INTRODUCTION
The global transition toward sustainable energy solutions emerges as a widely accepted strategy to tackle critical issues including structural shifts in energy supply, exhaustion of fossil fuel reserves, and escalating climate change impacts . Today, hydrogen is positioned as an energy carrier for a more sustainable future, primarily as a renewable energy carrier [2]. As highly efficient energy source, hydrogen energy plays the key role in energy revolution by virtue of its diversified energy conversion pathways, excellent energy conversion efficiency, and environmentally friendly characteristics [3,4]. Among them, PEMFC, as one of the core technologies for utilizing hydrogen energy, makes significant technological breakthroughs since the 1980s, and its commercial application value becomes increasingly prominent. It uses hydrogen and oxygen as reaction raw materials, so they only discharge water, making it a key low-carbon technology [6]. PEMFC demonstrates superior performance characteristics, particularly in terms of energy conversion efficiency [6], rapid response, and low noise [7]. In the field of transportation, the application of PEMFC can not only effectively alleviate environmental pollution and energy shortage. It is a solution for sustainable energy in urban transportation, but also has important strategic significance for bolstering energy security, and advancing high-end manufacturing capabilities [8-11].
PEMFC operates as a converter, transforming chemical energy into electrical energy via electrolyte-mediated electrical conduction [10,12]. Within the structural configuration of fuel cell stacks, the bipolar plate constitutes an essential element whose flow field pattern significantly influences the overall electrochemical performance of the system. The primary design considerations for PEMFC channel focus on three critical functions: facilitating the delivery of oxidant and fuel to the electrode interface, maintaining homogeneous gas dispersion throughout the catalyst-coated membrane, and effectively eliminating byproduct water to minimize performance degradation in the electrochemical device.
Three predominant channel configurations were extensively employed in fuel cell applications: parallel, serpentine, and interdigitated channel. Each type of flow field structure had its own unique characteristics, in which the parallel channel, as a usual channel structure, was simpler and more regular compared to other channel and had significant advantages in practical applications [13,14]. Nevertheless, the parallel configuration exhibited significant limitations, particularly manifesting as non-uniform gas distribution patterns during high-current-density operation. The liquid water generated by the reaction tended to form large droplets in the channel, which could not be discharged in time, affecting the stable operation of the PEMFC. Therefore, it was urgent to develop a new parallel channel structure to improve the overall efficiency [15-17].
Previous studies established solid base for advancing fuel cell technology, with notable progress made particularly in the optimization of flow field structures. These efforts could be broadly categorized into two directions: structural modification of flow fields, application of baffles, and adjustment of basic geometric parameters.
The first was the innovations of macroscopic structural. For instance, Shen et al. investigated the efficiency of PEMFC with parallel, serpentine, and pressurized parallel channels and found that the latter exhibited superior performance and more uniform gas distribution with uniform water-gas distribution [18]. Arun Saco et al. found that serrated parallel channels enhance PEMFC efficiency, while Emad Farokhi et al. demonstrated that gradually tapered the main channel increased reactant mass transport [19,20]. Beneficial flow resistance could also be generated by introducing baffles. Zhang et al. found that aligned baffle distributions can increase net power density by 47.8% and improve drainage performance. Perng et al., highlighted the synergistic effect of combining tapered channel with baffle blockage [21,22]. Similarly, the discontinuous rib structure proposed by Dong et al. improved gas transport and electrochemical properties [23]. Thitakamol et al. proposed new baffle-inserted finger runner design, and the experimental findings indicated that the PEMFC performance with baffle at the middle position of the runner was superior to that of conventional inserted finger type channel [24]. These studies collectively emphasized that strategically placed obstacles enhanced forced convection from the GDL to the CL. Additionally, there were studies focusing on tapered or convergent channel designs. Ghasabehi et al. and Zhong et al. proposed new type of tapered parallel channel. Research showed that this inlet tapering design enhanced gas transport in the channel, improved current density uniformity, and thus significantly boosted the performance of PEMFC [25-27]. Mancusia et al. revealed the underlying mechanism through gas-liquid two-phase flow simulations. It was found that the increasing in inclination angle accelerated H2O removal and improved O2 distribution [28]. Ashrafi et al. improved the uniformity of two-phase flow by gradually thinning the sub-channels and achieved enhancement in the stability of PEMFC [29].
