This paper proposes a validation scheme for the effect of wind tunnel blockage on decaying grid-generated turbulence. This validation scheme was derived from the governing equations of the k-∈ model. Analytical solutions for the validation scheme were derived by introducing a model of the difference between the rate of change of the effect of fluid acceleration on the turbulent kinetic energy and that of the effect on its dissipation. The derived solutions include a decay exponent that excludes the acceleration effect, a parameter characterizing the acceleration, the initial anisotropy, and the model coefficient of the k-∈ model, and can be quantified by parameters which can be known. The physical meaning of the model was clarified. The derived solutions and model were confirmed to be accurate through numerical simulation. An equation for the decay exponent, which is also affected by the fluid acceleration, was developed using the derived solutions. This scheme was applied to the examination of the reduced fluid acceleration effect in a moderate-sized wind tunnel to measure the grid-generated turbulence. The fluid acceleration effect in the wind tunnel was confirmed to be small using the derived equations. The decay characteristics of the grid-generated turbulence in the wind tunnel were measured and were found to agree with those obtained in previous experiments.
A micro size particle behavior considering thermophoretic and drag forces are simulated by using direct simulation Monte Carlo (DSMC) method. The computation time is too high to compute the micro particle movement by conventional DSMC method because the computation time is proportional to a particle diameter. In this paper, the molecule-particle collision model, which computes the collision between a particle and multi molecules in a collision event, is considered. The momentum transfer to the particle is computed with a collision weight factor, where the collision weight factor means the number of molecules colliding with a particle in a collision event. The large time step is adopted by considering the collision weight factor. Therefore, the computation time becomes fifty thousandth times for micro size particle computation theoretically. We simulate the particle motion considering thermophoretic and drag forces by DSMC-Neutrals (Particle-PLUS neutral module) with above molecule-particle collision model, where DSMC-Neutrals is commercial software adopting DSMC method. The thermophoretic velocity with molecule-particle collision model is verified by comparison with Waldmann's model. Furthermore, it is shown that the DSMC method with molecule-particle collision model reproduces completely the conventional DSMC method. The behavior of a particle, which is polystyrene latex (PSL), is simulated.
Effect of the Reynolds number on the torque and power characteristics of a small straight-bladed vertical axis wind turbine has been investigated experimentally under various wind velocity. The maximum mean torque coefficient and the maximum mean power coefficient increase with increasing the Reynolds number based on the wind velocity and representative length of the wind turbine, and the dependence of these coefficients on the Reynolds number can be successfully approximated in the logarithmic function. The tip speed ratio for the maximum mean torque coefficient is almost independent of the Reynolds number. Otherwise, the tip speed ratio for the maximum power coefficient increases as increasing Reynolds number, and the dependence of the maximum mean torque coefficient on the Reynolds number can be approximated in the logarithmic function. When the curvature parameter, the aspect ratio, and the solidity represented forms of the wind turbine are same, the wind turbine performance can be successfully explained by an semi-empirical formula including simple analytical functions, namely, the mean torque and the mean power coefficients can be represented well by the logarithmic functions of the Reynolds number and quadratic or cubic function of the tip speed ratio. The proposed approximate equations successfully predict experimental data for the particularly higher tip speed ratio.
This paper proposes a new riblet configuration. A traditional sinusoidal riblet was modified with the aim of reducing pressure drag. The height of the side wall in the newly configured riblet is lowered toward the node of the sinusoidal curve to reduce the large pressure drag, while the riblet height is maintained at the anti-node position as it has been reported to be the most effective for straight or traditional sinusoidal riblets. In this paper, effective configuration parameters and the drag-reducing performance of the modified sinusoidal riblet are investigated through parametrically conducting direct numerical simulation (DNS). First, the optimal combination of the wavelength and amplitude of the spanwise sinusoidal curve configuration is determined with a fixed riblet spacing s and height h, which are located at around the design point of straight riblets. The effective combination obtained is similar to that reported in previous studies on traditional sinusoidal riblets. Second, the optimal vertical amplitude a of the modified sinusoidal riblet is determined with the effective settings mentioned above. Unfortunately, with a/h < 0.2, the drag-reducing performance is similar to that of traditional sinusoidal riblets. Furthermore, the performance deteriorates with a/h ≥ 0.2. However, the drag-reducing performance of the modified sinusoidal riblet is improved when s+ deviates from the design point compared to that of comparable straight or traditional sinusoidal riblets. This indicates that flow-condition robustness is improved for the modified sinusoidal riblet. Finally, the mechanism of robustness is investigated. The contributions of the bottom and side walls to drag are calculated, and the difference of each contribution between the traditional and modified sinusoidal riblets is shown to discuss the influence caused by the lowered side wall. In addition, a quadrant analysis is conducted to examine the change in momentum transfer above the modified sinusoidal riblet.
