Large eddy simulation of smoke blocking by water sprays in a tunnel fire

Abstract

This study presents a large eddy simulation (LES) of smoke blocking by water sprays in tunnel fires to investigate the effect of water sprays on the characteristics of the smoke flow. This study uses an LES solver – FireFOAM. In addition, fire experiments using a 1/10 scale model tunnel with two water spray nozzles are conducted to validate the LES results. The length of the model tunnel is 12 m, the height and width are 0.5 m, respectively, and the heat release rate of the fire is 7.38 kW (2.33 MW in full scale). Each spray nozzle operates at a pressure of 0.3 MPa, and the volume median diameter of spray droplets is 226 μm. First, a grid sensitivity analysis on the gas temperatures under free-burn conditions is performed. A 1.25 cm uniform mesh is chosen in the tunnel domain for all subsequent simulations. Then, FireFOAM predictions are validated against experimental data on the distribution of water spray mass flux under conditions without the fire and the gas temperatures measured under conditions with and without the water sprays. The water spray mass flux and gas temperatures predicted by FireFOAM show good agreement with the measurements. Next, the characteristics of the smoke flow are investigated in the presence of the water sprays. It is found that the water sprays block part of the smoke flowing under the ceiling and force it to descend to the ground. The descended smoke flows back in the lower part of the fire zone, and a vitiated air layer is formed. The vitiated air layer is composed of the smoke flowing back and fresh air drawn from the portal. On the other hand, some smoke leaks through the water spray zone and flows along the ceiling on the portal side. It is also found that the spray cooling delays the propagation of the leaked smoke toward the tunnel portal. Furthermore, it is quantitatively shown that the buoyant flow stratification destroyed by the water sprays becomes thermally and chemically stabilized again on the portal side, for the specific case under this study. Last, the smoke blocking performance is quantified based on the mass flow rates of the smoke flowing under the ceiling in the fire zone and the smoke flowing back toward the fire. A comparison of the smoke blocking rates between the predictions (time dependent: 40–50%) and measurements (steady: 39.4%) shows good agreement.