Researchers at the Korea Institute of Energy Research (KIER) have developed a new method for analyzing the microstructure of carbon fibre paper, a crucial component in hydrogen fuel cells. This innovative approach utilizes digital twin technology and artificial intelligence (AI) to conduct analyses at speeds 100 times faster than traditional methods.
The carbon fibre paper plays a vital role in facilitating water discharge and fuel supply within hydrogen fuel cell stacks. The research team, led by Dr. Chi-Young Jung, employed X-ray diagnostics combined with an AI-based image learning model to achieve precise analysis without the need for an electron microscope. This advancement allows for near real-time condition diagnosis.
The study involved extracting 5,000 images from over 200 samples of carbon fibre paper, which were used to train a machine learning algorithm. The resulting model can predict the 3D distribution and arrangement of the paper's key components—including carbon fibers, binders, and coatings—with an accuracy exceeding 98%.
Dr. Jung emphasized the significance of this study, stating that it enhances analysis technology by merging AI with virtual space utilization. This development clarifies the relationship between the structure and properties of energy materials, showcasing its practical applicability. The research team anticipates that this technology will also benefit related fields such as secondary batteries and water electrolysis in the future.