C. S c h r 6 d e r 1 a n d A.

D u c h o w ~

Heavy Truck Handling Performance

Analysis in Vehicle Test and Computer
Simulation 2

REFERENCE: Schr6der, C. and Duchow, A., "Heavy Truck Handling Performance

Analysis in Vehicle Test and Computer Simulation," Tire Science and Technology,
TSTCA, Vol. 25, No. 2, April-June, 1997, pp. 119-136.

ABSTRACT: The dynamic stability of heavy trucks, as spinout, jackknifing, and rollover,
is highly dependent on vehicle configuration, driving maneuver, and the force and moment
characteristics of tires. Increasing safety requirements on the handling performance of
heavy trucks demand tools that allow a tire design engineer to predict tire influences on
the tire/vehicle system dynamic behavior. The computer simulation of handling perfor-
mance of vehicles offers possibilities of evaluating influences of tire design changes on
handling properties in any developing stage of new tire lines. Thus, modem simulation
techniques may contribute to the building and testing of tires in an early design stage.
This paper presents results from a recent program of tire/vehicle system research, applying
tire/vehicle testing and simulation techniques to a 40 ton truck-semitrailer combination.
The goal of this work is to visualize the possibilities of state-of-the-art simulation
technologies on the tire design process. Tire force and moment characteristics can be
calculated from the tire layout by an advanced tire model. The tire model for this type
of calculation is a multibody system. Calculated and measured dynamic tire characteris-
tics are used for the full vehicle handling simulation in ADAMS. Extensive tire charac-
teristic testing on the road and test stand was done to improve and validate the tire
model. Vehicle handling tests as steady state circular and lateral transient response tests
were done for the empty and laden vehicle with different tires to prove the vehicle
model. With the use of the simulation of the tire and vehicle behavior, the tire design
engineerwill be able to judge tire characteristics of different variants in an early design
stage. Vehicle dynamic simulation studies up to instability as spinout, jackknifing,
and rollover can be performed using modern CAE methods without harming man and
environment, but subjective and objective tire evaluation still remains necessary for
approving and validating the predicted results.

KEY WORDS: tire force, cornering stiffness, moment, self-aligning torque, truck tire

T h e n e e d to c l e a r l y u n d e r s t a n d w h a t a f f e c t s h e a v y v e h i c l e s t a b i l i t y is p a r a -
mount for the design and development of the whole vehicle system. Factors
s u c h as c e n t e r o f g r a v i t y , r o l l t h r e s h o l d , p a y l o a d c o n f i g u r a t i o n , a n d tire p r o p e r -
ties c o n t r o l v e h i c l e s t a b i l i t y . O n e o f t h e m o s t i m p o r t a n t a i m s in t h e tire d e s i g n

Continental AG, Tire Research/Tire Mechanics, P.O. Box 169, D-30001 Hanover, F. R. Ger-
many.
2 Presented at the fifteenth annual conference of The Tire Society at The University of Akron,
Akron, Ohio, March 19-20, 1996.

119
120 TIRE SCIENCE & TECHNOLOGY

process is an optimal fit of a tire to the vehicle. The confidence with which
a vehicle can be driven in any situation is of great importance for the tire
and vehicle development.
During the development of new truck tire lines, the design engineer has to
meet many requirements in low rolling resistance, endurance, tread wear,
high mileage, tire/road noise, wet grip performance, traction and braking
performance, and tire/vehicle safety (stability).
Different simulation technologies are available for the design process. For
instance, the FEM is used to optimize the tire endurance, rolling resistance,
etc. In the investigation of the handling performance of commercial vehicles,
simulation technologies are useful for predicting tire influences on the tire/
vehicle system behavior before the first prototype stage. Other advantages
are the possibility of varying parameters of the investigated tire and the
evaluation of the effects of the variations on the system behavior in an objec-
tive manner.
Thus, modern simulation techniques may contribute to the reduction in
cost and time for the development of new commercial vehicle tires. This can
be reached, when the applied simulation models contain enough detailed
information on the fire/vehicle system to be able to deliver realistic simulation
results. For our investigation, we use a Magic Formula-based tire model for
the force and moment representation, a complex MBS tire model for calculat-
ing fire dynamic properties from the tire design, and the ADAMS simulation
software for the full vehicle handling simulation. Calculated and measured
dynamic tire characteristics are used for the full vehicle handling simulation.
The goal of this work is to visualize the possibilities of state-of-the-art simula-
tion techniques in comparison to tire/vehicle testing using a tractor-semitrailer
combination.

