the finer mesh, the higher stress?

[gmarco]

Hi, I'm new to this forum, so be patient with me. I've got some problems with costraints. I took a cylindrical beam subjected to bending stress and fully constrained it in a portion of its lateral surface, as if to simulate a bearing.
As the mesh becomes finer, the Von Mises stress f.i. is higher and higher, without any possibility of convergence.
Can somenone help me?

 

[lawson]

This explains it better and in more detail than I can - check out the plate in tension part of the article.

http://www.machinedesign.com/asp/vi...=&strSponsorerUrl=&strSecondPath=&myCatID=361

 

[gmarco]

The article was very very interesting and useful. It seems clear that I have to change the model, because the stress analysis in the area of interest is unreliable. Do you know how can I simulate a 3D hinge constrained by a bearing with a reasonable computational effort?

 

[maxPE]

What you have simulated is a sharp corner. True that the finer the mesh the higher the load/stress to infinite. I suggest placing contact surfaces at the bearing location with a radius on the contact edge otherwise you have a discontinuity.

Max