In recent years, the widespread adoption of lithium-ion batteries in various applications, including civil aviation, has raised significant safety concerns due to the risk of thermal runaway. Thermal runaway in lithium-ion batteries can lead to catastrophic events such as fires or explosions, posing severe threats to personnel, property, and operational safety. The early detection of thermal runaway is critical for implementing timely countermeasures, thereby ensuring flight safety and stability. Traditional methods for analyzing gases emitted during lithium-ion battery thermal runaway, such as gas chromatography, mass spectrometry, and spectroscopic techniques, often suffer from limitations like delayed response, poor real-time capability, and low usability. These drawbacks make them unsuitable for the demanding environments of civil aviation, where rapid and accurate detection is paramount. To address these challenges, we have focused on developing advanced sensing technologies, with a particular emphasis on bionic electronic nose systems. These systems utilize arrays of gas sensors to detect multiple gases simultaneously, offering advantages such as high sensitivity, fast response, strong anti-interference capabilities, and cost-effectiveness. This makes them ideal for detecting characteristic gases from lithium-ion battery thermal runaway. In this study, we present the design and optimization of a bionic electronic nose gas chamber specifically tailored for enhancing the detection accuracy and reducing the response time for lithium-ion battery thermal runaway gases. Our approach draws inspiration from the canine nasal structure, known for its exceptional olfactory capabilities, and employs orthogonal experiments and computational fluid dynamics (CFD) simulations to validate the design. The goal is to create a gas chamber that facilitates rapid, uniform gas flow to the sensor array, ensuring full contact and improved performance in monitoring lithium-ion battery safety.
The importance of lithium-ion battery safety cannot be overstated, especially in aviation contexts where these batteries are used in aircraft systems, portable electronic devices, and emergency power sources. Thermal runaway in lithium-ion batteries is a complex process initiated by factors such as overcharging, internal short circuits, or mechanical damage, leading to a rapid increase in temperature and the release of hazardous gases like carbon monoxide (CO), hydrogen, and hydrocarbons. Early warning systems that can detect these gases promptly are essential for preventing accidents. While electronic nose technology has been explored in other fields—such as food quality assessment, environmental monitoring, and fire safety—its application in aviation for lithium-ion battery monitoring remains limited. Previous studies have highlighted the role of gas chamber design in electronic nose performance, as the chamber’s structure directly influences airflow patterns, gas distribution, and sensor response characteristics. For instance, optimizing chamber geometry can enhance gas-sensor interaction, control flow velocity, and protect sensors, thereby improving overall system reliability. In this work, we aim to bridge this gap by designing a bionic gas chamber that mimics the efficient gas-handling capabilities of canine nostrils, offering a simple, cost-effective solution with superior performance for lithium-ion battery thermal runaway detection.
Our design philosophy is rooted in bionics, where we analyze the nasal anatomy and airflow dynamics of dogs, which are renowned for their keen sense of smell. Canine nostrils feature a narrow-front and wide-back structure, with internal components like the nasal septum and turbinates that create turbulent flows, slow down gas movement, and increase pressure to enhance odor detection. By emulating these features, we developed a gas chamber with a rectangular exterior for stability and ease of integration, an inlet designed as a “narrow-front, wide-back” configuration to accelerate and then decelerate airflow, and internal baffles to simulate nasal structures. These baffles guide gas分流, induce vortices, and prolong gas-sensor contact time, thereby increasing gas concentration in the sensor array region. To optimize the chamber dimensions, we employed orthogonal experiments, considering factors such as wall plate length, internal baffle spacing, and inlet velocity. CFD simulations were then conducted to verify the chamber’s effectiveness, comparing it with a conventional square chamber. The results demonstrate that our bionic design significantly improves gas mass fraction at the sensor array, accelerates response times, and ensures uniform gas distribution, making it a promising tool for safeguarding lithium-ion battery applications in aviation and beyond.
