Professional Documents
Culture Documents
A R T I C LE I N FO A B S T R A C T
Keywords: This article investigates the formation of cone-jet structure in an electrospray process based on a two-phase
Cone-jet mode numerical simulation. The numerical approach takes account of the coupled governing equations of fluid flow
Electrospray and electrostatics in conjunction with the charge conservation equation and a VOF interface tracking method on
Electric potential the basis of a CSF model. The temporal and spatial evolutions of the cone-jet mode are examined in connection
Flow rate
with the operating parameters, i.e. liquid flow rate and electric potential. Under the influence of these para-
Numerical simulation
meters, this study elucidates the physical aspects of the geometrical growth and extension along with the electric
charge dispersion within the cone-jet structure. Furthermore, the flow patterns developed in the two-phase flow
are studied revealing how orderly the operating parameters can alter the flow configuration. The results are
compared with experimental data indicating good agreements, which, in turn, confirm the effectiveness of the
simulation methodology concerning the electrospray phenomenon.
2ɛ 0 ⎝ de ⎠ (1)
deposition (Salata, 2005; Jaworek and Sobczyk, 2008; Yoon et al.,
2011; Sweet et al., 2014) and electrohydrodynamic printing (Park where de and L, respectively represent the emitter diameter and the
et al., 2007). The functioning of an electrospray process is dependent on emitter to ground electrode distance, γ is the surface tension coefficient
various factors. These factors would be divided into four groups and ɛ0 denotes vacuum permittivity equal to 8.854 × 10−12 CV−1m−1.
⁎
Corresponding author.
E-mail address: m.jahannama@isrc.ac.ir (M.R. Jahannama).
https://doi.org/10.1016/j.ijheatfluidflow.2018.02.012
Received 1 November 2017; Received in revised form 2 February 2018; Accepted 22 February 2018
Available online 20 March 2018
0142-727X/ © 2018 Elsevier Inc. All rights reserved.
H. Dastourani et al. International Journal of Heat and Fluid Flow 70 (2018) 315–335
In addition, the minimum liquid flow rate would be estimated using the with a pointed apex. This also led to a profounder study on the elec-
following relation (Rosell-Llompart and De La Mora, 1994); trospray physics by Gañán-Calvo (1997) who described the transition
γ ɛr ɛ 0 between Taylor cone and jet regions proposing asymptotic universal
Qmin = scaling laws for both the jet size and the issued electric currents.
ρK (2)
The first simulations of electrically charged jets were inspired by
where ρ, ɛr and K are the density, relative permittivity and electrical numerical models on uncharged liquid jets (Jeong and Moffatt, 1992;
conductivity of liquid, respectively. Eggers and Dupont, 1994; Brenner et al., 1997). Hartman et al. (1999b)
The research work on electrospray can be divided into experimental developed a physical model to simulate the electrospray cone-jet mode
and theoretical studies. The experimental work of Zeleny (1914, 1917) based on a steady state one-dimensional axial momentum equation.
would be acknowledged as the first systematic study on the electrospray This equation was established over a balance among the hydrodynamic
whereon the following research studies to-date are based and evolved potential and kinetic sources of energy, tangential electric stress and the
(Taylor, 1969; De La Mora and Loscertales, 1994; Gañán-Calvo et al., dissipation viscous stress, which could ultimately determine the cone-
1997; López-Herrera et al., 2004; Yu et al., 2016). In contrast, the jet shape. Although they also proposed another model using a La-
theoretical study of Taylor (1964) can be distinguished as a leading grangian approach to predict drop dynamics (Hartman et al., 1999a),
methodical attempt that founded a robust basis for the subsequent the model was not an extension of their cone-jet model and solely relied
analytical and numerical investigations until present. on the experimental data as the input information.
Although the succeeding theoretical efforts in the electrospray in- Yan et al. (2003) simulated the formation of a liquid meniscus and
itially focused on analytical methods, particularly oriented towards the the arising liquid jet due to an electric field in the cone-jet mode. This
instability of liquid jets (Chaudhary and Redekopp, 1980; Setiawan and model would be considered as an extension of the work of Hartman
Heister, 1997; Cherney, 1999; Hartman et al., 2000), this was the nu- et al. (1999b) to an axisymmetric two-dimensional model based on
merical simulation which has drawn the attention of research studies employing the Navier–Stokes equations and the Gauss law (except for
during the recent decades mainly owing to the tremendous advance- the liquid-gas interface). The interface was dealt with by a current
ments achieved in computing facilities. Nevertheless, it seems that the balance to accommodate the jump in the normal electric field.
first numerical simulations on electrically charged liquids can be Lastow and Balachandran (2006) employed the commercial code
tracked back to three decades ago with a main focus on the formation of CFX to numerically model the cone-jet mode using the Navier–Stokes
a stable liquid meniscus from which no jet was emerged. In fact, this equations in conjunction with the Laplace equation. The simulation did
viewpoint would be thought as a zero flow rate limit for the cone-jet not include the parts of current and conductivity in the governing
mode that does not encounter the singularity at the cone apex due to equations implying the insert of a dielectric body in an electric field
the liquid jet emanation. In this connection, Joffre et al. (1982) initiated with no charge flow.
