A Study on the Influence of Mobile Fans on the Smoke Spreading Characteristics of Tunnel Fires

Abstract

Mobile fans, as flexible and convenient new longitudinal ventilation and smoke extraction equipment for tunnels, demonstrate more significant effectiveness in an emergency response to tunnel fires compared to traditional smoke extraction methods. This study employs computational fluid dynamics simulation methods, selecting two fire scenarios to investigate the effects of fan inclined angles and fan airflow volumes on the longitudinal temperature distribution and smoke back-layering length in tunnels. The results indicate that when using mobile fans for longitudinal ventilation in tunnels, at a lower fan airflow volume, the temperature distribution along the longitudinal axis is nearly symmetrical. The fire source and the fan installed in the upstream are within a certain range, and it is more effective to use the horizontal angle for longitudinal ventilation. As the fan airflow volume increases, the back-layering length significantly decreases (210,000 m³/h < V < 270,000 m³/h). However, as the fan flow volume continues to increase (270,000 m³/h < V < 300,000 m³/h), the reduction in the back-layering length becomes less pronounced, the smoke spread distance of the latter is only 11% of that of the former. Therefore, selecting appropriate fan airflow volumes and fan inclined angles them can effectively enhance the performance of tunnel smoke extraction systems. Moreover, by comparing with traditional fans, we find that mobile fans provide an alternative effective strategy during firefighting by allowing adjustments in distance from the fire source and fan inclination angles, enhancing fire suppression effectiveness while reducing energy losses. The research findings can serve as a reference for tunnel fire prevention design.

1. Introduction

In the past few decades, highway tunnels have emerged as an efficient form of infrastructure in modern transportation systems, meeting the growing demand for more efficient urban traffic systems in densely populated megacities [1]. Despite the lower probability of fire incidents in tunnels compared to open spaces, the significant number of tunnels worldwide and the high traffic density result in a considerable number of tunnel fire accidents each year. By the end of 2019, China had completed 19,067 tunnels, with a total length of 18,967 km. Among them, there were 1175 extra-long tunnels with a total length of 5218 km and 4784 long tunnels [2]. While the construction of these tunnels has provided significant convenience to urban transportation systems, the consequences of a fire without prompt and effective control measures could be catastrophic. Not only could it potentially cause extensive damage to tunnel structures [3], but it also significantly increases the risk of casualties [4].

Research indicates that smoke generated by fires is the primary cause of casualties in tunnel fires [5]. Therefore, researchers often focus on developing effective smoke extraction technologies to ensure the safe evacuation of individuals during tunnel fires. Tunnel ventilation methods typically include natural ventilation and mechanical ventilation. However, the impact of natural ventilation on tunnel fires is uncertain and limited, while fixed mechanical ventilation equipment is prone to failure in high-temperature smoke environments. To ensure safe evacuation and firefighting during tunnel fires, mobile fans play a crucial role in controlling smoke and serve as auxiliary ventilation during tunnel emergency rescue operations [6]. Mobile fan smoke extraction prevents smoke backflow and suppresses the spread of smoke from downstream fire areas to upstream areas by generating longitudinal airflow and positive pressure, providing more time for the evacuation of individuals in upstream areas [7]. Due to its excellent operability and low-cost advantages, mobile fan smoke extraction has been widely recognized and adopted as a new type of tunnel ventilation method in practical applications.

Experiments conducted by Ingason et al. [8] in an abandoned highway tunnel demonstrated that deploying a mobile fan inside and outside the tunnel can generate a stable airflow exceeding 3 m/s, which decreases to 2.5 m/s after ignition, confirming the feasibility of the mobile fan system. Kim et al. [9] conducted a series of scale model experiments using mobile fans and found an optimal distance between the mobile fan and the fire source, ensuring the stability of the smoke layer. Zhou et al. [10] carried out full size experiments to investigate the spread characteristics of hot smoke, installing mobile fans at different distances from the fire source. The results showed that if the distance between the mobile fan and the fire source is either too large or too small, the performance of the smoke extraction system (in terms of reducing the maximum temperature and suppressing smoke backflow) is significantly reduced compared to when the distance is within a critical range. Jiang and Ingason [11] developed a one-dimensional lumped theoretical model to describe the transient behavior of flow development inside the tunnel based on full-size experiments carried out in railway tunnels. In the field of tunnel fire research, the temperature distribution of smoke [12,13,14,15] and the length of smoke back-layering [16,17,18,19] are key areas of focus for researchers. However, the above-mentioned research results indicate a relative scarcity of studies on these aspects in fire scenarios involving the use of mobile fans. Therefore, this study aims to utilize numerical simulation methods to investigate the changes in temperature distribution and smoke back-layering length inside tunnels under longitudinal ventilation using mobile fans, and Figure 1 illustrates the flowchart of the methodology employed in the study.