The second topic was systematically examined the influence of basic geometric dimensions, such as channel width and rib spacing. The research by Wang et al. and Zhang et al. confirmed that optimizing these parameters enhanced liquid water discharge and oxygen transport, thereby boosting performance [30,31]. Orhan et al. investigated the relationship between voltage loss and the rib-to-channel width ratio, suggested that the ratio should be maintained between 1 and 2 [32]. Kumar’s research also supported this view, indicating that smaller rib width helped for improving the utilization efficiency of reactive gases [33]. Scholta et al. pointed out that when the channel or rib width was controlled within the range of 0.7-1 mm, PEMFC performance reached its optimal state [34]. The conclusion confirmed that the above geometric parameters would significantly change the working characteristics of PEMFC, which in turn affected its overall output efficiency [35,36].
Although the independent effects of geometric parameter adjustments and tapered channel design had been fully verified, research gap remained regarding key parameters: the synergistic interaction between inlet dimensions and rib width ratio had not been systematically investigated. Additionally, the synergistic optimization relationship between these two parameters had not yet been systematically explored. In particular, there was lack of mechanistic explanations for why specific combinations of inlet dimensions and rib width ratios yielded optimal performance. Therefore, this research aimed to fill this gap by simultaneously varying inlet dimensions and rib width ratios, analyzing their effects on reactive gas distribution, water management, and current density uniformity, which providing optimization basis. Therefore, in this study, three-dimensional model created. There were two main aspects studied. Firstly, the study examined how variations in channel dimensions affected PEMFC performance. Secondly, the rib-width ratio of the channel was changed, and double optimization was carried out to obtain the new flow field.
3D NUMERICAL MODELING FOR PEMFC
This study, employed COMSOL software to construct a model, and the fuel cell module within it was utilized for simulation and analysis. The PEMFC was composed of cathode-anode bipolar plates and membrane electrode assembly (MEA). The MEA included GDL and CL at both the cathode and anode, with a proton exchange membrane (PEM) in the middle. The structure was shown in Fig. 1. In this study, several tapered parallel channel and conventional tapered parallel models with different inlet sizes and rib width ratios were constructed. The tapered parallel flow field featured an effective area measuring 39 mm by 39 mm. The flow field comprised 25 individual channels, each with a width of 1 mm, and the spacing between adjacent channels was 0.5 mm. The structural parameters of the novel tapered parallel channel PEMFC were shown in Table 1.
MODEL SOLUTION AND RATIONALITY VERIFICATION
Model solving
Model assumption
Given the intricate nature of the reaction mechanism within PEMFC and the interconnected pathways involved, this research employed a streamlined modeling approach to simplify the reaction process. The goal was to enhance computational speed, minimize resource usage, and maintain the reliability of the results without compromising their precision. Given the complexity of the reaction mechanism inside the PEMFC and the close correlation between the reaction paths, a simplified assumption approach was used to model the reaction process., aiming to improve the computational efficiency, reduce the consumption of computational resources, and ensure that the accuracy of the obtained results would not be affected.
(1) PEMFC was operated in steady state.
(2) The reactant gas followed laminar flow and stable transmission.
(3) Reactive gas species were treated as ideal gases in the computational model.
(4) Gravitational effects were disregarded in the reaction analysis.
The operating parameters of the PEMFC were exhibited in Table 2.
To verify the generality of the model under practical operating conditions, additional extended simulations were conducted with combinations of different relative humidity (RH = 60%, 80%, 100%) and back pressures (0-0.2 MPa). The results show that within the range of commonly used operating conditions (RH ≥ 60%, back pressure ≤ 0.2 MPa), the core conclusions of this study remain consistent. To eliminate the interference of “membrane drying” on performance and to more easily highlight the functional regularity of changes in the rib width ratio, this study employs RH = 100% with no back pressure.
