Faculty Profiles - ISHIHARA Takashi

Papers - ISHIHARA Takashi

Division display 1 - 20 of about 46 ／ All the affair displays >>

Direct numerical simulations of compressible turbulence in a periodic box: Effect of isothermal assumptions on turbulence statistics Reviewed

Yoshiki Sakurai, Takashi Ishihara

Physics of Fluids 36 ( 8 ) 2024.8

Authorship：Last author Publishing type：Research paper (scientific journal) Publisher：AIP Publishing

Direct numerical simulations of compressible nonisothermal turbulence in a periodic box with up to 40963 grid points were conducted by varying Reynolds numbers and resolution levels. The results were compared with those of compressible isothermal turbulence by Sakurai and Ishihara [“Direct numerical simulations of compressible isothermal turbulence in a periodic box: Reynolds number and resolution-level dependence,” Phys. Rev. Fluids 8, 084606 (2023)] to study the effect of isothermal assumption on turbulence statistics. The turbulent Mach number and ratio of the dilatational to solenoidal root mean square velocities were fixed at approximately 0.3 and 0.4, respectively. A comparison under approximately equal flow conditions showed that the dilatational component of the energy spectra for the nonisothermal case exhibited approximately equal k−3 scaling at kη&gt;1 as observed for the isothermal case and was consistently smaller in the wavenumber range 0.05&lt;kη&lt;0.6 than that for the isothermal case, where η is the Kolmogorov length. The dilatational energy is mainly dissipated around kη≈0.3, the same wavenumbers as the solenoidal energy dissipation irrespective of the isothermal assumption. As the Reynolds number increased, the dilatational energy dissipation caused by shocklets around kη≈2 became larger, especially in the nonisothermal case. It was found that the isothermal assumption weakened the intermittency of the velocity divergence. No significant differences were observed in the normalized mean energy dissipation rates and pressure statistics. The local flow topology was also marginally affected by the isothermal assumption; however, the difference was significantly less than the changes owing to the different values of parameters such as the Reynolds number and Mach number.

DOI： 10.1063/5.0216518

researchmap

File I/O Cache Performance of Supercomputer Fugaku Using an Out-of-Core Direct Numerical Simulation Code of Turbulence Reviewed

Yuto Hatanaka, Yuki Yamane, Kenta Yamaguchi, Takashi Soga, Akihiro Musa, Takashi Ishihara, Atsuya Uno, Kazuhiko Komatsu, Hiroaki Kobayashi, Mitsuo Yokokawa

Computational Science – ICCS 2024 173 - 187 2024.6

An Asymptotic Parallel Linear Solver and Its Application to Direct Numerical Simulation for Compressible Turbulence Reviewed

Mitsuo Yokokawa, Taiki Matsumoto, Ryo Takegami, Yukiya Sugiura, Naoki Watanabe, Yoshiki Sakurai, Takashi Ishihara, Kazuhiko Komatsu, Hiroaki Kobayashi

Computational Science – ICCS 2024 383 - 397 2024.6

Intermittency across Reynolds numbers – the influence of large-scale shear layers on the scaling of the enstrophy and dissipation in homogenous isotropic turbulence Reviewed

G.E. Elsinga, T. Ishihara, J.C.R. Hunt

Journal of Fluid Mechanics 974 2023.10

Publishing type：Research paper (scientific journal) Publisher：Cambridge University Press (CUP)

Direct numerical simulations up to Re_λ = 1445 show that the scaling exponents for the enstrophy and the dissipation rate extrema are different and depend on the Reynolds number. A similar Reynolds number dependence of the scaling exponents is observed for the moments of the dissipation rate, but not for the moments of the enstrophy. Significant changes in the exponents occur at approximately Re_λ ≈ 250, where Re_λ is the Taylor based Reynolds number, which coincides with structural changes in the flow, in particular the development of large-scale shear layers. A model for the probability density functions (PDFs) of the enstrophy and dissipate rate is presented, which is an extension of our existing model (Proc. R. Soc. A, vol. 476, 2020, p. 20200591) and is based on the mentioned development of large-scale layer regions within the flow. This model is able to capture the observed Reynolds number dependencies of the scaling exponents, in contrast to the existing theories which yield constant exponents. Moreover, the model reconciles the scaling at finite Reynolds number with the theoretical limit, where the enstrophy and dissipation rate scale identically at infinite Reynolds number. It suggests that the large-scale shear layers are vital for understanding the scaling of the extrema. Furthermore, to reach the theoretical limit, the scaling exponents must remain Reynolds number dependent beyond the present Re_λ range.

