Javi's Final Project

From WikIRI

Back to Main Page

Computational Thermal and Pressure drop model of a Polymer Electrolyte Membrane Fuel Cell

File:Copy articlejam mo03.doc Word document with the current version of the paper.

Abstract

The aim of this paper is to present a simple computation model of a Polymer Electrolyte Membrane Fuel Cell (PEMFC) that allows for the simulation of heat distribution and the properties of the reactant gas flows in and out of the cell (including pressure drop, humidity and liquid water content). Although the theoretical model can be adapted to any type of fuel cell, in order to present some results and verify it, it has been adapted to the single cell test fixture. Some of the results herein presented are: Temperature distribution inside the cell, critical current at which liquid water is created and pressure drop for both reactant, under a verity of operating conditions. JAVI State some conditions and there results

1. Introduction to the model

The model described here is based on the numerical resolution of heat transfer problems. The challenges imply the resolution of various equilibrium differential equations, which can be accomplished using numerical iterative methods. Both reactant gases streams can be modeled separately and interact with the thermal model through convection heat transfer in channels (1996 Incropera). The pressure drop in the channels is determined using both linear and singular(WHAT DOES SINGULAR MEAN? minor is better then singular) losses for laminar flows (1999 White). Water and heat created inside the fuel cell due to the reaction is determined by commonly used equations 2005 Barbir Book.

2. Description of the model

The model can be conceived as two sub models working together, a thermal and a pressure drop one.

2.1 Thermal Model:

The thermal equilibrium equation (equations do not work yet in the wiki)[2.1.1] can be approximately solved by creating a mesh of the fuel cell and assuming constant temperature in each cube inside the mesh (each cube is represented by a node, situated in its midpoint). It should be noted that a appropriate mesh is crucial to obtain good results, and using the following finite difference explicit numerical method: [2.1.2], were the thermal conductivity between nodes can be obtained by the equations: 0 for non adjacent nodes, [2.1.3] or [2.2.4] for adjacent nodes and [2.1.5] between a surface node and the exterior.

In order to build a program that computes the model, it is recommended to us the matrix notation: [2.1.6]. Although it seems to be a quadratic expression, it can be computed as a linear expression due to the diagonal matrix MC and MK which have a maximum of 7 non-zero values per file and column. This structure reduces significantly the computational time needed by the program.

The channels may be included as isolated nodes with all thermal properties equal zero. The heat transfer between gas and channel is introduced as a thermal generation inside the adjacent nodes. Thermal generation inside the active area is distributed between its nodes.

2.1.1 Thermal generation at active area

Quantity of heat created by the reaction inside the fuel cell can be obtained by its efficiency, using the equation: where [2.1.7]

4. Experimental verification of the model

4.1 Experimental methodology

In this section we present the single cell test fixture, the experimental setups, and the LabView hardware and software used for data acquisition.

4.1.1 Cell hardware

The single cell test fixture used in this paper is an ElectoChem® (model # EFC-05EFC-05-02-02) 5 cm2 active area fuel cell. It is equipped with a Nafion® 115 membrane with 1 mgPt/cm2 catalyst loading and Toray® carbon fiber paper (Type TGP-H-060) gas diffusion layers, serpentine/straight channel flow field with groups of three channel that come back together at each turn in a uniting manifold. There are five straight sections in the flow field which makes is similar to a three pass serpentine configuration. INSERT FIGURE OF FLOW FIELD. The dimensions of the channel are 0.78mm wide, 0.78mm deep, and a space between each channel is 0.78mm. The length of straight section of the channel is 23.25mm. The gases enter the flow field through two 1.6mm diameter holes and exit through the same size holes. The flow fields are machined into POCO® graphite plates that are 19.1mm thick. The Teflon® fitting which connect the fuel cell to the test station are also screwed into the graphite plates. The current collector bus plates are gold plated copper for enhanced surface conductivity. Resistive square planner 60W heaters with adhesive are in the center of each bus plate with the dimensions of 50mm wide.

4.1.2 Experimental setup

The test station consists of two reactant (anode and cathode) gas subsystems. Each subsystem contains a Bronkhorst® mass flow controller, membrane based humidification with dew point sensors for control, inlet line heaters to prevent condensation, absolute pressure transducers at the inlet, differential pressure transducers between the inlet and outlet of each reactant, and back pressure regulator at the outlet of the fuel cell to control system pressure. The mass flow controllers are each calibrated for there specific gas (Hydrogen for the anode and synthetic air for the cathode).

There are 8 temperature reading that come from the fuel cell by way of K Type thermal couples. Four of the measurements are of the graphite plates (two in each plate) with one connected to a RedLion® PID controller model #T4810105, which controls the temperature of the fuel cell. The cooling of the cell is attained mainly by natural convection. The other four temperature measurements are of the reactants inlets and outlets. The inlet temperature measurements are close to the outlet of the gas line heaters but out side of the cell. The outlet temperatures are measured inside the fuel cell in the outlet manifold. All the measurement and the control are made with in real time through LabView® which is explained in more detail in the following section.

4.1.3 Data acquisition system

Two computers define the data acquisition and control system. The first one is called host and has the responsibility to allow the user to start, finish, change configurations and control settings during the operation. All of this by means of a graphical interface developed by our laboratory through LabView.

Second computer runs under a real time operating system and works in deterministic mode offering consistent and stable functionality, implementation of controllers and stores the data. Inside real time computer there are CompactRIO cards with input/output modules that contain configurable signal conditioning, isolation and screw terminals to provide direct connections to our sensors and actuators.

References Javi´s Project

Journal Publication References

Books References

Fuel Cell Systems Explained Second Edition

Fuel Cell Theory and Practice