rec.autos.simulators

lines below from the [physics] stanza of the gpl_ai.ini file:

nominal_downforce_coeff = 5700.657227           ; ????????? coeff = v^2 /
X.XXXG. = XXXMPH^2 / X.XXXG = 200MPH^2 * 7 / ONE_G
nominal_drag_coeff = 2638.590088                ; ???????? coeff = v^2 /
0.XXXXG. = XXXMPH^2 / 0.XXXXG

It's only a peek under the hood, but they're paying attention it seems.  Wonder
what some fudging would bring there?  Invisible wings, anyone?

Can you say Katmai?  600+mhz?  Voodoo3?  That's pretty much the same argument
Papy used for not putting in rain or force feedback - too much overhead.  It's
all about evolution, and the future looks promising, if costly.

Cheers,

Steve B.

remove "edy" from address for email

(snip)

If, by the term "simulators," you mean CFD (computational fluid dynamics), I
think that it would be safe to say that both are used. I spoke to someone at
Ford (I think) a couple of months back who mentioned that they were in the
process of building a new wind tunnel. We have also talked to a NASCAR team
that mentioned that they had scheduled some time in a wind tunnel next year.

Of course, there have been quite a few published reports about people doing
aerodynamic studies on cars using CFD. I would imagine that the car
industry, like the aviation industry, has seen an increase in the use of
computational methods over the past several years.

Although I have never been involved with aerodynamics in the automotive
industry, I used to work in the aerodynamics group at a missile development
company. I know that in that environment, wind tunnel testing and CFD are
used together, and replacing the wind tunnels won't happen any time soon.
The two methods complement each other.

If I'm not mistaken, judging by the cigar shape of the cars, 1967 was before
aerodynamics was leveraged to play a crucial role in racing.
It is safe to say that the aerodynamic modeling in N3 is much more elaborate
than in GPL. We have spent a great deal of time developing a full 6-DOF
aerodynamic model, using as much wind tunnel data as we could get. (In my
opinion the model is better than some flight sim models :) ) The model
includes drafting and damage, and was designed to be accurate at any car
orientation. As a matter of fact, in the current version of N3 with roof
flaps "turned off," during a spin the car exhibits the pre-1994 behavior of
liftoff.

The aerodynamics model is easily extensible to the aerodynamic
characteristics of different car shapes. As far as modeling wind eddies.....
maybe N4.

Jeremy Furtek
Papyrus Design Group

Both. They use scaled down models, and fluid dynamics. I remember an
issue of  Racecar engineering that had an article on how Bennetton did
their aero.

Actually, why should it be so difficult to implement? In N2 and ICR2 I'd
say they had a pretty good hang on making the wing adjustments feel
real. In reality, not many factors change; it's all about "adding
weight" - simplified, of course. In those early years, say without
Venturis, and the wings doing all the extra downforce (the chassis were
not implemented as a part of the areo package at that stage). The
calculations would need speed, how much the wing "pushes" at different
speeds, and the effect on drag. Then this could be used to calculate
increased grip. Remember, the model is very advanced; feel the surface
of Zandvoort, and compare it to Silverstone....

I do not think the eddies can be simulated to easily; just look at the
flop McLaren had with their central wing a few years back. I do not
think wind is implemented now either; a car without wings is still
vulnerable to crosswinds..

   Interesting topic! Let's hear more views!

Matt Knutsen

<snip>

I hate to say it, but I think that's by far overestimating the depth of GPL.
Downforce will probably just be a function of wing angle, speed, and
maybe area.  But we'll probably never know for sure just how they do
whatever they do.

rob.