Numerical simulation of two‐phase cross flow in the gas diffusion layer microstructure of proton exchange membrane fuel cells

The cross flow in the under‐land gas diffusion layer (GDL) between 2 adjacent channels plays an important role on water transport in proton exchange membrane fuel cell. A 3‐dimensional (3D) two‐phase model that is based on volume of fluid is developed to study the liquid water‐air cross flow within the GDL between 2 adjacent channels. By considering the detailed GDL microstructures, various types of air‐water cross flows are investigated by 3D numerical simulation. Liquid water at 4 locations is studied, including droplets at the GDL surface and liquid at the GDL‐catalyst layer interface. It is found that the water droplet at the higher‐pressure channel corner is easier to be removed by cross flow compared with droplets at other locations. Large pressure difference Δp facilitates the faster water removal from the higher‐pressure channel. The contact angle of the GDL fiber is the key parameter that determines the cross flow of the droplet in the higher‐pressure channel. It is observed that the droplet in the higher‐pressure channel is difficult to flow through the hydrophobic GDL. Numerical simulations are also performed to investigate the water emerging process from different pores of the GDL bottom. It is found that the amount of liquid water removed by cross flow mainly depends on the pore's location, and the water under the land is removed entirely into the lower‐pressure channel by cross flow.     

Numerical investigations on liquid water removal from the porous gas diffusion layer by reactant flow

Water removal from the gas diffusion layer (GDL) is crucial for the efficient operation of proton exchange membrane (PEM) fuel cell. Static pressure gradient caused by the fast reactant flow in the flow channel is one of the main mechanisms of water removal from GDL. Reactant can leak or cross directly to the neighboring channel via the porous GDL in the cells with serpentine flow channel and many of its modifications. Such cross flow plays an important role for the removal of liquid water accumulated in the GDL especially under land area. To investigate the characteristics of liquid water behavior in the GDL under pressure gradient, the fibrous porous structure of the carbon paper is modeled by three dimensional impermeable cylinders randomly distributed in the in-plane directions and unsteady two-phase simulations are conducted. It is shown that the permeability from the numerical model matches well the experimental measurements of the common GDLs in the literature. The contact angle and pressure gradient are the key parameters that determine the initiation and the process of liquid water transport in the GDL which is initially wet with stagnant liquid water. It has been observed that the larger contact angle results in faster water removal from the GDL. Numerical simulations are performed for a wide range of pressure gradient with different contact angles to determine the minimum pressure gradient that initiates the liquid water transport in the GDL. It is found that the amount of pressure gradient caused by the cross flow is sufficient and effective to get rid of the liquid water accumulated in the GDL. The simulation results are also compared with experimental data in literature showing a good agreement. The characteristics of liquid water discharging from the gas diffusion layer are also described.     

Dynamics of liquid water in the anode flow channels of PEM fuel cells: A numerical parametric study

A 3D volume of fluid (VOF) model for an anode channel in a PEM fuel cell has been built. The effects of the initial position of the water droplet, its size as well as the wettability of the gas diffusion layer (GDL) are investigated under different operating conditions. It is found that the initial position of the relatively small water droplet in the channel has almost no effect on the pressure drop and the time taken for the liquid water to move out from the channel; however, such effects become more profound as the size of the water droplet increases. Also, when the droplet is placed at the side wall of the channel, then it develops into pockets of water that are mainly located at the upper corners of the channel, thus causing a smaller pressure drop compared to the cases in which the water droplet is placed either on the surface of the GDL or on the top wall of the channel. Furthermore, the hydrogen velocity is found to have a negligible effect on the dynamics of liquid water; however, the pressure drop and removal time are significantly influenced by the hydrogen velocity. Moreover, as the size of the water droplet increases, the pressure drop increases and the time required for the liquid water to move out of the channel decreases. Finally, the pressure drop in the channel decreases and the removal time of the liquid water increases as the contact angle of the GDL decreases.     

