Numerical Simulation on Rupture Behavior and Selection of Safety Membranes

This present study deals with a new mechanical device consisting of a set of safety membranes, which has been successfully applied in several middle and small hydropower stations in China instead of a surge tank. Safety membranes are installed on the penstock near the powerhouse as controlled weak points. When the pressure caused by load rejection rises to the preset explosive value, one or more membranes rupture, protecting the penstock and the unit from damage. The device is simple, reliable and economical. The method of characteristics is employed to establish numerical model of safety membranes to simulate their rupture behavior, which is then introduced to investigate how to determine the number and diameter of membranes from two aspects, large fluctuation and hydraulic disturbance. The results show that the diameter of the membranes depends on the negative pressure along pipeline under hydraulic disturbance while the number of the membrane depends on the maximum water hammer pressure under large fluctuation during load rejection of the unit. The conclusion of membrane selection can perfect the present theory of safety membranes, and provide the theoretical guidance and practical basis for membrane device design and safety operation.

Comparison of Sharp Interface and Smoothed Profile Methods for Laminar Flow Analysis Over Stationary and Moving Boundaries

This study is a comparison of two techniques for simulation of particulate flows on fixed Cartesian grids: Sharp interface Method (SIM) (Udaykumar et al., 2001, 2002, 2003) and a modified version of Immersed Boundary Method (Peskin, 1977) (IBM) known as Smoothed Profile Method (SPM) (Nakayama and Yamamoto, 2005; Luo et. al, 2009). Different cases were studied includes flow over one or two moving and stationary particles. Predictions of the drag coefficient shows that SPM and SIM are very close to the experiments. SIM slightly under-predicts the value of the drag coefficient while SPM has a small over-estimation. Moreover, SPM is more accurate on coarse grids. However, with refinement of the grid SIM approaches the exact values very fast leading to better results on fine grids. Flow pattern and vortex structures of SPM and SIM are almost the same. Both methods are capable of analyzing the wake flow. Unlike SIM, SPM is able to simulate the flow when two particles are in contact. When two particles are in motion and are very close in a way that the two interfaces overlap, SPM shows a repulsion force between two spheres which reduces the accuracy in comparison with SIM. However, SPM can achieve the collision of two particles without problem.

Numerical Simulations of Particle-Laden Flow Based on Given Friction Reynolds Number and Mean Reynolds Number Respectively

Multiphase flow with particles covers a wide spectrum of flow conditions in natural world and industrial applications. The experiments and the direct numerical simulation have become the most popular means to study the dilute particle-laden flow in the last two decades. In the experimental study, the mean Reynolds number is often adjusted to the value of single-phase flow for each set of particle conditions. However, the friction Reynolds number usually keeps invariable in the direct numerical simulation of the particle-laden flows for convenience. In this study the effect of the difference between given mean Reynolds number and friction Reynolds number was investigated. Two simulations were performed for each set of particle parameters, and the mean Reynolds number and friction Reynolds number were kept invariant respectively. From the results it can be found that the turbulence intensity and the dimensionless velocities are larger when keeping the friction Reynolds constant. And the results calculated from the cases of keeping the mean Reynolds number invariable agree with the experiment results better. In addition, the particle distribution along the wall-normal coordinate was found to be unchanged between two simulation conditions. As a suggestion, keeping the same mean Reynolds number in the direct numerical simulation of particle-laden flow is more appropriate.

Bubble Entrapment in Sprayed Films

A compressed air sprayer was used to spray model paint onto two glass substrates at the same time. Afterwards, one glass substrate was placed on a LED light source and still photographs were taken from the top using a DSLR camera with a timer system. The other substrate was put on a balance to record weight. Pictures and weight measurements were taken at 5 second intervals for 15 minutes. The sprayed film thickness was varied. The pictures were analyzed using ImageJ software. Bubble Count vs. Time, Sauter Mean Diameter (SMD) of Bubbles vs. Time as well as Weight vs. Time was plotted. It was seen that the pace of weight loss was faster for thinner films. The rate of bubble escape also depended on film thickness. It took a longer time for thicker films to lose the bubbles entrapped in them. In the first 30 seconds, large bubbles escaped due to buoyancy forces and afterwards surface-tension driven flows became dominant. There was also a lot of bubble movement in thicker films. The effect of gravity was studied as well. Gravity did not affect the bubble escape rate since a downward facing film had the same bubble count as an upward facing film confirming that bubble motion was not due to buoyancy forces alone. However, the SMD of bubbles in a downward facing film was larger than an upward facing film. Buoyancy is not a factor in bubble escape from the downward facing film and only surface-tension driven flows play a role.

Effect of Hole Pitch Ratio and Compound Angle on the Thermal Flow Fields of Double-Jet Film Cooling Hole

In this study, the effect of a row of double-jet film-cooling hole configurations on the thermal-flow characteristics of gas turbine blades was examined. To investigate the effect of the interference of anti-kidney vortices, the ratios of the pitch distance and hole diameter (P/d=5, 6.25, 8.333) were considered with two different compound angles (λ=0°, 4°). The film cooling performance and the generated losses were studied. Then, the relevant mechanisms were identified and explained. A numerical study was performed using ANSYS CFX 14.5 with the shear stress transport (SST) turbulent model. The blowing ratio was kept at a constant value of M=1.5. The film cooling effectiveness and temperature distribution are graphically depicted with various geometrical configurations.