an axisymmetric equilibrium approach based on a balance among the The aforementioned models did not comprise the liquid jet breakup
forces arising from the surface tension, hydrostatic pressure and electric into drops and, thus, appeared to require subsequent extensions with a
field, which could determine the liquid shape profile. They further particular focus on intricacies of the liquid free surface. These goals
extended the model to include the corona discharge from the meniscus were pursued by taking account of various numerical schemes devel-
surface inserting the space charge effects in the meniscus formation oped for the multiphase flows to scrutinize the moving interfaces be-
process (Joffre and Cloupeau, 1986). Following on from these works, tween the different fluid phases (Puckett et al., 1997; Tryggvason et al.,
Pantano et al. (1994) employed the same electrohydrostatic equili- 2001). In this respect, Lim et al. (2011) simulated the cone-jet mode
brium strategy by taking account of a small perturbation analysis to involved with the jet breakup and drop formation based on a two
overcome the mathematical singularity in the liquid meniscus profile
316
H. Dastourani et al. International Journal of Heat and Fluid Flow 70 (2018) 315–335
dimensional axisymmetric approach. They utilized a front tracking needs solution of the governing equations on fluid flow (conservation of
method to treat the liquid-air interface further to assuming a constant mass and momentum) along with the liquid-gas interface tracking. The
surface charge density on the whole liquid-air interface. The interfacial mass conservation equation for an incompressible fluid flow is given by
charge density was chosen on the basis of a trial-and-error method by (Lim et al., 2011);
fitting the numerical data to experimental results to attain the best
suitability. This assumption highly simplified the charge transport ∇→u =0 (3)
phenomenon on the interface degrading the accuracy of the electric →
where u is the fluid field velocity vector. The Navier–Stokes equations
model established. ⎯→
⎯
including the forces of surface tension on the liquid-gas interface ( FST ),
Herrada et al. (2012) proposed a numerical scheme to simulate a ⎯→
⎯
electrical charging of liquid ( FES ) and gravity can be expressed by
steady axisymmetric electrospray cone-jet mode excluding the jet
(Ouedraogo et al., 2017)
breakup. They implemented a tracking method by setting an equili-
brium condition for tangential and normal stresses at the liquid-air ∂→
u → + ⎯→⎯ ⎯→
⎯
interface. This led to an interface shape function, and in turn, to explicit ρ⎡ +→u ·∇→u ⎤ = −∇P + μ∇2→ u + ρg FES + FST
⎢ ∂t ⎥ (4)
relations of the electric field components in terms of that function. The ⎣ ⎦
electric charge was assumed to be confined to the liquid free surface →
where P is the pressure, g the gravity vector, while ρ and μ denote the
implying a zero space charge density, which notably simplified the ⎯→
⎯ ⎯→
⎯
fluid density and viscosity, respectively. Moreover, FES and FST re-
computations. The results were verified in comparison with an open present the electric and surface tension forces, respectively. The electric
source code, i.e. GERRIS, which was developed and extended by López- force in Eq. (4) can be given as follows (Tomar et al., 2007):
Herrera et al. (2011) to simulate electrohydrodynamic processes using
Volume of Fluid (VOF) tracking method. Furthermore, Ferrera et al. ⎯→
⎯ ⎯→
⎯ 1 ⎯→
⎯ 2
FES = ρe E − ( E ) ∇ɛ
(2013) employed the similar approach to simulate the dynamical be- 2 (5)
havior of electrified pendant drops, i.e. a zero flow rate electrospray ⎯→
⎯
where E is the electric field, ρe the volume charge density and ε the
process, using the GERRIS whose results found good agreement com-
permittivity of fluid (ɛ = ɛr ɛ 0 with ɛ0 as vacuum permittivity and ɛr as
pared with experimental data.
Wei et al. (2013) simulated the electrospray cone-jet mode in-
cluding the jet breakup using the OpenFOAM as an open source com-
putational fluid dynamics (CFD) software toolbox. The simulation was
carried out using an unsteady approach towards the fluid flow gov-
erning equations while the liquid-gas interface was treated using the
VOF model. However, the role of electric field was treated using a
leaky-dielectric model, which simplified the charge conservation to a
steady state equation.
Xu et al. (2013) developed a numerical model to simulate the
electrospray cone-jet mode for a core-shell configuration using the
commercial code Fluent. The interfaces of fluids were tracked using the
VOF method while a modified leaky- dielectric model was proposed to
determine the interface charge density. However, the modified model
was based on a tuning factor, which was tailored according to a trial-
and-error procedure in order to provide the best fit with their experi-
mentally observed cone-jet profile. Although this work (within a group
study) was further employed by Davoodi et al. (2015) and Yan et al.
(2016) to examine impacts of the nozzle tip configurations on the flow
formation, no essential development would be realized in the overall
structure of the simulation approach.
The present study is concerned with the numerical simulation of
electrospray process in the cone-jet mode and the subsequent jet
breakup into drops. The study is intended to elucidate the underlying
physics involved with the process via the role of operating parameters.
This is carried out in a transient axisymmetric two dimensional fra-
mework taking simultaneous account of the fluid and electrical coupled
governing equations. In particular, the charge transfer process is sur-
veyed using the full charge conservation equation with no simplifica-
tion.
The study is introduced and discussed in two main parts. In the first
part, the fluid flow, interface tracking and electrical charging are
mathematically formulated and the solution method, computational
domain and grid dependency analyses are described. In the second part,
the results are given and the effects of liquid flow rate and electric
potential on the electrospray structure are investigated further to being
validated.
2. Mathematical formulation
317
H. Dastourani et al. International Journal of Heat and Fluid Flow 70 (2018) 315–335
relative permittivity). The first term on the right hand side of Eq. (5) is Table 1
the Coulomb force, which implies the interaction between the electric Values of geometrical parameters.
charges acting in the direction of the electric field. The second term
di,e (mm) do,e (mm) Led (mm) ddisk (mm)
represents dielectric force that due to ∇ɛ acts perpendicular to the in-
terface. 0.12 0.45 30 24
The surface tension force can be determined using the continuum
surface force (CSF) model introduced by Brackbill et al. (1992). In the
CSF model, the surface tension is presented as a volumetric force that method is used for the interface tracking in this study. The VOF method
based on a constant surface tension coefficient would be stated as fol- is based on a volume fraction C where C = 0 for the cells filled with gas,
lows: 0 < C < 1 for the cells filled with both gas and liquid and C = 1 for the
⎯→
⎯ cells filled with liquid. The volume fraction C is a scalar function whose
→
FST = γκn (6) transport equation in a standard form is as follows (Hirt and Nichols,
→ 1981):
where κ is the curvature of interface and n a normal vector to the in-
terface. According to Eq. (6), the representative force of surface tension ∂C
operates perpendicular to the interface of liquid-gas. In the CSF model, + ∇ ·(→
u C) = 0
∂t (8)
the curvature of the interface may be defined by κ = −∇ ·n ̂ where n ̂ is a
unit normal vector to the interface and is defined by n ̂ = → n/→ n . Thus, Since the mass and volume properties are interchangeable in an
Eq. (6) can be rewritten as; incompressible two- phase flow, this makes the use of volume con-
⎯→
⎯ servation of any fluid as an important advantage in the VOF method
FST = −γ (∇ ·n )̂ →
n (7) (Hirt and Nichols, 1981). If the function of volume fraction C is defined
∼
In the CSF model, the interface between the phases is assumed as a as a characteristic function (C = C ) representing the interface of liquid-
→
very thin layer with the order of computational grid through which gas, the normal vector would be n = ∇C whose substitution in Eq. (7)
physical properties change smoothly from one phase to another. If the leads to;
∼
characteristic function representative of the interface is denoted by C ,
→ ∼
the normal vector to the interface can be defined as n = ∇C .
