In recent years, the rapid advancement of lithium-ion battery technology has significantly transformed the automotive industry, with electric vehicles (EVs) becoming increasingly prevalent. However, this shift has been accompanied by a rise in fire incidents, particularly involving electric SUV models, which pose unique challenges due to their larger size and complex internal structures. As a researcher focused on automotive safety, I have conducted an in-depth numerical simulation to analyze the fire dynamics of a pure electric SUV, specifically examining scenarios initiated by thermal runaway in ternary lithium-ion batteries. This study aims to provide a comprehensive understanding of the combustion characteristics, flame propagation, and smoke behavior in such vehicles, which is critical for enhancing safety protocols and evacuation strategies.
The simulation model was developed based on a full-scale electric SUV, incorporating key combustible components such as seats, door panels, tires, and the battery pack. The battery pack, composed of 192 ternary lithium-ion cells arranged in 32 modules, was identified as the primary ignition source. To accurately represent the fire scenario, I considered the influence of various materials on heat transfer and combustion. The governing equations for fluid dynamics, heat transfer, and combustion chemistry were solved using a large eddy simulation approach, which divides the spatial domain into control volumes with assumed uniform physical properties. The fundamental equations include the continuity equation, momentum equation, energy equation, species transport equation, and the ideal gas law, expressed as follows:
$$ \frac{\partial \rho}{\partial t} + \nabla \cdot (\rho \mathbf{V}) = 0 $$
$$ \frac{\partial (\rho \mathbf{V})}{\partial t} + \nabla (\rho \mathbf{V} \mathbf{V}) + \nabla p = \rho \mathbf{g} + \nabla \tau $$
$$ \frac{\partial}{\partial t} (\rho h) + \nabla \cdot (\rho h \mathbf{V}) = \frac{dp}{dt} + \dot{q}”’ – \nabla \cdot \mathbf{q} + \Phi $$
$$ \frac{\partial}{\partial t} (\rho Y_i) + \nabla \cdot (\rho Y_i \mathbf{V}) = \nabla \cdot (\rho D_i \nabla Y_i) + \dot{m}_i”’ $$
$$ p = \frac{\rho R T}{M} = \rho R T \sum_i \left( \frac{Y_i}{M_i} \right) $$
Here, \( \rho \) represents gas density, \( \mathbf{V} \) is the velocity vector, \( p \) is pressure, \( \mathbf{g} \) is gravitational acceleration, \( \tau \) is the viscous stress tensor, \( h \) is specific enthalpy, \( \dot{q}”’ \) is the volumetric heat source, \( \mathbf{q} \) is the radiative heat flux vector, \( \Phi \) is the dissipation function, \( Y_i \) is the mass fraction of species \( i \), \( D_i \) is the diffusion coefficient, \( \dot{m}_i”’ \) is the generation or consumption rate per unit volume, \( T \) is temperature, \( R \) is the universal gas constant, and \( M \) is the molecular weight. The boundary conditions were set to an initial environmental temperature of 20°C, with open boundaries to allow heat exchange with the surrounding air, mimicking real-world fire conditions.
For the mesh generation, I employed a multi-zone approach to balance computational accuracy and efficiency. The domain was divided into six regions with varying grid densities, as summarized in Table 1. The battery pack and central vehicle areas, where combustion is most complex, used a fine grid size of 0.05 m, calculated based on the characteristic flame diameter formula:
$$ D^* = \left( \frac{Q}{\rho_{\infty} c_p T_{\infty} \sqrt{g}} \right)^{2/5} $$
where \( Q \) is the heat release rate, \( \rho_{\infty} = 1.2 \, \text{kg/m}^3 \) is air density, \( c_p = 1016 \, \text{J/(kg·K)} \) is specific heat capacity, \( T_{\infty} = 293 \, \text{K} \) is ambient temperature, and \( g = 9.81 \, \text{m/s}^2 \) is gravitational acceleration. Other regions, such as the front and rear power compartments, used a coarser grid of 0.1 m, while the upper air spaces utilized a 0.2 m grid to focus on smoke dispersion.
| Mesh Zone | Region Description | Grid Size (m) | Purpose |
|---|---|---|---|
| Mesh1 | Central vehicle and battery area | 0.05 | High-resolution combustion analysis |
| Mesh2 | Front power compartment | 0.1 | Temperature and smoke diffusion |
| Mesh3 | Rear power compartment | 0.1 | Temperature and smoke diffusion |
| Mesh4 | Upper air space (front) | 0.2 | Smoke蔓延 observation |
| Mesh5 | Upper air space (central) | 0.2 | Smoke蔓延 observation |
| Mesh6 | Upper air space (rear) | 0.2 | Smoke蔓延 observation |
The combustion model adopted was a finite-rate combustion approach, which directly defines the heat release rate per unit area and thermal parameters for equivalent combustibles. This model is suitable for complex scenarios like battery and vehicle fires, as it accounts for the intricate reaction processes without requiring detailed elemental composition. Key materials, including the battery electrolyte, seat polyurethane foam, door panel PVC, and tires, were assigned properties based on empirical data, as detailed in Table 2. For instance, the battery electrolyte had a density of 1290 kg/m³, a thermal conductivity of 0.45 W/(m·K), and a heat of combustion of 13,200 kJ/kg, which contributed to the rapid fire development.