Viscous Droplet Interaction With Micro-Textured Solid Surfaces

The dynamics of viscous droplets near solid surfaces, especially micro-textured surfaces, and the interaction between them are of great importance in industrial applications, biochemical processes, and fundamental materials research on surface wettability. In this work, a three-dimensional spectral boundary element method has been employed to investigate the dynamics of a viscous droplet falling under gravity influence to micro-textured solid surfaces. The droplet size, in this study, is comparable to the size of the surface texture. The influences of the Bond number, relative size of the droplet with respect to the surface features, and the topological characteristics of the substrate on the droplet motion and deformation are investigated. The stress exerted on the substrate due to droplet motion is also explored.

Three Dimensional Simulation of Blood Flow in the Small Arteries of a Truncated Vascular System With Three Levels of Bifurcation

Three dimensional blood flow in a truncated vascular system is investigated numerically using a commercially available finite element analysis and simulation software. The vascular system considered in this study has three levels of symmetric bifurcation. Geometric parameters for daughter vessels, such as their diameters and their angles of bifurcation, are specified according to Murray’s law based on the principle of minimum work. The ratio of blood vessel length to diameter is based upon experimental data found in the literature. An experimentally obtained velocity profile, available in the literature, is used as the inlet boundary condition. An outflow boundary model, consisting of a contraction tube to represent the pressure drop of the small arteries, arterioles, and capillaries that would follow the truncated vascular system, is used to specify the boundary condition at the eight outlets. The results show that although the blood flow velocity experiences a sudden decrease after the bifurcation points due to the higher total cross-sectional area of the daughter vessels as compared to the parent vessel, this decrease in velocity is partially recovered due to the tapering of the blood vessels as they approach the next bifurcation point. The results also show that the secondary flow which is typical after the bifurcation of large arteries does not develop after the bifurcation of small arteries due to the presence of laminar blood flow with very low Reynolds number in the small arteries. The numerical model yields pressure distributions and pressure drops along the vascular system that agree quite well with the physiological data found in the literature. Finally, the results show that, immediately following a bifurcation, the blood flow velocity profile is not symmetrical about the longitudinal axes of blood vessel. However, symmetry is recovered as the blood flow proceeds down the vessel.

Design and Experimental Study on the Surge Performance Improvement by the Effects of Bleed Slot Casing of a Centrifugal Compressor

The primary design goal of a compressor is focused on improving efficiency. Secondary objective is to widen the compressor’s operating range. This paper presents a numerical and experimental investigation of the influence of the bleed slot to enlarge operating range for the 1.2MW class centrifugal compressor installed in a turbocharger. The main design parameters of the bleed slot casing are upstream slot position, inlet pipe slope, downstream slot position and width. The DOE (design of experiment) method was carried out to optimize the casing design. Numerical analyses were done by the commercial code ANSYS-CFX based on the three dimensional Reynolds-averaged Navier-Stokes equations. From the analysis, as the downstream slot position and width are smaller and upstream position is located away from impeller inlet, efficiency and pressure ratio are increased. Experimental works were done with and without the bleed slot casing. The simulation results were in good agreement with the test data. In case without the bleed slot casing, the surge margin value came out to be only 11.8% but with the optimized bleed slot design, the surge margin reached 23%. Therefore, the surge margin increase of 11.2% was achieved.

The Effect of Viscosity Ratio on the Hydrodynamics of Separation From an Oil-Coated Particle

Separating oil from solid particles is of great importance in many industrial processes including the extraction of bitumen from oil sands, and the remediation of oil spills. The usual approach is to separate the oil from the solid by introducing another liquid (e.g. water). Separation is often assisted by fluid mixing, and chemical addition. Yet while oil-water-particle separation has been well studied from a chemical standpoint, little research has taken into account the effect of hydrodynamics on separation. In this work, the separation of oil from a single oil-coated spherical particle falling through an aqueous solution was evaluated as a function of viscosity ratio. Solvents were used to modify the viscosity of the oil. The experiments were recorded using a high-speed camera and post-processed using the MATLAB image-processing toolbox. A CFD model has also been developed to study this phenomenon. The results indicate that when viscous forces are strong enough, the oil film deforms, flows to the back of the sphere, and forms a tail that eventually breaks up into a series of droplets due to a capillary wave instability. When the viscosity ratio is small (i.e. the oil is less viscous than the solution), a thin tail forms quickly, the growth rate of the instability is high, and hence the tail breaks very quickly into smaller droplets. When the viscosity ratio is high (i.e. the oil is more viscous), more time is required for the deformation/separation to initiate, and the tail is thicker and breaks more slowly into larger droplets. It was observed that when the viscosity ratio is close to 1, the rate of separation is increased and maximum separation is achieved.

CFD Modelling of Bubbly Flow in Adiabatic Upward Pipe Using a Solver Based on OpenFOAM® With the Quadrature Method of Moments

An Eulerian-Eulerian approach was used to model adiabatic bubbly flow with CFD techniques. The OpenFOAM® solver twoPhaseEulerFoam was modified to predict upward bubbly flow in vertical pipes. Interfacial force and bubble induced turbulence models are studied and implemented. The population balance equation included in the two-fluid model is solved to simulate a polydisperse flow with the quadrature method of moments approximation. Two-phase flow experiments with different superficial velocities of gas and water at different temperatures are used to validate the solver. Radial distributions of void fraction, air and water velocities, Sauter mean diameter and turbulence intensity are compared with the computational results. The computational results agree well with the experiments showing the capability of the solver to predict two-phase flow characteristics.