GENERATIVE DESIGN과 MESHGRAPHNET을 활용한 패널 적층 구조 최적화 요소 기술 개발

February, 2023 · 1 min read

Development of an AI Model for Mobile Display Panel Stress Prediction Using MeshGraphNet (2022.5 - 2023.2)

Drop Impact

  • We achieved over 80% accuracy in predicting maximum panel stress compared to commercial simulation programs (e.g., ANSYS, ABAQUS).

  • We reduced cost and time while improving speed and efficiency by applying the AI model compared to conventional dynamic analysis methods.

  • We aim to optimize thin film thickness, materials, and PO Mobile panel stacking structure through Ball Impact simulations.

AI-Based Surrogate Model Spatio-Temporal System Graph Neural Network Meshgraphnet Drop Impact Test
Jiyong Kim
Authors
Jiyong Kim
Ph.D. student at KAIST