Performance characteristics of an axial flow fan having distorted inlet flow have been investigated using numerical analysis. Two kinds of hub-cap, rounded and right-angled front shape, are tested to investigate the effect of inlet flow distortion on the fan performance. In case of right-angled front shape, axisymmetric distorted inflow is induced by flow separation at the sharp edge of hub-cap, and the characteristics of the inflow depend on the distance between hub-cap and blade leading edge. Three-dimensional Reynolds-averaged Navier-Stokes equations are introduced to analyze the flow characteristics inside the blade passage. Numerical solutions are validated in comparison with experimental data measured by a five-hole probe downstream of the fan rotor. It is found from the numerical results that non-uniform axial inlet velocity profile near the hub results in the change of inlet flow angle. Large recirculation flow upstream the fan rotor for the right-angled hub-cap induces separated flow on the blade surfaces near the hub region, and thus deteriorates the performance of fan rotor. The effect of the distance between hub-cap and blade leading edge on the efficiency is also discussed.
Large-eddy simulations (LES) are applied to particle-laden swirling jets, and effect of the swirl on particle dispersion is investigated. The trajectories of all particles are individually pursued with a Lagrangian method. The particles with different diameters are uniformly injected into a non-swirling flow or swirling flows with different swirling numbers. The results show that the trajectories of the particles largely differ depending on their diameters. In swirling jets, the peak of particle number density for small particle is located on the central axis, whereas that for larger particles is shifted outward by the centrifugal force. However, the larger particles, which migrate outward in the upstream region, tend to gradually migrate inward toward downstream, and this trend is remarkable as the particle diameter decreases. This is due to the fact that the direction of the particle migration in the downstream region is dominated by the turbulent motions, which act to transport the smaller particles inward in wide range of the swirling jets.
The effects of nozzle geometry on cavitation in the nozzle of pressure atomizers and the liquid jet are examined using various two-dimensional (2D) nozzles with different geometries. Then, whether or not the conventional cavitation numbers can be used to predict the formation of supercavitation, in which liquid jet atomization is enhanced, is examined. As a result, we confirm that (1) the thickness of the cavitation zone increases with the ratio Cu of the cross-sectional area upstream of the nozzle to that of the nozzle, (2) the spray angle increases with Cu, (3) the formation of supercavitation can be predicted using the cavitation number σc' in which the effects of the flow contraction and the frictional pressure drop are taken into account, and (4) the conventional cavitation numbers σ, σ2 and σ3 cannot predict the formation of supercavitation in nozzles with different geometries.
Cavitation in a cylindrical nozzle and liquid jets discharged from the nozzle are simultaneously visualized using a high-speed camera to investigate the mechanism of liquid jet atomization induced by the fully-developed cavitation. Three mirrors are used to capture the front view of liquid jet interfaces and the side view of cavitation clouds within a frame of the camera. The high-speed visualization confirms that the collapse of a cavitation cloud near the nozzle exit induces a ligament formation not only for 2D nozzles but also for cylindrical nozzles. The visualization also finds that the collapse of a large cavitation cloud tends to cause the formation of a large ligament, while that of a small cloud results in a small ligament.
In order to measure unsteady flow rate of the order of less than 1μl/sec, a new flowmeter consisting of a capillary and a pressure gauge has been developed. When a target flow passes through the capillary, the measured pressure loss in the capillary gives the flow rate according to the Hagen-Poiseuille equation, which indicates the flow rate is proportional to the pressure loss. Investigating prototype flowmeter characteristics in the case of steady flow of water, we confirmed that the flow rate given by the Hagen-Poiseuille equation derived from the measured pressure loss accords with the flow rate estimated by the gravimetric technique. The accuracy is within ± 1 %. When the flow rate decreases gradually, the measured flow rate accords with the theoretical value at each moment. This flow meter enables to measure time varying flow rate. In the case that the flow rate abruptly changes, the measured pressure has a time delay because of the property of the pressure gauge used in this study. Compensation for this time delay is demonstrated.