To verify the impact of membrane thickness on the conclusions, additional extended simulations were conducted with different PEM thicknesses (12 μm, 13 μm, 25 μm, 100 μm). The results show that the core conclusions of this research remain consistent under these membrane thickness. To rule out the possibility that performance improvements stem from the channel rib width ratio rather than the optimization of the membrane’s own parameters, membrane thickness of 100μm was used in this research.
Mathematical model
A PEMFC finite element model for PEMFC was constructed based on the COMSOL Multiphysics platform. The simulation included secondary current distribution, dense matter transfer, and the Brinkman equation. To maintain the balance between the physical fields, the model also introduced series of auxiliary control equations, thereby enabling the accurate simulation of the PEMFC working process [21,37]. For the basic control equation, it mainly included:
Equation of mass conservation
Where, was the density of gas (kg/m3), ε and u −
Momentum-conservation equation
where, μ and P were dynamic viscosity (N·s/m2) and pressure (Pa), respectively; Su was the momentum-source term (N·s), which was a physical quantity describing the change of momentum in unit time.
Energy conservation equation
where cP represented the specific heat capacity (J·kg-1·K-1); t was temperature (K); k was the thermal conductivity (W·m-1·K-1); SQ was the energy source term, which represented the rate of energy generation or consumption per unit time caused by electrochemical reaction, energy conversion, and thermodynamic process, in units of (W/m3); the subscript eff denoted the porous medium’s effective properties. Component conservation equation
This equation could more accurately describe the mass transport and conversion processes of reactive gases (such as H2, O2) and water.
where ck and D k eff
For CL, the component source terms of H2, O2, and H2O were:
Current conservation equation
It could be decomposed into two potential equations:
Where c and m denoted the solid and membrane phases. σ represented the conductivity (S/m), ϕ represented the potential (V), and S represented the current source contribution.
Butler-Volmer equation
Among them, j a ref j c ref
Transport equation of liquid water
In the formula, Pwv was the liquid water pressure (Pa);Psat referred to the partial pressure (Pa) of water vapor in the gas phase. 1 was water density (kg/m3); rw represented the gas-liquid phase change rate of water vapor [kg/(m3·s)]. s was the phase saturation of water, which was a dimensionless quantity, usually expressed in volume fraction, with a range of 0 to 1.
Parasitic loss (Ppara) (i.e. pumping power) formula
Where, η was the fan efficiency, taken as 0.8 [21]; Q was the air volumetric flow rate (m3/s), which was determined by Faraday’s law and the ideal gas state equation, considering the air stoichiometric ratio λ=2 and the oxygen volume fraction of 0.21; and ΔP was the channel pressure drop (Pa).
Net output power (Pnet)
Herein, Pcell is the output power of the battery (W); A was the effective cell area (cm2); and was P max '
Calculations in this study showed that the parasitic loss caused by channel pressure drop was three orders of magnitude lower than PEMFC output power, and its impact could be neglected. Therefore, the net power in this study was the PEMFC output power, which did not affect the conclusions regarding performance comparisons between different channel designs.
Model rationality verification
Grid independence verification
Since the shape and size of each part of the PEMFC varied greatly, it was important to conduct grid independence test during the construction of the grid to ensure that the results of the PEMFC model simulation would not be overly affected by changes in the number of grids. The grid independence validation model used traditional parallel flow field, which was meshed in COMSOL software. The details of grid division were shown in Fig. 2. The detailed grid parameters of all schemes were listed in Table 3.
The table revealed that once the grid count fulfilled the criteria of option 3, there was no significant increase in the value of current density with the increase in the number of grids. The current density difference between option 3 and option 5 was only 0.24%, which could be ignored. Because the calculation time of option 3 was shorter than that of option 4 and option 5, considering the calculation cost and calculation accuracy, option 3 was ultimately adopted as the grid partitioning strategy for the following research in other subsequent calculation models.
Model validation consistent with experimental conditions
Mehrdad Ghasabehi et al. verified the impact of tapered parallel channel structure on PEMFC performance through experiments. This study validated the tapered parallel channel under boundary conditions and operational parameters established by Ghasabehi et al. Table 4 was the operating parameters set by the experiment. Fig. 3 showed the comparison of the experimental data with the simulation data. The experimental data exhibited strong alignment with the simulation results, indicating high reliability of the simulations.
