When interpreting results from FEMAP, how do you determine if a part is safe under applied load. What if a few elements or a single node in each element is over the yield limit of the material by far? Please see the attached picture. Node 1701 reads 277 kpsi stress. Is this a singularity point? Average solid von mises as you can see from the picture is 100889 psi. The yield of the material is 138 kpsi.
How do I conclude my findings? Should I be worried about any failure just because of a single node with un –reasonable high stress?
Any help is very much appreciated.
Solved! Go to Solution.
The nodal stress in a sharp corner doesn't reflect a realistic stress level. Normally classification societies or industry standards have stress limits that are related to the stressed area, so I would start with finding the appropriate standard to use for your design. Even if the material isn't yielding there might still be fatigue issues so you should check that too.
From the picture it is the typical issue in FEM. I guess it is a welded location:
- from strenght point of view, normally if the surrounding are under yield, the stresses can be redistributed and no risk of failure exists.
- from fatigue point of view, it could be a problem.
Anyway for FEM this kind of stresses are singularties and the stress values are not "real" there. It is a geometrical abrupt change, stress concentration. If you remesh the stresses tend to infinite at this location being the surrounding are at the same stress level
In fact, for fatigue, the guidelines recommend not to read out the stress values at these locations but at some nodes far way from this (hot spot method for example). As I said , i guess the location represent a weld.
A follow up question.
Can I fix this by adding a very small radius to sharp corners? In reality the parts do have a radius (even if it's spec'd as sharp). Would this help to eliminate the singularity stress nodes?
It is not a welded location. But yes it’s a location with abrupt geometrical change. I tried to attach the .modfem file to better explain the issue I am having. But it seems like those extensions are not accepted.
Thank you for the help.
If you have a corner without raduis, it represents a singulaty by FEM. The stresses are
not real, like when you applied a nodal force or fixed displacemnets. This corner shape (90º) cause changes in stress directions. As you introduced the radius the singularity dissapears, and the stresses normally tends to decrease as the radius increases. ( I said normally beacause I do not know the loading etc...). This is the reason that once the piece is finished , all corner shapes are rounded with some radius.