Table 2
Accordingly, the curvature of interface and, in turn, the surface tension
Physical properties of fluids used in simulations.
force can be calculated. The characteristic function related to the in-
terface tracking method used in the present study is based on the vo- ρ (kgm−3) µ (mPa.s) K (Sm−1) ɛr γ (Nm−1)
lume of fluid (VOF) method.
Heptane 684 0.42 1.4 × 10−6 1.93 0.0186
Air 1.225 0.0183 1.05 × 10−15 1 –
2.2. Interface tracking
Fig. 3. Maximum magnitude of electric field versus minimum cell size on different z lines.
Fig. 2. A schematic of (a) physical domain, (b) computational domain. Fig. 4. Liquid profiles at the emitter exit for various minimum cell sizes.
318
H. Dastourani et al. International Journal of Heat and Fluid Flow 70 (2018) 315–335
⎯→
⎯ ∇C ⎞ ⎤ ρ = ρl + ρg (1 − C )
FST = −γ ⎡∇ ·⎛ ⎜ ∇C
⎟
⎢ ⎝ ∇C ⎠ ⎥ (9) μ = μl + μg (1 − C )
⎣ ⎦
K = Kl + K g (1 − C )
In implementation of the interface tracking method for a two-phase
ɛ = ɛl + ɛ g (1 − C ) (10)
flow, two immiscible fluids are considered as a single effective fluid in
the whole computational domain. Hence, the physical properties of the
effective fluid can be calculated by a weighted average of the volume
2.3. Electrical equations
fraction as given below (Tomar et al., 2007);
⎯→
⎯
To calculate the electric force ( FES ) in Eq. (4), the electric field due
Table 3
Boundary conditions.
P U C ρe Φ
⁎ →
This is a fixed value condition calculated from specified total pressure p0 and local velocity u .
** This is a zero gradient condition when flow is outwards and is a fixed value when flow is inwards.
Fig. 5. Typical patterns of equipotential (color lines) and electric field lines (arrowed black lines) in the computational domain for Q = 3.4 mLh−1 and Φ0 = + 4000 V at t = 1.2 ms.
319
H. Dastourani et al. International Journal of Heat and Fluid Flow 70 (2018) 315–335
Fig. 6. Temporal Cone-jet formation for Φ = + 4000 V and (a) Q = 0.25 mLh−1, (b) Q = 0.50 mLh−1, (c) Q = 0.80 mLh−1, (d) Q = 0.93 mLh−1, (e) Q = 1.20 mLh−1.
320
H. Dastourani et al. International Journal of Heat and Fluid Flow 70 (2018) 315–335
Fig. 7. Temporal Cone-jet formation for Φ = + 4000 V and (a) Q = 2.60 mLh-1, (b) Q = 3.40 mLh-1, (c) Q = 4.90 mLh-1, (d) Q = 6.20 mLh−1, (e) Q = 10.0 mLh−1.
321
H. Dastourani et al. International Journal of Heat and Fluid Flow 70 (2018) 315–335
Fig. 7. (continued)
322
H. Dastourani et al. International Journal of Heat and Fluid Flow 70 (2018) 315–335
accordance with Eq. (4) have been modified. The solution procedure 4. To determine the velocity field in the computational domain, the
used in the current simulation follows the flowchart as shown in Fig. 1 Navier–Stokes equations are solved. It should be noted that the
and has the main features as given below: coupled fields of velocity and pressure are solved using the PIMPLE
algorithm which is a combination of SIMPLE and PISO algorithms.
1. Initial and boundary conditions, e.g. liquid flow rate, electric po- 5. Since the numerical solution is performed in a transient state, the
tential, volume fraction distribution and physical properties, are time step is recalculated based on the maximum Courant number in
applied to the computational domain. each step. The solution continues to achieve a sustainable process.
2. To determine the volume fraction as well as the correction of ef-
fective physical properties, the volume fraction equation (Eq. (16))
is solved. 4. Solution domain and conditions
3. To compute the electric force, Eqs. (12) and (15) are solved.
The solution domain and conditions used in the present simulation
Fig. 8. Temporal distribution of electric charge density for Φ = + 4000 V and (a) Q = 0.25 mLh−1, (b) Q = 0.93 mLh−1, (c) Q = 2.60 mLh−1, (d) Q = 4.90 mLh−1.
323
H. Dastourani et al. International Journal of Heat and Fluid Flow 70 (2018) 315–335
Fig. 9. Distribution of electric charge density for various liquid flow rates and Φ = + 4000 V at t = 1.2 ms.
Fig. 10. Radial distributions of (a) volume fraction and (b) electric charge density in z = 0.03 mm for various liquid flow rates and Φ = + 4000 V at t = 1.2 ms.
324
H. Dastourani et al. International Journal of Heat and Fluid Flow 70 (2018) 315–335
Fig. 11. Cone-jet shapes for various liquid flow rates and Φ = + 4000 V at t = 1.2 ms.
Fig. 12. Variations of acting forces on liquid-gas interface versus liquid flow rate for Φ = + 4000 V at t = 1.2 ms on (a) z = 0.03 mm and (b) z = 0.05 mm.