RESULTS AND DISCUSSION
All the following comparative analyses of the channel geometries were conducted based on the standard operating conditions of PEMFC (353.15 K, 100% RH).
Impact of tapered parallel channel on PEMFC performance
This section discussed the influence of the same-size tapered and traditional parallel channel for PEMFC. Each of the channel with 25 single runners, had inlet dimensions, channel widths, and rib widths of 1 mm. Only the primary channel in the tapered parallel flow field was tapered, with other geometric parameters and operating parameters were the same.
The polarization curve was a key reference index when evaluating PEMFC performance. Fig. 4 displayed the polarization and power density curves for the tapered as well as the traditional parallel channel.
The improved tapered parallel channel exhibited more uniform current density distribution and higher power density compared with traditional parallel channel. The performance of PEMFC was ultimately governed by the intensity of electrochemical reactions, which in turn depended strongly on the reactant transport efficiency determined by channel design. As shown in Fig. 4, the uniformity of current density distribution in the tapered channels were significantly improved, which was directly attributed to optimized oxygen transport and alleviated local flooding, thereby reducing the mass transfer overpotential caused by uneven reactant distribution. Moreover, the results demonstrated that the tapered parallel channel achieved a current density of 0.61415 A·cm-2, while its maximum power density reached 0.36849 W·cm-2, representing a 43.66% increase in comparison to the traditional parallel channel.
Oxygen concentration, water concentration and current density distribution on CL at 0.7 V were shown in Fig. 5. From Fig. 5 (I) and (II), it could be seen that the distribution uniformity of oxygen and water on the CL surface at the cathode side of the fuel cell had been significantly improved, and the mass transfer effect of the fuel cell under the tapered parallel flow field was the best. In the traditional parallel channel design, the presence of numerous internal passages caused a reduction in gas mobility and drop in pressure difference, ultimately leading to a decline in the concentration of reacting gases on the CL. As the reaction progressed, water continuously formed in the channel. Notably, the traditional parallel channel's central region had a higher water molarity. It could be clearly seen from Fig. 5 (I) and (II) that the phenomenon of the tapered parallel flow field was obviously improved. The gradually tapering design of the inlet in the tapered parallel channel essentially created convergent-divergent geometric structure. According to the law of mass conservation, when reactive gas flowed through the convergent section with reduced cross-sectional area, its flow velocity increased. This acceleration effect produced two key impacts: (1) it enhanced the convective transport capacity in the middle section of the channel, enabling more effective “pushing” of oxygen to the surface of the gas diffusion layer (GDL). (2) it strengthened the shear and stripping effect on liquid water that might block the channels, thereby reducing the risk of flooding and creating more unobstructed diffusion path for oxygen. Therefore, the findings demonstrated that the tapered parallel channel design improved the oxygen concentration distribution and water removal ability.
Fig. 5 (III) displayed the current density distribution on the CL at 0.7 V it revealed that the traditional parallel channel exhibited minimum current density at the channel's central region in the PEMFC. The tapered parallel channel adopted the channel contraction design, and by improving the oxygen distribution along the channel direction, the active area of the whole catalyst layer was more fully and evenly utilized, which increased the catalytic performance of the catalyst layer. The chemical energy of the reactant converted into electrical energy was more efficiently, which made the gas convection range under the rib larger and the current density increased significantly, thus greatly improving the defect of low current density in the middle channel. Hence, the tapered parallel channel was found to be superior for enhancing fuel cell performance to the traditional parallel channel.
Impact of entrance size on PEMFC performance
This section examined how the tapered parallel channel’s geometric features affect PEMFC performance. Initially, the effect of geometric parameters on fuel cell performance was examined by altering the channel inlet dimensions. The optimization scheme altered the main channel’s structure and dimensions, and the channel width was 1 mm wider than the rib width. In Model 1, the main channel’s width shrank in a linear fashion, decreasing from 2 mm to 1 mm. In Model 2, the main channel’s width shrank in a linear fashion, decreasing from 3 mm to 1 mm. In Model 3, the main channel’s width shrank in a linear fashion, decreasing from 4 mm to 2 mm. In Model 4, the main channel’s width shrank in a linear fashion, decreasing from 4 mm to 1 mm.
