Detailed Modelling Of Droplet Heating And Evaporation Engineering Essay

Detailed Modelling Of Droplet Heating And Evaporation Engineering Essay

There are two methods that have been developed for multi-dimensional simulation of droplet vaporization, one-fluid method and two-fluid method [ 1 ] . For one-fluid method, a set of regulating equations are solved, while for two-fluid method, the regulating equations for each stages are solved.

In the last decennaries, two-fluid method has been widely used to look into the job of droplet vaporization. The two-fluid method normally assumes that the droplet maintains the stiff spherical form during the procedure of droplet vaporization. Masoudi and Sirignano [ 1997 The influence of an advecting whirl on the heat transportation to a liquid droplet ] [ 2 ] investigated the 3-dimensional interaction between the advecting whirl and an evaporating droplet. The droplet convective heat transportation figure is sensitive to the initial place and construction of the advecting whirl. Furthermore, the perpendicular constructions consequence the mass and heat transportation mechanisms in spray burning systems. The survey by Zhang [ 3 ] used two-fluid theoretical account to imitate a vaporizing fuel droplet in a forced convective environment. Good understanding was observed between numerical anticipations and experimental informations.

There have been a few surveies utilizing one-fluid method for analysis of droplet vaporization. The chief advantage of one-fluid method is it can see the distortion of vaporizing droplet due to interface trailing and interface capturing methods.

In order to take into history the distortion of vaporizing droplet, Deng et Al. [ 4 ] developed a method for droplet kineticss and vaporization utilizing Arbitrary Lagrangian-Eulerian ( ALE ) method. They found that the droplet could be deformed under a little comparative speed between gas stage and liquid stage. The ALE method has a low efficiency and the regrid is needed when a big distortion of the interface occurs [ 1 ] , so it is non widely used.

The Volume-Of-Fluid ( VOF ) method [ 5 ] has been introduced in the surveies of droplet vaporization in recent old ages. The VOF method is one of the interface capturing methods. Nikolopoulos et Al. [ 6 ] presented the VOF attack for the calculation of the vaporization procedure of a liquid droplet encroaching onto a hot substrate utilizing an adaptative local grid polish technique at the liquidi??Cgas interface. They focus on chiefly on the droplet form development during droplet encroachment on a het surface. The computations have been done in both planar axisymmetric and to the full 3-dimensional spheres. The algorithm of evaporated mass flux considers the kinetic theoretical account. Strotos et Al. [ 2008 Numerical probe on the vaporization of droplets lodging on het surfaces at low Weber Numberss ] made a numerical comparing against Ficki??i??s jurisprudence, kinetic theoretical account and conventional hydrodynamic theoretical account for vaporization of droplets encroaching and lodging on heated solid walls utilizing the VOF method. A conjugate theoretical account based on kinetic theoretical account and conventional hydrodynamic theoretical account was formulated for the calculation of droplet vaporization in contact with the hot wall. A method for 3-dimensional direct numerical simulation of vaporizing droplets based on VOF attack was presented by Schlottke and Weigand [ 7, 8 ] . They used the PLIC method to capture the interface of droplet and the vaporization rate was computed based on Ficki??i??s jurisprudence.

This paper is focused on the survey of droplet vaporization procedure utilizing VOF method. First, patterning of droplet warming and vaporization utilizing VOF theoretical account is described. Second, the comparing of simulation consequences of different theoretical accounts is studied. Third, the comparing of simulation consequences between droplets array and a individual droplet is discussed. Fourthly, the effects of different droplets spacing on droplet vaporization is showed. Finally, we compare the droplet distortion in different ambient force per unit areas.

2. Modeling of droplet warming and vaporization utilizing VOF method

The VOF ( Volume of Fluid ) method is proposed on the footing of the MAC method, which can be used to cover with the issue of any free interface. The VOF describes two or more non-miscible fluid ( stage ) . The chief thought of VOF method is: The volume fraction of the fluid is set to a map in a cell and the stages interface is captured through the computation of the map. The stage interface capturing job is transformed into the simulation of flow field conveyance through building the equations of the fluid volume fraction. The equations determine the place of the traveling interface, form and distortion of the fluid. The place and form of the interface is determined by the ratio of volume of fluid and cell.

