Research
My research focuses on advancing the assessment and design of infrastructure systems subjected to extreme loading through the development of real-time hybrid simulation (RTHS) frameworks that integrate physical experiments with high-fidelity computational models. My work emphasizes multi-physics cyber-physical simulations incorporating soil–structure, fluid–structure, and aeroelastic interaction effects through the use of neural-network-based surrogate models and real-time digital twins. My broader research interests include structural behavior under extreme loading, real-time hybrid (pseudo-dynamic) simulation, structural dynamics and control, nonlinear structural analysis, numerical methods for structural systems, soil–structure interaction, fluid–structure interaction, and machine learning applications for nonlinear structural dynamics.
Research Projects
Multi-physics Real-time Hybrid Simulations
Development of RTHS frameworks integrating neural-network surrogate models of soil-foundation and fluid domains with experimental substructures for predicting response of structures under extreme loading. The research also focuses on including hydro/aeroelastic effects using scaled models of building facade in a wind tunnel/wave flume.
Real-time Digital Twins of Physical Systems
Online neural-network surrogate models and cyber-physical model updating frameworks for RTHS experimental substructures. The research incorporates data-driven and physics-based models that utilize measured responses from physical devices at one location in a building and creates a digital twin of these devices at other locations in the building.
Seismic Resilience of SMA-based Friction Dampers
Computational and experimental investigations on seismic performance assessment of SMA-based damping systems in reduced strength steel MRFs. The research includes seismic design and performance assessment.
Design and Performance Assessment of CLT Rocking Walls
Computational and experimental investigations on seismic performance assessment of mass timber CLT rocking walls. The research includes characterization testing, development of computational models, seismic desing and performance assessment including fragility analysis.
Research Interests
Research Demonstrations
Real-Time Hybrid Simulation of Fluid-Structure Interaction Systems
Experimental demonstration of multi-physics fluid-structure interaction using neural-network-based models. Numerical simulation clocktime using ALE-SSM computational fluid-dynamics framework is approximately 45 hours. RTHS clocktime using neural network model of fluid domain is 48 seconds. The simulation is run for a total of 48 seconds.
Real-Time Hybrid Simulation of Soil-Structure Interaction Systems
Experimental demonstration of multi-physics soil-structure interaction using neural-network-based models. Numerical simulation clocktime using OpenSees continuum model is approximately 2 hours. RTHS clocktime using neural network model of the soil-foundation domain is 40 seconds. The simulation is run for a total of 40 seconds.