2. Numerical Simulation

Currently, numerical simulation, due to its ability to effectively address the limitations of experimental conditions in reality, has been widely utilized in the field of fire safety engineering and has gained universal adoption and experimentation. FDS (Fire Dynamics Simulator, version 6.7.1) is a computational fluid dynamics software used to simulate fluid motion in fire scenarios. It numerically solves N-S equations, which are applicable to low-speed thermally driven flow of smoke and heat transfer in fires. For large constructions such as highway tunnels, railway tunnels, and drifts, Large Eddy Simulations (LESs) can effectively handle the relationship between turbulence and buoyancy. Relevant scholars have also verified the accuracy of FDSs in tunnel fire simulations and have obtained relatively ideal results [20,21]. Therefore, this study employs the Large Eddy Simulation (LES) method for numerical simulation calculations.

2.1. Model in FDS

In this study, FDS was used to simulate fire scenarios and various operating states of fans in a certain highway tunnel under longitudinal forced ventilation conditions. Additionally, to simplify the model construction, the entire tunnel was approximated as a straight structure. The tunnel cross-section was approximated as a semicircle with a radius of 7 m, resulting in a maximum width and height at the top of 14 m and 7 m, respectively. Considering that the focus of the study is on the fire accident section and its upstream and downstream areas, a portion of the tunnel with a length of 120 m was selected, with the fire source positioned on the tunnel centerline and the fan arranged upstream of the fire source. A layout diagram of the fire scenario in the numerical simulation model is shown in Figure 2.

2.2. Simulation Conditions

In numerical studies of tunnel models, the use of the FDS multi-grid method is more common, and typically, the grid around the fire source location is refined [22,23,24]. In order to conserve computational resources while ensuring simulation accuracy, this study allocated more refined grids near the fire source, with grid sizes set to twice the size of the fire source area for regions further away from the fire source. Therefore, near the fire vicinity, the grid size is set to 0.2 m × 0.2 m × 0.2 m, covering the Y-axis range of 60–80 m. Subsequently, the regions of 0–60 m and 80–120 m along the Y-axis are assigned grid sizes of 0.2 m × 0.4 m × 0.2 m, resulting in a total of approximately 957,600 grid cells. The materials for the tunnel walls and ground are set as concrete, with density, thermal conductivity, and specific heat set to 2280 $\mathrm{k}\mathrm{g}/{\mathrm{m}}^3$, 1.8 $\mathrm{W}/(\mathrm{m}·\mathrm{K}),$ and 1.04 $\mathrm{k}\mathrm{J}/(\mathrm{k}\mathrm{g}·\mathrm{K})$ respectively. The environmental temperature and pressure under all simulation conditions are set to 20 °C and 101.325 kPa, and the smoke production rate is set to 0.015. Both ends of the tunnel are open.

In the design of fire scenarios, to realistically recreate the combustion of flammable materials such as fuel, vehicle interiors, and transported goods during tunnel fires, a propane combustion model and a rapid t² fire are used to simulate the fire source, with a fire growth factor of 0.04689. The fire source is located on the centerline of one side of the tunnel exit, with the fire source area set to 18 m² (for fires involving vehicles such as trucks). Based on research on the equipment of some fire departments, the area of the air supply outlet is set to 6 m² (for high-volume smoke extraction vehicles). The mobile ventilation device is placed on the centerline of the tunnel, with the inclined angles of the mobile ventilation device selected within the range of 0–20 degrees.

The heat release rate (HRR) is a crucial parameter that must be considered during the process of designing numerical simulation cases. Many fire engineers have indicated that the HRR is not only one of the key factors for determining the severity of a fire but also an important criterion for assessing the scale of the fire [25]. In order to comprehensively simulate larger-scale fire accidents such as the burning of trucks, multi-passenger vehicles, and the ignition of fuel, we refer to the standards of the National Fire Protection Association (NFPA) and set 15 MW as the heat release rate for medium-sized and larger fire accidents.

