A probabilistic framework to define the design stress and acceptable defects under combined-cycle fatigue conditions

Probabilistic routines are necessary to determine the permissible design stress when the operational conditions of traditional gas turbines are further challenged to reach higher performances in terms of flexibility. This paper focuses on the definition of a probabilistic tool aimed to simulate the damage development under combined high and low cycle fatigue conditions considering the stochastic nature of some of the variables introduced in the model. The variables considered in the present study are the long crack threshold ΔKth,LC and the slope of the crack growth law C in the so-called Paris region, the endurance limit Δσw0 and the variability of the applied stress as it cannot be considered a fixed variable in the complex loading scenario of gas turbines. A broad experimental campaign was performed to characterize the material properties in terms of short and long crack resistance under constant and combined cyclic conditions. The data were used to define the average and variance of the materials parameters to feed Montecarlo simulations and, at a target probability of failure, establish the design stress. This work points to determine safety margins which strictly depend on the natural variability of the variables that govern the damage accumulation in the mechanical component.