Thermal Management of Fuel Cells in Hydrogen-Powered Unmanned Aerial Vehicles

Abstract

Hydrogen-powered unmanned aerial vehicles (UAVs) offer significant advantages, such as environmental sustainability and extended endurance, demonstrating broad application prospects. However, the hydrogen fuel cells face prominent thermal management challenges during flight operations. This study established a numerical model of the fuel cell thermal management system (TMS) for a hydrogen-powered UAV. Computational fluid dynamics (CFD) simulations were subsequently performed to investigate the impact of various design parameters on cooling performance. First, the cooling performance of different fan density configurations was investigated. It was found that dispersed fan placement ensures substantial airflow through the peripheral flow channels, significantly enhancing temperature uniformity. Specifically, the nine-fan configuration achieves an 18.5% reduction in the temperature difference compared to the four-fan layout. Additionally, inlets were integrated with the fan-based cooling system. While increased external airflow lowers the minimum fuel cell temperature, its impact on high-temperature zones remains limited, with a temperature difference increase of more than 19% compared to configurations without inlets. Furthermore, the middle inlet exhibits minimal vortex interference, delivering superior thermal performance. This configuration reduces the maximum temperature and average temperature by 9.1% and 22.2% compared to the back configuration.

1. Introduction

Hydrogen fuel cells generate electrical power for loads by converting chemical energy into electricity through an electrochemical reaction between hydrogen and oxygen. They offer significant advantages, including high energy density, low noise emission, and environmental sustainability [1]. Compared to other fuel cell types, such as solid oxide fuel cells (SOFCs) and methanol fuel cells, hydrogen fuel cells exhibit higher efficiency, lower weight, faster dynamic response, and shorter startup time [2]. Furthermore, hydrogen fuel cells possess a higher energy density than conventional lithium-ion batteries. Consequently, hydrogen-powered UAVs can achieve significantly longer endurance than traditional battery-powered UAVs, demonstrating considerable application potential. However, hydrogen fuel cells generate substantial heat during operation, leading to significant thermal management challenges. Moreover, the fuselages of HUAVs are characterized by compact space and complex structures. During cruise missions, the fuel cells operate continuously for extended periods, producing considerable heat within a confined space available for heat dissipation. Thus, an efficient thermal management system (TMS) is essential to dissipate heat from the fuel cells. This maintains the cells within their optimal operating temperature range, ensuring the operational efficiency of long-endurance UAVs throughout their mission profile.

During operation, a fuel cell converts approximately 50% of its input energy into electrical power [3], with the majority of the remaining energy dissipated as waste heat. The operating temperature range for hydrogen fuel cells typically lies between 60 °C and 80 °C [4,5]. Temperatures that either exceed or fall below this optimal range are detrimental to performance. Specifically, excessively high temperatures cause excessive water evaporation within the cell, leading to membrane dehydration [6] and potential damage to fuel cell components [5]; conversely, excessively low temperatures result in water accumulation, causing “flooding” [6,7]. Consequently, both insufficient and excessive cooling can thus harm hydrogen fuel cells. Furthermore, non-uniform temperature distribution and excessive internal temperature gradients can lead to issues such as non-uniform current density distribution [8], which also significantly impairs normal operation, making an efficient and appropriate cooling strategy crucial. Currently, the predominant cooling methods for hydrogen fuel cells are liquid cooling and air cooling. Research on liquid cooling primarily focuses on numerical models for liquid-cooled fuel cell systems [9,10], the design of liquid cooling flow channels [8,11,12,13], and investigations into the influence of coolant temperature on fuel cell power output and operating temperature [14,15,16]. Similarly, research on air cooling encompasses modeling studies of air-cooled fuel cells [6,7,17,18,19], studies on air cooling flow channels [20], and research on associated cooling components such as cooling fans [21,22,23,24,25] and heat sinks [26,27,28,29]. Although liquid cooling generally offers superior thermal performance, it requires additional auxiliary equipment which increases the overall volume and weight of the fuel cell cooling system (including the fuel cell and liquid-cooling device). In contrast, air cooling utilizes ambient air directly as the coolant, simultaneously supplying oxygen for the electrochemical reaction and removing waste heat, while occupying less space and exhibiting a simpler structure.