2.1 Volume of Fluid equation

This survey uses VOF method to look into the droplet vaporization and two stages are set, liquid stage and gas stage. The gas stage includes the vapor and the air.

If the value is 0, it indicated that the cell is full of gas stage. If the value is 1, it indicated that the cell is full of liquid stage. Furthermore, if the value is 0 & lt ; i??i?? & lt ; 1, it indicated that the cell is filled with gas stage and liquid stage. The incline and curvature in the stage interface is calculated by the value of volume fraction of the next cells, thereby the distinct equations can be solved.

The fluid flow must obey the jurisprudence of preservation, including the jurisprudence of preservation of mass, the jurisprudence of preservation of impulse and the jurisprudence of preservation of energy and so on. The basic equation of VOF method is the preservation equation of volume of fluid. The VOF equation has the undermentioned signifier,

Where is the mass beginning for the qth stage. The beginning footings for liquid stage and gas stage are,

For the two-phase cell, the conveyance belongingss are determined by the volume fraction mean method of liquid stage and gas stage. All the belongingss are computed in this mode. For illustration, the denseness is given as,

2.2 Momentum equation

One set of impulse equation for both stages is solved in the computational sphere. The speed field is shared in each stage and the impulse equation is calculated based on the volume fraction of each stage. Due to the consequence of vaporization, the liquid stage lost the impulse and the gas stage derive the impulse [ 9 ] . The impulse equation and beginning term is expressed as,

2.3 Energy equation

The energy equation is shared between the liquid stage and the gas stage. The energy equation is similar to the impulse equation, the beginning term caused by the vaporization can be obtained by the by the vaporization rate and latent heat of vaporization. The radiation effects are ignored. The energy equation and beginning term is expressed as,

2.4 Species conveyance equation

Speciess preservation jurisprudence represents that the clip rate of alteration of the chemical composing in the cell is equal to the net diffusion flux in the interface and the coevals rate due to the chemical reaction. In this survey, because the gas stage is composed of air and n-heptane vapor, the transport equation of the n-heptane vapor is solved. In the liquid stage, the conveyance equation for n-heptane is solved. The species equation and beginning term is given by

2.5 Turbulence equation

Large eddy simulation ( LES ) was foremost proposed by Smagorinsky in 1963 [ 10 ] . In LES theoretical account, big Eddies are resolved by the N-S equations straight. However, little Eddies can non be resolved straight, Subgrid-Scale Models are introduced to pattern the little Eddies. In this survey, Smagorinsky-Lilly Subgrid-Scale Model [ 11 ] is used for the computations of droplet vaporization in forced convection.

2.6 Volumetric mass beginning term

There are a big figure of mathematical theoretical accounts depicting droplet vaporization procedure, such as empirical theoretical accounts [ 12 ] , a pure diffusion theoretical account [ 13 ] and molecular kineticss model [ 14 ] . In this survey, we use pure diffusion controlled theoretical account to carry on the survey of the droplet vaporization procedure. This theoretical account assumes that the vapour-liquid interface is in stage equilibrium province and the vaporization rate is calculated by Fick ‘s jurisprudence. The volumetric mass beginning of vapor is calculated as [ 7 ]

Where is the local interface denseness.

2.7 Saturate vapor force per unit area

At the interface of two stages, the droplet vaporization rate can be calculated utilizing this equation.

There are many methods to cipher the saturate vapor force per unit area, which are based on integral of the Clausius-Clapeyron equation [ 15 ] . This equation presents gas stage and liquid stage is in the position of stage equilibrium ; therefore the temperatures, force per unit areas and chemical potency of two stages are tantamount:

( 16 )

( 17 )

Clapeyron equation used for ciphering saturate vapour force per unit area of fuel droplets:

( 18 )

( 19 )

( 20 )

( 21 )

2.8 Geometric Reconstruction Scheme

PLIC-VOF ( piecewise linear interface building ) method [ 16 ] is 2nd order. The two-phase interface is composed of some consecutive lines utilizing PLIC-VOF. ( Fig. 1 )

Figure 1. Comparison

3. ETC theoretical account

Effective thermic conduction ( ETC ) theoretical account takes into history finite liquid thermic conduction [ 12, 17 ] . The theoretical account considers Hill ‘s type of internal circulation inside the droplets, so it leads to an increasing of the sphere thermic conduction by a rectification factor. The rectification factor was approximated as [ 12 ]

The rectification factor varies over the scope of 1-2.72. The transient heat conductivity equation is described as

The thermic radiation effects are neglected [ 18 ] . where is the liquid thermic diffusivity, is the liquid thermic conduction, is specific heat capacity, is liquid denseness, is the droplet temperature, R is the distance from the Centre of the droplet.

