(108e) A Fuel Cell Motivated Model for the Decay of Bubbles Due to Coalescence in Small Channels

The direct methanol fuel cell (DMFC) offers the advantage of a high energy density liquid as a fuel over similar sized hydrogen fuel cells however; the result of the anode reaction is gaseous carbon dioxide (CO2) that must be removed from the cell stack. Since the bubbles are evolved within the liquid fuel at rates of 10's of ml/min per Watt of power, the result is a gas-liquid flow that must carry the gas to a place where the fuel and CO2 can be separated. Several problems arise from the quest for increased power density. The pressure drop of pumping the gas ?liquid mixture can be excessive. If the bubbles are not continually removed from the anode, there can be mass transfer limitations of the fuel to the surface due to a lack of available contact area, particularly at high current densities.

Removal of gas is complicated because while bubbles formed in the gaseous diffusion layer initially appear in the flow field as small spherical bubbles, these coalesce as they travel through the cell, forming large gas slugs. The larger slugs can effective block the flow at the pressure drops that are reasonably available to operate these devices.

In this work the bubbles in a small rectangular channel similar to that found in a DMFC were studied under laminar flow conditions using high speed imaging and computer image processing algorithms. Using data obtained from experiments conducted under various flow conditions, fluid properties, and channel orientations, the change in population and size of bubbles with respect to time were modeled using a second order model derived from kinetic theory. Other items to be discussed include how fluid properties affect the film drainage time essential to coalescence, how that affects bubble contact time requirements for a coalescence event to occur, and some techniques for preventing CO2 bubble formation in a DMFC.