While numerous studies have investigated air cooling for hydrogen fuel cells, the majority focus on thermal management and temperature control of the fuel cell stack itself. Research specifically addressing the cooling challenges of hydrogen fuel cells within the practical application context of UAVs remains relatively limited. The primary air cooling methods employed for hydrogen-powered UAVs include fan cooling, heat sink cooling [30], fuselage venting [31], and air inlet cooling [32]. Additionally, Zhou et al. [33] positioned the hydrogen fuel cells and propulsion system at the nose of the UAV, utilizing a direct air intake design for cooling. The cooling effectiveness varies significantly among these different approaches. This study focuses on investigating the air cooling performance of the fuel cell system in a hydrogen-powered UAV. A numerical model of the UAV TMS is developed, and transient computational fluid dynamics (CFD) simulations using ANSYS Fluent 2023R1 are performed to analyze the heat dissipation performance under various flight conditions. Building upon simulations of fan-only cooling, the research further examined the combined cooling strategy integrating fans with air inlets. In summary, this work provides valuable insights for the thermal management research of fuel cells in hydrogen-powered UAVs.

2. Thermal Management Model for Hydrogen-Powered UAVs

2.1. Introduction to the Hydrogen-Powered UAV Model

The TMS geometry for the hydrogen-powered UAV is constructed in the three-dimensional modeling software CATIA V5R20 (Figure 1). The fuselage dimensions are 1539 mm in length and 564.4 mm in width, with the fuel cell stack positioned at the back section. Initially, a configuration employing four cooling fans is implemented. Given the complexity of electrochemical models for fuel cells, developing a comprehensive electrochemical heat generation model for a multi-cell hydrogen fuel cell stack would be both intricate and computationally demanding. Since this study focuses on the influence of different UAV flight conditions on fuel cell cooling performance rather than the intricate internal electrochemical processes, a simplified approach is adopted. Airflow channels within the fuel cell are represented by perforations machined into a solid block, treating the entire assembly as a volumetric heat source. This simplification effectively captures the essential heat dissipation characteristics. The overall dimensions of the modeled fuel cell block are 250 mm × 250 mm × 80 mm. It features a uniformly distributed 10 × 10 array of perforations, each with dimensions of 10 mm × 10 mm.

2.2. Simulation Methodology

Governing Equations

The governing equations for the fluid domain of UAV in the simulations—comprising the continuity Equation (1), momentum conservation Equation (2), and energy conservation Equation (3)—are given as follows [34]:

$${\displaystyle \frac{\partial \rho }{\partial t}}+\nabla \cdot (\rho \overrightarrow{u})=0$$

(1)

$${\displaystyle \frac{\partial \overrightarrow{u}}{\partial t}}+(\overrightarrow{u}\nabla )\overrightarrow{u}=-{\displaystyle \frac{\nabla p}{\rho }}+{\displaystyle \frac{\mu }{\rho }}{\nabla }^2\overrightarrow{u}$$

(2)

$$\rho c_p\left({\displaystyle \frac{\partial e}{\partial t}}+u_x{\displaystyle \frac{\partial e}{\partial x}}+u_y{\displaystyle \frac{\partial e}{\partial y}}+u_z{\displaystyle \frac{\partial e}{\partial z}}\right)=k\left({\displaystyle \frac{{\partial }^{2}e}{\partial x^2}}+{\displaystyle \frac{{\partial }^{2}e}{\partial y^2}}+{\displaystyle \frac{{\partial }^{2}e}{\partial z^2}}\right)$$

