Really high stresses in thermal analysis

[dktoao]

Hi all. I have just started doing FEA for our company and the main structural engineer is out today. So I have a somewhat simple question.

I am running a thermal analysis on a part that is constrained such that intuitively, stresses caused by thermal expansion should be minimal. However I am seeing a factor of safety of about 1/2.

I made sure that the thermal expansion on all rigid elements is the same as the materials that it is attached to. My units are consistent, and looking at the displacements they seem reasonable to me. I think it may be an issue with the units since I am inputting a temperature and length units into the analysis but not force units (although I suppose maybe the mass of the material factors into this?). Does anyone have any experience with something else that might be causing this?

-Thanks

 

[corus]

How can a rigid element have thermal expansion?
If you restrain/constrain something then stresses from thermal expansion aren't necessarily minimumm, but probably more significant.
With thermal loads you can do rough calculations based upon differential thermal expansion that will give you an idea of what stresses to expect, even if it's just E.a.dT. Try that to see if your units are wrong. I think it's more to do with the way you've set the model up though.

Tata

 

[dktoao]

Sorry, I guess I should have specified, technically I am not using rigid elements, just very stiff elements with the same thermal expansion as the material, it is an automatic feature of the software that I am using (FEMAP). I will try the hand calc though, thanks for the suggestion.

 

[40818]

If your running nastran, there is a lagrangian option for rigid elements, which i believe you will need to switch on.