Seminar Schedule
[MAT Seminar]
 Date:
 2019/12/17 (Tuesday) 13:00  15:00
 Place:
 MAT theater at 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Yuzuru Yamamoto (MAT)
[MAT Seminar]
 Date:
 2019/12/04 (Wednesday) 13:00  15:00
 Place:
 MAT theater at 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Furuichi Mikito
 Title:
 Simulating yield response with the Discrete Element Method (DEM) in the numerical slump flow experiment
 Abstract:

We are developing the parallelized Discrete Element Method (DEM) code that is originally developed for the research of the granular material. Lagrangian nature of DEM also offers a great advantage for simulating the crustal deformation, landslide, and deformation of concrete and mud, over the meshbased method. Thus, recently, the demand for such DEM applications is increasing. However, to perform such simulations, the basic interaction forces for granular material are not sufficient and extra models are needed for representing the continuum constitutive law. Here, we introduce the yield forces to represent Bingham like plastic deformation into the DEM. The quality of the model is validated by numerical slump test following the proposed theory of yield stress and yield force conversion. We successfully demonstrated the scaling behaviors of the yield force model against the particle sizes (up to about 20M particles), which is not presented in the earlier work. We also present the new orthogonal parallel domain decomposition technique to avoid the highly distorted aspect ratio problem which frequently found in the practical engineering simulation settings.
[MAT Seminar]
 Date:
 2019/11/27 (Wednesday) 13:00  15:00
 Place:
 MAT theater at 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Seyed Reza Amini Niaki
 Title:
 Effects of inter particle friction on the mesoscale hydrodynamics of dense gassolid fluidized flows
 Abstract:

Gassolid fluidized bed reactors are widely applied in chemical and energy industries, and their design and scaleup are virtually empirical, extremely expensive and time consuming. This scenario has motivated the development of alternative theoretical tools, and twofluid modeling, where gas and particulate are both treated as interpenetrating continuum phases, has appeared as a most promising approach. Owing to the large domains to be resolved in realscale fluidized bed reactors, only filtered modeling approaches are feasible, and closure models become necessary to recover subgrid effects that are filtered by the very coarse numerical grids that are imposed owing to computational limitations. Those closure models, which in hydrodynamic formulations account for filtered interphase momentum exchanges and filtered and residual stresses in the phases, can be derived from results of highly resolved simulations (HRS) performed over small size domains under refined numerical grids. One widely practiced approach consists of applying twofluid modeling under microscale defined closures, generally known as microscopic twofluid modeling. This approach includes microscopic closures for solid phase stresses derived from the kinetic theory of granular flows (KTGF), which accounts for kineticcollisional effects only, and is adequate to dilute flows. Otherwise, the conventional KTGF does not account for interparticle friction effects, and its application to dense flow conditions is quite questionable. In this work a literature available modified version of KTGF is applied which also accounts for interparticle friction, and highly resolved simulations are performed for dense flow conditions in order to evaluate the effects of friction over relevant filtered parameters (namely effective drag coefficient, filtered and residual stresses). Ranges of domain average solid volume fractions and gas Reynolds numbers are considered (macroscale conditions) embracing dense gassolid fluidized flows from suspensions up to pneumatic transport. The MFIX open source code is used in all the simulations, which are performed over 2D periodical domains for a unique monodisperse particulate. The HRS results (i.e. mesoscale flow fields) are filtered over regions compatible with grid sizes in large scale simulations, and the relevant filtered parameters of concern are derived and classified by ranges of other filtered parameters taken as independent variables (filtered solid volume fraction, filtered slip velocity, and filtered kinetic energy of solid velocity fluctuations, which are referred to as markers). Results show that the relevant filtered parameters of concern are well correlated to all of those filtered markers, and also to all of the imposed macroscale conditions. Otherwise, interparticle friction showed no significant effects over any filtered parameter. It is recognized that this issue clearly requires further investigation notably regarding the suitability of the markers that were assumed for classifying the filtered results. The current work is intended as a contribution for future developments of more accurate closure models for large scale simulations of gassolid fluidized flows.
[MAT Seminar]
 Date:
 2019/11/20 (Wednesday) 13:00  15:00
 Place:
 MAT theater at 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Jian Chen
 Title:
 Discrete element simulation of a (drillingand) mixing process
 Abstract:

Soils beneath seafloors may find to be important resources of minerals. A better understanding of the drilling and mixing process is one of the first steps to achieve efficient retrieval of mineral resources from seafloors. The designs of drill bit are mainly relied on empirical knowhows and the model experiments are costly and time consuming. Aiming to shed light upon the design of drill bit in the long run, in this study we explore the capability of a DEM code for simulating the drilling and mixing processes. The simulations are compared with a couple of experiments on color sands. Consistent results, in terms the torques on the drill bit, the increases of surface height and the surface mixing patterns, have been obtained. As future works, further validations will be carried out for the DEM code and the current force models will be extended for clayey soils.
[MAT Seminar]
 Date:
 2019/11/13 (Wednesday) 13:00  15:00
 Place:
 MAT theater at 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Arthur Bauville (MAT)
 Title:
 A comprehensive study of pressuretodepth conversion models for (ultra)highpressure metamorphic rocks
 Abstract:

Pressure estimated from metamorphic rocks is the main tool of geodynamic reconstructions. The pressuretemperature path of UHP metamorphic rocks typically shows a linear increase of P and T followed by a rapid drop of Pressure at nearconstant temperature. The geological history can be reconstructed by using the metamorphic pressure as a proxy for depth. Researchers often base their geodynamic reconstruction on a simple linear mapping of pressure to depth, by considering that the pressure is the weight of the overlying column of rock or lithostatic pressure. In recent years, an increasing corpus of evidence demonstrates that rocks can experience pressures that deviate from the lithostatic state on the order of GPa. These deviations can be at the scale of the orogen (Petrelli and Podladchikov, 2002), the outcrop (Jamtveit et al., 2018; Luisier et al., 2019); or even at the grainscale (Tajcmanova, 2015). Thus, these studies raise the concern that metamorphic pressures may not be reliable proxies for depth, and therefore could not be used for geodynamic reconstructions. The objective of this contribution (1) to review the various models proposed in the literature for metamorphic pressure, (2) to formulate analytical models with simple assumptions that can be used to convert metamorphic pressure to depth even in the case where pressure deviates significantly from the lithostatic pressure. We use our pressuretodepth conversion models to estimate the depth of ~50 samples from various orogens worldwide. The prediction of the different models varies widely. Some models predict depth as deep as 160km for specific samples, while other models predict depth <75 km (i.e. deepest depth of the Moho) for all samples. We discuss the limits of applicability and the geodynamic implications of each model. We review available petrological and geophysical data and geodynamic models for the Alps to determine the bestsuited models for the sample recovered from this orogen.
[MAT Seminar]
 Date:
 2019/10/30 (Wednesday) 13:00  15:00
 Place:
 MAT theater at 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Takane Hori (FEAT)
 Title:
 Trial to revise governing equations of earthquake generation cycle simulations, II
 Abstract:

In my previous MAT seminar presentation, I introduced the problem that the governing equations of earthquake generation cycle simulation have too nonlinear and unstable characteristics for numerical calculations and too much parameters to be defined although we do not have enough data to constrain them. These become intrinsic problems when we use the equations to monitor and forecast earthquake generation cycle processes. I have still been struggling how to revise the governing equations, especially the fault constitutive law. I will introduce the remaining problems, approach to overcome the problems, and what I have found during the struggling, such as an idea of a unified model for regular and slow earthquakes.
[MAT Seminar]
 Date:
 2019/10/23 (Wednesday) 13:00  15:00
 Place:
 MAT theater at 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Shigenobu Hirose (VAiG)
 Title:
 Brief overview of line broadening calculations
 Abstract:

Hydrodynamics of multispecies gas mixture with radiative heat transport is relevant in many research areas including atmospheric science, engineering and astrophysics. To perform simulations of such hydrodynamics accurately, a reliable database of gas opacity is required, which we are developing now. In this talk, I'll give a brief overview of line broadening calculation, which is necessary to evaluate gas opacity. Generally, spectral lines are not infinitely sharp but broadened due to many kinds of physical processes, called "line broadening". Line broadening is important because it not only affects the efficiency of radiative transport but also contains information about the gas properties including temperature and density.
[MAT Seminar]
 Date:
 2019/10/16 (Wednesday) 13:00  15:00
 Place:
 MAT theater at 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Makiko Ohtani (ERI)
 Title:
 Synchronization of megathrust earthquakes to periodic slow slip events in a singledegreeoffreedom springslider model
 Abstract:

Recently recognized slow slip events (SSEs) recurring in the deeper extensions of seismogenic zones along plate boundaries are drawing attention to their potential for triggering megathrust earthquakes that rupture the entire seismogenic zone. We describe how earthquakes simulated in a singledegreeoffreedom model are synchronized to the rhythm of imposed periodic SSEs. The time lag t_Q from the one most recent SSE to a seismic event varies with system parameters and may take a broad range of eligible values between 0 and T_SSE (SSE recurrence period). Earthquakes were found to synchronize with SSEs in various patterns depending on the proportion of SSEdriven loading within an SSE cycle, the recurrence period of the SSEs, and the duration of the SSEs, although synchronization itself remained a prevalent feature. Asynchrony was found only for long SSE durations.
[MAT Seminar]
 Date:
 2019/10/09 (Wednesday) 13:00  15:00
 Place:
 MAT theater at 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Leonardo Ventura (Queen Mary University of London)
 Title:
 Folds “a priori”
 Abstract:

Folds have been recently employed as a cheap and straightforward solution to complex problems, like manufacturing microscale 3D components or enhancing the energy storage capabilities of polymerbased materials (Meng, 2019).
In particular, folds creation, postfolding deformation and fracture prevention are of critical importance for the success of the market of foldable electronics.
Folds in plates are jump discontinuities in the rotations, and the mechanical behaviour of folds depends on two parameters: rest opening angle and fold rotational stiffness (Barbieri, 2019). These two physical properties change with the material, plate thickness and folding process.
Therefore, in principle, one should measure the rest opening angle and the rotational stiffness for each fold.
To reduce the amount of experimental work needed to model folds, we developed an analytical model able to predict these two values knowing only the elastoplastic properties of the material. We validated our rest opening angle predictions against experimental results. Knowing a priori the rest opening angle and the fold stiffness values will help to investigate complex interactions between folds and cracks.
REFERENCES:
2019 E Barbieri, L Ventura, D Grignoli, E Bilotti. A meshless method for the nonlinear von Kármán plate with multiple folds of complex shape. Computational Mechanics, 119
2019 N Meng, X Ren, G Santagiuliana, L Ventura, H Zhang, J Wu, H Yan, M Reece, E Bilotti  Ultrahigh βphase content poly(vinylidene fluoride) with relaxorlike ferroelectricity for high energy density capacitors  Nature Communications, 2019  accepted
[MAT Seminar]
 Date:
 2019/10/02 (Wednesday) 13:00  15:00
 Place:
 MAT theater at 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Takashi Minoshima (MAT)
 Title:
 Improving numerical method for shock waves in magnetized gas
 Abstract:

Shock wave, generated when a supersonic flow collides with an obstacle, is a fundamental physics in both science (e.g., astrophysics) and engineering (e.g., aerodynamics), and a computer simulation plays an important role to understand its dynamics. Capturing the shock in computer simulations is a challenging issue, and many studies have been devoted to develop a robust and accurate numerical method. This kind of scheme may be classified into two types, FVS (flux vector splitting) and FDS (flux difference splitting) schemes. While FDSbased schemes are widely adopted in computational astrophysics (e.g., Kritsuk et al. 2011), FVSbased schemes seem to be also utilized in aerospace computational engineering field in order to overcome numerical difficulties in FDSbased schemes (Shima 2015). Although advantages of recent FVSbased schemes are useful for computational astrophysics as well, their application is quite limited thus far. Therefore, we try to propose a new FVSbased scheme that can be an alternative to FDSbased schemes in computational astrophysics. This talk will present the detail and current status of the scheme.
[MAT Seminar]
 Date:
 2019/09/25 (Wednesday) 13:00  15:00
 Place:
 MAT theater at 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Hisazumi Akai (Institute for Solid State Physics, The University of Tokyo)
 Title:
 Halfmetallic antiferromagnets and their possible device applications
 Abstract:

Halfmetallic antiferromagnts are the materials that exhibit halfmetallicity and antifirromagnetism (compensated ferrimagnetism) simultaneously. Such materials are especially useful for spintronics devices since they have 100% spinpolarized Fermi surfaces and the direction of polarization is insensitive to external field. Also, in many cases they have a rather high magnetic transition temperature. We found that (XY)Z_{2}, where X and Y are transition metal elements and Z is a chalcogens or a pnictigen, show halfmetallic antiferromagnetism when the sum of effective d electron numbers of X and Y is 10. Examples are (CrFe)S_{2} and (CrFe)Se_{2}. We report a systematic investigation of the electronic structure of these materials calculated by the KorringaKohnRostoker (KKR)Green’s function. We have found that they are chemically stable and their calculated magnetic transition temperatures are sufficiently higher than the room temperature. The transport properties of the systems are also calculated by the KKR method combined with the KuboGreenwood formula. The results show that those systems can function as spintronics materials if used as components of giant magnetoresistance (GMR) or tunneling magnetoresistance (TMR) sensors, magnetic randomaccess memory. We propose a possible new structure for GMR sensor that is far simpler and more compact than ever existing ones.
[MAT Seminar]
 Date:
 2019/09/18 (Wednesday) 13:00  15:00
 Place:
 MAT theater at 5th floor, IT Building, Yokohama Institute
 Language:
 Japanese
 Speaker:
 Yasuko Yamagishi (MAT)
[MAT Seminar]
 Date:
 2019/09/11 (Wednesday) 13:00  15:00
 Place:
 MAT theater at 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Manabu Morishige (MAT)
 Title:
 Bayesian inference of model parameters governing subduction zone dynamics
 Abstract:

Temperature is one of the most fundamental information to understand the dynamics of Earth’s interior, and it is controlled by many model parameters. In this presentation we will demonstrate how we can constrain these parameters from the given observations based on Bayesian inference. This approach has several advantages, including (1) we can vary a large number of model parameters at a time, and (2) we can obtain model parameters which explain the observation in a probabilistic manner. As an example, we focus on the surface heat flow observed in Tohoku, Japan. We predict surface heat flow by using 2D steadystate thermalmechanical modeling. We investigate the following three model parameters governing subduction zone dynamics: slabmantle decoupling depth, activation energy for viscosity, and radiogenic heat production rate in the upper continental crust. We find that the obtained values for activation energy and radiogenic heat production rate are different from those generally accepted. This may suggest that the thermal structure in this region is affected by a recent opening of Japan Sea and possibly by the presence of smallscale convection, and does not reach steady state. The approach presented here provides us with a new framework for understanding the dynamics of Earth’s interior.
[MAT Seminar]
 Date:
 2019/09/04 (Wednesday) 13:00  15:00
 Place:
 MAT theater at 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Yoji Kawamura (MAT)
 Title:
 Metachronal waves in nonlocally coupled oscillators
 Abstract:

Ciliates such as Paramecium are covered with an array of cilia. The cilia array exhibits a wavy movement called metachronal waves, which can be considered as stable plane waves in coupled oscillators. In this talk, we consider a system of nonlocally coupled phase oscillators and perform a linear stability analysis of plane wave solutions of the system. Consequently, we demonstrate the stability of the solution associated with a particular wave vector and also the robustness of the stability against the heterogeneity of natural frequencies. The theoretical results are confirmed by direct numerical simulations. Finally, we discuss the theoretical derivation and experimental realization of the system.
[MAT Seminar]
 Date:
 2019/08/28 (Wednesday) 13:00  15:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Ryoichiro Agata (FEAT)
 Title:
 Introduction of covariance components in slip inversion of land and seafloor geodetic data and application to slip deficit rate estimation in the Nankai Trough subduction zone
 Abstract:

When spatial distribution of observation stations has bias in geodetic slip inversion, modeling errors in the inversion scheme may result in significant unnatural shortwave components in estimated slip distribution, which overfit to data. Combined use of both land and seafloor geodetic data in slip inversion often leads to such situations. To avoid overfitting, I proposed a method to incorporate proper covariance components in the data covariance matrix originated by modeling errors in the geodetic slip inversion. Because linearity of the inversion problem is retained, widelyknown approaches to introduce prior constraints to the inversion, which assumes linear inversion, are easily applied to the proposed method. Synthetic tests showed that the introduction of covariance components allowed for the estimation of slip distributions closer to the true one, avoiding overfitting to geodetic data in biasedly distributed observation stations, compared to a conventional approach that does not introduce covariance components. I applied the proposed method to the estimation of the slip deficit rate in the Nankai Trough subduction zone, using geodetic data of displacement rates provided by land GNSS stations and seafloor GNSSAcoustic stations. The proposed method estimated a reasonably smooth distribution of slip deficit rate compared to the conventional approach. Spatial distribution of residuals in the displacement rates suggested that the proposed method successfully avoids overfitting to the block motion at the south of the Median Tectonic Line, which the physical model I assumed here cannot describe.
[MAT Seminar]
 Date:
 2019/08/21 (Wednesday) 13:00  15:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Muneo Hori (MAT)
 Title:
 Most suitable coordinate system for numerical analysis of wave equation
 Abstract:

Development of better numerical analysis of wave equation is a fundamental issue of continuum mechanics. Most of numerical analysis methods employ Cartesian coordinate system, mainly because physics laws from which the wave equation is derived do not depend on coordinate system and Cartesian coordinate system with vertical axis and horizontal axes is a natural choice. There is a possibility of the presence of the coordinate system that is most suitable for numerical analysis. Starting particle discretization scheme, I explain recent trail of using a particular coordinate system to solve wave equation. I make an excuse that using a particular coordinate system is practically useful in identifying errors in input parameters of a numerical analysis model.
[MAT Seminar]
 Date:
 2019/08/07 (Wednesday) 13:00  15:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Dmitry Kolomenskiy (CEIST)
 Title:
 A waveletbased data compression method for threedimensional numerical
simulations on regular grids
 Abstract:

Highperformance computing can produce large volumes of output data. Even though data reduction may be performed during the course of computation in order to store only the quantities of interest such as statistical moments, it is often necessary to store full threedimensional fields for purposes including simulation restart, timeresolved visualization and exploratory analyses. In this talk, I will present a waveletbased method for compression of numerical simulation data on regular structured grids. It is inspired by image compression, and it consists of discrete wavelet transform, quantization adapted for floatingpoint data, and entropy coding. I will discuss different aspects of these numerical techniques, present an opensource software WaveRange, and show example numerical tests, ranging from idealized configurations to realistic global weather simulation data.
[MAT Seminar]
 Date:
 2019/07/31 (Wednesday) 13:00  15:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Gen Li (MAT)
 Title:
 New understanding on the optimising strategies in fish swimming from CFD and engineering perspectives
 Abstract:

(Attention Please: this talk substantially overlaps my presentation in the JAMSTEC international day 2019/07/16 )
In this talk, I would like to introduce two recent CFDrobotic hybrid projects: stingrayinspired roundplate robot and magnetic field driven flexible swimmer, as well as an development of fluidstructureinteraction (FSI) solver assorted with the magnetic swimmer.
Also, I would like to introduce my discovery of the optimising strategies in fish swimming. Drag is a key factor to explain behaviours and predict strategies in fish, but extremely difficult to be measured in experiment. By utilizing computational approach, interesting characteristics of drag in a swimming fish were found: the (scaling) trend of drag follows that of a threedimensional object, while the magnitude of drag depends on the kinematics of the undulatory wave, much exceeding that of a rigid fish gliding at the same speed. The duality in swimming fish determines its swimming strategy. In continuous swimming at a specific speed, fish need to wisely choose its frequency and amplitude to prevent excessive energy consumption due to drag. Meanwhile, since gliding drag is much less than undulatory drag at the same speed, properly switching between undulatory propulsion and gliding (burstandcoast) may become a feasible strategy for fish to save energy.
[MAT Seminar]
 Date:
 2019/07/24(Wednesday) 13:0015:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Ettore Barbieri (MAT)
 Title:
 A Numerical Method for the Mechanics of Folds in Plates
 Abstract:

Folds occur in a variety of physical phenomena and engineering applications: from chevron folds and mountain ridges in geology to the ribbed shells of marine crustaceans to deployable antennas for satellites. Even in the arts, famous examples of folding are origami sculptures. Interestingly, the folding problem shares a mathematical trait with the fracture problem. Mathematically speaking, fractures are discontinuities on the displacements. Folds, instead, are discontinuities in the displacements gradients. This presentation will show how to treat such discontinuities in numerical simulations of folds for thick and thin plates, drawing from the author's experience in computational fracture mechanics. Furthermore, this talk will present newly obtained analytical solutions for the nonlinear folding problem in 1D, necessary for the verification of the
method.
REFERENCES:
Curvature tuning in folded strips through hyperstatic applied rotations, E Barbieri (2019), Frontiers in Materials 6, 41
A meshless method for the nonlinear von Kármán plate with multiple folds of complex shape, E Barbieri, L Ventura, D Grignoli, E Bilotti, (2019) Computational Mechanics, 119
[MAT Seminar]
 Date:
 2019/07/17(Wednesday) 13:0015:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Richard Spitz (University of Lausanne)
 Title:
 3D numerical modelling of nappe emplacement during orogenic wedge formation
 Abstract:

Foldandthrust belts are common to orogens like the Canadian Rocky Mountains or the Swiss Alps. One prominent kmscale structural feature of such foldandthrust belts are socalled foldandthrust nappes. The emplacement of nappes is an ongoing research topic since the late 19th century in geology. Therefore there is a manifold of studies utilizing different analysis tools like geological crosssections, geometrical, theoretical, analog, and numerical models. These models often rely on individual twodimensional (2D) cross sections. Series of cross sections separated by 10100 km often reveal significant structural variations. These variations likely indicate lateral changes in the geometry, material properties, and physical conditions before deformation (i.e., initial conditions). In this study, we investigate the control of initial lateral geometrical variation on the evolution of geological structures. We perform 2D and 3D numerical simulations of the emplacement of tectonic nappes. Results show that crosssections from the 3D simulations differ significantly from 2D simulations with the same initial crosssectional geometry. We systematically test different geometrical transition between two crosssections, and we compare the results with natural examples from the Helvetic nappe system in the Western Swiss Alps.
[MAT Seminar]
 Date:
 2019/07/10(Wednesday) 13:0015:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Ryo Onishi (CEIST)
 Title:
 SuperResolution Simulation for RealTime Prediction of Urban Micrometeorology
 Abstract:

We propose a superresolution (SR) simulation system that consists of a physicsbased meteorological simulation and an SR method based on a deep convolutional neural network (CNN). The CNN is trained using pairs of highresolution (HR) and lowresolution (LR) images created from meteorological simulation results for different resolutions so that it can map LR simulation images to HR ones. The proposed SR simulation system, which performs LR simulations, can provide HR prediction results in much shorter operating cycles than those required for corresponding HR simulation prediction system. We apply the SR simulation system to urban micrometeorology, which is strongly affected by buildings and human activity. Urban micrometeorology simulations that need to resolve urban buildings are computationally costly and thus cannot be used for operational realtime predictions even when run on supercomputers. We performed HR micrometeorology simulations on a supercomputer to obtain datasets for training the CNN in the SR method. It is shown that the proposed SR method can be used with a spatial scaling factor of 4 and that it outperforms conventional interpolation methods by a large margin. It is also shown that the proposed SR simulation system has the potential to be used for operational urban micrometeorology predictions.
[MAT Seminar]
 Date:
 2019/07/03(Wednesday) 13:0015:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Ippei Oshima (MAT)
 Title:
 Mechanism of Liquid Sheet Atomization Induced by Air Flow
 Abstract:

