Analytical and neural network-based approaches for mechanical postbuckling analysis of simply-supported functionally graded graphene origami-enabled auxetic metamaterial plates

Peng Shi, Zixuan Wang, Vu Ngoc Viet Hoang, Wei Zhao, Hang Xie, Raj Kiran, Jian Yang

Research output: Journal PublicationArticlepeer-review

Abstract

This study investigates the nonlinear postbuckling behavior of functionally graded graphene origami-enabled auxetic metamaterial (FG-GOEAM) plates with a negative Poisson's ratio under in-plane compressive loads. The FG-GOEAM plates are modeled with layer-wise graphene origami (GOri) distributions across their thickness, utilizing genetic programming-assisted micromechanical models to accurately capture their complex material behavior. The theoretical framework is established based on Reddy's higher-order shear deformation theory and von Kármán's geometric nonlinearity, while Galerkin method and Airy's stress function are employed to determine the critical buckling loads and postbuckling load–deflection curves. An innovative approach was developed by integrating artificial neural networks (ANNs) with analytical modeling to enhance prediction accuracy and computational efficiency. Three advanced training algorithms—Bayesian Regularization (BR), Levenberg–Marquardt (LM), and Scaled Conjugate Gradient (SCG) backpropagation—are employed to develop ANN models capable of accurately predicting the nonlinear postbuckling load–deflection response. These models are trained on an extensive dataset generated from the analytical solution, ensuring robust generalization while mitigating overfitting. A comparative analysis between the analytical results and ANN predictions demonstrates exceptional agreement, validating the reliability of the proposed approach. Notably, errors for BR and LM consistently remain below 0.5%, with BR often yielding the smallest deviations, while SCG errors can approach 3% in certain cases. Additionally, an in-depth parametric analysis is conducted to elucidate the effects of thermal environments, material distributions, and geometric variations on the nonlinear postbuckling responses of FG-GOEAM plates. Increasing the GOri weight fraction markedly enhances both critical and postbuckling loads, whereas a higher degree of GOri folding reduces these loads by inducing auxetic behavior and increased deformability.

Original languageEnglish
Article number113606
JournalThin-Walled Structures
Volume216
DOIs
Publication statusPublished - Nov 2025

Keywords

  • Artificial neural network
  • Auxetic metamaterial
  • Higher-order shear deformation theory
  • Kerr elastic foundation
  • Postbuckling analysis

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Building and Construction
  • Mechanical Engineering

Fingerprint

Dive into the research topics of 'Analytical and neural network-based approaches for mechanical postbuckling analysis of simply-supported functionally graded graphene origami-enabled auxetic metamaterial plates'. Together they form a unique fingerprint.

Cite this