Here, the expression “X-Y mm” was defined as follows: “X” referred to the initial width of the channel inlet section, corresponding to the width before contraction, with the unit of mm; “Y” denoted the final width of the channel inlet section, i.e., the final width after contraction, with the unit of mm. To eliminate interference from other variables, the axial distance (inlet section length) over which all inlet dimension schemes contracted from the initial width X to the final width Y was fixed at 37.5 mm, ensuring that differences between schemes arose solely from width changes before and after contraction. The specific parameters were summarized in Table 5.
Fig. 6 illustrated the polarization and power density curves of PEMFC of Models 1 to 4. The expansion of the channel’s inlet size extended the duration of the peak power density peak area. By increasing the flow rate of the reaction gas, the gas distribution became more uniform, thereby diminishing the concentration disparity within the channel and postponing the reduction in power density. Model 4 exhibited superior performance in tapered parallel channels featuring varied inlet and outlet dimensions. It exhibited a current density of 1.31 A·cm-2 at a voltage of 0.4. And the 0.53685 W·cm-2 maximum net power density demonstrated 45.6% enhancement relative to the original tapered parallel channel in the previous section.
Oxygen concentration, water concentration and current density distribution on CL at 0.6 V was shown in Fig. 7. From Fig. 7 (I), it’s clear that the area with insufficient oxygen supply, namely the blue region, was mainly located in the central and lower part of the whole flow field. However, it could be observed that Model 1-3 had only slight improvement in oxygen concentration distribution. Excessively small dimensions (e.g., 2-1 mm, 3-1 mm) led to a limited increase in flow velocity, which was insufficient to effectively maintain the oxygen transport intensity in the middle and rear sections along the entire length of the channel. In contrast, excessively large dimensions or improper design (e.g., 4-2 mm, with a relatively large outlet cross-section) weakened the acceleration effect in the outlet region. Fig. 7 (II) illustrated how water concentration was distributed across the entire CL’s surface. The results indicated that enlarging the inlet size enhances gas flow uniformity, which minimized localized water buildup and improved drainage efficiency within the channel. The optimal-performing Model 4 (4-1 mm) design provided an optimal solution: its significant inlet cross-section (4 mm in width) ensured initial low flow resistance, while its steep taper, contracting to 1 mm outlet, generated the most significant increase in flow velocity in the middle section of the channel. This, on one hand, enhanced the convective mass transfer coefficient of oxygen from the center of the channel to the GDL surface. On the other hand, the increased gas flow shear force more effectively removed liquid water from the GDL surface, significantly alleviating flooding in the GDL pores and within the channel. Thus, the findings demonstrated that modifying the channel’s inlet dimensions enhanced fuel cell water management.
Fig. 7 (III) depicted the current density distribution on CL at 0.6 V. It was evident that Model 4 exhibited a significant difference from the conventional tapered parallel channel. Fig. 7 (III) illustrated that variations in current density across various inlet dimensions predominantly occur at the channel split. This was because the increasing of inlet size increased the gas flow, ensuring more uniform gas dispersion throughout the channel. This effectively minimized concentration gradients of reactant gases, allowing all regions of the CL to receive adequate gas supply. As mentioned previous, the Model 4 ensured more uniform supply of oxygen along the entire length of the channel through its strong convection in the middle and rear sections, allowing the catalyst active area to be more fully utilized. Its excellent oxygen transport and water removal capabilities directly alleviated mass transfer limitations, enabling electrochemical reactions to proceed at high rate under lower overpotentials. Therefore, at the same voltage, this channel could output higher current density.
This further demonstrated that altering the channel’s inlet size enhanced PEMFC performance. Thus, the analysis of the four models revealed that the PEMFC featuring a 4-1 mm tapered parallel channel exhibited superior electrochemical characteristics. Next, the rib width ratio was optimized to determine the ideal channel.
