A Semi-explicit Large Eddy Simulation (SeLES) is proposed to help handle interactions between discontinuities and turbulence when using conventional Sub-Grid Scale (SGS) models of large eddy simulation. For the grid-scale simulation, the numerical viscosity becomes effective in the limited region detected by the discontinuity sensor, whereas a central difference scheme is applied elsewhere. In our SeLES, the discontinuous properties are considered in the SGS portion. Numerical tests are demonstrated for compressible coaxial jet flows of gas fuel and air without reaction. A comparison of the numerical results and experimental data confirms that this SeLES successfully prevents the overlap of numerical viscosity and SGS viscosity.

Some deep learning-based methods, Deep Galerkin Method (DGM), Physics Informed Neural Networks (PINNs), and others, have been suggested for solving partial differential equations. In these methods, a loss function for training a deep neural network is formulated, so that differential operators, boundary conditions, and initial conditions of the intended partial differential equation are satisfied. This study applied a deep learning-based method to solve viscous and resistive magnetohydrodynamic (MHD) equations. In particular, the deep learning-based method is applied to the 1D Brio-Wu shock wave problem, 2D Balsara-Spicer MHD Rotor problem, 2D Orszag-Tang MHD Vortex problem, and 2D magnetosphere problem. The approximated solution obtained by the deep learning-based method generally agrees with that obtained by conventional numerical methods.

The performance of a superconducting linear accelerator (SLA) system for a pellet injection to fuel nuclear fusion reactors has been investigated numerically. For this purpose, a numerical code has been developed to analyze the shielding current density and dynamic motion of a high-temperature superconducting thin film by means of finite element method. In the present study, an exponentially increasing current was employed to drastically accelerate a pellet in the container. Computational results indicate that the acceleration distance required for the SLA system to reach the target speed of 5 km/s was approximately 2.5 km, it is reduced to approximately 88%. It was therefore concluded that the proposed current has the potential to improve acceleration performance significantly.

Volume rendering is useful for visualizing computer fluid dynamics (CFD) data, and its VR visualization helps to understand complex 3D data. For visualizing large-scale data in remote locations, the remote volume rendering using VR is an important issue in the visualization field. A remote visualization application CS-PBVR can interactively visualize the large-scale datasets in remote locations with volume rendering. In order to extend CS- PBVR into VR-PBVR which is applicable to a head mount display (HMD) Oculus rift S, we added a stereo image generation function, a gesture control function, and a renewed processing flow. VR-PBVR achieved interactive visualization of remotely located test dataset (2M cells) with 90 fps.

In this paper, a simple method is proposed that allows two planar inverted-F antennas (PIFAs) maintain their bandwidth while simultaneously reducing the mutual coupling between them. By creating a slit on each PIFA patch and then connecting the two PIFAs by a bridge line (BL), the bandwidth can be maintained and the mutual coupling between them can be simultaneously reduced to less than －10 dB. Therefore, total antenna efficiency can be significantly improved over a wide band. The proposed model is fabricated and measured, and the simulation results are validated by the measurement results.

Time-frequency analysis is the main method used to analyze acoustic signals, most notably the short-time Fourier transform and the continuous wavelet transform. In contrast, the wavelet-synchrosqueezing transform improves the frequency resolution by calculating the angular frequency from the result of the continuous wavelet transform and its time derivative. This paper proposes an improved method for calculating angular frequencies using time shifts in the wavelet-synchrosqueezing transform. The simulation results are compared with the previous research, and it is confirmed that the speed is increased and the followability to the frequency change is improved.

X-ray phase contrast image has higher sensitivity to materials with light elements than the conventional absorption contrast. Phase image measurement technique using diffraction gratings can be used in the laboratory. Furthermore, X-ray phase microscope using the diffraction gratings has been also developed. However, phase measurement with diffraction gratings has a drawback that the spatial resolution is limited to the grating period. To overcome this problem, a scanning system using a triangular phase grating has recently been proposed, which is expected to archive a large field of view and a high spatial resolution. This study aims to investigate the feasibility of the system and a deep understanding of the optical system by simulation.

This paper provides a topology optimization (TO) of motor stator using an on/off method with considering vibration. The on/off method using evolutionary optimization tends to yield complex shapes, which are impossible to produce in view of engineering. To solve this problem, two methods are introduced: a on/off method with filtering and a normalized Gaussian function network (NGnet) method. Optimization results show that the effectiveness of these methods is confirmed in a TO of an interior permanent magnet motor.

This study considers inclusion methods for the multiplication of three point matrices. These methods are essential for verification methods such as for a matrix determinant, generalized eigenvalue decomposition, singular value decomposition, square roots and inverse square roots of a matrix, and Sylvester equations. We consider three methods that involve five matrix multiplications and compare the tightness of enclosures and numerical experiments. The proposed methods produce tighter results than the previous method with the same computational cost. We also show methods for obtaining the upper bound of the maximum norm of a product of three point matrices.

