NUREG report CR-6909 [1] proposed in 2007 new rules for evaluating environmental effects in fatigue analyses of new reactors components. These new rules, based on simple correlations, consider that Fen penalty factors (ratio of fatigue life in air at room temperature to that in water at service temperature) are mainly function of strain rate, temperature, sulfur content and dissolved oxygen concentrations. In order to evaluate the conservatisms included in the NUREG report CR-6909, discriminating Low Cycle Fatigue (LCF) tests were performed on a 304L austenitic stainless steel in PWR environment using various modified loading signals deduced from a representative loading strain history as close as possible to actual transients. Using the strain rate integral method recommended in the NUREG/CR-6909, the expected Fen penalty factors evaluated for each modified representative loading signal was close to 6, while the experimental Fen penalty factors measured were strongly dependent of the shape of the loading signals. Experimental Fen penalty factors obtained for the various modified loading signals vary from ∼ 1.5 to ∼ 4. These discriminating LCF tests performed on polished specimens demonstrate that the detailed strain integral method cannot correctly predict the actual environmental effects for the various shapes of loading signals deduced from a representative loading strain history. Other LCF tests were also performed for various constant strain amplitudes using fully reverse triangle signals for comparison purpose with tests performed by other laboratories. It appears that, contrary to published results, environmental effects decrease when the strain amplitude decreases from 0.6% until 0.2%. In the case of the strain amplitude of 0.6%, the Fen penalty factor formulation proposed in the NUREG report CR-6909 is accurate while, for lower strain amplitude of 0.2%, the Fen penalty factor formulation is very severe. It appears that the application of the NUREG/CR-6909 including the Fen model proposed by ANL (Argonne National Laboratory) for austenitic stainless steel provides excessive margins compared to penalty factors as observed experimentally. From this experimental program, conservatisms included in the NUREG/CR-6909 methodology appear to be excessive and can lead to fatigue design issues.
Effect of Strain Rate on the Low Cycle Fatigue Behavior of 316L(N) Stainless Steel Weld Joints
