Forums

Full Version: ring2007 and BumpAmplitude
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Ever wondered why the cars are sinking into ground and behave violently at places like Schwalbenschwanz. This is not a bug. It is due to BumpAmplitude settings for the track surfaces.

Code:
[surface-03]
Type = asphalt
BumpWaveLength = 2.0
BumpAmplitude = 0.25
FrictionNonTread = 0.8
FrictionTread = 0.9
RollResistanceCoefficient = 2.5
RollingDrag = 0.0

A bump amplitude of 0.25m is way to much for the cars we've got, not to speak of F1-02.
Another thing I've noticed are the BumpWavelengths up to 10 meters. Such surface deviations can be modeled by track geometry / bezier patches.

I think the surface bump amplitude, wavelength should be used to simulate high frequency details of the smooth surface with amplitudes somewhere between 0 - 0.05 and wavelengts of up to 0.5 meters.
NaN Wrote:A bump amplitude of 0.25m is way to much for the cars we've got, not to speak of F1-02.

that was a typo (it should have been 0.025) and it's fixed in svn. thanks for catching it.

--alex--
Looks like the simulation has some problems with the curbs. Will look into it.
We have problems with wheels approaching inclined surfaces at hight speeds and shallow angles. Due to normal force and camber relative to the curb surface, tire friction gets big enough to flip the car into air. I am not sure how to fix this. Limiting surface normals(camber)? Are there actually limits for camber in Pacejka?

A hacky workaround would be, to set friction for curbs to zero.