(3)

where ρ, t, u, p, μ, c_p, and e represent the air density (kg·m⁻³), time (s), velocity vector (m·s⁻¹), pressure (Pa), dynamic viscosity (kg·m⁻¹·s⁻¹), specific heat capacity (J·kg⁻¹·K⁻¹), and total energy (J), respectively. The k-omega SST turbulence model is chosen for its enhanced stability in predicting air flow and its high accuracy in simulating flow near wall surfaces, with its transport equations are given by [35]:

$${\displaystyle \frac{\partial (\rho k)}{\partial t}}+{\displaystyle \frac{\partial \left(\rho U_{i}k\right)}{\partial x_i}}={\stackrel{\sim }{P}}_k-{\beta }^{*}\rho k\omega +{\displaystyle \frac{\partial k}{\partial x_i}}\left[\left(\mu +{\sigma }_k{\mu }_t\right){\displaystyle \frac{\partial k}{\partial x_i}}\right]$$

(4)

$${\displaystyle \frac{\partial (\rho \omega )}{\partial t}}+{\displaystyle \frac{\partial \left(\rho U_i\omega \right)}{\partial x_i}}=\alpha \rho S^2-{\beta }^{*}\rho {\omega }^2+{\displaystyle \frac{\partial }{\partial x_i}}\left[\left(\mu +{\sigma }_{\omega }{\mu }_t\right){\displaystyle \frac{\partial \omega }{\partial x_i}}\right]+2\left(1-F_1\right)\rho {\sigma }_{\omega 2}{\displaystyle \frac{1}{\omega }}{\displaystyle \frac{\partial k}{\partial x_i}}{\displaystyle \frac{\partial \omega }{\partial x_i}}$$

(5)

where the blending function F₁ is defined by:

$$F_1=\tanh \left\{{\left\{\min \left[\max \left({\displaystyle \frac{\sqrt{k}}{{\beta }^{*}\omega y}},{\displaystyle \frac{500v}{y^2\omega }}\right),{\displaystyle \frac{4\rho {\sigma }_{\omega 2}k}{C{D}_{k\omega }{y}^2}}\right]\right\}}^4\right\}$$

(6)

And y represents the distance to the nearest wall. The cross-diffusion term CD_kω is given by:

$$C{D}_{k\omega }=\max \left(2\rho {\sigma }_{\omega 2}{\displaystyle \frac{1}{\omega }}{\displaystyle \frac{\partial k}{\partial x_i}}{\displaystyle \frac{\partial \omega }{\partial x_i}},{10}^{-10}\right)$$

(7)

2.3. Computational Grid Configuration

The complete model is imported into ANSYS SpaceClaim for fluid domain extraction, followed by the generation of an unstructured mesh using Fluent Meshing (Figure 2). Local grid refinement is applied specifically to the fuel cell region to ensure computational accuracy. The physical fan assemblies are replaced with fan boundary condition models, which simulate airflow generation by applying a pressure jump across designated planar surfaces representing fan locations [36]. This modeling approach replicates the aerodynamic effects of actual fans through momentum source terms that increase static pressure across the interface.

2.4. Assumptions and Boundary Conditions

During the computational process, the following assumptions are made: air is treated as an incompressible fluid; thermal radiation effects are neglected; the fuel cell casing is ignored; and the hydrogen fuel cell is simplified as a constant-power heat source with a uniform volumetric heat generation rate of 120,000 W/m³, based on the heat generation conditions of the actual UAV fuel cell. Relevant thermophysical properties [31] are assigned as specified in

Table 1. Physical parameters of hydrogen fuel cells.

Specific Heat Capacity/(J∙(kg∙K)−1)	Density/(kg∙m−3)	Heat Conductivity Coefficient/(W∙(m∙K)−1)
710	1850	10

.

