rspe
Mechanical
- Jan 30, 2006
- 1
I want to determine the stresses in a structure that is subjected to random vibration so that I can evaluate the life of that structure (up to crack initiation, i.e. not including crack propagation). I am using Abaqus version 6.5. I have set up and solved the random response analysis successfully, but I have found that the only stress outputs available to me are the RMS values of the stress components: RS11, RS22, RS33, RS12. If I have a 3-D state of stress, I don't see how these outputs provide me with useful information.
The fatigue methodology that I had planned to use was (1) calculate the von Mises stress PSD, (2) calculate RMS von Mises stress, (3) calculate the statistical frequency (i.e. zero stress crossing with positive slope), (4) convert the RMS stress to an equivalent damage, constant amplitude stress, and (5) compare this to the S-N curve. It appears that I can't even get through step (1) using Abaqus because it won't output von Mises stresses for a random response analysis.
I expect that I must be misunderstanding this process. Surely, Abaqus provides the output that I need to solve this problem. Can anyone help me understand how to get to my answer from what is available in Abaqus?
The fatigue methodology that I had planned to use was (1) calculate the von Mises stress PSD, (2) calculate RMS von Mises stress, (3) calculate the statistical frequency (i.e. zero stress crossing with positive slope), (4) convert the RMS stress to an equivalent damage, constant amplitude stress, and (5) compare this to the S-N curve. It appears that I can't even get through step (1) using Abaqus because it won't output von Mises stresses for a random response analysis.
I expect that I must be misunderstanding this process. Surely, Abaqus provides the output that I need to solve this problem. Can anyone help me understand how to get to my answer from what is available in Abaqus?