CFD predictions of fire spread over wood cribs in large open-plan compartments: New insights

A CFD-based numerical simulator, coupled with a multi-component pyrolysis model established for an isolated wood crib fire test, successfully predicts fire spread behaviours and the evolution in compartment fire conditions in a large-scale test with an extended, uniform fuel bed. Five distinct fire spread phases are identified, showing a gradual transition from 2D to 1D planar fire spread patterns. The kinetically controlled pyrolysis model permits an analysis of the drivers of fire spread at the level of the individual wood sticks, in relation to the evolution of the heat transfer boundary conditions on each stick surface. The analysis shows linear correlations of burning rates with the incident heat fluxes at the fire’s leading edge, and similar trends within the crib during the main travelling fire phase, but higher rates later in the fire when the effects of preheating become more dominant. Burning at the trailing edge is more decoupled from thermal exposures, perhaps due to the insulating effects of the char layer. The simulation also reveals complex behaviours inside the wood crib, with a progression from the initial fuel controlled burning through the travelling fire phase where fire spread on the crib surface dominates, to the final rapid growth phase where in-depth burning rapidly catches up. Thus, we establish the value of linking the detailed heat transfer controlled pyrolysis within individual wood sticks to the macroscopic behaviours of fire spread seen at compartment scale. Such models have potential for more general deployment as numerical simulators which can be used to explore the coupling of fire behaviours to design parameters. • Numerical simulator established for fire spread prediction in a large compartment • Multi-component pyrolysis model successfully predicts fire spread behaviours • Model discriminates five distinct fire spread phases with transition from 2D to 1D spread • Drivers of fire spread assessed at level of individual wood stick via correlation to surface heat fluxes • Pyrolysis model captures strong enhancement in burning rates during growing phase of fire