Pacejka formula vs surface type

[frca]

Hi, I'm looking at this article about the Pacejka's tire model: http://www.racer.nl/reference/pacejka.htm
It computes tire contact forces, but there is no mention about surface type. How to combine pacejka with the fact that it is different whether one rides on the asphalt and on the gravel?
I have a theory that I would compute lateral+longitudinal tire force without any surface information -> resulting horizontal tire contact force and then I would limit it according to the http://en.wikipedia.org/wiki/Circle_of_forces . What do you think? And how would I compute the radius of the circle for various surface types in case my assumption is right?
Thanks.

[FXCarl]

in Pacejka Formula:
y(x) = D sin[ C arctan { Bx - E ( Bx - arctan(Bx))}]

you can use friction factor to multiply D.

for example
D = (b1* Fz + b2) * Fz
=>
D = friction_factor * (b1 * Fz_squared + b2 * Fz);

[bone]

Hi, I'm looking at this article about the Pacejka's tire model: http://www.racer.nl/reference/pacejka.htm
It computes tire contact forces, but there is no mention about surface type. How to combine pacejka with the fact that it is different whether one rides on the asphalt and on the gravel?

The first question I might ask is: what is your desired level of accuracy in tire modelling?

Pacejka is designed to only work with one set of conditions at a time (surface type and/or dampness, tire pressure, wear, and temperature). You can certainly go with FXCarl's scaling solution to "simulate" changing conditions, and it will be fairly believable. For more accuracy, you would need to alter the shape of the slip responses based on the conditions (not an easy task).

I have a theory that I would compute lateral+longitudinal tire force without any surface information -> resulting horizontal tire contact force and then I would limit it according to the http://en.wikipedia.org/wiki/Circle_of_forces . What do you think? And how would I compute the radius of the circle for various surface types in case my assumption is right?
Thanks.

Again, this would be somewhat believable. Just FYI (and I just noticed this is mentioned near the bottom of the Racer site), there are quite a few versions of Pacejka's Magic Formula, the newest ones have even more parameters to handle combined slip situations.

But if you do just need a simple equation for the radius of the circle of forces would be:

force_radius = coefficient_of_friction * D, where:

the coefficient_of_friction (CoF) depends on the surface type and,
D is calculated as in FXCarl's post

There's a significant problem here, though, and maybe you've already noticed, but the D can (and should, in real life) be different between the lateral and longitudinal cases. If you're just going for a believable model, you could simply lock them to be the same (in other words, the b1 and b2 parameters need to be the same for both the lateral and longitudinal case ... technically what this is doing is making the load sensitivity curve the same).

Hope this helps.