MUlMCOlMla MlaTa

and five for the cathode

where M represents moles of a given component. Experimental

A unique single cell fixture and test stand were designed and constructed to provide tight control on all testing parameters, including mass flow control of gas streams, active humidification control using computer controlled liquid chromatography pumps with constant temperature vaporizers. Both the anode and cathode exhaust streams were condensed and the total water content measured gravimetrically. Water mass balances were obtained with 2% accuracy. Cell temperature was controlled using circulation of a constant temperature fluid through heat exchangers built into the single cell. A data acquisition and control system allowed unattended operation with programmed gas stoichiometrics and humidification (6,7). High frequency resistance of the MEA, the "membrane resistance", was measured at 1 kHz by a 4 probe method.

Membrane electrode assemblies (MEAs) were assembled by a procedure similar to that described by Wilson et. al. (8). Two decals were coated with a carefully controlled thickness of catalyzed ink in the tertiarybutyl amine form and dried in an oven. One decal was placed on each side of the membrane and pressed under controlled conditions. The catalyzed membrane was then sandwiched between two pieces of wet-proofed graphite paper and placed in the test cell. Active area was 46 cm2.

Model Validation

A series of PEM fuel cell tests provided a data base for validating the model. While the test stand was designed with the ability to perform constant stoichiometry polarization curves, the tests shown here were operated at constant flow to minimize the scan time. At high flow rate with saturated gas streams, Fig. 1(a) shows that a good fit between the model and data is obtained. Two adjustable parameters are required to obtain this match: 1.) pore volume occupied by water in the catalyst layer and 2.) water in the saturated region of the cathode backing layer. At lower flow with an unsaturated cathode feed, performance near the mass transfer limiting current is over estimated by the model (Fig. 1(b)). Apparently, flow distribution, flow field lands, or partial flooding at the backing-membrane interface occurs. These processes are not accounted for in the model at lower stoichiometrics.

Current

Figure 1. Single Cell Validation: (a) 85°C, Anode: H2, 2.5 atm, 100% RH, 1.0 slpm, Cathode: Air, 2.5 atm, 100% RH, 5.0 slpm (b) same except Anode: 0.34 slpm Cathode: 0.77 slpm, 50% RH (flow rate corresponds to a Stoichiometry of 1 @ 1 amp/cm2).

Current

Figure 1. Single Cell Validation: (a) 85°C, Anode: H2, 2.5 atm, 100% RH, 1.0 slpm, Cathode: Air, 2.5 atm, 100% RH, 5.0 slpm (b) same except Anode: 0.34 slpm Cathode: 0.77 slpm, 50% RH (flow rate corresponds to a Stoichiometry of 1 @ 1 amp/cm2).

0 0

Post a comment