Abstract
Design codes and standards are used for new designs and also for management of operation and in service inspection of existing NPP primary circuit pressure boundaries. The current codes — ASME, KTA, RCC-M, PNAE-G and JSME — have been originally rooted to the ASME Boiler and Pressure Vessel Code, Section III, Nuclear Vessels, which was published in 1963. Article 4, N-415 “Analysis for cyclic operation” instructed calculation of stress intensities for fatigue transients and provided two design curves for basic material types — one for ferritic, the other for austenitic steels.
The design curves were based on strain-controlled LCF tests, which measured the allowable numbers of cycles as function of plastic strain. The obtained material performance was then processed to strain-life and design curves (Sa = Ex εa). This local strain fatigue approach was found applicable for both ferritic and austenitic steels though their elastic plastic cyclic responses are much different. Fatigue data for stainless steels extended to ¼ million cycles and the design curve was extrapolated to one million cycles. Later on — supported by load-controlled HCF tests — the original LCF approach has been extended even beyond HCF to VHCF regime.
Our strain-controlled HCF results for alloy types 347 and 304L are in conflict with the reference mean curve behind current ASME curve for stainless steels. We assume that this reflects a generic issue related to extrapolation of the LCF methodology by Coffin, Langer and other pioneers. Furthermore, analysis of cyclic responses and variable amplitude testing to millions of cycles give reasons to assume that the concept of an endurance limit (Se) is applicable also for variable amplitude straining. Variable amplitude HCF straining was not studied for ferritic steels and we propose the concept of enforced endurance limit to austenitic stainless steels only.
We propose a critical review on relevance of the current ASME III design curve for stainless steels.