Our aim is to understand the complicated flow inside a suction sump in the vertical-wet-pit-pump configuration. The information of unsteady and three-dimensional flow structures is useful, not only for this complicated flow but also for other complicated flows which are commonly seen in various practical applications. To acquire such information, a three-dimensional particle tracking velocimetry (3D-PTV) technique is one solution. However, this technique includes more unknown factors in reliability and accuracy than other well-established measuring techniques. In the present study, we examine the simultaneous measurement using both of the 3D-PTV and another velocimetry, namely, an ultrasonic velocity profiler (UVP) with common tracer particles. This simultaneous measurement is expected to become an effective method to dissolve the above fatal defect of the 3D-PTV. As a result, we have successfully found out the suitable condition for the simultaneous measurement with high accuracy. Then, under this condition, we have revealed the time-mean and instantaneous velocity vectors of the unsteady and three-dimensional flow inside the suction sump.
We investigate the accuracy of the laminar boundary layer on a flat plate in the simulation by an immersed boundary-lattice Boltzmann method (IB-LBM). In this study, we use the single relaxation time-lattice Boltzmann method combined with the multi direct forcing method, which can enforce the no-slip boundary condition accurately by determining the body force iteratively. The simulations of the laminar boundary layer on a flat plate at the Reynolds number of 1000 are performed by using the IB-LBM, and it is found that in order to obtain a reasonably accurate result such that the error from the solution of the boundary layer equations is within 5%, the boundary layer has to be resolved by about 50 lattice spacings. In addition, it is found that the IB-LBM has the same accuracy whether the flat plate is coincident with the lattice or not. In order to improve the accuracy of the boundary layer calculated by the IB-LBM, we discuss the effective thickness of the boundary caused by the body force distributed near the boundary. Also, we find that the accuracy is much improved by using a finer lattice only around the flat plate.
The moving particle semi-implicit (MPS) method has been used in a wide range of industrial fields with the free surface. Many machines use oil for lubrication as well as cooling. In this study, stirred fluid flow of various oil types with rotating cam-shafts was calculated by using the MPS method. Moreover, experiments were conducted in order to validate the calculation results by comparing the torque values. When the oil viscosity is 30 cs or lower than 30 cs, the calculated torques agree well with those of the experiment. When the oil viscosity is larger than 30 cs, the calculated torques are lower than those of the experiment. This is due to the involvement of small air bubbles. The air bubbles are retained for a long time when the oil viscosity is high, which increases the oil surface and eventually causes a higher torque. This phenomenon is not calculated because the air bubbles are not modeled in this study. In the case when the oil viscosity is 30 cs, the calculation results of the torque values were in good agreement with those of the experimental results. The calculation results show convergence with respect to the particle size.
The extended Bernoulli equation is formulated in an exact form for a microscopic and small Reynolds number Jeffery-Hamel flow in a two-dimensional convergent or divergent channel. The friction loss and the friction coefficient derived from the extended Bernoulli equation are also obtained for the purpose of engineering applications. The assumption of microscopic and low Reynolds number flow enables us to make the analysis simple, and the results obtained are expressed in forms easy to use. The zeroth- and first-order approximate solutions of velocity distribution in the channel are obtained by solving the nonlinear ordinary differential equation with the optimal homotopy asymptotic method. The zeroth-order solution is shown to be the same function form as that in the two-dimensional parallel flow, i.e., the two-dimensional Poiseuille flow. The extended Bernoulli equation, the friction loss, and the friction coefficient in a finite region of the channel, which are indispensable for applications, are reasonably derived along a stream line and also expressed by cross-sectional average quantities. The cross-sectional average formulae of the friction loss and the friction coefficient are expressed by the geometry of the channel, i.e., the convergent or divergent angle, the channel length, channel widths at inlet and exit. These formulae include the corresponding well-known ones for the two-dimensional parallel flow as a special case where the angle is zero. The friction coefficient drastically increases according to the increase in the angle, especially in a narrow channel region, and attains more than ten times of the friction coefficient for the parallel flow.