Application of machine learning method to estimate bending properties of advanced composite elastic system subjected to external mechanical loading

Abstract

This work introduces the novel analysis of the bending behavior of a composite doubly curved panel made of graphene origami (GOri)-enabled auxetic metal metamaterials (GOEAMs) subjected to external mechanical force. The motivation for this study stems from the many applications of such structures in the field of aeronautics. These materials have mechanical and structural properties that continuously vary. This study is the first to examine the effects of an instantaneous external shock on the bending response of a composite doubly curved panel made of GOEAMs due to the extraordinary properties of these advanced materials. The mathematical governing equations are obtained from the higher-order shear deformation theory and are solved using the Laplace transform method (LTM) and the analytical solution procedure (ASP). After solving the equations, the results of the current system are compared to those of a previously published work, revealing a significant level of concurrence between the two sets of data. This study presents a machine learning approach that utilizes a deep neural network (DNN) with input, hidden, and output layers, as well as independent variables and other relevant factors. The aim is to provide an efficient computational technique for solving engineering issues. This is achieved by the use of mathematical modeling and the verification of the current output’s outcomes. Additionally, a database is supplied for an in-depth examination of this structure’s suitability for aeronautical purposes. The database contains comprehensive bending information, including normal, shear, and displacement fields in various directions, specifically for the GOEAMs subjected to external shock loading.