Research
I have carried out research in structural and geotechnical engineering with a focus on uncertainty and risk calculation. I am particularly interested in employing smart sampling and advanced computation methods for efficient regional scale simulations. In my PhD, I studied computational modeling of soil-structure interaction in underground constructions. In my postdoc at SimCenter, I extended to analyze regional-scale interdependent infrastructure systems, such as building stocks, transportation, and pipeline networks, under multiple natural hazards, including earthquakes, soil liquefaction, landslides, and floods. I am leading the development of the Regional Resilience Determination (R2D) Tool at SimCenter.
My current research activities involve post-disaster transportation and lifeline systems assessment, community resilience simulation, efficient uncertainty quantification for regional-scale risk assessment, machine learning and remote sensing-based soil-infrastructure monitoring, and soil-structure interaction in urban excavations.
Post-disaster transportation and lifeline systems assessment
- Mentored a project at 2024 NHERI Computational Academy about transportation systems assessment for the San Francisco Bay Area.
- Conduct seismic risk assessment for the water conveyance and distribution system of the Metropolitan Water District (MWD) of Southern California.
- Participate in the California Energy Commission-funded OpenSRA 2 project to develop performance-based monitoring and risk assessment tools for gas pipelines.
Community resilience simulations
- Developed an integrated computational workflow spanning from hazard, exposure, and vulnerability to recovery modeling based on R2D and open-source package pyrecodes.
- Developed a testbed based on the Island of Alameda in California.
- Presented at the 2024 Engineering Mechanics Institute (EMI) Conference and the EMI Objective Resilience Committee’s meeting.
Efficient uncertainty quantification for regional risk assessment
- Proposed an importance sampling-based earthquake scenario downsampling algorithm.
- Develop adaptive importance sampling for regional risk assessment (ongoing).
- Develop hazard-consistent aftershock sequences sampling algorithms (ongoing).
Machine learning and remote sensing-based soil-infrastructure monitoring
- Machine-learning-based algorithms are used to identify earthquake fault locations using InSAR sensing.
- Remote sensing informs pipeline risk assessment (ongoing).
- Geotechnical and geospatial model informed machine-learning algorithms for ground deformation reconnaissance.
Soil structure interactions in urban excavations
- Developed a two-stage finite element method for 3D soil-structure interaction.
- Proposed a probabilistic assessment method for building safety assessment in tunneling and deep excavation-induced ground deformations.
- Published open-source Python package and ArcGIS toolbox for soil-structure interaction analysis.
