I am performing a static structural analysis on a cynindrical roller bearing. The contact stress and Von - Mises stress on the roller surface needs to be aquired. The mesh on the contact region is quite dense (0,06 mm - element size). The problem is solved in 50 substeps.
When checking the results, both on Von Mises and Contact Pressure they seem sensible. The contact pressure is maximized along the contact line and gets smaller away from it, as expected considering Hertz theory. However, there are sudden, rapid increments of the pressure magnitude on very small regions of the contact area that do not seem realistic and make me think that there is something wrong with the model.
Your mesh along the horizontal line where pressure is 4025.5 have very bad ratios. Moreover these elements seem to be disconnected as there are elopements with five sides ? try to define a rectangular area where you expect model to contact. There then break the lines in subdivision to impose a nice even mesh with rectangles of nice ratios. Then let ansys mesh the rest at the boundaries. See if this works.
Hello Laodamas,
ERESX,NO command may help you to resolve the stress peak issue. Please see more information at the below links:
https://forum.ansys.com/discussion/2294/the-equivalent-stress-over-ultimate
https://forum.ansys.com/discussion/1374/stress-singularity-due-to-interference-fit#comment-550243a4-9ffd-4d41-bd53-a93100d77512
- Badges
- Science topic