Concentrate stress issue

[pelillom]

I have a very simple model of a bifricated pin. I am applying a known deflection(in this case .203mm) to the nose of the pin, and mechanica is saying that the pin is yielding. But when I look at the stress on the part, the high stress in right where the deflections are placed and not where I would expect them.

Is there a way to setup the model to eliminate the stress concentrations? I can do a manual query, but really wanted to try the optimization routine.

View attachment 3148

 

[ankarl]

Seems like you have constrained the same surfs, or very close, as were your deflection are. That will give you high stress concentrations in that area.


Try splitting the part in half and use symmetry constraints, that will help you getting better results.

 

[pelillom]

The deflection is tied to the top edge, not the surface. As the nose of the pin collapses during installation in to a hole, the nose does have some rotation to it. This rotation causes the deflection to be different at the surface top and bottom.

Here is the output of the model with symmetry. All of the model except that edge is reporting little stress.View attachment 3155

 

[Kaz Z06]

Clearly the stress on the edge is artificial since the BC is applied only to this edge - yet contact doesn't occur over the entire edge. There are two things you can do.Before anything else,you can cut the model one more time and take advantage of 1/4 symmetry in the other plane. Then you can eithercreate a contact model that will provide a much more accurate, and realistic, model of the real-world situation, or simply ignore the stress on the BC edge and evaluate the stresses in the area you know is being stressed most heavily (near root of the split). It is very easy to explain away any "red" color on this edge in a report or presentation - besides, the stress on this edge is predominately compressive anyway! Engineers must always check max and min principle stresses and not just von Mises as von Mises does not indicate compression or tension states!


Your approach should be chosen based on what your goal is - are you performing a "post-mortem" analysis to a failed part that broke in the root, or a design exercise? A post-mortem model would be what you have done while ignoring the high stress on the BC edge. A design problem should include a contact model as there may be a galling issue at your constraint edge based on material properties and actual edge geometry (broken, radiused, etc).


Hope this helps
Edited by: Kaz Z06