Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Watanabe, Seiya*; Kawahara, Jun*; Aoki, Takayuki*; Sugihara, Kenta; Takase, Shinsuke*; Moriguchi, Shuji*; Hashimoto, Hirotada*
Engineering Applications of Computational Fluid Mechanics, 17(1), p.2211143_1 - 2211143_23, 2023/00
Times Cited Count:1 Percentile:52.66(Engineering, Multidisciplinary)In tsunami inundations or slope disasters of heavy rain, a lot of floating debris or driftwood logs are included in the flows. The damage to structures from solid body impacts is more severe than the damage from the water pressure. In order to study free-surface flows that include floating debris, developing a high-accurate simulation code of free-surface flows with high performance for large-scale computations is desired. We propose the single-phase free-surface flow model based on the cumulant lattice Boltzmann method coupled with a particle-based rigid body simulation. The discrete element method calculates the contact interaction between solids. An octree-based AMR (Adaptive Mesh Refinement) method is introduced to improve computational accuracy and time-to-solution. High-resolution grids are assigned near the free surfaces and solid boundaries. We conducted two kinds of tsunami flow experiments in the 15 and 70 m water tanks at Hachinohe Institute of Technology and Kobe University to validate the accuracy of the proposed model. The simulation results have shown good agreement with the experiments for the drifting speed, the number of trapped wood pieces, and the stacked angles.
Hasegawa, Yuta; Imamura, Toshiyuki*; Ina, Takuya; Onodera, Naoyuki; Asahi, Yuichi; Idomura, Yasuhiro
Proceedings of 13th Workshop on Latest Advances in Scalable Algorithms for Large-Scale Heterogeneous Systems (ScalAH22) (Internet), p.10 - 17, 2022/00
The ensemble data assimilation of computational fluid dynamics simulations based on the lattice Boltzmann method (LBM) and the local ensemble transform Kalman filter (LETKF) is implemented and optimized on a GPU supercomputer based on NVIDIA A100 GPUs. To connect the LBM and LETKF parts, data transpose communication is optimized by overlapping computation, file I/O, and communication based on data dependency in each LETKF kernel. In two dimensional forced isotropic turbulence simulations with the ensemble size of and the number of grid points of
, the optimized implementation achieved
speedup from the naive implementation, in which the LETKF part is not parallelized. The main computing kernel of the local problem is the eigenvalue decomposition (EVD) of
real symmetric dense matrices, which is computed by a newly developed batched EVD in EigenG. The batched EVD in EigenG outperforms that in cuSolver, and
speedup was achieved.
Hasegawa, Yuta; Aoki, Takayuki*; Kobayashi, Hiromichi*; Idomura, Yasuhiro; Onodera, Naoyuki
Parallel Computing, 108, p.102851_1 - 102851_12, 2021/12
Times Cited Count:3 Percentile:41.82(Computer Science, Theory & Methods)The aerodynamics simulation code based on the lattice Boltzmann method (LBM) using forest-of-octrees-based block-structured local mesh refinement (LMR) was implemented, and its performance was evaluated on GPU-based supercomputers. We found that the conventional Space-Filling-Curve-based (SFC) domain partitioning algorithm results in costly halo communication in our aerodynamics simulations. Our new tree cutting approach improved the locality and the topology of the partitioned sub-domains and reduced the communication cost to one-third or one-fourth of the original SFC approach. In the strong scaling test, the code achieved maximum speedup at the performance of 2207 MLUPS (mega- lattice update per second) on 128 GPUs. In the weak scaling test, the code achieved 9620 MLUPS at 128 GPUs with 4.473 billion grid points, while the parallel efficiency was 93.4% from 8 to 128 GPUs.
Ebihara, Kenichi
JAERI-Research 2005-004, 121 Pages, 2005/03
This report is the JAERI's report version of the doctor thesis by the author. In this report, first, the validity and usefulness of the application of the two-phase fluid model of the lattice-gas method and the lattice Boltzmann method(LBM) are examined. On the basis of the examination, next, the horizontal stratified two-phase flow that is the fundamental and important flow is simulated by the HCZ model which is one of the two-phase fluid model of the LBM. It is seen that the interfacial growth of the HCZ model satisfies the Kelvin-Helmholtz instability theory and reproduces the theoretical two-phase flow regime map of Taitel and Dukler(T-D map). It is found that more superficial flow velocity of the rare phase is necessary in the channel with the narrow width. The HCZ model can also simulate the droplet generation accompanying more complex interfacial phenomena and reproduce the experimental correlation of Ishii and Grolmes in the range of the distribution of the experimental data.
Ebihara, Kenichi; Watanabe, Tadashi
Proceedings of 2004 ASME International Mechanical Engineering Congress and Exposition (IMECE '04) (CD-ROM), 9 Pages, 2004/11
This paper describes the numerical simulation of the interfacial growth of the stratified wavy two-phase flow in the horizontal rectangular channel. The influence of the channel width upon the growth of the interfacial wave was evaluated by the several simulations for the different channel width. The numerical simulation model adopted in this paper is a one-component two-phase fluid model of the lattice Boltzmann method. The wave growth was observed and the dimensionless numbers that characterize the two-phase flow state were measured during the computations. The relation between the wave growth and the dimensionless numbers, which characterize the flow state, was compared with that in the flow pattern map proposed by Taitel and Dukler. It was verified in the case of the wide channel width that the simulated relation was almost in agreement with that in the flow pattern map. It was shown that the narrower the channel width became, the more mass flow rate of the rare phase the interfacial growth needed and the obtained relation deviated from that in the flow pattern map.
