*2.3. Experimental Verification 2.3. Experimental Verification*

A 1:10 model test with the same tunnel section and length was conducted to verify the reliability of the numerical model. The model tunnel is made of aluminum sheet and fireproof glass, and each smoke vent has a push–pull steel plate to control the opening state of the vent. The outlet on one side of the exhaust duct is connected with the smoke exhaust fan, and the outlet of the exhaust duct on the other side is sealed, as shown in Figure 2. The jet fan in Figure 2 was not installed during the test conditions in this paper. The fire source of the model test is liquefied petroleum gas (LPG), and the HRR of the fire source is controlled by a rotameter [39], as shown in Figure 2. Details of the small-scale model tunnel can be found in the author's previous research [20,40]. The HRR and smoke exhaust rate are converted by Froude similarity criterion. A 1:10 model test with the same tunnel section and length was conducted to verify the reliability of the numerical model. The model tunnel is made of aluminum sheet and fireproof glass, and each smoke vent has a push–pull steel plate to control the opening state of the vent. The outlet on one side of the exhaust duct is connected with the smoke exhaust fan, and the outlet of the exhaust duct on the other side is sealed, as shown in Figure 2. The jet fan in Figure 2 was not installed during the test conditions in this paper. The fire source of the model test is liquefied petroleum gas (LPG), and the HRR of the fire source is controlled by a rotameter [39], as shown in Figure 2. Details of the small-scale model tunnel can be found in the author's previous research [20,40]. The HRR and smoke exhaust rate are converted by Froude similarity criterion.

**Figure 2.** Experimental device diagram. **Figure 2.** Experimental device diagram.

In this paper, the numerical model will be verified from the temperature profile and smoke spread. The temperature profile in the tunnel is obtained under the same conditions. The temperature distribution of the numerical results is approximately the same as the model results, as shown in Figure 3a. The temperature difference between the numerical and model results is very small near the tunnel ceiling and floor but relatively large in the middle of the tunnel. The difference between the results of the model test and the numerical simulation is within 10.8 ◦C under the tunnel ceiling. Under the same condition, the smoke spread length in the numerical calculation is like that in the model test, and the maximum difference is 5.4 m. Maximum errors of temperature and back-layering length are 10.3% and 5.1%, respectively, which can be ignored for tunnel engineering. Therefore, the accuracy of the mesh size of the numerical model can be guaranteed.

**Figure 3.** Comparison of experimental and numerical results at 30 MW of Case "4". (**a**) Temperature distribution. (**b**) Smoke spread length. **Figure 3.** Comparison of experimental and numerical results at 30 MW of Case "4". (**a**) Temperature distribution. (**b**) Smoke spread length.

In this paper, the numerical model will be verified from the temperature profile and smoke spread. The temperature profile in the tunnel is obtained under the same conditions. The temperature distribution of the numerical results is approximately the same as the model results, as shown in Figure 3a. The temperature difference between the numerical and model results is very small near the tunnel ceiling and floor but relatively large in the middle of the tunnel. The difference between the results of the model test and the numerical simulation is within 10.8 °C under the tunnel ceiling. Under the same condition, the smoke spread length in the numerical calculation is like that in the model test, and the maximum difference is 5.4 m. Maximum errors of temperature and back-layering length are 10.3% and 5.1%, respectively, which can be ignored for tunnel engineering. Therefore,

the accuracy of the mesh size of the numerical model can be guaranteed.

### **3. Results and Discussion 3. Results and Discussion**

### *3.1. Smoke Spread 3.1. Smoke Spread*

In a tunnel fire, the opening state of the smoke vent is 3, 4, 5, and 6. When opening three smoke vents, there are two situations (Figure 4): (1) Case "3A": two smoke vents are opened near the axial flow fan (upstream) and one is opened near the closed end (downstream); (2) one smoke vent is opened near the axial flow fan and two are opened near the In a tunnel fire, the opening state of the smoke vent is 3, 4, 5, and 6. When opening three smoke vents, there are two situations (Figure 4): (1) Case "3A": two smoke vents are opened near the axial flow fan (upstream) and one is opened near the closed end (downstream); (2) one smoke vent is opened near the axial flow fan and two are opened near the closed end (Case 3B). There are also two situations when opening five smoke vents: (1) three smoke vents are opened near the axial flow fan (upstream) and two are opened near the closed end (downstream) (Case "5A"); (2) two smoke vents are opened near the axial flow fan and three are opened near the closed end (Case 5B).