The foundation of our bionic electronic nose lies in understanding the canine nasal cavity, which serves as a natural model for efficient gas sampling. Dogs possess a highly specialized olfactory system characterized by a complex internal structure that maximizes gas exposure to olfactory receptors. As shown in prior research, the canine nasal cavity includes regions like the nasal vestibule, maxilloturbinate area, and ethmoid region, which collectively create a labyrinthine pathway for inhaled air. This pathway induces turbulence, reduces flow speed, and increases pressure, allowing for enhanced odorant absorption and discrimination. From a fluid dynamics perspective, the Reynolds number (Re) is a key parameter for characterizing flow regimes. It is defined as:
$$Re = \frac{\rho v L}{\mu}$$
where $\rho$ is the gas density, $v$ is the flow velocity, $L$ is the characteristic length, and $\mu$ is the dynamic viscosity. In canine nostrils, the flow often transitions to turbulence (Re > 4000), which promotes mixing and improves odor detection. We applied these principles to our gas chamber design, aiming to replicate similar flow conditions to benefit lithium-ion battery gas detection. The chamber’s internal layout was conceptualized to mimic the nasal septum and turbinates through strategically placed baffles, which split the incoming gas stream and create recirculation zones. This not only extends the residence time of gases near the sensor array but also elevates local pressure, effectively increasing the concentration of target gases like CO released during lithium-ion battery thermal runaway. The design process involved balancing multiple parameters to achieve optimal performance, as detailed in the following sections.
To systematically optimize the bionic gas chamber, we identified three critical factors that influence its performance: the length of the outer wall plate (Factor 1), the spacing between internal horizontal baffles (Factor 2), and the gas inlet velocity (Factor 3). Each factor was assigned three levels based on preliminary designs and miniaturization requirements, as summarized in the table below. The orthogonal experiment method was chosen for its efficiency in exploring multiple variables with a reduced number of trials, allowing us to determine the best combination for enhancing gas detection in lithium-ion battery monitoring.
| Level | Factor 1: Wall Plate Length (mm) | Factor 2: Baffle Spacing (mm) | Factor 3: Inlet Velocity (m/s) |
|---|---|---|---|
| 1 | 70 | 18 | 4.5 |
| 2 | 80 | 22 | 5.0 |
| 3 | 90 | 26 | 5.5 |
Using an orthogonal array, we conducted nine experiments, with the average pressure in the sensor array region at a simulation time of 0.5 seconds as the performance metric. Higher pressure indicates greater gas concentration, which is desirable for improving sensor response. The experimental setup and results are presented in the following table, where the factors are coded according to their levels.
| Experiment No. | Factor 1 (Level) | Factor 2 (Level) | Factor 3 (Level) | Average Pressure in Sensor Array Region (Pa) |
|---|---|---|---|---|
| 1 | 1 | 3 | 1 | 101350.12 |
| 2 | 3 | 2 | 3 | 101355.22 |
| 3 | 1 | 1 | 3 | 101356.86 |
| 4 | 3 | 3 | 2 | 101350.22 |
| 5 | 2 | 3 | 3 | 101356.50 |
| 6 | 1 | 2 | 2 | 101353.10 |
| 7 | 2 | 2 | 1 | 101346.09 |
| 8 | 2 | 1 | 2 | 101350.45 |
| 9 | 3 | 1 | 1 | 101345.57 |
We performed range analysis on the data to assess the influence of each factor. The range (R) values indicate the magnitude of effect, with larger values signifying greater importance. The results, calculated using statistical software, are shown in the table below.
| Factor | Average Effect at Level 1 (Pa) | Average Effect at Level 2 (Pa) | Average Effect at Level 3 (Pa) | Range (R) (Pa) | Optimal Level |
|---|---|---|---|---|---|
| Factor 1: Wall Plate Length | 101353.36 | 101351.01 | 101350.34 | 3.02 | 1 (70 mm) |
| Factor 2: Baffle Spacing | 101350.96 | 101351.47 | 101352.28 | 1.32 | 3 (26 mm) |
| Factor 3: Inlet Velocity | 101347.26 | 101351.26 | 101356.19 | 8.93 | 3 (5.5 m/s) |
Based on the range analysis, the order of factor significance is: inlet velocity > wall plate length > baffle spacing. The optimal combination was determined as a wall plate length of 70 mm, baffle spacing of 26 mm, and inlet velocity of 5.5 m/s. This configuration maximizes pressure in the sensor array region, thereby enhancing gas concentration for improved detection of lithium-ion battery thermal runaway gases. The finalized bionic gas chamber design features a compact rectangular structure with internal baffles arranged to guide airflow, as illustrated in the conceptual diagrams. The chamber dimensions are detailed below, with all values derived from the optimization process to ensure peak performance in monitoring lithium-ion battery safety.
