超並列ボクセル有限要素解析を用いたアルミニウム鋳造合金の介在物の応力集中評価

抄録

Crack initiation in cast aluminum alloy under low-cycle fatigue loading is mainly caused by the fracture of Si particles. In this study, the large-scale image-based finite element analysis was conducted to reveal the mechanism of Si particle fracture. The fatigue test and in situ CT scanning were performed at SPring-8. The fracture of Si particle was identified by chronological CT observation. The CT images used for the model was extracted into quarter of horizontal cross section where the fractured Si particle for crack initiation was included. The image-based finite element model with voxel element was generated semi-automatically using the extracted CT images. The number of elements for the finite element model was about 130 million. To evaluate the incremental rate of stress for Si particles under cyclic loadings, 10 cycles of loading were applied in the analysis. The large-scale finite element analysis and its post-processing were performed on the supercomputers by the massively-parallel computing. The result of the finite element analysis showed that the value of the first principal stress for Si particles and its incremental rate under cyclic loadings did not only depend on their shapes but also their position in the specimen. Moreover, it was observed that the increment of the first principal stress for the fractured Si particle near a pore under cyclic loadings was drastically changed.