You are on page 1of 13

Analysis of Automobile Suspension

AE 3515
Computer Project II


Dale Arney
and
Chirag Talati

13 July 2004
AE 3515 Dale Arney and Chirag Talati
Computer Assignment II 13 July 2004
Page 1 of 13
Table of Contents

1. Introduction and Purpose 2
2. Part A: Half Vehicle Suspension Model 3
3. Part B: Half Vehicle Suspension Model with Damping Effects 5
4. Part C: Natural Responses to Given Initial Conditions 7
5. Part D: Responses to Inputs and Suitable Absorber Coefficients 11

AE 3515 Dale Arney and Chirag Talati
Computer Assignment II 13 July 2004
Page 2 of 13
Introduction and Purpose

For computer assignment 2 the purpose is to approximate a half vehicle
suspension model as a body attached to two springs (see diagrams on following pages)
and then as a mass spring damper system.
We derived the state space representation of the system. This representation was
put into MATLAB to perform various simulations concerning the natural frequencies and
modes of vibration of the system.
Next, we simulated a bump in the road and viewed the response of the system to
this input. Finally, we determined the ideal range of absorber coefficients to provide the
most comfortable ride for passengers in the car.



AE 3515 Dale Arney and Chirag Talati
Computer Assignment II 13 July 2004
Page 3 of 13
Part A: Half Vehicle Suspension Model


As shown above, the automobile suspension can be pictured as a mass (the automobile)
and a pair of springs (the suspension system). Since the center of mass is not always at
the geometric center of the body, two lengths are created to compensate for this.

The given constants are:
k
1
= k
2
= k = 2.5x10
4
N/m
l
1
= 1.5 m
l
2
= 2 m
m = 1000 kg
J = 1250 kg-m
2


The equations of motion for this body (there are two equations because this is a two-
degree-of-freedom body) are as follows:


{ } { }
1 1 1 2 2 2
1 1 1 1 2 2 2 2
[( sin ) ] [( sin ) ]
[( sin ) ] [( sin ) ]
mx k x l u k x l u
J l k x l u l k x l u


= +
= + +



where u
1
and u
2
are the input displacements on spring 1 and 2 respectively (determined
by the road profile)

Taking into consideration the given constants, small angle approximations, and putting
the system into matrix form:


1 2 1
2 2
1 2 1 2 1 2 2
2 ( ) 0
( ) ( ) 0
k k l l k k u m x x
k l l k l l kl kl u J

+ =

+



AE 3515 Dale Arney and Chirag Talati
Computer Assignment II 13 July 2004
Page 4 of 13
By solving the generalized eigenvalue problem, the resultant eigenvalues are the natural
frequencies and the eigenvectors are the modes of vibration. This is done using the eig
function in MATLAB:


126.63 0
0 48.369

=



3.13 0.46
0.41 2.80
V

=




The diagonal values of the matrix are the squares of the natural frequencies (
n
) and
the V matrix contains the modes of vibration (each column is a mode which corresponds
to the natural frequency in the same column in the matrix).

What this means is that, if the system were to be perturbed by the given mode (x
increased to 3.13 and increased to 0.41, etc.), it would oscillate at the natural frequency
of 11.253.

Mode 1 Mode 2
Initial height (x) 3.13 m -0.46 m
Initial angle () 0.41 rad 2.80 rad
Resultant frequency 11.253 rad/s 6.955 rad/s

If converted into a state-space representation, this system would look like this:


1 1
2 2 1
1 1
3 3 2
2 2
4 4 1 2 2 1 2
1
2
3
4
0 0 1 0 0 0
0 0 0 1 0 0
( ) 2
0 0
( ) ( )
0 0
1 0 0 0
0 1 0 0
0 0 1 0
0 0 0 1
l
x x
x x u
k l l k k k
x x u
m m m m
x x kl kl k l l k l l
J J J J
x
x
y
x
x








= +






+






=




where
1
2
3
4
x x
x
x x
x


AE 3515 Dale Arney and Chirag Talati
Computer Assignment II 13 July 2004
Page 5 of 13
Part B: Half Vehicle Suspension Model with Damping Effects

The procedure used to solve this problem is very similar to the previous system without
damping, except a damping term is added into the equations of motion.


