Pacejka Simulation

[Boston]

I'm trying to write a straight-line acceleration simulation using the Pacejka tire data provided by the TTC. As advised by you guys, I got Tire and Vehicle Dynamics by Pacejka. I understand Pacejka takes normal load, slip ratio and camber angle and outputs longitudional thrust. I've read a good portion of the book, but I still have one big question: How do I relate engine torque to slip ratio for the entire curve? I know for SMALL slip ratios, slip ratio = wheel thrust/longitudional stiffness, but this isnt sufficient for the entire Pacejka curve. Can somebody please spoon feed me just a little bit?

[Jersey Tom]

Take a look at what thrust value where the longitudinal Pacejka model peaks (with Fz taking account longitudinal weight x-fer). If you were getting that much thrust at both wheels, how many g's would you be accelerating at? Do you think youre gonna get up near that peak? Or must of the time are you gonna be in the linear range?

[Boston]

Most of the time I expect we would be in the linear range. In the accel event, we slip a lot off the line and a little bit durring shifts. I want my program to optimize shifting and gear ratio for the accel event.

[Garlic]

This can be very complicated on initial launch. I would recommend just using a fixed RPM for initial launch, until a certain speed. You can get your slip ratio from this.

This isn't too unlike real life, as many cars have fixed RPM launch control.

Once you get going the proper way to do it is to take the power to the wheels, and calculate the slip ratio based on a standard COF (best gotten from testing) and the inertia of the drivetrain.

If you try to do this from a stop you will have a hard time controlling the sim and realize the benefit of my first comment. http://fsae.com/groupee_common/emoticons/icon_smile.gif

[Boston]

Thank you Garlic, that sounds like a good solution.

[JKap]

I've got a question that I was hoping somebody could help me out with because I can't seem to find the answer anywhere else. I'm trying to gain an understanding of the magic formula tire model for the first time by looking at the TTC data and I'm not entirely sure how the "user scaling factors" are supposed to be used. As I understand it, the entire set of Pacejka coefficients are determined from the given set of tire data with the assumption that all of the scaling factors are equal to 1 (or 0 in the case of the muV factor). How and why would a user change these values? I understand that some of their values are modified to account for environmental differences such as road conditions and weather, and physical suspension characteristics, but wouldn't you need more data describing the function of the tire under these conditions in order to change the scaling factors appropriately? In that case, couldn't you just calculate a separate set of coefficients? Or are these scaling factors used to account for inherent flaws in the magic formula, such as the shapes of the curves changing dramatically in certain conditions, which would make it impossible to create a proper fitting set of coefficients? Any help would be appreciated. Thanks.

[Jersey Tom]

You would need more data, yes. I leave everything at 1.

[Boston]

I found I need to change the Road Surface Coefficient in the MRA tire data from 1 to between .7 and .8 in order to get reasonable curves. The Pacejka formula has a similar coefficient that should probably be changed.

[ben]

Originally posted by JKap:
I'm not entirely sure how the "user scaling factors" are supposed to be used.

Ahh, the fudge factors http://fsae.com/groupee_common/emoticons/icon_wink.gif

Ben

[SR-Mike]

Depending on what you are trying to do with the simulation, i would recommend that for a simple longitudinal acceleration simulation that you do not require a model anywhere near as detailed as the Pajeka model. It will save lots of time and hassle. And do remember that factors like temperature which if i remember right is not taken into account in the Pajeka model will affect real results significantly.

Enjoy your reading Boston, its riveting stuff!

Mike