Multiphysics modeling and experimental analysis of corrosion-assisted degradation in industrial pressure transducer packages under thermomechanical fatigue

This study focuses on investigating the fatigue behavior of pressure transducer packages subjected to combined thermal-mechanical and corrosion stresses. To analyze the stress distribution within the package, a comprehensive multi-physical finite element model was developed. Experimental testing involved subjecting sample packages to temperature and humidity cycling, along with exposure to NaCl solution. Scanning electron microscopy was utilized to observe any cracks and surface deformations that occurred during the testing period. The results highlighted an increasing measurement deviation with each cycling period, reaching over 14 % after 720 h when exposed to 90 °C and 30 % NaCl solution. The results highlight the adverse influence of corrosion on the thermal stress-induced performance degradation of packaging. To effectively capture and assess the observed trends in the experimental data, an artificial neural network (ANN) model was utilized. The ANN exhibited remarkable predictive abilities, attaining a significantly high coefficient of determination value of 0.997 across diverse conditions. [Display omitted]