mikycampo
Mechanical
- Jan 3, 2007
- 17
Dear all,
i'm reading the new ASME 2007 sec VIII div 2 code and i'm facing some problems with the elastic-plastic stress analysis.
In particular, chapter 5.5.4 describes two method to perform fatigue calculation in an elastic-plastic stress analysis: the "cycle-by-cycle analysis" and the "twice yield method".
Second one is enough clear and similar to the "strain life" approach described in many books, that uses the ramberg-osgood curves for taking into account plasticity, performing a linear analisys.
First method is not so clear. It describes that with this method the "cyclic plasticity algorithm with kinematic hardening" shall be used, after some lines it prescribes to use Annex 3.D (par 3.4.D) where the ramberg osgood curve is described. It seems that the two methods use the same material model! Which are the differences between the two methods?
What I've understood is that the"cycle-by-cycle" analysis must be performed by means of a FE model with a elasto-plastic material model (i.e. bilinear with kin hardening) and I've to perform a load/unload cycle for each bins in the rainflow matrix. Is it correct? or I've to use a more cpmlex unload curve as in 3.D.14 formula?
Someone knows about examples which compare the two methods on a real application?
Thanks a lot,
michele
i'm reading the new ASME 2007 sec VIII div 2 code and i'm facing some problems with the elastic-plastic stress analysis.
In particular, chapter 5.5.4 describes two method to perform fatigue calculation in an elastic-plastic stress analysis: the "cycle-by-cycle analysis" and the "twice yield method".
Second one is enough clear and similar to the "strain life" approach described in many books, that uses the ramberg-osgood curves for taking into account plasticity, performing a linear analisys.
First method is not so clear. It describes that with this method the "cyclic plasticity algorithm with kinematic hardening" shall be used, after some lines it prescribes to use Annex 3.D (par 3.4.D) where the ramberg osgood curve is described. It seems that the two methods use the same material model! Which are the differences between the two methods?
What I've understood is that the"cycle-by-cycle" analysis must be performed by means of a FE model with a elasto-plastic material model (i.e. bilinear with kin hardening) and I've to perform a load/unload cycle for each bins in the rainflow matrix. Is it correct? or I've to use a more cpmlex unload curve as in 3.D.14 formula?
Someone knows about examples which compare the two methods on a real application?
Thanks a lot,
michele