Yarn-scale fabric simulation with crease image validation for shape retention analysis

Fabric shape retention is a critical property that directly determines the functional performance and aesthetic quality of textile products; existing numerical simulation methods have limitations in simultaneously capturing the geometric deformation and internal stress distribution of fabrics during shape retention processes. This paper proposes a yarn-scale finite-element simulation model specifically designed for fabric shape retention detection, which covers three sequential stages of fabric shape change: the lifting and arching process, the compression process, and the release and recovery process. Three common fabric types were tested under different compression durations to verify the model’s applicability. To ensure the reliability of the proposed simulation model, an image-based verification method was developed. This method introduces crease image technology into yarn-scale simulation validation, representing an innovative application. Results confirm high simulation accuracy and mechanical reliability. Pearson correlation coefficients between simulated and experimental curvature values are 0.967 (cotton), 0.978 (polyester), and 0.969 (linen), with all fabrics showing satisfactory fitting precision (P < 0.01). Stress analysis reveals distinct gradient distributions in the crease regions. Cotton exhibits the highest residual stress, reaching 37 MPa under 90-second compression. Residual stress increases nonlinearly with arching duration. The yarn-scale simulation model integrated with the image-based validation method enables internal stress evaluation without physical crease tests. This provides key theoretical support for fabric structure optimization and high-performance textile development.