Fig. 13. Variations of cone-jet dimensions based on various liquid flow rates for Φ = + 4000 V at t = 1.2 ms.
are introduced in this section. The geometrical configurations of elec- Fig. 2 parameters di,e, do,e, Led and ddisk denote emitter inner diameter,
trodes as well as the physical properties employed in this study are emitter outer diameter, emitter to disk distance and disk diameter, re-
corresponding to the experimental work of Tang and Gomez (1996). spectively, and their values are given in Table 1. Working fluids are
Fig. 2 shows schematics of the physical and computational domains. In considered Heptane as the liquid and air as the gas whose physical
325
H. Dastourani et al. International Journal of Heat and Fluid Flow 70 (2018) 315–335
Fig. 14. Temporal variation of cone-jet length for various liquid flow rates at Φ = + 4000 V.
In this part, effects of the liquid flow rate on the electrospray cone-
Fig. 15. Variation of cone-jet length versus liquid flow rate for Φ = + 4000 V at jet formation are studied. Referring to Eqs. (1) and (2), the onset vol-
t = 1.2 ms.
tage and minimum flow rate required for the formation of a cone-jet
mode can be estimated. Using the geometrical and physical properties
properties are given in Table 2. tabulated in Tables 1 and 2, the corresponding onset voltage and
In the present study, the simulations are performed in an axisym- minimum flow rate are found to be +2452 V and 1.20 mLh−1, re-
metric state. The computational domain is discretized using a struc- spectively. In this regard, the simulations associated with the liquid
tured and non-uniform grid so that the minimum grid size is positioned flow rate are performed based on various liquid flow rates at a constant
at the emitter exit. To examine the grid independency on the solution electric potential +4000 V.
data, seven different mesh sizes with the minimum cell size 1, 1.5, 2, Fig. 6 depicts a temporal view of the electrospray process for the
2.5, 3, 5 and 10 µm are tested, respectively. The corresponding simu- liquid flow rates smaller than and equal to the estimated minimum flow
lations are performed for a flow rate of 12 mLh−1 and an electric po- rate. According to the figure, the liquid flow emerged from the emitter
tential of + 4000 V. The maximum electric field versus the minimum steadily being affected by the electric field, initially shaping a meniscus
cell size is plotted in Fig. 3 for different z lines. Moreover, liquid profiles progressing towards a convex conical shape from whose apex fine
in the emitter exit for the various minimum cell sizes are shown in droplets are ejected. More increase in the flow rate converts the me-
Fig. 4. According to these figures, the mesh with the minimum cell size niscus to a concave conical shape in addition to stretching it out.
of 2 µm is chosen for the following simulations. For the selected mesh However, the electrospray process is short of producing a full cone-jet
spacing, the total number of grids is 735,050. mode at the flow rates lower than 0.93 mLh−1 implying insufficiency of
The boundary conditions with a reference to Fig. 2, (b) are treated these flow rates for the jet emergence. Since the evolution of electro-
as presented in Table 3. spray process attains a cone-jet structure at 0.93 mLh−1, this reveals a
Since the simulations are carried out in a transient situation, all lower limit than the estimated minimum flow rate. This would lie in the
dependent variables, i.e. velocity, pressure, electric potential and derivative conditions of Eq. (1), which arguably covers the liquids with
charge density, are set to zero at the starting time of liquid injection K > 10−4 Sm−1 (Chen and Pui, 1997). This should be noted that the
into the domain. At this time, the emitter is assumed to be fully filled main goal of using Eq. (1) in this study is to gain an estimate of the
with the liquid, i.e. C = 1, whereas rest of the domain is filled with the minimum flow rate in order to quantitatively tune the parameter
air, i.e. C = 0. around the cone-jet point formation. In this regard, the results confirm
appropriateness of the estimation predicting a right magnitude of order
5. Results and discussion for the formation flow rate.
Fig. 7 shows the temporal process of cone-jet formation for the flow
The simulation results of electrospray process are presented in this rates larger than the minimum flow rate required for the cone-jet for-
section. It is worth mentioning that the present study is built upon a mation. Based on this figure, the evolution of liquid flow, in addition to
numerous number of simulation cases of which a proportionate number comprising the main features of the process in Fig. 6, directs towards
of pertinent cases is reported. the jet emergence from the cone apex by stretching it down to the point
326
H. Dastourani et al. International Journal of Heat and Fluid Flow 70 (2018) 315–335
Fig. 18. Total surface charge density on cone-jet interface versus flow rate at t = 1.2 ms.
Fig. 17. Velocity of liquid-gas interface versus cone-jet length for various liquid flow rates Fig. 19. A typical pattern of Streamlines formed in whole physical domain for
and Φ = + 4000 V at t = 1.2 ms. Q = 6.20 mLh−1 and Φ = + 4000 V at t = 1.2 ms.
327
H. Dastourani et al. International Journal of Heat and Fluid Flow 70 (2018) 315–335
Fig. 20. Patterns of streamlines formed in the emitter exit for various liquid flow rates and Φ = + 4000 V at t = 1.2 ms.
surface tension and electrostatic forces. In fact, the increase in the li- leading to a higher cone-jet length as drawn in Figs. 14 and 15. This
quid flow rate originated from the rising trend of the hydrodynamic implies that an increase in the flow rate by inflicting a larger amount of
pressure weakens the surface tension and electrostatic effects ex- liquid to a constant electric potential necessitates more liquid flow
panding the cone structure and, in turn, leading to the straight cone development to reach the electrohydodynamic instability required for
generator at the larger flow rates. It should be noted that the hydro-
dynamic pressure mainly acts as a creating force for the fluid motion
and, thus, the cone expansion is a lateral effect of this force with an
order of magnitude higher than the equal orders of surface tension and
electrostatic forces. In addition, the increase in the flow rates lower
than the minimum flow rate largely suppresses the surface tension ef-
fects causing this force to approach to the electrostatic force. This im-
plies a twofold role for the liquid flow rate including that its rise at
small magnitudes results in the formation of the cone-jet structure and
at large magnitudes in the expansion of that structure.
