Hydrothermal aging effects on mechanical degradation of short fiber reinforced polyamide composites

Hydrothermal aging of the polymer matrix can significantly impact the long-term performance of short fiber reinforced thermoplastic composites (SFRTs), yet the extent of this influence remains unclear due to complex contributing factors. In this study, thermostatic immersion aging followed by tensile test is performed for polyamide (PA) and its composites reinforced with short carbon fiber and glass fiber, with variations in fiber content, geometric characteristics, and aging temperature. Despite differing diffusion rates and saturation levels among materials, all systems follow Fickian diffusion kinetics. To model the modulus degradation due to aging, a two-step homogenization (TSH) framework is developed by sequentially applying the Bridging Model and a hybrid method. The aged modulus is estimated by replacing the pristine matrix modulus with its moisture-degraded counterpart. However, predicting strength is more challenging due to the complex interplay between matrix plasticization and pore growth, as revealed by comparative CT imaging and tensile test results. Therefore, a Random Forest regression model is trained using 181 data points to predict the elastic modulus, tensile strength, and failure strain of SFRTs based on six input features. All predictions achieved R 2 values above 0.96 on the testing dataset, confirming the adequacy of the selected input features. Regarding feature importance, SHAP analysis identifies fiber volume fraction, moisture content, and fiber-to-matrix modulus ratio as the most influential variables. The minimal effect of aging temperature further supports the TSH assumption that moisture-driven changes in the matrix govern the mechanical degradation behavior of SFRTs in this system.