Tire Force and M o m e n t M e a s u r e m e n t s

The tire force and moment testing was done at the Continental proving
ground with the University of Hanover mobile tire tester. The mobile tire
tester is explained by Pflug and Weber [1] and Hahn and Weber [2]. It is a
tractor-semitrailer with a gross weight of 38 tons. Tire testing was done on
a 500 m straight and even lane of the high speed oval. The fires tested were
315/80R22.5, 315/60R22.5, and 385/65R22.5. The tires can be steered from
a - 2 ~ to + 11 ~ slip angle for the measurement of the tire lateral force behavior;
longitudinal forces are caused by hydraulic disk brakes. For each tire size,
measurements for longitudinal and lateral force and self-aligning moment
characteristics under pure slip conditions and the tire properties for the com-
bined slip situation were done. All tire tests were repeated four times. Addi-
tionally, the dynamic slip angle characteristics of the tires were measured.
Therefore, a random slip angle was applied to the tires. All tires ran through
a brake-in procedure before testing. Figure 1 shows the lateral force and self-
 SCHRs ET AL. ON TRUCK HANDLING PERFORMANCE 121

aligning torque characteristics of one of the test tires as measured in the tire
tester coordinate system.
The tire properties around 0 ~ slip angle are characterized by the cornering
stiffness which is very important for vehicle handling. The cornering stiffness
increases as a function of wheel load for small wheel loads in a linear way
and for higher loads in a digressive way. On dry roads, a maximum friction
coefficient of around 0.8 can be reached for small wheel loads; for high loads
it drops down to 0.6.
The tire lateral force transient response is shown in Fig. 2. The measure-
ments show a significant decrease in lateral force amplitude for frequencies
higher than 2.0 Hz. Similar to automobile tires, the value of the lateral force
response on truck tires is dependent on the driving speed; this can be seen
in amplitude and phase. The vehicle speeds normally driven on highways
and normal roads narrow the differences in the lateral force response below
2.0 Hz slip angle frequency.

Tire Force and Moment Modeling

Tire Model for Handling Simulations
Vehicle motions are caused primarily by tire forces and moments that
result from powertrain, braking, and steering inputs. Tire forces represent a
significant part of the vehicle dynamic behavior. A comprehensive tire model
is therefore of considerable importance.
Tire forces and moments interact with wheel spin, steering system behavior,
and vehicle lateral, longitudinal, and inertial dynamics. Thus, the tire model
should properly represent tire force and moment properties to inputs of lateral
and longitudinal slip, wheel loads and camber.
To complete the tire force and moment model, the delay in force production
should additionally be taken into account. The delay in the tire force can be
explained by the tire rolling through some distance to produce the desired
slip angle.
For the representation of the steady state force and moment properties of
tires within full vehicle handling simulations, the empirical Magic Formula
[3,4] tire model is used. The tire force and moment delay is modeled as a
second-order filter applied to tire model inputs.
The tire model is coupled to the ADAMS vehicle model by an ADAMS
aggregate element statement (TIRE) to represent a tire, and user-written sub-
routines, which are set up so that each tire of a vehicle model can have
individual tire properties. Figure 3 shows the lateral force and self-aligning
torque characteristics of a tire as measured and computed by the Magic
Formula.

