Development of a Procedure for Estimating the High Cycle Fatigue Strength of Some High Temperature Structural Alloys

Abstract

The generation of strain controlled fatigue data, for the standard strain rate of 4 × 10−3 s−1 , presents a problem when the cycles to failure exceed 105 because of the prohibitively long test times involved. In an attempt to circumvent this difficulty an evaluation has been made of a test procedure involving a fast cycling rate (40 Hz) and load contolled conditions. The validity of this procedure for extending current fatigue curves from 105 to 108 cycles and beyond, hinges upon the selection of an appropriate strain value, since the strain range can change rapidly during the early stages of fatigue. Results from annealed 2 1/4 Cr-1 Mo, Type 304 stainless steel, Incoloy 800 H and Hastelloy X, tested over a wide range of temperatures, show that the strain measured at Nf /2 is a reasonable estimate since it gives an excellent correlation between the strain and load controlled tests in the 105 range where the data overlap. It seems clear that the differences in cycling rate and early stress-strain history for the two tests do not significantly affect the correlation. It may, therefore, be concluded that such load control test procedures may be used as a valid fast way for extending currently available fatigue curves from 105 to 108 cycles, and beyond.