Development of a new stability theory for general vortices in transitional and turbulent flows based on hierarchical vortex clustering

Reference No.	2022a023
Type/Category	Grant for General Research- Short-term Visiting Researcher
Title of Research Project	Development of a new stability theory for general vortices in transitional and turbulent flows based on hierarchical vortex clustering
Principal Investigator	Kazuo Matsuura（Graduate School of Science and Engineering, Ehime University・Associate Professor）
Research Period	January 17,2023. ～ January 18,2023.
Keyword(s) of Research Fields	Fluid mechanics, Vortex, Machine learning, laminar-turbulent transition, Compressible Navier-Stokes equation, Direct numerical simulation
Abstract for Research Report	The design of transition and turbulence based on the accurate prediction and reliable laws of nonlinear dynamics during the process of laminar-turbulent transition is a central problem of fluid mechanics. Boundary-layer transition becomes the problem in various industrial applications such as gas turbines, airplanes, pipe lines regarding efficiency, body stability, noise and vibration. A preliminary design regarding how turbulent states are utilized, which is supported by robust laws, becomes important because turbulence is associated with momentum transfer and entropy increase due to energy dissipation, and has ambivalent effects such as the prevention of boundary-layer separation, the suppression of discrete tone noise, the acceleration of heat and mass transfer, and drag increase. Although transition process has been investigated by various linear & nonlinear stability analyses, and also direct numerical simulation (DNS), ambiguity has been remaining regarding the onset of turbulence and its sustenance due to the strong nonlinearity and the large degree of freedom. On the other hand, it is expected that further fusion of transition simulation based on the Navier-Stokes equation and big data science such as machine learning will lead to the further findings and understandings of transition and successive turbulence. Therefore, in this study, compressible DNS is conducted, the data mining of its resultant large-scale unsteady data is then conducted, and then machine learning theory that enables a unified stability analysis in addition to finding universal mechanisms governing the dynamical behaviors of the group of unsteady vortices appearing in laminar-turbulent transition is developed.
Organizing Committee Members (Workshop) Participants (Short-term Joint Usage)	Yasuhide Fukumoto（IMI, Kyushu University・Professor）