Parametric Sensitivity of Predicted Fuel Fire in a Building

The ability of current modeling and simulation tools to accurately predict a building fire of practical size and duration is at issue. Modeling is challenged by computational cost, fidelity of assumed physics, and correct knowledge of initial and boundary conditions. A series of simulations has been conducted to compare with experiments for a fuel fire in a facility. The purpose of the study was to understand the importance of simulation parameters. The test geometry is sufficiently large and the fire of long enough duration to present a challenge to model in detail. Several computational parameters have been varied at magnitudes consistent with the uncertainty in the parameter to determine the parametric sensitivities. The predicted heat flux inside the facility was sensitive to varying degrees to the parameters selected for the study, with those related to the fuel source being the most important physical parameters.