Fuchen Jia
1999
The main objective of this thesis is to develop relatively simple but reasonable engineering models within a CFD software framework to simulate fire in a compartment and fire growth and propagation in enclosures in which solid combustibles are involved through wall or ceiling linings. Gas phase combustion, radiation and solid fuel combustion are addressed in this study. At the heart of this study is the integration of the three sub-models epresenting the key elements mentioned above in compartment fire development and other auxiliary calculations such as the evaluation of the radiative properties of gas-soot mixture, temperature calculation for non-burning solid surfaces, etc. into a complete fire spread model.
Shortcomings in the conventional six-flux radiation model are highlighted. These were demonstrated through a simple artificial test case and corrected in the modified six-flux model. The computational cost and accuracy of the six-flux model and the discrete transfer method (DTM) using different number of rays are also investigated.
A simple empirical soot model is developed based on experimental observations that soot formation occurs in the fuel rich side of the chemical reaction region and the highest soot concentration is found in the same region. The soot model is important to evaluate the radiative properties of the gas-soot mixture in fires.
By incorporating the gas-phase combustion model, the radiation models and the soot model, substantial improvement in the predicted upper layer temperature profiles is achieved in the simulations of one of the Stecklers room fire test. It is found that radiation plays an important, perhaps dominant role in creating the nearly uniform temperature distribution in the upper layer.
The integral method to calculate temperatures of non-combustible solids is extended to be capable of dealing with the non-linearity of the reradiation at the solid surface(top surface) exposed to a fire and the convective heat loss at the opposite surface. The integral method is economic and simple for the calculation of temperatures of non-combustible solids.
Pyrolysis models for noncharring and charring solid combustibles are developed. The mass loss rates produced by the noncharring model for PMMA are in excellent agreement with experimental data. The charring model produced predictions for the mass loss rates and temperature distribution of a wood sample in very close agreement to that measured.
Finally, qualitative and quantitative verifications for the integrated fire spread model are carried out. The model is demonstrated to be capable of both qualitatively and quantitatively predicting fire, fire growth and development within compartment fire scenarios.