An analytical analysis on the cross flow in a PEM fuel cell with serpentine flow channel

A serpentine flow channel can be considered as neighboring channels connected in series, and is one of the most common and practical channel layouts for polymer electrolyte membrane (PEM) fuel cells, as it ensures the removal of liquid water produced in a cell with good performance and acceptable parasitic load. During the reactant flows along the flow channel, it can also leak or cross directly to the neighboring channel via the porous gas diffusion layer (GDL) due to the high‐pressure gradient caused by the short distance. Such a cross flow leads to a larger effective flow area resulting in a substantially lower amount of pressure drop in an actual PEM fuel cell compared with the case without cross flow. In this study, an analytical solution is obtained for the cross flow in a PEM fuel cell with a serpentine flow channel based on the assumption that the velocity of cross flow is linearly distributed in the GDL between two successive U‐turns. The analytical solution predicts the amount of pressure drop and the average volume flow rate in the flow channel and the GDL. The solution is validated over a wide range of the thickness and permeability of the GDL by comparing the results with experimental measurements and 3‐D numerical simulations in literature. Excellent agreement is obtained for the permeability less than 10^?9 m², which covers the typical permeability values of the GDLs in actual PEM fuel cells. The solution presents an accurate and efficient estimation for cross flow providing a useful tool for the design and optimization of PEM fuel cells with serpentine flow channels. Copyright © 2010 John Wiley & Sons, Ltd.     

The role of the gas diffusion layer on slug formation in gas flow channels of fuel cells

Water drops emerge from large pores of the hydrophobic Gas Diffusion Layers (GDL) into the cathode gas flow channel of Polymer Electrolyte Membrane (PEM) Fuel Cells. The drops grow into slugs that span the cross-section of the flow channels. The slugs detach and are forced out the gas flow channel by the air flow. An acrylic micro-fluidic flow cell with a 1.6 mm gas flow channel and a 100 μm liquid pore through a carbon paper GDL has been used to quantitatively determine slug volumes, velocity of slug motion, and the force required to move slugs as functions of the gas and liquid flow rates. In a channel with 4 acrylic walls, slugs detach immediately upon formation. A porous GDL wall allows gas flow to bypass the slugs, thus allowing slugs to continue to grow after spanning the open area of the channel. The differential pressure to detach and move slugs is equal to the dynamic interfacial force on a slug normalized by the cross-sectional area of the channel. The dynamic interfacial force is equal to the difference between the downstream (advancing) and upstream (receding) contact lines of the water with the channel walls. Slugs will stop moving if the differential pressure drop for gas flow to bypass the slug and flow through the GDL under the rib separating the channels is less than the differential pressure required to move the slug. The results improve our physical insight into the state of water hold up in PEM fuel cells.     

Numerical study of water management in the air flow channel of a PEM fuel cell cathode

The water management in the air flow channel of a proton exchange membrane (PEM) fuel cell cathode is numerically investigated using the FLUENT software package. By enabling the volume of fraction (VOF) model, the air–water two-phase flow can be simulated under different operating conditions. The effects of channel surface hydrophilicity, channel geometry, and air inlet velocity on water behavior, water content inside the channel, and two-phase pressure drop are discussed in detail. The results of the quasi-steady-state simulations show that: (1) the hydrophilicity of reactant flow channel surface is critical for water management in order to facilitate water transport along channel surfaces or edges; (2) hydrophilic surfaces also increase pressure drop due to liquid water spreading; (3) a sharp corner channel design could benefit water management because it facilitates water accumulation and provides paths for water transport along channel surface opposite to gas diffusion layer; (4) the two-phase pressure drop inside the air flow channel increases almost linearly with increasing air inlet velocity.     

