Realization of image-based massively-parallel finite element analysis using synchrotron radiation CT for metallographic structures

2017 Fiscal Year Final Research Report

• Project/Area Number: 16K13735
• Research Category: Grant-in-Aid for Challenging Exploratory Research
• Allocation Type: Multi-year Fund
• Research Field: Computational science
• Research Institution: University of Fukui
• Principal Investigator: KUWAZURU Osamu 福井大学, 学術研究院工学系部門, 准教授 (40334362)
• Co-Investigator(Renkei-kenkyūsha): TERANISHI Masaki 福井大学, 学術研究院工学系部門, 助教 (80798322)
• Project Period (FY): 2016-04-01 – 2018-03-31
• Keywords: 大規模計算 / マイクロメカニクス / 放射光CT / 超並列 / 有限要素解析 / 疲労 / アルミニウム合金

Outline of Final Research Achievements

A feasibility of large-scale image-based simulation using a supercomputer and a synchrotron radiation CT was examined. To overcome the difficulty in the mesh generation based on images, which encumber the large-scale simulation, we employed the voxel element in line with the voxel-to-element strategy. As a result, it became possible to perform a large-scale finite element analysis considering microscopic metallographic structures and we clarified that the elastic-plastic and finite strain analysis can be done up to about one hundred million elements. Moreover, we introduced a heterogeneous voxel element which can use different material property at each integration point within a single element, and examined that this element could reduce the numerical error due to the inaccuracy in shape which is specific to the voxel element. We conducted the micromechanical fatigue simulation for the cast aluminum alloy and discussed the mechanism of fatigue crack initiation from material surfaces.

Free Research Field

計算固体力学