**2. Experimental Setup**

These experiments were carried out in the scaled engine cabin test platform (Figure 1). The top and bottom plates of cabin were 8.5 m (L) × 7.5 m (W) and 8.5 m (L) × 6 m (W). The height of cabin was 1.9 m. Inside the cabin there were two gas turbine models with a height of 1.25 m. The experimental bench smoke exhaust system was based on the original ventilation system, based on the principle of smoke control, and improved by adding

pipelines and valves. There were four air supply inlets and two smoke exhaust outlets in the engine room. The schematic diagram of the air supply inlet was shown in Figure 1, its size could be adjusted, and the opening of the air inlet was controlled by the valve (0.2 m × 0.2 m, 0.15 m × 0.15 m, 0.1 m × 0.1 m). The air inlet could be set to three different heights—0.1 m, 0.4 m, 0.7 m—from the bottom plate. The smoke exhaust outlet was located at the top of the cabin, and its size was 0.4 m × 0.4 m. pipelines and valves. There were four air supply inlets and two smoke exhaust outlets in the engine room. The schematic diagram of the air supply inlet was shown in Figure 1, its size could be adjusted, and the opening of the air inlet was controlled by the valve (0.2 m × 0.2 m, 0.15 m × 0.15 m, 0.1 m × 0.1 m). The air inlet could be set to three different heights— 0.1 m, 0.4 m, 0.7 m—from the bottom plate. The smoke exhaust outlet was located at the top of the cabin, and its size was 0.4 m × 0.4 m.

*Fire* **2023**, *6*, 16 3 of 14

height of 1.25 m. The experimental bench smoke exhaust system was based on the original ventilation system, based on the principle of smoke control, and improved by adding

**Figure 1.** Front view (**left**) and top view (**right**) of the experimental cabin. **Figure 1.** Front view (**left**) and top view (**right**) of the experimental cabin.

To compare the effect of smoke control in the cabin under different forced air conditions, the experiment considered four different air supply configurations of 2.646 m³/s, 2.352 m³/s, 2.058 m³/s and 1.764 m³/s, and three different air inlet heights of 0.1 m, 0.4 m and 0.7 m. The smoke exhaust volume was based on the existing ventilation system design and is fixed at 0.9 times the air supply volume. To compare the effect of smoke control in the cabin under different forced air conditions, the experiment considered four different air supply configurations of 2.646 m3/s, 2.352 m3/s, 2.058 m3/s and 1.764 m3/s, and three different air inlet heights of 0.1 m, 0.4 m and 0.7 m. The smoke exhaust volume was based on the existing ventilation system design and is fixed at 0.9 times the air supply volume.

When comparing the cabin smoke control effects under different forced air volumes, the first principle should be to ensure the consistency of the fire source in each test. Therefore, the heptane oil pool fire with good repeatability was used as the experimental fire source. The diameter of the oil pan is D = 40 cm, and the amount of heptane in a single experiment is 3.5 L. The thermophysical properties of heptane are shown in Table 1. When comparing the cabin smoke control effects under different forced air volumes, the first principle should be to ensure the consistency of the fire source in each test. Therefore, the heptane oil pool fire with good repeatability was used as the experimental fire source. The diameter of the oil pan is D = 40 cm, and the amount of heptane in a single experiment is 3.5 L. The thermophysical properties of heptane are shown in Table 1.



Effective absorption coefficient 1.1 (±0.3)

During the experiment, the Deant ES60K electronic balance was used to measure the mass loss of the pool fire fuel with a range of 60 kg and an accuracy of 0.5 g. Thermocouples were used as temperature sensors in the experiment, and K-type armored thermocouples with a diameter of one millimeter were used for the thermocouples. A total of five thermocouple trees are arranged. The thermocouples are arranged vertically, one at a distance of 0.2 m downward from the height of 1.8 m at the top of the cabin. The specific layout is shown in Figure 2. Thermocouple tree 1, 2 and 3 are placed at the quarter positions on the longitudinal axis of the cabin to monitor the temperature at the fire source in the cabin and on both sides of the fire source. Thermocouple tree 4 is placed at the position of the manhole to monitor the temperature distribution at the manhole as the only escape During the experiment, the Deant ES60K electronic balance was used to measure the mass loss of the pool fire fuel with a range of 60 kg and an accuracy of 0.5 g. Thermocouples were used as temperature sensors in the experiment, and K-type armored thermocouples with a diameter of one millimeter were used for the thermocouples. A total of five thermocouple trees are arranged. The thermocouples are arranged vertically, one at a distance of 0.2 m downward from the height of 1.8 m at the top of the cabin. The specific layout is shown in Figure 2. Thermocouple tree 1, 2 and 3 are placed at the quarter positions on the longitudinal axis of the cabin to monitor the temperature at the fire source in the cabin and on both sides of the fire source. Thermocouple tree 4 is placed at the position of the manhole to monitor the temperature distribution at the manhole as the only escape route after a fire occurs in the cabin. Thermocouple tree 5 is used to monitor the temperature distribution at the left and right ladder openings after a fire occurs in the cabin. Since the left and right are basically symmetrically distributed, only one bunch of thermocouple trees is provided.

