Session: CS-01-02/03-01 Structural Integrity of Pressure Components & Environmental Fatigue Issues (Joint M&F, D&A)

Paper Number: 61507

Start Time: Tuesday, July 13, 2021, 09:00 AM

61507 - Material and Temperature Effects in Low and High Cycle EAF of Austenitic Stainless Steels 

Fatigue design of nuclear power plant pressure boundary components includes the use of design curves, where the allowable number of cycles is a function of the applied stress intensity. Differences between laboratory specimens and plant components are considered with the use of transferability factors, which are applied to the smooth specimen air best-fit curve. Whether or not the reactor coolant water environment makes up a part of the overall transferability factor has been debated since the 1970’s.

A concise database of strain-life low cycle fatigue data (mainly AISI 347) was used by Langer, Coffin et al. to construct the 1963 ASME III design curve for 18-8 stainless steels and Ni-Cr-Fe alloys up to 427 °C and applicable to 10⁶ cycles. In the 1970’s, multiple non-stabilized stainless steel fatigue results suggested the Code best-fit curve needed a critical re-evaluation for non-conservatism. In 2009, the ASME III stainless steel design curve was changed to reflect a similar database and applied up to 10¹¹ cycles with a single curve.

A closer evaluation of individual material heats reveals that the best-fit curve of Langer can in some cases be unconservative. On the contrary, the same detailed analysis repeated for the much larger databases used by Jaske and O’Donnell and later by Chopra et al. reveals that individual material heats rarely behave as predicted by the proposed global best-fit curve. Both conservative and unconservative cases are identified.

This paper continues the series of papers most recently published in PVP2020-21136. At VTT, heat specific strain-life data has been obtained. In this paper, using our data and literature data we demonstrate the significance these heat and temperature specific best-fit curves can have on the predicted number of cycles, when also considering the detrimental effect of environment. The example calculations show that in most cases, a simple change of the reference curve from the Code best-fit curve to a more realistic one adds considerably to the number of predicted cycles. Mainly for high cycle fatigue at high temperature, the mean curve may actually decrease. However, refinement of the F_en methodology is considered to provide significant relief to these cases so that the net predicted number of cycles may still be increased.

Presenting Author: Tommi Seppanen VTT Technical Research Centre of Finland Ltd.

Authors:

Tommi Seppanen VTT Technical Research Centre of Finland Ltd.
Jouni Alhainen VTT Technical Research Centre of Finland Ltd.
Esko Arilahti VTT Technical Research Centre of Finland Ltd.
Jussi Solin VTT Technical Research Centre of Finland Ltd.