Turbulent flow simulation of liquid jet emanating from pressure-swirl atomizer

The performances of three linear eddy viscosity models (LEVM) and one algebraic Reynolds stress model (ARSM) for the simulation of turbulent flow inside and outside pressure-swirl atomizer are evaluated by comparing the interface position with available experimental data and by comparing the turbulence intensity profiles at the atomizer exit. It is found that the turbulence models investigated exhibit zonal behaviors, i.e. none of the models investigated performs well throughout the entire flow field. The turbulence intensity has a significant influence on the global characteristics of the flow field. The turbulence models with better predictions of the turbulence intensity, such as Gatski-Speziale’s ARSM model, can yield better predictions of the global characteristics of the flow field, e.g. the reattachment lengths for the backward-facing step flow and the sudden expansion pipe flow, or the discharge coefficient, film thickness and the liquid sheet outer surface position for the atomizer flows. The standard k–ε model predicts stronger turbulence intensity as compared to the other models and therefore yields smaller film thickness and larger liquid sheet outer surface position. In average, the ARSM model gives both quantitatively and qualitatively better results as compared to the standard k–ε model and the low Reynolds number models.     

Controlled atomization using a twin-fluid swirl atomizer

This paper presents the results of an experimental study of a twin-fluid internally mixed swirl atomizer. In this type of injectors, atomization is attained by injecting a small amount of air (i.e. of the order of less than 16% of the mass flow rate of liquid) into a liquid stream within the injector and the two-phase air liquid mixture is passed through a swirling passage to impart a swirling motion to the flow. Since most of the energy for atomization is supplied to the liquid by the atomizing air, a significantly small pressure drop can produce very fine spray with a small amount of atomizing air. At low values of air–liquid mass ratio (ALR), the appreciable tangential component of velocity with respect to the axial velocity provides a hollow cone spray structure, which turns into a solid cone spray with the increase in axial momentum, through either an increase in ALR or the liquid supply pressure. The results presented in this paper suggest that the investigated injector could be used to control the flow rate and spray characteristics (e.g. spray cone angle, spray solidity, breaking distance, and the droplet diameter) independent of each other by simultaneously varying the supply pressure of the liquid and the atomizing air flow rate. The controlled atomization studied in this paper for a twin-fluid internally mixed swirl atomizer makes it attractive to be used for various commercial applications as the atomizer is capable of providing various spray characteristics depending upon the application requirement.     

An experimental investigation of swirl atomizer sprays

In our previous studies (Chu et al. in Heat Mass Transf 43(11):1213–1224, 2007), a theoretical model of swirl atomizers was successfully established. From the analysis, the equations for the droplet size, velocity components, the boundary layer thickness and the spray cone angle were deduced based on the fundamental governing equations. The purpose of this study is to further compare the experimental result with the theoretical one already gained by a satisfactory embodiment of series of experiments. The aim is to corroborate the analytical results of the influence of atomizer construction and controlled pressure difference on typical swirl chambers. The results provide the droplet diameter as a function of pressure difference, swirl atomizer geometry, flow rate, spray cone angle. The experimental outputs also show a good confirmation of theoretical results and can also be used for further optimization on existing swirl chambers. Based on the results obtained, an optimization methodology on characteristics of swirl atomizers is proposed with the adjustment of individual design parameter and the matching flow number.     

Experimental investigation of swirl atomizer spray defections

Introducing, distinguishing and correcting the observed defections in swirl-pressurized atomizers, applicable in cruise missiles, are the main purpose of this experimental work. Twenty sets of twelve reversed engineering atomizers infected with various defections which cannot perform correctly were considered. A set of eight original atomizers without any defections is used to not only identify and illuminate the defections, but also to provide and approve of the atomizers’ performance. In this work, over 350 tests performed for both two sets mostly repeated tests to identify the problems and make the corrections to perform as well as the original ones.     

