bod312
Mechanical
- Jul 21, 2009
- 1
I made a model of a cylindrical shaped device. It is symmetrical and contains slots, sleeves, bands and a cap. I am trying to run Cosmos on the model for static loading case and buckling scenario. The mesh appears to be great except for 1 thing. On the top of the model where the cap is located, there are 3 areas equally spaced, about 120° apart, where the sizes of the elements shrink. These areas are narrow but extend from the edge where the cap meets the O.D. of the cylindrical shell inwards. The cap is sandwiched between 2 bands and these bands along with the shell for a small distance contain these “hot spots.” The elements are uniform except for these 3 “hot spots” which cause neighboring element size to differ greatly.
The static analysis shows that there are higher stress levels where these “hot spots” are located. I believe this could be due to slight error since the neighboring elements have large discrepancies in their size. There is no reason for these “hot spots” everything is symmetrical and there is no pattern that can be discerned from the geometry of the model.
If anyone has any idea of the cause or solution to this issue it would be greatly appreciated.
After running the buckling analysis using these meshes it appears that the analysis is diverging. After decreasing the element size by 5 mm the F.O.S. dropped 3 and the next decrease of 5 mm decreases the F.O.S. by 13. I am wondering does this mean that FEA cannot be performed on the assembly or if the mesh issues could be tainting the converging diverging results.
Any help on these issues would be appreciated.
The static analysis shows that there are higher stress levels where these “hot spots” are located. I believe this could be due to slight error since the neighboring elements have large discrepancies in their size. There is no reason for these “hot spots” everything is symmetrical and there is no pattern that can be discerned from the geometry of the model.
If anyone has any idea of the cause or solution to this issue it would be greatly appreciated.
After running the buckling analysis using these meshes it appears that the analysis is diverging. After decreasing the element size by 5 mm the F.O.S. dropped 3 and the next decrease of 5 mm decreases the F.O.S. by 13. I am wondering does this mean that FEA cannot be performed on the assembly or if the mesh issues could be tainting the converging diverging results.
Any help on these issues would be appreciated.