News on www.pragtic.com

Von Mises stresses and strains in fatigue analyses  

by jan
Fatigue Calculations - Dec 19, 2008

Me and my colleagues repeatedly asked ourselves, whether the choice of the signing of von Mises stress (by the sign of the first tensor invariant) was the right one. This option often leads to very conservative results, well, conservative in such a scale that I am doubting their quality. I implemented the new signing method to betaJ version of PragTic - the sign can be now derived also from the sign of the principal stress/strain with biggest absolute value. For some cases even the removal of any sign worked. Because all this harassment is not based on any detailed comparison of different available methods, I would like to ask all users - if you know about any better solution or some comparison of different solutions, please, let me know.
The pressure vessels standards often use vM stress from the upper value of the cycle and compute the range of vM stress from ranges of individual components. This leads to the amplitude then and the mean stress is generated from upper vM value subtracted by the amplitude. It is a nice method, which does not necessitae any signing, but alas, what will be the equivalent stress used for decomposition of the cycles processed by this way? Even though I could try some solutions, I need somebody who would like to test the various combinations on experimental data to get some understanding. Is not there somebody willing to help?

papuga@pragtic.com