The inlet boundary condition for the fluid domain is defined as a velocity inlet with a magnitude of 10 m/s oriented in the nose-to-tail direction. The inlet air temperature is set to 20 °C, consistent with the initial temperature of the hydrogen fuel cell. The outlet boundary is specified as a pressure outlet. All wall boundaries are modeled with stationary no-slip conditions. Simulations are conducted using the k-omega SST turbulence model with the energy equation enabled.

2.5. Grid Independence and Model Validation

To ensure solution accuracy independent of spatial discretization, a grid independence study is carried out. Four distinct meshes with varying resolutions are generated for the entire fluid domain. Simulations are conducted under cruise conditions at a 0° angle of attack and an airspeed of 10 m/s. As shown in Figure 3, the predicted thermal parameters exhibit minimal variation when the mesh count increases from 1.8 million to 2.5 million elements. Consequently, the 1.8-million-element mesh is chosen for all subsequent computations, as it achieves an optimal balance between numerical accuracy and computational efficiency.

3. Research on Influential Factors of Hydrogen Fuel Cell Cooling Effectiveness

3.1. Effect of Fan Distribution Density on Fuel Cell Temperature at Constant Total Active Area

Transient thermal simulations are performed to analyze fuel cell temperature evolution over 900 s under three different fan density configurations: one-fan, four-fan, and nine-fan configurations. To isolate the effects of fan density configurations, the total active cooling area is held constant across all configurations (Figure 4), with an identical pressure jump of 50 Pa applied uniformly on all fan surfaces. Under the condition of only fan cooling, air enters the interior of the UAV fuselage through the gaps between the fans, the hydrogen fuel cell, and the UAV. As shown in Figure 5, while the average fuel cell temperature exhibits minimal sensitivity to fan density during this period, the maximum temperature decreases progressively from 64.3 °C (one-fan) to 63.3 °C (four-fan) and 62.1 °C (nine-fan). More significantly, the temperature non-uniformity varies across configurations. The four-fan configuration yields the highest temperature difference of 18.9 °C, whereas the nine-fan configuration achieves the most uniform thermal configuration with a difference of only 15.4 °C—representing an 18.5% reduction compared to the four-fan case.

Figure 6 illustrates the temperature distributions under different fan configurations. In Figure 6a, the one-fan configuration results in significantly lower temperatures at the center of fuel cell compared to peripheral regions, with maximum temperatures occurring at the front lateral positions. This configuration exhibits substantially greater thermal non-uniformity compared to the four-fan and nine-fan configurations (Figure 6b,c). Figure 6b shows a localized hotspot in one quadrant under four-fan cooling setup. Cross-sectional flow analysis (Figure 7) identifies a recirculation zone formation at this elevated temperature region, where reduced air mobility leads to heat accumulation and consequent temperature rise. Conversely, Figure 6c demonstrates maximum temperature concentration at the central front section, spatially distinct from the one-fan pattern.

Figure 8 illustrates the temperature and streamline distributions at the 900-s mark on the fuel cell symmetry plane and peripheral regions (positions indicated by black lines in Figure 6c). Unlike the four-fan configuration, where the central zone lacks active cooling, both one-fan and nine-fan configurations enable effective heat dissipation in this critical area, maintaining lower temperatures than the four-fan case. In the one-fan configuration (Figure 8a), cooling concentrates in the central region, achieving sufficient airflow and uniform surface temperatures across the fuel cell.

For the four-fan configuration (Figure 8c), the absence of fan influence on the symmetry plane allows external air to enter beneath the fuel cell, cooling through direct convective heat transfer. This airflow primarily affects posterior and inferior regions, creating a thermal distribution characterized by lower temperatures inferiorly and posteriorly, but elevated temperatures anteriorly and superiorly. The larger high-temperature zone compared to Figure 8a,e demonstrates limitations of relying solely on external airflow for cooling during flight.

