| Fire is one of the most major threats to ships, and smoke management in fire is important to fire safety, as well as the survivability of ships. However, due to the lack of knowledge of fire characteristics under mechanical ventilation conditions on ships, the development of smoke management is conservative and slow, compared to that of fire detection and fire extinguishment. Therefore, the study on the necessity and feasibility of smoke management on ships is important to promote its development. Meanwhile, the guidelines for smoke management system on ships are extremely necessary.The requirements and aims of smoke management in different fire stages and ship areas were proposed firstly, and then engineering requirements of smoke management were translated into two fire research issues:the effects of mechanical ventilation on smoke management in cabins, and the smoke management methods in complex corridors. The full-scale experimental platform for smoke management on ships was built. With numbers of experiments, the fire characteristics in fire cabin and adjacent corridor under different ventilation conditions were studied, and the methods for calculating key parameters of smoke management systems on ship were proposed. The principal conclusions obtained from our experiments under the tested conditions are as follows:The effects of vent configuration on ship cabin fires were revealed, and appropriate style of the vent configuration for smoke management in ship cabins was proposed. Compared with the no ventilation test, the mass loss rate (MRL) of fuel was enlarged in the mechanical ventilation tests, and the MRLs in one-inlet tests were smaller than those in two-inlet tests. With the increase of inlet height, there was a "sudden drop" in MRLs, and the critical inlet heights were 0.43H and 0.76H in one-inlet and two-inlet tests in the conducted experiments, where H is the height of compartment. The cabin temperature was influenced by the heat production increase due to enlarge of MRL by ventilation and the cooling effect of cabin caused by the hot smoke ejection. Compared with the no ventilation test, the gas temperatures in one-inlet tests were lower, while those in two-inlet tests became higher. It means that the one-inlet and two-inlet configuration mitigated and aggravated the cabin thermal hazard, respectively. The gas temperatures in two-vent tests initially increased with the elevation of the air inlet, but suddenly dropped at when the inlet reached to 0.76H. However, the effects of inlet elevation on the gas temperatures in one-vent tests could be ignored. By revising the calculate method of the lower layer temperature, a modified three-layer model was proposed for analyzing the thermal stratification in mechanical ventilation enclosure fires. The smoke stability parameter Ψ=(Y∞-Yl)/(Y∞-Yu) was calculated based on oxygen concentration measurements. According to the thermal stratification height and the smoke stability parameter, elevating the air inlet and reducing the number of air inlets increased smoke disturbance and destroyed the formation of the smoke layer. Based upon the aim of smoke management in ship cabins and the tolerance of firefighters on fire productions, gas temperature was selected as the primary evaluation parameters affecting the smoke management. Based on the experiment results, one-inlet configuration with the inlet height lower than 0.26H was appropriate for smoke management in ship cabins.On the basis of the appropriate vent configuration for smoke management, the effects of mechanical ventilation rate on enclosure fires were examined, and the gas temperature models with different smoke distributions were established. The MRL of fuel increased initially and then decreased with the ventilation rate, and the critical ventilation rate was 40 air changes per hour (ACPH). The heat release rate and combustion efficiency were calculated according to the oxygen consumption method and the oxygen conservation equations. The combustion efficiency of fire was increased linearly with the oxygen concentration around fire, and the heat release rate had the similar trend with the MRL of fuel. The gas temperature increased with the ventilation rate firstly, and then decrease when it reached 20-30 ACPH. According to the smoke stability parameter Ψ, the increase of ventilation rate made for the formation of smoke layer in the conducted experiments. The smoke stability parameter Ψ was related to the dimensionless parameter ω=i/(?)2. In the conducted experiments, we got Ψ<0.5, which meant that the stable smoke stratification was formed in the cabin when ω>0.014. In the smoke stratification tests, the smoke stability parameter Ψ decreased linearly with the parameter ω. The temperature model for mechanical ventilation enclosure fires under smoke stratification assumption was established, and the mass transfer rate between the upper and lower layers, which could not be negligible in the mechanical ventilation condition, was introduced. The gas temperature calculated by method proposed in this study was higher than that calculated by the method based on the traditional smoke well-stirred assumption. By comparing with the experimental results, the method proposed in this study was appropriate for the tests satisfied ω>0.014, i.e., the smoke stratification tests according to Ψ. The method under traditional smoke well-stirred assumption showed better prediction for the tests with ω<0.014. With the aim of limiting the cabin temperature under the acceptable criterion, the ventilation rate should meet with ω>0.014 firstly, then design by the proposed temperature model based upon the smoke stratification assumption.Smoke movement characteristics in the connected two-deck corridor were investigated. The model for calculating gas temperature in the corridor with irregular geometry was established. Full-scale experiments on smoke movement in the two-deck corridor were conducted. Results indicated that smoke spread into the adjacent deck quickly when fire source was located at the lower deck (Deck 2). The fire hazard in the adjacent deck corridor was higher than that in the fire deck, according to the lower space temperature and thermal layer height. When the fire source was located at the upper deck (Deck 1), smoke descended quickly in the fire deck, but there was no smoke spread into the adjacent lower deck. The methods for calculating gas temperature at typical structures in complex corridor, including the horizontal opening, vertical opening, turn structure and fork structure, were investigated. In the openings, the spill plume were translated into the ideal point plume approximately, and the relationship between the temperatures of smoke inside and outside opening (ceiling level of the adjacent space) was established based upon the theory of the ideal point plume. In the turn and fork structures, the temperature decay was discussed based on the exponent assumption model. The temperature decay coefficients after smoke flow through the two structures were correlated to that in the straight corridor:KRa=0.74Ks, Kf,I=∑W/(2Wi)·Ks. By using the idea of fire network model, the model for calculating smoke temperature in irregular geometry corridors was established on the basis of the above methods for the four typical structures, and it was validated by the full-scale experiments. Results showed that the proposed model can be used for fast estimation on the thermal hazard in the corridor with irregular geometry on ships.The model for calculating the critical velocity of opposite airflow used for preventing smoke spread out of openings in corridor was established, and the calculation model was validated with the full-scale experiments. By large numbers of numerical simulations, we derived the critical velocities of opposite airflow in different fire scenarios. The correlation between the critical Froude number Frc=Vc/(2gh△T/T)1/2 and the dimensionless fire-opening distance l*=l/L for smoke prevention in horizontal and vertical openings was constructed, respectively. For horizontal opening, Frc initially declined linearly with l*:Frh,c=-0.14l*+0.8, and then became constant at about 0.38 when l*> 3. For vertical openings, Frc was insensitive with l*:Frv,c=0.365. Full-scale experiments on smoke prevention at ceiling aperture and door in corridor were conducted with the ventilation parameters designed by the above models. Results indicated that smoke can be prevented at both the ceiling aperture and the door in corridor with the designed opposite airflow velocities. Combined with the smoke temperature model in irregular geometry corridors and the critical Froude number models, the critical velocity of opposite airflow used for smoke prevention at horizontal and vertical openings in ship corridors can be predicted. |
PDF Full Text Request