DOI： 10.1017/jfm.2023.799

researchmap

Direct numerical simulations of compressible isothermal turbulence in a periodic box: Reynolds number and resolution-level dependence Reviewed

Yoshiki Sakurai, Takashi Ishihara

Physical Review Fluids 8 ( 8 ) 084606-1 - 084606-24 2023.8

I/O Performance Evaluation of a Memory-Saving DNS Code on SX-Aurora TSUBASA Reviewed

Mitsuo Yokokawa, Yuki Yamane, Kenta Yamaguchi, Takashi Soga, Taiki Matsumoto, Akihiro Musa, Kazuhiko Komatsu, Takashi Ishihara, Hiroaki Kobayashi

2023 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW) 2023.5

Performance evaluation of parallel direct numerical simulation code on supercomputer SX-Aurora TSUBASA Reviewed

Mitsuo Yokokawa, Yujiro Takenaka, Takashi Ishihara, Kazuhiko Komatsu, Hiroaki Kobayashi

Computers and Fluids 261 105913 - 105913 2023.4

Non-local dispersion and the reassessment of Richardson's t3-scaling law Reviewed

G.E. Elsinga, T. Ishihara, J.C.R. Hunt

Journal of Fluid Mechanics 932 A17-1 - A17-32 2022.2

Publishing type：Research paper (scientific journal) Publisher：Cambridge University Press (CUP)

The Richardson-scaling law states that the mean square separation of a fluid particle pair grows according to <italic>t</italic>³ within the inertial range and at intermediate times. The theories predicting this scaling regime assume that the pair separation is within the inertial range and that the dispersion is local, which means that only eddies at the scale of the separation contribute. These assumptions ignore the structural organization of the turbulent flow into large-scale shear layers, where the intense small-scale motions are bounded by the large-scale energetic motions. Therefore, the large scales contribute to the velocity difference across the small-scale structures. It is shown that, indeed, the pair dispersion inside these layers is highly non-local and approaches Taylor dispersion in a way that is fundamentally different from the Richardson-scaling law. Also, the layer's contribution to the overall mean square separation remains significant as the Reynolds number increases. This calls into question the validity of the theoretical assumptions. Moreover, a literature survey reveals that, so far, <italic>t</italic>³ scaling is not observed for initial separations within the inertial range. We propose that the intermediate pair dispersion regime is a transition region that connects the initial Batchelor- with the final Taylor-dispersion regime. Such a simple interpretation is shown to be consistent with observations and is able to explain why <italic>t</italic>³ scaling is found only for one specific initial separation outside the inertial range. Moreover, the model incorporates the observed non-local contribution to the dispersion, because it requires only small-time-scale properties and large-scale properties.

DOI： 10.1017/jfm.2021.989

researchmap

Statistics of local Reynolds number in box turbulence: ratio of inertial to viscous forces Reviewed

Yukio Kaneda, Takashi Ishihara, Koji Morishita, Mitsuo Yokokawa, Atsuya Uno

Journal of Fluid Mechanics 929 A1-1 - A1-19 2021.12

Publishing type：Research paper (scientific journal) Publisher：Cambridge University Press (CUP)