In terms of ventilation control effectiveness, the application of mobile ventilation devices in tunnel fire smoke extraction is mainly influenced by three important factors: the position of the fan, the airflow volume of the fan (the fan outlet area multiplied by the fan outlet velocity), and the fan inclined angles. To optimize the smoke control effectiveness of mobile ventilation devices, we propose the concept of a “characteristic air supply area” based on previous research, which refers to the ratio of the fan outlet area to the tunnel’s cross-sectional area. According to the Froude similarity criterion, a series of dimensionless numbers corresponding to numerical simulations were selected, such as the characteristic air supply area, the distance between the fan and the fire source, and the heat release rate of the fire source, to calculate the effective ventilation airflow of the mobile ventilation device under different fire intensities. Then, a series of preliminary experiments were conducted through numerical simulations to observe the smoke control effectiveness of the corresponding effective ventilation airflow under specific fire intensities. Considering the differences in fan control effectiveness at different inclined angles, four different volumes of fan airflow required to control smoke spread under the same fire intensity were determined. Additionally, we introduced two additional sets of conditions as a control group to investigate the differences between traditional fans and mobile fans in smoke control. The fans are positioned at the top of the tunnel, located 40 m from the fire source, with airflow volumes set at 240,000 m³/h and 300,000 m³/h, respectively. A total of 42 fire cases were simulated, as shown in

Table 1. Configuration of numerical simulation cases.

Test No.	Scenarios	Distance from Fire (m)	Fan Airflow Volume (m3/h)	Inclined Angle (Degrees)
1–20	Scenario 1	40	210,000, 240,000 270,000, 300,000	0, 5, 10, 15, 20
20–40	Scenario 2	50	210,000, 240,000	0, 5, 10, 15, 20
			270,000, 300,000
41–42	Scenario 3	40	240,000, 300,000	0 (traditional)

2.3. Mesh Grid Sensitivity Analysis

The accuracy of the FDS simulation results is closely related to the grid size. In theory, larger grid sizes lead to a poorer accuracy of computed results, while smaller grid sizes often yield more precise outcomes. However, reducing the grid size increases the number of grid cells, thereby potentially limiting computational speed due to constraints in computer performance. Therefore, when using FDS for fire simulation, it is crucial to select an appropriate grid size to achieve accurate simulation results and improve efficiency by reducing the computation time.

The parameter $D^{\mathrm{*}}/\delta x$ can be used to determine the appropriate grid size, where $D^{*}$ is the fire characteristic diameter and $\delta x$ is the grid size, which can be expressed as follows:

$$D^{*}={\left[{\displaystyle \frac{Q}{{\rho }_{\infty }\cdot c_p\cdot T_{\infty }\cdot \sqrt{g}}}\right]}^{2/5}$$

(1)

where $Q$ is the heat release rate of fire, $\mathrm{k}\mathrm{W};{\rho }_{\infty }$ is the density of ambient air, $\mathrm{k}\mathrm{g}/{\mathrm{m}}^3;c_p$ is the specific heat capacity of air at constant pressure, ${\displaystyle \frac{\mathrm{J}}{\mathrm{k}\mathrm{g}·\mathrm{K}}},T_{\infty }$ is the temperature of ambient air, $\mathrm{K}$, and $g$ is the gravitational acceleration, $\mathrm{m}/{\mathrm{s}}^2$.

According to the recommendations provided in the FDS user’s guide, where $D^{*}/\delta x$ typically ranges between 4 and 16, this range generally yields effective results in numerical simulations [26]. Ma and Quintiere [27] found that using a grid size of 0.05 $D^{*}$ in FDS for studying axisymmetric flames resulted in more accurate simulations. Yang et al. [28] showed that using a grid size of 0.1 $D^{*}$ produced flame plume temperatures and velocities that closely matched actual fire experiment results.

In this study, the grid size corresponding to a heat release rate of 15 MW should range from 0.1777 m to 0.7106 m. By comparing the downstream longitudinal temperature distribution of the ceiling under various grid sizes as shown in Figure 3—0.167 m (0.06 D*), 0.20 m (0.07 D*), 0.33 m (0.12 D*), 0.4 m (0.14 D*), and 0.5 m (0.18 D*) —it was observed that reducing the grid size to below 0.2 m had a negligible impact on result accuracy. It was observed that the results from the grid size of 0.2 m was the closest to those from the grid size of 0.1 m compared to the other three grid sizes, including both temperature variation trends and numerical values. The calculations reveal that the relative error between the results from the grid sizes of 0.2 m and 0.1 m ranges from 3% to 12%, with most measurement points showing relative errors within 10%. This discrepancy may arise from the method used to determine the HRR in the FDS input file and the settings of the thermocouple parameters. Additionally, the number of grid cells for the grid size of 0.1 m is approximately 3,006,720, resulting in a computation time almost 2–3 times longer than that of the grid size of 0.2 m. Considering both computational accuracy and computation time, the grid resolution of 0.2 m is selected for the following studies.