Viscous flow through microfabricated hyperbolic contractions

We study the flow of a Newtonian fluid through microfabricated hyperbolic contractions followed by a sudden expansion, with the aim of investigating the potential of this geometry to serve as an extensional microrheometer. A set of planar converging geometries, with total Hencky strains ranging from 1.0 to 3.7, were fabricated in order to produce a homogeneous extensional flow field within the contraction. The velocity field in various planes of the hyperbolic contraction was quantified by means of microparticle image velocimetry (μPIV) and the pressure drop across the converging geometry was also measured and found to vary approximately linearly with the flow rate. Additionally, an extensive range of numerical calculations were carried out using a finite-volume method to help assess the performance of this geometry as a microfluidic elongational rheometer. The measured velocity fields in the contraction and associated pressure drops compare very well (to within 10%) with the numerical predictions. For the typical dimensions used in the microfluidic devices, the steady viscous flow through the contraction is shown to be three-dimensional and it is demonstrated that regions with nearly constant strain rate can only be achieved using geometries with large total Hencky strains under Hele–Shaw (potential-like) flow conditions.     

Stability of a viscous subsonic swirl flow

The results of calculating the stability of a three-dimensional swirl flow of a viscous heat-conducting gas are presented. The stability characteristics are determined using the linear time-dependent theory of plane-parallel flow stability. The main undisturbed axisymmetric vortex flow was determined numerically using a quasi-cylindrical approximation for the complete set of Navier-Stokes equations. The circulation of the peripheral velocity in the cocurrent flow surrounding the viscous vortex core was assumed to be constant. In analyzing the stability, nonaxisymmetric perturbations in the shape of waves traveling along the vortex axis with both positive and negative wavenumbers were considered; in these two cases the perturbation rotation is either the same or opposite in sense to the rotation in the vortex core. Neutral stability curves are determined for various values of the swirling parameter and the cocurrent flow Mach number. Moscow. Translated from Izvestiya Rossiiskoi Akademii Nauk, Mekhanika Zhidkosti i Gaza, No. 3, pp. 50–59, May–June, 1998.     

Liquid flow in a simplex swirl nozzle

Pressure-swirl nozzles are widely used in applications such as combustion, painting, air-conditioning, and fire suppression. Understanding the effects of nozzle geometry and inlet flow conditions on liquid film thickness, discharge coefficient and spray angle is very important in nozzle design. The nozzle-internal flow is two-phase with a secondary flow which makes its detailed analysis rather complex. In the current work, the flow field inside a pressure-swirl nozzle is studied theoretically. Using the integral momentum method, the growth of the boundary layer from the nozzle entry to the orifice exit is investigated and the velocity through the boundary layer and the main body of the swirling liquid is calculated. A numerical modeling and a series of experiments have also been performed to validate the theoretical results. The effect of various geometrical parameters is studied and results are compared for viscous and inviscid cases. In addition, the condition in which the centrifugal force of the swirling flow overcomes the viscous force and induces an air core is predicted. The theoretical analysis discussed in this paper provides better criteria for the design and the performance analysis of nozzles.     

Viscous flow down a membrane trough

An inclined, gravity driven, open membrane trough is used as a low-cost fluid transport conduit. The membrane shape and the fluid velocity are determined numerically. The optimum opening width for maximum flow is found to be 0.651 of the membrane perimeter.     

Air flow and pressure inside a pressure-swirl spray and their effects on spray development

Air flow and pressure inside a pressure-swirl spray for direct injection (DI) gasoline engines and their effects on spray development have been analyzed at different injector operating conditions. A simulation tool was utilized and the static air pressure at the centerline of the spray was measured to investigate the static pressure and flow structure inside the swirl spray. To investigate the effect of static air pressure on swirl spray development, a liquid film model was applied and the Mie-scattered images were captured. The simulation and experiment showed that recirculation vortex and air pressure drop inside the swirl spray were observable and the air pressure drop was greater at high injection pressure. At high fuel temperature, the air pressure at the nozzle exit showed higher value compared to the atmospheric pressure and then continuously decreased up to few millimeters distance from the nozzle exit. The pressure drop at high fuel temperatures was more than that of atmospheric temperature. This reduced air pressure was recovered to the atmospheric pressure at further downstream. The results from the liquid film model and macroscopic spray images showed that the air pressure started to affect the liquid film trajectory about 3 mm from the nozzle exit and this effect was sustained until the air pressure recovered to the atmospheric pressure. However, the entrained air motion and droplet size have more significant influence on the spray development after the most of the liquid sheet is broken-up and the spray loses its initial momentum.     

