Laser interferometry has been combined with high-speed digital cinematography to measure time-averaged turbulent convective heat transfer rates. The method has been demonstrated for turbulent free convection in a tall vertical enclosure filled with air. Sample results have been obtained for a Rayleigh number (based on the enclosure width) of 2.8×10 5 and an enclosure aspect ratio of 17.1. An automated digital image processing algorithm has been used to calculate the instantaneous local heat flux from the sequence of interferograms captured by a high-speed camera. The effects of key experimental parameters, such as the camera frame rate and the total image capture time, have been investigated. The average Nusselt number for the entire enclosure was found to compare well with a widely used empirical correlation from the literature.

A numerical model was developed to study the effects of channel length and bend shape on the flow cross-over through the porous gas diffusion layer (GDL) and the pressure distribution in a PEM fuel cell flow plate with a serpentine channel flow system. Usually, on the cathode side of a PEM fuel cell, air flow through a flow plate with serpentine channels with certain lengths, to supply the oxygen to the catalyst layer for the chemical reaction. There is a porous GDL between the flow plate and the catalyst layer. Flow cross-over of air through the porous GDL from one part of the channel to another can occur because of the pressure differences existing between different parts of the channel. This cross-over causes the flow rate through the channel to vary with distance along the channel, and also has an influence on the pressure distribution through the plate and, eventually, the fuel cell performance. For the conventional channel flow, the pressure drop is proportional to the channel length. To study the importance of this channel length effect on the PEM fuel cell flow field with cross-over, the pressure distribution and flow rate variation along the channels have been examined by numerically solving for the flow through the plate and porous GDL assembly. Attention has been given here to serpentine channel flow systems with single and parallel channel patterns and with different numbers of passes. A 3-D, single-phase flow has been considered. It was assumed that the flow is steady and incompressible, and the flow through the porous diffusion layer can be described using the Darcy law. The governing equations have been written in dimensionless form using the channel width as the length scale and the mean velocity at the channel inlet as the velocity scale. The resulting set of dimensionless governing equations has been solved using the commercial finite element method (FEM) software package, FIDAP . The solution was obtained by simultaneously solving the equations for the flow in the channels and for the flow through the porous GDL. The solution depends on the following parameters, (1) the Reynolds number, Re , based on the channel width and on the mean velocity at the channel inlet, (2) the dimensionless GDL permeability, (3) the dimensionless channel length, (4) the bend shape, and (5) channel configurations. The main emphasis of this study was on the effect of channel length. The numerical results obtained indicate that the channel length has a significant effect on the flow cross-over through porous GDL and the pressure distribution in the flow plate.

One of the inter-connected factors that can lead to failures in the flow plates of PEM fuel cells is the pressure differences that exist between adjacent flow channels. These pressure differences lead to stresses in the channel supports, i.e., the ribs, which can be important in the presence of stresses arising due to other factors such as temperature gradients in the flow plates. In order to investigate the magnitudes of the pressure differences across the supports and the places where the maximum pressure differences occur, the flow and pressure variations in various forms of serpentine channels, these channels having a rectangular cross-sectional shape, have been numerically calculated. The presence of the diffusion layer has been ignored and the flow has been calculated using a commercial finite-element software package using the governing equations written in dimensionless form. Solutions have been obtained for various values of the Reynolds number for each of the flow geometries considered for two channel height-to width ratios (one and three). Except for the flow in the vicinity of the bends in channels, the pressure has been found, as is to be expected, not vary significantly across the channel cross-section. The difference between the center point dimensionless pressure in a given channel with those at the same longitudinal position in the adjacent channels has been determined. The dependence of the highest dimensionless pressure difference between channel on the input parameters has been examined.

Much effort has been expended in the past few years upon development of numerical models to obtain the detailed flow, current and temperature distributions in Polymer Electrolyte Membrane Fuel Cell (PEMFC). Therefore, the need for model validation also has increased to gain confidence in the accuracy of the numerical results. In the present work, a numerical model has been developed to study the pressure distribution in the flow field plate (FFP) and gas diffusion layer (GDL) assembly on the cathode side of a PEM fuel cell. The flow field plate has serpentine channels and the porous gas diffusion layer is adjacent to the flow field plate, to deliver the air to the catalyst layer where the electrochemical reaction occurs. Flow crossover of air through the porous GDL under the land from one part of the channel to another can occur, and this flow crossover affects the total pressure drop between the channel inlet and outlet, and the pressure difference between adjacent channels. The flow here has been assumed to be three-dimensional, steady, incompressible, isothermal and single-phase. The flow through the porous GDL has been described using the Darcy model. The governing equations have been written in dimensionless form and solved by using the commercial CFD solver, FIDAP . In parallel, experimental work has been conducted at the Queen’s-RMC Fuel Cell Research Center (FCRC), Canada, for comparison with the numerical results. The cathode FFP has a single serpentine channel. Flow of dry air at 20 °C and at 60 °C has been used for measuring pressure differences at specific locations in the flow field plate. The effects of Reynolds number, based on the mean channel width and the mean velocity at the channel inlet (values between 100 and 1500) have been studied. Other parameters that were considered are the land:channel width ratio (2:1 and 1:1) and the permeability of the GDL (values between 1.0E−19m 2 and 1.0E−10m 2 used). Good agreement was obtained between the numerical and experimental pressure distributions along the serpentine channel.

It is common in a PEM fuel cell to have the air flow through serpentine channels with a rectangular cross-sectional shape in a flow plate. There is a porous diffusion layer adjacent to this flow plate. Flow cross-over of air through the porous diffusion layer from one part of the channel to another can occur, as a result of the pressure differences between different parts of the channel, and it causes the flow rate through the channel to vary with distance along the channel and also has an influence on the pressure distribution along the channel. These changes in the pressure distribution as a result of cross-over can effect the fuel cell performance. In the present study the conditions under which cross-over occurs and the effects of the cross-over on the pressure distribution and local channel flow rates have been examined by numerically solving for the flow through the plate-porous layer assembly. Two flow channel arrangements have been considered: (i) a single serpentine channel flow system with different land widths between the channel sections (ii) a two-channel parallel serpentine flow system. A single phase flow has been considered. The governing equations have been written in dimensionless form using the channel width as the length scale and the mean velocity in the channel as the velocity scale. The resultant set of dimensionless equations has been numerically solved using a commercial finite element code, FIDAP. The solution was obtained by simultaneously numerically solving the dimensionless governing equations for the flow in the channels and for the flow through the porous gas diffusion layer. The numerical calculations were obtained using a commercial finite element code, FIDAP.