The combined impact of voids and thermal aging on the mechanical reliability of epoxy resin evaluated by statistical analysis

Epoxy resin is widely used in electrical engineering, and its long-term mechanical performance is directly related to the durability of the equipment. The issue might be aggravated when voids are introduced during manufacturing process and when the material is used under high temperature environment. This study aims to precisely reveal the effects of voids and thermal aging on mechanical strength of epoxy resin. Specimens with low and high void content were prepared and thermally aged at 105 ℃ for various lengths of time. Tensile test, scanning electron microscope (SEM) and the finite element simulation were employed to investigate the failure mechanism. The chemical change due to thermal aging was studied by Fourier Transform Infrared Spectroscopy (FTIR). The results illustrate the crack is initiated at the edge of the void due to stress concentration. With the increased void content, there was a slight decrease in the mean tensile strength but a large (108%) increase in the standard deviation. During thermal aging, the chain scission and oxidation were observed via FTIR, along with the inhomogeneity of chemical compounds. A defect density parameter ρ was proposed and integrated into the Weibull distribution to study the synergistic effect of voids and thermal aging on the mechanical properties. This statistical analysis quantitatively describes the decrease in average tensile strength and the increase in data scattering with ascending defect density ρ, due to higher void content and thermal aging. When 0.0005% failure probability is required, the predicted failure stress is significantly reduced from 49.6 MPa to 1.18 MPa for the specimen with high level of void after the long-term aging. In this work, we demonstrate that Weibull statistical analysis can quantitatively evaluate the impact of void defects and thermal aging and provide strength design for the high-reliability epoxy material.