2.4. Validation Work

To validate the reliability of using FDS for simulating tunnel fires, we replicated the full-scale experiments conducted by Hu et al. [29]. in an underground passage measuring 88 m in length, 8 m in width, and 2.65 m in height. The fire source consisted of a diesel pool fire formed by four identical disks, positioned along the centerline 9 m from the northern end, with a heat release rate of 1.6 MW and an ambient temperature of 28 °C. Figure 4 illustrates the longitudinal temperature distribution in the tunnel at a heat release rate of 1.6 MW, with the fire source located at “x = 0 m”. It can be observed that near the fire source, the experimental smoke temperature is slightly higher than the simulated smoke temperature, and this difference diminishes gradually as the distance from the fire source increases. Overall, the FDS simulation results for the longitudinal temperature distribution in the full-scale tunnel closely resemble those from the full-scale experiments, thereby validating the reliability of the numerical model.

3. Results

3.1. Smoke Movement Phenomenon Description

Tunnel fires generate large amounts of toxic and harmful hot gasses, which not only damage the tunnel structure but also pose a significant threat to personnel safety. Therefore, it is essential to study the distribution of smoke and temperature inside the tunnel. Figure 5 depicts the smoke spread along the longitudinal center plane of the tunnel under different fan airflow volumes for Scenario 2 (fan located 50 m from the fire source, HRR = 15 MW) When a tunnel fire occurs, the smoke hits the ceiling and dissipates in all directions. Tunnels have a distinct structural feature: they are typically very long in the longitudinal direction and very narrow in the lateral direction and the tunnel cross-section studied in this paper is arc-shaped, with the highest point of the ceiling directly above the centerline. This means that more smoke will spread along the longitudinal direction of the tunnel until it flows out. The entire smoke layer is divided into two parts: the upper flow layer and the lower stable layer. And under the longitudinal airflow provided by the mobile fan, the smoke disperses downstream. The smoke dispersion in the upstream area of the tunnel is significantly controlled. For the downstream region, the increase in fan airflow volume does not noticeably improve visibility, as the smoke almost spreads throughout the entire downstream area of the tunnel. As the fan airflow volume increases, the smoke spread distance gradually decreases. One reason is that the fan continuously supplies fresh air into the tunnel, expelling the high-temperature smoke and reducing its temperature, thereby decreasing the thermal pressure and slowing the spread of smoke. On the other hand, as the fan airflow volume increases, the negative pressure inside the tunnel becomes greater, leading to an increase in the amount of cold air entrained at the exhaust outlet and accelerating the smoke extraction speed, thus confining the smoke to a smaller area.

The temperature contour map on the longitudinal center plane of the tunnel is depicted in Figure 6, under the same conditions as Figure 5. Assuming gasses above 50 °C as the smoke layer [30], At the same heat release rate, the temperature field in the tunnel exhibits an asymmetrical distribution, with the temperature upstream being significantly lower than downstream due to the longitudinal airflow in the tunnel. When the fan airflow volume increases from 24,000 m³/h to 270,000 m³/h, there is a significant reduction in smoke spread distance, decreasing from 37 m to 2 m. However, when the fan airflow volume further increases from 270,000 m³/h to 300,000 m³/h, the smoke spread distance only decreases from 2 m to −2 m. The latter reduction is only 11% of the former, indicating the diminishing returns of increasing the same fan airflow volume. Therefore, from a cost-saving perspective, we should judiciously select the fan airflow volume.

To further analyze the impact of mobile fans on smoke spread within the tunnel, Figure 7 illustrates the longitudinal velocity distribution along the tunnel axis. It can be observed that at a fan deflection angle of 0 degrees, the high-speed airflow at the fan outlet propagates forward along the tunnel floor, gradually diffusing across the entire cross-section. Coupled with Figure 5c, this indicates that the airflow is sufficient to control smoke in the upstream section of the tunnel. As the airflow continues to advance, the obstruction caused by the fire source diminishes the entrainment effect of the fan, relieving the tendency of the high-speed airflow to descend. Consequently, the cross-sectional wind speed downstream of the fire source develops more uniformly, though it is significantly lower than the upstream wind speed.

