Past discussions have focussed on detecting penetration at the time of impact; in my case the surfaces are contacting and stationary at the start of simulation.
My Motion Study settings are:-
Wstiff (S12 fails with the usual message)
Error 1 E-4
Min step 1 E-10
Max step 1 E-4
Initial step 1 E-6
Jacobian at Maximum
Actual step is reported as 1 E-4
My model is a Floating rail wagon of 25,000 Kg mass on a Grounded rail track.
My objective is to satisfy myself and impress the customer with a video of the wagon rolling along the track.
I have not been able to persuade SW Premium 2009 SP3 to achieve this.
Penetration of wheel into rail seems to be a function of contact force and I ran a series of simulations to demonstrate this by varying only the value of Gravity.
In microgravity the wheels sit on top of the rails and the wagon rolls along the track.
In Moon gravity the wheels sink into the rails about 20 mm then reach equilibrium; the wagon rolls along the track like a car driving through soft sand.
In Earth gravity the wagon sinks rapidly out of sight as though on thin sea ice.
Whilst microgravity makes the kinematics look good, the dynamics are wrong.
I cannot see what else to control to achieve the desired result.