Particles in a liquid under standing ultrasonic waves have been known to aggregate. However, particle aggregation behavior remains unclear. Thus, ultrasonic waves horizontally irradiated particles in tap water or degassed water with a relatively large disk-type acoustic transducer. We observed the particle behavior and measured the sound pressure profiles. The following results were obtained. The behavior of particles in water under ultrasonic waves was classified as “band”, “point”, “particle clump”, and “non-aggregation”. Experimental conditions producing “band”, “point”, “particle clump”, and “non-aggregation” in tap water were found to be different from those in degassed water. Moreover, the point aggregations at a frequency f of 96.3 kHz were observed at many more locations (higher spatial density) than those at a frequency f of 23 kHz. The sound pressure profile for f = 96.3 kHz had many more peaks than that for f = 23 kHz in the vertical direction, which corresponds to the spatial densities of the point aggregation.
Dissolution of single carbon dioxide (CO2) bubbles in a vertical pipe of 25 mm diameter is measured to examine the effects of the ratio λ of the sphere-volume equivalent bubble diameter to the pipe diameter, the liquid Reynolds number and surfactants on mass transfer. The bubble diameter and liquid Reynolds number are varied from 5.0 to 26 mm (0.20 < λ < 1.0) and from 0 to 3100, respectively. Millipore water, tap water or water contaminated with Triton X-100 are used for the liquid phase. Dissolution processes are measured at atmospheric pressure and room temperature. Mass transfer coefficients and Sherwood numbers are evaluated from measured bubble diameters. Complicated capillary waves are formed at the clean bubble surface, whereas there are no capillary waves at the contaminated bubble surface. The disappearance of capillary wave results in the retardation of surface renewal, and therefore, Sherwood number decreases with increasing surfactant concentration. Empirical correlations of Sherwood numbers for bubbles rising in clean and contaminated liquids in a vertical pipe are proposed. The correlations are applicable not only to bubbles in stagnant liquid but also to bubbles in pipe flow, provided that the liquid Reynolds number is not so high.
Visualization of a brain neural pathway is a useful tool for supporting surgical planning. In this paper, we propose a method for visualizing a brain neural pathway by using both critical points and target regions in order to extract necessary pathways. Here, critical points are those points at which the magnitude of the vector vanishes, and we assume that a neural pathway consists of a set of streamlines in a vortical flow field and we find the start points of streamlines by classifying critical points. The user also specifies target regions, such as regions of interest, and visualizes the important streamlines that pass through the target regions.
The development of waves on a fluid-fluid interface, excited by a vertical, relative motion of a solid wall enclosing the fluids, is affected significantly by the mobility of the interface on the wall. The effective, cycle-averaged mobility depends on the amplitudes of excitation and waves, because of a non-linear dependence of the velocity of the fluid-fluid-wall contact line on the angle of the interface hitting the wall. At higher amplitudes of excitation and waves, moreover, the interface motion is tied to the low mobility of the contact line only for a limited fraction of each cycle; for the rest of the cycle the interface is virtually untied from the wall, being connected to the contact line only through the surface of a thin, flat liquid film left on the wall. An analytical model is developed on the wall surface boundary conditions for the interface profile, in terms of non-linear relationship between the interface velocity along the wall and the near-wall inclination angle of interface. The model is based on optical data taken in a quasi-static experiment where measurements of near-wall interface configuration were essentially unaffected by wave generation on the interface. Numerical simulation with this model reproduces successfully the changes in the near-wall configuration of the interface in the experiment, including development and depletion of a liquid film on the wall associated with relative motion between the contact line and the interface elevation away from the wall.
The analytical model developed in the previous paper is applied to excitation of an axisymmetric fundamental wave on a fluid-fluid interface enclosed by a vertical cylinder. The excitation is caused by a forced vertical motion of the fluids relative to the stationary solid wall. Model analyses reproduced experimental results on the interface wave amplitude and phase relationship to the forced excitation. It is found that the deformation of the interface, that is primary mechanism exciting the wave, takes place only in a part of each cycle. In this time period the interface is hinged directly to the fluid-fluid-wall contact line of low mobility and hence the near-wall interface profile is subjected to deformation as a result of the forced fluid motion relative to the wall. For the rest of the cycle the interface is virtually untied from the wall due to the presence of a liquid film between the moving interface and the contact line, which can stretch or shrink without imposing additional forces on the moving interface. The fraction of the time period when the interface is hinged on the wall without intervention by the liquid film decreases as the excitation amplitude is increased, or the wave amplitude increases. The effective, cycle-averaged mobility of the interface, dependent on this time fraction, affects the natural frequency of the interface as well as the efficiency of wave excitation.