3.2. Effect of Fan Angle on Ceiling Longitudinal Temperature

To investigate the impact of fan distance and angles on smoke exhaust effectiveness, the distribution of the maximum smoke temperature under the ceiling was used to quantitatively analyze the movement characteristics of smoke at varying distances and angles. The longitudinal temperature distribution along the ceiling of the tunnel is shown in Figure 8 (Scenario 2), and a positive longitudinal distance (“+”) indicates the downstream direction towards the tunnel exit, while a negative longitudinal distance (“−”) indicates the upstream direction. It is observed that the temperature beneath the ceiling decreases with increasing distance from the fire source. The highest temperature does not occur directly above the fire source due to the longitudinal airflow provided by the mobile fan, which directs it towards the downstream area. There exists a deflection length between the location of the highest temperature and directly above the fire source, a phenomenon explained by the theory proposed by Thomas et al. [31]. For the upstream area of the fire source, when the thermal buoyancy generated by the fire equals the resistance of the ventilation airflow, the smoke front ceases to spread, resulting in a sharp decrease in ceiling temperature, eventually approaching the ambient temperature. The longitudinal temperature decay downstream of the fire source is relatively slow. As the fan airflow volume in the tunnel increases, the maximum temperature along the ceiling decreases accordingly, and the location of the highest temperature shifts towards the downstream area, consistently occurring at an inclined angle of 20 degrees.

We attempted to study the effect of different fan inclined angles on the longitudinal temperature decay in the tunnel. For Scenario 2, we observed that at a lower airflow volume, the temperature distribution along the tunnel was nearly symmetrical longitudinally. This is because the longitudinal wind generated by the fan causes less disturbance to the smoke. For the upstream area of the fire source, the fan inclined angle significantly affects the smoke temperature distribution. Under four different airflow volumes, we found that when it was adjusted to the horizontal position (0 degrees), the upstream temperature in the tunnel decreases the slowest. Additionally, there is a significant temperature gradient decay near the fire source; as the airflow volume increases, this effect becomes more pronounced. In the downstream area from the fire source, the influence of different fan inclined angles on the temperature is less pronounced compared to the upstream area. When the fan airflow volume reaches 300,000 m³/h, the maximum temperature beneath the ceiling with a fan inclined angle of 5 degrees is the lowest, only half of that at 20 degrees. This suggests that in this fire scenario, a fan inclined angle of 5 degrees is most effective in suppressing the backflow of hot smoke in the tunnel.

To further analyze the impact of fan inclined angles on longitudinal temperature distribution and the significant temperature gradient decay near the fire source, we examined the temperature contour plots along the tunnel’s longitudinal center plane. Two airflow volumes were selected, one below the critical airflow volume and one above it. Figure 9 and Figure 10, respectively show the temperature contour plots for different fan inclined angles at airflow volumes of 240,000 m³/h and 300,000 m³/h under Scenario 2. As seen in Figure 9, when the fan airflow volume is 240,000 m³/h, it is insufficient to completely suppress the smoke backflow. For the upstream region of the fire source, with fan inclined angles of 10 degrees and 15 degrees, the temperature decay under the tunnel ceiling is more pronounced, resulting in a significant gradient decay near the fire source. This can also be observed in Figure 6b, where the longitudinal temperature upstream of the fire source decays to an ambient temperature more rapidly at these angles. When the angle increases to 20 degrees, the upstream temperature decay slows down again. The fan inclined angle has little effect on the temperature decay in the downstream region of the fire source, where smoke accumulates at the tunnel ceiling due to thermal buoyancy, leading to clear smoke stratification. When the fan airflow volume increases to 300,000 m³/h, only the 0-degree fan inclined angle fails to effectively suppress smoke backflow. The upstream smoke is effectively controlled at all other inclined angles. As analyzed in Figure 6d, the temperature decays fastest at a 5 degrees inclined angle, and smoke stratification in the downstream region is evident, indicating that a 5 degrees fan inclined angle is most effective under this fire scenario. This conclusion aligns with the results discussed earlier.

