Effect of tunnel slope on the critical velocity of densimetric plumes and fire plumes in ventilated tunnels

To control the harmful releases in a tunnel, a longitudinal ventilation flow has to be imposed. The concept of critical velocity has been the focus of numerous tunnel fire studies, with researches focus on horizontal tunnels and tunnels with inclination have received much less attention. Due to the stack effect, smoke movement in inclined tunnels (with inclination angle theta) is different from that in horizontal tunnels and the critical velocity can be affected. In this study, we tackle the problem of critical velocity theoretically, experimentally and numerically. The theoretical model relies on a top-hat plume impinging on the ceiling, and predicts a dependence, in case of buoyancy-driven releases, of the form V-c/V-c0 = root 1 + C(k)sin theta. Experiments are performed in a reduced scale tunnel, inclined from -5 degrees to + 5 degrees, and using light gas mixture (air/helium) to simulate the presence of light smokes. Numerical simulations are performed using FDS (Fire Dynamics Simulator) with hot-air plume and propane fire. Experimental results show that the dynamical condition at the source affects the critical velocity: when the buoyant plume is momentum-driven, the influence of slope is small; when the buoyant plume is buoyancy-driven, the influence of slope is large. This feature is conveniently reproduced by the analytical model adopting C-k = 2.1. Similar influence of Gamma(i) has been observed in the hot-air plume simulation with that in the experiment. Finally, in the numerical simulation of propane fire, results show that the influence of the slope on the ratio V-c/V-c0 is not affected by the heat release rate.