In Figure 8e, reverse airflow (tail-to-nose direction) shows wall-adherent front flow above the fuel cell. A large-scale vortex obstructs frontal cooling, causing elevated frontal temperatures, reduced posterior temperatures, and stepwise thermal distribution along the fan axis. Analysis of the peripheral regions reveals that Figure 8b depends exclusively on external airflow for cooling. Vortices obstruct multiple flow channel inlets (except one major vortex between the fuel cell and fuselage nose), significantly impairing heat dissipation. Figure 8f shows residual fan influence partially improving conditions: the nose vortex disappears, enhancing airflow compared to Figure 8b and reducing temperatures. However, an upper vortex traps heat in the back fuselage, causing a progressive temperature rise that is detrimental to fuel cell cooling. Conversely, Figure 8d is less affected by vortex interference at flow channel inlets than Figure 8b,f. Sustained high-velocity airflow maintains strong convective heat transfer, resulting in a lower temperature.

3.2. Effect of Inlet Height on Fuel Cell Temperature

Installing inlets on the hydrogen-powered UAV fuselage can increase airflow, potentially reducing fuel cell temperatures. An inlet with insufficient height restricts the inflow of external cooling air into the fuselage, thereby limiting cooling effectiveness. In contrast, an overly high air inlet increases aerodynamic drag and results in additional energy consumption. To address this trade-off, three inlet heights (20 mm, 30 mm, and 40 mm) are designed in fixed increments, guided by engineering practice and overall aircraft design considerations (Figure 9a), ensuring effective heat dissipation while avoiding significant impairment to the UAV’s flight performance. The inlet is positioned directly above the hydrogen fuel cell (the region indicated in Figure 9b). Numerical simulations compare temperature evolution over 900 s against a no-inlet baseline.

As shown in Figure 10, maximum temperatures vary minimally across configurations, while average temperatures decrease slightly with increasing inlet height due to larger inlet areas enhancing external cooling airflow into the fuselage, thereby lowering ambient temperatures around the fuel cell and improving overall heat dissipation.

However, temperature non-uniformity increases with the height of inlets: at 20 mm height, the temperature difference reaches 22.6 °C (19.6% higher than the baseline’s 18.9 °C), expanding to approximately 25 °C for 30 mm and 40 mm inlets (a further 10.6% increase compared to the 20 mm case). Figure 11 and Figure 12 show the resulting temperature and flow field distributions at 900 s, respectively.

Figure 11 demonstrates that increasing inlet height progressively reduces the high-temperature zones on the fuel cell, though elevated temperatures persist anteriorly. Figure 12a,c,e reveals enhanced external airflow through the symmetry plane at greater inlet heights, generating an expanded vortex in the back fuselage while maintaining high airflow velocity along the posterior fuel cell surface. This configuration sustains elevated heat transfer efficiency between the fuel cell and airflow, accelerating heat dissipation posteriorly and reducing temperatures in this region. The increased airflow in Figure 12c reduces the back-section temperature compared to (a). However, when the inlet height increases to 40 mm, the ambient air temperature remains constant, and the flow velocity behind the fuel cell ceases to increase significantly. Consequently, while higher air mass flow lowers the overall temperature in the back section, it negligibly affects the minimum temperature. This results in minimal difference in the fuel cell’s minimum temperature between Figure 12c,e.

However, as incoming air primarily flows toward the back section, minimal improvement occurs anteriorly where low airflow velocity persists, resulting in stable frontal high-temperature zones. Therefore, the temperature difference of the fuel cell exhibits negligible variation between the 30 mm and 40 mm inlet height configurations. Increased inlet height augments internal airflow (indicated by red dashed boxes), gradually enhancing cooling performance and expanding posterior low-temperature regions. On the cross-section through the fan axis, temperature variations remain negligible due to dominant fan-induced cooling effects.

3.3. Effect of Inlet Position on Fuel Cell Temperature

Given the superior thermal performance of the 40 mm-high inlet configuration—exhibiting lower average temperatures and enhancing heat dissipation without an increased temperature difference compared to the 30 mm variant—this optimal height is selected to evaluate positional effects. Identical inlets are then installed at back, mid-fuselage, and front positions (Figure 13).