We would like to explore whether similar patterns exist in Scenario 1. Figure 11 shows the longitudinal temperature distribution under Scenario 1 with fan airflow volumes of 210,000 m³/h and 240,000 m³/h, and at low ventilation volumes, the longitudinal temperature distribution in the tunnel is asymmetrical. The highest temperature within the tunnel appears at an inclined angle of 20 degrees, which is consistent with the previous conclusion, indicating that an inclined angle of 20 degrees is relatively ineffective for controlling smoke. When the fan is 40 m away from the fire source and the fan inclined angle is set to 0 degrees, the temperature of the smoke beneath the ceiling downstream of the tunnel is the lowest. The main reason for this result is that the high-speed airflow generated by the fan outlet needs some distance to dissipate. If the fan has an inclined angle at this point, the high-speed airflow will impact the tunnel ceiling, causing greater energy loss compared to a horizontal angle. This indicates that for mobile fans, using a horizontal angle within a certain range may be a better choice.

Figure 12 illustrates the centerline velocity distribution within the tunnel at a fan airflow rate of 270,000 m³/s under Scenario 2, focusing on different fan inclined angles downstream of the fire source. By comparing and analyzing the centerline velocity at various cross-sections downstream of the fire source, we can further explore the impact of mobile fan positions on the longitudinal temperature distribution within the tunnel. When the fan is located 50 m from the fire source, as the distance increases, the centerline velocity within the tunnel gradually stabilizes. However, higher wind speeds are observed at the top of the tunnel section; this phenomenon is attributed to the arc-shaped physical model used in this study, where the airflow flows along the arc at the tunnel’s top, and centrifugal force causes the fluid to move outward, leading to a decrease in airflow pressure at the top, resulting in an increase in wind speed. As shown in Figure 12b, when the fan inclined angle is set to 5 degrees, higher wind speeds are observed in the top of the tunnel close to the fire source, which helps lower the ceiling temperature downstream; this can be confirmed by Figure 8c.

3.3. Analysis of the Thermal Back-Layering Length

The primary function of the fan is to accelerate air movement, thereby preventing smoke backflow. Therefore, during the smoke exhaust process, we should focus more on the state of smoke backflow. Figure 13 shows a schematic diagram of smoke movement under longitudinal ventilation. After rising due to thermal buoyancy and impacting the ceiling, the smoke will spread longitudinally along the tunnel. The movement of smoke upstream of the tunnel will be suppressed by the longitudinal airflow from the fan. Since the kinetic energy provided by the constant heat release rate is fixed, part of the kinetic energy of the smoke will be converted into frictional work and dissipated during movement [32]. When the dynamic force of the smoke flow balances with the ventilation static pressure, the reverse flow of smoke stops, and the distance from the smoke front to the fire source is referred to as the back-layering length.

A common method used in numerical simulations to determine the smoke back-layering length is the ceiling temperature tracking method. Minehiro et al. proposed that the distance from the location where the temperature increases by 5 K upstream to the fire source is the smoke back-layering length [33]. When the fan airflow volumes are 210,000 m³/h and 300,000 m³/h, the back-layering lengths measured using the temperature tracking method are 28.5 m and 12.5 m, respectively, as shown in Figure 14.

As shown in Figure 15, we selected two fan airflow values the same as in Section 3.2, one below the critical airflow and one above it, to explore the influence of different fan inclined angles on the back-layering length. We observed that under both airflow volumes, different fan inclined angles significantly affected the backflow of smoke. An inclined angle of 0 degrees shows the weakest suppression effect on smoke, with a back-layering length of 50 m, even when the airflow volumes reaches 300,000 m³/h. The distance of the fan from the fire source seems to have little effect on the back-layering length, although we only examined two distances in this study. When the airflow volume is below the critical volume, the back-layering length decreases with increasing inclined angles, at an inclined angle of 10 degrees, the back-layering length is the shortest, and further increasing the angle results in minimal change in the back-layering length. When the fan airflow exceeds the critical airflow, the shortest back-layering length of the smoke occurs at an inclined angle of 5 degrees, even reaching negative values. This indicates that a reasonable fan inclined angle should be selected according to different fan airflow volumes.