An experimental study of fluid flow and heat transfer in decaying swirl through a heated annulus

The mean and turbulent structures of turbulent swirling flow in a heated annulus have been measured. Both forced and free vortex swirling flows were generated, and the outer wall of the test section was heated uniformly. The maximum swirl number was 1.39, Reynolds numbers were up to 200000, and heat input was 10.5 kW. Mean and turbulent velocity components, air and wall temperatures, and wall static pressures were all measured. Hot-film techniques were developed to measure turbulence. From these parameters, the flow and temperature fields, pressure distribution, and heat transfer coefficients were determined. The mechanisms of heat transfer were identified.     

Flow in a simple swirl chamber with and without controlled inlet forcing

Results are presented from a swirl chamber with and without controlled inlet forcing. The controlled inlet forcing is induced using arrays of vortex generators placed along one wall of the swirl chamber inlet duct. Flow visualization results are given, along with surveys of circumferential mean velocity, static pressure, and total pressure, at Reynolds numbers (based on inlet duct characteristics) as high as 8000. The controlled inlet forcing provides means to alter and control: (i) the spacing and number of Görtler vortices across the span of the swirl chamber, (ii) the amount of vortex development at a particular Reynolds number and circumferential location, (iii) the circumferential location and Reynolds number of initial Görtler vortex development, and (iv) the circumferential location and Reynolds number of Görtler vortex breakup into more chaotic flow.     

Viscous fluid flow around a semipermeable particle

The problem of hydrodynamic interaction between a laminar flow of a viscous fluid and a partially permeable spherical particle is formulated and solved analytically. The filtration flow inside the particle is assumed to obey the Darcy law. Expressions for the filtration flow velocity, drag, sedimentation velocity, and stream functions are obtained. The effect of the permeability of the particle on the flow characteristics is studied. Stream functions of the flow are constructed. __________ Translated from Prikladnaya Mekhanika i Tekhnicheskaya Fizika, Vol. 50, No. 4, pp. 48–53, July–August, 2009.     

Viscous flow in a mixed compression intake

Numerical simulation of the flow in a two‐dimensional mixed compression intake is carried out by solving unsteady viscous compressible equations using a stabilized finite element method. The effect of bleed in starting/unstarting of the intake and controlling the buzz instability is investigated in detail. Higher bleed leads to an increase in the ability of the intake to sustain larger back‐pressure for stable operation. The amount of bleed and its location is varied to understand its effect on the performance of the intake. Two kinds of unsteady oscillations are observed: ‘little’ and ‘big’ buzz. The frequency of the both kinds of buzz oscillations is found to be super‐harmonic of the fundamental acoustic frequency of intake modeled as an open‐closed organ pipe. The frequency as well as amplitudes of the big buzz cycles is larger than those of the little buzz. The little‐ and big‐buzz are found to occur for low‐ and high‐subcritical state of the intake and are very similar to Ferri and Dailey criteria, respectively. Buzz is eliminated when relatively high bleed is implemented, both, upstream and downstream of the throat. The effect of rate of change of back‐pressure on the start/unstart of the intake is investigated. Two situations are considered. The first case is that of an intake where the back‐pressure remains below the critical value. It is found that the intake remains started if the change in back‐pressure is gradual. However, it unstarts if the back‐pressure is changed relatively rapidly. The second set of simulations is an attempt to model the situation where the back‐pressure at the exit of the intake exceeds the critical value and a logic is incorporated in the feed back loop of the fuel modulation to start the intake. Low rate of change of pressure is unsuccessful in starting the intake. Relatively high rates result in either a quick starting of the intake or a slow unstart. Copyright © 2010 John Wiley & Sons, Ltd.     