Tire Model for Force and Moment Prediction

The tire model to calculate the six force and moment components acting
on the rim, which are dependent on various boundary conditions, is a
 122 TIRE SCIENCE & TECHNOLOGY

40000.0

z 30000.0
oe,e* 9 9 6/~ 'o
A~

9
o 20000.0
4~eDep ~ 9
a9
9 9
L 9 Oq
m 10000.0

oo

1500.00

E
Z
I,..,.i
oOO 4 ~O
Coo~
a) 1000.00 w--
Oq
OO0
o jmoOq Ira6
1- UOU -O
D Oo
01 DOOm DQO I
e. I 0111 9 9
. m
Oo~ m~ 9
e- 500.00 O 9O wu ~ ; , I I B I r O o 9 I'oOo
O1
.m O001 O0'
~I. O 9
I ~OoOq DO o
O~~ ,..IOOI 'O41~e oOOO~ '~UOU ~ 9
ru, DOOqll u9 9 9!
eeP e,ee II
I.,L. POoO4 b4t'O9 Ab

0.00
0.0 2.0 4.0 6.0 8.0 10.0 12.0
Slip Angle (x [Degrees]

9 Measurement
FIG, 1 - - Lateral force and self-aligning torque characteristics of a 315/80R22.5 tire.
 SCHR(gDER E T AL. O N T R U C K HANDLING PERFORMANCE 123

4000.0
i iiiliiii i iiiiiii! i i i!ili!
........ i--- -~- i--,;-,~ "] !~-,'....... i---i--i-i-i-!~-,~! ....... ,~---i---.'--i-!-i-ii-

.E
r
O
o
,,o
m
3875.0

3750.0 iiiiiiiiiiiiiiiiii
iiiiiiiiii
iiiiiiii
iiiiiiiiiii
....... i--i-i--i-i-iil ....... i---i-i-iTi-i-~i ....... ~--I-I-TI-I[
3625.0 ....... i----i--~--~-,~-i~ ....... ~----i---~-4-~-~ -.'-,~....... i----~--~--i-i-i-~.
-,I ..................... ~ .......................... , ............... , .............

3500.0
0.0 0.1 1.0 10.0
v = 40 kph
v = 60 kph
0.00 i ! i if:',',: ~ ! , '~"i i i i!i!i _ _ m v = 75 kph
........ .--.I.-J*.L-L '-L J: ....... L. . . . . J- ~ ........ J--J--~.L-,-t
i ! i iiliil i i :''ri::! ~! i !iiii:
C ......+---~--!--'-"-~.......D--i---F~-~-.'-~.... -.',,~--~--~-~-i-~
r
r
.1000 ........... i-i-!iiiii- . . . . .

D.

-15.00
,,o
u

E -2o.oo ....... i___i__i._Lii_iiJ ....... ::____L_!_Li.i!_ii ....... i____i__i._i.i_::_ii

3
: : ', ',::!i
-30.00
0.01 0.10 1.00 10.00
Frequency [Hz]
FIG. 2 - - Lateral force transient slip angle response of a 385/65R22.5 tire at Fz = 40 kN.

multibody system with lumped masses specially set up for tires. These as-
sumptions utilize the specific properties of the tire to get a small calculation
time. The mass points of the model are connected by linear springs and
velocity-proportional dampers. There are different types of elements in the
sidewall, belt, and tread region so that the anisotropic and nonlinear properties
relative to stiffness, geometry, and material can be considered. For the truck
tire calculations, a model with approximately 3000 mass points was used.
The mathematical description leads to a differential equation system that
is very stiff, because of the anisotropic properties of the tire. That means that
the computation of tire characteristics still requires long computational times.
For this reason the tire model cannot be used within the full vehicle handling
simulation. Therefore, the tire characteristic calculation was done by varying
the input data in intervals. For further data processing the Magic Formula
tire model was used, in which parameters are fitted to the calculated data by
optimization routines.
 124 TIRE SCIENCE & TECHNOLOGY

40000.0
!!
z 30000.0
.. ,,~,,,,~"- " -F
IJ. t

20000.0 " ~'a"~" --

10000.0

o.o iV
1500.00

0) 1000.00 I
O"
II-.