In practical engineering applications, the rapid control of smoke spread is crucial. When the fan airflow volume is 300,000 m³/h and the fan is 40 m away from the fire source, we observe a significant decrease in the smoke back-layering length when the fan inclined angle increases from horizontal to 5 degrees, dropping from 46 m to −3 m. As the inclined angle increased further, the smoke back-layering length also increased. At the critical airflow volume, an inclined angle of 5 degrees shows the most effective control of smoke. Increasing the angle beyond this point results in an increase in the back-layering length. “Fan 40” means the distance between the fire source and movable fan is 40 m. For the “Fan 50” condition, increasing the inclined angle from 5 degrees to 20 degrees led to an increase in the back-layering length from −2 m to 13 m, indicating a noticeable difference. At an inclined angle of 20 degrees, the distance of the fan from the fire source also significantly influences the smoke back-layering length; specifically, under “Fan 50” conditions, the back-layering length increases by 12 m compared to “Fan 40”. In conclusion, at the critical airflow volume, a fan inclined angle of 5° demonstrates the optimal control of smoke.

Figure 16 shows the change in the back-layering length of smoke under two fan inclined angles. When the heat release rate of the fire in the tunnel remains constant, it means that the total amount of smoke generated by the combustion and the kinetic energy driving the spread of smoke are fixed. Consequently, in the absence of longitudinal airflow, the velocity of smoke reverse flow remains constant. As the ventilation speed inside the tunnel increases, the resistance of longitudinal airflow to smoke also increases, leading to a more pronounced inhibition effect on smoke reverse flow. Therefore, for both “Fan 40” and “Fan 50” conditions, we observed a significant decrease in the back-layering length with increasing tunnel airflow volumes, but this decrease was more pronounced when the inclined angle was 5 degrees. At lower airflow volumes, an inclined angle of 10 degrees appeared to have a better smoke suppression effect, with the back-layering length almost half of that of 5 degrees. As the airflow volume increased, this “difference” gradually diminished, and eventually, the back-layering length at 5 degrees became even shorter than that at 10 degrees. Based on the analysis above, we can provide some recommendations for smoke control using mobile fans in tunnel fire scenarios: when a fire occurs in the tunnel, activating mobile fans and dynamically adjusting the fan inclined angle as the exhaust air volume increases will yield better control effects.

3.4. Comparison of Traditional and Mobile Fans in Smoke Control

The longitudinal ventilation smoke extraction method is one of the most common and simplest tunnel ventilation systems. This approach primarily relies on pressurized fans or jet fans installed beneath the tunnel ceiling to compress the air within the tunnel, facilitating its longitudinal flow to ventilate the space and expel high-temperature smoke. However, a significant drawback of this method is that the longitudinal airflow may cause flames and toxic smoke to spread downstream, thereby increasing the risk of fire spread. Additionally, excessively high airflow speeds can disrupt the stratification of smoke layers, negatively impacting personnel evacuation and rescue operations. Actual tunnel fires can occur at any location within the tunnel; however, in traditional longitudinal ventilation smoke extraction systems, fans can only be fixed at specific points within the tunnel. Additionally, high-temperature environments can damage fixed mechanical smoke extraction systems, such as exhaust fans and jet fans. Clearly, these fixed systems cannot meet the smoke control requirements during tunnel fires. In contrast, mobile fans offer greater flexibility; during firefighting operations, their position relative to the fire source and the angle of inclination can be adjusted according to the fire conditions. In the following sections, we will explore the smoke control effectiveness of traditional fans versus mobile fans based on these two factors.

In addition to the previously discussed ceiling temperature and smoke spread distance within the tunnel, the temperature distribution at a height of 2 m in the downstream space is also a critical factor affecting personnel safety. Figure 17 illustrates the temperature distribution at a height of 2 m with traditional fans and mobile fans in the downstream space. It is observed that relatively high temperatures occur near the fire source, while the temperature remains below 35 °C at a distance of 50 m from the fire source. It is generally accepted that temperatures exceeding 60 °C at a height of 2 m pose a threat to personnel safety [34]. Whether using traditional fans or mobile fans for ventilation and smoke extraction, the temperatures at a height of 2 m in the downstream space at different distances from the fire source are closely related to the airflow volume from the fan. When the smoke extraction airflow is relatively low, the temperature distribution in the downstream space is more uniform. As the distance from the fire source increases, the temperature gradually decreases, remaining below 35 °C at a distance of 10 m from the fire source. However, as the smoke extraction airflow increases, the temperature gradient near the fire source becomes more pronounced, leading to an overall rise in temperature in the downstream space of the tunnel. This results in a greater concentration of smoke settling in lower areas, resulting in a non-uniform distribution.

