Long-Life Fatigue of Type 316 Stainless Steel at Temperatures up to 593°C

Some energy-system structural components may be subjected to 10⁷ or more cycles of small-amplitude, strain-controlled vibrations. In such cases, information on the elevated-temperature, long-life (>10⁶ cycles to failure) fatigue resistance of the austenitic steels often used in these components is needed. Present design guidelines provide fatigue curves for these steels only up to 10⁶ cycles to failure. The objective of the present study was to evaluate the long-life fatigue resistance of one such steel, namely Type 316 stainless steel. Strain-controlled long-life (> 10⁶ cycles to failure) fatigue experiments were conducted on solution-annealed Type 316 stainless steel in air at temperatures from 21 to 593° C. These were all for continuous cycling of smooth specimens under fully reversed straining (no mean stress). Results of this work provided a major advance in understanding the fatigue behavior of this steel. Tentative best-fit fatigue curves have been developed, but more data are needed to establish needed statistical confidence in them. At 21°C, strain and load-controlled experiments gave similar fatigue-resistance values at 10⁸ cycles when inelastic straining was taken into account. However, at 427°C and above, strain-controlled cycling yielded fatigue-resistance levels at 10⁸ cycles about 15 to 25 percent above those for load-controlled cycling. This difference is related to the continually increasing stress levels observed under strain cycling at the higher temperatures. That is, cyclic hardening continues to occur for 10⁵ or more cycles of straining with accompanying two- to threefold increases in strength. This increased strength gives the increased fatigue resistance at long lives. Under load-controlled conditions, such cyclic hardening cannot occur, and the fatigue resistance is lower. Results of this work emphasize the need for considering the intended service conditions in carrying out laboratory experiments. The impact of these results on recommended experimental procedures for long-life fatigue testing of such alloys is discussed. Finally, considerations for application of these data in fatigue design are addressed.