To validate the effectiveness of our bionic gas chamber, we employed computational fluid dynamics (CFD) simulations using ANSYS Fluent software. The 3D model of the chamber was created in SolidWorks, focusing on the internal flow field where gases interact with the sensor array. For comparison, we also modeled a conventional square gas chamber with similar dimensions but without the bionic features. Both models were meshed with a grid size of 1 mm, incorporating boundary layers to capture near-wall effects. The total mesh elements for the bionic chamber amounted to 428,072, ensuring accurate resolution of flow dynamics. In the simulations, we considered a mixture of air and carbon monoxide (CO), as CO is a key indicator gas released during early stages of lithium-ion battery thermal runaway. The mass fraction of CO was set to 0.1, representing a typical concentration for detection scenarios. The inlet boundary condition was defined as a velocity inlet at 5.5 m/s, while the outlet was set to pressure-outflow conditions. The operating pressure was atmospheric, and we ignored temperature variations to focus on flow and concentration fields. The turbulence model selected was the standard k-epsilon model, which is suitable for high Reynolds number flows (Re > 4000) expected in the chamber. The simulation time was set to 0.5 seconds, sufficient to observe steady-state gas distribution.
The simulation results for the bionic gas chamber revealed favorable flow characteristics. The “narrow-front, wide-back” structure created a velocity gradient, with fast inflow at the inlet that decelerated as the chamber expanded, promoting uniform gas dispersion toward the sensor array. The internal baffles induced vortices in the front region, enhancing mixing and extending gas residence time. In the sensor array zone, flow velocity was moderate but consistent, ensuring stable gas-sensor contact. Pressure distribution showed a gradient with higher pressure in the front part and uniform pressure around the sensors, indicating balanced gas exposure. For the conventional square chamber, flow patterns were less optimal: gas tended to channel through the center, leaving corners with stagnant vortices, and pressure varied unevenly across the sensor array. To quantify performance, we analyzed the CO mass fraction in the sensor array region, defined as $Y_{CO} = m_{CO} / m_{total}$, where $m_{CO}$ is the mass of CO and $m_{total}$ is the total gas mass. The improvement in gas mass fraction was calculated using two metrics: first, the enhancement within the bionic chamber itself, comparing the sensor array region to the overall chamber average; second, the improvement relative to the conventional chamber. The formulas are as follows:
$$ \text{Internal Enhancement} = \frac{Y_{\text{sensor, bionic}} – Y_{\text{avg, bionic}}}{Y_{\text{avg, bionic}}} \times 100\% $$
$$ \text{Comparative Enhancement} = \frac{Y_{\text{sensor, bionic}} – Y_{\text{sensor, conventional}}}{Y_{\text{sensor, conventional}}} \times 100\% $$
At 0.5 seconds, the bionic chamber achieved a CO mass fraction of 0.1000 in the sensor array region, while the average across the entire chamber was 0.0690. This yields an internal enhancement of approximately 44.93%. In contrast, the conventional square chamber had a sensor array CO mass fraction of 0.0909 and an overall average of 0.0925, resulting in no significant internal improvement. The comparative enhancement showed that the bionic chamber outperformed the conventional one by about 10.01%. These values underscore the superiority of our design for lithium-ion battery gas detection. Additionally, we monitored CO mass fraction over time at specific points in the sensor array. The bionic chamber allowed CO to reach detectable levels (mass fraction 0.01) within 0.02 seconds and stabilized at 0.1 by 0.20 seconds, whereas the conventional chamber took 0.05 seconds to reach 0.01 and did not stabilize within 0.5 seconds. This demonstrates faster response times, crucial for early warning in lithium-ion battery applications.