The boundary status of the droplet is presented in the undermentioned signifier.

For the ETC theoretical account, the liquid thermic conduction is replaced by the alleged effectual thermic conduction [ 12 ] :

3. Bipartisan yoke method

The bipartisan yoke simulation is based on a yoke between the external flow calculations with the CFD package OPENFOAM, and an in-house codification work outing the clip dependent Navier-Stokes equations for the internal Fieldss of the droplet [ 19, 20 ] . The bipartisan yoke attack adopts two-fluid method and two sets of conservative equations are solved for liquid stage and gas stage. The energy equation of liquid stage takes the undermentioned signifier:

The effectual heat flux ensuing from the external flow calculation are used as non-uniform boundary conditions at the droplet surface.

The bipartisan yoke method adopts quasi-static vaporization hypothesis.

4. Consequences and treatment

The vaporization of droplets tandem is investigated utilizing ANSYS FLUENT 12.0. The computation of droplets vaporization performed utilizing a planar structured mesh. The computational sphere and boundary conditions of numerical simulation under this survey are sketched in Fig. 2, which is a speed recess status at the left side, free faux pas wall conditions on sidelong boundaries, a outflow status at the right side, a axis symmetricalness status on the axis line. The size of computational sphere is 4.66mm ( 20d ) i??i?? 2.33mm ( 10d ) .

Figure 2. Computational sphere and boundary conditions constellation for the numerical simulation of droplets tandem

Table 1 summarize the values of experimental parametric quantities. A additive monodisperse droplet watercourse is generated by Rayleigh decomposition of a liquid jet with the usage of a mechanical quiver obtained by a piezoceramic [ 21 ] .

Table 1 Valuess of experimental parametric quantities

Parameter Value

Injection temperature 297K

Droplets diameter 233i??i??m

Droplets spacing 4.0d

Ambient temperature 1150K

Droplet speed 10.1m/s

Surface tenseness coefficient 0.02N/m

3.1 Grid independency

The size of the grid has a major impact on the simulation consequences. In order to obtain the sure simulation consequences, it is necessary to transport out the grid independency survey. When grid size lessenings, the simulation consequences will be nearing a bound bit by bit. When the simulation consequences near to the bound, the influence of the grid size weakened quickly, and so the simulation consequences are independent of the grid size. Grid independency survey is the of import preparatory work of CFD.

In order to verify the independency of the grid, three different grids Numberss are set. The grids Numberss of the droplet occupied are 80, 316 and 716 severally. The computational clip of the three sorts of grids is 3.3 hours, 14.6 hours and 24.5 hours utilizing parallel calculating. Figure 3 shows a comparing of the droplets surface temperature of the three sorts of grids. The maximal mistake between harsh grid and all right grid is 0.55 % , the maximal mistake between medium grid and all right mesh is 0.28 % . In add-on, in the computation rhythm, the mean mistake between harsh grid and all right grid is 0.18 % , while the mean mistake between medium grid and all right grid is 0.04 % . As the mistake of the computation consequence additions with the grid size lessenings, the computation consequence is in a gradual convergence. Sing the computation clip and mistakes, the medium grid is selected as the computational grid paper and the size of medium grid is 11.65i??i??m i??i?? 11.65i??i??m.

Figure 3. Comparison of the droplets surface temperature of three sorts of grids

3.2 Comparison of different theoretical accounts

The subdivision is focused on the comparing of the effects of different theoretical accounts on droplet warming. The secret plans of droplet surface temperature versus clip for different theoretical accounts are shown in Fig.4. At the initial phase of droplet vaporization, the consequence of different theoretical accounts on clip development of surface temperature is comparatively little. However, the consequence becomes clearly seeable and the difference in surface temperature predicted by the theoretical accounts can make about 4 % at t = 2ms. The ETC theoretical account predicts lowest droplet surface temperatures compared with other attacks, while the bipartisan yoke attack predicts highest droplet surface temperature.