When the layout of the exhaust vent is Case "3A", the smoke upstream can be controlled within 50 m of the last smoke vent; when the exhaust volume exceeds 220 m3/s, the exhaust volume has little effect on smoke spread upstream. However, exhaust volume has a great influence on smoke spread downstream, and the length of the smoke spread decreases with the increase in exhaust rate, as shown in Figure 5a. When the layout of the smoke vent is Case "3B", the effect of exhaust volume on smoke spread is opposite to that

of Case "3A"; the exhaust volume mainly affects the smoke spread upstream, as shown in Figure 5b. Moreover, the total length of the smoke spread (the sum of the spread lengths of both sides of the fire) in the tunnel for Case "3A" is significantly longer than that of Case "3B" at the same exhaust rate, as shown in Figure 6. Therefore, if only three smoke vents near the fire source are opened, the smoke control effect of Case "3B" is better than that of Case "3A". closed end (Case 3B). There are also two situations when opening five smoke vents: (1) three smoke vents are opened near the axial flow fan (upstream) and two are opened near the closed end (downstream) (Case "5A"); (2) two smoke vents are opened near the axial flow fan and three are opened near the closed end (Case 5B).

*Fire* **2022**, *5*, x FOR PEER REVIEW 6 of 17

**Figure 4.** Layout of different smoke vent positions when 3 smoke vents are opened. **Figure 4.** Layout of different smoke vent positions when 3 smoke vents are opened.

spread length is much longer than when opening one or two smoke vents upstream, and **Figure 5.** *Cont*.

which will cause serious waste. When three smoke vents are opened upstream, the smoke

the total smoke spread length in the tunnel exceeds 300 m (the longest is 388 m) under different exhaust rates. The total length of the smoke spread is more than 200 m in all calculation conditions. From the perspective of smoke diffusion, for a single-side point

**Figure 5.** Smoke distribution with different number of smoke vents and exhaust rate (under each

smoke vent scheme, the smoke exhaust rate is increased from 200 to 280 m³/s).

**Figure 5.** Smoke distribution with different number of smoke vents and exhaust rate (under each smoke vent scheme, the smoke exhaust rate is increased from 200 to 280 m³/s). **Figure 5.** Smoke distribution with different number of smoke vents and exhaust rate (under each smoke vent scheme, the smoke exhaust rate is increased from 200 to 280 m3/s). *Fire* **2022**, *5*, x FOR PEER REVIEW 8 of 17

**Figure 6.** Smoke spreading length on both sides under different exhaust volume. **Figure 6.** Smoke spreading length on both sides under different exhaust volume.

*3.2. Temperature Distribution*  3.2.1. Maximum Temperature When the layout of the smoke vents is Case "5A" and Case "6", the maximum temperature decreases with the increase in exhaust volume; the maximum temperature under the tunnel ceiling is 719–918 °C, as shown in Figure 7. When the layout of the exhaust vent is Case "2", Case "3A", Case "3B", Case "4", and Case "5B", the maximum temperature first increases and then decreases with the increase in exhaust volume. This is because when the smoke is confined near the fire source and sinks to the bottom of tunnel, the fire is wrapped by the smoke and does not burn sufficiently. Increasing the exhaust volume strengthens the air convection in the tunnel and makes the maximum temperature rise. If When four smoke vents are opened, the smoke distribution on both sides of fire is basically symmetrical. When the exhaust volume is more than 220 m3/s, increasing the exhaust volume has no obvious effect on restraining the smoke spread, especially upstream. The smoke can be controlled within 150 m on both sides of the fire. When there are three smoke vents upstream and two downstream (Case 5A), the smoke spread length upstream increases obviously compared with Case "4". When the layout of the exhaust vent is Case "5B", the smoke spread length upstream decreases compared with Case "5A", as shown in Figure 5d,e. The influence of setting an exhaust fan on one side of the tunnel is highlighted. Due to the single-side point exhaust, the velocity of the third exhaust vent upstream is very high, which will help the smoke to spread to the third smoke vent. The smoke control effect of Case "5B" upstream is better than that of Case "5A".