In high-Reynolds-number turbulence the spatial distribution of velocity fluctuation at small scales is strongly non-uniform. In accordance with the non-uniformity, the distributions of the inertial and viscous forces are also non-uniform. According to direct numerical simulation (DNS) of forced turbulence of an incompressible fluid obeying the Navier–Stokes equation in a periodic box at the Taylor microscale Reynolds number <inline-formula>
<alternatives>
<inline-graphic xmlns:xlink="http://www.w3.org/1999/xlink" mime-subtype="png" xlink:href="S0022112021008065_inline1.png" />
<tex-math>$R_\lambda \approx 1100$</tex-math>
</alternatives>
</inline-formula>, the average <inline-formula>
<alternatives>
<inline-graphic xmlns:xlink="http://www.w3.org/1999/xlink" mime-subtype="png" xlink:href="S0022112021008065_inline2.png" />
<tex-math>$\langle R_{loc}\rangle$</tex-math>
</alternatives>
</inline-formula> over the space of the ‘local Reynolds number’ <inline-formula>
<alternatives>
<inline-graphic xmlns:xlink="http://www.w3.org/1999/xlink" mime-subtype="png" xlink:href="S0022112021008065_inline3.png" />
<tex-math>$R_ {loc}$</tex-math>
</alternatives>
</inline-formula>, which is defined as the ratio of inertial to viscous forces at each point in the flow, is much smaller than the conventional ‘Reynolds number’ given by <inline-formula>
<alternatives>
<inline-graphic xmlns:xlink="http://www.w3.org/1999/xlink" mime-subtype="png" xlink:href="S0022112021008065_inline4.png" />
<tex-math>$Re \equiv UL/\nu$</tex-math>
</alternatives>
</inline-formula>, where <inline-formula>
<alternatives>
<inline-graphic xmlns:xlink="http://www.w3.org/1999/xlink" mime-subtype="png" xlink:href="S0022112021008065_inline5.png" />
<tex-math>$U$</tex-math>
</alternatives>
</inline-formula> and <inline-formula>
<alternatives>
<inline-graphic xmlns:xlink="http://www.w3.org/1999/xlink" mime-subtype="png" xlink:href="S0022112021008065_inline6.png" />
<tex-math>$L$</tex-math>
</alternatives>
</inline-formula> are the characteristic velocity and length of the energy-containing eddies, and <inline-formula>
<alternatives>
<inline-graphic xmlns:xlink="http://www.w3.org/1999/xlink" mime-subtype="png" xlink:href="S0022112021008065_inline7.png" />
<tex-math>$\nu$</tex-math>
</alternatives>
</inline-formula> is the kinematic viscosity. While both conditional averages of the inertial and viscous forces for a given squared vorticity <inline-formula>
<alternatives>
<inline-graphic xmlns:xlink="http://www.w3.org/1999/xlink" mime-subtype="png" xlink:href="S0022112021008065_inline8.png" />
<tex-math>$\omega ^{2}$</tex-math>
</alternatives>
</inline-formula> increase with <inline-formula>
<alternatives>
<inline-graphic xmlns:xlink="http://www.w3.org/1999/xlink" mime-subtype="png" xlink:href="S0022112021008065_inline9.png" />
<tex-math>$\omega ^{2}$</tex-math>
</alternatives>
</inline-formula> at large <inline-formula>
<alternatives>
<inline-graphic xmlns:xlink="http://www.w3.org/1999/xlink" mime-subtype="png" xlink:href="S0022112021008065_inline10.png" />
<tex-math>$\omega ^{2}$</tex-math>
</alternatives>
</inline-formula>, the conditional average of <inline-formula>
<alternatives>
<inline-graphic xmlns:xlink="http://www.w3.org/1999/xlink" mime-subtype="png" xlink:href="S0022112021008065_inline11.png" />
<tex-math>$R_ {loc}$</tex-math>
</alternatives>
</inline-formula> is almost independent of <inline-formula>
<alternatives>
<inline-graphic xmlns:xlink="http://www.w3.org/1999/xlink" mime-subtype="png" xlink:href="S0022112021008065_inline12.png" />
<tex-math>$\omega ^{2}$</tex-math>
</alternatives>
</inline-formula>. A comparison of the DNS field with a random structureless velocity field suggests that the increase in the conditional average of <inline-formula>
<alternatives>
<inline-graphic xmlns:xlink="http://www.w3.org/1999/xlink" mime-subtype="png" xlink:href="S0022112021008065_inline13.png" />
<tex-math>$R_ {loc}$</tex-math>
</alternatives>
</inline-formula> with <inline-formula>
<alternatives>
<inline-graphic xmlns:xlink="http://www.w3.org/1999/xlink" mime-subtype="png" xlink:href="S0022112021008065_inline14.png" />
<tex-math>$\omega ^{2}$</tex-math>
</alternatives>
</inline-formula> at large <inline-formula>
<alternatives>
<inline-graphic xmlns:xlink="http://www.w3.org/1999/xlink" mime-subtype="png" xlink:href="S0022112021008065_inline15.png" />
<tex-math>$\omega ^{2}$</tex-math>
</alternatives>
</inline-formula> is suppressed by the Navier–Stokes dynamics. Something similar is also true for the conditional averages for a given local energy dissipation rate per unit mass. Certain features of intermittency effects such as that on the <inline-formula>
<alternatives>
<inline-graphic xmlns:xlink="http://www.w3.org/1999/xlink" mime-subtype="png" xlink:href="S0022112021008065_inline16.png" />
<tex-math>$Re$</tex-math>
</alternatives>
</inline-formula> dependence of <inline-formula>
<alternatives>
<inline-graphic xmlns:xlink="http://www.w3.org/1999/xlink" mime-subtype="png" xlink:href="S0022112021008065_inline17.png" />
<tex-math>$\langle R_{loc}\rangle$</tex-math>
</alternatives>
</inline-formula> are explained by a multi-fractal model by Dubrulle (<italic>J. Fluid Mech.</italic>, vol. 867, 2019, P1).

