Publications Details

Solar Receiver Design: Treatment of Creep-Fatigue Interaction

Jones, Wendell B.; Stephens, John J.

This report presents a provisional lifetime prediction method which attempts to account for creep- fatigue interactions typically encountered in the design of solar central receivers that spend a considerable fraction of their operating periods subjected to compressive stresses at elevated temperature. During its operating life, a solar central receiver will be exposed to a large number of startup/shut- down cycles (relative to other power-producing systems), along with only short periods (up to 10-12 hrs.) of steady-state operation during each daily cycle. As such, fatigue-related deformation is expected to dominate the damage leading to failure in the high temperature alloys used for such as receiver. Thus, the provisional method concentrates on a fatigue-based damage approach, with direct accounting for the effects of thermo-mechanical fatigue and hold times at elevated temperatures. Note that creep damage is treated in an implicit way only, by means of the hold time correction. The starting point for the methodology is the isothermal low cycle fatigue data set used to develop fatigue design curves for ASME Boiler and Pressure Vessel Code Case N-47. Since the original data were not available for materials of interest (316 Stainless Steel and Alloy 800H), we attempted to estimate the original data sets by stripping away the safety factors of 2 on Δϵ and 20 on N₁ from the N-47 design curves. These "baseline data curves" for N₁ versus Δϵ, which represent the mean low cycle fatigue properties for each alloy at a given temperature, are tabulated in the Appendix in both tabular format and by means of sixth-order polynomial equations. The baseline data curves are first reduced to account for the effects of frequency and hold time. Comparison of hold time data for both 316 SS and 800H have indicated that additional factors of safety are required to make the frequency and hold time reductions conservative for all data considered. Therefore, safety factors of 1.5 on Δϵ and 4.5 on N₁ are used, and these are shown to give generally conservative predictions. Finally, reductions for thermomechanical fatigue damage are made which are a function of f, the fraction of thermally imposed strain to the total imposed strain. It is expected that the resulting fatigue design curves should yield reasonable life predictions for the design of solar central receivers.