Date of Completion


Embargo Period



High-fidelity simulation; Thermal radiation; Pool fire;

Major Advisor

Xinyu Zhao

Associate Advisor

Baki Cetegen

Associate Advisor

Tianfeng Lu

Associate Advisor

Alexei Poludnenko

Associate Advisor

Wilson Chiu

Field of Study

Mechanical Engineering


Doctor of Philosophy

Open Access

Open Access


Buoyancy-driven diffusion flames have been widely studied as a canonical fire configuration due to practical and scientific interests. Numerical investigations are conducted in this dissertation to improve understandings of interactions and couplings among turbulence, chemistry, soot, and multiphase radiation in buoyancy-driven diffusion flames. A high-fidelity modeling framework based on OpenFOAM-5.x, including detailed models for chemistry, radiation, and soot, is developed to improve the numerical accuracy and the computational efficiency with scale-resolved simulations. A Monte Carlo ray tracing (MCRT) based radiation solver coupled with line-by-line databases is developed to describe gas and soot radiation. Detailed and efficient radiation models for water mists are developed and coupled with the MCRT solver. An adaptive hybrid integration chemistry solver is implemented to speed up finite-rate chemistry integration. A semi-empirical two-equation soot model is incorporated to describe soot dynamics.

The developed multi-physical platform is systematically verified through a series of combustion-radiation systems including a laminar ethylene diffusion flame and four laminar methane diffusion flames with good agreement. The developed platform is subsequently employed to investigate a laboratory-scale turbulent pool fire. Good agreement with experiments on radiative heat fluxes, and with theories on flame temperature, velocity and puffing frequency, is achieved. Detailed investigations on interactions among chemistry, soot, radiation, and turbulence are performed to gain physical insights on modeling chemistry, soot and radiation.

Drawn on the database from high-fidelity pool fire simulations, three physics-based reduced-order models including a flamelet model considering re-absorption, an optimized two-step mechanism for chemistry, and a simple soot model based on the laminar smoke point concept, are developed. Encouraging results are obtained using the reduced-order models with considerable savings in computational cost. Finally, to investigate radiative attenuation of water mists in fire suppression, a radiation model considering anisotropic scattering for water mists is developed and validated against theoretical values, and is adopted to obtain benchmark results for development of reduced-order radiation models.