Notably, when using a traditional fan with an airflow of 300,000 m³/h for smoke extraction, the temperature values remain relatively high, often exceeding 50 °C over considerable distances. This poses a risk of high-temperature injuries to evacuating personnel. The lower temperature at the exit is due to the heat exchange between the hot smoke and the cooler air from outside the tunnel. This indicates that a higher smoke extraction airflow is not always beneficial; excessively high longitudinal airflow can create inertial forces that disrupt smoke stratification in the downstream tunnel space, which is detrimental to the evacuation of individuals entering the downstream area in a state of panic. It is also possible that due to the Coanda effect, the high-speed airflow generated by the fan must travel a considerable distance along the tunnel ceiling to establish a uniform airflow across the entire cross-section of the tunnel [35].

When employing a traditional fan with an airflow of 240,000 m³/h at a distance of 40 m from the fire source, the downstream temperatures are generally slightly higher compared to those observed with a mobile fan at the same distance. This is attributed to the positioning of traditional mobile fans at the top of the tunnel, where the conical airflow generated by the smoke extraction fan experiences friction against the tunnel walls, resulting in energy loss. Consequently, the conical airflow fails to fully cover the tunnel cross-section, reducing the efficiency of the fan. Furthermore, the lowest temperatures are recorded when using a mobile fan positioned 50 m from the fire source, indicating that there is a critical range for the distance of the mobile fan from the fire source. This is related to the previously analyzed longitudinal temperature profile beneath the ceiling.

Figure 18 depicts the effect of different deflection angles of mobile fans on the temperature distribution at a height of 2 m in the downstream area, with traditional fans serving as a control group. To facilitate the comparison, all fans are located 40 m from the fire source, with a smoke extraction airflow of 240,000 m³/h. The deflection angle of the mobile fan significantly impacts the temperature distribution downstream; specifically, when the fan is set to a deflection angle of 20 degrees, the temperature at the tunnel exit reaches its highest point, exceeding that at a 0-degree angle by over ten degrees Celsius. This indicates that an excessively large deflection angle is unfavorable for effective smoke control. Furthermore, compared to traditional fans, the rate of temperature decrease at a height of 2 m with deflection angles of 0 degrees and 10 degrees is more pronounced. This phenomenon can be explained by the fact that when the fan’s outlet angle is adjusted, the high-speed airflow generated by the fan collides with the ceiling, resulting in increased energy losses, as previously discussed regarding vertical temperature distribution. Additionally, the Coanda effect can also lead to a significant reduction in the thrust produced by the fan [36]. Appropriately directing the jet fans towards the lower section can overcome the Coanda effect [37].

In summary, mobile fans, compared to traditional fixed smoke extraction facilities, provide greater flexibility during firefighting operations. They can be relocated within the tunnel to adjust their distance from the fire source, aiming to achieve the optimal extinguishment position. Additionally, the angle of inclination of the fans can be modified, facilitating effective fire suppression while reducing energy losses. The current research does not account for potential external conditions, such as shafts, that may influence the spread of smoke within the tunnel. Furthermore, it does not consider the potential presence of slopes or curvature in actual tunnel conditions. In the future, it is necessary to investigate the arrangement of multiple fans within the tunnel to explore how fan position can enhance smoke extraction efficiency.

4. Conclusions

Based on FDS (Fire Dynamics Simulator, version 6.7.1), this study established a full-scale numerical model of a highway tunnel to investigate the effects of different distances between the fan and the fire source, various fan airflow volumes, and fan inclined angles on longitudinal temperature distribution and smoke back-layering length. The main conclusions are as follows:

(1)

As the fan airflow volume increases, the smoke spread distance continuously decreases. However, the rate of reduction in smoke spread distance also diminishes, with the reduction rate dropping by 89% for the same increase in fan airflow volume.

(2)

The longitudinal temperature distribution within the tunnel is nearly symmetrical when a smaller fan airflow volume is used. When the fan airflow volume is 300,000 m³/h, the maximum smoke temperature at a fan inclined angle of 5 degrees is almost half of that at an inclined angle of 20 degrees. Within a certain range of distances between the fan and the fire source, a horizontal angle may be a better choice.

(3)

The inclined angle of the fan significantly impacts the length of smoke back-layering. When increasing the fan airflow volume, appropriately adjusting the inclined angle can enhance the control effect on the smoke.

(4)

In contrast to traditional fixed smoke extraction facilities, mobile fans provide enhanced adaptability in firefighting scenarios, allowing for adjustments in both their distance from the fire source and their tilt angles. This flexibility improves smoke control and contributes to more effective fire suppression strategies.