Abstract

During the operation of a proton exchange membrane (PEM) fuel cell water is produced as a byproduct of electrochemical reactions. As the produced water passes through the porous structure of the electrodes and enters the flow channels, liquid–gas two-phase flow forms within the flow channels. If a reliable model that can accurately predict the pressure drop is available, then the two-phase flow pressure drop can be used as an in-situ diagnostic tool to quantify the water content within the flow channels. While the two-phase flow pressure drop is investigated for straight sections of PEM fuel cell flow channels, few attention has been paid to evaluate the pressure drop across flow channel bends. In this study, liquid–gas two-phase flow across a flow channel bend is investigated with an ex-situ approach. A 90∘ flow channel bend was fabricated across a channel with a hydraulic diameter of 1 mm. Experiments were conducted by supplying air and hydrogen as the gas phase and injecting liquid water into the flow channel in order to represent the two-phase flow condition in the cathode and anode flow channels, respectively. The two-phase flow across the bend was visualized with a CCD camera and the pressure drop values were measured for the PEM fuel cell flow conditions. A comparison between the experimental data and predictions from some of the existing models prompted the need for a model that can accurately predict the two-phase flow pressure drop across PEM fuel cell flow channel bends. In this study, a new model to predict the frictional two-phase flow pressure drop across flow channel bends was developed based on the separated flow model. Because the two-phase flow in PEM fuel cell flow channels is dominated by the gas phase, the new model incorporated the gas minor loss which was defined as the pressure drop caused by the flow channel bend. In order to predict the gas minor loss, a correlation for the loss coefficient was obtained based on the gas single-phase pressure drop. In addition, a correlation was proposed for the C parameter as the two-phase flow across the flow channel bend featured an enhanced interaction between the two phases of liquid and gas. Results indicated that the proposed two-phase flow pressure drop model along with the proposed C correlation were able to predict the pressure drop with mean absolute error values of 7.74% for air–water runs and 10.77% for hydrogen–water runs.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call