Simmechanics Simulator

[eric922]

I've been working on suspension models in SimMechanics to calculate kinematics and wheel load distribution. Not having a lot of experience with multibody modeling, I have had to try what seems to me like an indirect way of simulating the effect of cornering accelerations.

Is it reasonable to think that I can calculate the wheel loads simply by changing the roll and pitch angles since vertical loads would change with suspension travel? Taking moments about a point on the ground, the vertical load moments should balance the acceleration moment on the car CG right? I would think that at least one of the missing factors are the anti roll effects from jacking.

Simulating what I have I end up with what seems like a lot of load transfer for just about 1g of lateral acceleration. The other thing I see is that the calculated acceleration after inputting FZ, IA, and SA are way higher than the acceleration I chose for calculating corner loads. I suppose this could be due to the FY scaling for the tire model for which I used 2/3 as a start. Maybe I'm missing factors that I would get from testing.

Any input is welcome, whether it's a push to the proper thread or a slap on the wrist...

[BillCobb]

Construct several versions of your 'suspension' which have characteristics that you can 'know' the results for. They don't have to be practical suspension geometries, just ones for which the anti-roll forces and moments , tire properties and deflection characteristics play no part (One at a time. etc.) and or have unity gain geometry. Suspensions are just linkages, so collaps yours into primitives.

The idea is to produce solutions and post processing which correctly summarizes the expected simplified results. Then graduate to more complicated models if 2D and then to 3D.

Then construct a 'perfect' vehicle with 4 equal suspensions and a F/A symmetric sprung mass. Do steady state ride and roll analysis first and foremost. Then do lateral. BTW: you don't have to or need to do dynamic analysis with SimMechanics. Research the way to do just static analysis with your model(s). The Matlab Forum will have methods for this too.

[eric922]

Thanks for the advice. It's always refreshing to receive a helpful nudge. I've been finding it quite interesting how much effort can go towards developing a method for finding a solution.