Fatigue is a major degradation mechanism and life-limiting factor for primary circuit piping. High temperature, pressurized reactor coolant aggravates fatigue damage with a suitable combination of loading parameters. Nonstandard test methods, incompatible with design codes and the peculiar material behaviour of austenitic stainless steel have been widely used. This complicates quantification of the effect of water environment, commonly referred to as the Fen factor.
Four test series in simulated PWR coolant were completed over four years with stainless steel alloys 347 and 304L, hypothesizing that . Linear strain waveforms were used with non-constant strain rate to represent simplified plant transients and non-realistic mirrored strain waveforms for comparison purposes. Applying multiple strain rates allows identifying potentially non-damaging effects of effectively elastic strain near the valley of a strain cycle and on the other hand the damaging effect of effectively plastic strain.
Results generated within this project were used to draft a replacement to the Fen methodology presented in NRC report NUREG/CR-6909. Initially, this model remains based on total strain rate and is presently limited to the narrow parameter window in which valid experiments were performed. It does however reduce the scatter and unnecessarily high conservatism associated with the NUREG Fen by a factor of approximately two. This paper presents new results for 304L and an outline of the draft model proposal for Fen evaluation in PWR water.