wise vehicle two has the best
characteristics. The reason for this includes the graphs displaying less overshoot and a quick
return to a steadier state.
For durability, and more so for the purpose of this assignment, vehicle one would be the best
suited as it is more heavier the predictability due to the centre of gravity located by
determining a + b shows that the vehicle responds better and provides a more adequate ride.

Further Work
To further develop this model, improvements can be made. Many factors still have not been
taken into consideration that could be included to provide a more accurate result. These
factors include tyre size, tyre pressures, road conditions and suspension specifications. All the
factors mentioned have a direct effect on a tyres behaviour therefore this leads to the
dynamics of how the car will behave.

References
i

About.com Inventors . 2002. History of Tires. [ONLINE] Available at:


http://inventors.about.com/library/inventors/bltires.htm. [Accessed 01 March 15].

Appendices
Mat lab code
Model 1 vehicle 1
m = 2000;
I = 3650;
a = 1.35;
b = 1.85;
Cf = 85000;
Cr = 75000;
Cz = 7.024;
U = [5 10 15 20 25 30 35 40 45 50];
df = 0.035;
h = 1;
t = 1.8;
T = 1.8;

Mu = 1;
g = 9.81;
Nf = (m*g*a)/a+b;
Nr = (m*g*b)/a+b;
Fzf = Nf
Fzr = Nr
A1 = (1*Nf);
A2 = (1*Nr);
B1 = Cf/A1;
B2 = Cr/A2;
D1 = Cz/A1;
D2 = Cz/A2;
F=a/(a+b)
R=b/(a+b)
open('model1')
sim('model1')
figure()
plot(simout1)
title 'Yaw Vs Time'
xlabel 'Time (sec)'
ylabel 'Yaw'
open('model1')
sim('model1')
figure()
plot(simout2)
title 'Lateral Velocity Vs Time'
xlabel 'Time (sec)'
ylabel 'Lateral Velocity'
open('model1')
sim('model1')
figure()
plot(simout3)
title 'Lateral Acceleration Vs Time'
xlabel 'Time (sec)'
ylabel 'Lateral Acceleration'

Model 1 vehicle 2
m = 1150;
I = 1200;
a = 1.05;
b = 1.15;
Cf = 120000;
Cr = 140000;
Cz = 11.024;
U = [5 10 15 20 25 30 35 40 45 50];
df = 0.035;
h = 0.5;
t = 2.0;
T = 2.0;
Mu = 1;

g = 9.81;
Nf = (m*g*a)/a+b;
Nr = (m*g*b)/a+b;
Fzf = Nf
Fzr = Nr
A1 = (1*Nf);
A2 = (1*Nr);
B1 = Cf/A1;
B2 = Cr/A2;
D1 = Cz/A1;
D2 = Cz/A2;
F=a/(a+b)
R=b/(a+b)
open('model1')
sim('model1')
figure()
plot(simout1)
title 'Yaw Vs Time'
xlabel 'Time (sec)'
ylabel 'Yaw'
open('model1')
sim('model1')
figure()
plot(simout2)
title 'Lateral Velocity Vs Time'
xlabel 'Time (sec)'
ylabel 'Lateral Velocity'
open('model1')
sim('model1')
figure()
plot(simout3)
title 'Lateral Acceleration Vs Time'
xlabel 'Time (sec)'
ylabel 'Lateral Acceleration'

Model 2 vehicle 1
m = 2000;
I = 3650;
a = 1.35;
b = 1.85;
Cf = 85000;
Cr = 75000;
Cz = 7.024;
U = [20 50];
df = 0.035;
h = 1;
t = 1.8;

10

T = 1.8;
Mu = 1;
g = 9.81;
Nf = (m*g*a)/a+b;
Nr = (m*g*b)/a+b;
Fzf = Nf
Fzr = Nr
A1 = (1*Nf);
A2 = (1*Nr);
B1 = Cf/A1;
B2 = Cr/A2;
D1 = Cz/A1;
D2 = Cz/A2;
F=a/(a+b)
R=b/(a+b)
open('model2final')
sim('model2final')
figure()
plot(simout1)
title 'Yaw Vs Time'
xlabel 'Time (sec)'
ylabel 'Yaw'
open('model2final')
sim('model2final')
figure()
plot(simout2)
title 'Lateral Velocity Vs Time'
xlabel 'Time (sec)'
ylabel 'Lateral Velocity'
open('model2final')
sim('model2final')
figure()
plot(simout3)
title 'Lateral Acceleration Vs Time'
xlabel 'Time (sec)'
ylabel 'Lateral Acceleration'