*Fire* **2023**, *6*, 16 4 of 14

**Figure 2.** Schematic diagram of fire source and thermocouple layout. **Figure 2.** Schematic diagram of fire source and thermocouple layout. **3. Results and Discussion** 

### **3. Results and Discussion** *3.1. Mass Loss Rate*

### **3. Results and Discussion**  *3.1. Mass Loss Rate* The mass loss rate is a parameter describing the mass loss of fuel during combustion.

trees is provided.

trees is provided.

*3.1. Mass Loss Rate*  The mass loss rate is a parameter describing the mass loss of fuel during combustion. it can be obtained by differentiating the mass change curve measured by the electronic balance. Through the change in mass loss rate, we can intuitively see the development process of the fire and the influence of boundary conditions on the compartment fire. Figure 3 shows the development law of the mass loss rate of pool fire under the condition of forced air supply and the effect of different air supply air volumes on the mass loss rate. Under the condition of forced ventilation, as the combustion progresses, the temperature of the gas in the cabin rose, and the heat radiation received by the oil pool gradually increased, which accelerated the evaporation process of the fuel and increased the mass loss rate. After burning for a certain time, the pool fire entered the boiling burning stage, and the mass loss rate increased suddenly, and then gradually decreased with the consumption of fuel. With the increase in the air supply volume, the development of pool fire was obviously accelerated, and the burning time was shortened. Figure 4 shows the change in The mass loss rate is a parameter describing the mass loss of fuel during combustion. it can be obtained by differentiating the mass change curve measured by the electronic balance. Through the change in mass loss rate, we can intuitively see the development process of the fire and the influence of boundary conditions on the compartment fire. Figure 3 shows the development law of the mass loss rate of pool fire under the condition of forced air supply and the effect of different air supply air volumes on the mass loss rate. Under the condition of forced ventilation, as the combustion progresses, the temperature of the gas in the cabin rose, and the heat radiation received by the oil pool gradually increased, which accelerated the evaporation process of the fuel and increased the mass loss rate. After burning for a certain time, the pool fire entered the boiling burning stage, and the mass loss rate increased suddenly, and then gradually decreased with the consumption of fuel. With the increase in the air supply volume, the development of pool fire was obviously accelerated, and the burning time was shortened. Figure 4 shows the change in the mass loss rate under the condition of different air supply port heights. With the change in air supply inlet height, mass loss rate change process was basically the same. it can be obtained by differentiating the mass change curve measured by the electronic balance. Through the change in mass loss rate, we can intuitively see the development process of the fire and the influence of boundary conditions on the compartment fire. Figure 3 shows the development law of the mass loss rate of pool fire under the condition of forced air supply and the effect of different air supply air volumes on the mass loss rate. Under the condition of forced ventilation, as the combustion progresses, the temperature of the gas in the cabin rose, and the heat radiation received by the oil pool gradually increased, which accelerated the evaporation process of the fuel and increased the mass loss rate. After burning for a certain time, the pool fire entered the boiling burning stage, and the mass loss rate increased suddenly, and then gradually decreased with the consumption of fuel. With the increase in the air supply volume, the development of pool fire was obviously accelerated, and the burning time was shortened. Figure 4 shows the change in the mass loss rate under the condition of different air supply port heights. With the change in air supply inlet height, mass loss rate change process was basically the same.

route after a fire occurs in the cabin. Thermocouple tree 5 is used to monitor the temperature distribution at the left and right ladder openings after a fire occurs in the cabin. Since the left and right are basically symmetrically distributed, only one bunch of thermocouple

route after a fire occurs in the cabin. Thermocouple tree 5 is used to monitor the temperature distribution at the left and right ladder openings after a fire occurs in the cabin. Since the left and right are basically symmetrically distributed, only one bunch of thermocouple

**0 Figure 3. Figure 3.** Variation in mass loss rate with time under different air volume conditions. Variation in mass loss rate with time under different air volume conditions.

