Some Advances on Understanding of High Cycle Thermal Fatigue Crazing

The aim of this paper is to improve the understanding of high cycle thermal fatigue crazing observed in some areas of residual heat removal (RHR) systems made of 304L stainless steel in PWR nuclear plants. High cycle thermal crazing and the absence of crazing under purely mechanical loading are explained through the arrest of cracks initiated at the surface in the thickness of the component for thermal fatigue. This arrest is due to high stress gradients in the case of thermal loading due to the high frequency of the thermal load. It is shown that close to the weld, the crack network configuration is related to the tensile weld residual stress field in accordance with the detrimental effect of tensile mean stress on fatigue life. However, these results are in apparent contradiction with the material uniaxial fatigue tests, where a beneficial effect of tensile mean stress is observed in load-controlled fatigue tests. Moreover, it is shown that near the weld, the absence of a crack network is related to the compressive weld residual stress field in accordance with the beneficial effect of compressive mean stress on fatigue life. This result is however, in apparent contradiction with the observation of crazing far from the weld under a highly compressive stress. These contradictions may be explained by the detrimental effect of prehardening in strain control and the beneficial effect of prehardening in stress control for stainless steels. From these results it is concluded that shot peening which is usually considered to be beneficial for ferritic steel may be detrimental in high cycle thermal fatigue for stainless steels. It may also be concluded that in high cycle thermal fatigue of stainless steel, improvement of fatigue life would be possible with heat treatment by reduction of strain-hardening rather than with shot peening.