Neural network modeling for anisotropic mechanical properties and work hardening behavior of Inconel 718 alloy at elevated temperatures

Inconel alloys are gaining a special attention for high temperature applications in service environment of aircraft structures, rocket engines, nuclear reactors, gas turbines and pressure vessels. This makes crucial to understand anisotropic material properties and work hardening behavior of a material. In this study, various mechanical properties such as ultimate strength σ u t s , yield strength σ y s , strain hardening exponent (n) and % elongation have been evaluated by using uniaxial tensile tests. The tensile tests have been conducted from room temperature to 600°C at an interval of 100°C with different slow strain rates (0.0001, 0.001, 0.01 s−1). Additionally, anisotropy of Inconel 718 alloy has been evaluated based on various measurable parameters such as normal anisotropy, planer anisotropy, in-plane anisotropy and anisotropic index. Furthermore, stress–strain response is analyzed by empirical work hardening equation by Hollomon, Swift, Ludwick and Voce. The Artificial Neural Network (ANN) models have been developed to predict various anisotropic mechanical properties and hardening behavior of Inconel 718 alloy. The ANN model is skilled by Levenberg–Marquardt algorithm and signifies a good accuracy of model with an excellent correlation coefficient and significantly low average absolute error. Validation for the accuracy of developed ANN model is confirmed with results from f-test and mean paired t-test.