I am doing a thermo-mechanical coupling simulation for the rotor carrier of an axial flux machine which has the permanent magnets glued to it. I applied my magnetic forces on the magnets and a centrifugal force. I added a thermal load of 90C to the model but suddenly stress in my titanium rotor jumped from 175MPa to 860MPa which didn’t seem reasonable. Upon studying the results, I noticed that high thermal stresses were formed in the constraint location. I studied the mechanism once again and thought about it. Upon thermal loading the rotor expands radially and henceforth I released the radial cylindrical coordinates in the constraints. Now the problem is that Nastran wouldn’t run the model because it throws the “max pivot ratio” error which means my model is under constrained. I added the MAXRATIO card but still failed. I want the model to run with this specific constraint in order to reduce my thermal stress. Any suggestions?
This might be something worth opening a customer support case so a more detailed discussion specific to your problem could be undertaken. If the real world component truly undergoes strain-free radial expansion, and you can't come up with a set of statically determinate constraints that don't create SPCFORCES and local 'hot spots', then you might need to use Inertia Relief and use the INREL,-1 option with a SUPORT entry. You can find information about Inertia Relief in the Linear Static Analysis User Guide. Depending on if your structure has 'real' physical constraints for the other DOF it might be easier to use INREL,-2... but only if you truly have exactly 6 rigid body modes.
Another consideration would be to try turning on PARAM,BAILOUT,-1 to force the solution to go ahead and try to solve. This is generally discouraged as you can force a solution with believable but incorrect results. But the solution 'can' also be correct depending on the situation. At least it might give a solution and allow you to do some investigation.
Here is an example using PARAM,INREL,-2 in a thermal stress analysis using SOL 101. Please note that you also need to add density in your materials in order for this to work.
Here is a word file that shows that using the PARAM,INREL,-2 and if you set the minimum structure constraints that you can take a away the rigid body motion, the reactions forces should equals to zero, this means teh material is freely to expand.