Fig. 1 Picture of the high-temperature NMR probe.53The temperature of the sample was varied from 296 to 724 K by a flow of heated nitrogen gas. The temperature of the nitrogen gas was monitored by a thermocouple. The accuracy of the temperature calibration was confirmed by 79Br NMR measurements of KBr.[1] The high-temperature NMR probe was applied for the measurement of the Li-ion diffusion in Perovskyte-type solid electrolyte. Non-Arrehenius behavior of the diffusion coefficients was revealed by the measurement. [2]P1-15Development and Application of High-temperature NMRKenjiro HashiCenter for Basic Research on Materials, National Institute for Materials Science (NIMS)Understanding the diffusion property of Li ions is crucial for improving the performance of Li-ion batteries. NMR is one of the most important techniques for characterizing the Li-ion mobility from microscopic points of view. However, the Li-ion mobility of solid electrolyte is too slow to explore with NMR around room temperature. To investigate the ion diffusion at high temperatures, we constructed a high-temperature pulsed-field-gradient (PFG) NMR probe capable of measurements at temperatures > 700K.[1] K. Hashi, et al., Analytical Sciences, 37, 1477 (2021). [2] G. Hasegawa, et al., Chemistry of Materials, 35, 3815 (2023). P1-16Poster Award NomineeCharacterization of the Tensile Properties of GAN-generated 3D Microstructures in Dual-Phase Steels Ta-Te Chen1, Ikumu Watanabe2, Keiya Sugiura1, Toshio Ogawa3, and Yoshitaka Adachi1 1 Graduate School of Engineering, Nagoya University 2 Center for Basic Research on Materials, National Institute for Materials Science (NIMS) 3 Department of Mechanical Engineering, Faculty of Engineering, Aichi Institute of Technology This study examines the validity of 3D microstructures generated by Generative Adversarial Networks (GANs) for dual-phase steels [1]. We compared GAN-generated structures with experimentally observed 3D microstructures for two types of ferrite-martensite dual-phase steels. Mechanical behaviors were analyzed using finite element analysis on representative volume elements, employing an image-based modeling approach with voxel coarsening. Results showed that the GAN algorithm successfully captured anisotropic responses caused by microscopic morphology, particularly in steel with higher martensite content. However, quantitative alignment with observed microstructures was limited due to inaccuracies in reproducing phase volume fractions. In the steel with lower martensite content, the algorithm struggled to replicate the connectivity of the martensite phase. This study highlights both the potential and limitations of GAN-based 3D microstructure generation, contributing to the development of computational methods for materials characterization and design. [1] I. Watanabe, K. Sugiura, T. Chen, T. Ogawa, and Y. Adachi, Science and Technology of Advanced Materials, 25, 2388501 (2024).
元のページ ../index.html#53