Computational Fluid Dynamics (CFD)

Corrected Volume Penalization Method for Direct Numerical Simulation of Aeroacoustic Sound

The aeroacoustic noises radiated from airplanes, high-speed trains, helicopters, wind turbines etc. are one of the great concerns these days since they have significant negative effects on our daily lives.  The problem is becoming more and more serious as the speed of the trains or the wings increases since the power of the noises increases significantly as M5 ∼M8 , where M is the Mach number based on the characteristic flow velocity and the exponent depends on the problem.  Thus there is increasing need for reduction of the aeroacoustic noises in a variety of industrial applications. In order to reduce the aeroacoustic noises we should under- stand the mechanism of their generation and propagation.

In this project we apply an immersed boundary method (corrected VP method) to DNS of the aeroacoustic sound. Our aim is to obtain acoustic waves and flow fields simultaneously with sufficient accuracy.  The results show that the present method is applicable to aeroacoustic problems in any complex geometry including practical engineering ones.  Application to three-dimensional problems is expected as the larger computational resources are becoming available.

• Ryu Komatsu, Wakana Iwakami and Yuji Hattori, “Direct numerical simulation of aeroacoustic sound by volume penalization method,” Comput. Fluids 130 (2016) 24-36.

Aeroacoustic Noise

Under construction

• Y. Hattori and Ryu Komatsu, “Mechanism of aeroacoustic sound generation and reduction in a flow past oscillating and fixed cylinders,” J. Fluid Mech. 832 (2017) 241-268.

Turbulence

Development of Turbulence Models by Machine Learning

Large-eddy simulation (LES) is used as an important tool of numerical simulation in a wide variety of fields where turbulent flows appear. In LES the turbulent flow fields are decomposed into resolved-scale or grid-scale (GS) flow field and small-scale or subgrid-scale (SGS) fluctuations by a filtering operation. In incompressible turbulent flows the effects of the fluctuations on the GS flow field appear as the residual or SGS stress tensor. How to model the SGS stress tensor using the GS flow field is the most important issue in LES. A new subgrid model which is much better than the existing ones should be developed.

The final goal of this project is to establish a new subgrid model for the SGS stress which performs better than the existing models.  As a first step toward this goal an artificial neural network (ANN) is tested as a tool for finding a new subgrid model of the subgrid-scale (SGS) stress in large-eddy simulation. An ANN is used to establish a functional relation between the grid-scale flow field and the SGS stress without any assumption of the form of function. Data required for training and test of the ANN are provided by direct numerical simulation of a turbulent channel flow. It is shown that an ANN can establish a model similar to the gradient model. The correlation coefficients between the real SGS stress and the output of the ANN are comparable to the existing models.

• Masataka Gamahara and Yuji Hattori, “Searching for turbulence models by artificial neural network,” Phys. Rev. Fluids 2 (2017) 054604.

Boundary layer is a thin layer of strong shear flow that develops around a body placed in fluid flow. When the laminar-to-turbulent transition occurs in a boundary layer, friction drag on the body surface increases drastically. For instance, about half of the total drag of an aircraft is due to this friction drag. Using theoretical and numerical approaches, we are seeking for efficient methods of reducing drag force by controlling boundary layer transitions.

The transitions are typically triggered when small disturbances grow up owing to some sort of instability. Therefore, it is important to avoid or suppress instability by elucidating the stability of boundary-layer flow. Roughness on the body surface is often the most influential disturbance source. In other words, there is room for altering the stability property by placing a specially-designed roughness. In our study, such a roughness shape is realized in DNS by applying an immersed boundary method (corrected VP method). We find that a specific roughness can suppress the crossflow instability which causes transition on the swept wings of aircrafts.

Following recent advancement of surface processing technology and micro flow-control devices, the practical use of such laminarization methods become more likely in the aircraft industry.

• Hirota, M., Hattori, Y., Ide, Y., Takami, H. & Yoshimoto, M. a patent-pending JP2019-069153

Vortex Dynamics

Curvature Instablity

Under construction

Magneto-HydroDynamics (MHD)

Plasma is an ionized gas that consists of ions and electrons. Their motions can be described by the fluid equations, respectively, and moreover interact with each other via electromagnetic fields. Since ion is more than a thousand times heavier than electron, a major research field in plasma physics has been magnetohydrodynamics (MHD) in which electron’s dynamics is mostly ignored. Following recent advance of computer performance, increasing attention is drawn to the study of “extended” MHD that deals with ions and electrons equally. We are developing a numerical solver of extended MHD and studying its fundamental feature.

The extended MHD equations include the electron-inertia and Hall effects that are neglected in MHD. Although these effects becomes dominant only in microscopic scales, they can also affect macroscopic dynamics and hence are indispensable. For example, our study shows that the explosive behavior of the solar flare, which is driven by the magnetic reconnection, can be explained by allowing for the electron inertia. We are also investigating self-organized flow and current that appear in consequence of such energy relaxation process.

For collisionless plasmas, the electron-inertia and Hall effects are more important than the dissipation effects (viscosity and resistivity). The extended MHD model is expected to be applied more and more to space plasmas and high-temperature fusion plasmas.

• Makoto Hirota, Yuji Hattori and Philip J Morrison, "Explosive magnetic reconnection caused by an X-shaped current-vortex layer in a collisionless plasma" Physics of Plasmas 22 (2015) 052114.
• Makoto Hirota, "An Explosive Scaling Law for Nonlinear Magnetic Reconnection and Its Insensitivity to Microscopic Scales" Plasma and Fusion Research 12 (2017) 1401010