Ebihara, Kenichi
Tsukuba Daigaku Daigakuin Shisutemu Joho Kogaku Kenkyuka Hakase Gakui Rombun, 134 Pages, 2004/09
In this thesis, first the liquid-gas models of the lattice method are examined by applying them to two-phase flow simulations. Next the liquid-gas model(the HCZ model) of the lattice Boltzmann method is applied to the three-dimensional simulation of the horizontal stratified two-phase flow. The following results are obtained. (1)The two- and three- dimensional interface simulated by the HCZ model satisfies the Kelvin-Helmholtz instability theory. (2)In the simulation of the interfacial growth in the rectangular channel, it is found that the relation between the interfacial growth and the flow state is in agreement with the flow regime map proposed theoretically by Taitel and Dukler. (3)It is also found that the three dimensionality becomes remarkable and the interfacial growth needs more flow rate of the rare phase than that of the theoretical flow regime map when the channel width is narrower. (4)In the droplet creation simulation, it is found that the relation between the droplet creation and the flow state simulates the experimental correlation proposed by Ishii and Grolmes.
Ebihara, Kenichi; Watanabe, Tadashi
Nihon Kikai Gakkai Rombunshu, B, 70(694), p.1393 - 1399, 2004/06
The horizontal stratified two-phase flow in the rectangular pipe whose width in smaller than the height is simulated by the one-component two-phase lattice Boltzmann method. The interfacial growth between two phases is measured for three cases with the different pipe width and the measured dimensionless number charactering the two-phase flow is compared with the flow regime map proposed by Taitel and Dukler. It is found that the boundary separating the interfacical growth from the non-growth which is obtained by the simulations is larger in the flow regime map when the pipe width is narrower.
Hasegawa, Yuta; Onodera, Naoyuki; Idomura, Yasuhiro
no journal, ,
Towards the plume dissipation prediction simulation in urban area with dense buildings, we are developing the urban wind simulation code CityLBM based on a lattice Boltzmann method with GPU computing. In this work, we develop memory reduction implementations for CityLBM code by optimizing the size of MPI buffer, and by removing calculation grids inside the buildings or the underground of undulating terrain. We also implement an ensemble simulation based on MPI communicator splitting, which manages all ensemble members by single mpirun execution. The ensemble simulation for the field tracer dissipation experiment in Oklahoma-city was performed, and it is confirmed that the results of ensemble simulation agreed with the experimental result, while a single simulation could not reproduce the experiment. These results showed that the ensemble simulation improves the reliability of plume dissipation prediction simulations.
Hasegawa, Yuta; Aoki, Takayuki*; Kobayashi, Hiromichi*; Idomura, Yasuhiro; Onodera, Naoyuki
no journal, ,
We developed a block-structured static adaptive mesh refinement (AMR) CFD code for the aerodynamics simulation using the lattice Boltzmann method on GPU supercomputers. The data structure of AMR was based on the forest-of-octrees, and the domain partitioning algorithm was based on space-filling curves (SFCs). To reduce the halo data communication, we introduced the tree cutting approach, which divided the global domains with a few octrees into small sub-domains with many octrees, leading to a hierarchical domain partitioning approach with the coarse structured block and the fine SFC partitioning in each block. The tree cutting improved the locality of the sub-divided domain, and reduced both the amount of communication data and the number of connections of the halo communication. In the strong scaling test on the Tesla V100 GPU supercomputer, the tree cutting approach showed 1.82 speedup at the performance of 2207 MLUPS (mega-lattice update per second) on 128 GPUs.
Onodera, Naoyuki; Idomura, Yasuhiro; Hasegawa, Yuta; Nakayama, Hiromasa; Shimokawabe, Takashi*; Aoki, Takayuki*
no journal, ,
This paper presents a novel data assimilation method in realistic turbulent boundary layer simulations for the realization of a wind digital twin. We have developed a plume dispersion simulation code named CityLBM based on a lattice Boltzmann method. CityLBM enables a real time ensemble simulation for several km square area by applying locally mesh-refinement method on GPU supercomputers. Mesoscale wind boundary conditions produced by a Weather Research and Forecasting Model are given as boundary conditions in CityLBM by using a nudging data assimilation method. In this study, we propose a dynamic nudging data assimilation method, where a particle filter optimizes the nudging coefficient based on the observation data. This approach gave reasonable agreements in vertical profiles of the wind speed, the wind direction, and the turbulent intensity compared to the observation data throughout the day, and enabled all-day simulations, where atmospheric conditions change significantly.
Hasegawa, Yuta; Onodera, Naoyuki; Idomura, Yasuhiro
no journal, ,
To reduce memory usage and accelerate data communication in the locally-refined lattice Boltzmann code, we tried an intra-node multi-GPU implementation using Unified Memory in CUDA. In the microbenchmark test with uniform mesh, we achieved 96.4% and 94.6% parallel efficiency on weak and strong scaling of a 3D diffusion problem, and 99.3% and 56.5% parallel efficiency on weak and strong scaling of a D3Q27 lattice Boltzmann problem, respectively. In the locally mesh-refined lattice Boltzmann code, the present method could reduce memory usage by 25.5% from the Flat MPI implementation. However, this code showed only 9.0% parallel efficiency on strong scaling, which was worse than that on the Flat MPI implementation.
Onodera, Naoyuki
no journal, ,
A plume-dispersion simulation of the environmental dynamics of radioactive sub-stances is very important from the viewpoint of nuclear security. Since air-flows in large cities are turbulent with Reynolds numbers of several million, large-scale CFD simulations are needed. We developed a CFD code based on the adaptive mesh-refined Lattice Boltzmann Method (AMR-LBM) on GPU-rich supercomputers. In this presentation, we realize both high-resolution and real-time wind simulations in urban areas on the NVIDIA DGX-2 supercomputer.