Knowledge about the margins against fatigue failure in nuclear piping systems is essential. Therefore, the margins in the ASME Boiler and Pressure Vessel Code (section III) fatigue design procedure for austenitic steels, type 304/316, as applied to real components have been investigated in this literature study. It is concluded that the current knowledge about the margins in ASME for realistic austenitic steel components is limited, since experimental evaluations of these margins are scarce in the literature.
It is concluded that data for fully reversed constant amplitude load on small smooth specimens are unable to provide reliable information about the fatigue margins in a real nuclear piping system. Thus previous discussions and analyses by for example ANL are of limited value. The test conditions with small scale smooth specimens deviate too much from the real fatigue situation and the ability to transfer these laboratory data to the real situation remains uncertain. Moreover, results from such testing tend to be biased by elasto-plastic phenomena, which play a key role in fatigue in austenitic steels. It is noteworthy that significant plastic deformation occurs in austenitic steels even for very long lives, which distinguishes the material from carbon steel and aluminum. As a consequence of the elasto-plastic properties, the effects of variable amplitude and differences between load and displacement control become important even in high cycle fatigue, thus having significant impact on margins. There is a clear need to investigate the transferability of the ASME data to fatigue in real components in operation. Hence, relevant component testing is necessary where conditions for elasto-plastic deformation must be realistic.
A convenient testing program for pressurized welded straight pipes loaded in four point bending is proposed. The testing will be performed under as realistic conditions as possible in a laboratory environment. Modern technique, combined with a rigorous statistical treatment, enables the establishment of SN-curve with prediction limits under variable amplitude. Thus direct evaluation of the margins in ASME is enabled and would significantly improve the knowledge about the margins in the ASME fatigue design procedure.