Abstract

Based on the constructed natural suede moisture permeability visualisation model, the backward elimination method was employed for iterative term screening, resulting in an optimised model containing 10 key terms. Through systematic statistical validation, mechanistic interpretation, and parameter optimisation, a closed-loop validation of the model was completed, and its predictive performance and generalisation ability were comprehensively evaluated. The response surface model is shown to maintain excellent fitting characteristics (R² = 0.7706, Predicted R² = 0.7372). Analysis of variance indicates the model is highly significant (P < 0.0001). Contribution rate analysis identifies temperature as the most influential factor affecting moisture permeability (contribution rate 16.37%), followed by thickness (14.89%). The dominant role of the temperature factor is successfully validated by this method, and the interaction effect between temperature and thickness is confirmed to be the most significant, reflecting the synergistic mechanism of thermal driving force and diffusion path resistance. Based on model interpretation, the optimal process parameters for suede are determined as a thickness of 0.90mm, a temperature of 60°C, and a time of 1.00 hour, with an optimal moisture permeability rate of 12.5436kg/m²·24h. Validation experiments indicate that the relative error between predicted and measured values is less than 5%, and under conditions of parameter fluctuation, the optimised solution maintains 96.45% performance stability. Through a complete statistical modelling-optimisation-validation workflow, rigorously verified process conditions and a theoretical basis for regulating the moisture permeability performance of suede are provided. The established response surface model validation and optimisation methodology is considered to offer reference value for the process development of functional porous materials