Electromechanical impedance-based deterioration monitoring and tensile strength prediction of epoxy resin under humid-thermal and alkaline environments

Deterioration monitoring and performance prediction under humid-thermal and alkaline environments are crucial for epoxy resins used as structural adhesives and composite matrices. The physical properties (water absorption, functional groups, and surface morphology) and mechanical properties (dynamic thermo-mechanical analysis and tensile behavior) of two resins exposed to alkaline conditions (60 °C, pH 13.6) were characterized. The adhesive and matrix reached water absorption saturation on days 15 and 7, respectively, with saturation levels of 12 % and 0.5 %. Optical scanning of the resin surface showed that pitting bumps appeared on the deteriorated adhesive, while reticulated cracks formed on the deteriorated matrix. The tensile strength of adhesive and matrix, as measured by tensile testing, initially increased briefly before continuously declining. By day 120 of aging, the tensile strength retention of the adhesive and matrix was 49.4 % and 96.5 %, respectively. To facilitate in-situ monitoring of epoxy resin deterioration, an electromechanical impedance (EMI)-based sensor was proposed. A piezoelectric ceramic (PZT) sheet was embedded within the resin to establish an electromechanical coupling system. Experimental results indicate that resin aging causes regular variations in the conductance peak frequency and conductance-based statistical indicators. Regression models correlating conductance characteristic indicators with resin strength were established, achieving coefficients of determination (R²) greater than 0.93, indicating strong correlations. By monitoring sensor conductance, quantitative predictions of the resin’s mechanical properties can be achieved. This study elucidates the deterioration of two commercial epoxy resins under humid-thermal and alkaline conditions, providing an effective technical solution for monitoring resin deterioration and predicting performance.