An engineering model for ignition and extinction of wood flames using bench-scale data

Abstract

Predicting the fire spread in wooden structure buildings widely uses engineering tools such as computational fluid dynamics (CFD) simulation softwares. These softwares mix several physical models to address a wide range of phenomena, such as heat and mass transfer, pyrolysis, combustion or fluid dynamics. These models must be kept as simple as possible to limit the modelling costs. The behaviour of wooden materials under fire conditions has been extensively studied for several years. The existing studies mainly focuses on the characterization of ignition of these materials. Several physical parameters have been examined, notably ignition time and temperature or critical ignition flux. In particular, ignition temperature has been shown a very elusive quantity (Babrauskas [1]). Some more recent work (Shen [2]) stated on the complexity to use theoretical models to predict ignition phenomena at an engineering level. Wood extinction has been much less studied. This issue raises as a major question in recent work for building fire safety studies, as it is a key parameter to demonstrate robustness against fire. It is commonly agreed that self-extinction of wood occurs in most cases when the external thermal aggression stops (Emberley [3]). The physical phenomena leading to extinction are still under study. This work proposes a criterion-based model of wood products ignition, pyrolysis and extinction with a very simple approach. The proposed criterion is built on test data measured on cross laminated timber and plywood panels, and identifies a burning region in the domain (Surface temperature) vs. (Net heat flux at the surface). This model has been incorporated into a CFD simulation software and used to model a large-scale fire of a plywood facade. It produced good outcomes for plywood fire spread predictions with very few modelling costs.