Mesh convergence issue (Abaqus/Explicit)

Hi, I am looking for reasons why the curve of force do not converge with mesh refinement. (Each calculation converges!) I am doing a crush analysis of a multi-cell structure for finding peak load and energy absorption using Abaqus Explicit. My setup is two rigid shell planar plates and a shell extrusion part between them. Some details of my analysis: - Tie at both the ends surface to surface - Bottom plate reference point encastre and top plate rp was able to translate in only one direction. I set an initial velocity and a mass for the top plate. - General contact (all with self). Friction coefficient=0.2. - Get the displacement values from the top plate rp and reaction force values from bottom plate rp. When I refine the mesh the absorbed energy decays. I am already using a really small m
 esh size and the energy still does not converge.