AI-Driven Microstructure Design and Simulation for Ceramic Filter Materials Using GeoDict

To realize a carbon-neutral society, advanced separation technologies are indispensable, and ceramic filters are attracting increasing attention as key components in these systems. In this context, we are developing a wide range of ceramic filters. The performance of these filters—such as permeability, separation efficiency, and durability—strongly depends on the microstructure of the porous materials. Therefore, material design and evaluation at the microscale are essential for achieving optimal performance.

To address this challenge, we focus on microstructure analysis and material property simulation using GeoDict. Furthermore, by combining artificial intelligence (AI) technologies, including Generative Adversarial Networks (GANs), we aim to accelerate the development of innovative ceramic materials.

This presentation introduces our recent efforts to enhance the efficiency of the design and optimization processes for ceramic filter materials through the combined use of simulation and AI technologies.