Surface electromyogram (sEMG) has been used to estimate human motions. Feature extraction (FE) has been proposed as a process to decrease noise and extract information about muscle activities from measured raw sEMG signals. Electromechanical delay (EMD) is a time delay between a generation of sEMG and subsequent occurrence of human motions, and sEMG can be obtained in advance of motions occurrence by using this time delay. The purpose of this study is to propose a real time human motion prediction method using a neural network model and FE-processed sEMG signals utilizing EMD. Experimental results show that the proposed method has better prediction results among other methods.

Recently, we have proposed an efficient method for solving saddle point problems using their block structure. Instead of solving an original problem directly, this method solves a linear system with multiple right-hand sides. Then it computes solution vectors of the original problem using a solution matrix of this linear system. Since solving a linear system with multiple right-hand sides is the most time-consuming part of our method, it is desirable to solve that as fast as possible. In this paper, a parallelization strategy of our method is proposed, and our method is implemented on a parallel computation environment. Moreover, to achieve further speed-up of our method, this paper also proposes an implementation strategy using GPUs. Numerical experiments illustrate the parallel performance of our method on a GPU cluster.

The effect of hydrogen on the structural change of vacancies in tungsten is analyzed by molecular dynamics in order to clarify the interaction between vacancies and hydrogen in tungsten. Simulations are performed at different temperatures (573, 773, 1073 K) and with different numbers of hydrogen in the vacancies (0, 18, 36, 54). Evaluating (1) isopotential surface for the increase of total potential energy, (2) the root mean square deviation (RMSD) of tungsten atoms, (3) the root mean square fluctuation (RMSF) of tungsten atoms and (4) density distribution in radial direction, we found that the presence of a large number of hydrogen in a vacancy at each temperature caused changes in the structure. This fact partially supports the experimental fact that in the coalescence of two tungsten vacancies, the retention of hydrogen activates them.

A numerical method has been presented to compute the time structure of the radiation emitted from a relativistic electron moving inside an undulator, which plays important roles in the cutting-edge synchrotron radiation applications. The method is based on the Lienard-Wiechert potential in conjunction with the Lorentz equation to calculate the electron motion in retarded time. In this method, we used the Fourier transformation to calculate the spectral structure of the radiation and cut off high harmonic components, which do not propagate in the optical components in the experiment. Then, by using inverse Fourier transformation the time structure of the fundamental radiation has been obtained. We show some numerical examples under some experimental conditions. An interesting finding is that, by applying an external uniform magnetic field and deflecting the electron orbit, the waveform is chirped which may be interesting for some applications.

For the automatic measurement of DNA length in the video captured by fluorescence microscope, image processing method using pix2pix is developed. To increase the accuracy of the measurement in low-resolution video, a method of inputting multiple time-series frames to pix2pix are proposed. To generate the training data, molecular dynamics simulation is employed to prepare dummy video. The trained pix2pix by multiple frames input can generate more accurate images in a scene of DNA breakage than the pix2pix trained by single frame input.

Origami-based structures utilize the ancient art of origami to solve various engineering problems, particularly in energy absorption applications. This paper presented a newly designed truncated origami structure based on the Miura-ori pattern. As the truncated ratio e decreased, the truncated structure gradually transformed from a Miura-ori structure to a condensed solid. The behavior of truncated origami structures under quasi-static compressive loading in the in-plane direction was analyzed through numerical simulations. The experimental results confirmed the validity of the simulation models. The truncated origami structure with varying e was investigated, and the structure with e = 0.6 exhibited the best overall energy absorption performance. The truncated origami structure was compared to honeycomb structures with the same relative density, and it performed better in specific energy absorption (SEA) in two in-plane directions.

The behavior ‘Turn alternation’ that is seen in pull bug and others is investigated as one of intelligence of natural creatures in this paper. Some numerical simulations are carried out on the square lattice with some obstacles. As the number of obstacles increases, the reachable rate to the goal largely decreases. In order to investigate the effectiveness of this strategy, we compare it with that of random walk, it is clarified that it takes more time to reach the goal as obstacles increases. Moreover that tendency is more remarkable in the larger simulation space size. However under some obstacle allocation, the strategy of turn alternation enables us to move in the wider range. Therefore it has some possibility to be a promising search algorithm if we combine it with other algorithms.

The digitalization of diagnostic tools is progressing in the medical field, with all hospitals storing big data, such as patient data including diagnostic data. Programming techniques are needed to analyze these big data, yet few researchers in the medical field are skilled at programming. In this study, we established a cross-disciplinary research team for the data analysis of Kawasaki disease (KD). We conducted a KD patient trend analysis using data from the 25^th and 26^th rounds of the Nationwide KD Survey in Japan. We found that there were fewer KD patients than in previous years and that there was a slight correlation between the percentage reduction in the weekly number of patients with KD and the percentage reduction in the traveling population at major stations.