| Material | Density (kg/m³) | Thermal Conductivity (W/(m·K)) | Specific Heat Capacity (J/(kg·K)) | Ignition Temperature (°C) | Heat of Combustion (kJ/kg) |
|---|---|---|---|---|---|
| Battery Electrolyte | 1290 | 0.45 | 133.9 | 110 | 13,200 |
| Seat Polyurethane Foam | 28 | 0.05 | 1.0 | 350 | 30,000 |
| Door Panel PVC | 1380 | 0.16 | 1.4 | 507 | 19,000 |
| Tire Rubber | 950 | 0.134 | 1.88 | 350 | 32,000 |
The ignition source was a single battery cell with a heat release rate of 1535.82 kW/m², validated against experimental data to ensure accuracy. As shown in Figure 1, the simulated heat release rate curve for the battery cell closely matched empirical results, confirming the model’s reliability for full-scale simulation. The electric SUV model, with dimensions of 5022 mm × 1962 mm × 1756 mm, was constructed by simplifying CAD data to focus on essential components, and the battery pack was integrated into the chassis, with the ignition point located in the upper-left module to replicate a common thermal runaway scenario.
To monitor the fire progression, I installed multiple thermocouples and heat sensors at critical locations, such as the battery pack, driver’s head, driver’s floor, front power compartment, and rear power compartment. Additionally, 2D slices were placed between the cabin and power compartments to visualize temperature distribution and smoke spread dynamically. The simulation results revealed that the fire in the electric SUV evolved through three distinct stages, characterized by varying heat release rates and flame propagation patterns.
In the first stage, spanning approximately 0 to 60 seconds, thermal diffusion occurred within the battery pack. Following the ignition of a single cell, heat rapidly propagated to adjacent cells, leading to localized fires and visible flames emerging from cracks in the battery casing. Within 10 seconds, intense combustion generated significant smoke, which escaped through wheel arches and gaps in the power compartments. By 35 seconds, flames became apparent beneath the vehicle’s front section, indicating the onset of external fire spread. The heat release rate during this phase increased gradually, as the battery’s internal structure confined the combustion, limiting upward heat transfer.
The second stage, from 60 to 120 seconds, involved the combustion of interior combustibles in the cabin. As the battery thermal runaway intensified, flames and heat ignited materials such as floor coverings and door panels, with fire spreading more rapidly along the roof compared to the floor. Smoke accumulated at the vehicle’s top, gradually obscuring visibility, while flames became prominent in the rear power compartment. At around 70 seconds, elevated temperatures caused window glass to shatter, allowing external air to enter and accelerate combustion. This influx of oxygen led to a phenomenon known as backdraft, significantly increasing the combustion intensity and heat release rate.
The third stage, from 120 to 150 seconds, marked full vehicle involvement. With the cabin engulfed in flames, the fire extended to the power compartments, where ample oxygen supply facilitated vigorous burning. The peak heat release rate reached 5100 kW, consistent with documented ranges for electric vehicle fires, and temperatures exceeded 800°C in critical areas. The flame propagation exhibited a distinct pattern, moving faster toward the front of the electric SUV due to the ignition source’s location, resulting in more severe damage in the front sections compared to the rear.
To quantify the temperature distribution, I analyzed data from the thermocouples, as summarized in Table 3. The battery pack recorded the highest temperature, peaking at 900°C, while the front power compartment reached 650°C, and the rear compartment stabilized at 400°C initially before rising in later stages. The cabin temperature showed a sharp increase during the backdraft phase, highlighting the rapid escalation of fire hazards.
| Location | Peak Temperature (°C) | Time to Peak (s) | Notes |
|---|---|---|---|
| Battery Pack | 900 | 50 | Rapid rise due to thermal runaway |
| Driver’s Head Area | 830 | 110 | Influenced by cabin combustion |
| Driver’s Floor | 300 (initial), 800 (peak) | 120 | Two-phase increase from floor heating |
| Front Power Compartment | 650 | 80 | Early involvement due to proximity to battery |
| Rear Power Compartment | 400 (initial), >650 (late) | 130 | Gradual rise with fire spread |
The heat release rate profile, illustrated in Figure 2, further delineates these stages. The initial slow increase corresponds to battery internal diffusion, followed by a steep climb during cabin combustion, and a plateau as the fire encompassed the entire electric SUV. The total energy released was substantial, emphasizing the severe thermal hazards associated with such incidents.
Smoke analysis revealed critical insights into occupant safety. A smoke detector placed at the driver’s position indicated that smoke entered the cabin within 15 seconds of ignition, reaching 60% concentration in just 5 seconds and completely filling the occupant compartment by 25 seconds. This rapid smoke spread, combined with the enclosed space prior to window breakage, poses a significant threat to evacuation efforts. The smoke density continued to rise even after window failure, due to enhanced combustion, underscoring the importance of early detection and escape in electric SUV fire scenarios.
In conclusion, this numerical simulation effectively captures the complex fire dynamics of an electric SUV initiated by battery thermal runaway. The three-stage combustion process, characterized by progressive flame spread and intense heat release, aligns with real-world observations, validating the model’s accuracy. The findings highlight the elevated temperatures in the battery and front sections, as well as the swift smoke infiltration that endangers occupants. For future safety measures, I recommend integrating advanced fire suppression systems and designing quicker egress protocols for electric SUV models. This study not only enhances the understanding of thermal risks in electric vehicles but also serves as a foundational reference for improving fire safety standards in the automotive industry.
Further research could explore the impact of different battery chemistries or vehicle designs on fire behavior, potentially incorporating experimental validation to refine the models. As the adoption of electric SUVs continues to grow, such insights are vital for mitigating fire-related risks and ensuring passenger safety.