O
I-
o'1
.m
r
500.00
.m
n
<C
q,,,
m

03
0.00
0.0 2.0 4.0 6.0 8.0 10.0 12.0
Slip Angle ~ [Degrees]

9 Measurement
Magic Formula
FIG. 3 -- Lateral force and self-aligning torque characteristics of a 315/80R22.5 tire from
measurement and Magic Formula.
 SCHRODER ET AL. ON TRUCK HANDLING PERFORMANCE 125

40000.0

............. .,............ ~......................

ii'1 ,;20
Fz 30
F. ,o k.,
k.k.c.,cu,a,,on
F z = 50 kN
30000.0
9 Force M e a s u r e m e n t ............. i . . . . . . . . . . i--=*, -~-'"
z
u . ~

o 20000.0 .......... ~........... ,.......... ~........ , ~~' - - . --~ ~:.,.-

-- ~. o . - ~- . .- ', ~
1,1.
m
L_

.......... ........... ...... . . , , . . ~ . . . . ~ _ _ . ~ ..........

_1 i i/ ",,, i i i ~ , i
10000.0 ..............
~;;,,{ ! ........ ~ . . . . . . . . . . . .
~ ...... ! ...................... i .......... ~ .........

...... iiiiiiiiiiiiiiiiiiill iiiiiiiiiill

il;iiiiill
0.0
L0 2.0 4.0 6.0 8.0 10.0
Slip Angle (x [ D e g r e e s ]
FIG. 4 - - C o m p a r i s o n b e t w e e n M B S calculated a n d m e a s u r e d lateral f o r c e characteristics o f
a 3 1 5 / 8 0 R 2 2 . 5 drive axle tire.

For the validation of the multibody tire model, calculated tire characteristics
were compared to measurements from tire testing on the road and the drum
test stand Figure 4 shows a good agreement between tire measurement and
calculation for the example of the lateral force characteristic on the r o a d
This tire model is used to calculate the influence of tire design parameters
or tireroad contact conditions on tire force and moment characteristics

Vehicle Handling Testing

The vehicle handling tests were conducted on a Mercedes Benz 1850 LS
tractor and a K~issbohrer SB 10-24 L semitrailer. All tests were done for the
empty and laden vehicle on different sets of tires.
The vehicle was equipped with 36 sensors from which 32 signals were
recorded by a measuring device. Steering wheel angle and torque, steer angle,
lateral and longitudinal acceleration, yaw velocity, sideslip angle, and tire
126 TIRE SCIENCE & TECHNOLOGY

