Fundamental Study of Knit Wearing Simulation Using Multi-Fidelity, Multi-Scale Methodology

Abstract

Recently, there has been a demand to reduce the number of prototypes when developing knit products to realize a sustainable society. In response, attention has focused on knit-fitting simulators. Simulators have now modeled a single knit with a spring-mass model and mapped it onto a virtual garment shape. However, simulating knitwear, which can have several million knits, requires enormous computation time and has been challenging. Therefore, this study focused on the multi-fidelity multi-scale method as a top-down approach. It aimed to propose a framework for a simulator that efficiently calculates the shape of an entire knit product and the algorithms required for it. The results revealed the possibility of calculating the approximate shape of a garment shape with large-scale shape changes by changing the fidelity and scale in the simulation process.