Here's an email that I write to a customer almost every year, in one form or another. Usually, the customer asks about an exceptionaly high stress in a corner of his model, and the stress value is often much higher than that computed by another FEA programs. Until I get permission to post it, you'll have to imagine what the customer's result window fringe plot looks like.
This situation is fairly common, and is sometimes a source of confusion, so I hope my response below will help others in similar situations.
Dear Customer,
From the images in your e-mail, it appears that there are re-entrant corners in your model, which is leading to very high stresses. According to the theory of linear elasticity, the stress in that re-entrant corner is infinite. Thus any stress that is computed in that corner by any FEA program (including Creo Simulate, Abaqus, ANSYS, etc.) that is using linear elasticity theory is fictitious. In fact, for your problem, it appears that Creo Simulate did a better job of computing an infinitely high stress than FEA-program-X did.
Of course, in the real world, there is no such thing as an infinitely high stress, and linear elasticity is only an approximation of the true physics. Faced with your situation, you have a couple of options.
- If you are certain that the stresses in those areas are unimportant because they do not affect the integrity of your design, you can choose to ignore them. If you don't want your analysis stress measures (such as max_prin_mag) to be affected by these fictitious stresses, then you should create an isolate for exclusion AutoGEM control in that area.
- If you want to compute the true stresses in those areas, you should remove the re-entrant corner by modeling the fillet that actually exists in that area.
- If you really have a re-entrant corner, or if the fillet is too small to model, then you need to account for the plastic yielding that would occur in that corner by using an elasto-plastic material and performing a non-linear, large deformation analysis.
I hope that helps. Please let me know if you have any questions.
Tad Doxsee
PTC
