Fire Dynamics in Ship Enclosures Considering the Effects of Ceiling Vent and Fire Locations


Student thesis: Doctoral Thesis

View graph of relations


  • Jiaqing ZHANG


Awarding Institution
Award date28 Aug 2014


Fire is one of the most terrible accidents for ships, and always leads to ship damage and casualties. Understanding the effects of opening and fire locations on ship enclosure fires is a significant aspect of ship fire safety, and can provide essential data and a theoretical basis for safety design, fire fighting and onboard rescue. In order to examine the effects of XY and Z factors on ship enclosure fires, experimental and theoretical analysis of ceiling-vented compartment fires with various fire sources and vent locations, and elevated fires in compartments without vertical openings was conducted. Fire models for ship enclosures are proposed and the effects of XY and Z factors are characterized. The principal conclusions obtained from our experiments under the tested conditions are as follows:

The XY factor of fire sources, i.e., the impact of horizontal fire location in a ceiling-vented compartment, was revealed. Placing the fire off the center of the floor increased the length of the steady burning stage and decreased the average mass loss rate. The space-averaged oxygen concentration in the compartment and combustion efficiency of fires during the entire burning process decreased when being located horizontally off the ceiling vent, and the critical vent sizes were different for fires at different locations. The concept of dimensionless flame height is proposed to facilitate comparison of the flame stretch of fires with different heat release rates. A small but measurable increase in flame stretch occurred when the fire source was located off the center of the floor. The horizontal expansion of flames during boiling burning is described. A dimensionless temperature rise reflecting heat storage capacity is proposed to enable the comparison of the temperature rise of smoke for fires at different locations. The differences in smoke temperature could have been caused by both of the entrainment restriction and the horizontal distance between the fire and the ceiling vent. The average smoke density decreased with ceiling vent size, with a sharper decreasing trend for center fires than for wall and corner fires. The smoke descent rate of the center fire was the largest, due to the greater HRR. The XY factor of vent location, i.e., the impact of horizontal vent location in ceiling-vented compartment fires, was examined. The MLRs of the center-vented fires were slightly larger than those of corner-vented fires. The combustion efficiency of center-vented fires was greater than that of corner-vented fires, due to the larger oxygen concentration in the fire zone, as well as the heat release rate. With a lower heat storage capacity, the dimensionless temperature rises in the compartments of fires with center vents were smaller than those with fires with corner vents. The smoke descent rate of center-vented fires was larger than that of corner-vented fires. Numerical studies on the effects of vent location on ceiling-vented compartment fires with constant HRRs were conducted. The results obtained agreed well with the conclusions from the experiments using dimensionless variables.

The Z factor, i.e., the impact of elevation on fires in ceiling-vented and closed compartments, was established experimentally. The light extinction coefficient, oxygen concentration and gas temperature all showed distinct stratification. For a higher fire, the average fuel loss rate and the light extinction coefficient were smaller, the oxygen concentration was higher, the gas temperature was lower and the smoke descent rate was slower. As regards those parameters, the fire was less hazardous if elevated. Smoke filling of closed compartments with elevated fire sources is described. Parameters such as the light extinction coefficient, oxygen concentration and gas temperature showed distinct stratification, and the interface of the stratification was the fuel surface level. The results indicate that the smoke layer descended to the fuel surface level, but did not descend directly to the floor in the center of the compartment: rather, it continued the filling process by forming wall jets. The visualization showed that the wall jets penetrated the interface, traveled along the wall, concentrated at the floor, and then rose from the center of the floor. The impact of elevation on pool fire behavior was investigated on the basis of the distinctive stratification phenomenon. A method of calculating the oxygen depletion rate is proposed. The combustion efficiency and carbon conversion ratio decreased linearly with an increase in the height of the fire source. Considering heat release rate alone, elevated fires where the flame impinged on the ceiling were more hazardous at the early stage. There was no noticeable difference in the average heat release rate of fires impinging on the ceiling, however.

Fire models for ship enclosures, including the overall heat transfer correlations, a fire-induced temperature model, correlations of average oxygen concentration and smoke density are proposed, and the effects of XY and Z factors on established models are characterized. The average overall heat transfer coefficient was in direct proportion to one third of the power of the heat release rate. Use and misuse of the ceiling vent flow are discussed. Care should be taken in employing Cooper algorithms to calculate the mass flow out of a ceiling-vented compartment for fires directly beneath the ceiling vent. To avoid errors, an alternative method of calculating ceiling vent flow for fires positioned directly under the ceiling vent is proposed. Based upon the widely accepted horizontal vent flow formula and the energy balance in the room, the correlations of fire-induced temperature in ceiling-vented compartment fires are described. These correlations show good agreement with the experimental results. The impact of XY factor is reflected by the two newly-established dimensionless terms, which represent the ratio of total energy to energy released through the ceiling vent, and the ratio of energy lost through the wall to energy released through the ceiling vent. For fires horizontally off the ceiling vent, the temperature rise was in direct proportion to two thirds of the power of the heat release rate, while for fires beneath the ceiling vent, the temperature rise was in direct proportion to four thirds of the power of the heat release rate, and was inversely proportional to one sixth of the power of the ceiling vent size. The elevation of fires had no significant impact on our temperature model. Correlations for average oxygen concentration and mass density of the smoke in ceiling-vented compartments are proposed on the basis of a dimensionless opening factor that accounts for the compartment volume, ceiling vent size, air pumping ability of the fire and heat release rate.

    Research areas

  • ship enclosure fire, ceiling vent, closed compartment fire, fire location, vent location, fire models