Fig. 8 and 9 showed the pressure distribution and corresponding performance comparison under different inlet dimension projects, respectively. It could be seen from Fig. 8 that Model 1 and Model 2 projects exhibited steeper pressure gradients with higher flow resistance, while Model 4, which had the optimal performance, showed a gentler pressure decay. This indicated that its channel structure had lower flow resistance, facilitating the uniform distribution of reactant gases. Fig. 9 further showed that Model 4 achieved the highest net power density, which was approximately 17.3%, 8.5%, and 6.9% higher than that of Model 1, Model 2, and Model 3, respectively, significantly outperforming the other projects. There was no simple negative correlation between pressure drop and net power density among the projects. This phenomenon reconfirmed that the parasitic loss caused by channel pressure drop had a negligible impact on net power. Therefore, the differences in cell performance mainly stemmed from the optimization effect of channel geometry on mass transfer processes.
In summary, optimizing the channel inlet dimensions to improve mass transfer efficiency was an effective strategy for enhancing PEMFC performance, with its benefits far outweighing the minor flow resistance losses. The tapered design of Model 4 effectively improved reactant supply and catalytic layer utilization, thereby significantly enhancing the overall performance of the PEMFC.
Impact of rib width ratio on PEMFC performance
In this section, four distinct combinations of runner widths and rib widths were designed, with their specific parameters presented in Table 6. The remaining parameters followed the Model 4 in the previous section. Model 4 was the scheme of this section. The impact of rib width ratio on PEMFC performance was studied by changing the channel and rib widths.
Fig. 10 illustrated the polarization and power density curves of PEMFC in projects (a) through (d). The plots illustrated the polarization behavior and power density of the PEMFC under diverse configurations of channel and rib widths. The outcomes of projects (b)-(d) were juxtaposed against those of project (a) for comparison. Fig. 10 clearly showed that when the rib width was held steady, expanding the channel width resulted in progressively worse performance for the PEMFC with a tapered parallel flow design. The data reveals that this performance drop became more pronounced as the channel width continued to grow. With a fixed channel width, decreasing the rib width greatly enhanced the tapered parallel channel PEMFC’s performance. The findings demonstrated that the tapered parallel channel PEMFC under scheme (d) exhibited the best performance with the maximum current density of 1.3497 A·cm-2.
Oxygen concentration, water concentration and current density distribution on CL at 0.6 V was shown in Fig. 11 As depicted in Fig. 11 (I), oxygen levels steadily decreased from inlet to outlet during the reaction. Projects (a) and (d) showed better oxygen molar concentration than projects (b) and (c). The project (d) with the rib width of 0.5 mm, the channel width of 1 mm, and the rib width ratio of 2 showed the highest oxygen molar concentration. Projects (a)-(c) revealed that the increasing in channel width increased the area of internal channels present within the tapered parallel channels and decreased the gas flow rate, which led to the reduction in the concentration of reactive gases within the CL. Projects (a) and (d) showed that reducing the width of the rib plate and the interval between channel significantly enhanced reaction gas diffusion into the GDL and CL regions and make the oxygen concentration distribution more uniform. Fig. 11 (II) clearly illustrated that projects (a) and (d) exhibited a lesser molar concentration of water compared to projects (b) and (c). The discrepancy stemmed from the channel's increased width, which led to a reduction in gas flow rate. Consequently, the reduced gas flow rate was incapable of effectively transporting water away, leading to a higher likelihood of water pooling within the channel. Notably, projects (d) showcased the lowest water molar concentration at a mere 0.34 mol/m3. In contrast, the original conical parallel flow field boasted a maximum water molar concentration of 0.38 mol/m3 and demonstrated an 11% enhancement in water removal capacity. For project (d), the narrow ribs meant larger area beneath, the GDL was directly exposed to the channel, which optimized the direct diffusion path for oxygen. More critically, this design altered the under-rib convection effect: the pressure difference from one channel, passing under the rib, to the adjacent channel drove stronger under-rib convection. This convection pumped reactants from the center of the channel into the under-rib region, significantly improving oxygen supply to the under-rib area and alleviating local starvation issues. Meanwhile, this enhanced permeation flow also helped carry away liquid water generated in the under-rib region, preventing local flooding. These findings revealed that adjusting the rib width ratio could positively affect the distribution of O2 and the system’s water removal capabilities.
