AbstractAluminum‐lithium alloys are a class of advanced materials designed to reduce weight and improve performance in aerospace and other high‐tech applications. This paper presents a research investigation on the in‐phase and out‐of‐phase multiaxial fatigue behaviors of the third‐generation AW2099‐T83 aluminum‐lithium alloy that have not been addressed before. Additional hardening was observed under nonproportional loading condition at high strain amplitudes. Fatigue lives were estimated using von Mises equivalent strain and two critical plane models: the Fatemi‐Socie (FS) and the Smith‐Watson‐Topper (SWT). In addition, a supervised machine‐learning model (support vector machine—SVM) was employed to predict the fatigue life under the above‐mentioned loading conditions. The FS criterion was found to yield better fatigue life predictions than SWT. The estimations of FS model mostly fall within ±3× scatter bands with some data falling within the conservative and non‐conservative regions. The SVM model resulted in excellent predictions within ±2× scatter bands.
Read full abstract