I am trying to model the flattening of a spherical asperity tip by a rigid flat surface. I know that plastic deformation initiates under the surface, according to Hertz theory. Then it propagates to the surface and continues till the whole surface is plastically deformed. When I look in distribution of PEEQ on the surface of asperity, it seems that right on the surface there is a big gradient of PEEQ. Plastic strain on the surface is quite small but within a very small depth (one layer of elements) it grows rapidly. Is it a numerical / meshing problem or something normal?

I have used CAX4R element (Axisymmetric 4-node element) in Abaqus software. Material is elastic perfect plastic. Surface to surface contact algorithm has been used.