deflections are just a few of the measurements taken. Depending on the vehicle
test, different sensor signals were detected. Each set of new tires was broken
in before any vehicle test was done.
The following vehicle tests were done:
1. Steady state cornering to determine the oversteer/understeer behavior
of the vehicle as a function of lateral acceleration,
2. Sinusoidal steering wheel input, frequency response measurement to
evaluate the extent and promptness of vehicle responses as a function
of steering wheel angle input, and
3. Single lane change to evaluate the steering effort and the effect of a
lane change as an excitation of yaw oscillations.
The steady state cornering test is the most frequently used procedure to
supply data on the steering tendency of vehicles. In our case it was done for
R = 30 m and R = 100 m right and left turn. Lateral accelerations up to 6
rn/s 2 could be achieved on the dry courses.
Figure 5 shows three different test procedures, which were tested at the
beginning of the evaluation. Variant 3 has been proven to be the one with
the best reproducibility, but is the most time-consuming one. One of the
major influences on the vehicle behavior obtained from this test is the change
of the tire properties with increasing tire wear caused by cornering. Therefore,
our tests were done with new sets of tires for right and left turn.
Different test procedures as shown in Fig. 5 cause test data that are affected
by the tire changes due to wear. Procedures one and two do not deliver
reproducible results, because the driving of the vehicle at high lateral accelera-
tions with high lateral forces cause extreme tire wear. This tire wear changes
the cornering stiffness and self-aligning torque the most. That means that,
already, the second run of the vehicle test starts with different tire characteris-
tics than the first. Test procedure three repeats each lateral acceleration step
and, thus, the change of tire properties for each lateral acceleration step in
the lower range can nearly be neglected. However, the influence of tire wear
has to be considered for higher lateral accelerations.
A sinusoidal steering wheel input is used to evaluate the frequency response
properties of a vehicle as the intensity of periodic steering wheel input to
vehicle reaction. The procedure assumes that the response characteristics of
a vehicle can be examined in a linear way with good approximation. The
frequency response function describes the amplitude ratio of output to input
value and phase lag between both values in steady state condition depending
on the frequency.
In comparison to other methods as pulse- or pseudo-random steering input,
this method provides the most reliable results for commercial vehicles. The
actual tests were done for v = 60 kph and v = 75 kph starting from a steering
wheel input frequency of 0.2 Hz with an increment of 0.2 Hz up to the limit
of the driver's performance, which was 2.0 Hz. The steering wheel amplitude
was set to a value that caused a lateral acceleration of 2.5 ntis 2 at 0.2 Hz.
 SCHRC)DER ET AL, ON TRUCK HANDLING PERFORMANCE 127

t~
o

~k
S Stepwise increased
velocity, kept constant
for measurement

Continuously increased
velocity, measurement
in certain regions
interval for
Measurement

Measurement was
repeated four times
> /, : : ; i i for every velocity

time

FIG. 5 -- Different testing procedures for steady state cornering.

The single lane change maneuver is applied to analyze the amount of

required steering to be performed by the driver to remain on the track and
to stabilize the vehicle. This test was carried out by driving the vehicle on a
desired path that was marked by pylons at the right and left boundaries of
the path. Figure 6 shows the testing conditions for the single lane change.
The vehicle handling tests were started with a velocity of 50 kph and increased
by an increment of 5 kph up to the vehicle stability limit. These tests were
also carried out for the empty vehicle on a wet road.

Vehicle Model
The vehicle model of the tractor-semitrailer combination was written in
ADAMS. The model of the tractor is divided into a front part and a rear part.
The front part contains a driver cabin, drive unit, steering system, and front
axle mountings. The rear part contains rear axle mountings and the coupling
unit. The front and rear part of the tractor are coupled by a torsional stiffness
to consider the frame torsion of the tractor. The front and rear axle units
including springs, dampers, and antiroll bars are coupled to the tractor parts
accordingly.
128 TIRE SCIENCE & TECHNOLOGY

9.0 I I I

6.0
A A

3.0

9 9 9 ~ r

0.0

-3.0

-6.0

-9.0 I t I
30.0 50.0 70.0 90.0 110.0
Driving Distance x [m]
FIG. 6 - - Path for the single lane change.

The semitrailer can be divided in up to five parts to evaluate the influence

of frame torsion on the vehicle handling. The three axle units of the semitrailer
are coupled to the frame. The vehicle model is shown in Fig. 7.

Validation o f t h e Tire/Vehicle System

The test data obtained from the vehicle handling tests were used to validate
the vehicle model. The validation is not meant to fit the simulated data
exactly to the measured data, but to gain confidence that the vehicle handling
simulation is giving insight into the behavior of the simulated vehicle. Geo-
metrical and physical data for the vehicle model were taken either from
drawings, technical data sheets, or directly from the vehicle. For the investiga-
tion of the influence of tire properties on the vehicle stability, it is important
to have an appropriate vehicle model for the maneuvers of interest. The test
 SCHRODER ET AL. ON TRUCK HANDLING PERFORMANCE 129

A
I
A

A
A

A
A

A
A

A
FIG. 7 -- Model of the tractor-semitrailer.