Effect of GDL compression on pressure drop and pressure distribution in PEMFC flow field

Improving reactant distribution is an important technological challenge in the design of a PEMFC. Flow field and the Gas Diffusion Layer (GDL) distribute the reactant over the catalyst area in a cell. Hence it is necessary to consider flow field and GDL together to improve their combined effectiveness. This paper describes a simple and unique off-cell experimental setup developed to determine pressure as a function of position in the active area, due to reactant flow in a fuel cell flow field. By virtue of the experimental setup being off-cell, reactant consumption, heat production, and water generation, are not accounted as experienced in a real fuel cell. A parallel channel flow field and a single serpentine flow field have been tested as flow distributors in the experimental setup developed. In addition, the interaction of gas diffusion layer with the flow distributor has also been studied. The gas diffusion layer was compressed to two different thicknesses and the impact of GDL compression on overall pressure drop and pressure distribution over the active area was obtained using the developed experimental setup. The results indicate that interaction of GDL with the flow field and the effect of GDL compression on overall pressure drop and pressure distribution is more significant for a serpentine flow field relative to a parallel channel flow field.     

Interface resolving two-phase flow simulations in gas channels relevant for polymer electrolyte fuel cells using the volume of fluid approach

With the increased concern about energy security, air pollution and global warming, the possibility of using polymer electrolyte fuel cells (PEFCs) in future sustainable and renewable energy systems has achieved considerable momentum. A computational fluid dynamic model describing a straight channel, relevant for water removal inside a PEFC, is devised. A volume of fluid (VOF) approach is employed to investigate the interface resolved two-phase flow behavior inside the gas channel including the gas diffusion layer (GDL) surface. From this study, it is clear that the impact on the two-phase flow pattern for different hydrophobic/hydrophilic characteristics, i.e., contact angles, at the walls and at the GDL surface is significant, compared to a situation where the walls and the interface are neither hydrophobic nor hydrophilic (i.e., 90° contact angle at the walls and also at the GDL surface). A location of the GDL surface liquid inlet in the middle of the gas channel gives droplet formation, while a location at the side of the channel gives corner flow with a convex surface shape (having hydrophilic walls and a hydrophobic GDL interface). Droplet formation only observed when the GDL surface liquid inlet is located in the middle of the channel. The droplet detachment location (along the main flow direction) and the shape of the droplet until detachment are strongly dependent on the size of the liquid inlet at the GDL surface. A smaller liquid inlet at the GDL surface (keeping the mass flow rates constant) gives smaller droplets.     

Numerical investigation of liquid water distribution in the cathode side of proton exchange membrane fuel cell and its effects on cell performance

A three-dimensional unsteady two-phase model for the cathode side of proton exchange membrane fuel cell (PEMFC) consisting of gas diffusion layer (GDL) with hybrid structural model is developed to investigate liquid water behaviors under different operating and geometrical conditions and to quantitatively evaluate effects of liquid water distribution on reactant transport and current density distribution. Simulation results reveal that liquid water transport processes and distributions are significantly affected by inlet air velocity, wall wettability and water inlet position, which in turn play a prominent role on local reactant transport and cause considerable disturbances of the current density. Liquid water film spreading on the gas channel (GC) top wall is identified as the most desirable flow pattern in the GC based on overall evaluations of current density magnitude, uniformity of current density distribution and pressure drop in the GC. Modification to GDL structure is proposed to promote the formation of the desirable flow pattern.     

Numerical investigation of liquid water dynamics in wave‐like gas channels of PEMFCs

For the air feed in proton exchange membrane fuel cells (PEMFCs), the wave‐like gas channel (GC) shows obvious advantages over the straight GC because the former enhances collision of secondary flow and diffusion in the gas diffusion layer (GDL). However, it is prone to water flooding, which brings greater pressure drop, larger pressure oscillation, and blocking of reaction area. In the present study, numerical models of the water dynamic processes, including water droplets emerging from micropores on the GDL surface and removing through the GC, are established based on the volume of fluid (VOF) method. Water coverage ratio and pressure drop are calculated to evaluate the water flooding. The effects of the dimensional parameters of wave‐like GC and contact angle of channel walls on the water accumulation are studied. The emergence and removal of liquid water is a quasiperiodic and oscillating process. Multicycle simulations show that channel pressure drop increases linearly with greater growth rate than channel length. The equilibrium position of water droplet is strongly dependent on the relative wettability of the GDL and bipolar plate (BPP) surfaces. And the geometric parameters of GC have a significant impact on the pressure, water removal behavior and detachment time. Smaller bent angle brings bigger pressure drop, and larger cycle length is helpful for relieving the oscillation of pressure.     

