Novel Planar Auxetic Metamaterial Perforated with Orthogonally Aligned Oval‐Shaped Holes and Machine Learning Solutions
Hui Wang, Si-Hang Xiao, Chong Zhang
Abstract
Auxetic metamaterials with negative Poisson's ratio have attracted much attention due to their counterintuitive deformation behavior over the conventional engineering materials. However, it is difficult to describe the complex correlation between microstructure parameters and auxeticity by analytical or empirical solutions in the form of math expressions. Herein, the machine learning (ML) model with artificial neural network (ANN) is developed to analyze a novel planar auxetic metamaterial designed by introducing orthogonally aligned oval‐shaped perforations in solid base material, and its feasibility is demonstrated through the experimental and finite element method (FEM) solutions. It is found that the proposed structure involving less design parameters exhibits the best performance at the aspects of auxetic behavior and stress level than those with peanut‐shaped holes and elliptic holes. Moreover, the results of parameter analysis demonstrate that the present ML solution model can provide accurate predicting results rapidly for this problem, without the limitations of explicit solution expressions which are typically not available in practice. The ML model allows one to obtain the desired auxetic property by tailoring the geometric parameters effectively and accelerate auxetic metamaterial design.