Fatigue Life Prediction and Optimization of Gfrp Composites Based on Failure Tensor Polynomial in Fatigue Model With Exponential Fitting Approach

Abstract

In this study, a new fatigue life prediction and optimization strategy utilizing the Failure Tensor Polynomial in Fatigue (FTPF) model with exponential fitting and numerical bisection method for fiber reinforced polymer composites has been proposed. Within the experimental stage, glass/epoxy composite laminates with (Formula presented.), (Formula presented.), and (Formula presented.) lay-up configurations were fabricated, quasi-static and fatigue mechanical behavior of GFRP composites was characterized to be used in the FTPF model. The prediction capability of the FTPF model was tested based on the experimental data obtained for multidirectional laminates of various composite materials. Fatigue life prediction results of the glass/epoxy laminates were found to be better as compared to those for the linear fitting predictions. The results also indicated that the approach with exponential fitting provides better fatigue life predictions as compared to those obtained by linear fitting, especially for glass/epoxy laminates. Moreover, an optimization study using the proposed methodology for fatigue life advancement of the glass/epoxy laminates was performed by a powerful hybrid algorithm, PSA/GPSA. So, two optimization scenarios including various loading configurations were considered. The optimization results exhibited that the optimized stacking sequences having maximized fatigue life can be obtained in various loading cases. It was also revealed that the tension-compression loading and the loadings involving shear loads are critical for fatigue, and further improvement in fatigue life may be achieved by designing only symmetric lay-ups instead of symmetric-balanced and diversification of fiber angles to be used in the optimization.