Fig. 13, corresponding to Fig. 11, shows the variation of the cone-jet
radius versus the cone-jet length providing a quantitative view into the
cone-jet dimensions. It is obvious that the increase in the liquid flow
rate in addition to thickening the cone-jet radius elongates the jet by Fig. 21. Displacement of vortex center versus liquid flow rate for Φ = + 4000 V at
transferring its instability, i.e. surface undulations, further downstream t = 1.2 ms.
328
H. Dastourani et al. International Journal of Heat and Fluid Flow 70 (2018) 315–335
Fig. 22. Radial variation of axial velocity on a sectional plane passing through the vortex center for various flow rates and Φ = + 4000 V at t = 1.2 ms.
the jet disintegration. contracting and shifting it further downstream as also quantitatively is
Fig. 16 shows the variation of electric charge density along the li- shown in Fig. 21. Moreover, Fig. 22 shows the radial variation of axial
quid-gas interface using various liquid flow rates for a constant electric velocity on a plane passing through the center of a center. According to
potential at t = 1.2 ms. As seen in the figure, the electric charge density this figure, the shrinkage of a vortex at the larger flow rates is accom-
exhibits an ascending-descending trend along the cone-jet interface on panied with the higher velocities at the liquid-gas interface since the
which the maximum charge density coincides with the cone-jet neck vortex takes place at a far-off axial section compared with a lower flow
having the highest concentration of electric charge. Fig. 17 corre- rate.
sponding to Fig. 16 depicts the velocity of the liquid-gas interface along Fig. 23 shows the distribution of Sauter mean diameter (D32) of
the cone-jet structure. The figure shows that an increase in the flow rate droplets against the liquid flow rate for the present simulation in
leads to a reduction in the interface velocity. This is associated with the comparison with the work of Tang and Gomez (1996). As seen in the
lower levels of surface charge density at the larger flow rates, as shown
in Fig. 18, which in turn decelerates the flow owing to the electrostatic
effect.
Fig. 19 illustrates a typical pattern of streamlines formed in the
whole physical domain while Fig. 20 provides a closer insight into the
streamlines at the emitter exit. In general, the formation of streamlines
in the surrounding air stems from the liquid flow in which the accu-
mulation of electric charge on its surface accelerates the liquid-gas in-
terface by the electric field. This leads to the creation of two vortices
including a clockwise vortex within the liquid meniscus (Fig. 20) and a
counter-clockwise one in the surrounding air (Fig. 19). Fig. 20 also
shows that an increase in the liquid flow rate after the establishment of
the cone-jet structure, i.e. Q > 0.93 mLh−1, in addition to altering the
cone profile from a convex shape to the concave and straight cones as Fig. 23. Variation of Sauter mean diameter of droplets versus liquid flow rate for
mentioned before, shrinks the vortex within the meniscus by Φ = + 4000 V and t = 1.2 ms.
329
H. Dastourani et al. International Journal of Heat and Fluid Flow 70 (2018) 315–335
Fig. 24. Reynolds number versus Capillary number for jets in various liquid flow rates.
Fig. 25. Temporal Cone-jet formation for Q = 3.4 mLh−1 and electric potentials (a) Φ = + 2000 V, (b) Φ = + 2250 V, (c) Φ = + 2452 V.
figure, the simulation results agree well both qualitatively and quanti- results in connection with the axisymmetric model. This necessitates to
tatively with the experimental data. This confirms the validity of the ensure that the formation of the stable cone-jet mode acquired on the
results and effectiveness of the simulating model used in this study. flow rates leading to the jetting-dripping transition is valid under the
Moreover, Fig. 23 shows that the increase in the liquid flow rate results axisymmetric approach used. In this respect, an analysis would be taken
in the larger droplets since the electric potential as the main source of into account based on the dimensionless numbers as given by (Gañán-
liquid instability and disintegration is held constant. Calvo and Montanero, 2009),
In the present study, the attention is also paid to the simulation
330
H. Dastourani et al. International Journal of Heat and Fluid Flow 70 (2018) 315–335
Fig. 26. Temporal Cone-jet formation for Q = 3.4 mLh−1 and electric potentials (a) Φ = + 3000 V, (b) Φ = + 4000 V, (c) Φ = + 5000 V.
Fig. 27. Cone-jet shapes for various electric potentials and Q = 3.4 mLh−1 at t = 1.2 ms.
331
H. Dastourani et al. International Journal of Heat and Fluid Flow 70 (2018) 315–335
Fig. 28. Cone-jet dimensions for various electric potentials and Q = 3.4 mLh−1 at t = 1.2 ms.
r j ɛ 0 E02
χj =
γ
ρUj r j ρQ
Rej = =
μ πμr j
μUj μQ
Caj = =
γ πγr j2
Wej = Rej ·Caj (17)
where χj, Rej, Caj and Wej denote Taylor, Reynolds, Capillary and
Weber numbers, respectively. Gañán-Calvo and Montanero (2009) and
Fig. 29. Magnitude of electric field against electric potential for Q = 3.4 mLh−1 at
López-Herrera et al. (2010) collected and analyzed a considerable
t = 1.20 ms.
number of experimental data from the literature for various liquids
electrosprayed or electrospun in a stable and steady cone-jet mode.
They also developed and validated an analytical model showing that
there are theoretical critical curves in the Rej-Caj plane above which a
stable and steady cone-jet would be shaped. In this regard, Fig. 24
compares the present simulation data with the critical curve limits
showing that the results are located above the curves. This confirms
that the stable and steady cone-jet structures attained in this study for
0.93 ≤ Q ≤ 12.0 mLh−1 at a constant electric potential, with Weber
and Taylor numbers ranged in 2.1 ≤ Wej ≤ 46.7 and χj < 0.003, are in
agreement with the instability limits proposed in the literature.
Fig. 31. Patterns of streamlines in emitter exit for various electric potentials for Q = 3.4 mLh−1 at t = 1.2 ms..
