Nowadays, it is well known that the low cycle fatigue (LCF) life of austenitic stainless steels can be affected in specific conditions of temperature, strain rate, strain amplitude or dissolved oxygen concentration by the effect of Pressurized Water Reactor (PWR) primary coolant environment. Nevertheless, questions remain about the best methodology that must be used to consider environmental effects for nuclear power plant licensing and for operating lifetime extensions. These environmental effects are most commonly evaluated from a mean fatigue curve based on tests conducted in air at room temperature. However, it is well established that air is not a neutral environment for metallic alloys and its effect can be highly dependent on the temperature level. Thus, in order to evaluate the intrinsic fatigue resistance of a 304L austenitic stainless steel at 300°C and the importance of complex fatigue – environment interactions in air or in PWR water, LCF tests were performed in both environments and specifically designed ones were conducted in secondary vacuum. Tests were performed on 304L cylindrical specimens at 20 or 300°C in vacuum or in air and only at 300°C in PWR water, under total axial strain control using a triangular waveform at strain amplitudes of ±0.3 or ±0.6% and strain rates of 4 × 10−3, 1 × 10−4 or 1 × 10−5 s−1. It was found that compared with vacuum, air is responsible for a strong decrease in fatigue lifetime in this steel, especially at 300°C and low strain amplitude. The PWR water coolant environment is still more active than air and leads mainly to increased damage kinetics, with slight effects on initiation sites or propagation modes. More precisely, the decreased fatigue life in PWR water is essentially attributed to an enhancement of both crack initiation and “short crack” micropropagation stages. Furthermore, a detrimental influence of low strain rates on the fatigue lifetime at 300°C was observed in PWR water environment or in air, but also in vacuum without environmental effects, and was in the last case exclusively attributed to the occurrence of the dynamic strain aging (DSA) phenomenon. So, the use of data obtained in a neutral environment as a reference allows the evaluation of the intrinsic effect of each environmental or loading condition. Moreover, in an active environment such as air or PWR primary water, damage evolutions as well as fatigue lives cannot be predicted by a simple multiplication of each parameter effect taken separately because they are the result of numerous interactions. The last conclusion is supported by complementary results showing that the PWR water environment effect as well as the ground surface finish effect can be attenuated when LCF tests are performed with a more representative loading signal shape.

The small crack problem is addressed within the applicability of Linear Elastic Fracture Mechanics as the result of crack closure phenomenon. The variation of crack closure stress intensity factory Kop as a function of crack length, a, was determined in two materials, namely a A508 steel and a 2024A1 alloy. These results were obtained on two-dimensional small cracks (a 3 0.1 mm) which were machined from long fatigue cracks. These measurements of Kop in addition to data published in the literature on a nodular cast iron and a 9Cr-1Mo steel yield to a unique characteristic function: K,,/K, = 1 -exp(-ka) in which k is the only parameter to characterize the small crack effect. A prediction of the threshold behaviour of small and long cracks on A508 steel is made using the results of crack closure measurements.