Model 2 Vehicle 2
m = 1150;
I = 1200;
a = 1.05;
b = 1.15;
Cf = 120000;
Cr = 140000;
Cz = 11.024;
U = [20 50];
df = 0.035;

11

h = 0.5;
t = 2.0;
T = 2.0;
Mu = 1;
g = 9.81;
Nf = (m*g*a)/a+b;
Nr = (m*g*b)/a+b;
Fzf = Nf
Fzr = Nr
A1 = (1*Nf);
A2 = (1*Nr);
B1 = Cf/A1;
B2 = Cr/A2;
D1 = Cz/A1;
D2 = Cz/A2;
F=a/(a+b)
R=b/(a+b)
open('model2final')
sim('model2final')
figure()
plot(simout1)
title 'Yaw Vs Time'
xlabel 'Time (sec)'
ylabel 'Yaw'
open('model2final')
sim('model2final')
figure()
plot(simout2)
title 'Lateral Velocity Vs Time'
xlabel 'Time (sec)'
ylabel 'Lateral Velocity'
open('model2final')
sim('model2final')
figure()
plot(simout3)
title 'Lateral Acceleration Vs Time'
xlabel 'Time (sec)'
ylabel 'Lateral Acceleration'

Model 3 Vehicle 1
m = 2000;
I = 3650;
a = 1.35;
b = 1.85;
Cf = 85000;
Cr = 75000;
Cz = 7.024;

12

U = [20 50];
df = 0.035;
h = 1;
t = 1.8;
T = 1.8;
Mu = 1;
g = 9.81;
Nf = (m*g*a)/a+b;
Nr = (m*g*b)/a+b;
Fzf = Nf
Fzr = Nr
A1 = (1*Nf);
A2 = (1*Nr);
B1 = Cf/A1;
B2 = Cr/A2;
D1 = Cz/A1;
D2 = Cz/A2;
F=a/(a+b)
R=b/(a+b)
open('model3')
sim('model3')
figure()
plot(simout1)
title 'Yaw Vs Time'
xlabel 'Time (sec)'
ylabel 'Yaw (Deg/Sec)'
open('model3')
sim('model3')
figure()
plot(simout2)
title 'Lateral Velocity Vs Time'
xlabel 'Time (sec)'
ylabel 'Lateral Velocity (m/s)'
open('model3')
sim('model3')
figure()
plot(simout3)
title 'Lateral Acceleration Vs Time'
xlabel 'Time (sec)'
ylabel 'Lateral Acceleration (m/s^2)'

Model 3 Vehicle 2
m = 1150;
I = 1200;
a = 1.05;
b = 1.15;
Cf = 120000;
Cr = 140000;

13

Cz = 11.024;
U = [20 50];
df = 0.035;
h = 0.5;
t = 2.0;
T = 2.0;
Mu = 1;
g = 9.81;
Nf = (m*g*a)/a+b;
Nr = (m*g*b)/a+b;
Fzf = Nf
Fzr = Nr
A1 = (1*Nf);
A2 = (1*Nr);
B1 = Cf/A1;
B2 = Cr/A2;
D1 = Cz/A1;
D2 = Cz/A2;
F=a/(a+b)
R=b/(a+b)
open('model3')
sim('model3')
figure()
plot(simout1)
title 'Yaw Vs Time'
xlabel 'Time (sec)'
ylabel 'Yaw (Deg/Sec)'
open('model3')
sim('model3')
figure()
plot(simout2)
title 'Lateral Velocity Vs Time'
xlabel 'Time (sec)'
ylabel 'Lateral Velocity (m/s)'
open('model3')
sim('model3')
figure()
plot(simout3)
title 'Lateral Acceleration Vs Time'
xlabel 'Time (sec)'
ylabel 'Lateral Acceleration (m/s^2)'

14

You might also like