To determine the factors contributing to large-scale blue tide occurrences, blue tide distributions captured from satellite images were estimated using the estimation model for blue tides. This model was developed based on the observation results of the optical properties of Tokyo Bay and numerical simulation by a three-dimensional hydrodynamics and ecological model. By complementing the information from the results of both satellite image analysis to estimate the blue tide reflectance and the numerical simulation to calculate the dissolved oxygen (DO) concentration, the upwelling areas of sulfide-containing anoxic water were distinguished from those of non-sulfide-containing anoxic water.
In order to analyze effectively the vibration response of the structure members of railway reinforced concrete viaducts, we have developed a new analysis method. It divides the whole railway system into vehicles/track model, and the track/structure model. Using these models, we examined the influence of various parameters of the vehicle, track and structure on the structure member vibration. As a result, the following has become clear. For example, for 20Hz or more, unsprung wheelset mass has a great influence on response of structure member. For 20-100Hz and 150Hz or more, rail surface roughness, for 60Hz or more, the stiffness of the rail pad, for less than 60Hz, the stiffness of the slab of structure and so on, have great influence on structure member vibration respectively.
As one of the next generation FEL technologies, staggered array undulator (SAU) made by high Tc superconductor (HTSC) is developed to achieve small size and high intensity magnetic field undulator. In use of superconductor magnet, it is very difficult to adjust the magnet array of the undulator after HTSC magnetization since the entire undulator is installed inside a cryostat, therefore appropriate magnet alignment has to be determined before the magnetization. This paper presents a development of a numerical simulation code of the SAU magnetization process for assisting the design of an optimal size and alignment of the SAU magnets.
Radiotherapy is one method of treatment for cancer. To achieve safe, accurate and efficient irradiation to the diseased portion of a patient during radiotherapy, simulation software, treatment-planning software and various treatment devices for radiotherapy have been developed. Our visualization software for radiotherapy simulations, gMocren, is described in this paper. It was developed in combination with a unique Monte Carlo simulator for radiotherapy that calculates fully physical behaviors in both the complex beam-delivery devices and the patient’s body. The visualization software creates images of the geometrical data of the beam-delivery devices, calculated physical quantities, particle trajectories, and the patient volume data set from the radiotherapy simulation. gMocren has been used to assist in the development of beam-delivery devices and simulation software. It assists users in visually and intuitively understanding the results of such simulations. The manual, an online tutorial and downloadable files for gMocren are also available to users on the gMocren website.
In this paper, we present a particle-based fluid simulation method with deforming obstacles. For defining the geometric information of time-varying obstacles, we employ the implicit function form that varies along the time axis. Our fluid simulation is based on the Moving Particle Semi-implicit (MPS) method, one of the typical particle-based methods for incompressible flow, and we add a new formulation for implicitly defined deforming obstacles. Although many of the particle-based algorithms require a set of particles on the obstacles, our method works without generating particles on the obstacles. We define the particle motion at the vicinity of the obstacles and generate a new algorithm specific to the implicit deforming obstacles by incorporating approximations in the processes of determining particle forces affected by the obstacles.
In this paper, we propose a visualization development framework that is a C++ class library for easily and efficiently rendering three-dimensional datasets, such as numerical simulation results, medical image datasets, and measurement datasets. Our framework provides a modular programming environment that supports the construction and execution of a visualization pipeline and allows the user to readily implement custom visualization algorithms using our simple module-based execution model. In addition, we also provide a simple implementation of a visualization environment that can handle multiple volumes and semi-transparent polygons in a single scene, which we call a fused visualization environment. Although many visualization software packages have been proposed thus far, such an environment has not previously been supported because of the visibility ordering problem. To confirm the effectiveness of our proposed visualization framework, we demonstrate several visualization applications implemented using this framework.
Ubiquitous computing is an important technology in medicine that is predicted to support doctors anywhere and anytime. To help achieve it, this paper develops the Interactive Segmentation and Visualization System for Medical images on Mobile devices (ISVS_M2), which originally was designed to work on workstations, but also works on a wide range of mobile devices via a mobile client-server platform. The developed ISVS_M2 basically consists of three modules: asegmentation module that is implemented on a server; commutation modules on both the server and mobile device; and interactive and visualization modules on the mobile device that not only give visualization of internal information of an organ model, but also interactively refine organ segmentation according to user experience. With interaction via a computer graphic interface on the mobile device, and communication via the mobile client-server platform, ISVS_M2 offers users a novel and efficient approach to computer-aided medicine.