Error of Darcy's law for serpentine flow fields: An analytical approach

Serpentine flow fields and other flow fields with partial under-land cross-flow are commonly used in various energy devices, such as proton exchange membrane (PEM) fuel cells and redox flow batteries, due to their higher mass transfer rate to reaction sites and better product removal capability. Accurately predicting the under-land cross-flow rate and pressure drop in such flow fields is crucial in flow field design optimizations. Darcy's law is the most commonly used model in predicting the under-land cross-flow and pressure drop in such flow fields. However, since the Darcy's law neglects inertial effect, its validity in different designs and operating conditions needs to be carefully studied. In this work, mathematical models for a serpentine flow field are developed based on both the Darcy's law and a modified Darcy's law that includes the inertial effect. Both models are solved and analytical solutions are obtained. The predicted pressure drops and under-land cross-flow rates from the two models are compared with experimental data and the results show that under some conditions, both the Darcy's law and the modified Darcy's law can predict pressure drop and under-land cross-flow rate reasonably well. However, under other conditions the Darcy's law can result in significantly large errors in predicting both pressure drop and under-land cross-flow rates. Further studies provide the variations of errors from the Darcy's law with different parameters, including channel length, gas diffusion layer (GDL) thickness, land width, inlet flow rate, GDL permeability and GDL inertial coefficient.     

Effect of air flow on liquid water transport through a hydrophobic gas diffusion layer of a polymer electrolyte membrane fuel cell

The transport of liquid water through an idealized 2-D reconstructed gas diffusion layer (GDL) of a polymer electrolyte membrane (PEM) fuel cell is computed subject to hydrophobic boundary condition at the fibre–fluid interface. The effect of air flow, as would occur in parallel/serpentine/interdigitated type of flow fields, on the liquid water transport through the GDL, ejection into the channel in the form of water droplets and subsequent removal of the droplets has been simulated. Results show that typically water flow through the fibrous GDL occurs through a fingering and channelling type of mechanism. The presence of cross-flow of air has an effect both on the path created within the GDL and on the ejection of water into the channel in the form of droplets. A faster rate of liquid water evacuation through the GDL (i.e., more frequent ejection of water droplets) as well as less flooding of the void space results from the presence of cross-flow. These results agree qualitatively with experimental observations reported in the literature.     

An experimental and numerical investigation on the cross flow through gas diffusion layer in a PEM fuel cell with a serpentine flow channel

A serpentine flow channel is one of the most common and practical channel layouts for a polymer electrolyte membrane (PEM) fuel cell since it ensures the removal of water produced in a cell with acceptable parasitic load. During the reactant flows along the flow channel, it can also leak or cross to neighboring channel via the porous gas diffusion layer due to the high pressure gradient caused by the short distance. Such a cross flow leads to a larger effective flow area altering reactant flow in the flow channel so that the resultant pressure and flow distributions are substantially different from that without considering cross flow, even though this cross flow has largely been ignored in previous studies. In this work, a numerical and experimental study has been carried out to investigate the cross flow in a PEM fuel cell. Experimental measurements revealed that the pressure drop in a PEM fuel cell is significantly lower than that without cross flow. Three-dimensional numerical simulation has been performed for wide ranges of flow rate, permeability and thickness of gas diffusion layer to analyze the effects of those parameters on the resultant cross flow and the pressure drop of the reactant streams. Considerable amount of cross flow through gas diffusion layer has been found in flow simulation and its effect on pressure drop becomes more significant as the permeability and the thickness of gas diffusion layer are increased. The effects of this phenomenon are also crucial for effective water removal from the porous electrode structure and for estimating pumping energy requirement in a PEM fuel cell, it cannot be neglected for the analysis, simulation, design, operation and performance optimization of practical PEM fuel cells.     