Velocity measurements in a confined swirl driven recirculating flow

Laser-doppler measurements of the three components of mean velocity and their intensities have been made in an axisymmetric swirling isothermal turbulent combustion chamber flow. Swirl is generated by a gas-turbine aerodynamic swirler and the flow is representative of that found in the primary zone of many practical combustors. Two configurations were studied. In the first the fuel injector was removed and a central core jet entered the chamber via the resulting circular hole in the centre of the swirler. In the second case the injector was retained but a circular baffle was located at the exit plane: this latter device being necessary to prevent an inflow — not present for combusting flow — arising at the exit plane. The velocity data is sufficiently detailed to aid the testing and further development of methods for calculating combustion chamber flows.     

Eddy viscosity in decaying swirl flow in a pipe

Prediction of heat transfer coefficient for swirling flows can be made provided the values of the eddy viscosity are available. In the present work the axial and tangential velocity fields are surveyed in a pipe for the determination of eddy viscosity. The data thus obtained were utilised to determine the influence of the axial Reynolds number and swirl number on the eddy viscosity. An empirical relationship is suggested to determine the eddy viscosity as a function of Reynolds number and swirl number.Nomenclature A _T angular momentum, equation (10) - a coefficient, equation (1) - b coefficient, equation (1) - D pipe diameter - f friction factor - F(y) initial condition function, equation (8) - J ₀ Bessel's function of the first kind of order zero - J ₁ Bessel's function of the first kind of order one - R pipe radius - Re Reynolds number, u _av D/ - r radial coordinate - S _n swirl number, equation (6) - (S _n )_in swirl number at the inlet of the test pipe - u axial velocity - u _av mean axial velocity in pipe - W non-dimensional local tangential velocity, w/u _av - w tangential velocity - X non-dimensional axial coordinate, x/D - x axial coordinate - y non-dimensional radial coordinate, r/R - z non-dimensional parameter, 4(1+/)/Re(x/D) - kinematic eddy viscosity - _n eigenvalues, equation (7) - kinematic viscosity - density     

Turbulence measurements in decaying swirl flow in a pipe

Turbulence in decaying swirl flow through a pipe is studied experimentally by using a single rotating inclined hot-wire probe and the results are presented in the form of profiles of the three mean velocities and the six components of the stress tensor. The results indicate that all the nine components are very weakly influenced by the Reynolds number, but strongly depend on the initial swirl. The swirling stream develops two regions of flow which are characterized by the variation of the tangential velocity and the Reynolds stresses. It was also found that the turbulence intensities are higher at the core and with the decay of the swirl their magnitudes reduce markedly at the core while they change slightly near the wall.Nomenclature A, A ₁–A ₄ constants in equations (1–9) - B constant in equation (10) - C constant in equations (1–9) - D pipe diameter - E hot-wire anemometer output voltage - e fluctuating component of anemometer voltage - k yaw parameter - R pipe radius - r radial coordinate - S _n swirl number, 2 ₀ ^R uwr ²dr/R2 ₀ ^R ur ² dr - (S _n )_in swirl number at inlet - T _m maximum moment of tangential velocity about axis, (W×Y)_max - U non-dimensional axial velocity, u/u _av - u, v, w time mean velocities in x, r and directions - u, v, w fluctuating components of velocity in x, r and directions - u _av average axial velocity - V non-dimensional radial velocity, v/u _av - v ₁ instantaneous velocity vector - v _x , v _y , v _z instantaneous velocity components in x, y and z directions - W non-dimensional tangential velocity, w/u _av - W _max maximum tangential velocity at a section - x axial coordinate - Y non-dimensional radius, r/R - Y ⁺ radius at which maximum tangential velocity occurs - probe angle - the angle between the wire axis and the instantaneous velocity vector - azimuthal angle of rotation of the probe about its axis