data were used to check whether or not the input parameters for the vehicle
model are reasonable. The steady state cornering test will now be taken as
an example for the validation of the vehicle model. The test was repeated
several times to gain confidence in the test data.
Figure 8 shows the steering wheel angle, sideslip angle of the tractor, and
the lateral acceleration of the semitrailer as a function of the lateral accelera-
tion of the tractor. In the lower lateral acceleration range of up to 3.0 m/s 2
both measurement and simulation exhibit a nearly linear increase in steering
wheel angle. For lateral accelerations higher than 3.0 m/s 2 a progressive
increase of the steering wheel angle can be seen. From 4.5 rn/s 2 on, the
simulated steering wheel angle increases more than the measured one. This
effect may be caused by tire wear, which changes tire properties during the
vehicle handling test. In general, measurement and simulation agree with
good accuracy.
In the graph for the tractor sideslip angle we can find similar effects.
Measurement and simulation show a nearly linear behavior up to 3.5 m/s 2
lateral acceleration and then a progressive behavior. The difference in the
sideslip angle above 4.5 m/s 2 corresponds with the difference in steering
wheel angle.
130 TIRE SCIENCE & TECHNOLOGY

150.0 ~ I I i / ,

100.0
C
50.0

0.0 , I , I , I , , I ,

3.0 i , r ' i i

2.0
C
1.o

0.0

-1.0

-2.0
6.0 . . . . . .

~ 4.0

~ 2.O
i
I-- 0.0 * V e h i c l e T e s t Data
-2.0 ~ I , I ~ I , I , I
0.0 1.0 2.0 3.0 4.0 5.0 6.0
Tractor Lateral A c c e l e r a t i o n [m/s 2]

FIG. 8 -- Calculated and measured vehicle handling characteristics of the unladen vehicle.

Tractor and semitrailer lateral accelerations are linearly correlated for the
whole lateral acceleration range. Again, measurement and simulation agree
with good accuracy.
The next example for the validation of the vehicle model is the single lane
change maneuver of the vehicle at 40 ton gross weight and at a speed of 60
kph. As input for the simulation, the steering wheel angle was taken from a
measurement. A traction controller within the vehicle model kept the velocity
constant at 60 kph. The steering wheel input for this single lane change maneuver
can be compared to a complete period of sinusoidal steering wheel input with
 SCHRC)DER ET AL. ON TRUCK HANDLING PERFORMANCE 131

j
100.0 I 20.0

50.0 10.0

~
o'--'
I,,i,I
0.0 0.0

F
o
.m
I--
-50.0 -10.0

-100.0 , t , -20.0 i I , I i r
5.0 7.0 9.0 11.0 5.0 7.0 9.0 11.0

4.0 20.0

9 2.0 10.0

~
0.0 0.0
o .m

I--
-2.o -10.0

-4.0 , i , t , t -20.0 , I r I t I
5.0 7.0 9.0 11.0 5.0 7.0 9.0 11.0
T i m e t [s] T i m e t [s]

Simulation
* Measurement
FIG. 9 - - Comparison of measurement and simulation, single lane change, vehicle at gross
weight, 40 ton, v = 60 kph.

an amplitude of 80 ~ and a frequency of around 0.5 Hz and then a 10~ input to

stabilize the vehicle. For this lane change a maximum lateral acceleration of
3.0 m/s 2 was reached with only little time delay at the tractor. The yaw velocity
of the tractor follows the steering wheel input immediately and reaches nearly
15~ the reaction of the semitrailer is delayed by 0.4-0.5 s. As for the steady
state cornering, again measurement and simulation agree with good accuracy.
The results of the measurement and simulation are shown in Fig. 9.
 132 TIRE SCIENCE& TECHNOLOGY

40000.0

'/~I Fz
Fz== 403020
kNkNkN'
dry road .....................................................