An investigation of two-phase flow pressure drop in helical rectangular channel

Numerical simulations have been carried out to evaluate the two-phase frictional pressure drop for air-water two-phase flow in horizontal helical rectangular channels by varying configurations, inlet velocity and inlet sectional liquid holdup. The investigations performed using eight coils, five different inlet velocity and four different inlet sectional liquid holdups. The effects of curvature, torsion, fluid velocity and inlet sectional liquid holdup on two-phase frictional pressure drop have been illustrated. It is found that the two-phase frictional pressure drop relates strongly to the superficial velocities of air or water, and that the curvature and torsion have some effect on the pressure drop for higher Reynolds number flows in large-scale helical rectangular channel; the inlet sectional liquid holdup only increases the magnitude of pressure drop in helical channel and has no effect on the development of pressure drop. The correlation developed predicts the two-phase frictional pressure drop in helical rectangular channel with acceptable statistical accuracy.     

Numerical investigation of water droplet removal characteristics in novel block channels of PEMFC using dynamic wettability model

Liquid water transport and removal is one of the critical issues in the proton exchange membrane fuel cell (PEMFC) for achieving good performance and durability. In this study, two novel channels with different blocks are designed to study their effects on water removal using the volume of fluid (VOF) model considering the dynamic contact angle effect. It is found that compared with the conventional straight channel, both the one-block and two-block channels can promote liquid water removal. The one-block channel leads to faster water movement and removal on the gas diffusion layer (GDL) surface, but results in a much higher pressure drop. The separated two-block channel directly drags water away from the GDL surface by the capillary wicking effect of the block surface, achieving both faster water removal and smaller pressure drop. Effects of the droplet size, air velocity and static contact angle of GDL surface on water removal are investigated comprehensively in both the novel channels, as well as the conventional straight channel, with particular attention on the variations of water removal time, water coverage ratio and pressure drop.     

Dynamic water transport and droplet emergence in PEMFC gas diffusion layers

The dynamics of liquid water transport through the gas diffusion layer (GDL) and into a gas flow channel are investigated with an ex situ experimental setup. Liquid water is injected through the bottom surface of the GDL, and the through-plane liquid pressure drop, droplet emergence and droplet detachment are studied. The dynamic behaviour of water transport in and on the surface of the GDL is observed through fluorescence microscopy, and the through-plane liquid pressure drop is measured with a pressure transducer. With an initially dry GDL, the initial breakthrough of liquid water in the GDL is preceded by a substantial growth of liquid water pressure. Post-breakthrough, droplets emerge with a high frequency, until a quasi-equilibrium liquid water pressure is achieved. The droplet emergence/detachment regime is followed by a transition into a slug formation regime. During the slug formation regime, droplets tend to pin near the breakthrough location, and the overall channel water content increases due to pinning and the formation of water slugs. Droplets emerge from the GDL at preferential breakthrough locations; however, these breakthrough locations change intermittently, suggesting a dynamic interconnection of water pathways within the GDL. The experiments are complemented by computational fluid dynamics (CFD) simulations using the volume of fluid method to illustrate the dynamic eruption mechanism.     

Three dimensional numerical simulation of gas-liquid two-phase flow patterns in a polymer-electrolyte membrane fuel cells gas flow channel

Water management in polymer-electrolyte membrane fuel cells (PEMFCs) has a major impact on fuel cell performance and durability. To investigate the two-phase flow patterns in PEMFC gas flow channels, the volume of fluid (VOF) method was employed to simulate the air-water flow in a 3D cuboid channel with a 1.0 mm × 1.0 mm square cross section and a 100 mm in length. The microstructure of gas diffusion layers (GDLs) was simplified by a number of representative opening pores on the 2D GDL surface. Water was injected from those pores to simulate water generation by the electrochemical reaction at the cathode side. Operating conditions and material properties were selected according to realistic fuel cell operating conditions. The water injection rate was also amplified 10 times, 100 times and 1000 times to study the flow pattern formation and transition in the channel. Simulation results show that, as the flow develops, the flow pattern evolves from corner droplet flow to top wall film flow, then annular flow, and finally slug flow. The total pressure drop increases exponentially with the increase in water volume fraction, which suggests that water accumulation should be avoided to reduce parasitic energy loss. The effect of material wettability was also studied by changing the contact angle of the GDL surface and channel walls, separately. It is shown that using a more hydrophobic GDL surface is helpful to expel water from the GDL surface, but increases the pressure drop. Using a more hydrophilic channel wall reduces the pressure drop, but increases the water residence time and water coverage of the GDL surface.     

