名古屋大学 大学院工学研究科 土木工学専攻 構造・材料工学講座 構造解析学グループ

We study topology optimization methods and Eulerian finite volume methods suitable for massively parallel structure and fluid dynamics simulation. These researches can be categorized as computational mechanics, a critical technology for future science and engineering.

​

Since computational mechanics can be applied to almost all engineering fields, we conduct cross-disciplinary research in civil, automotive, aerospace, robotics, and material engineering. We also investigate numerical methods related to informatics, such as machine learning.

​

Structural dynamics are rarely determined by the structure alone. We also need to consider the interaction with the surrounding fluid, such as a high-rise building swaying in strong wind. Therefore, in our laboratory, we also study a unified simulation method for structure and fluid dynamics suitable for massively parallel computers such as the supercomputer Fugaku.

 

The computational performance has been increasing exponentially. For example, the iPhone 13 released in 2021 has almost the same computational performance as the US supercomputer Intel ASCI RED, the fastest supercomputer in the world in 1997. In other words, the present supercomputer is a future personal computer. Thus research on numerical methods suitable for supercomputers is indispensable for digital innovation.

​

3D printing or additive manufacturing has been causing technological innovation in recent years. Topology optimization is suitable for additive manufacturing, so novel manufacturing can be overcome by integrating topology optimization and additive manufacturing.