332
H. Dastourani et al. International Journal of Heat and Fluid Flow 70 (2018) 315–335
333
H. Dastourani et al. International Journal of Heat and Fluid Flow 70 (2018) 315–335
Fig. 35. Reynolds number versus capillary number for jets in various electric potentials.
6. Conclusions in association with the simulation results. The analysis of the results
with regard to the cone-jet formation in a jetting-dripping transition
A two-phase numerical simulating model was developed to study reveal that the magnitudes of liquid flow rates and electric potentials
the electrospray formation in the cone-jet mode taking account of the explored in this study fall in the convective instability region wherein a
liquid flow rate and electric potential as the operating parameters. To stable and steady cone-jet structure can be formed.
have a magnitude basis for the parameters, the adoption of two avail-
able correlations proposed for the minimum liquid flow rate and onset References
voltage revealed that the correlations could appropriately predict the
establishment of the cone-jet mode. To ensure the validity of the si- Agarwal, S., Greiner, A., Wendorff, J.H., 2013. Functional materials by electrospinning of
mulation methodology, the numerical results were compared with the polymers. Prog. Polym. Sci. 38 (6), 963–991.
Banerjee, S., Mazumdar, S., 2012. Electrospray ionization mass spectrometry: a technique
existing experimental data, which showed good correspondence as- to access the information beyond the molecular weight of the analyte. Int. J. Anal.
serting the suitable capability of the method for the study of electro- Chem. 2012 (1), 1–40.
spray process. This capability was also evidenced by the stability of the Barrero, A., Gañán-Calvo, A.M., Davila, J., Palacios, A., Gomez-Gonzalez, E., 1999. The
role of the electrical conductivity and viscosity on the motions inside Taylor cones. J.
computational procedure, which was attained during the subsequent Electrost. 47 (1-2), 13–26.
transient solutions for a substantial number of cases. Brackbill, J.U., Kothe, D.B., Zemach, C., 1992. A continuum method for modeling surface
The evolution of the electrospray structure illustrates that the liquid tension. J. Comput. Phys. 100 (2), 335–354.
Brenner, M.P., Eggers, J., Joseph, K., Nagel, S.R., Shi, X.D., 1997. Breakdown of scaling in
flow rate and electric potential have minimum values beyond which a droplet fission at high Reynolds number. Phys. Fluids 9 (6), 1573–1590.
cone-jet profile can be shaped. The profiles also indicate that the ac- Castellanos, A, 1998. Electrohydrodynamics, second ed. Springer, New York.
cumulation of the electric charge occurs in the liquid-gas interfacial Chaudhary, K.C., Redekopp, L.G., 1980. The nonlinear capillary instability of a liquid jet.
Part 1. Theory. J. Fluid Mech. 96 (2), 257–274.
layer making the interior of the liquid flow electrically uncharged. In
Chen, D.R., Pui, D.Y., 1997. Experimental investigation of scaling laws for electro-
this respect, the electric charge density reveals an ascending-des- spraying: dielectric constant effect. Aerosol Sci. Technol. 27 (3), 367–380.
cending trend along the cone-jet interface assigning the maximum Chen, D.R., Pui, D.Y., Kaufman, S.L., 1995. Electrospraying of conducting liquids for
concentration to the cone-jet intersection. However, the effect of the monodisperse aerosol generation in the 4 nm to 1.8 µm diameter range. J. Aerosol
Sci. 26 (6), 963–977.
electric potential on the electric charge density is more obvious than for Cherney, L.T., 1999. Structure of Taylor cone-jets: limit of low flow rates. J. Fluid Mech.
the liquid flow rate, which stems from the ascending trend of electric 378, 167–196.
field strength against the electric potential. Chetwani, N., Cassou, C.A., Go, D.B., Chang, H.C., 2010. High-frequency AC electrospray
ionization source for mass spectrometry of biomolecules. J. Am. Soc. Mass Spectrom.
The results show that the increase in the liquid flow rate leads to the 21 (11), 1852–1856.
longer jet tails with a transformation of the cone generator from a Cloupeau, M., Prunet-Foch, B., 1989. Electrostatic spraying of liquids in cone-jet modes.
concave curve to a straight line. However, the larger liquid flow rates J. Electrostat. 22 (1), 135–159.
Cloupeau, M., Prunet-Foch, B., 1990. Electrostatic spraying of liquids: main functioning
are associated with the lower levels of surface charge density, which, in modes. J. Electrostat. 25 (2), 165–184.
turn, reduce the liquid-gas interfacial velocity. In contrast, the cone-jet Davoodi, P., Feng, F., Xu, Q., Yan, W.C., Tong, Y.W., Srinivasan, M.P., Sharma, V.K.,
length indicates an ascending-descending trend versus the electric po- Wang, C.H., 2015. Coaxial electrohydrodynamic atomization: microparticles for drug
delivery applications. J. Control. Release 205, 70–82.
tential.
De La Mora, J.F., Loscertales, I.G., 1994. The current emitted by highly conducting Taylor
The fluid flow patterns reveal the creation of vortices within the cones. J. Fluid Mech. 260, 155–184.
liquid flow at the emitter exit. The liquid flow rate and electric potential Deshpande, S.S., Anumolu, L., Trujillo, M.F., 2012. Evaluating the performance of the
two-phase flow solver interFoam. Comput. Sci. Discov. 5 (1), 014016.
affect the characteristics of the vortices. The patterns indicate that the
Eggers, J., Dupont, T.F., 1994. Drop formation in a one-dimensional approximation of the
sizing and positioning of the vortices are drastically dependent on the Navier–Stokes equation. J. Fluid Mech. 262, 205–221.
liquid flow rate as compared with their moderate dependency on the Emad, V., 2014. Evaluating the Performance of Various Convection Schemes on Free
electric potential. As anticipated, the liquid flow rate and electric po- Surface Flows by Using Interfoam Solver. Doctoral dissertation. Eastern
Mediterranean University.
tential play opposite roles on the mean droplet size. Fenn, J.B., Mann, M., Meng, C.K., Wong, S.F., Whitehouse, C.M., 1989. Electrospray io-
The validity of axisymmetric approach used in this study is assessed nization for mass spectrometry of large biomolecules. Science 246 (4926), 64–71.
