論文

The assessment of turbulence models including a second-moment closure and the standard k-ε model has been performed for calculating a turbulent separated flow over a surface mounted obstacle. The results indicate more complex characteristics in this problem compared to backward facing step flow which is often used as the test case of turbulence models. Although the required computational effort is much higher than that of the k-ε model, the superior performance of the second-moment closure model is evident in representing the velocity field immediately above the obstacle as well as the pressure variation along the channel. It is indicated that the shortcomings of the k-ε model are attributable to the erroneous behaviour of the production rate of turbulent kinetic energy as calculated according to the eddy viscosity assumption. © 1992, The Japan Society of Mechanical Engineers. All rights reserved.

Simultaneous measurement of the fluctuating velocity and pressure is performed combining a miniature staticpressure probe and a time-resolved PIV aiming at further improvement in evaluation of the velocity-pressure correlation. Two static-pressure probes were placed in the measurement domain of the time-resolved PIV and the instantaneous velocity field and the fluctuating velocity are measured simultaneously. The instantaneous pressure filed is estimated from the measured velocity field by numerically solving a discrete Poisson equation for pressure. The values of the estimated pressure are corrected based on the pressure values which are directly measured by the static-pressure probe.

鋸歯形状の壁面を有する流路における伝熱特性と速度場の関係について実験と数値解析により流路幅の影響などについて論じた。(7頁)

周期的外乱によるはく離乱流の制御効果について、速度場のレーザ計測結果から導かれる運動方程式の収支を基に議論した。(7頁)

A novel experimental technique to measure static pressure fluctuation was applied to evaluate velocity-pressure correlation in a turbulent mixing layer. The developing region of the mixing layer was found to be out of equilibrium state, indicating the oscillatory velocity fluctuation due to vortex shedding and remarkable production of turbulence. A simple model for pressure diffusion by Lumley gave satisfactory approximation when the flow approached to the self-similar state. The contribution of pressure-related terms in the transport equation of Reynolds stress is found to be well captured by the present method. The analogy between pressure diffusion and turbulent diffusion becomes weaker when the flow exhibits characteristics which are not found in equilibrium state.

衝突噴流が回転円盤上に衝突して発達する放射状壁面噴流の速度場計測をLDVにより実行し、円盤の回転による遠心力と境界層のねじれが乱流構造に及ぼす影響について明らかにした。

Flow inside a hard-disk drive (HDD) is investigated using a simplified 3.5″ HDD model with shroud opening and read-and-write arm (RWA). The model is designed to serve as a benchmark of HDD flow study both in experiments and in numerical simulations. The complex flow behavior in the narrow disk-to-disk space is focused with the RWA inserted into the space. The velocity statistics are acquired at totally 8 different locations using a velocity profile sensor with a spatial slot resolution of 30 μm along the axial direction. The measurement was carried out at the disk Reynolds number Red=2. 4×105 which corresponds to the rotation speed of 7700 rpm at a real 3.5″ HDD for desktop computers. The resulting mean velocities and Reynolds stress components are influenced by the shroud opening and especially by the RWA, compared to an axisymmetric model without an RWA. The highly sheared flow between the RWA and adjacent upper/lower disk was successfully captured in the experiment. Copyright © 2011 by ASME.

This research introduces the Duct Water Current Turbine Triple Helix with very low head less than 2m and water current at river or in the ocean, economical ecological use for small hydro power rating between 100 to 1000 kW still represents an unsolved problem. Unlike conventional hydro installation, water current turbine in open flow and generate power from flowing water with almost zero in environmental impact. Developments in water current turbine design are reviewed and some potential advantages of duct or "diffuser augmented" current turbine and extremely low head turbine will be explored. For the output expected from the project is helical turbine with control flow on duct. The research aims to apply the helical turbine inside the duct to find the parameter of the power coefficient against tip speed ratio λ, pitch angle and twist angle y which is also the optimum parameter to design a helical cross flow turbine with duct. Parameters are obtained from numerical simulation and through the experimental result. For ducted turbines the theoretical limit depends on (i) the pressure difference that can be created between duct inlet and outlet, and (ii) the volumetric flow through the duct. © 2010 American Institute of Physics.

Recent advances in the parallelizability of fast N-body algorithms, and the programmability of graphics processing units (GPUs) have opened a new path for particle based simulations. For the simulation of turbulence, vortex methods can now be considered as an interesting alternative to finite difference and spectral methods. The present study focuses on the efficient implementation of the fast multipole method and pseudo-particle method on a cluster of NVIDIA GeForce 8800 GT GPUs, and applies this to a vortex method calculation of homogeneous isotropic turbulence. The results of the present vortex method agree quantitatively with that of the reference calculation using a spectral method. We achieved a maximum speed of 7.48 TFlops using 64 GPUs, and the cost performance was near $9.4/GFlops. The calculation of the present vortex method on 64 GPUs took 4120 s, while the spectral method on 32 CPUs took 4910 s. © 2009 Elsevier B.V. All rights reserved.

分子動力学専用計算機MDGRAPE2を渦法による流体シミュレーションに応用し、その際に発生する問題点について吟味した。衝突渦輪に対して適用したところ、ファンクションテーブルの値域を適切に設定することで、MDGRAPE2のライブラリに含まれる単精度計算による誤差の増幅を抑える手法について述べた。

We present a splitting-free variant of the vorticity redistribution method. Spatial consistency and stability when combined with a time-stepping scheme are proven. We propose a new strategy preventing excessive growth in the number of particles while retaining the order of consistency. The novel concept of small neighbourhoods significantly reduces the method's computational cost. In numerical experiments the method showed second order convergence, one order higher than predicted by the analysis. Compared to the fast multipole code used in the velocity computation, the method is about three times faster.

