Implementation of fiber optic temperature sensors for enhanced performance in fuel cell systems

Abstract

Proton exchange membrane (PEM) fuel cells represent a forefront technology in the realm of sustainable mobility, characterized by their impressive efficiency and minimal emissions. Nonetheless, internal thermal gradients significantly impact their performance and durability. Localized hotspots can accelerate material degradation and reduce electrochemical efficiency. Conventional temperature sensors often lack the spatial resolution required to detect these crucial variations. This study addresses this limitation by integrating distributed fiber optic temperature sensors into the gas channels of a PEM fuel cell stack to enable high resolution thermal monitoring. Eight U shaped optical fiber loops are embedded within the bipolar plate, enabling temperature measurements with sub millimeter spatial resolution. A Python based data processing algorithm was developed to calibrate raw sensor output using reference RTD measurements conducted during an oven test. The tool achieved a reduction in temperature error from 17.86 °C to 0.41 °C. The corrected dataset showed a noise reduction of 96.7%, with 84.67% of the values falling within ±0.30 °C of the RTD reference and a mean absolute error of 0.41 °C. Post-processing steps include isolating thermally relevant gas channel regions and removing invalid data near structural interfaces. Temperature fields that have been validated are mapped onto a finite element mesh and visualized with ParaView, resulting in better 3D thermal profiles of the stack. To assess the effects of integration on fluid dynamics, computational fluid dynamics (CFD) simulations were conducted in ANSYS Fluent for three designs: a reference model, a previous integration layout, and an optimized geometry. The results demonstrate that the improved sensor configuration preserves flow uniformity and reduces turbulence,low pressure loss, maintaining efficient reactant distribution at test conditions corresponding to a flow velocity of 10 meters per second. This integrated framework combining calibrated fiber optic sensing, automated correction, and simulation validation offers a robust platform for advanced thermal diagnostics in PEM fuel cells. It supports future developments in real time monitoring, intelligent thermal control, and digital twin technologies, enabling more efficient and reliable fuel cell systems.