DOI： 10.1017/jfm.2021.806

researchmap

Effects of the Compressibility of Turbulence on the Dust Coagulation Process in Protoplanetary Disks Reviewed

Yoshiki Sakurai, Takashi Ishihara, Hitomi Furuya, Masayuki Umemura, Kenji Shiraishi

The Astrophysical Journal 911 ( 2 ) 140 - 140 2021.4

Authorship：Last author Language：English Publishing type：Research paper (scientific journal) Publisher：American Astronomical Society

Planetesimals are believed to be formed by the coagulation of dust grains in the protoplanetary disk turbulence. However, the bouncing and fragmentation barriers have not been completely solved, particularly for silicate dust. To circumvent these barriers, the turbulent clustering of dust particles must be properly treated. According to the minimum-mass solar nebula (MMSN) model, the Mach number of the turbulence ranges from Mrms ; 0.01–0.32, and thus the turbulence is often regarded as essentially incompressible. However, it has not been quantitatively investigated whether the incompressible limit is adequate for protoplanetary disk simulations. We therefore compare in this study the motions of inertial particles in direct numerical simulations (DNSs) of the Navier–Stokes equation between weakly compressible turbulence and incompressible turbulence. In the DNSs of compressible turbulence, we use an external force to set the total dissipation and the dilatational-to-solenoidal dissipation ratio. The DNSs reveal that despite the small Mach number Mrms(≾ 0.3), the compressible turbulence field notably differs from the incompressible field in terms of the density fluctuations, pressure fluctuations, and shocklet generation, depending on the ratio of the dilatational forcing. However, we quantitatively confirmed that these effects on the particle collision statistics are weak and that the motion of inertial particles in weakly compressible turbulence is dominated by the solenoidal velocity components. Therefore we can conclude that the incompressible assumption is appropriate for an investigation of the dust coagulation process in protoplanetary disk turbulence, as assumed in the MMSN model.

DOI： 10.3847/1538-4357/abe9ba

Scopus

researchmap

Other Link： https://iopscience.iop.org/article/10.3847/1538-4357/abe9ba/pdf

Thin Shear Layers in High-resolution Direct Numerical Simulations of Turbulence

Takashi Ishihara, Koji Morishita, J. C.R. Hunt

Notes on Numerical Fluid Mechanics and Multidisciplinary Design 147 517 - 526 2021

Publishing type：Part of collection (book)

Recent studies based on high-resolution direct numerical simulations (DNSs) of turbulence have shown that there are three different types of shear layers in homogeneous and inhomogeneous turbulent flows; at the outer edge (T/NT), in the interior (T/T) and within the buffer layer near the wall (T/W). All the layers play important roles in various turbulence phenomena in the fields of natural sciences and engineering applications. Data analyses showed that all the shear layers act as a barrier of the velocity fluctuations by blocking velocity fluctuations from the one side to the other side. It was suggested that the blocking mechanism of thin turbulent wall (T/W) layers can be used to control the turbulent wake of aerofoils. For high Reynolds number, flows over typical aerofoils at low angle of incidence in the thin turbulent boundary layers (TBLs) have a conditional structure with thin T/W layers which act as a barrier to eddies in the outer part of the TBL. Recent studies indicate that using disturbance—devices at the airfoil surface enhances the blocking effect of the TW barrier and leads—unexpectedly-to reduced shear fluctuations at the wall. This suggests that deeper understanding of the structures and properties of the T/W layers may provide ideas to consider the better strategies for aerofoils designs. In this paper, we review the recent studies on the three different types of shear layers, which are based on the high-resolution DNSs of homogeneous isotropic turbulence (HIT), turbulent boundary layer (TBL), and turbulent channel flow (TCF).