the exhaust volume continues to increase, the velocity at the vent nearest to the fire will make the high-temperature smoke unable to gather in the vault, and the smoke vent far away from the fire will restrict air convection, which will reduce the maximum temperature. When the layout of the exhaust vent is Case "5A" and Case "6", the smoke within the range of the smoke vents will sink to the bottom of the tunnel, as shown in Figure 4. It is difficult for fresh air to reach the fire source with the increase in exhaust volume, so the maximum temperature will not increase suddenly. When six smoke vents are opened, the smoke spreading length is the longest among all the smoke exhaust opening schemes. Most importantly, when the smoke exhaust rate is greater than 240 m3/s, the smoke downstream cannot spread to the last smoke vent, which will cause serious waste. When three smoke vents are opened upstream, the smoke spread length is much longer than when opening one or two smoke vents upstream, and the total smoke spread length in the tunnel exceeds 300 m (the longest is 388 m) under

**Figure 7.** Maximum temperature with different exhaust volume.

different exhaust rates. The total length of the smoke spread is more than 200 m in all calculation conditions. From the perspective of smoke diffusion, for a single-side point exhaust tunnel, the best smoke exhaust effect can be obtained when four exhaust vents are opened, especially for a two-way tunnel. **Figure 6.** Smoke spreading length on both sides under different exhaust volume.

*Fire* **2022**, *5*, x FOR PEER REVIEW 8 of 17

### *3.2. Temperature Distribution 3.2. Temperature Distribution*

### 3.2.1. Maximum Temperature 3.2.1. Maximum Temperature

When the layout of the smoke vents is Case "5A" and Case "6", the maximum temperature decreases with the increase in exhaust volume; the maximum temperature under the tunnel ceiling is 719–918 ◦C, as shown in Figure 7. When the layout of the exhaust vent is Case "2", Case "3A", Case "3B", Case "4", and Case "5B", the maximum temperature first increases and then decreases with the increase in exhaust volume. This is because when the smoke is confined near the fire source and sinks to the bottom of tunnel, the fire is wrapped by the smoke and does not burn sufficiently. Increasing the exhaust volume strengthens the air convection in the tunnel and makes the maximum temperature rise. If the exhaust volume continues to increase, the velocity at the vent nearest to the fire will make the high-temperature smoke unable to gather in the vault, and the smoke vent far away from the fire will restrict air convection, which will reduce the maximum temperature. When the layout of the exhaust vent is Case "5A" and Case "6", the smoke within the range of the smoke vents will sink to the bottom of the tunnel, as shown in Figure 4. It is difficult for fresh air to reach the fire source with the increase in exhaust volume, so the maximum temperature will not increase suddenly. When the layout of the smoke vents is Case "5A" and Case "6", the maximum temperature decreases with the increase in exhaust volume; the maximum temperature under the tunnel ceiling is 719–918 °C, as shown in Figure 7. When the layout of the exhaust vent is Case "2", Case "3A", Case "3B", Case "4", and Case "5B", the maximum temperature first increases and then decreases with the increase in exhaust volume. This is because when the smoke is confined near the fire source and sinks to the bottom of tunnel, the fire is wrapped by the smoke and does not burn sufficiently. Increasing the exhaust volume strengthens the air convection in the tunnel and makes the maximum temperature rise. If the exhaust volume continues to increase, the velocity at the vent nearest to the fire will make the high-temperature smoke unable to gather in the vault, and the smoke vent far away from the fire will restrict air convection, which will reduce the maximum temperature. When the layout of the exhaust vent is Case "5A" and Case "6", the smoke within the range of the smoke vents will sink to the bottom of the tunnel, as shown in Figure 4. It is difficult for fresh air to reach the fire source with the increase in exhaust volume, so the maximum temperature will not increase suddenly.

**Figure 7.** Maximum temperature with different exhaust volume. **Figure 7.** Maximum temperature with different exhaust volume.

When the layout of the exhaust vent is Case "4", the exhaust volume has a great influence on the maximum temperature. When the exhaust volume increases from 220 m3/s to 240 m3/s, the maximum temperature increases by 97 ◦C. Although the maximum temperature decreases rapidly with the increase in the exhaust volume, it is still much higher than that of the other schemes. When the exhaust volume is less than 220 m3/s, the maximum temperature is almost the same as that of the other schemes. If only four smoke vents are opened, the exhaust rate should not be more than 220 m3/s.
