Sensitivity Analysis on the Parameters of a Fire Spread Model

A physical model of a linear flame front spreading over a pine needle fuel bed was developed in our previous work. The model agrees well with the experimental data on rate of fire spread (ROS) for lower slope angles, while the ROS values are overestimated under higher slope angles. This is possibly due to the overestimate of the radiative heating or underestimate of the convective cooling, caused by the errors in the evaluation of model parameters. In this paper, the effects of the model parameters on the predicted ROS are investigated by model calculations, by which the sensitivity of the prediction of ROS to different model parameters is revealed. The results indicate that the ignition temperature and flame tilt angle have minor effects on the predicted ROS, while the flame length and fuel consumption efficiency have significant influences on the prediction especially for high slope angles. The predicted ROS increases with increasing flame temperature for all slope angles. It is also suggested that for higher slope angles, the natural convection cannot independently account for the convective cooling mechanisms, and higher heat convection coefficients should be used in the model to include the convective cooling due to the reverse air inflow around the flame front.