Output the element load resultants (N's) and divide by the thickness. Actually I doubt that the element stress option output are "wrong"; its just that users don't understand what they mean or don't mean for composites.

Thanks @SW. I did not have this confusion before, however, to double check I read about the NOCOMPS card that is used to output the stress for an equivalent homogenous section. The item in the list that bothered me was that "when Z0 is not defined as -0.5 times the thickness of the laminate, equivalent stresses output are incorrect". Hence the confusion. If you could explain in short (if that is a possibility!) how to interpret the result, it would be of much help to me (and maybe more people).

Thanks again!

That sounds like an issue with how the bending stresses are calculated (inner surface versus outer surface). A z0=-0.5 would be at the midplane and anything else might have a neutral axis shift. This would not affect in-plane results, but may affect the bending results.

You can probably run some simple test cases. I would try it with and without bending loads and at two different z0 values.

Brian
www.espcomposites.com