Large-scale numerical simulations on modern leading-edge supercomputers have been continuously generating tremendous amount of data. In-Situ Visualization is widely recognized as the most rational way for analysis and mining of such large data sets by the use of sort-last parallel visualization. However, sort-last method requires communication intensive final image composition and can suffer from scalability problem on massively parallel rendering and compositing environments. In this paper, we present the Multi-Step Image Composition approach to achieve scalability by minimizing undesirable performance degradation on such massively parallel rendering environments. We verified the effectiveness of this proposed approach on K computer, installed at RIKEN AICS, and achieved a speedup of 1.8× to 7.8× using 32,768 composition nodes and different image sizes. We foresee a great potential of this method to meet the even larger image composition demands brought about by the rapid increase in the number of processing elements on modern HPC systems.
In this paper, we discuss the device performance of atomic monolayer semiconductor FETs composed of Si, Ge and C elements. First, we present the performance potentials of silicene nanoribbon (SiNR), germanene nanoribbon (GeNR) and graphene nanoribbon (GNR), which all have a sufficient band gap to switch off, as a field-effect transistor (FET) channel material. We demonstrate that comparing at the same band gap of ~ 0.5 eV, GNR FET maintains the advantage over SiNR or GeNR FETs under an ideal transport situation, but SiNR and GeNR are attractive channel materials for high performance FETs as well. Next, we compute the electronic band structure and the electron mobility of germanane, which is a hydrogen-terminated Ge monolayer. We demonstrate that germanane has a band gap larger than 1 eV without a nanoribbon structure, and an effective mass smaller than bulk Ge. Therefore, germanane is also a promising two-dimensional material as a FET channel.
In this paper, a Fermi-Dirac statistics based quantum energy transport (FDQET) model is developed for numerical simulations of high mobility MOSFETs. The QET model allows simulations of carrier transport including quantum confinement and hot carrier effects. Fermi-Dirac statistics are further considered for the analysis of device characteristics with high degeneracy material such as In0.53Ga0.47As. Numerical stability and convergence are achieved by developing an iterative solution method used when Fermi-Dirac statistics are modeled. Numerical results for Si, Ge and In0.53Ga0.47As bulk n-MOSFETs are presented. The FDQET model allows us to evaluate the device characteristics with high degeneracy material such as In0.53Ga0.47As.
Ballistic performance of graphene, silicene, and germanene-nanoribbon field-effect transistors (FETs) with a gate-length of 10 nm has been numerically investigated. The graphene-nanoribbon FET is found to have the largest ON-current when one compare FETs with a nanoribbon channel having a nearly equal band-gap Eg ≈ 0.5 eV. The graphene device exhibits the largest OFF-current due to the smallest effective-mass enhancing the source-drain direct tunneling.
In this study, we theoretically optimize a two-dimensional (2D) channel doping profile of metal-oxide-semiconductor field-effect transistors (MOSFETs) with given current voltage (I-V) characteristics by using a geometric programming (GP) technique. Inverse modeling of channel doping profile for device characteristics with simultaneously considering the short-channel effect (SCE) and random-dopant-fluctuation-induced threshold voltage fluctuation (RDF-induced σVt) is advanced. The formulated model of doping profile is a GP problem which can be transformed into a convex optimization problem and solved globally and efficiently. Constrains of I-V characteristics with including the RDF-induced σVt are included to optimize desired doping profiles. The optimization methodology is applied for 45-nm MOSFET devices and the results are validated with 2D numerical device simulation. This approach provides an alternative way to design doping profile for various technologies of MOSFETs.
We investigate the mechanism of off-leakage current in InGaZnO (IGZO) thin-film transistors with the help of a two-dimensional device simulator. The deep donorlike states probably originating from the oxygen vacancies are introduced in the IGZO channel, and it is shown that these trap states significantly affect Id–Vg characteristics in the off-state region through the pinning of the channel potential. A simple analytical model to explain the simulation results is proposed, which suggests that the off-leak characteristics is controlled by the amount and the depth of the deep donorlike states as well as the thicknesses of IGZO and SiO2 layers in TFT.
We demonstrate that the parallel computing with graphic processing unit (GPU) effectively accelerates a particle-based carrier transport simulation called EMC/MD method. The simulation speed is increased by parallelizing the point-to-point Coulomb’s force calculation, which is sufficient to accomplish a device characteristic simulation of nanostructured metal-oxide-semiconductor field effect transistor (MOSFET) including source and drain diffusion regions. The EMC/MD simulation powered by GPU computing is a useful tool to investigate the statistical variability analysis of nano-scale transistors.