Forecasting High Cycle and Very High Cycle Fatigue Through Enhanced Strain-Energy Based Fatigue Life Prediction

Abstract This study applies a novel strain-energy based fatigue life prediction method to low cycle fatigue (LCF) data of conventionally wrought and additive manufactured (AM) Inconel 718 round dog-bones, specifically focusing on cycles to failure less than 105. The objective is to forecast high cycle fatigue (HCF, 106–107) and very high cycle fatigue (VHCF 107), aiming to approximate stress versus cycles to failure (SN) behavior and the inherent variability in the fatigue life of the material system. This tool is motivated by the need to characterize an expanding and diverse array of materials produced by additive manufacturing methods, process modifications, and optimizations, which can significantly impact fatigue life and overall structural reliability. The fatigue life prediction method achieves this by generating a fatigue life distribution as a function of stress level. Bayesian statistical inference and stochastic sampling techniques are employed iteratively for each test data point, incorporating stress, cycles to failure, and total strain-energy dissipated leading to fatigue failure. Experimental LCF fatigue data are acquired using a standard axial servo-hydraulic load frame equipped with an axial extensometer for strain and strain-energy data collection. HCF and VHCF data is generated utilizing an ultrasonic fatigue test frame capable of cycling at 20kHz. The fatigue data obtained from both servo-hydraulic and ultrasonic fatigue testing are presented and compared. The energy-based fatigue life prediction framework, utilizing only LCF data, is deployed to forecast HCF and VHCF fatigue behavior for both conventionally wrought and AM Inconel 718 fatigue. The HCF and VHCF data generated are then used to validate the forecasts generated from the novel energy-based fatigue life prediction method. Comparisons are made against Random Fatigue Limit curve estimations to assess the effectiveness and accuracy of the proposed methodology.