CRACK GROWTH IN ALLOY 718 UNDER THERMAL-MECHANICAL CYCLING (FATIGUE, FRACTURE, NICKEL, SUPERALLOY).

摘要

An investigation was conducted to evaluate and model the crack growth rates in a nickel-base superalloy under load controlled thermal-mechanical cycling. Experiments were conducted on center-cracked panel specimens of Inconel 718 with temperature limits of 427(DEGREES)C to 649(DEGREES)C. Closed loop temperature control in the cracked region of the specimen was maintained by a microcomputer and four quartz heating lamps. A D.C. electric potential drop method was used to monitor crack lengths. The elastic stress intensity factor, K, was used to correlate all crack growth data.All thermal and mechanical cycles used during thermal-mechanical fatigue (TMF) testing were symmetric, triangular, and 96 seconds long. Crack growth rates were determined over a range of (DELTA)K using a stress ratio, R, of 0.1. Tests were conducted with the maximum load leading the maximum temperature by phase angles of 0(DEGREES), 90(DEGREES), 180(DEGREES), 225(DEGREES), 270(DEGREES), and 315(DEGREES). The in-phase test (0(DEGREES)) produced the highest crack growth rates, with the 315(DEGREES), 270(DEGREES), 225(DEGREES), 180(DEGREES), and 190(DEGREES) tests following in order. The 0(DEGREES) and 90(DEGREES) crack growth rates were separated by over a factor of ten at all (DELTA)K values tested. All TMF crack growth rates were bracketed by the isothermal growth rates at 427(DEGREES)C and 649(DEGREES)C.The mixed-mode damage term did not significantly contribute to the linear cumulative damage model crack growth rate predictions. The original model integrated sustained load crack growth over the entire loading portion of the thermal-mechanical cycle and overpredicted TMF crack growth rates by a factor of up to four (270(DEGREES) test). The modified model integrated sustained load crack growth over the loading portion of the cycle as the sustained load crack growth rate is increasing. Two proof tests were conducted to evaluate the applicability of the modified model. All modified model predictions were within a factor of two of the experimental results.A linear cumulative damage model was developed which sums cycle-dependent, mixed-mode, and time-dependent damage terms to predict thermal-mechanical fatigue crack growth rates. The model was developed entirely from isothermal fatigue and sustained load crack growth test data.