Ruud van Gaal
Free car sim : http://www.racesimcentral.net/
Pencil art : http://www.racesimcentral.net/
I noticed at Spa that the F1 cars have equal tires in front and rear,
unlike in Cart. How are they doing things there? (or is there a rule
stating you must use equal tires in front and rear?)
Ruud van Gaal
Free car sim : http://www.marketgraph.nl/gallery/racer/
Pencil art : http://www.marketgraph.nl/gallery/
"tu quoque fili" ;o)
> > Damn ! :o)
> > So my first impression was the good one, i didn't understand anything !
> > What the meaning of "Et tu Brutus" ???
> "You too, Brutus" ?
> (Uttered by Caesar as he was stabbed to death according to Shakespeare)
> --
> -asbjxrn
anyway, i don't think that load distribution is relevant here, since
i've tried just about anything from 100/0 to 0/100, however the problem
(uncontrollability in oversteer situations) remains.
it's not that i don't like general oversteer/understeer attitude of my
engine. if i implement an FWD car it gives a lot of understeer just
like IRL. the problem is that once the rear brakes out i never manage
to regain control.
i have not implemented roll centers (thought they would result from the
other parameters, but read that they are necessary), so understanding
and implementing them is the next suggestion i will try.
jonas
They don't, AFAIK.
A couple of years ago, Tom Walkinshaw tried rear tyres all round at
the Hungaroring until Dunlop banned it on "safety grounds"; he said
afterwards that what he'd /really/ wanted to do was use front tyres
all round at Hockenheim!
It's probably just the perspective but I'm pretty sure the rears are
wider.
Jonny
Things to check off the top of my head would include your longitudinal force
combining--try setting your slip ratio to zero and coast into a turn to
simplify the situation, also check to make sure the "wrong end" of your tire
curve checks out correctly, with no sine(x) = x approximations starting to
bite you as becomes large, also check that you are correctly factoring in
the angular velocity of the car body at the wheel point, that you aren't
getting overlarge integration errors, esp from wheel hop or wheel rotation
chatter, etc, etc. It can be very hard to get a good chassis setup even for
a sim you wrote yourself--it's about as hard as setting up a real car, which
takes years of experience, but you should be able to throw on a stiff front
anti-roll bar on your vehicle and at least get it to plow like a pig into
corners... keep on trying, the first time you can make your own car that
feels right is an amazing experience!
-Dave P.
i just read some info on this, and it seems that tire-load sensitivity
and camber angles do seem to have an important impact on dynamics (i
didn't know about that before, nor did i understand, why a front heavy
car should be understeer in first place). the problem is that the
pacejka formulae i use seem to take these things into consideration,
however the pacejka constants i have (taken from Brian Beckmann's
articles) zero out the effect. they don't seem to be suitable for the
car i'm modelling anyways - at maximum efficiency the Beckmann tires
project 1.7 times their normal Force. i though that in school physics
they tought that coefficients of friction above 1 are not possible. if
i use the unmodified numbers from Beckmann my car is actually likely to
flip over in curves due to the high lateral forces.
unfortunatly these seem to be the only pacejka numbers one can find on
the net. any suggestions on where to get better numbers?
jonas
Most other popular and wrong theories taught in physics are probably
found in fluid dynamics, especially with respect to mishandling the
Bernoulli's equation, probably the most abused physics formula ever :).
-Gregor
> i though that in school physics
> they tought that coefficients of friction above 1 are not possible.
Hmmm.. Ummm.. Well... I dunno?? Is it? lol If it is, I sure went to a
lot of extra trouble for nothing! I think it might be the same except you and
I would be off by one time step probably, as you're taking the last time step's
velocity and I'm taking the current one... Right? Guess I was just so proud
of myself for coming up with a vector method all on my own :-)
Todd Wasson
---
Performance Simulations
Drag Racing and Top Speed Prediction
Software
http://PerformanceSimulations.Com
Yeah, you're right. Guess I thought (wrongly) too quickly on that one. In
that case, the load sensitivity issue would be the only factor, right?
Todd Wasson
---
Performance Simulations
Drag Racing and Top Speed Prediction
Software
http://PerformanceSimulations.Com
Yeah, you're probably right. As Mike Stanley pointed out, the yaw torque
cancels out my point. Maybe load sensitivity isn't the problem either. Guess
I can't help much there, sorry :-)
I don't think that will fix it. Mostly, putting in roll centers at this
point would effect transient behavior and total roll angle and that's about it.
Once the car breaks away, it'll most likely still have the same problem.
The only thing left I can think of to try is using less grip at the front.
Todd Wasson
---
Performance Simulations
Drag Racing and Top Speed Prediction
Software
http://PerformanceSimulations.Com
As Gregor said, friction coefficient can go well over one. Top fuel
dragster tires have peak grip coefficients of over 4. If they can't go over 1,
dragsters would never run better than about 9 seconds at 200 mph (they run
4.7-8 at 330+ these days, I think.) One of the big things about tires I think
is that at very low load (nothing but tire and rim), even a skinny street tire
may have a 1.7 coefficient, but it drops a little with every kg you add.
Todd Wasson
---
Performance Simulations
Drag Racing and Top Speed Prediction
Software
http://PerformanceSimulations.Com
> > well, i ve a little knowlege of Italien and i think the correct word is not
> > "Et" but "i" that mean "And" in Italian.
> > "I tu brutus" ...
> > but maybe we had to continue ond rec.latin.qutation ;o)))))
> I think the correct spelling is "Et tu, Brutu", actually. This is latin,
> not italian.
--
Gunnar
#31 SUCKS#015 Tupperware MC#002 DoD#0x1B DoDRT#003 DoD:CT#4,8 Kibo: 2
gnus don't kill people. gnus-summary-lower-score kills people.
today i accidently i tried something interesting:
i link wheel rotational velocity to the road velocity (allowing only a limited difference), thus allowing only a limited amount of slip. this lets me go through curves with a constant drift angle, without requiring much user input (counter-steering or throttle regulation), giving a very arcardy feel. i mention this because a few weeks ago there was a thread on how to create convining arcarde modes. this is something one might want to try out. of course there would have to be many things to take care of (donuts, oversteer not caused by engine power).
unfortunatly this gets me nowhere closer to a solution to my original problem. once i begin counter-steering (to get a wider curve radius) the car turns around into the other direction and gets out of control ;-(
jonas
-Gregor
> unfortunatly this gets me nowhere closer to a solution to my original problem. once i begin counter-steering (to get a wider curve radius) the car turns around into the other direction and gets out of control ;-(
> jonas