334
H. Dastourani et al. International Journal of Heat and Fluid Flow 70 (2018) 315–335
Ferrera, C., López-Herrera, J.M., Herrada, M.A., Montanero, J.M., Acero, A.J., 2013. 120–128.
Dynamical behavior of electrified pendant drops. Phys. Fluids 25 (1), 012104. Pantano, C., Gañán-Calvo, A.M., Barrero, A., 1994. Zeroth-order electrohydrostatic so-
Gamero-Castano, M., 2008. The structure of electrospray beams in vacuum. J. Fluid lution for electrospraying in cone-jet mode. J. Aerosol Sci. 25 (6), 1065–1077.
Mech. 604, 339–368. Park, J.U., Hardy, M., Kang, S.J., Barton, K., Adair, K., kishore Mukhopadhyay, D., Lee,
Gañán-Calvo, A.M., 1997. Cone-jet analytical extension of Taylor's electrostatic solution C.Y., Strano, M.S., Alleyne, A.G., Georgiadis, J.G., Ferreira, P.M., 2007. High-re-
and the asymptotic universal scaling laws in electrospraying. Phys. Rev. Lett. 79 (2), solution electrohydrodynamic jet printing. Nat. Mater. 6 (10), 782–789.
217–220. Puckett, E.G., Almgren, A.S., Bell, J.B., Marcus, D.L., Rider, W.J., 1997. A high-order
Gañán-Calvo, A.M., Davila, J., Barrero, A., 1997. Current and droplet size in the elec- projection method for tracking fluid interfaces in variable density incompressible
trospraying of liquids. Scaling laws. J. Aerosol Sci. 28 (2), 249–275. flows. J. Comput. Phys. 130 (2), 269–282.
Gañán-Calvo, A.M., Montanero, J.M., 2009. Revision of capillary cone-jet physics: elec- Raees, F., Van der Heul, D.R., Vuik, C., 2011. Evaluation of the Interface-Capturing
trospray and flow focusing. Phys. Rev. E 79 (6), 066305. Algorithm of OpenFoam for the Simulation of Incompressible Immiscible Two-Phase
Ghelich, R., Mehdinavaz Aghdam, R., Torknik, F.S., Jahannama, M.R., 2016. Low tem- Flow. Delft University of Technology.
perature carbothermal reduction synthesis of ZrC nanofibers via cyclized electrospun Rosell-Llompart, J., De La Mora, J.F., 1994. Generation of monodisperse droplets 0.3 to
PVP/Zr (OPr)4 hybrid. Int. J. Appl. Ceram. Technol. 13 (2), 352–358. 4 µm in diameter from electrified cone-jets of highly conducting and viscous liquids.
Gueyffier, D., Li, J., Nadim, A., Scardovelli, R., Zaleski, S., 1999. Volume-of-fluid interface J. Aerosol Sci. 25 (6), 1093–1119.
tracking with smoothed surface stress methods for three-dimensional flows. J. Rusche, H., 2003. Computational Fluid Dynamics of Dispersed Two-Phase Flows at High
Comput. Phys. 152 (2), 423–456. Phase Fractions, Ph.D. dissertation. Imperial College.
Hartman, R.P.A., Borra, J.P., Brunner, D.J., Marijnissen, J.C.M., Scarlett, B., 1999a. The Salata, O.V., 2005. Tools of nanotechnology: electrospray. Curr. Nanosci. 1 (1), 25–33.
evolution of electrohydrodynamic sprays produced in the cone-jet mode; a physical Saville, D.A., 1997. Electrohydrodynamics: the Taylor–Melcher leaky dielectric model.
model. J. Electrostat. 47 (3), 143–170. Ann. Rev. Fluid Mech. 29 (1), 27–64.
Hartman, R.P.A., Brunner, D.J., Camelot, D.M.A., Marijnissen, J.C.M., Scarlett, B., 1999b. Setiawan, E.R., Heister, S.D., 1997. Nonlinear modeling of an infinite electrified jet. J.
Electrohydrodynamic atomization in the cone–jet mode physical modeling of the li- Electrostat. 42 (3), 243–257.
quid cone and jet. J. Aerosol Sci. 30 (7), 823–849. Sussman, M., Almgren, A.S., Bell, J.B., Colella, P., Howell, L.H., Welcome, M.L., 1999. An
Hartman, R.P.A., Brunner, D.J., Camelot, D.M.A., Marijnissen, J.C.M., Scarlett, B., 2000. adaptive level set approach for incompressible two-phase flows. J. Comput. Phys. 148
Jet break-up in electrohydrodynamic atomization in the cone-jet mode. J. Aerosol (1), 81–124.
Sci. 31 (1), 65–95. Sussman, M., Fatemi, E., Smereka, P., Osher, S., 1998. An improved level set method for
Herrada, M.A., López-Herrera, J.M., Gañán-Calvo, A.M., Vega, E.J., Montanero, J.M., incompressible two-phase flows. Comput. Fluids 27 (5), 663–680.
Popinet, S., 2012. Numerical simulation of electrospray in the cone-jet mode. Phys. Sussman, M., Smereka, P., Osher, S., 1994. A level set approach for computing solutions
Rev. E 86 (2), 026305. to incompressible two-phase flow. J. Comput. Phys. 114 (1), 146–159.
Hirt, C.W., Nichols, B.D., 1981. Volume of fluid (VOF) method for the dynamics of free Sweet, M.L., Pestov, D., Tepper, G.C., McLeskey, J.T., 2014. Electrospray aerosol de-
boundaries. J. Comput. Phys. 39 (1), 201–225. position of water soluble polymer thin films. Appl. Surf. Sci. 289 (1), 150–154.
Jaworek, A., Krupa, A., 1999a. Classification of the modes of EHD spraying. J. Aerosol Sci. Tang, K., Gomez, A., 1996. Monodisperse electrosprays of low electric conductivity li-
30 (7), 873–893. quids in the cone-jet mode. J. Colloid Interface Sci. 184 (2), 500–511.
