Experimental and numerical study of proton exchange membrane fuel cell with spiral flow channels

• Published in: Applied energy Vol. 99; pp. 67 - 79
• Main Authors: Jang, Jiin-Yuh, Cheng, Chin-Hsiang, Liao, Wang-Ting, Huang, Yu-Xian, Tsai, Ying-Chi
• Format: Journal Article
• Language: English
• Published: Kidlington Elsevier Ltd 01.11.2012; Elsevier
• Subjects: Applied sciences; catalysts; cells; Channel design; Energy; Exact sciences and technology; fuel cells; heat; mass transfer; mathematical models; prediction; Proton exchange membrane fuel cell; Spiral channels; Spiral channels; Channel design; Proton exchange membrane fuel cell
• ISSN: 0306-2619; 1872-9118
• DOI: 10.1016/j.apenergy.2012.04.011

► Numerical and experimental study of the fuel cell with spiral channels is performed. ► Secondary vortices in cross section of the spiral channels are found. ► Enhancement in the performance of the fuel cell by the secondary vortices is discussed. ► The spiral channels also lead to a reduction in the pressure drop of the gas flow. Numerical simulation of the performance of a proton exchange membrane fuel cell (PEMFC) with spiral channels is performed in this study. Experiments are also conducted to verify the numerical predictions. The spiral channel pattern produces secondary vortices which lead to enhancement in heat and mass transfer in the curved channels and appreciably improves the performance of the fuel cell. In addition, the spiral channels may also lead to a reduction in the pressure drop of the gas flow through the fuel cell. When the sizes of the outlet channels are designed to be smaller than those of the inlet channels, water flooding in the catalyst layers can be further improved. In the present study, the spiral channel pattern consists of five inlet channels and five outlet channels. Radius and area of the active zone are 28.2mm and 2500mm2, respectively. A comparison between the spiral and the serpentine channels shows that the average current density with the former is higher than that with the latter by 11.9%. It is found that numerical predictions are in close agreement with the experimental results.