Reference is made to the problem discussed in more detail in Case CS0209634.

Basically, a bug has been noticed in the graphic post processor, resulting in a misleading representation of the stresses in quadrilateral plate elements. In a symmetric FE model, i.e. with symmetric geometry, loads and restraints, this leads to an unsymmetric stress color distribution, which is nonsense. The stresses calculated by SACS are correct, the problem is only in the way the colors are assigned by the post processor. The issue can be removed by considering the stresses not only at the corners but also in the middle of the plate, which however would take time to be implemented. Another way, apparently simpler, is the one proposed below.

The graphic processor considers the quadrilateral plate as divided in two triangles, which is fine, but it always slashes the plate in one direction, e.g. top left to bottom right. This arbitrary choice eventually generates the unsymmetric result. The problem can be eliminated just considering the other slashing direction as well, i.e. the slash from top right to bottom left, and then taking the average of the two results.