1 1 1 2 2 2
[( sin ) ] [( sin ) ] mx k x l u k x l u = +
1 1 1 2 2 2
[( cos ) ] [( cos ) ] b x l u b x l u +


{ } { }
1 1 1 1 2 2 2 2
[( sin ) ] [( sin ) ] J l k x l u l k x l u = + +


{ } { }
1 1 1 1 2 2 2 2
[( cos ) ] [( cos ) ] l b x l u l b x l u + +



Again, taking into considerations the given constants (also assuming that both of the
damping coefficients are equal to b) and putting the equations into matrix form:


1 2 1 2
2 2 2 2
1 2 1 2 1 2 1 2
2 ( ) 2 ( ) 0
( ) ( ) ( ) ( ) 0
b b l l k k l l m x x x
b l l b l l k l l k l l J

+ +

+ +





1 1
1 2 2 1 2 2
b b u k k u
bl bl u kl kl u

= +




AE 3515 Dale Arney and Chirag Talati
Computer Assignment II 13 July 2004
Page 6 of 13
In order to use MATLAB to do analysis, the state-space representation of this system will
be useful:


1 1
1 2 2 2 1
1 1 1 2
3 3 2
2 2 2 2
31 32
4 4 2 1 2 1 2 1 2
41 42
0 0 1 0
0 0 0 1
( ) ( ) 2 2
( ) ( ) ( ) ( )
l
b b
x x
m m
bl bl x x u
k l l b l l k b
J J
x x u
m m m m
b b
x x k l l k l l b l l b l l
b b
J J J J








= +






+ +


1
2 1
3 2
4 1 2
0 0
1 0 0 0
0 0
0 1 0 0
0 0 1 0
0 0 0 1
x
x u
b b
y
x u
m m
x bl bl
J J








= +













This time however,
1
2
1 2
3
4 1 2
1 2
x
x
x
b b
x u u
x
m m
x bl bl
u u
J J

and
2 2
1 1 2
31 2
2 2
2 1 2
32 2
2 2 2 2
1 1 2 1 2
41 2
2 2 2 2
2 2 1 2 1 2
42 2
( ) 2
( ) 2
( ) ( )
( ) ( )
b l l l k b
b
m m mJ
b l l l k b
b
m m mJ
b l l l b l l k
b
J J mJ
kl b l l l b l l
b
J J mJ

= + +


= +

= +

= +



The usefulness of this representation will become apparent in the next section when it is
used to make visualizations of motion.

Furthermore, when the damp command is used in MATLAB to find the natural
frequencies of this damped system, two results are given:

RESULTS Angular Frequency Frequency
First Result 6.9548 rad/s 1.107 Hz
Second Result 11.2531 rad/s 1.791 Hz

AE 3515 Dale Arney and Chirag Talati
Computer Assignment II 13 July 2004
Page 7 of 13
Part C: Natural Responses to Given Initial Conditions

By representing the half vehicle suspension model with a state-space representation,
MATLABs Controls Toolbox can be used to analyze the system. The following 3 pages
show the results of these analyses.

The first page of graphs shows the outputs of the first simplified model (suspension
consists of only springs). The initial conditions given are the initial conditions given by
the modes of vibration.

The second page of graphs shows the outputs of the second model (both springs and
dampers). Again, the initial conditions are the modes of vibration. For these graphs, the
absorber coefficient is the same: b
1
=b
2
=b=500N-s/m. Notice on both of these pages that
the outputs all have the same frequency (characteristic of the modes of vibration-natural
frequency relationship found in Part A).

In order to compare with other initial conditions, the third page of graphs shows the
outputs given random initial conditions and shows the odd responses resulting from them.
As the reader can see, there is not a uniform frequency across all of the outputs or within
the same output for that matter.

AE 3515 Dale Arney and Chirag Talati
Computer Assignment II 13 July 2004
Page 8 of 13



AE 3515 Dale Arney and Chirag Talati
Computer Assignment II 13 July 2004
Page 9 of 13



AE 3515 Dale Arney and Chirag Talati
Computer Assignment II 13 July 2004
Page 10 of 13



AE 3515 Dale Arney and Chirag Talati
Computer Assignment II 13 July 2004
Page 11 of 13
Part D: Responses to Inputs and Suitable Absorber Coefficients

Now we will look at the response of the system given an input (i.e. a bump). We
selected an impulse of 0.1m which is roughly equal to 4 inches. The speed of the car is
7m/s which translates to the time between inputs u
1
and u
2
to 0.5s. Using the lsim and
gensig commands in MATLAB we were able to view the results of this response. On
the next page is a graph of the response of the car to the inputs of the bump.


Max Position
(m)
Position Settling
Time (s)
Max Angle
(rad)
Angular Settling
Time (s)
500 (N*s/m) .08 7 .5 4
750 (N*s/m) .1 6 1 3
875 (N*s/m) .11 4.5 1.1 3
1000 (N*s/m) .125 4 1.25 3
1125 (N*s/m) .14 4 1.35 2.5
1250 (N*s/m) .15 4 1.5 2

From the information in the chart as the absorber coefficient increases the
maximum position and maximum angular position increase while the settling time
decreases. From these observations we conclude that the optimal absorber coefficient
lies between 875 (N*s/m) and 1125 (N*s/m). A selection of an absorber coefficient in
this range would allow for a compromise between displacement and settling time to
provide for a more comfortable ride for the passenger.

On the following page these results are shown graphically.
AE 3515 Dale Arney and Chirag Talati
Computer Assignment II 13 July 2004
Page 12 of 13

You might also like