An account to this may be provided as follows. In the ETC theoretical account, the droplet warming is controlled by the

In the VOF theoretical account,

Figure 4. Comparison of different theoretical accounts on droplets surface temperature

The relation between surface temperature and mean temperature in ETC theoretical account can be presented in the undermentioned [ 22 ] .

where

Figure 5. Comparison of the droplets mean temperature of different theoretical accounts

Figure 6. Comparison of experimental values & A ; simulation values of n-heptane droplet vaporization at 550K

The temperature distributions of liquid stage

Figure 7. Comparison of experimental values & A ; simulation values of n-heptane droplet vaporization at 550K

55 52 48 45 41 38 34 31 27 24 20

Figure 4 shows the temperature distributions and streamline Fieldss of liquid stage. The difference of temperature distribution between lead droplet and in-between droplet or downstream droplet is comparatively big, while the difference of temperature distribution between in-between droplet and downstream droplet is comparatively little.

Figure 8. The temperature distributions and streamline Fieldss of liquid stage at clip =1ms

3.3 Comparison between droplets array and stray droplet

The temporal development of droplets surface temperature, shown in Fig.9, indicates noteworthy differences between the in-between one of droplets tandem and stray droplet. At the really initial phase of droplet warming and vaporization, the values of droplet surface temperature are comparatively insensitive for the interaction of droplets. At intermediate clip, the surface temperature of droplets tandem is lower than the stray droplet. The difference between both anticipations is due to the consequence of the vapour aftermath of the lead droplet. Because the temperature of vapor is lower than that of air and the in-between droplet is at the place of vapour aftermath of lead droplet, the vapors wake affects the surface temperature of the in-between droplet.

Figure 9. Comparison of the droplets surface temperature between droplet tandem and stray droplet

Figure 10. Comparison of the droplets mean temperature between droplet tandem and stray droplet

3.4 The effects of different droplets spacing

For different spacing of the droplets, the downstream droplet is located at different places of the lead droplet vapors wake, hence different droplets spacing will impact the vaporization procedure of downstream droplet. This subdivision will concentrate on this subject and the droplets spacing is from 2d to 8d.

Vorticity magnitude is the magnitude of the vorticity vector. Vorticity is a step of the rotary motion of a unstable component as it moves in the flow field, and is defined as the coil of the speed vector [ FLUENT 12 Eµi??i??i??i?? ] :

It can be seen from Fig.9, the distribution of gas stage vorticity at the forepart sides of lead droplets is same with different droplets spacing. However the distribution of vorticity at the forepart sides of downstream droplets is different with different droplets spacing. At a smaller droplets spacing of 2d, there is a small vorticity at the front side of the downstream droplet. With the addition of droplets spacing, the distribution country of vorticity at front side of downstream droplet go larger. There is a small consequence of distribution of vorticity at front side of downstream droplet with a droplets spacing 6d and 8d. Furthermore, the distribution country of vorticity at front side of downstream droplet is near to that of lead droplet with a droplets spacing 6d and 8d.

Figure 11. Vorticity magnitude of different droplets spacing at

Fig.10 shows the surface temperatures of downstream droplet predicted by different droplets spacing. An obvious addition in surface temperature due to an addition in droplets spacing, is attributed to the places of downstream droplet located. The larger the droplets spacing is, the smaller the consequence of lead droplet is. The surface temperature is lowest with the droplets spacing of 2d while the surface temperatures are about same with the droplets spacing of 6d and 8d. This can be explained as follows. The surface temperature is chiefly controlled by the interface temperature of gas-liquid stage, which is effected by the vapour aftermath of lead droplet. The temperature of vapour aftermath is lower than ambient temperature and the temperature of vapour aftermath is relative to droplets spacing.

Figure 12. Vorticity magnititude of different downstream droplets spacing

The drag coefficient Cadmium for the droplet is determined from the aerodynamic force and a dynamic force per unit area based on a comparative speed. The retarding force force Fd is given by [ 23 ]

Fig. 13 compares the retarding force coefficient b the downstream droplet with isolate droplet. As can be seen from Figure, the different droplet spacing, when the droplet spacing more hours, after the droplets are greater the opposition ; when the droplets the greater the spacing, the smaller the opposition after droplets are.

