No abstract

We have studied the corrosion-fatigue crack initiation and propagation of different types of steels: two austenitic, one austenitic-ferritic, and one ferritic. The testing program had been established in order to define, particularly, the part in corrosion fatigue of electrochemical effects and mechanical damage. Tests were carried out on smooth specimens in rotating bending and on precracked specimens in tension in air, and in a 3 percent aerated solution of sodium chloride. The capacity of the metal for dissolution and repassivation seems largely responsible for the behavior of the metal in the corrosive solution. This capacity varies notably with the composition of the metal, the loading conditions of the specimens, and the characteristics of the corrosive medium.

An experimental program has been conducted on Type 316 stainless steel uniaxial specimens to determine the main characteristics of their behavior under repeated loading. The cyclic stress-strain curve measured under constant repeated load, without high preloads, is independent of the level of the mean stress. Two different behaviors with respect to the occurrence of progressive deformation have been observed: (1) at room temperature, progressive deformation occurs when the maximum value of the tensile stress exceeds a critical value, and (2) at 320°C and for loading with temperature cycling between ambient temperature and 320°C, progressive deformation occurs when the amplitude of stress exceeds a critical value. These results show that Section III of the ASME Code, which limits the range of variation of the primary plus secondary stress intensity to the value of 3 Sm, gives an effective guarantee against progressive deformation for operating conditions in the range of 20 to 320°C. It is recalled that the absence of progressive deformation is a necessary condition for the validity of conventional low-cycle fatigue analysis.

In order to predict long-term behavior of a high-creep resistant low-carbon Type 316L stainless steel under low-cycle fatigue with long hold times, a series of tests of fatigue relaxation was undertaken at 550, 600, 650, and 700°C for medium strain ranges (Δεt = 0.7, 1.2, and 1.6 percent). Hold times up to 5 h were introduced at the maximum tensile strain. It has been shown that a reduction of fatigue life occurred, generally associated with intergranular cracking when hold times increased. A maximal effect was observed at 600°C. Different methods for extrapolating results for very long hold times, such as those encountered on fast breeder reactor components (∼1000 h) were proposed. These methods were based on a time-temperature equivalence comparable to those used for extrapolating creep rupture data. A correlation between reduction of fatigue life with the amount of stress relaxation during hold times was also used. Predictions by these methods are compared with ASME N47 fatigue design curves.

No abstract