I am simulating a simple cantilever/frame consisting of two pieces of 60" long, 2" x 1" x 13 gauge rectangular tubing, and a 3/16" gusset. There is a 50lbf load at the end of one beam, the other is fixed, they are bonded with 45 degree miters. I've made 3 models: 1x with beams for the tube and shells for the gusset, and 2x with solids for the tube and shells for the gusset.
In the beam model, I've chosen one face of the gusset and used "define shell from face", then created an offset. The von mises stress on the "top" and "bottom" shell surfaces of the gusset are approximately equal, as they should be.
In the 2nd model, I've defined the gusset shells in exactly the same way, but modeled the tubes as solids as opposed to beams. Despite symmetry, the von mises stress results show asymmetric stress on the shell faces. I have tried redefining them over and over, the offset is correct, and I still get asymmetric stress, as if the gusset were not correctly centered (this is not the case, it's the same model as above).
In the 3rd model, I used a surface version of the gusset, mated it to the center-plane of the tube and created shells directly from this. The stress is again correctly symmetrical.
Can anyone explain why I might be having this asymmetrical result using the offset? Thanks.