**Figure 3.** Variation in mass loss rate with time under different air volume conditions.

**0 100 200 300 400**

**Time (s)**

**12**

*Fire* **2023**, *6*, 16 5 of 14

**Air inlet height h=0.1m Air inlet height h=0.4m Air inlet height h=0.7m**

**Figure 4.** Variation in mass loss rate with time under different air supply inlet heights. **Figure 4.** Variation in mass loss rate with time under different air supply inlet heights. The mass loss rate is time-averaged, and the forced air volume and mass loss rate are

The mass loss rate is time-averaged, and the forced air volume and mass loss rate are dimensionless based on the experimental results in Figure 3 and the experimental results of predecessors [7]. The dimensionless mass loss rate ሶ <sup>∗</sup> can be defined as ሶ <sup>∗</sup> = ሶ /ሶ ௧, where ሶ is experimental mass loss rate, ሶ ௧ is the mass loss rate of the oil pool of diameter *D* in the free atmosphere, ሶ ௧ = ሶ <sup>ஶ</sup> ᇱᇱ (1 − −). The number of air changes N can be defined as = 3600௦/, where ௦ is the supply air volume, is the cabin volume. By introducing a dimensionless length (, ) =∙ ିଵ/ଷ, the relationship between the dimensionless mass loss rate and the number of air changes in this experiment and previous research results can be drawn on a graph, as shown in Figure 5. It can be seen from the figure, under the conditions of the experimental cabin and ventilation configuration, that the fire source mass loss rate and the supplementary air volume have the The mass loss rate is time-averaged, and the forced air volume and mass loss rate are dimensionless based on the experimental results in Figure 3 and the experimental results of predecessors [7]. The dimensionless mass loss rate . *m* ∗ can be defined as . *m* <sup>∗</sup> = . *m*/ . *mth*, where . *<sup>m</sup>* is experimental mass loss rate, . *mth* is the mass loss rate of the oil pool of diameter *<sup>D</sup>* in the free atmosphere, . *mth* = . *m* 00 ∞ 1 − *e* −*kβD Af* . The number of air changes *N* can be defined as *N* = 3600*Vs*/*Vc*, where *V<sup>s</sup>* is the supply air volume, *V<sup>c</sup>* is the cabin volume. By introducing a dimensionless length *f*(*D*, *Vc*) = *D*·*V<sup>c</sup>* <sup>−</sup>1/3, the relationship between the dimensionless mass loss rate and the number of air changes in this experiment and previous research results can be drawn on a graph, as shown in Figure 5. It can be seen from the figure, under the conditions of the experimental cabin and ventilation configuration, that the fire source mass loss rate and the supplementary air volume have the following relationship: dimensionless based on the experimental results in Figure 3 and the experimental results of predecessors [7]. The dimensionless mass loss rate ሶ <sup>∗</sup> can be defined as ሶ <sup>∗</sup> = ሶ /ሶ ௧, where ሶ is experimental mass loss rate, ሶ ௧ is the mass loss rate of the oil pool of diameter *D* in the free atmosphere, ሶ ௧ = ሶ <sup>ஶ</sup> ᇱᇱ (1 − −). The number of air changes N can be defined as = 3600௦/, where ௦ is the supply air volume, is the cabin volume. By introducing a dimensionless length (, ) =∙ ିଵ/ଷ, the relationship between the dimensionless mass loss rate and the number of air changes in this experiment and previous research results can be drawn on a graph, as shown in Figure 5. It can be seen from the figure, under the conditions of the experimental cabin and ventilation configuration, that the fire source mass loss rate and the supplementary air volume have the following relationship:

$$\dot{m}^\* f(D\_\prime V\_c) = -0.21 \text{exp}\left(-\frac{N}{36 \times 3.25}\right) + 0.24\tag{1}$$

**Figure 5.** Relationship between dimensionless mass loss rate and dimensionless air volume. **Figure 5.** Relationship between dimensionless mass loss rate and dimensionless air volume. **Figure 5.** Relationship between dimensionless mass loss rate and dimensionless air volume.

### *3.2. Temperature Distribution* tween the measurement points increased rapidly. This is because the smoke layer accu-

*3.2. Temperature Distribution* 

The heat transfer of the flame to the space gas in the cabin has three ways: conduction, convection and radiation. Among them, the convective heat transfer is the most important, and the convection heat transfer depends on the movement of the smoke and the entrainment formed by the combustion of flame. In the region, the temperature increase is mainly because of smoke. Therefore, analyzing the distribution law of space gas temperature in the combustion process can reflect the distribution characteristics of smoke to a certain extent. Taking the air supply volume of 2.058 m3/s as an example, Figure 6 shows the temperature changes at different heights in position 1 in the cabin. It can be seen from the figure that the temperature changes at different heights were basically the same. The temperature difference between the measurement points below the height of about 1.2 m in the cabin was not large. Above approximately 1.2 m, the temperature difference between the measurement points increased rapidly. This is because the smoke layer accumulates near the fire source, resulting in a higher temperature in the upper part of the cabin, whereas the temperature of the lower cold air layer does not change much. The cold air layer has a relatively obvious interface. mulates near the fire source, resulting in a higher temperature in the upper part of the cabin, whereas the temperature of the lower cold air layer does not change much. The cold air layer has a relatively obvious interface. Figure 7 shows the variation in vertical temperature with time. At the initial moment, the vertical temperature in the cabin was basically the same. After ignition, with the filling of smoke, the temperature of the upper layer gradually increased, whereas the temperature of the lower layer did not change much. The temperature distribution from top to bottom in the cabin could be regarded as a "double-zone distribution". The temperature inflection points of the upper and lower layers decreased gradually with the settlement of the smoke layer. Because of forced air supply and exhaust, the smoke did not settle to the bottom of the cabin, and the vertical temperature maintained a consistent "double-zone distribution". Afterwards, with the consumption of fuel, the power of the ignition source decreased, the height of the smoke layer gradually increased, and the height of the temperature inflection point of the upper and lower layers gradually increased.