I have joined MAT (Center for Mathematical Science and Advanced Technology) in June, 2019. So, in this seminar, I’ll introduce my previous activities of research and development and future work in JAMSTEC.
My research interest includes fluid mechanics, multiphase flow and atomization. I have researched airblasted atomization of a liquid sheet induced by air flow for gas turbine engines. To clarify the mechanism of the liquid sheet atomization by air flow and develop the atomization model based on the mechanism, I have performed visualization and optical experiments and numerical simulation and theoretical analysis. These experiences and techniques will be useful for JAMSTEC’s work.
In the research topic, I’ll introduce overview of atomization and the background of my research. I also explain how droplets are produced through the atomization process in gas turbines. Based on that, in this presentation, I’ll briefly explain following two small topics;
1) Is the liquid sheet deformation injected from the injector governed by pure KelvinHelmholtz instability?
2) What kind of correlation is proposed on the longitudinal wavelength of the liquid sheet deformation?
In the development topic, I’ll introduce one of missions from my current hobby work of developing a small satellite.
Finally, work plan in JAMSTEC which is concerned with geothermal power plant will be introduced.
[MAT Seminar]
 Date:
 2019/06/26(Wednesday) 13:0015:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Osamu Kuwano (MAT)
 Title:
 Evolution of size and shape distributions of comminuted particles under shear deformation
 Abstract:

A faults zone contains fine rock powders called gouge that have been ground up by past fault motions.
Particle size distribution and particle shape of gouge particles affect the frictional properties of the fault and reflect the comminution process by the past fault motions. It is well known that particle size distribution of fault gouge show powerlaw distributions. Exponent of this power law is considered to reflect the style and degree of deformation. We did several shear experiments using a rotary shear apparatus with the shear displacement ranges from 10mm to 1m. In this seminar, we will discuss about the relationship between the particle size distribution and the degree of comminution of model particles by particle image analysis.
[MAT Seminar]
 Date:
 2019/06/19(Wednesday) 13:0015:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Natsuki Hosono (MAT)
 Title:
 SPH Simulations on AIspecific processors
 Abstract:

The Smoothed Particle Hydrodynamics method is a widely accepted particlebased numerical simulation method.
Since it is a particlebased method, a straightforward implementation of SPH has O(N^2) computational cost.
Due to this expensive computational cost, in order to carry out large scale numerical simulations, we need to apply some technique to accelerate calculations.
One of the standard techniques is to employ manycore devices.
Recently, AIspecific processors, such as NVIDIA's V100 and Google's TPU, are a hot topic.
They often are manycore devices with low precision arithmetics.
AIspecific processors can be attractive candidates to carry out SPH simulations.
However, to obtain high efficiency of AIspecific processors, we need to use low precision.
In other words, there is a tradeoff between the precision and calculation speed.
However, it is not well known that low precision can be used for SPH simulations.
Hence, as the first step to effectively use AIspecific processors on SPH, we aimed at investigating the dependence of simulation accuracy on the precision.
We chose the Couette flow as a benchmark test and quantified its simulation accuracy by L2error.
In this seminar, we will report the ongoing status of the above work.
[MAT Seminar]
 Date:
 2019/06/12(Wednesday) 13:0015:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Shun Nomura(MAT)
 Title:
 Sedimentation process of experimentally produced turbidity current
 Abstract:

Turbidity current is a particleladen current driven by density difference due to the presence of suspended sediment particles. Since it moves down along submarine channels with large amount of sediment for a long distance, it is one of the most important factors to understand the submarine systems.
To clarify its sediment transport systems, the sedimentation process of an experimentally produced turbidity current in an inclined flume was investigated by the electrical resistance based depositmeter.
After introducing the measurement technique of the depositmeter, the experimental results performed in three different initial fluid densities are introduced. Then the semi empirical model for spatiotemporal sedimentation process is explained and its applicability is discussed.
[MAT Seminar]
 Date:
 2019/06/05(Wednesday) 13:0015:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Masaru Nakano (FEAT)
 Title:
 Discrimination of seismic signals from earthquakes and tectonic tremor by applying a convolutional neural network to running spectral images (SRSpecCNN)
 Abstract:

The recent development of dense highsensitivity geophysical observation networks has led to many observations of seismic and geodetic signals known as the slowslip earthquakes. Monitoring the slow deformation at plate boundaries as the Nankai trough would provide insights into deformation processes in the source regions of megathrust earthquakes. Tectonic tremor is one of signals radiated from such slow deformations and recorded by seismometers. Their signals occur dominantly in the 2–8 Hz frequency band and can last for tens of seconds to several minutes, in contrast to typical earthquakes that produce seismic signals at frequencies up to several tens of hertz and last less than a minute. Because tremor is caused by stochastic processes, the resultant waveforms are represented by a stochastic function and construction of deterministic measures to discriminate tremor signals from earthquakes is very difficult. In this study, we used a convolutional neural network (CNN) to discriminate the signals of tectonic tremor from those of local earthquakes in running spectral images of these signals. Ordinary methods of image recognitions are designed insensitive to the location where the objects appear in the target image. In the discriminations of earthquake and tremor signal, on the other hand, the absolute frequency of signal appearance is important which reflects the physical properties of the signal source. Therefore, we developed a method (SRSpecCNN) that is sensitive to the absolute frequency of signal appearance, but is insensitive to the time of signal onset. Based on simultaneous analyses of the frequency contents and durations of the signals, we achieved 99.5% accuracy for our identifications of signals from tectonic tremor, local earthquakes, and noise. Because running spectra clearly differentiate the characteristic features of these signals, we were able to achieve this high accuracy by using a CNN of simple architecture.
[MAT Seminar]
 Date:
 2019/05/22(Wednesday) 13:0015:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Daisuke Nishiura (MAT)
 Title:
 SPH simulation of wave channel Wave impact edition featuring Davide
 Abstract:

Dr. Davide Wüthrich has experimentally studied about tsunamilike wave hydrodynamics by using a wave channel at Laboratory of Hydraulic Constructions (LCH) of Ecole Polytechnique Fédérale de Lausanne (EPFL) in Switzerland. On the other hand, the MAT has developed the large scale SPH simulation for the tsunami study. However, we need to validate the accuracy and reproducibility of our SPH simulation to apply the numerical model for actual engineering problems. So, there are many good benefits for us to collaborate with each other. Previously, fundamental wave characteristics such as a wave height and velocity profiles without buildings were validated by comparing with many experimental data. This time, we investigated the wave impact forces acting on the buildings with and without openings. The reproducibility of our SPH simulation was validated in the case of buildings with different geometries (opening size and ratio) and then some issues for the computational accuracy of our SPH simulation were pointed out.
[MAT Seminar]
 Date:
 2019/05/15(Wednesday) 13:0015:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Takehiro Miyagoshi (MAT)
 Title:
 Experiments of horizontal convection
 Abstract:

We have studied horizontal thermal convection by experiments with fluids. Horizontal convection is interesting in terms of dynamics in the interior of the planet. At the coremantle boundary, stagnant slabs produce temperature difference in horizontal direction. We want to know whether those temperature differences can drive vigor core convection and dynamo. In massive superEarths, it is known that the tidal lock often occurs. Then, the temperature difference at the surface between the noon side and night side becomes very large. The temperature difference may drive the horizontal convection in the mantle of superEarths. The experimental vessel is rectangular box with the aspect ratio of 4:1:1. We use water with glycerin or pure water for working fluid. We have started this experimental study from December in the last year, and we would like to introduce our initial results in the MAT seminar.
[MAT Seminar]
 Date:
 2019/05/08(Wednesday) 13:0015:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Mikito Furuichi (MAT)
 Title:
 Progress of numerical modeling of the riser pipe flow for slurry transportation
 Abstract:

For the efficient design of the mining system for deepsea resources, prediction of the performance of pipe flow is important. Then, we develop a numerical model for evaluating the relationship between the flux by the slurrypumps, property of mud (size, fraction and nonNewtonian), injection of mud at the bottom and so on. By using the model, we discuss the risk of stuck the pipe and efficiency of the energy cost of mining system. The obtained data should be also useful for the optimization of the operation.
[MAT Seminar]
 Date:
 2019/04/24(Wednesday) 10:0011:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Takatoshi Yanagisawa (VERC)
 Title:
 Structural change from roll to largescale cell in turbulent thermal convection of a liquid metal layer
 Abstract:

To understand the expected nature of turbulent flow in liquid outer core of the Earth, we performed laboratory experiments on RayleighBenard convection with a liquid metal in a square box geometry having a moderate aspect ratio. Horizontal velocity profiles of flow were measured at several lines by using ultrasonic velocity profiling. By combining the information from profiles, we can reconstruct organized flow structures.
Systematic variation of the structure was detected with increasing the Rayleigh number; a quasitwodimensional roll changes to a cell having a relatively larger horizontal scale. We found that the flow structure, whether it is roll or cell, show quasiperiodic oscillation whose representative period is approximately same as the circulation time of the flow.
We also performed numerical simulations of convection with the same geometry as the experiments by setting a small Prandtl number like a liquid metal. Quantitative comparison on the velocity profiles between experiments and simulations provided quite satisfactory agreement. By integrating these results, we propose a scaling low on the Rayleigh number dependence of horizontal size of flow structure.
[MAT Seminar]
 Date:
 2019/04/17(Wednesday) 13:0014:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Yuzuru Yamamoto (MAT)
[MAT Seminar]
 Date:
 2019/04/01(Monday) 13:0015:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Stephan Rudykh (University of Wisconsin Madison)
 Title:
 Instabilityinduced Microstructure Transformations for Switchable Soft Metamaterials
 Abstract:

Nature actively uses sophisticated designs of microstructures to
achieve astonishing material properties and functionalities. Thus,
microstructures give rise to the incredible toughness of
motherofpearl. Another example is an octopus, an amazingly effective
soft machine created by the nature. The creature can squeeze its whole
body through an extremely narrow space while preserving a large
variety of functionalities. The nature created soft machine comprises
highly deformable composites that are characterized by different
dynamically tunable microstructures and phase properties, depending on
the required functionalities. Indeed, such materials are highly
desirable for humaninteractive soft robotics, and novel actuators and
sensors, and biomedical applications.
In this presentation, I will specifically focus on the role of
microstructures in the performance of deformable multifunctional
composites. We will consider how large deformations and elastic
instabilities can be used to trigger dramatic pattern transformations,
and to control a large variety of functionalities; in particular, the
design of switchable acoustic metamaterials will be
discussed. Recently discovered new type of instabilityinduced domain
formations in soft composites will be presented.
Analytical and numerical findings, as well as experimental results of
3Dprinted soft composites will illustrate the ideas.
Upcoming Seminar schedule
※title and abstract of each seminar will be informed when they are decided.
[MAT Seminar]
 Date:
 2019/03/19(Tuesday) 13:0015:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Jian Chen
 Title:
 Mechanics of soils  discrete and continuum approaches and a trial to bridge the two approaches
 Abstract:

From a “microscopic” viewpoint, soils consist of granular particles of varying sizes and shapes. Micromechanicsbased numerical methods, especially discrete element method (DEM), are popular for studying the mechanics/dynamics of such granular assemblies. A simple homebrew experiment would show that a heap can be made on a smooth mirror by pouring a pack of nonspherical particles, while it is extremely difficult using spherical particles. The geometry of individual particle plays an important role on the properties of granular assemblies, at least at the scales of sand heaps or laboratory experiments on soil samples (based on which constitutive laws derived). In this talk, we will introduce briefly a DEM approach which applies nonspherical particles for studying mechanics of soils.
Though simulations based on micromechanics are useful tools to investigate mechanics of soil samples, it is usually not applicable for engineering problems of much larger scales, e.g., assessing the occurrence of soil liquefaction caused by earthquakes. For engineering problems, soils are generally treated (homogenized) as continuum media and the corresponding partial differential equations (PDEs) are solved for proper initial and boundary conditions. Physically, soil liquefaction is an unstable phenomenon caused by an excessive increase of pore water pressure. Mathematically, it should correspond to some unstable solutions of the (linearized) partial differential equations for soil dynamics, giving that the constitutive law and governing equation are adequate for describing the behaviors of soils of interest. In this talk, we will present briefly a simple stability analysis for the solutions of the PDES derived from a continuum approach for soil mechanics.
In discrete approaches, the main material parameters are essentially the spring constants and the basic math tool is ordinary differential equation (ODE). In continuum approaches, the material parameters are the constitutive tensors and the basic math tool is PDE. In this talk, we will also discuss briefly how to drive a continuumnized governing equations (and the corresponding material tensors), which is a set of PDEs, from a system of ODEs for individual particles, considering the translations and rotations.
[MAT Seminar]
 Date:
 2019/03/12(Tuesday) 13:0015:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Takane Hori (CEAT)
 Title:
 Trial to revise governing equations of earthquake generation cycle simulations
 Abstract:

We have simulated earthquake generation cycles for decades basically based on the governing equations proposed by Rice (1993) (e.g., Hori et al., 2004). We have modified the algorithms to solve the equations, for example, to demonstrate much larger scale simulations introducing matrix approximation as Hierarchical matrix (Ohtani et al, 2011), to apply curved fault plane introducing triangular subfaults (Hyodo et al, 2013), to include the effects of heterogeneous medium property combining with FEM (Agata et al., in press), and so on. However, we have never changed the basic concept and equations. Recently, we have realized that the governing equations, especially the fault constitutive law, have too nonlinear and unstable characteristics for numerical calculations and too much parameters to be defined although we do not have enough data to constrain them. These become intrinsic problems when we use the equations to monitor and forecast earthquake generation cycle processes. In the seminar, I will explain above and introduce a concept how to revise the governing equations to overcome these difficulties.
[MAT Seminar]
 Date:
 2019/03/05(Tuesday) 13:0015:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Arthur Bauville (MAT)
 Title:
 Control of fault weakening on the structural styles of underthrustingdominated noncohesive tectonic wedges
Arthur Bauville, Mikito Furuichi, and Muriel Gerbault
 Abstract:

Underthrusting is a typical process at compressive margins responsible for nappe stacking and sediment subduction. Despite being essential for orogenesis, and the subduction cycle, the mechanics of underthrusting is still incompletely understood. The critical taper theory predicts that underthrusting occurs only in the limit where μ ′ = μ_{b}′ , with μ ′ and μ_{b}′ the inner and basal effective coefficient of friction, respectively. Paradoxically, in nature, underthrusting is associated with weak faults. For instance, at erosive subduction margins, sediments are dragged in subduction, while normal faulting in the middle prism suggests the plate boundary is weak. Wang et al., (2010) have proposed that dynamic fault strength variations during the earthquake cycle can reconcile these observations. In their model, underthrusting occurs during earthquakes when the fault is strong and normal faulting occurs in the prism during the interseismic period when the fault is weak. However, recent evidence (Gao and Wang, 2016) suggests that plate boundaries are relatively weak also during earthquakes.
In this contribution, we investigate the control of permanent fault weakening on the dynamics of a strongbased (μ′ = μ_{b}′ ) noncohesive tectonic wedge. We control the wedge material strength by a spatially constant fluid overpressure factor (λ_{ov}), and fault strength by a strain weakening factor (χ). We investigated the problem in two steps. First, we used the critical taper theory to determine a mechanical mode diagram that predicts structural styles. This diagram is deduced from the function ∆α(β,χ,λ_{ov}) which is the difference in taper angle at a given β, between the compressively critical taper of a strongbased wedge and the extensionally critical taper of a weakbased wedge (μ_{b}′ = (1 − χ)μ′). Then, we performed numerical simulations of accretionary wedge formation to establish the characteristics of these structural styles regarding wedge and fault geometry. We determined there is a continuum of structural styles with three endmembers occurring at the theoretically determined mechanical mode transitions.
Style 1 is characterized by thin tectonic slices and little to no underthrusting. Surface angles are bimodal with a peak corresponding to the strongbased wedge stability, and another corresponding to the compressively critical taper of a weakbased wedge. Style 2 is characterized by thick slices and nappe stacking. Surface angles are multimodal and vary cyclically within the stability of the weakbased wedge; with part of the wedge being asymptotic to the strongbased stability line. Style 3 is characterized by the complete underthrusting of the incoming sediments. Sediments are exhumed when they reach the backstop to form a weak upper wedge whose base is the paleosurface of the incoming sediments. Surface angles are asymptotic to the stability line of a fully weakened wedge (μ′ =μ_{b}′ =(1−χ)μ_{0}′, with μ_{0}′ a reference friction coefficient).
The model gives a mechanical explanation that reconciles underthrusting with weak faults. In agreement with observations, the model predicts the occurrence of style 2 in strongbased sandbox experiment (μ = 0.6, λ_{ov} = 0, χ = 0.1 − 0.2). Finally, we argue that style 3 is triggered by any mechanism, such as material weakening, erosion or sedimentation, that limits the surface angle such that the incoming material is always significantly undercritical.
[MAT Seminar]
 Date:
 2019/02/26(Tuesday) 13:0015:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Takashi Minoshima (MAT)
 Title:
 A divergencefree shockcapturing scheme for magnetohydrodynamics
 Abstract:

The magnetohydrodynamic (MHD) modeling has been extensively applied to various macroscopic dynamics of magnetized gas and fluid, for example, the planetary magnetosphere, the Sun, an accretion disk around a protostar, and so on. The MHD simulation is an indispensable tool to study these nonlinear phenomena. Since a supersonic flow is frequently observed in these circumstances, the numerical simulation needs to solve fully compressible MHD equations.
The socalled upwindtype shockcapturing method developed for compressible hydrodynamic simulations has been extended to compressible MHD simulations. However, the multidimensional MHD simulation based on the shockcapturing method should take special care of the divergencefree condition for the magnetic field in order to avoid an unphysical solution. The onedimensional shockcapturing method does not necessarily satisfy the divergencefree condition, and the simulation without any special divergence control could be easily crashed. Therefore, numerous studies have been devoted to develop novel methods so as to satisfy the shockcapturing property and the divergencefree condition. Here, we propose a new numerical scheme for MHD that satisfies the following criteria: (1) the scheme is based on a onedimensional shockcapturing method, (2) the multidimensional scheme is divergencefree within a machine precision and reduces to the base onedimensional scheme for onedimensional problems, (3) the order of accuracy of the scheme can be improved to an arbitrary level, and (4) the scheme is easy to implement to existing codes. We present details of the proposed scheme and its application to ideal MHD simulations, and then discuss its extension to HallMHD simulations that is a minimal model to take into account the kinetic effect.
[MAT Seminar]
 Date:
 2019/02/05(Tuesday) 13:0015:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Shigenobu Hirose (MAT)
 Title:
 Code Development for Calculating Chemical Equilibrium of Ideal Gas Mixture and its Opacity
 Abstract:

Hydrodynamics of multispecies gas mixture with radiative heat transport is relevant in many research areas including atmospheric science, engineering and astrophysics. To perform simulations of such hydrodynamics accurately, reliable databases of chemical equilibrium and opacity are required. However, public databases are available only in limited cases and researchers usually have to make databases by themselves for their own purposes. To reduce such tasks, I'm going to develop, as a project in the next midterm, a public code that calculates chemical equilibrium and opacity for wide ranges of pressure and temperature (T = 10 ~ 10^6 K) as well as for many kinds of chemical species. In this talk, I'll briefly explain the basic concept of developing such numerical codes.
[MAT Seminar]
 Date:
 2019/01/29(Tuesday) 13:0015:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Yasuko Yamagishi (MAT)
 Title:
 Improvement of database system for cruise information of research vessels
 Abstract:

We have developed a database system for cruise information of research vessels belonging to JAMSTEC to provide useful information for planning future research cruises. The database is available now on LAN of JAMSTEC. We are also developing an application to predict the downtime of future research cruise using the data provided by the database. At present, we are improving the database system because on developing the application it is revealed that the current database system cannot provide the information necessary for data analysis. In this seminar we will report the current status of the database improvement and the application development. The data stored into the database system is based on the “vessel operation report” which is Email sent from the vessel to the ground management department for research curises. At present, we are making the data manually from the Email. We are also developing an application for automatic creation of vessel operation reports to produce the data to be stored into the database system. In addition, to know the situation of the vessel operation report creation, I boarded on the research vessel. We will talk about the developing the application and the first boarding on the research vessel.
[MAT Seminar]
 Date:
 2019/01/22(Tuesday) 13:0015:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Yoji Kawamura (MAT)
 Title:
 Phase reduction of oscillatory thermal convection with a spatially translational mode
 Abstract:

We formulate a theory for the phase reduction of oscillatory thermal
convection in a twodimensional incompressible NavierStokes system
with laterally periodic boundary conditions. This system possesses
spatial translational symmetry in the lateral direction as well as
temporal translational symmetry. Oscillatory thermal convection in
this system is described by a limittorus solution that possesses two
phase modes; one is a spatial phase and the other is a temporal phase.
The spatial and temporal phases indicate the position and oscillation
of the thermal convection, respectively. The theory formulated in
this study can be considered as a phase reduction method for
limittorus solutions to partial differential algebraic equations
representing oscillatory thermal convection with a spatially
translational mode. The theory can also be considered as a
generalization of our earlier work [Kawamura and Nakao, Physica D
295296, 1129 (2015)], in which we studied oscillatory
cylindricalHeleShaw convection. In this talk, we derive the phase
sensitivity functions for spatial and temporal phases; these functions
quantify the spatiotemporal phase responses of oscillatory thermal
convection to weak perturbations applied at each point and at each
time. Using the phase sensitivity functions, we analyze the
spatiotemporal phase synchronization between weakly coupled systems of
oscillatory thermal convection.
[MAT Seminar]
 Date:
 2019/01/15(Tuesday) 13:0015:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Ryoichiro Agata (CEAT)
 Title:
 Segmentation of earthquake rupture caused by threedimensional material heterogeneity
 Abstract:

Subduction zone earthquakes are known to occur repeatedly in an identical portion of the entire focal region. This phenomenon is referred to as segmentation of earthquake rupture zone. The Nankai region in the southwest Japan hosts typical earthquake segmentations characterized by welldocumented earthquakes in the past several hundred years. Estimated elastic properties of rocks around the segmentation boundary of this region implies that local heterogeneity of elastic structure in the vicinity of the earthquake fault plane contributes to the earthquake segmentation (Kodaira et al. 2006). In this study, we examine how such local heterogeneous elastic structures affects earthquake generation patterns using an idealized subduction zone model by directly considering such heterogeneity in largescale numerical simulations of earthquake generation, which has not been done before due in part to the associated computation cost. We show that under certain conditions, earthquake segmentation is seen in the simulation results when a local stiff domelike structure, which resembles one found by Kodaira et al (2006) off Kiipeninsula, is located in the overriding plate on the fault plane.
[MAT Seminar]
 Date:
 2019/01/08(Tuesday) 11:0012:00
 Place:
 MAT theater on the 5th floor, IT Building, Yokohama Institute
 Language:
 English
 Speaker:
 Muneo Hori (MAT)
 Title:
 REBEL WITH CAUSE  CORRECTION OF ELASTOPLASTICITY THEORY
