I modeled a structure using FEMAP and run the NASTRAN and it gives me a high value of epsilon (E+14). There's no singularities, nor free edge. What else could cause this problem?
If you run a modal analysis make sure you have the correct density in mass units. Then see if there are any near zero modes. I would expect the frequencies to be above 5Hz for most typical structures. Regardless, the mode shapes will tell all. You may want to exagerate the deformation past the 10% FEMAP has. Look for any parts moving as a rigid body. When epsilon is that high it means there are large, unrealistic displacements as well.