Refined nonlinear micromechanical models using artificial neural networks for multiscale analysis of laminated composites subject to low-velocity impact. (1st March 2023)

• Record Type: Journal Article
• Authors: Hochster, Hadas; Bernikov, Yevheniia; Meshi, Ido; Lin, Shiyao; Ranatunga, Vipul; Waas, Anthony M.; Shemesh, Noam N.Y.; Haj-Ali, Rami
• Is Part Of: International journal of solids and structures. Volume 264(2023)
• Publication Date: 2023-03-01
• Subjects: Micromechanics -- PHFGMC -- Composite -- Artificial Neural Network -- Low-Velocity Impact; Mechanics, Applied -- Periodicals; Structural analysis (Engineering) -- Periodicals; Elastic solids -- Periodicals; Mécanique appliquée -- Périodiques; Constructions, Théorie des -- Périodiques; Solides élastiques -- Périodiques; Elastic solids; Mechanics, Applied; Structural analysis (Engineering); Periodicals; 624.18
• DOI: 10.1016/j.ijsolstr.2023.112123
• Languages: English
• ISSNs: 0020-7683

Abstract

Abstract: The parametric high fidelity generalized method of cells (PHFGMC) is an advanced micromechanical method that can be used for the nonlinear and failure analysis of several composite materials. The computational effort required for studying the nonlinear and damage multiaxial behavior is relatively small, depending on the size of the discretized repeating unit cell (RUC). However, it is computationally challenging, if not impossible, to integrate refined nonlinear micromechanical models within a multiscale analysis of composite structures. This is due to the thousands or more RUC models required at the integration points within a multiscale finite-element (FE) model of laminated structures. To that end, we propose a new artificial neural network (ANN) based micromechanical modeling framework, termed ANN-PHFGMC, for exploring the nonlinear behavior of fiber-reinforced polymeric (FRP) materials. Pre-simulated mechanical stress–strain responses and behaviors are determined using the PHFGMC to generate a multiaxial training database for the ANN micromodel. The simulated training data is founded on the PHFGMC-RUC results based on a hexagonal RUC. The PHFGMC effective stress–strain responses for different applied multiaxial strain paths are divided into two sets of data; one for the training and the other for verifying the trained ANN-PHFGMC model. The resulting trained ANN-PHFGMC is accurate, with less than 5% error in the verified predictions. The ANN-PHFGMC can be used as a stand-alone or embedded as a surrogate proxy model within a multiscale analysis of composite structures. Next, the ANN-PHFGMC model is integrated within a commercial explicit FE code for low-velocity impact (LVI) analysis of laminated composite plates. Multiscale LVI analyses are performed for two composite plates with different layups. Further, results are compared to experimental data to demonstrate the new model's ability to integrate refined nonlinear micromechanical models within a multiscale analysis.