A Multi-Scale Performance Prediction Method for 3D Printed Composite Materials Based on FEA and GNN

By combining finite element analysis and graph neural networks, a multi-scale prediction framework was constructed, which solved the problem of predicting the mechanical properties of 3D printed composite materials at the microstructure level. This enabled efficient and accurate mechanical property evaluation of continuous fiber reinforced composite materials and reduced computational costs.

CN121211797BActive Publication Date: 2026-05-26NANJING UNIV OF SCI & TECH
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202511173043.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-08-21
Publication Date
2026-05-26
Estimated Expiration
2045-08-21

Smart Images

  • Figure CN121211797B_ABST
    Figure CN121211797B_ABST
Patent Text Reader

Abstract

This invention discloses a multi-scale performance prediction method and system for 3D-printed composite materials based on FEA and GNN. The method includes: acquiring stress-strain data at the microscale under different fiber volume fractions and random distributions; converting a representative volume element finite element model into a graph structure model, and forming a graph structure dataset by constructing edge connections; constructing and training a graph attention network model containing a multi-head attention mechanism based on the graph structure dataset; constructing a temperature-dependent thermo-mechanical coupled macroscopic finite element simulation model based on the macroscopic filament geometry model to obtain stress responses at different temperatures; and constructing a joint deep learning model based on the stress responses at different temperatures to predict the mechanical behavior of composite materials under multiple temperature fields. Its significant advantage is that it comprehensively considers material properties, temperature effects, and structural randomness, and can efficiently and accurately simulate and predict the mechanical response of composite materials under different working conditions.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Thermodynamic property prediction method and device based on artificial neural network

    CN116206708A

  • Macroscopic progressive damage intelligent evaluation method driven by multi-scale microscopic damage observation

    CN120163006A