(719f) Stochastic Simulation of Fire Spread in Solid Fuel Beds

Flame propagation through solid fuel beds relevant to wildland fires is a challenging and important problem.  The mechanism of flame spread in general involves a combination of radiative and convective heating coupled to local fluid motions induced by buoyant acceleration and wind.  Recent experiments and observations have shown the importance of direct flame contact for fire spread in relatively fine fuels.  This is due to convective cooling of the fuel that counteracts radiative heating.  To understand and model fire spread under these conditions requires detailed information about the intermittency of turbulent flamelet contact with solid fuel particles.  The only simulation approach that can fully resolved all flow and flame structures is direct numerical simulation (DNS).  This approach is computationally prohibitive for all but canonical flows at low Reynolds numbers.  Reynolds averaged Navier-Stokes (RANS), and Large eddy simulation (LES) approaches capture larger-scale flow, but do not resolve intermittent flames at the fine dissipation scales.  The one-dimensional turbulence model is a proven method for capturing multi-scale phenomena.  Intermittent flames are resolved in a single dimension, making the approach computationally affordable.  The ODT model is multi-scale.  It is well-suited to canonical flows that are well-represented by boundary layer assumptions (such as jets, plumes, and channel flows).  Whereas RANS and LES capture larger scales and model fine scales, ODT captures fine scales and models the turbulent advection through stochastic mapping processes (eddy events) that rearrange fluid elements on the domain in a manner consistent with turbulent scaling laws.

We present application of the ODT model to fire spread in solid fuel beds to investigate the interaction between intermittent flames and the fuel characteristics.  The ODT model solves transport equations for mass, momentum and energy in the gas phase with heat and mass transfer to a reacting solid fuel.  The fuel is heated and pyrolyzed.  The off gases react quickly in the gas phase relative to the transport timescales and are treated with an equilibrium model.  This eases the reaction of chemically complex off gases.  Buoyant effects are included, and radiative transport is modeled using a two-flux formulation with grey gases.  We introduce and present results of the model including sensitivity of fire spread rates to fuel particle size, and density, along with turbulent intensity, and wind speed.  Results will allow quantification of modeling approximations used in RANS and LES simulations.