Jaworek, A., Krupa, A., 1999b. Jet and drops formation in electrohydrodynamic spraying Taylor, G., 1969. Electrically driven jets. Math. Phys. Eng. Sci. 313 (1515), 453–475.
of liquids: a systematic approach. Exp. Fluids 27 (1), 43–52. Taylor, G.I., 1964. Disintegration of water drops in an electric field. Math. Phys. Sci. 280
Jaworek, A., Sobczyk, A.T., 2008. Electrospraying route to nanotechnology: an overview. (1382), 383–397 1964.
J. Electrostat. 66 (3), 197–219. Tomar, G., Gerlach, D., Biswas, G., Alleborn, N., Sharma, A., Durst, F., Welch, S.W.J.,
Jeong, J.T., Moffatt, H.K., 1992. Free-surface cusps associated with flow at low Reynolds Delgado, A., 2007. Two-phase electrohydrodynamic simulations using a volume-of-
number. J. Fluid Mech 241 (1), 1–22. fluid approach. J. Comput. Phys. 227 (2), 1267–1285.
Joffre, G., Prunet-Foch, B., Berthomme, S., Cloupeau, M., 1982. Deformation of liquid Tryggvason, G., Bunner, B., Esmaeeli, A., Juric, D., Al-Rawahi, N., Tauber, W., Han, J.,
menisci under the action of an electric field. J. Electrostat. 13 (2), 151–165. Nas, S., Jan, Y.J., 2001. A front-tracking method for the computations of multiphase
Joffre, G.H., Cloupeau, M., 1986. Characteristic forms of electrified menisci emitting flow. J. Comput. Phys. 169 (2), 708–759.
charges. J. Electrostat. 18 (2), 147–161. Wei, W., Gu, Z., Wang, S., Zhang, Y., Lei, K., Kase, K., 2013. Numerical simulation of the
Klostermann, J., Schaake, K., Schwarze, R., 2013. Numerical simulation of a single rising cone–jet formation and current generation in electrostatic spray—modeling as re-
bubble by VOF with surface compression. Int. J. Numer. Methods Fluids 71 (8), gards space charged droplet effect. J. Micromech. Microeng. 23 (1), 1–11.
960–982. Xu, Q., Qin, H., Yin, Z., Hua, J., Pack, D.W., Wang, C.H., 2013. Coaxial electro-
Lastow, O., Balachandran, W., 2006. Numerical simulation of electrohydrodynamic hydrodynamic atomization process for production of polymeric composite micro-
(EHD) atomization. J. Electrostat. 64 (12), 850–859. spheres. Chem. Eng. Sci. 104, 330–346.
Lim, L.K., Hua, J., Wang, C.H., Smith, K.A., 2011. Numerical simulation of cone‐jet for- Yan, F., Farouk, B., Ko, F., 2003. Numerical modeling of an electrostatically driven liquid
mation in electrohydrodynamic atomization. AIChE J. 57 (1), 57–78. meniscus in the cone–jet mode. J. Aerosol Sci. 34 (1), 99–116.
López-Herrera, J.M., Barrero, A., Boucard, A., Loscertales, I.G., Marquez, M., 2004. An Yan, W.C., Davoodi, P., Tong, Y.W., Wang, C.H., 2016. Computational study of core‐shell
experimental study of the electrospraying of water in air at atmospheric pressure. J. droplet formation in coaxial electrohydrodynamic atomization process. AIChE J. 62
Am. Soc. Mass Spectrom. 15 (2), 253–259. (12), 4259–4276.
López-Herrera, J.M., Gañán-Calvo, A.M., Herrada, M.A., 2010. Absolute to convective Yoon, H., Woo, J.H., Ra, Y.M., Yoon, S.S., Kim, H.Y., Ahn, S., Yun, J.H., Gwak, J., Yoon,
instability transition in charged liquid jets. Phys. Fluids 22 (6), 062002. K., James, S.C., 2011. Electrostatic spray deposition of copper–indium thin films.
López-Herrera, J.M., Popinet, S., Herrada, M.A., 2011. A charge-conservative approach Aerosol Sci. Technol. 45 (12), 1448–1455.
for simulating electrohydrodynamic two-phase flows using volume-of-fluid. J. Youngs, D.L., 1982. Time-dependent multi-material flow with large fluid distortion.
Comput. Phys. 230 (5), 1939–1955. Numer. Methods Fluid Dyn. 24 (1), 273–285.
Morris, T., Malardier-Jugroot, C., Jugroot, M., 2013. Characterization of electrospray Yu, J., Qiu, Y., Zha, X., Yu, M., Rafique, J., Yin, J., 2008. Production of aligned helical
beams for micro-spacecraft electric propulsion applications. J. Electrostat. 71 (5), polymer nanofibers by electrospinning. Eur. Polym. J. 44 (9), 2838–2844.
931–938. Yu, M., Ahn, K.H., Lee, S.J., 2016. Design optimization of ink in electrohydrodynamic jet
Osher, S., Sethian, J.A., 1988. Fronts propagating with curvature-dependent speed: al- printing: effect of viscoelasticity on the formation of Taylor cone jet. Mater. Des. 89,
gorithms based on Hamilton-Jacobi formulations. J. Comput. Phys. 79 (1), 12–49. 109–115.
Ouedraogo, Y., Gjonaj, E., Weiland, T., De Gersem H., SteinhausenC., Lamanna, G., Zeleny, J., 1914. The electrical discharge from liquid points, and a hydrostatic method of
Weigand, B., Preusche, A., Dreizler, A., Schremb, M., 2017. Electrohydrodynamic measuring the electric intensity at their surfaces. Phys. Rev. 3 (2), 69–91.
simulation of electrically controlled droplet generation. Int. J. Heat Fluid Flow 64, Zeleny, J., 1917. Instability of electrified liquid surfaces. Phys. Rev. 10 (1), 1–16.
335