Large Deflection Analysis of Functionally Graded Reinforced Sandwich Beams With Auxetic Core Using Physics-Informed Neural Network

Abstract

This paper aims to investigate the large deflection behavior of a sandwich beam reinforced with functionally graded (FG) graphene platelets (GPL) together with an auxetic core, rested on a nonlinear elastic foundation. The nonlinear governing equations of the problem are derived using Hamilton's principle based on the Euler-Bernoulli beam theory for large deflections. Five different distributions are considered to describe the dispersion of GPL in the top and bottom faces of the sandwich beam. The Physics-Informed Neural Network (PINN) method is employed to model the nonlinear deflection of the beam under various boundary conditions. This study highlights the effectiveness of PINN in handling the complexities of nonlinear structural analyses. The findings underscore the impact of the core auxeticity, GPL amount and distribution, and elastic foundation coefficient on the nonlinear deflection of the sandwich beam under different loading scenarios. For instance, using Type I configuration can reduce the deflection of the beam by nearly half compared to using Type IV. Furthermore, a nonlinear foundation with a unit coefficient results in a 48% reduction in deflection compared to the scenario without an elastic foundation.