Flaw Tolerance Assessment for Low-Cycle Fatigue of Stainless Steel

Abstract

According to Appendix L of the Boiler and Pressure Vessel Code Section XI, flaw tolerance assessment is performed using the stress intensity factor (SIF) even for low-cycle fatigue. On the other hand, in Section III, the fatigue damage is assessed using the design fatigue curve (DFC), which has been determined from strain-based fatigue tests. Namely, the stress is used for the flaw tolerance assessment. In order to resolve this inconsistency, in the present study, the strain intensity factor was used for crack growth prediction. First, it was shown that the strain range was the key parameter for predicting the fatigue life and crack growth. The crack growth rates correlated well with the strain intensity factor even for the low-cycle fatigue. Then, the strain intensity factor was applied to predict the crack growth under uniform and thermal cyclic loading conditions. The estimated fatigue life for the uniform cyclic loading condition agreed well with that obtained by the low-cycle fatigue tests, while the fatigue life estimated for the cyclic thermal loading condition was longer. It was shown that the inspection result of “no crack” can be reflected to determining the future inspection time by applying the flaw tolerance analysis. It was concluded that the flaw tolerance concept is applicable not only to the plant maintenance but also to plant design. The fatigue damage assessment using the design fatigue curve can be replaced with the crack growth prediction.

Fatigue Damage Management According to Performance-Based Maintenance Concept

“…It was also shown that the curve was not affected by the crack growth rate when the exponential constant, which was 2.25 in Equation ( 4), was the same. Furthermore, the curve was almost the same for different rise times and heat transfer coefficients (Kamaya, 2017b). Namely, the curve was the same even if there were errors in estimating the crack growth driving force (stress and strain range) and in the crack growth rate (for the same exponential constant 2.25).…”

“…The initial and final depths were 50 μm and 10 mm, respectively. It was shown that the change in crack depth plotted for N/N f was completely identical for different temperature changes ΔTf and strain ranges (Kamaya, 2017b). It was also shown that the curve was not affected by the crack growth rate when the exponential constant, which was 2.25 in Equation ( 4), was the same.…”