Data-Driven and Topological Design of Structural Metamaterials for Fracture Resistance

Daicong Da, Liwei Wang, Yu-Chin Chan, Wei Chen

Research output: Contribution to conferencePresentation

Abstract

Overcoming the conflict between toughness and strength is a major challenge in the design of material microstructures. Data science as the fourth paradigm provides novel and diverse opportunities for materials design attaining exceptional mechanical properties. From both single- and multi-scale perspectives, this work aims to improve toughness and strength of structural brittle materials through precisely controlling the stress and combining topological optimization (TO) with data-driven techniques. First, mixed TO is developed to maximize the strain energy under displacement loading, while a maximum threshold is imposed on the local stress as a constraint [1]. Fracture response is tailored and compared between optimized structural materials by setting the threshold and adopting the phase field modeling method. Secondly, we propose a data-driven design framework for fracture resistance of the metamaterials with aperiodic microstructures. A deep neural network model consisting of a variational autoencoder (VAE) and a regressor for property prediction is trained to allow complex designs [2]. The latent space of the constructed VAE enables easy manipulation between different types of microstructures and generates new ones with tailored material properties. Thereafter, it is integrated in a topology-like distribution optimization for each component of the effective elasticity matrix. Finally, the targeted unit cells from the database are assembled into the macroscopic global structure to achieve the optimal distribution of the material properties. Consistency between different unit cells is guaranteed via a graph-based optimization method. We show the fracture resistance of porous structures can be significantly enhanced by the present method [3].
Original languageAmerican English
StatePublished - 28 Jul 2021
Externally publishedYes
Event16th U.S. National Congress on Computational Mechanics - Virtual
Duration: 28 Jul 2021 → …

Conference

Conference16th U.S. National Congress on Computational Mechanics
Period28/07/21 → …

EGS Disciplines

  • Biomedical Engineering and Bioengineering
  • Mechanical Engineering

Fingerprint

Dive into the research topics of 'Data-Driven and Topological Design of Structural Metamaterials for Fracture Resistance'. Together they form a unique fingerprint.

Cite this