Comprehensive Analysis of Hydrogen-Induced Material Degradation Mechanisms through Dynamic Calculations

2023 Fiscal Year Final Research Report

• Project/Area Number: 20K15013
• Research Category: Grant-in-Aid for Early-Career Scientists
• Allocation Type: Multi-year Fund
• Review Section: Basic Section 26010:Metallic material properties-related
• Research Institution: The University of Tokyo
• Principal Investigator: Shimizu Koji 東京大学, 大学院工学系研究科(工学部), 助教 (00838378)
• Project Period (FY): 2020-04-01 – 2024-03-31
• Keywords: 機械学習ポテンシャル / 第一原理計算 / 水素脆化

Outline of Final Research Achievements

This study aimed to understand the mechanism of material degradation induced by hydrogen from atomic-scale using machine-learning interatomic potentials constructed using first-principles calculations based on density functional theory. Specifically, we accessed the accuracy of predicting hydrogen diffusion in aluminum and performed dynamical calculations considering the quantum effects of hydrogen nuclei. Furthermore, stress tests were conducted through molecular dynamics simulations using relatively large-scale computational models, revealing the hydrogen aggregation within aluminum which accelerates material fracture. Additionally, we conducted further analysis on the alumina-aluminum system, obtaining preliminary results on hydrogen behavior near interfaces.

Free Research Field

計算科学

Academic Significance and Societal Importance of the Research Achievements

水素による材料劣化機構の理解は水素社会の実現に向けた重要な課題であり、材料中における水素の拡散や凝集、欠陥との相互作用等について原子レベルでの解析をより一層推し進める必要がある。本研究では、アルミニウム系材料に着目しそこでの水素の動的過程を調査するために、機械学習を用いた計算手法の開発と応用に取り組んだ。本計算手法を用いた多角的な水素挙動の解析によって、材料劣化につながる要因について様々な知見を得ることができた。