各校計畫成果

活動簡介

This project seeks to develop an integrated computer simulation and machine learning algorithms, advanced sensor materials and monitoring technologies for the prediction and damage assessment of critical infrastructures subjected to natural disasters, and mitigation of potential damage to geo-structure systems. The proposed technologies will also provide guidance to the local government for building responsive warning systems and designing enhanced Standard Operating Procedures (SOP) for emergency response to natural disasters promptly.

The first 1.2 years of efforts focused on the fundamental development of computational methods, high-performance computing, machine learning algorithms, damage assessment techniques, and investigation of 3D-printed ultra-high performance concrete (UHPC) for infrastructural damage prediction and protection under natural disasters. In conjunction with these fundamental research developments is the software implementation for Digital Twin application, including the a Reproducing Kernel Particle Method (RKPM) code for hydro-mechanical coupled slope failure modeling, the MoSES_2PDF code for landslide runout dynamics simulation, the thermodynamics-constrained convolutional neural network (TCCNN) for modeling complex path-dependent materials behavior under extreme loading conditions , for accelerate the prediction of damage modes in infrastructure components and slope instabilities. The team also worked closely with LAND ENGINEERING CONSULTNATS in collecting field data for training and validation of the computational code (RKPM and and MoSES_2PDF) and machine learning models (TCRNN, TCCNN) for the Guanziling (關子嶺) slope Digital Twin application.

J. S. Chen and the project team have established initial collaboration with Center for Extreme Events Research (CEER), Natural Hazards Engineering Research Infrastructure (NHEIR), Scripps Institute of Oceanography (SIO), and Department of Structural Engineering (SE) at UCSD on Infrastructure Digital Twins by co-organizing a UCSD-NCKU Workshop on Infrastructure Digital Twins for Natural Disaster Protection held in January 15-16, 2025, at University of California, San Diego (UCSD), to facilitate collaboration on the development of research and technologies relevant to infrastructure digital twins for natural disaster protection, sharing research results, and identifying the emergent needs for the rapid advancement of this critical field of research. During this 2 day workshop, a UCSD-NCKU MOU has been signed by UCSD JSOE Dean Albert Pisano and NCKU College of Engineering Dean Chyan Deng Jan.

Fellow J. S. Chen received the Raymond D. Mindlin Medal in 2024 and the John von Neumann Medal in 2025, the highest honor awarded by the U.S. Association for Computational Mechanics (USACM). These recognitions highlight his outstanding contributions and long-term dedication to the field of computational mechanics. His pioneering achievements in meshfree methods, multiscale modeling, and the simulation of engineering materials fundamentally transformed the way engineers understand complex structures and materials, thereby establishing his position as a global leader in computational mechanics.