DOI： 10.1007/978-3-030-55594-8_43

Scopus

researchmap

Optimizations of DNS Codes for Turbulence on SX-Aurora TSUBASA

Yujiro Takenaka, Mitsuo Yokokawa, Takashi Ishihara, Kazuhiko Komatsu, Hiroaki Kobayashi

Sustained Simulation Performance 2019 and 2020 51 - 59 2021

Drastic Changes of Turbulence in the Ignition Process of an n-Heptane/Air Mixture

Takashi Ishihara, Ryousuke Kuno

Lecture Notes in Mechanical Engineering 205 - 209 2021

Authorship：Lead author Publishing type：Research paper (international conference proceedings)

Two-dimensional turbulent combustion with rapid temperature and pressure growth in the ignition process of a homogeneous n-heptane/air mixture is studied by high-resolution direct numerical simulations (DNSs) of the flow field of the mixture with up to 20482 grid points. The DNSs show that the turbulent flow mixes the temperature field so that it affects on the ignition delay. The DNSs show also that the rapid temperature and pressure growth due to the heat release at ignition makes drastic changes in the turbulent velocity field, especially in its compressible component, and causes pressure oscillations in some cases.

DOI： 10.1007/978-981-33-4960-5_32

Scopus

researchmap

Extreme dissipation and intermittency in turbulence at very high Reynolds numbers Reviewed

Gerrit E. Elsinga, Takashi Ishihara, Julian C. R. Hunt

Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences 476 ( 2243 ) 20200591 - 20200591 2020.11

Language：English Publishing type：Research paper (scientific journal) Publisher：The Royal Society

Extreme dissipation events in turbulent flows are rare, but they can be orders of magnitude stronger than the mean dissipation rate. Despite its importance in many small-scale physical processes, there is presently no accurate theory or model for predicting the extrema as a function of the Reynolds number. Here, we introduce a new model for the dissipation probability density function (PDF) based on the concept of significant shear layers, which are thin regions of elevated local mean dissipation. At very high Reynolds numbers, these significant shear layers develop layered substructures. The flow domain is divided into the different layer regions and a background region, each with their own PDF of dissipation. The volume-weighted regional PDFs are combined to obtain the overall PDF, which is subsequently used to determine the dissipation variance and maximum. The model yields Reynolds number scalings for the dissipation maximum and variance, which are in agreement with the available data. Moreover, the power law scaling exponent is found to increase gradually with the Reynolds numbers, which is also consistent with the data. The increasing exponent is shown to have profound implications for turbulence at atmospheric and astrophysical Reynolds numbers. The present results strongly suggest that intermittent significant shear layer structures are key to understanding and quantifying the dissipation extremes, and, more generally, extreme velocity gradients.

DOI： 10.1098/rspa.2020.0591

Scopus

researchmap

Other Link： https://royalsocietypublishing.org/doi/full-xml/10.1098/rspa.2020.0591

Second-order velocity structure functions in direct numerical simulations of turbulence with Rλ up to 2250 Reviewed

Takashi Ishihara, Yukio Kaneda, Koji Morishita, Mitsuo Yokokawa, Atsuya Uno

Physical Review Fluids 5 ( 10 ) 104608-1 - 104608-15 2020.10

Authorship：Lead author Language：English Publishing type：Research paper (scientific journal) Publisher：American Physical Society (APS)

We report results of a series of high-resolution direct numerical simulations (DNSs) of forced incompressible isotropic turbulence with the number of grid points and the Taylor scale Reynolds number Rλ up to 122883 and ∼2250, respectively. The DNSs show that there exists a scaling range (approximately 100<r/η<400), at which the second-order two-point velocity structure functions S2(r) fit well with a simple power-law, S2(r)/(ϵr)2/3=C2(r/L0)ζ, where r is the distance between the two points, η is the Kolmogorov length scale, ϵ is the mean rate of energy dissipation per unit mass, and L0 is the integral length scale. The exponent ζ is constant independent from Rλ. However, the coefficient C2 is dependent of Rλ or the viscosity. This implies that the power-law scaling range of 100<r/η<400 for Rλ up to ∼2250 is not the so-called "inertial subrange"in the sense that the statistics in the range are independent from the viscosity, as assumed in various turbulence theories. This suggests that the constancy of the scaling exponent of a structure function within a certain range does not necessarily mean that the exponent is the scaling exponent in "the inertial subrange."

