Effects Of MEA Compression On Fuel Cell Stress, Resistances And Water Transmission

Polymer electrolyte membrane (PEM) fuel cells are energy conversion devicessuitable for automotive, stationary and portable applications. A membrane electrodeassembly (MEA) is one of the core components of PEM fuel cells (PEMFCs)comprising a catalyst coated membrane, gas diffusion electrodes, frames and seals.An engineering challenge that is hindering the widespread use of PEM fuel cells isMembrane electrode assembly and its effects on water transmission issues. Theseissues are addressed by this dissertation through simulation and experiments.(1) By employing a2-D finite element model the mechanical stresses inmembrane caused during the cell assembly procedure are analyzed for different MEAframe materials, frame structures and contact behaviors. A zone with strongnonuniform stresses in the membrane under the end edge of frame/membrane isobserved. The stepped frames assembly (SFA) and the bonded contact behaviors havemore uniform stress distributions than the aligned frames assembly (AFA) does. Theframe materials are not the main factors affecting the stress concentration inmembrane for the SFA, while for the AFA the mechanical properties of framematerials should be CLose to those of the membrane.(2) A multi-electrode probe is employed to distinguish the bulk and contactresistances of the catalyst layer (CL) and the gas diffusion layer (GDL) with thebipolar plate (BPP). Resistances are compared for Vulcan carbon catalyst layers,carbon paper and carbon CLoth GDL materials, and GDLs with microporous layers(MPL). The Vulcan carbon catalyst layer bulk resistance is100times greater than thebulk resistance of carbon paper GDL. Carbon CLoth has bulk and contact resistancestwice those of carbon paper. Compression of the GDL decreases the GDL contactresistance, but has little effect on the bulk resistance. Treatment of the GDL withpolytetrafluoroethylene increases the contact resistance, but has little effect on thebulk resistance. A microporous layer (MPL) added to the GDL decreases the contactresistance, but has little effect on the bulk resistance. An equivalent circuit model shows that for channels less than1mm wide the contact resistance is the majorsource of electronic resistance and is about10%of the total ohmic resistanceassociated with the membrane electrode assembly.(3) By seting experiment model to analyze the path of water flow correspondingto the flow resistances. Liquid water travels both laterally (interface flow) andtransversely through the largest pores of the porous GDL structure. Narrow aperturesin the largest pores are the primary resistance to liquid water penetration. Aftersufficient hydrostatic pressure is applied, water penetrates the limiting aperture andflows through the pore. The pressure required for water to flow through the pores isless than the pressure to penetrate the limiting aperture of the pores. There is lessresistance to lateral liquid water flow at the interface between the MPL andmembrane surface than through the GDL.(4) Water generating in the catalyst layer flows through the GDL and emerges inthe microfluidlic channels of fuel cells. Characters of liquid water motion in a singlemicrofluidlic channel with bend are investigated corresponding to various Reynoldsnumber and different compression ratios of GDL. Gas can bypass the rib to the nextdownstream part of channel and flow through the GDL underneath the slug. In orderto remove slug, the percentage of bypass distance and the gas flow rate should belarger than13%. The cross-section shape of the slug in1hydrophobic GDL wall and3hydrophilic acrylic walls is an analogous upside-down trapezoid. It is suggested thatthe compression ratio of GDL should be over13%to eliminate liquid water instead ofsimply increase the volumetric gas flow. Residual droplets were received at thevertical corner compared to the arc corner in the channel. The regimes of liquid waterflow can be summarized as slug motion, compressed drop and elongate dropletmotion, drop oscillation and cap drop motion.(5) The effects of gas diffusion layer compression on gas bypass and water slugmotion in parallel channels were studied. Water slugs in gas flow channels can forcegas to flow laterally through GDLs in PEMFCs. Slugs ocCLude the cross-sectionalarea of the gas flow channel and force gas to flow underneath water slugs or beneaththe ribs between gas flow channels. Gas bypass underneath the ribs between channelsresults in synchronous motion of slugs in parallel channels. The gas flow through the GDL depend the GDL permeability, which is reduced when the GDL is compressedduring the assembly of fuel cells. The characteristics of slug motion and gas bypassare analyzed in microfluidic parallel channels with one wall comprised of a GDL;compression of the GDL by the rib between parallel channels mimics thecompression underneath ribs of a fuel cell bipolar plate. The physics of coupled gasflow in flow channels and through the GDL are described. It is shown that gas flow inthe GDL between adjacent channels results in synchronous slug motion. Anexperimental procedure is introduced in which the volume of slugs and the distancethey move past an obstruction can be employed to determine the GDL permeabilityunder the channel and the GDL permeability under the rib.