30000.0 F, = 50 kN .............................. ~ .......... ; . . . . . . . . . .

wet road ~
Z
i !ti i _i . . . .
0t_
0 20000.0 ..........~............ ~.......... ~........... ;..........t - ........'~....
i i i i /i =-
. ......... ~..........~'.........
! i
tt
i i i !/ '1 i ' : !

', ', ', #S S ' ' ', ', ',

......... : ........... '. . . . . . . . . . . " . . . ~ . . . ~ _ .... ~ .................... '. . . . . . . . , .......

10000.0 m

0.0
0.0 2.0 4.0 6.0 8.0 10.0
Slip Angle ~ [Degrees]
FIG. 10 - - Calculated lateral force characteristics on dry and wet roads f o r a 315/80R22.5 HS
45 tire.

Application
After the vehicle model and tire model were proven to be of good accuracy,
the full vehicle model was used to predict the behavior of the tractor-semi-
trailer combination performing the single lane change m a n e u v e r on a wet
road. Therefore the tire characteristics on the wet road were calculated by
the MBS tire model. The frictional properties of the MBS tire model were
set up for a wet road. The MBS tire model calculated the lateral force charac-
teristics of a steer axle tire on dry and wet roads (Fig. 10). The m a x i m u m
lateral force on the wet road was reduced to 80% of the value on the dry
road. A change in cornering stiffness could not be found. As input for the
simulation, the steering wheel angle was taken from one of the measurements.
A traction controller within the vehicle model kept the velocity constant at
60 kph. Figure 11 shows the vehicle characteristics of the unladen vehicle
on the wet road as measured and calculated. A steering wheel amplitude o f
100 ~ had to be used to keep the vehicle in the lane and to reach a lateral
 SCHRODER ET AL. ON TRUCK HANDLING PERFORMANCE 133

100.0 20.0 ' I " I " I " I

50.0 10.0

o~ 0.0 o.o

I,,-
-50.0 -10.0

-100.0 9 I , I 9 I I I -20.0 . i . , . , . ,
5.0 7.0 9.0 11.0 13.0 5.0 7.0 9.0 11.0 13.0

4.0 . , 9 , 9 , 9 , 20.0

2.0 10.0

o~-
0.0 ~ 0.0
o

I--
-2.0 -10.0

, I , i l l i l
-4.0 -20.0 . ' 9 ~ , = , I
5.0 7.0 9.0 11.0 13.0 5.0 7.0 9.0 11.0 13.0
Time t [s] Time t [s]

Simulation
9 Measurement

FIG. 11 - - Calculated vehicle handling characteristics on a w e t road f o r the unladen vehicle v

= 60kph,

acceleration of 3.0 ITI/S 2. The yaw velocity of the vehicle reaches its maximum
with 12~ for the tractor and 10% for the semitrailer. The reaction of the
trailer follows the tractor reaction with a time delay. Measurement and simula-
tion agree with good accuracy.
The next example for the application of full vehicle handling simulation
is the driving of the unladen vehicle on a handling course. For this simulation
we coupled the IPG or ADAMS/driver to the vehicle model. The vehicle
134 TIRE SCIENCE & TECHNOLOGY

.... V e h i c l e Path
C o u r s e C e n t e r Line

FIG. 12 - - Calculated vehicle path on a handling course including a driver model

was driven from standstill. Limits for driving speed, lateral and longitudinal
acceleration, and longitudinal deceleration were set up. For example, the
maximum vehicle speed was set to 80 kph; maximum lateral acceleration
was set to 6.0 m/s 2.
The centerline of the handling course in comparison to the path of the
vehicle is shown in Fig. 12. The driving of the vehicle on the handling course
combines the following different parts of vehicle handling evaluation:
1. Step steer
2. Alternating curves
3. Steady state cornering
4. Straight ahead driving
5. Acceleration/deceleration
Therefore the driver model has to be seen as a complex controller to fulfill
the above-mentioned tasks. Results like lateral force as a function of slip
angle, wheel load, camber angle, and longitudinal force can be evaluated
 SCHRODER ET AL. ON TRUCK HANDLING PERFORMANCE 135

