Analysis of Formability in Stamping of Metallic Bipolar Plates with Parallel Flow Field for Proton Exchange Membrane Fuel Cells using Adaptive Neuro-fuzzy Inference System

Abstract

Bipolar plates (BPPs) play an important role in PEM fuel cells in terms of weight and cost points of view. In this paper, the manufacturing of titanium BPPs with parallel flow field was experimentally and numerically studied. In this regard, a stamping die with a parallel pattern is conducted to perform the experiments. Then, the process was modeled via the finite element (FE) simulation. By comparing simulation and experiment results, it was found that the results are in good agreement and hereupon, the accuracy of the FE model was verified. To evaluate the sheet formability, a set of FE experiments was designed through the response surface methodology (RSM). The die clearance, forming velocity, and friction coefficient were considered input parameters, and the maximum thickness reduction (MTR) of the sheet was assumed to be the output. The results revealed that a lower friction coefficient causes an increase in thickness reduction and finally tearing in the formed BPPs. Moreover, changing the forming velocity has no remarkable influence on the MTR. Afterward, an Adaptive Neuro-Fuzzy Inference System (ANFIS) was trained for predicting the output of the MTR with the three mentioned inputs.

Keywords

• adaptive neuro-fuzzy inference system
• bipolar plate
• proton exchange membrane fuel cell
• thickness reduction
• titanium sheet