Discussions on Effects of Temperature in Fatigue Assessment of Stainless NPP Primary Piping

Abstract

Fatigue performance of Nuclear Power Plant (NPP) primary circuit components and laboratory specimens depends on temperature. Temperature plays a particular role in environmental fatigue, but affects also generic fatigue evaluations. Historically, fatigue data was generated in air at room temperature using strain-control testing. For the purpose of engineering calculations, the strain was then transformed into “stress intensity”. The elastic-plastic strains were scaled to pseudo-stress units (psi or MPa) via the use of an elastic modulus. To make a single curve useable over a range of temperatures an adjusting factor was implemented in the codified stress analysis. The ratio of elastic modulus in the design and room temperatures was considered applicable to adjust the stress intensity (total strain) used in fatigue assessment. Later on, temperature dependent factors were proposed for environmental effects {Fen = Nf(RT,air) / Nf(T,environment)}. Also environment independent temperature effects are covered by Fen factors, which are used to multiply fatigue usage. This means that temperature is supposed to be accounted for twice in fatigue assessments: first for stress intensity, then for allowable cycles. Moreover, fatigue data in high temperatures have been included in the data set behind the current design curve for stainless steels. Incompatible corrections and code updates can lead to some over-conservatism when selecting adequate values for the elastic modulus and Fen for the fatigue calculations. Given the construction of the design curve as an integral part of the original codified assessment procedure, new developments in numerical stress analysis and experiments together with aim to perform calculations as consistent as possible with the physics at hand, a proposal is made in this paper to define a method to evaluate fatigue using strain amplitude rather than stress intensity amplitude. A concern on double counting of temperature effects and inaccuracies in fatigue assessment was raised by the current authors [1] [2], but this issue has not yet been studied and discussed in depth. This paper will discuss effects of temperature in fatigue experiments and assessment in terms of transferability of laboratory data for fatigue assessment. Among others, a statistical study based on laboratory test results in Pressurized Water Reactor (PWR) environment will help identify possible gaps between the recorded fatigue lives and the values provided by the codified rules. The aim is to quantitatively analyze the ranges of inaccuracies and unknowns affecting fatigue assessment of NPP components in various operational temperatures.