Natsumi Iwata

Flow and transport in non-equilibrium high energy density plasmas

Grant No.：JPMJPR21O1

Researcher

Natsumi Iwata

Outline

When a matter is irradiated with intense light, a high energy density plasma is created. Using high-power lasers, we can achieve energy densities equivalent to the interior of a star. The laser-driven high energy density plasma can be a platform to explore non-equilibrium physics where microscopic particle acceleration and macroscopic hot plasma flow proceed simultaneously. This project aims to develop a statistical model of microscopic processes in high energy density plasmas and extend the model to fluid descriptions in order to understand flow and transport phenomena in the high energy density state of matter.

Yuki Uematsu

Creation of ionic fluid mechanics at micro- and nano-interface systems

Grant No.：JPMJPR21O2

Researcher

Yuki Uematsu

Outline

Based on recent social changes and advances in science and technology, I will conduct experiments of fluid dynamics in porous membrane systems, bubbles and water droplet systems. Furthermore, combining the experimental data with theoretical analyses, I will find the novel principles of ionic complex flows in micro- and nano-interfacial systems. In addition, I will elucidate unsolved problems in these systems using the effects of ions at interfaces, and prepare a knowledge base for further understanding and innovation of related promising analytical, industrial, and energy technologies.

Kazumichi Kobayashi

Molecular fluid dynamics of AWAI

Grant No.：JPMJPR21O3

Researcher

Kazumichi Kobayashi

Outline

This work is concerned with the development of the mean-field kinetic theory based on the Enskog-Vlasov equation. The mean-field kinetic theory is extended to multicomponent gas-liquid mixtures in this work. Further, a new field of study called “Molecular Fluid Dynamics of AWAI” is presented. Here, AWAI refers to a traditional Japanese word, and it denotes the interfacial region between the bulk liquid and gas phases in this work. From the extended mean-field kinetic theory, the fluid-dynamics-type equations are derived with the aim of establishing the dynamics of non-equilibrium gas-liquid flows; these results will further be extended to several fields.

Shintaro Sato

A posteriori fluid geometry

Grant No.：JPMJPR21O4

Researcher

Shintaro Sato

Outline

The dramatic development of experimental measurement techniques for flow field and computational fluid dynamics enables us to attain data of complex flow fields including turbulence; however, a complete description of the state space of a fluid dynamical system has remained a significant challenge. In this project, an approach to describe the state space of the complex fluid dynamics on a manifold is developed based on experimental and numerical simulation results in a posteriori manner using data-driven and mathematical insights. The universal nature of fluid dynamics is explored using the data-driven geometry of fluid dynamics. Also, a novel active flow control technique is developed in view of a posteriori fluid geometry.

Yoshiyuki Tagawa

Force-oriented fluid science

Grant No.：JPMJPR21O5

Researcher

Yoshiyuki Tagawa

Outline

Aiming to create a new fluid science based on the fluid stress field, we propose a new measurement system for the fluid stress field. This will provide a path to fundamental elucidation and control of complex flow and transport phenomena, and build a new multidisciplinary platform for fluid science and other fields. During this research period, we focus on the phenomenon related to an aneurysm. To overcome the non-linearity of stress measurement, we automate our photoelastic stress measurement system and use machine learning which is trained by theoretical and numerical calculations.

Yuji Tasaka

Establishment of new fluid science based on rheological evaluation of multiphase/complex fluids

Grant No.：JPMJPR21O6

Researcher

Yuji Tasaka

Outline

This project aims to establish the base technology of rheological evaluation on multiphase and complex fluids for understanding and controlling the behavior of the fluids. Advancement and application of ultrasonic spinning rheometry (USR), which can evaluate transient rheological properties of the complex fluids, achieve extension of flow instability problems to multiphase fluids. The application studies of USR on problems of fluid mechanics, material science, chemical engineering, and medicine provide a paradigm shift from “evaluation” to “prediction and control” of fluid flows in rheology science.

Motoki Nakata

Activation of self-organized structure formations and transport phenomena in turbulence explored by integrated approach of turbulence physics, mathematics, and informatics

Grant No.：JPMJPR21O7

Researcher

Motoki Nakata

Outline

Based on an integrated approach of turbulence physics, mathematics, and informatics, theoretical and computational methods to activate self-organized flows, vortex structures, and associated energy and mass transport phenomena in turbulence immersed in external fields with geometric degrees of freedom are explored in this project. Through the study of nonlinear interactions among vortices, flows, and magnetic fields in plasma turbulence, new interdisciplinary aspects of fluid science are developed.

Daiki Nishiguchi

Exploration of 3D structure and control methodology of active turbulence

Grant No.：JPMJPR21O8

Researcher

Daiki Nishiguchi

Outline

The research aims to directly observe the spatio-temporal structure of three-dimensional active turbulence using high-speed confocal microscopy. We will experimentally elucidate the statistical properties of active turbulence and develop a statistical mechanical methodology to comprehensively understand the macroscopic boundary conditions of active fluids from the microscopic level. To unravel the behavior of active turbulence in the presence of the mean background flow, simultaneous observation of the viscous response and collective motion of active turbulence in a microchannel will be performed. This will contribute to designing novel devices that utilize protocol-dependent negative viscosity.

Masanobu Horie

Fluid modeling using a combination of numerical analysis and machine learning

Grant No.：JPMJPR21O9

Researcher

Masanobu Horie

Outline

Machine learning is a promising technology to accelerate physical simulations on complex flow and transport phenomena. However, for reliable prediction using machine learning models, it is inevitable to fulfill the requirements coming from physics, such as equivariance under the coordinate transformation and fulfillment of the conservation law widely observed in the flow and transport phenomena. In this research, I develop a physical simulation method based on machine learning that satisfies the equivariance and conservation law. To that end, I use graph neural networks capable of dealing with general meshes and the discontinuous Galerkin method that guarantees conservation. I expect the method developed here to be a key to understanding, predicting, and controlling complex flow and transport phenomena.

Daiki Matsunaga

Analysis of intracellular fluid-structure interactions by experiment/physics

Grant No.：JPMJPR21OA

Researcher

Daiki Matsunaga

Outline

The cellular contractile force is driven by motor proteins such as myosin and kinesin, and the intracellular fluid dynamics can be considered as a physics of fluid-structure interactions triggered by the activity of the motor proteins. In order to clarify a part of the mechanical phenomenon that spans a broad spatiotemporal hierarchy by combining experiments and mathematical techniques: this project aims to reveal how the local molecular activities relate to the contractile forces and eventually create the cellular scale fluid dynamics.

Yuki Minamoto

Clean combustion technology with physico-data-driven approaches

Grant No.：JPMJPR21OB

Researcher

Yuki Minamoto

Outline

This project focuses on the development of physical models, required for practical computational fluid dynamics (CFD) to be employed during the development and optimization of next-generation clean combustion devices utilizing non-carbon fuels such as hydrogen. Data-driven approaches are extensively applied during the model development based on the foundation of physical understandings of the phenomena. The architectures of these models are designed by considering both usability and robustness, as well as accuracy. For additional usability of data-driven physical models, a machine-learning (ML) platform is developed where CFD engineers can perform various ML development processes, regardless of their ML expertise. The outcome of this research will contribute to the establishment of a society where usage of data-driven models prevails among R&D phases.