Fatigue of stabilised steel in NPP primary piping - discussion on design curves

The ASME Code Section III fatigue design methodology is based on local strain approach and transferability of material performance measured using small scale laboratory samples. Strain controlled tests data shall be transferred to fatigue assessment in a compatible way to provide relevant prediction and follow-up of fatigue usage. Transferability to real components is secured by appropriate margins and full scale testing. Transferability can be studied and improved through better simulation of component operation conditions such as loading mode and sequence, temperature and environment. But the design curves shall be based only on standard tests compatible with the design procedure. A new design curve for stainless steels has been adopted into the ASME Code Section III. However, it is suspected that derivation of the curve is not fully compatible with the design procedure. Variable temperature data has been mixed inappropriately. Furthermore, the code curve is not applicable to all grades of stainless steel. This paper reports contradictory data for stabilised austenitic stainless steels extending up to 10 million cycles. Niobium and titanium stabilised stainless steel specimens were sampled from 100% relevant material batches fabricated to be used in NPP primary piping. Fatigue tests periodically interrupted for holds indicated time and temperature dependent hardening during holds at 25°C to 325°C. Notable extension of fatigue life was measured when loading patterns consist of cyclic deformation in lower temperatures than hold annealing. Many NPP piping thermal transients separated by normal operation belong to this category and fatigue assessment based on standard fatigue data seems to underestimate fatigue endurance.