In February/March 2007, The NRC issued Regulatory Guide “RG1.207” and Argonne National Laboratory issued NUREG/CR-6909 that is now applicable in the US for evaluations of PWR environmental effects in fatigue analyses of new reactor components. In order to assess the conservativeness of the application of this NUREG report, Low Cycle Fatigue (LCF) tests were performed by AREVA NP on austenitic stainless steel specimens in a PWR environment. The selected material exhibits in air environment a fatigue behavior consistent with the ANL reference “air” mean curve, as published in NUREG/CR-6909. LCF tests in a PWR environment were performed at various strain amplitude levels (± 0.6% or ± 0.3%) for two loading conditions corresponding to a simple or to a complex strain rate history. The simple loading condition is a fully reverse triangle signal (for comparison purposes with tests performed by other laboratories with the same loading conditions) and the complex signal simulates the strain variation for an actual typical PWR thermal transient. In addition, two various surface finish conditions were tested: polished and ground. This paper presents the comparisons of penalty factors, as observed experimentally, with penalty factors evaluated using ANL formulations (considering the strain integral method for complex loading), and on the other, the comparison of the actual fatigue life of the specimen with the fatigue life predicted through the NUREG report application. For the two strain amplitudes of ± 0.6% and ± 0.3%, LCF tests results obtained on austenitic stainless steel specimens in PWR environment with triangle waveforms at constant low strain rates give “Fen” penalty factors close to those estimated using the ANL formulation (NUREG/6909). However, for the lower strain amplitude level and a triangle loading signal, the ANL formulation is pessimistic compared to the AREVA NP test results obtained for polished specimens. Finally, it was observed that constant amplitude LCF test results obtained on ground specimens under complex loading simulating an actual sequence of a cold and hot thermal shock exhibits lower combined environmental and surface finish effects when compared to the penalty factors estimated on the basis of the ANL formulations. It appears that the application of the NUREG/CR-6909 in conjunction with the Fen model proposed by ANL for austenitic stainless steel provides excessive margins, whereas the current ASME approach seems sufficient to cover significant environmental effects for representative loadings and surface finish conditions of reactor components.
Effect of surface finish on the failure mechanisms of flip-chip solder joints under electromigration