The retarding force coefficients of different droplets spacing is shown in Fig.13. As follows from this figure,

I?13i??E?i??i??E?i??O?i??i??i??i??i??bO?i??Iµi??i??i??fIµi??i??i??i??I?i??i??i??O?i??i??i??i??i??U?i??I¬i??i??O?i??I?i??a?¬i??i??O?i??I?i??i??O???E±i??i??i??i??O?i??i??i??i??i??Uµi??i??i??i??i??fO?i??o?»¶i??O?i??I?i??i??O?i??i??E±i??i??i??i??O?i??i??i??i??i??Uµi??i??i??i??i??fO???i??i??

Figure 13. Droplet coefficients of the downstream droplets tandem

3.5 Droplets distortion

An of import difference between the droplets and solid atoms is that the droplet will significantly deform in the class of motion. Furthermore, the droplets besides occur in the air flow under rotary motion [ 24 ] . Droplet vaporization, the theoretical theoretical account assumes that the droplet is stiff spherical form and distortion of the droplets is ignored. However, in existent droplet vaporization procedure, the country of vaporizing droplets surface is altering. In the instance of the same volume, the surface country of the spherical droplets is minimum.

The axis length of droplet perpendicular to the airflow way is defined as a, and the axis length of droplet analogue to the airflow way is defined as B ( Fig. 12 ) .

Figure 14. Comparison of experimental values & A ; simulation values of n-heptane droplet vaporization at 550K

The distortion factor of the droplets is defined as follows:

( 4-16 )

The distortion factor characterizes the alteration in form of the vaporizing droplets and reflects the alteration of the droplets shape in different waies. When & gt ; 1, it indicates that the distortion of droplets in the perpendicular to the airflow way becomes larger ; When = 1, it indicates that the droplet doesni??i??t deform and keeps spherical ; when & lt ; 1, it indicates that the distortion of droplets parallel to the airflow way becomes larger.

Fig.13 shows the form of droplets in different ambient force per unit areas at 2 ms. Droplet is well spherical when the ambient force per unit area 1bar. As the ambient force per unit area additions, the eccentricity of the oval becomes larger and the droplet form bit by bit becomes level, which shows the influence of the ambient force per unit area on the droplet form is more obvious.

Figure 15. Comparison of experimental values & A ; simulation values of n-heptane droplet vaporization at 550K

It can be seen from Fig.16, the droplet distortion factors increase with increasing environmental force per unit area. The distortion factors of the droplets vary sporadically. At a higher ambient force per unit area of, the maximal distortion factors of droplets is 2.35. The form of the droplet is comparatively long length spherical from the axial way. However at the ambient force per unit area of, the maximal distortion factors of droplets is 1.15. As follows from this figure, at higher ambient force per unit areas, the distortion factors of the droplets are larger. The ambient force per unit area affects the velocity of the droplet. It can be seen that the greater the ambient force per unit area is, the shorter the clip of the droplet get away the computational sphere is.

Figure 16. Comparison of experimental values & A ; simulation values of n-heptane droplet vaporization at 550K

5. Decisions

Recognitions

[ 1 ] NICHITA B A. An Improved CFD Tool to Simulate Adiabatic and Diabatic Two-phase Flows [ M ] . 2010.

[ 2 ] MASOUDI M, SIRIGNANO W A. Collision of a whirl with a zaping droplet [ J ] . Int J Multiphas Flow, 2000, 26 ( 12 ) : 1925-49.

[ 3 ] ZHANG H. Numerical research on a zaping fuel droplet in a forced convective environment [ J ] . Int J Multiphas Flow, 2004, 30 ( 2 ) : 181-98.

[ 4 ] & lt ; 1979Quasi-steady droplet stage alteration in the presence of convection.pdf & gt ; [ J ] .

[ 5 ] HIRT C W, NICHOLS B D. Volume of fluid ( VOF ) method for the kineticss of free boundaries [ J ] . J Comput Phys, 1981, 39 ( 1 ) : 201-25.

[ 6 ] BERGELES G, NIKOLOPOULOS N, THEODORAKAKOS A. A numerical probe of the vaporization procedure of a liquid droplet encroaching onto a hot substrate [ J ] . Int J Heat Mass Tran, 2007, 50 ( 1-2 ) : 303-19.

