Glosey

Research Overview

My research focuses on understanding and mitigating the effects of fire on the built environment across two complementary thrusts: Structural Fire Engineering and Wildfire/WUI Engineering. It investigates the mechanics of structural response, ignition vulnerability, and infrastructure-scale risk, forming a unified framework that links component-level behavior to community-scale resilience.

Fire is among the most severe and complex hazards affecting buildings and urban environments—one increasingly amplified by climate change, drought, heatwaves, and the expanding wildland–urban interface (WUI). In 2022, more than 522,000 structure fires occurred in the United States, causing an estimated US $15 billion in direct losses. The 2025 Los Angeles wildfires—including the Palisades and Eaton Fires—damaged over 12,300 structures and produced insured losses exceeding US $20 billion, underscoring the urgent need for science-based, multi-scale fire resilience strategies.

Research Vision

My work integrates structural mechanics, fire dynamics, and data science to develop performance-based, climate-aware fire engineering tools. A central goal is to couple detailed modeling of heat transfer and structural response with community-scale wildfire exposure and infrastructure vulnerability, enabling fire-resilient and low-carbon design solutions at multiple scales.

A particular emphasis is on structure-to-structure fire spread in the WUI. Current codes largely address vegetation-to-structure ignition, but once a building ignites, its long-duration radiative output dominates local fire behavior. Timber-framed structures can burn for hours, releasing 2–3× the energy of the interior fuel load alone. My research quantifies these effects through fragility functions linking ignition probability to radiant heat flux, ember density, exposure duration, and building geometry.

Three Core Research Directions

(1) Computational Fire Mechanics
I develop physics-informed and AI-assisted computational frameworks that couple thermo-mechanical finite element analysis with surrogate modeling to predict structural behavior under realistic fire scenarios. This includes new methods implemented and taught in my Stanford course CEE 284F: Fire Engineering Design for Buildings.

(2) Wildfire Exposure & Community-Scale Resilience
I build spatial risk models combining radiant-heat networks, ember transport, and probabilistic ignition to assess structure vulnerability in WUI neighborhoods. This work supports hazard mapping, prioritization of mitigation strategies, and land-use planning, and is strengthened by my role as Co-PI on the NSF RAPID project IRIS-Wildfire, which documents ignition pathways and building-to-building spread during the 2025 Los Angeles wildfires.

(3) Low-Carbon Fire Design & Life-Cycle Performance
I investigate the trade-offs between fireproofing efficiency, structural performance, and embodied carbon. My recent development of Fireproof Optimizer (Stanford OTL S25-399) provides an AI-enabled platform for optimizing fire protection thicknesses while quantifying carbon reduction, enabling sustainable performance-based fire engineering.

Past Research: Advanced Modeling of Steel Connections under Fire

My earlier research advanced the understanding of how steel connections and composite floor systems behave under fire through a combination of full-scale testing, such as insights drawn from the Cardington frame experiments, and high-fidelity nonlinear finite element modeling developed during my FP7 Marie Curie Fellowship. This work demonstrated that composite slabs, continuity, catenary action, and tensile-membrane effects can significantly enhance structural fire resilience—mechanisms often overlooked in prescriptive design. The broader goal was to improve the prediction and performance of buildings, bridges, and large urban infrastructure under severe fires, generating fundamental knowledge on the fire behavior of connections and composite systems and informing performance-based design approaches across Europe. Through coordinated computational and experimental studies, the research clarified key phenomena such as connection behavior, composite action, thermal gradients, and global stability, while also contributing to hazard-mitigation priorities, developing validated design methodologies, and creating educational and Eurocode-aligned guidance for structural fire engineering practice.

This work built upon landmark large-scale fire experiments—such as the Cardington Frame Tests—revealing that composite floor slabs and continuity effects can significantly enhance fire performance through catenary action and membrane forces. These mechanisms are often not captured by prescriptive design methods but are critical for realistic assessment of structural robustness under natural fires.

Past Research: Nonlinear Transient Heat Transfer Code – FEHEAT

The overall research goal was to develop an effective finite element program, FEHEAT, for simulating two-dimensional nonlinear transient heat transfer in structural components subjected to fire. Implemented in MATLAB, FEHEAT solves the parabolic transient heat conduction equation with both convective and nonlinear radiative boundary conditions—an essential capability for thermal analysis in structural fire engineering. Unlike simplified analytical methods, FEHEAT captures the nonlinear effects of radiation from hot fire gases and enables accurate computation of temperature distributions across structural cross-sections, particularly steel I-shapes commonly used in practice.

FEHEAT was designed as a research and educational tool for performing detailed thermal analyses of beams, columns, and other components, where two-dimensional cross-section temperature fields are required for subsequent structural analysis under fire. The code incorporates consistent formulations for conduction, convection, and radiation, enabling users to compute fiber temperatures for integration into thermo-mechanical finite element models. The project included the development of a clear user manual, documentation of model capabilities and limitations, and several benchmark examples to validate the finite element formulation. This work supported the broader goal of improving structural fire modeling workflows by providing an open, accessible, and technically rigorous thermal analysis tool.