There are many benefits to tunnels, but they can also be dangerous. One of these is the chance of tunnel fires. As technology for building tubes improves and more people move into cities, the number of tunnels in China is growing quickly. The CFD methods are substituted with upmarket experimental methods in this field. This is why, on account of the reference paper entitled “Extended CFD models for numerical simulation of tunnel fire under natural ventilation: Comparative analysis and experimental verification”, the present CFD study is conducted.

• Reference [1]: Xu, Tong, et al. “Extended CFD models for numerical simulation of tunnel fire under natural ventilation: Comparative analysis and experimental verification.” Case Studies in Thermal Engineering31 (2022): 101815.

Figure 1: Schematic diagram of three physical models of tunnel [1]

Simulation Process

The model geometry is primarily created using Design Modeler. By taking proper blocking inside ICEM meshing software, a structured grid is performed. The tunnel was 9.0 m long, 0.6 m, 3 wide and 0.45 m high with a rectangular cross-section. The fire source was located in the center of the tunnel. A square ethanol pool fire with a side length of 0.2 m was used to simulate the fire source. The fuel thickness was maintained at 5 cm before ignition.

The combustion rate calculation method figures out the real rate of a chemical process. It uses the Eddy Dissipation model (ED). The chemical reaction can be described by Species Transport Module as:

C2H5OH(l) + 3(O2 + 3.76N2) == 2CO2 + 3H2O(g) + 11.28N2

As a simplification, air is ideally considered as oxygen.

Figure 2: Illustration of AEM (mass sink and mass source) [1]

Post-Processing

The velocity contour visualization reveals critical flow dynamics in the tunnel fire simulation, with maximum velocities reaching approximately 2.88 m/s. As shown in the top image, the fire source at the tunnel center generates a complex buoyancy-driven flow pattern with highest velocities (cyan to light green) concentrating around the fire region and along the ceiling. This demonstrates the classic ceiling jet phenomenon where hot gases spread horizontally along the tunnel ceiling before being guided by the longitudinal ventilation. The visualization reveals recirculation zones forming near the tunnel boundaries, particularly evident in the blue regions where velocities approach zero. These circulation patterns are significant for understanding smoke propagation and potential backlayering effects that could impede evacuation during actual tunnel fire emergencies. The three-dimensional flow field captured by this simulation provides essential insight into the complex turbulent structure that develops under natural ventilation conditions.

Figure 3: Smoke propagation in the tunnel in terms of stream and velocity contours

The CO2 mass fraction contour provides crucial insight into combustion product distribution throughout the tunnel environment. Peak CO2 concentrations (reaching 0.171) appear directly above the fire source where the ethanol undergoes complete combustion according to the simplified reaction equation (C₂H₅OH + 3O₂ → 2CO₂ + 3H₂O). The visualization shows how these combustion products spread through the tunnel via both convective transport and diffusion mechanisms, with the highest concentrations following the ceiling flow paths established by the buoyancy-driven currents. This concentration gradient has significant implications for fire toxicity assessment and tenability conditions within the tunnel during a fire emergency. The longitudinal distribution pattern follows the expected behavior for natural ventilation conditions, where the pressure differences between tunnel ends creates a predominant flow direction that carries combustion products toward the tunnel exits.