Comparative analysis of fuel cell temperatures across configurations (Figure 14) reveals significant positional influence. The back-positioned inlet yielded maximum values: maximum temperature (63.8 °C), average temperature (54.9 °C), and temperature difference (25.2 °C). Conversely, the middle configuration achieved minimal maximum (58 °C) and average (50.8 °C) temperatures—representing reductions of 9.1% and 22.2%, respectively, compared to the back position—while maintaining a difference comparable to the front position and substantially lower than the back configuration.

Temperature contours (Figure 15) demonstrate distinct spatial patterns: back-positioned inlets produce extensive high-temperature zones concentrated anteriorly with poor uniformity; mid-position inlets reduce both thermal magnitude and spatial extent of hotspots while enhancing posterior cooling; front-position inlets exhibit moderate temperature increases relative to mid-position but remain significantly superior to back-position performance. The mid-configuration notably reduces temperatures in posterior cooling-enhanced regions.

Temperature contours and streamlines for mid- and front-positioned inlets at 900 s (Figure 16) reveal critical airflow modifications relative to the back configuration (Figure 12e,f). Anterior inlet migration shifts accompanying vortices forward at Location 1 (Figure 16a), reducing their impact on the fuel cell’s anterior region while significantly improving local airflow characteristics. This enhancement increases heat dissipation and reduces temperatures. However, transitioning from back to mid-position generates a new vortex at Location 2 (Figure 16a), which subsequently migrates posteriorly to the fuel cell’s anterior surface when inlets move forward. This repositioned vortex impedes surface airflow, reducing anterior velocity vectors to predominantly vertical downward trajectories. As air flows downward along the fuel cell surface, its temperature rises progressively. This diminishes the temperature gradient between the air and the cell, thereby reducing heat transfer efficiency. This degradation results in localized hotspots in the anteroinferior region. Conversely, on fan-aligned sections (Figure 16b,d), anterior inlet placement improves airflow adequacy and mitigates vortex interference compared to the back configuration (Figure 12f), resulting in measurable temperature reduction.

4. Conclusions

The performance of hydrogen fuel cells is intrinsically linked to their thermal state, necessitating an effective thermal management system (TMS). Integration within UAV fuselages imposes stringent cooling requirements due to spatial constraints and prolonged operational heat loads. This study conducts 900-s transient simulations of hydrogen fuel cell cooling in UAV applications, yielding three principal conclusions:

(1) In a constant total active cooling area, fan density configurations have a significant impact on thermal performance. The increased number of fans enables a wider distribution across the fuel cell surface, dispersing airflow more uniformly through the flow channels. This ensures effective cooling at peripheral regions, reducing the temperature difference. Compared with the four-fan configuration, a nine-fan arrangement reduces the temperature difference by 18.5% while enhancing temperature uniformity.

(2) Both inlet height and position strongly influence the thermal behavior of the fuel cell. Following inlet integration, the external airflow primarily affects the back section of the fuel cell, reducing its temperature while exerting minimal influence on high-temperature frontal zones. Consequently, the temperature difference increases significantly compared to the no-inlet configuration. As inlet height increases, a larger volume of external cooling air enters the fuselage, lowering the average fuel cell temperature. However, airflow velocity near the back low-temperature zone remains nearly constant, and ambient temperature does not change. Thus, the minimum temperature does not decrease further; instead, the low-temperature zone expands. This explains why the temperature difference stabilizes when inlet height increases from 30 mm to 40 mm.

(3) Mid-positioned inlets demonstrate optimal cooling efficacy, reducing maximum and average temperatures by 9.1% and 22.2%, respectively, versus back-positioned inlets. The high-temperature zones experience minimal vortex interference, enabling unimpeded airflow that significantly enhances heat dissipation. The temperature difference remains comparable to front-positioned inlets while being significantly lower than back configurations.