Page 50 - Computational Fluid Dynamics for Engineers

1.5 Aerodynamics of Ground-Based Vehicles 35

the vehicle. There are, of course, exceptions; this is obvious from the flight of
VSTOL-aircraft and helicopters.
The influence of aerodynamic forces depends strongly on the vehicle speed
and weight. For example, automobile aerodynamics is not of great concern in
traffic on city streets, and a heavy train operator usually ignores the side wind ef-
fects. While the aerodynamics of motor vehicles has been investigated for many
years, it did not attain prominence until improved roads allowed for greater
speeds and fuel crises inspired demand for improved fuel economy. Early at-
tempts to streamline vehicles were mostly based on experience gained from
aircraft aerodynamics, and the resulting designs were not always practical or
accepted by the general public. As early as the 1920's, it was demonstrated that
a drag coefficient of 0.15 was attainable under ideal conditions, which should
be compared to the then-prevalent box-design drag coefficient of about 0.8. Im-
provements have been slow in coming, and the drag coefficient of post-World
War II automobiles remained around 0.5 until the fuel crises in the 1970's. As
some critics claim, the reduction from 0.8 to around 0.5 was more due to styling
than conscious aerodynamic development. Since that time, however, drag coeffi-
cients have been reduced to around 0.30 by systematic attention to aerodynamic
details, and there is promise for further improvements. Automobile aerodynam-
ics, however, will always be subject to constraints imposed by utility, styling,
and public acceptance.
The principal tool used to study automobile aerodynamics has been the
wind tunnel. Testing began with small-scale models in aeronautical facilities
and has evolved into the use of special full-scale wind tunnels run by the larger
automobile manufacturing companies. Testing in a wind tunnel creates its own
problems because the boundary layer on the ground plane interferes with the
simulation of the actual flow conditions. Several remedies have been proposed
such as reflection models, tangential blowing, a moving ground plane, etc., of
which the moving ground plane provides the best correlations with road tests.
Since flow details underneath the vehicle; and in the wheel well are related to
drag, a refined test may include provisions for spinning the wheels, which adds
a further complication.
In general, the drag of a typical passenger automobile is essentially pressure
drag or is due to local flow separation. For this reason, the shape of the sharp
edges from which the flow separates has a definite effect on drag. Reference
[29] gives an example of a "detail-optimization" which reduced the drag coef-
ficient from 0.48 to 0.32 without noticeable changes in the appearance of the
automobile. Since an automobile is basically a blunt body, flow around its longi-
tudinal edges sets up vortical flow and causes vortex drag that is not necessarily
associated with lift or induced drag, although well-rounded shapes resembling
half bodies have considerable lift and consequently induced drag. The flow un-
derneath the car has a tendency to diverge to the sides, creating low pressure