Current density distribution of projects (a)-(d) at 0.6V illustrated Fig. 11 (III). The results demonstrated that the current density distribution of project (d) was more uniform compared to projects (a)-(c). This resulted from project (d) narrowing the rib plate’s width while maintaining the channel’s width. For configuration with rib width of 0.5 mm, the difference in current density between the under-channel and under-rib regions narrowed significantly, resulting in more uniform distribution. This demonstrated that the narrow rib width successfully activated the catalyst in the under-rib region by shortening the diffusion distance and enhancing under-rib convection, thereby substantially improving the overall utilization efficiency of reactants and reducing instances of excessively high or low local current density. Therefore, project (d) was considered as the best one, resulting in doubly optimized novel channel.
It could be seen from Fig. 12 that project (a) and (d) exhibited significant pressure gradients, indicating relatively high flow resistance. In contrast, the pressure curves of project (b) and (c) were relatively flat, reflecting that their channel structures had the lowest flow resistance. Fig. 13 further showed that project (d) achieved the highest net power density, which was approximately 4.3%, 7.2%, and 20% higher than that of project (a)-(c), respectively. The above results indicated that although project (d) and (c) had the same rib-width ratio, project (d) halved the absolute dimensions of rib width and channel width. This resulted in the increasing in pressure drop while significantly enhancing the transport efficiency of reactant gases and the utilization of active sites at the electrode interface, thereby achieving the optimal electrochemical performance. In contrast, due to excessively wide channels, project (c) suffered from uneven reactant distribution, which limited the improvement of its power density.
In summary, for PEMFC, the design of channel structures required trade-off between reducing transport resistance and enhancing reactant supply. Therefore, the reasonable rib-width ratio design was of great significance for optimizing the performance of PEMFC.
CONCLUSIONS
This research investigated how geometric structure parameters of tapered parallel channel affected PEMFC performance. Furthermore, a three-dimensional multi-constant temperature steady-state model of multiple tapered parallel channel was developed and subjected to numerical simulation. Parameters of the flow field structure (including inlet dimensions and rib width ratios) were investigated for their impact on cell efficiency. The conclusions were as follows:
(1) When compared to the traditional parallel flow field, the tapered parallel flow field could obviously promote the transport of reactive gas and elevated its concentration within the channel. The maximum power density was 0.36849 W·cm-2, representing a 43.66% enhancement compared to the traditional parallel channel (0.2564925 W·cm-2).
(2) Model 4 showed excellent output performance. The current density was 1.31 A·cm-2 at 0.4 V, and the maximum net power density was 0.53685 W·cm-2, which was 45.6% higher than that of the initial tapered parallel channel.
(3) In the optimization design of different rib width ratios, the PEMFC output performance of project (d) was ideally improved (with the rib channel width ratio of 2:1 (1mm: 0.5 mm)), and the effect was the best. In comparison to the original tapered parallel channel design, the proposed design in this study achieved an approximate 52% enhancement in net power density and an approximately 11% improvement in water removal capacity.
(4) While this study systematically investigated two parameters-the channel inlet dimensions and rib width ratio. It did not consider the comprehensive impact of other key parameters on PEMFC performance. This may lead to one-sided nature of the research conclusions. Future studies could further expand the research scope based on the Project (d) model, systematically exploring the effects of operating parameters (anode/cathode stoichiometric ratios, cell temperature, relative humidity (RH), and backpressure) on PEMFC performance. This would form multi-parameter synergistic analysis framework for PEMFC, further enhancing the comprehensiveness of the research. This research also explores the influence of channel depth (0.8, 1.0, 1.2 mm). The results show that while the channel depth alters local transport characteristics, it does not affect the optimal rib width ratio-this confirms the robustness of the latter as the key design parameter. To focus on the core research theme, the systematic parametric analysis of channel depth will be conducted in future work.
PDF Links
PubReader
ePub Link
Full text via DOI
Download Citation
