Advice needed
Hello,
I would like to have your expert opinion in one of the problems I am currently working on. Part of the simplified geometry is shown in the attached picture. As seen I have two cylindrical parts in contact with an interference fit of 0.001 inches. In addition to that a pin is installed and goes through (with a size to size connection) both cylindrical parts.
I decided to perform a quasi static non-linear analysis, firstly by solving the interference fit and pin connection load case and then continuing with the load on the structure as a second load case. When I solved the press fit problem (without the pins), I had no problems, both the displacements and the stresses looked reasonable. But when the pins are included into the solution I am having a band of peak stresses (well beyond the material's yield point) where the pin contacts both cylinders (see red oval in the attached pic). This happened with almost every type of contact connection I tried between the pins and the cylinders (freeze, slide, friction etc). Same results when I replaced the 3D pins with 1D elements with RBE2 connections (to connect them with the cylinders). One solution is to ignore the pins and continue with my solution and then perform a detail stress analysis of the pins based on my results, but before doing so, I would like to know that I have tried all possible scenarios with the pins' contact.
Any suggestions?
Thank you in advance.
Answers
-
I guess this has to be solved with continuous or finite sliding contact.Please refer attached deck from OptiStruct Example guide.
0 -
Thank you very much for the files attached. I'll have a look and let you know how it went.
Best regards
0 -
Hello again. Your files provided a great inside into the finite sliding contact, however I am not sure if I can apply these settings since we are using an old version of Hyperworks. Anyway, I managed to work around the problem by using 1d elements to represent the pins and link them to the remaining structure with RBE2 elements. The only change to my previous attempts was that I connected the 1d bolts with the upper part and the lower part of the cylinders, thus avoiding the area of their contact (where I had peak stresses). It worked, so I might think to proceed with that solution.
0