Copyright © 2018 ASME Power generation of laboratory-scaled marine hydrokinetic (MHK) cross-flow (vertical axis) turbines in counter-rotating configurations was scrutinized both experimentally and numerically. A tabletop experiment, designed around a magnetic hysteresis brake as the speed controller and a Hall-effect sensor as the speed transducer was built to measure the rotor rotational speed and the hydrodynamic torque generated by the turbine blades. A couple of counter-rotating straight-three-bladed vertical-axis turbines were linked through a transmission of spur gears and timing pulleys/belt and coupled to the electronic instrumentation via flexible shaft couplers. A total of 6 experiments in 3 configurations, with various relative distances and phase angles, were conducted in the water channel facility (3.5 m long, 0.30 m wide, and 0.15 m deep) at rotor diameter base Reynolds number of 20,000. The power curve of the counter-rotating turbines (0.068-m rotor diameter) was measured and compared with that of a single turbine of the same size. Experimental results show the tendency of power production enhancement of different counter-rotating configurations. Additionally, the two-dimensional (2D) turbine wakes and blade hydrodynamic interactions were simulated by the shear stress transport k-omega (SST k-omega) model using OpenFOAM. The computational domain included a stationary region and two rotating regions (for the case of counter-rotating turbines) set at constant angular velocities. The interface between the rotating and stationary region was modeled as separated surface boundaries sliding on each other. Velocity, pressure, turbulent kinetic energy, eddy viscosity, and specific dissipation rate field were interpolated between these boundaries.

Copyright © 2018 ASME. The present study proposes a new device for the experiment of self-propelling bodies in the water. As opposed to the studies in the past whose experiments were often carried out in a water channel with a given freestream velocity, the new device allows the model to swim under an actual self-propelling condition. The adopted model mimics the caudal fin of various shapes and made of elastic material, and the self-propelling speed is investigated primarily as a function of the forcing frequency. The influence of the amplitude of forced vibration and the materials of different elasticity is also investigated. The flow field around the model fin has been measured by PIV to characterize the flow pattern produced by the fin-motion.

This study focuses on surveying different turbulence models and dynamic mesh techniques to simulate a marine hydrokinetic (MHK) crossflow turbine at Re ˜ 7,000. While several research projects have shown that studies of MHK devices in low Re flow could still yield interesting and significant results, existing computational fluid dynamic (CFD) simulations were conducted at the chord based Re of 10 ~ 10 . The wake and power production of a laboratory-scaled MHK crossflow turbine were numerically simulated and compared with relevant experimental data. The vertical axis turbine operated in a small flume with 20% blockage ratio and was fabricated by mounting three NACA0012 (2.54 cm chord length) straight blades at a radius of 3.41 cm and 15 pitch angle. Within OpenFOAM environment, blade-resolved models were built with Spalart-Allmaras, k-omega shear stress transport (SST), and k-kl-omega unsteady Reynolds-averaged Navier-Stokes simulation (URANS) in both two and three dimensions. Results from each model were compared with the experimental power measurement and flow field obtained by monoscopic particle image velocimetry (2D PIV). Additionally, four different techniques for moving the solid boundaries (turbine blades) in the unsteady simulation were presented and compared in terms of solution consistency and required computational power. Overset mesh, time-deforming mesh, and moving immersed boundary are all available in this open source environment, beside the common rotating mesh technique, and possess the potential to be applied to a more complicated configuration of turbines. c 5 6 ?

Unsteady and asymmetric behavior of separation of nominally axisymmetric and steady turbulent boundary layer in an axisymmetric expansion is investigated. The test diffusers have the total divergence angle from 8 to 50deg and length-to-diameter ratio of 8. Bulk Reynolds number is ranging between 2.5 × 104and 2.4 × 105. The wall static pressure fluctuation increases monotonically with increasing divergence angle up to 25° and then decreases. The wall static pressure exhibits large scale law frequency fluctuations, occurring almost concurrently in the entire passage. Distributions of both mean velocity and reverse flow fraction in a cross-stream plane show axial-asymmetry of the separated flow, suggesting the rapid growth of circumferential non-uniformity upstream. The velocity measurements employing a split film probe to distinguish the flow directions exhibits intermittent flow reversal with higher velocity as compared with forward flow. The conditional average of wall hot film signals reveals the steep decrease at the incipience of flow reversal.

Effects of periodic perturbation on the turbulent separating flows in a plane two-dimensional asymmetric diffuser have been experimentally investigated. The perturbation was accomplished by alternating injection and suction through a two-dimensional slit located upstream of the expected detachment point. The results indicate that the pressure recovery along the diffuser depends on the frequency as well as the amplitude of the injection/suction flow ; the maximum improvement was achieved at St∼0.03, the Strouhal number St being based on the channel width and channel center velocity at the inlet of the diffuser. The phase-averaged velocity distributions at each phase of the oscillation were evaluated by means of synchronized sampling of the velocity signal employing direction-sensible hot-wire anemometry. A wavy pattern of the instantaneous streamlines, obtained from the velocity profiles, extends throughout the entire region of the diffuser, confirming an enhanced momentum transport due to the organized oscillatory fluid motion. © 1993, The Japan Society of Mechanical Engineers. All rights reserved.

旋回失速の予兆現象を現わす壁面の微小な圧力変動を感知し、それをDSPでデジタル処理した信号にしてフィードバックすることではく離の制御効果を得ることができた。(6頁)

グリッドフリー渦法による乱流のシミュレーションに際し、従来粘性拡散に用いられるコアスプレッディングモデルではスペクトルが正しく表せないが、MPSラプラシアンモデルの導入によりDNSとの良い一致を得ることができた。