Good question

Hello again Vigneswari -

Suppose you have a fully reversed fatigue load case. For example, you'd have a bar subjected to an axial set of forces: F=100 (time=1) and F=-100 (time=2). What you could do is create a finite element model in ABAQUS where you apply only F=1 (the direction doesn't matter you are going to reverse it anyway). Then in fe-safe you create a load block where you associate that dataset (you had created just an unit load case in ABAQUS) with some scale factores (100,-100). You don't need to create two STEPS in ABAQUS. You can, of course, if you want, but it is not necessary.

One word of caution: You don't compute the SN Curve in fe-safe or ABAQUS. The SN curve needs to be evaluated by running physical fatigue tests in a laboratory. The material needs to tell you how it behaves - you can't run a simulation and predict that. However, as you're learning, you can estimate the fatigue curves (SN or eN curve) if you roughly know some monotonic properties. Read the papers written by Professor Ali Fatemi.

Regards

Giovanni.

Thankyou sir for your response!

Thankyou Giovanni De Morais Teixeira if i using XFEM Can export a file inp to Fe-saf

I would add that a static structural analysis with F=1 works fine in so far as the system is linear and dynamic effects (for example, resonance) are negligible, otherwise the frequency of the input load may also have an effect on the output stress

Thanks Denis Benasciutti for the reminder. Just translating to the Abaqus' folks then, what Denis is saying is that if your loads are alternating at a frequency somehow closed to one of the resonant frequencies of your component, you have to adopt either *Steady State Dynamics or *Modal Dynamics instead. In case of local plasticity (or notch plasticity if you will) you can still get away with using unit loads and performing Neuber Correction (or Glinka if you prefer) as long as there isn't other significant non-linearity e.g. contact or geometric.