Flame spread over solid fuels with simple geometries has been extensively studied in the past, but few have investigated the effects of complex fuel geometry. This study uses numerical modeling to analyze the flame spread and burning of wavy (corrugated) thin solids and the effect of varying the wave amplitude. Sensitivity to gas phase chemical kinetics is also analyzed. Fire Dynamics Simulator is utilized for modeling. The simulations are two-dimensional Direct Numerical Simulations including finite-rate combustion, first-order pyrolysis, and gray gas radiation. Changing the fuel structure configuration has a significant effect on all stages of flame spread. Corrugated samples exhibit flame shrinkage and break-up into flamelets, behavior not seen for flat samples. Increasing the corrugation amplitude increases the flame growth rate, decreases the burnout rate, and can suppress flamelet propagation after shrinkage. Faster kinetics result in slightly faster growth and more surviving flamelets. These results qualitatively agreement with experiments.

Theoretical models have been developed to address several important aspects of numerical modeling of turbulent combustion and flame spread. The developed models include a pyrolysis model for charring and non-charring solid materials, a fast narrow band radiation property evaluation model (FASTNB) and a turbulence model for buoyant flow and flame. In the pyrolysis model, a completely new algorithm has been proposed, where a moving dual mesh concept was developed and implemented. With this new concept, it provides proper spatial resolution for both temperature and density and automatically considers the regression of the surface of the non-charring solid material during its pyrolysis. It is simple, very efficient and applicable to both charring and non-charring materials. FASTNB speeds up significantly the evaluation of narrow band spectral radiation properties and thus provides a potential of applying narrow band model in numerical simulations of practical turbulent combustion. The turbulence model was developed to improve the consideration of buoyancy effect on turbulence and turbulent transport. It was found to be simple, promising and numerically stable. It has been tested against both plane and axisymmetric thermal plumes and an axisymmetric buoyant diffusion flame. When compared with the widely used standard buoyancy-modified k-e model, it gives significant improvement on numerical results. These developed models have been fully incorporated into CFD (Computational Fluid Dynamics) code and coupled with other CFD sub-models, including the DT (Discrete Transfer) radiation model, EDC (Eddy Dissipation Concept) combustion model, flamelet combustion model, various soot models and transpired wall function. Comprehensive numerical simulations have been carried out to study soot formation and oxidation in turbulent buoyant diffusion flames, flame heat transfer and flame spread in fires. The gas temperature and velocity, soot volume fraction, wall surface temperature, char depth, radiation and convection heat fluxes, and heat release rate were calculated and compared with experimental measurements. In addition to provide comprehensive data for comparison, experiments on room corner fire growth were undertaken, where the gas temperature, solid fuel surface temperature, radiative heat flux, char depth and heat release rate were all measured.

Material flammability and burning behaviors of thin solids in concurrent flows in normal and microgravity are studied using a previously-developed transient numerical model. The model consists of an unsteady gas phase and an unsteady solid phase. The gas phase solves full Navier-Stokes equations including mass, momentum, energy and species equations, using Direct Numerical Simulation. A one-step, second-order overall Arrhenius reaction is adopted. Gas phase radiation is considered by solving the radiation transfer equation with a discrete ordinates SN approximation. In the solid phase, conservation equations of the energy and mass are solved. A cotton-fiberglass-blend fabric is considered as the solid material in this research. Test-based two-step decomposition reactions are implemented for the solid pyrolysis. In this work, the following efforts are made: (1) enhancement to the Adaptive Mesh Refinement (AMR) scheme and (2) development of a two-dimensional version of the program (based on the original three-dimensional program). The first effort allows the program to simulate and resolve multiple flames spreading along the surface of the solid combustible. The second effort dramatically reduces the computational cost when simulating flame spread over wide samples. The model is applied to simulate three scenarios: (1) upward flame spread in normal gravity, (2) purely-forced concurrent flow flame spread over a large and wide sample (41 cm wide 94 cm long), and (3) purely-forced concurrent flow flame spread over a moderate size (5 cm wide, 30 cm long) sample. In the first scenario, upward flame spread in normal gravity, the simulations follow the dimension/configuration of a standard test, NASA-STD-6001 Test #1. This test is the current ground-based screening test for materials that are intended for use in space exploration. The tested sample is 5 cm wide and 30 cm long. In the simulation, ambient pressure is the main parameter. At low pressures, the conventional upward flame spread process is observed. As the pressure increases, a special flame splitting phenomenon is observed. The splitting process is presented in details using the solid and gas profiles. It is concluded that the two-step solid pyrolysis is the cause of this special phenomenon. For the second and third scenarios, simulations are performed to support an on-going NASA project Saffire, which consists of a series of large-scale microgravity burning experiments. Concurrent flow speeds at 20 and 25 cm/s are simulated for both large and moderate sized samples. The results of both Saffire experiments and the simulations are presented and compared in detail. The numerical results are also used to interpret the phenomena observed in the experiments. For the wide sample (scenario 2), a parametric study on the sample width (5-41 cm) is conducted, and additional simulations (using the two-dimensional version of the program) at various flow conditions (different flow speeds, ambient pressures, and oxygen percentages) are performed. Based on the simulation results, analytical analysis is conducted and formulations are proposed for flame spread rate and flame length. The proposed formulation for flame spread rate is evaluated using literature data of microgravity experiments and shows seasonable performance.

