Contact Elements: Accuracy? 2

Hi,
well, yu're asking the impossible, I fear, since CW has only very marginal contact controls. I suppose it is using internal algorythms in order to set proper compenetration tolerance, pinball factor, normal rigidity factor, tangent stiffness factor and so on.
I would consider CWcontact results accurate as far as no high stress gradients are involved (I would be a bit suspicious in cases where surface/surface contact evolves towards edge/surface or even corner/surface contact because of the components' deformation: in these cases, at least the contact stiffness should be updated between timesteps, with consequent non-linear analysis, which is not the case with CW where the contact stiffness is calculated "at start" and then never updated).

Regards