10.0

5.0

c
0

_r 0.0
o
o
<

.J
-5.0
Tractor
Trailer

-10.0
0.0 50.0 100.0 150.0 200.0
Time t [s]
FIG. 1 3 - - Calculated vehicle lateral acceleration on a handling course, simulation including
a driver model.

under realistic driving and loading conditions. These simulation results can,
for example, be used for tire wear and durability prediction.
As an example, the lateral acceleration of the tractor and the semitrailer
while the vehicle is driven around a handling course is shown in Fig. ! 3. In
some regions on the course the vehicle model is driven at the stability limit,
which can be identified at those points in the diagram where lateral accelera-
tions greater than 6.0 m/s 2 were reached.
To evaluate a vehicle's handling behavior using simulation tools like the
driver model requires the judgement of subjective evaluation engineers evalu-
ating plots like Fig. 13 and studying the animation of the vehicle going around
the simulated test course.

Summary and Conclusions

In this paper possibilities for simulating heavy truck handling performance
as a function of tire force and moment properties are shown. The simulation
136 TIRE SCIENCE & TECHNOLOGY

predictions are qualitatively compared to the measured data. Tire measure-

ments, which were done on the Continental proving ground "Contidrom"
could be accurately represented within the vehicle handling simulation by the
empirical Magic Formula tire model. Tire transient effects were covered by
applying first- and/or second-order filters to the tire model inputs.
Complex tire models like the MBS with lumped masses make it possible
to calculate tire characteristics from the tire design.
The full vehicle handling simulation shows, by the example of steady state
cornering of the empty vehicle and single lane change of the laden vehicle,
that the input data for the vehicle model is reasonable and the vehicle model
represents the vehicle handling characteristics of interest with good accuracy.
Finally, applications of the full vehicle handling simulation also including
a driver model are shown.
The conclusions from this paper are that tire/vehicle handling simulations
can be performed with reasonable accuracy. Validated models can, further-
more, be used to give feedback on parameter changes and the influences of
design variables on the vehicle handling. Objective and subjective testing still
remains necessary for approving and validating the simulation results.

References
[1] Pflug, H.-Chr. and Weber, R., "Seitenfiihrungskr~ifte von Nutzfahrzeugreifen im echten
StraBenbetrieb," Automobil-lndustrie 1984, Heft 4, S. 475.
[2] Hahn, W.-D. and Weber, R., "Messung der Fiihrungseigenschaften von Nutzfahrzeu-
greifen," VDI-Berichte, Nr. 741, 1989, S. 25.
[3] Pacejka, H. B. and Bakker, E., "The Magic Formula Tyre Model," Proceedings ofthefirst
International Colloquium on Tyre Models for Vehicle Dynamics Analysis, Vehicle System
Dynamics, Volume 21, 1993, pp. 1-19.
[4] Bayle, P., Forissier, J. F., and Lafon, S. " A New Tire Model for Vehicle Dynamics Simula-
tions."
[5] "Reifeneigenschaften und Fahrsicherheit," BMFT-Forschungsbericht TV 7672.
[6] Heydinger, G., et al., "The Importance of Tire Lag on Simulated Transient Vehicle Re-
sponse," SAE Paper No. 910235.
[7] Heydinger, G., et al., "Pulse Testing Techniques Applied to Vehicle Handling Dynamics,"
SAE Paper No. 930828.
[8] Loeb, J. S., et aL, "Lateral Stiffness, Cornering Stiffness and Relaxation Length of the
Pneumatic Tire," SAE Paper No. 900129.
[9] McRuer, D. and Klein, R. H., "Effects of Automobile Steering Characteristics on Driver/
Vehicle Performance for Regulation Tasks," SAE Paper No. 760778.
[10] Mitschke, M., "Dynamik der Krafffahrzeuge," Band C, Fahrverhalten, Springer-Verlag
1990, ISBN 0-387-15476-0.