Study on stress-based fatigue monitoring methodology for Class 1 nuclear pipe based on the RCC-M code

Abstract

On-line fatigue monitoring technology plays a key role in the LTO of nuclear power plants, for which the main challenges are: how to gain the thermal shock stress quickly and reliably, how to order the random stress history, and how to do the code-coforming fatigue assessment (eg. the RCC-M Code and ASME Code). Compared to the calculation process of FEM, the numerical technique based on the Green’s function is efficient and reliable. The stress due to a linear radial temperature distribution which was previously classified as a secondary stress is re-classified as a peak stress in the ASME Code while this requirement is not supported in the mainstream FEM software (such as ANSYS and ABAQUS). Traditionally, detection of peaks and valleys in the stress history is simply based on the change on sign of the slope of the major stress component (1-D rain flow) which is not strict compliance with the requirement in the codes. Therefore, a refined strategy for detecting peaks and valleys needs to be developed. In this paper, the Green’s function is used to calculate the thermal shock stress for the TECMAN on-line fatigue monitoring software which is developed by the author's team, and the method to separate the stress due to the linear radial temperature distribution and the 3-D rain flow technology for the random stress history with a multiaxial stress state are discussed. In the study case of a prototypical nozzle under rapid thermal shock, the results show that the simple integration method of the Green's function gives identical results as the conventional FEM, and the maximum axial stress caused by the radial linear temperature distribution under the reference transient accounts for 7.03% of the design specification limit 3Sm. Compared to the design requirements in the ASME Code, the criteria in the RCC-M Code is unnecessarily conservative. When the extreme values of the stress components are obtained almost at the same time, the nonconservative for the 1-D rain flow is limited, while the nonconservative cannot be negligible under a multiaxial stress state.