CLARIFICATION - MECHANICA Buckling - Lanczos or Subspace Iteration?
A few months ago I posted a request relating to MECHANICA buckling. A Nastran model had found a lower mode than MECHANICA and switching to Lanczos in MECHANICA made no difference. Nastran was still finding a lower mode than MECHANICA (Nastran was using Lanczos as well). Although we have completed our engagement at Boeing, we did not get a chance to complete our investigation on this issue. I would like to clarify what was going on so that I leave no wrong impressions of MECHANICA. Although incomplete, I will explain what I do know.
The model is planar and we were using solids for a thin plate (doubler). I later learned that the Nastran model was a shell model. All loads were in-plane. The Nastran model was identical to the MECHANICA model, the only difference being that it was a shell model.
We were using out-of-plane edge constraints on the part boundary of the MECHANICA solid model, not surface/face constraints. Only one edge (curve) was being constrained - either the front or rear edge.
I was told by one of the leads at Boeing (after I had left) that he believed that Nastran and MECHANICA would agree if on our solid MECHANICA model a curve located at the center of the part's edge, were constrained, rather than the front or rear edge. Unfortunately, I did not get a chance to confirm this and I no longer have access to Nastran. He also said that a solid model in Nastran, with only the front or rear edge constrained out-of-plane, would most likely not find the lower mode.
Apologies to John if my post caused a stir.
Randy Speed, Principal Speed Consulting, LLC 2871 Howard Road Waxahachie, TX 75165 (972) 938-0490 ph (972) 937-2319 fax www.speedconsulting.com