Estimating shear modulus of yarn on impact by lazy learning

Zishun Yuan,Jie He,Yaru Yang,Pinghua Xu,Zhenqian Lu,Wang Xu
DOI: https://doi.org/10.1016/j.ijmecsci.2024.109074
IF: 7.3
2024-02-02
International Journal of Mechanical Sciences
Abstract:The shear modulus ( G ) of a yarn model is of paramount importance for simulating the ballistic behaviours of the yarn-level models in finite element (FE) modelling. However, the G of the filament yarn is difficult to measure in the experiment. This study aims to propose the interpolation-based lazy learning methodology to estimate the G for a homogeneous yarn model under a high-speed impact in FE modelling. A two-step process has been developed, each of which contains an interpolation and a lazy learning approach with the 1-NN (1-nearest neighborhood) algorithm. A Dyneema® yarn model under a high-speed impact is initially developed and the transverse deflections of the model with three different values of G are collected as the input in each step. The input for the second step is based on the G predicted in the first. The transverse deflections of the final estimated G are highly consistent with the analytical counterparts. The methodology has been validated in estimating the G of yarn models with different materials and the G of crimped yarn models in the fabric model, which verifies the universality of the methodology.
engineering, mechanical,mechanics
What problem does this paper attempt to address?