Studies On Characteristics Of Tunnel Fire Development And Natural Ventilation Mode Using Shafts

Tunnel is an effective way to solve traffic problems. Nowadays, more and more tunnels are under construction all over the world. However, owing to the narrow structure of tunnels, smoke induced by fires, which is the most fatal hazard to the people, will not easily be discharged. The smoke will hamper safe evacuation of occupants and affect firefighters from extinguishing the fire. In recent years, disastrous tunnel fires have occurred, thus more and more attentions are being paid to the fire protection of tunnels. There are already some researches on tunnel fires at home and abroad. Considering the problems existing in the tunnel fire research, the following aspects were studied in this work:In former studies, fires were always assumed to occur at the longitudinal centerline of tunnels. In fact, fires will occur at any location in a tunnel, with different distances to the tunnel sidewall. A set of small scale experiments was carried out, to investigate the influence of transverse fire location on maximum smoke temperature under the tunnel ceiling. Results show that the restriction effect of sidewall causes the maximum smoke temperature rise under the ceiling to increase compared with the unconfined space, even fires occur at the longitudinal centerline. The maximum smoke temperature rise above the fire kept almost unchanged with the fire moving closer to the sidewall at the beginning and then increased significantly after the distance between the fire and the sidewall decreased to a certain value. Under all fires, the smoke temperature rise under the ceiling decreased exponentially as the longitudinal distance from fire increased. Correlations for related parameters were proposed. Comparison for the transverse and longitudinal smoke temperature distributions near the fire was conducted and the difference was researched. A simplified correlation determining the transverse smoke temperature distribution under the ceiling was developed by taking the fire location into account.To study the influence of sidewall effect on flame characteristics and burning rate, a series of experiments with heptane pools was conducted. The results showed that as the fires were placed close to the sidewall, the flames inclined to the sidewall due to the restriction on air entrainment, and the burning rate increased on the whole, which could be mainly due to the enhanced radiation from the heated sidewall and ceiling flame. However, regardless of fuel shape, the burning rate obtained the peak value when the fire was near the sidewall, rather than attached to the sidewall, resulting from less flame radiation from the vertical flame part to the fuel in the latter case. The ratio of longitudinal ceiling flame length to transverse length tended to decrease with the fire moving close to the sidewall. For cases with the largest length and wall fires, the ratio was nearly0.5, which could be explained according to the theory of mirror effect. Also, due to the non-monotonous sidewall effect, a higher burning rate did not necessarily lead to a larger ceiling flame length.Model scale fire tests were performed in tunnels with varying tunnel widths and heights in order to study the effect of tunnel cross-section and ventilation velocity on the heat release rate (HRR), ceiling gas temperatures and heat fluxes downstream of the fire source for both liquid pool fires and solid fuel fires. The results showed that for well ventilated heptane pool fires, the tunnel width nearly has no influence on the HRR while a lower tunnel height clearly increases the HRR. For well ventilated solid fuel fires, the HRR increases by approximately25%relative to a free burn test but the HRR is not sensitive to either tunnel width, tunnel height or ventilation velocity. For solid fuel fires that are not well ventilated, the HRRs could be less than those in free burn laboratory tests. In the case of ventilation controlled fires, the HRRs approximately lie at the same level as for cases with natural ventilation. The maximum temperature under the tunnel ceiling is a weak function of HRR and ventilation velocity for large fires with HRR more than100MW at full scale. It clearly varies with the tunnel height and is a weak function of the tunnel width. With a lower tunnel height, the ceiling is closer to the base of continuous flame zone and the temperatures become higher. Overall, the gas temperature beneath the ceiling downstream of the fire decreases with the increasing tunnel dimensions, and increases with the increasing longitudinal ventilation velocity. The HRR is also an important factor that influences the decay rate of excess gas temperature, and a dimensionless HRR integrating HRR and other two key parameters, tunnel cross-sectional area and distance between fire source center and tunnel ceiling, was introduced to account for the effect. An equation for the decay rate of excess gas temperature downstream of the fire, considering both the tunnel dimensions and HRR, was developed. Moreover, a larger tunnel cross-sectional area will lead to a smaller heat flux downstream of the fire.A set of burning experiments with n-heptane pool fire was conducted to investigate the air entrainment mode with natural ventilation using shafts in tunnel fires. The disturbance of smoke exhausting on the smoke-air interface, which causes different amounts of fresh air exhausted directly and indirectly, was investigated. Results show that using shafts to discharge smoke leads to a strong mixing process between the smoke layer and the fresh air. The influence of longitudinal wind on natural ventilation with shaft in a tunnel fire was investigated numerically by Large Eddy Simulation. Various longitudinal wind velocities have obvious influence on the smoke stratification stability in the tunnel and natural ventilation performance of shaft. The plug-holing phenomenon occurs at small longitudinal velocity and brings adverse factors to the smoke exhausting process. A large longitudinal wind velocity will enhance the air entrainment in the tunnel and engenders an evident boundary layer separation along with a vortical region in the shaft, resulting in poor exhaust capacity. There exists a medium longitudinal wind velocity, in which the plug-holing is avoid and there is no obvious boundary layer separation occurring, thus better smoke exhausting effect can be obtained. The influence of shaft arrangement on natural ventilation performance during tunnel fires was investigated numerically by Large Eddy Simulation. As a whole, the total mass flow rate of smoke exhausted by shafts increases with the shaft amount under a given total area of shafts. The case with maximum shafts for natural ventilation can gain the best ventilation performance in spite of the longitudinal wind. The case with the largest longitudinal wind velocity will gain the minimum total mass flow rate of smoke exhausted in spite of the shaft amount, due to the fact that a very obvious boundary layer separation occurs. It is suggested that the cross-section of one shaft opening in the actual engineering design is oversize in general, which is not in favor of exhausting more smoke. Replacing the right-angle connection with the bevel-angle connection was proposed to split one separation point into two separation points, to attenuate the negative effect of boundary layer separation. The detailed characteristics of the separation phenomenon were analyzed and the proposition was verified by Large Eddy Simulation. Results show that there are no relatively large-scale vortexes in shafts with bevel-angle connections, resulting in improved natural ventilation effectiveness. For lower shafts, the advantage of using the bevel-angle connection is more significant.