Advancing Blast Fragmentation Simulation of RC Slabs: A Graph Neural Network Approach

The flowchart of FGN

In this talk, I present the Fragment Graph Network (FGN) [slides, code, paper], a novel approach to simulating blast fragmentation in reinforced concrete slabs using Graph Neural Networks (GNNs). The FGN model addresses the limitations of traditional experimental and numerical methods, offering a more efficient and scalable solution for predicting the dynamic response of structures under extreme conditions. This talk showcases the capabilities of GNNs in structural engineering, emphasizing their potential for enhancing protective design and ensuring public safety.