The heat transfer of the flame to the space gas in the cabin has three ways: conduction, convection and radiation. Among them, the convective heat transfer is the most important, and the convection heat transfer depends on the movement of the smoke and the entrainment formed by the combustion of flame. In the region, the temperature increase is mainly because of smoke. Therefore, analyzing the distribution law of space gas temperature in the combustion process can reflect the distribution characteristics of smoke to a certain extent. Taking the air supply volume of 2.058 m3/s as an example, Figure 6 shows the temperature changes at different heights in position 1 in the cabin. It can be seen from the figure that the temperature changes at different heights were basically the same. The temperature difference between the measurement points below the height of about 1.2 m in the cabin was not large. Above approximately 1.2 m, the temperature difference be-

*Fire* **2023**, *6*, 16 6 of 14

**Figure 6.** Temperature changes at different heights of the 1# thermocouple tree in the cabin (௦ = 2.058 mଷ/s). **Figure 6.** Temperature changes at different heights of the 1# thermocouple tree in the cabin (*V<sup>s</sup>* = 2.058 m3/s).

Figure 7 shows the variation in vertical temperature with time. At the initial moment, the vertical temperature in the cabin was basically the same. After ignition, with the filling of smoke, the temperature of the upper layer gradually increased, whereas the temperature of the lower layer did not change much. The temperature distribution from top to bottom in the cabin could be regarded as a "double-zone distribution". The temperature inflection points of the upper and lower layers decreased gradually with the settlement of the smoke layer. Because of forced air supply and exhaust, the smoke did not settle to the bottom of the cabin, and the vertical temperature maintained a consistent "double-zone distribution". Afterwards, with the consumption of fuel, the power of the ignition source decreased, the height of the smoke layer gradually increased, and the height of the temperature inflection point of the upper and lower layers gradually increased.

**2.0**

*Fire* **2023**, *6*, 16 7 of 14

**Figure 7.** Vertical temperature change in 1# thermocouple tree (௦ = 2.058 mଷ/s). **Figure 7.** Vertical temperature change in 1# thermocouple tree (*V<sup>s</sup>* = 2.058 m3/s ). Figure 8 shows the variation in the vertical average temperature at different positions

Figure 8 shows the variation in the vertical average temperature at different positions in the cabin. The average temperature here is the average temperature of all thermocouples vertically at this position. It can be seen from the figure that the temperature in the middle part of the cabin was obviously higher than the temperature at the position of the manhole and the entrance of the ladder. Figure 9 shows the variation law of the vertical average temperature at position 1 under different supply air volume conditions. With the Figure 8 shows the variation in the vertical average temperature at different positions in the cabin. The average temperature here is the average temperature of all thermocouples vertically at this position. It can be seen from the figure that the temperature in the middle part of the cabin was obviously higher than the temperature at the position of the manhole and the entrance of the ladder. Figure 9 shows the variation law of the vertical average temperature at position 1 under different supply air volume conditions. With the increase in air supply air volume, the average temperature in the cabin shown a gradually decreasing trend. in the cabin. The average temperature here is the average temperature of all thermocouples vertically at this position. It can be seen from the figure that the temperature in the middle part of the cabin was obviously higher than the temperature at the position of the manhole and the entrance of the ladder. Figure 9 shows the variation law of the vertical average temperature at position 1 under different supply air volume conditions. With the increase in air supply air volume, the average temperature in the cabin shown a gradually decreasing trend.

increase in air supply air volume, the average temperature in the cabin shown a gradually

**0 100 200 300 400 500 Figure 8.** Temperature changes at different positions in the cabin (௦ = 2.058 mଷ/s). **Figure 8.** Temperature changes at different positions in the cabin (*V<sup>s</sup>* <sup>=</sup> 2.058 m3/s ).

**Figure 8.** Temperature changes at different positions in the cabin (௦ = 2.058 mଷ/s).

**Time (s)**

**Figure 9.** Temperature change at position 1 under different air supply volumes. **Figure 9.** Temperature change at position 1 under different air supply volumes.
