Stress Higher Than Yield?

Whag

I am running a non-linear CCX solution and setting the yield stress equal to 200MPa (assuming perfectly plastic model). The solution gives me stresses which are higher than yield stress. I understood this could be an error due to the translation of calculation point solution to nodal points. Can anyone advise how to solve this issue? Can we get the solution at the integration points instead?

Thanks

Whag

I am using shell elements by the way.

disla

I understood this could be an error due to the translation of calculation point solution to nodal points.Can anyone advise how to solve this issue?

NLMAterial better with solid elements. Check if refinement improves solution. Take a look at this example. Review material properties definition.

https://mecway.com/forum/discussion/1190/prof-martin-kraska-tensile-test-of-a-wire

You can get some variables at the integration point if you request ccx to write them on a file. (With custom cards). Check ccx manual 2.19 7.51 *EL PRINT
If you work with solid element you can also use the interpolation tool.

Victor

A workaround to get integration points closer to the surface of shells might be to use laminate material with two thing layers at the surface since each layer has its own integration points. I've heard of this but not tried it myself.

Whag

Thanks All, will try some options and let you know whether it was succesfull!

disla

@Victor

¿Is that possible?

I discarded that option for some days ago according to this post: https://mecway.com/forum/discussion/comment/6572#Comment_6572

¿Do you mean that if we make the effort to write the custom material properties and Plastic cards by hand MECWAY would be able to represent the solution properly?
If yes, I still don’t see the benefit compared with a solid element with 3 layers per thickness.

Victor

I'm not sure if it works or not. But you also pointed out in that thread that plastic materials get 5 integration points through the thickness so it looks like some are already closer to the surface than the usual 2 for elastic materials.

I agree solids are just better all round if you can mesh them. Sometimes the geometry is too complex to do it easily though.