Numerical investigation of drop dynamics in presence of wettability gradient inside a serpentine channel of proton exchange membrane fuel cell

Water management significantly affects the performance of a proton exchange membrane fuel cell (PEMFC). Therefore, interest is felt to numerically investigate water droplet movement and slug formation inside the microchannel (gas) of PEMFC. Two important parameters—water coverage ratio and pressure drop have been studied in detail. A U-shaped geometry with a round corner is used for this purpose. 3D unsteady-state models are used to study the drop dynamics using commercial CFD software ANSYS FLUENT 18. For tracking of water drop dynamics, the volume of fluid model is used. Two different situations are simulated. In the first case, the investigation of hydrodynamics of the 0.4 mm drop adhered to the surface of the gas diffusion layer (GDL) has been done. In the second case, simulation of air-water slug flow has been done. GDL surfaces at upstream and downstream of bend are modified using user-defined functions, such that the GDL surface has a dynamic contact angle with respect to the direction of flow. This makes it a continuously hydrophilic surface at upstream and continuously hydrophobic surface at downstream with respect to the direction of flow. The impact of GDL wettability on water retention and removal has been discussed. It is noted that the presence of a gradient facilitates the removal of water drop adhered to the GDL surface. For the case of a suspended drop with an increase of 1°/mm in the magnitude of the gradient, a decrease of 30% is observed in water coverage ratio and pressure drop observed in the channel. Such modified surfaces aids in the conversion of slugs to film at the downstream of bend that reduces maldistribution. The pressure fluctuations and average pressure drop are reduced by 66% when subjected to the aforementioned hybrid gradient.     

Numerical simulation of liquid water emerging and transport in the flow channel of PEMFC using the volume of fluid method

Three-dimensional numerical simulation of liquid water emerging from the gas diffusion layer (GDL) surface to the gas flow channel in the proton exchange membrane (PEM) fuel cell (PEMFC) is carried out using the volume of fluid (VOF) method. The effects of the water velocity in the GDL hole, the airflow velocity and the wettability of the channel surfaces on the water emerging process and transport in the flow channel are investigated. It is found that at low water velocity, the water detaches from the water hole, forming discrete water droplets on the GDL surface, and is transported downstream on the GDL surface until removed from the GDL surface by the U-turn part of the flow channel; whereas at high water velocity, the continuous water column impinges the hydrophilic channel surface counter to the GDL surface, being directly removed from the GDL surface. The airflow velocity affects water detachment and impact process in the channel corner, and water droplet breakup is observed under high airflow velocity. The channel surface wettability influences water droplet shape and its transport in the channel. Rather than forming corner water films at the U-turn for hydrophilic channel surface, water maintains the droplet shape and smoothly passes through the U-turn for hydrophobic channel surface. The importance of the U-turn to the water removal is also discussed. The U-turn promotes water removal from the GDL surface at low water velocity and water breakup at high airflow velocity.     

Effects of needle orientation and gas velocity on water transport and removal in a modified PEMFC gas flow channel having a hydrophilic needle

Water management is critical to the performance and operation of the proton exchange membrane fuel cell (PEMFC). Effective water removal from the gas diffusion layer (GDL) surface exposed to the gas flow channel in PEMFC mitigates the water flooding of and improves the reactants transport into the GDL, hence benefiting the PEMFC performance. In this study, a 3D numerical investigation of water removal from the GDL surface in a modified PEMFC gas flow channel having a hydrophilic needle is carried out. The effects of the needle orientation (inclination angle) and gas velocity on the water transport and removal are investigated. The results show that the water is removed from the GDL surface in the channel for a large range of the needle inclination angle and gas velocity. The water is removed more effectively, and the pressure drop for the flow in the channel is smaller for a smaller needle inclination angle. It is also found that the modified channel is more effective and viable for water removal in fuel cells operated at smaller gas velocity.