Flame spread over solid fuel presents distinctive characteristics in reduced gravity, especially when the forced flow velocity is low. The lack of buoyancy allows a blue, dim flame to sustain where the induced velocity would otherwise blow it off. At such low velocities, a quenching limit exists where the soot content is low and the effect of radiative heat loss becomes important. The objective of this study is to establish a high fidelity numerical model to simulate the growth and extinction of flame on solid fuels in a reduced gravity environment. The great importance of the spectral dependency of the gas phase absorption and emission were discovered through the model development and therefore, Statistical Narrow-Band Correlated-k (SNB-CK) spectral model was implemented. The model is applied to an experimental configuration from the recent space experiment, Burning And Suppression of Solids (BASS) project conducted aboard the International Space Station. A poly(methyl methacrylate) (PMMA) sphere (initial diameter of 2cm) was placed in a small wind tunnel (7.6cm x 7.6cm x 17cm) within the Microgravity Science Glovebox where flow speed and oxygen concentration were varied. Data analysis of the BASS experiment is also an important aspect of this research, especially because this is the first space experiment that used thermally thick spherical samples. In addition to the parameters influencing the flammability of thin solids, the degree of interior heat-up becomes an important parameter for thick solids. For spherical samples, not only is the degree of internal heating constantly changing, but also the existence of stagnation point, shoulder, and wake regions resulting in a different local flow pattern, hence a different flame-solid interaction. Parametric studies using the numerical model were performed against (1) chemical reaction parameters, (2) forced flow velocity, (3) oxygen concentration and (4) amount of preheating (bulk temperature of the solid fuel). Flame Spread Rate (FSR) was used to evaluate the transient effect and maximum flame temperature, standoff distance and radiative loss ratio were used to evaluate the spontaneous response of the gas phase to understand the overall response of the burning solid fuel. After evaluating the individual effect of each parameter, the efficacy of each parameter was compared. Selected results of this research are:[1]Experimental data from BASS and numerical simulation both showed that within the time periodbetween ignition until the flame tip reaches the shoulder of the sample, the flame length and timehave almost a linear relation.[2]Decreasing forced flow velocity increases the radiative loss ratio whereas decreasing oxygen molefraction decreases the radiative loss ratio. This finding must be considered in the effort to replicatethe behavior of flame spread over thick solid fuels in microgravity on earth.[3]Although the standoff distance will increase when the forced flow velocity is decreased as well aswhen the oxygen mole fraction is decreased, the forced flow velocity has a much stronger effect onthe standoff distance than the oxygen mole fraction.[4]Unlike the previous two comparisons, the effect of forced flow velocity and oxygen mole fraction onthe maximum flame temperature was at similar level, reduction of either parameter would result inlowering the maximum flame temperature.[5]The effect of preheating on the flame spread rate becomes stronger when either the oxygen flowrate or forced flow velocity becomes larger. Depending on which element is more important, we candistinguish oxygen flow rate driven flame spread from preheating driven flame spread. Findings of this research are being utilized in the design of the upcoming space experiment, Growth and Extinction Limits of solid fuel (GEL) project. This research is supported by the National Aeronautics and Space Administration (NASA). This work made use of the High Performance Computing Resource in the Core Facility for Advanced Research Computing at Case Western Reserve University and the Ohio Supercomputer Center.