A Phenomenological Predictive Model for Thermo-Mechanical Fatigue of Notched Type 304 Stainless Steel

The behavior of parts subjected to simultaneous thermal and mechanical fatigue loads is an area of research that carries great significance in the power generation, petrochemical, and aerospace industries. Machinery with expensive components undergo varying applications of force while exposed to variable temperature working fluids. An example case is found in steam turbines, which subject stainless steel blades to cyclic loads from rotation as well as the passing of heated gases. Accurate service life prediction is especially challenging due to the thermo-mechanical loading being present on the complex geometric profile of the blades. This research puts forth a method for determining crack initiation lifetimes in variably-notched type 304 austenitic stainless steel specimens subjected to differing fatigue and thermo-mechanical fatigue conditions. A base analytical model and genetic algorithm were used to develop phenomenology-informed predictions that fall within a factor of two of the actual crack initiation times.