A team led by Dr. Chi-Young Jung at the Korea Institute of Energy Research (KIER) has developed a groundbreaking method to quickly identify problems in hydrogen fuel cells. This method uses AI and digital twin technology to analyze the microstructure of carbon fiber paper, a key material in fuel cells, 100 times faster than before.
Carbon fiber paper is essential in hydrogen fuel cells as it helps with water removal and fuel supply. Over time, its structure and coating change, which lowers fuel cell performance. Traditional methods for analyzing this material involved damaging samples and using electron microscopes, which was time-consuming and inefficient.
The new technology uses X-ray diagnostics and an AI-based model to analyze the material without damaging it. By training the AI with 5,000 images from over 200 samples, the system achieves over 98% accuracy in identifying the 3D structure of carbon fiber paper. This allows researchers to detect damage or performance issues in seconds instead of hours.
Additionally, the team used this model to study how factors like thickness and binder content affect fuel cell performance. They proposed design improvements to enhance efficiency.
Dr. Jung highlighted the significance of this research, saying it combines AI and virtual space technology to better understand energy materials. He believes this innovation could also benefit other fields like battery technology and water electrolysis in the future.
Source: newswise