DOI： 10.1103/physrevfluids.5.104608

Scopus

researchmap

Other Link： http://harvest.aps.org/v2/journals/articles/10.1103/PhysRevFluids.5.104608/fulltext

Thin shear layers in high Reynolds number turbulence - DNS results

T. Ishihara, J. C.R. Hunt, Y. Kaneda

ETC 2013 - 14th European Turbulence Conference 2020

Publishing type：Research paper (international conference proceedings)

Data analysis of high resolution DNS of isotropic turbulence with the Taylor micro-scale Reynolds number Rλ = 1131 on 40963 grid points shows that there are significant thin shear layers of thickness of the order of Taylor micro-scale, consisting of a cluster of strong vortex tubes with typical diameter of order 10η (η is the Kolmogorov length scale). There are big velocity jumps of the order of the rms of the fluctuation velocity across the layers. The locally averaged energy dissipation rate in the layers is much greater than the mean energy dissipation rate, and contributes disproportionately to the overall mean dissipation rate. Up scale and down scale energy transfer to small eddies is largest within and just outside these layers (where the most intense vortices and dissipation occur). The statistics (based on a systematic study) of the layers will be presented and their Reynolds number dependence will be discussed.

Scopus

researchmap

Thin shear layers in high Reynolds number turbulence - A coherent-structure model

J. C.R. Hunt, T. Ishihara, Y. Kaneda

ETC 2013 - 14th European Turbulence Conference 2020

Publishing type：Research paper (international conference proceedings)

Direct numerical simulations of homogeneous isotropic turbulence at the Taylor microscale Reynolds number of order 103 show that the most energetic small scale motions and places with the most intense dissipation in the homogeneous isotropic turbulence are highly inhomogeneous and are located inside thin shear layers. We propose a local mechanism for how the most intense small scale and dissipative motions are produced in the vicinity of well separated thin shear layers. A model is developed based on a local rapid distortion analysis and the Townsend-Obukhov energy transfer hypothesis which is consistent with the data of the numerical simulation.

Scopus

researchmap

Dynamic geometrical analysis of high-enstrophy structures in isotropic turbulence

Yuji Hattori, Takashi Ishihara

ETC 2013 - 14th European Turbulence Conference 2020

Publishing type：Research paper (international conference proceedings)

The time evolution of fine-scale tube-like vortical structures in isotropic turbulence is investigated by dynamic geometrical analysis. Our method successfully captures series of events like deformation, creation, extinction, merging and fission of the vortical structures. It is found that extinction and fission are more frequent than creation and merging, respectively.

Scopus

researchmap

Estimation of turbulence-development by a multifractal theory

Toshihico Arimitsu, Naoko Arimitsu, Kohei Takechi, Yukio Kaneda, Takashi Ishihara

ETC 2013 - 14th European Turbulence Conference 2020

Publishing type：Research paper (international conference proceedings)

It is shown for the first time how the degree of development in turbulence, i.e., if it is fully developed or if not how much developed, can be estimated by the multifractal theory called multifractal probability density function theory (MPDFT) [1]. Precise and self-consistent analyses of 40963 DNS turbulence [2] by means of MPDFT reveal that the turbulence with Rλ = 675 is in the fully developed turbulent state, but that the one with Rλ = 1132 is not fully developed yet. The degree of turbulence-development of the latter is estimated.

Scopus

researchmap

Performance of a Two-Path Aliasing Free Calculation of a Spectral DNS Code Reviewed

Mitsuo Yokokawa, Koji Morishita, Takashi Ishihara, Atsuya Uno, Yukio Kaneda

ICCS 2019. Lecture Notes in Computer Science 11539 587 - 595 2019.6

Publishing type：Research paper (international conference proceedings)

A direct numerical simulation (DNS) code was developed for solving incompressible homogeneous isotropic turbulence with high Reynolds numbers in a periodic box using the Fourier spectral method. The code was parallelized using the Message Passing Interface and OpenMP with two-directional domain decomposition and optimized on the K computer. High resolution DNSs with up to grid points were performed on the K computer using the code. Efficiencies of 3.84%, 3.14%, and 2.24% peak performance were obtained in double precision DNSs with grid points, respectively. In addition, a two-path alias-free procedure is proposed and clarified its effectiveness for some number of parallel processes.

DOI： 10.1007/978-3-030-22747-0_44

Scopus

researchmap

　 PREV - NEXT 　

< ISHIHARA Takashi >

Papers - ISHIHARA Takashi