SPCD +right stress? 1

[airmail]

I have a model in PATRAN which is loaded using SPCD in border nodes.

The forced displacement method (as far as I know) move these nodes according to inputs and then, clamp the node. So, I have a feeling that stress in the elements are not true (maybe clamped nodes take load), and only strains results are reliable. Does anybody know more about "forced displacement method" and the right use of it?

 

[airmail]

Well, maybe there is confussion... I said "forced displacement method" when I wanted to say "enforced displacement method". sorry.

 

[johnhors]

Airmail

Stress and strain go together, they are either both reliable or both untrue, you can't split them apart.

Basically clamping nodes (with or without enforced displacement) produces unrealistic stresses in the neighbouring mesh, and cannot be believed. In the real world nothing is clamped, any object will have some degree of flexibility. A fully clamped condition is a mathematical curiousity only. Ideally, your model should be fully load balanced with minimal supports which do not react any load and are only there to remove rigid body motion.

There are special cases which involve fully fixing nodes, for example implementing planes of symmetry, enforcing displacements on a sub-model using results from a previous full model analysis.