A Life Prediction Model for Thermomechanical Fatigue Based on Microcrack Propagation

Abstract

A thermomechanical fatigue (TMF) life prediction model is developed based on the concept of microcrack propagation. The model is used to correlate high temperature fatigue lives of two nickel-base superalloys. Due to its complexity, a well-accepted framework for correlation of TMF life has been elusive. Various approaches have been taken, ostensibly nonisothermal generalizations of isothermally derived models. The proposed model explicitly accounts for damage from all three high temperature fatigue (HTF) damage mechanisms prevalent in metals, i.e. fatigue, oxidation, and creep. The general form of the microcrack propagation equation is dadN=dadN|fatigue+dadN|creep+dadN|ox No crack growth is actually monitored in the present analysis; rather the equation is integrated between an appropriate initial and final crack size. The constants and exponents in this integrated form of the equation are then fit from test data. The fatigue and oxidation components are correlated using the ΔJ parameter with an additional time and temperature dependence included in the oxidation term. The ΔJ parameter has demonstrated excellent versatility in the correlation of fatigue macrocrack growth and also implicitly contains the elastic deformation information that is so important in the life of high-strength Ni-base superalloys. The creep component of microcrack propagation is correlated using C, which is similar to the creep crack growth parameter Ct. Isothermal as well as TMF test data for MAR-M247 are analyzed. The mechanical strain versus temperature relationships for the TMF tests include in-phase (IP), out of-phase (OP), and counterclockwise diamond history (DH). Lives of the isothermal, IP, and OP tests are correlated within a factor of ± 2 of the median life. Lives of the DH tests are predicted within a factor ± 2 as well. Isothermal, hold time, and bithermal test data for as-cast MAR-M246 are also correlated within ± 2 of the median life using the approach.