[ 7 ] SCHLOTTKE J, WEIGAND B. Direct numerical simulation of vaporizing droplets [ J ] . J Comput Phys, 2008, 227 ( 10 ) : 5215-37.

[ 8 ] SCHLOTTKE J, DULGER E, WEIGAND B. A VOF-based 3D numerical probe of vaporizing, deformed droplets [ J ] . Advancement in Computational Fluid Dynamics, an International Journal, 2009, 9 ( 6 ) : 426-35.

[ 9 ] R B. Turbulent conjugate heat and mass transportation from the surface of a binary mixture of ethanol/iso-octane in a rip stratified two-phase flow system [ J ] . Int J Heat Mass Tran, 2008, 51 ( 25-26 ) : 5958-74.

[ 10 ] SMAGORINSKY J. GENERAL CIRCULATION EXPERIMENTS WITH THE PRIMITIVE EQUATIONS [ J ] . Monthly Weather Review, 1963, 91 ( 3 ) : 99-164.

[ 11 ] LILLY D K. A proposed alteration of the Germano subgrid-scale closing method [ J ] . Physicss of Fluids A: Fluid Dynamics, 1992, 4 ( 3 ) : 633-5.

[ 12 ] ABRAMZON B, SIRIGNANO W. Droplet vaporisation theoretical account for spray burning computations [ J ] . Int J Heat Mass Tran, 1989, 32 ( 9 ) : 1605-18.

[ 13 ] T. KIM M S B, P. V. FARRELL. Evaporating spray concentration measurementsfrom little and medium dullard Diesel injectors [ J ] . SAE, 2002-01-0219,

[ 14 ] SERGEI S S. Advanced theoretical accounts of fuel droplet warming and vaporization [ J ] . Prog Energ Combust, 2006, 32 ( 2 ) : 162-214.

[ 15 ] JING-SHAN T. THERMOPHYSICAL PROPERTIES OF FLUIDS: Principal and Calculation [ M ] . Beijing: China Petrochemical Press, 2008.

[ 16 ] L Y D. Time-Dependent multii??i??mateiral flow with big unstable distotrion [ M ] . New York: AcademicPress, 1982.

[ 17 ] SAZHIN S S, KRISTYADI T, ABDELGHAFFAR W A, et Al. Models for fuel droplet warming and vaporization: Comparative analysis [ J ] . Fuel, 2006, 85 ( 12-13 ) : 1613-30.

[ 18 ] SAZHIN S S, ELWARDANY A, KRUTITSKII P A, et Al. A simplified theoretical account for bi-component droplet warming and vaporization [ J ] . Int J Heat Mass Tran, 2010, 53 ( 21-22 ) : 4495-505.

[ 19 ] FRACKOWIAK B, LAVERGNE G, TROPEA C, et Al. Numeric analysis of the interactions between vaporizing droplets in a monodisperse watercourse [ J ] . Int J Heat Mass Tran, 2010, 53 ( 7i??C8 ) : 1392-401.

[ 20 ] CASTANET G, FRACKOWIAK B, TROPEA C, et Al. Heat convection within vaporizing droplets in strong aerodynamic interactions [ J ] . Int J Heat Mass Tran, 2011, 54 ( 15i??C16 ) : 3267-76.

[ 21 ] CASTANET G, LABERGUE A, LEMOINE F. Internal temperature distributions of interacting and zaping droplets [ J ] . Int J Therm Sci, 2011, 50 ( 7 ) : 1181-90.

[ 22 ] DOMBROVSKY L A, SAZHIN S S. A Parabolic Temperature Profile Model for Heating of Droplets [ J ] . Journal of heat transportation, 2003, 125 ( 3 ) : 535-7.

[ 23 ] SIRIGNANO W A, REVIEWER A C F E. Fluid Dynamics and Transport of Droplets and Sprays [ J ] . Journal of Fluids Engineering, 2000, 122 ( 1 ) : 189-90.

[ 24 ] i??i??A?i??i?? . i??i??E?i??i??i??i??i??E?i??i??N§ [ M ] . i??i??l: i??i??li??i??i??N§i??i??i??i??i?? , 2005.