Influence of Load Ratio, Rise Time and Waveform on the Corrosion Fatigue Crack Growth of Austenitic Stainless Steel in a PWR Primary Coolant Environment

Abstract

The rate of growth of flaws in reactor circuit components by fatigue is usually determined using the reference crack growth curves in Section XI of the ASME Boiler and Pressure Vessel Code. These curves describe the rate of crack propagation per cycle (da/dN) as a function of the applied stress intensity factor range (ΔK). No reference curves for water-wetted defects in austenitic stainless steels are currently available. This paper describes the results of testing of austenitic stainless steel and weld metal in simulated PWR primary coolant over a range of temperatures and mechanical loading conditions. Previous data presented by the authors on wrought stainless steel demonstrated that crack growth rates can be significantly enhanced by the PWR primary environment at temperatures between 150°C and 300°C. The current study extends these data to weld metal and also investigates the impact of other loading waveforms (e.g. trapezoidal loading) on the degree of environmental enhancement. The environmental enhancement increases significantly with reducing loading frequency and decreases with decreasing water temperature. The environmental influence on fatigue is shown to be independent of load ratio over the range R = 0.1 to R = 0.8. The level of enhancement is frequently smaller at very high R ratio (≥0.85) with the enhanced rates of fatigue frequently being unsustained at these high load ratios. There is a strong correlation between the rise time and the level of enhancement of crack growth rate over inert crack growth rates at all temperatures tested. Weld metal has been shown to exhibit similar behavior to wrought material over the whole temperature range studied although the apparent rates of enhancement relative to average inert crack growth rates are lower than found for wrought material. For complex loading waveforms (e.g. trapezoidal loading with hold periods at maximum or minimum load) it is possible predict the level of enhancement on the basis of the test data generated using simpler saw tooth loading regimes.