The following table summarizes key simulation outcomes, highlighting the advantages of the bionic gas chamber for lithium-ion battery thermal runaway detection.
| Parameter | Bionic Gas Chamber | Conventional Square Chamber | Improvement |
|---|---|---|---|
| CO Mass Fraction in Sensor Array at 0.5 s | 0.1000 | 0.0909 | +10.01% |
| Average CO Mass Fraction in Entire Chamber | 0.0690 | 0.0925 | -25.41% (lower, but focused on sensor region) |
| Internal Enhancement (Sensor vs. Chamber Average) | 44.93% | -1.73% | Superior gas concentration at sensors |
| Time to Reach CO Mass Fraction 0.01 at Sensors | 0.02 s | 0.05 s | 60% faster |
| Time to Stabilize at CO Mass Fraction 0.1 | 0.20 s | Not achieved in 0.5 s | Significantly quicker response |
| Flow Uniformity in Sensor Region | High (uniform pressure and velocity) | Low (gradients and vortices) | Better sensor accuracy and consistency |
Our bionic gas chamber design offers several mechanistic benefits for detecting lithium-ion battery thermal runaway. By mimicking canine nasal anatomy, we have engineered a system that optimizes gas dynamics to enhance sensor performance. The “narrow-front, wide-back” geometry accelerates incoming gases, creating inertial forces that promote penetration into the chamber, while the subsequent expansion slows flow, allowing more time for gas-sensor interaction. The internal baffles serve multiple functions: they split the gas stream to cover a broader area, induce turbulence for improved mixing, and generate localized high-pressure zones that increase gas density near the sensors. From a fluid dynamics perspective, these features can be described using the Navier-Stokes equations for incompressible flow:
$$ \rho \left( \frac{\partial \mathbf{v}}{\partial t} + \mathbf{v} \cdot \nabla \mathbf{v} \right) = -\nabla p + \mu \nabla^2 \mathbf{v} + \mathbf{f} $$
where $\mathbf{v}$ is the velocity vector, $p$ is pressure, $\rho$ is density, $\mu$ is viscosity, and $\mathbf{f}$ represents body forces. In our chamber, the baffle-induced vortices add complexity to the flow field, enhancing convective mass transfer of gases like CO to the sensor surfaces. The mass transfer rate can be approximated by Fick’s law:
$$ J = -D \nabla Y $$
where $J$ is the diffusion flux, $D$ is the diffusion coefficient, and $Y$ is the mass fraction. By increasing $Y$ in the sensor region through optimized flow, we boost $J$, leading to faster sensor response. This is particularly important for lithium-ion battery monitoring, where early detection of thermal runaway can prevent cascading failures. The chamber’s design also mitigates common issues in electronic nose systems, such as sensor poisoning or drift, by ensuring even gas distribution and reducing stagnant zones where contaminants might accumulate.
The implications of this work extend beyond aviation to any application involving lithium-ion battery safety, such as electric vehicles, energy storage systems, and consumer electronics. Our bionic chamber provides a scalable template for improving electronic nose accuracy in real-time gas detection. Future research could explore adaptive designs that adjust baffle configurations dynamically based on flow conditions or integrate multi-sensor arrays for detecting a wider range of gases emitted from lithium-ion batteries. Additionally, environmental factors like temperature, humidity, and vibration—common in aviation—should be incorporated into simulations to assess robustness. We also plan to fabricate a prototype and conduct experimental validations with actual lithium-ion battery thermal runaway tests, comparing sensor signals between our chamber and conventional designs. This will further verify the practical benefits for lithium-ion battery safety management.
In conclusion, we have successfully designed and optimized a bionic electronic nose gas chamber for enhanced detection of lithium-ion battery thermal runaway gases. Drawing inspiration from canine nasal structures, we employed orthogonal experiments to determine optimal dimensions: a wall plate length of 70 mm, internal baffle spacing of 26 mm, and inlet velocity of 5.5 m/s. CFD simulations confirmed that this design facilitates rapid, uniform gas flow to the sensor array, increasing the CO mass fraction by 44.93% compared to the chamber average and by 10.01% compared to a conventional square chamber. The chamber also reduces response times, with CO detectable within 0.02 seconds and stable readings achieved by 0.20 seconds. These improvements are critical for early warning systems in lithium-ion battery applications, where timely intervention can avert disasters. Our work underscores the potential of bionic approaches in sensor technology and paves the way for more reliable monitoring of lithium-ion battery safety across various industries. As lithium-ion batteries continue to proliferate, advancements in detection methods like ours will play a vital role in ensuring operational safety and preventing thermal runaway incidents.