The Pressurized Water Reactor (PWR) primary coolant environment is known both to significantly reduce the fatigue life of austenitic stainless steels and to lead to enhanced fatigue crack propagation rates. Relationships for the impact of the PWR coolant environment on fatigue life have been presented in NUREG/CR-6909 using an environmental fatigue correction factor (Fen), which is a function of temperature. Fatigue crack growth behavior has been codified in ASME Code Case N-809 in terms of parameters such as rise time, stress intensity factor, load ratio and temperature. However, plant performance suggests that the application of these predicted environmental effects using current assessment procedures for fatigue for plant transient loading may be unduly pessimistic. One potential reason for this over-conservatism is thought to be that, although the majority of plant design transients result from variations in thermal loading, most available data are derived from isothermal testing.
For the calculation of fatigue initiation life, NUREG/CR-6909 gives guidance on the effective temperature to be used in assessments of thermal transients. Recent results from Thermo-Mechanical Fatigue (TMF) testing on stainless steels in PWR coolant show that this guidance is conservative for out-of-phase cycling of temperature and loading, and potentially non-conservative for in-phase thermal loading. In contrast, code case N-809 gives no guidance on the effective temperature for fatigue crack growth assessments, resulting in maximum temperatures frequently being adopted for assessments of thermal transients. There is therefore a need for a clearer understanding of the impact of variable temperatures during transients on the predicted levels of environmental fatigue.
This paper describes test facilities developed to permit measurements of both thermo-mechanical fatigue life and fatigue crack growth rates in pressurized water reactor environments. Initial test results obtained using these facilities are presented. The fatigue life data have been generated for a range of applied strain amplitudes, 0.45% to 1%, using temperature cycling between 100°C and 300°C. These data, for both in- and out-of-phase temperature and loading, are compared to the predictions of the “weighted Fen” model which is detailed in a separate paper, PVP2017-66030. Similarly, crack growth rate data generated for cycles between 140°C and 280°C are presented and comparisons made against the predictions of the “weighted K rate” (WKR) method detailed in paper PVP2017-65645. In both cases, the test results suggest that the weighted models are able to provide good predictions of an effective temperature to be used in fatigue assessment methods, which offer a significant improvement in the treatment of variable temperatures compared to current assessment practice.