Questionable frame deformation

Question asked by Jason Smith on Nov 1, 2016

I am trying to simulate a relatively simple frame under load. The model consists of beam elements with 3 different cross-sections; the legs are 2” x 1” x 0.095” rectangular tube, the upper and lower rectangular frames consist of 2” x 1.5” x 1/8” angle plug welded to 2” x 1” x 0.095” tube, and the cross members 2” x 1” x 0.095” rect tube with a 1/8” strip plug welded across the top. The load currently being simulated is a distributed 500lbf total vertical down load at each of the cross members, along with gravity. It’s a static study, material is plain carbon steel.

I’m trying to go through and carefully verify to a reasonable degree all of the simulation calculations with hand calcs. However, looking at the exaggerated deformed shape, I’m not seeing symmetry despite the load and problem geometry being symmetric. When looking at the frame from the side as shown in the attached pic, the lower part swings out to the left, whereas the upper part swings out to the right. Since the beam profiles oppose each other, the neutral axes of the upper and lower side beams are such that the beams should want to bend away from each other, inducing an axial load in the cross members. Instead, I'm getting a near zero axial stress in those members, and both lower and upper rectangular frames bending as if their neutral axes were oriented in the same plane.

I've tried taking the advice from another similar post, where it was recommended to separate each structural member into a separate group. Despite doing this, both simulations give the same results.  Can anyone explain what is going on here?