Welded contact and non-zero SPC Vector Resultant

[ksaw]

I'm running a NEiNASTRAN linear static solution and FEMAP for pre/post.
I have a model of an aluminum assembly (2 different alloys) and I'd eventually like to know the precise deformed shape with a uniform temperature load for further optical analysis. For troubleshooting purposes, I have set CTE to the same values for the two materials. I have symmetric welded connectors defined between the mating regions of the components. The only load applied is the uniform temperature. The only constraint applied is the fixed base node. The base node is attached to one of the aluminum components in 4 areas with an RBE2 with thermal expansion activated for the rigid element (using same CTE). I initialized the temperature throughout the model with a default temp body load, which is applied in the initial conditions field. The displacements look reasonable, but not the stress. There should be no (or very low) stress throughout the model but there are seemingly random pockets and areas of stress up to as high as 50MPa. All of the highest stress is near the contact regions. In addition, I am getting a large SPC vector resultant >200N and a maximum applied loads that are on the order of 4.e8N. I get the same results with AUTOSPC on or off. Any ideas for why I get might see this lack of equilibrium and odd stress results?

 

[Tuw]

Hi ksaw,

Sorry mate I am not a NeiNastran user, but just trying to help.
When you said the highest stress is near the contact regions, does the definitions there (connectors, mesh) have the same thermal expansion rate?
If you are running linear static analysis, for diagnosis purpose, maybe you can modify/remove the contact area, by uniting the 2 bodies into 1. I believe this would remove the high stress concentration.

Regards,
Tuw

 

[ksaw]

Thank you Tuw.
I checked the input file directly and the thermal expansion rate is the same for the parts in question. The only parts with different expansion coefficient are structurally isolated by CBUSHs. The final goal is that the materials will be different, so I'd like to keep the welded contact. I just assigned the same CTE for troubleshooting because I saw incorrect SPC resultant.
-Kent