Abaqus shell elements unreliable: does anyone have Adina?

If so, I'd appreciate your help. I have a suspicion the first-order shear deformation theory (FSDT) for shells in Abaqus is not implemented correctly. The shell element S4 is a "version" of the MITC4 element in Adina. 

I wonder if I could send you an .inp file to run in Adina? It is a square plate with [45/-45] layup, 10x10 elements (so a very small model)...

Thanks in advance.

Petr Krysl

pkrysl@ucsd.edu