Professional Documents
Culture Documents
Performance Vehicle Dynamics - Chapter - 9
Performance Vehicle Dynamics - Chapter - 9
Performance Vehicle Dynamics - Chapter - 9
In designing a high-performance car, there are clearly some design aims that
stem from common sense. We need an engine with plenty of power, tyres
with good grip and a low overall weight. If we have a knowledge of the
competition, we might be able to list some targets that will give our car
an advantage. This is all well and good, but there is a distinct limit to
how much one can improve a car this way. We will never know all the
relevant details about the opposition, and if we simply try to emulate them,
we shall always be behind either commercially or on the track. In addition,
there is always the problem of resources. There is only ever a finite length of
time before the next deadline, whether it’s a race or a launch and always a
limit on expenditure too.
If we are to be rational about designing a car, we need to put numbers on
these quantities. When we say we would like to improve the design of a rac-
ing car, what we really mean is that we want to know which activities will
produce the greatest improvements in performance per unit effort in making
them happen. We need to know, for example, whether spending half our
time and budget on improving the engine will yield as much improvement
as spending it on improving the chassis dynamics. In a racing situation, it is
lap-time simulation that is used to answer these questions, and in this
chapter, we will develop an understanding of how it is done. However,
for performance road cars, we still need the same ability to predict the whole
vehicle performance, and full-vehicle simulation is the process by which this
is done. In a lap-time simulator, one enters all the relevant vehicle dynamic
data including the track, and the package will determine the lap time.
Changes can then be made and their effect determined. In full-vehicle sim-
ulation, again, all the vehicle dynamic data are entered, and the package is
then used to simulate manoeuvres. The performance of the vehicle in terms
of ride quality and comfort can be determined as well as dynamic perfor-
mance. The effect of changes in the design can be studied and improvements
made. In both cases, validation is an important step, meaning comparison of
the simulation with test data to confirm that the simulation is acceptably
accurate before using it for optimisation.
Velocity
50 m Straight – 20 m
constant radius
A
V1
B C
V2
S 1 S2 S3 Distance
produce the line from the origin to point ‘A’. We also know the maximum
deceleration, 1:0 μ, and therefore the gradient of the deceleration line
from S1 to S2. Since we know the speed with which the car must enter
the corner, we can draw in the two lines, and the point in time at which
they cross is the point at which the driver must go from wide open throttle
to full braking, normally called the ‘braking point’.
First determine V2:
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
V2 ¼ ðLa^tacc RÞ
Fig. 9.2 Any racing circuit can be divided up into a series of J-turns.
all the knowledge we gained in Chapters 1–8 because one needs to have that
level of understanding to know why any given vehicle behaves as it does.
The limit on turn in response is a good example; you can’t exceed the the-
oretical limit. However, armed with all this understanding, the availability of
packages such as Adams Car makes for rich simulation possibilities and no
doubt goes a long way towards explaining why current levels of refinement
on road cars are so good and why racing cars are so fast.
Adams Car is divided up into two modes of use, the ‘template builder’
and the ‘template user’. As a template user, you can call up different subas-
semblies from the library such as the one in Fig. 9.4 and link them together
to build an entire vehicle.
Fig. 9.5 shows a full-vehicle assembly in Adams Car. It is made by linking
a collection of subassemblies together. Once ‘assembled’, the vehicle can be
put through a whole range of manoeuvres and events. It can be made to
perform a step steer, for example, or a lane change or follow a prescribed
path. A user enters the details for the manoeuvre and the package then deter-
mines the vehicle response, and this can be viewed as a movie or studied in
detail with graphs of any parameter one might care to track. If you become a
template builder, then it is possible to produce new vehicles of whatever
design you wish to analyse. Fig. 9.1 shows a vehicle on a rough road and
is a very advanced simulation.
Fig. 9.4 A front wishbone assembly in Adams Car. Adams, Adams Car, Adams View and
other MSC product names are trademarks or registered trademarks of MSC Software
Corporation and/or its subsidiaries in the United States and/or other countries. Image
provided courtesy of MSC Software. 2017 MSC Software Corporation.
324 Performance Vehicle Dynamics
Fig. 9.5 A full-vehicle assembly in Adams Car. Adams, Adams Car, Adams View and other
MSC product names are trademarks or registered trademarks of MSC Software Corporation
and/or its subsidiaries in the United States and/or other countries. Image provided courtesy
of MSC Software. 2017 MSC Software Corporation.
Fig. 9.6 Six-wheel off-road vehicle manufactured by Arctic Trucks. Reproduced by kind
permission of Nolan McCann and Arctic Trucks.
Fig. 9.7 Simulation in Adams Car of the truck in Fig. 9.8. Reproduced by kind permission
of Nolan McCann and Arctic Trucks.
Fig. 9.8 Correlation between test and simulation for truck yaw velocity. Reproduced by
kind permission of Nolan McCann and Arctic Trucks.
326 Performance Vehicle Dynamics
when the changes are also made on the real car, it too will improve as the
simulation did. This process is applied equally to racing as we see in the next
section. The two huge advantages offered are that firstly a vast number of
parameter setups can be examined, far more than practical by testing alone,
and secondly it is much cheaper than extensive testing. With simulations of
this quality made possible, testing should really be a process of confirmation
rather than experiment.
Fig. 9.9 The front end GUI for the lap-time simulator ChassisSim. Reproduced by kind
permission of Danny Nowlan—ChassisSim.
The graphical user interface shown in Fig. 9.9 allows the user to call up
existing library circuits, cars, etc. and edit them to suit the case in hand.
After this, the control panel on the right is used to run simulations from
which the results can be studied. Fig. 9.10 shows the track map for a sample
simulation.
Fig. 9.11 Plotted data for the simulation. Reproduced by kind permission of Danny
Nowlan—ChassisSim.
Fig. 9.11 shows output data for the simulation in hand, and the distance
along the bottom axis refers to the distance from the start of the above
circuit. In this case, speed, steering input and throttle are shown, but all
dynamics parameters of interest can be plotted.
In the top line, we see the speed trace for the racing car, and it is clear that
the car is never in equilibrium, accelerating all the time, and that it spends a
minority of the time at terminal velocity. Using tools such as ChassisSim,
vehicle dynamicists are able to determine exactly what set of vehicle param-
eters will result in the shortest possible lap times, something that is not
otherwise easily answered. Issues such as damper settings bring advantages
in one situation but not in another; exactly how all these play out for a par-
ticular circuit is not easy to anticipate. The existence of transient lap-time
simulators such as ChassisSim makes for excellent design opportunities in
racing and goes some way to explaining why racing packs are generally
so close in performance.
The final step in the process of being a successful racing vehicle dynami-
cist is to overlap computer-simulated data with real-life telemetry from the
car in hand, just as above from road car simulation. This is shown using
Lap-Time, Manoeuvre and Full-Vehicle Simulation 329
ChassisSim in Fig. 9.12. In the figure, we see graphs, including throttle posi-
tion, the steering wheel position for a neutral car overlaid with the actual
position and roll angle. Once a good level of agreement has been developed
between measured data and simulated data, then one can make adjustments
in the model, find changes that bring improvements and then reasonably
expect that when these changes are made on the real car, the same improve-
ments will result.
Thus, lap-time simulation can be used to determine the overall best
package for a given circuit. This certainly makes it seem as if, in vehicle
dynamics, only lap-time simulation is needed, but of course, this is not
the case. Lap-time simulators determine what the lap time will be, whereas
vehicle dynamics determines why.
And finally, to quote from the introduction at the start of this book,
‘The issue is that until you put numbers on things you’re wasting your time, you’re
just playing about’