See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/352912378

NUMERICAL STUDY OF VACUUM-AIDED SPRAY FLASH DESALINATION

Conference Paper · January 2021

DOI: 10.1615/TFEC2021.fip.036274

CITATIONS READS

3 152

4 authors:

Guangyu Guo Hongling Deng

New Jersey Institute of Technology New Jersey Institute of Technology
14 PUBLICATIONS 114 CITATIONS 13 PUBLICATIONS 96 CITATIONS

SEE PROFILE SEE PROFILE

Chao Zhu Zhiming Ji

Independent Researcher New Jersey Institute of Technology
67 PUBLICATIONS 1,149 CITATIONS 69 PUBLICATIONS 1,395 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Guangyu Guo on 10 January 2022.

The user has requested enhancement of the downloaded file.

 6th Thermal and Fluids Engineering Conference (TFEC)
May 26–28, 2021
Virtual

TFEC-2020-36274

NUMERICAL STUDY OF VACUUM-AIDED SPRAY FLASH

DESALINATION
Guangyu Guo, Hongling Deng, Chao Zhu*, Zhiming Ji

Department of Mechanical and Industrial Engineering

New Jersey Institute of Technology, Newark, NJ 07102

ABSTRACT

Spray flash desalination is an advanced distillation method by space flash evaporation of over-
saturated multi-component droplets and/or ligaments to generate the distilled condensable vapor.
Vacuum-aided spray flash is spraying the superheated and atomized droplets into a depressurized
chamber whose low pressure is maintained by a continued vacuum extraction of vapors, a process of
high thermal non-equilibrium and of high non-uniformity in phase changes and transport. This study
presents a CFD numerical simulation to quantify some of the non-equilibrium and non-uniformity
characteristics in vacuum-aided spray flash distillation, especially on the temperature difference and
the thermal re-equilibrium tendency between vapor and liquid residue. Since ANSYS Fluent lacks
the mathematic model of self-heat supplied evaporation in discrete phase modeling (DPM), a
pressure-driven flash evaporation model is proposed and implemented into CFD via User-defined
functions (UDF). Preliminary results meet a good agreement with experiments. It revealed the non-
uniformed characteristics of the vapor phase (i.e., distributions of temperature and velocity), as well
as the thermal non-equilibrium between the two phases. It also suggests that the salt fraction within
droplets will result in negative impacts on evaporating efficiency and evaporating rate during the
process.

KEYWORDS: spray flash desalination, multi-component evaporation, non-equilibrium process, CFD, UDF

1. INTRODUCTION

Spray flash desalination is an advantaged thermal distillation approach realized by flash evaporation of
superheated droplets within an evaporating chamber[1]. Flash evaporation is an intensive evaporation process
when liquid experiencing an abrupt state transition of becoming superheated under depressurized operating
conditions. Comparing with traditional flash desalination such as pool flash desalination and film flash
desalination, the spray flash desalination dramatically increases the effective evaporation area of the liquid by
atomization, which results in a much faster evaporation rate and better evaporation efficiency [2]. It also has
other unique advantages such as better controllability of the flash process, less reaction volume, no scaling,
etc.[1]

The spray flash desalination is a very complicated and heavily coupled phenomenon that involves the
integrated exchanges of mass, momentum, and energy between components (volatile fraction and non-violate
fraction) and phases (liquid and gas) [3,4]. Multi-component droplet evaporation within a spray is mainly
governed by several coupled mechanisms: the heat transfer to the interface from the inner liquid side; the heat

*Corresponding Author: chao.zhu@njit.edu

1
 TFEC-2020-36274

and mass transfer on the evaporating interface from the gaseous ambient; and the mass diffusion of the volatile
substances inside the droplet due to the concentration gradient created by the evaporation [5]. Depending on
different kinds of droplet compounds and operating conditions, the evaporation rate can either be limited by
the heat transfer inside the droplet, the mass transfer through the evaporating interface, or the concentration of
the volatile substances on the interface [5,6]. For most cases of spray flash desalination in which the superheat
level of inlet liquid is relatively low, the local salt concentration and the mass transfer of steam in the vapor
side will be the primary limitations for its evaporation rate [4,6].

One unique phenomenon of spray flash evaporation that has been reported in previous studies is it will
effectively generate a thermal non-equilibrium between two phases (vapor and liquid) during the evaporation
[4,7,8]. Specifically, in an isolated spray flash evaporation, in which latent heat is mainly self-supplied by the
droplet itself, the temperature of residue liquid will continuously decrease along with the evaporation, and it
will always be colder than the freshly generated vapor. As a result, a thermal re-balance tendency will appear
due to the formed thermal non-equilibrium between two phases [8]. Meanwhile, accelerating the vapor
extraction during the evaporation process by introducing an active vacuuming source to the evaporation
chamber will promote such a thermal non-equilibrium due to the reduced residue time of vapor within the
chamber [4]. However, due to the introduced momentum by the spray and vapor extraction, the convective
gaseous phases within the evaporation chamber will make the spray flash process even more complicated to
predict. For instance: the convective boundary layer enhances the heat and mass transport on the droplet’s
interface; The vaporization, the relative surface forces, and body forces will cause a varying density of gas
around the droplet and complicates the flow characteristics [3,5,8].

Therefore, to consider the strong coupling and highly non-equilibrium of heat and mass transfers, as well as
the strong non-uniformity in spray characteristics and gaseous phases, a full-scale three-dimensional CFD
study is desired to simulate the transport phenomena of a spray flash desalination process under active
vacuuming. So far, several CFD studies that involved spray flash have been reported based on different
applications [9–11]. Especially, a CFD simulation of spray flash evaporation with regular tape water is reported
by Guo [8]. The spray flash evaporation in all these CFD studies is based on the lumped heat capacity modeling
to the droplet, which is due to the simplified point-treatment of droplets in the discrete phase modeling [12–
14].

In this work, a three-dimensional CFD simulation of spray flash desalination in an isolated chamber with active
vapor extraction is performed via ANSYS Fluent. To simulate the self-heat supplied flash evaporation
mechanism, the non-volatile fraction (salt) effect to the evaporation, as well as the coupled heat and mass
transfer between the liquid and gas, a multi-component evaporation model based on the Lagrangian-Euler
approach has been developed and compiled into discrete phase modeling (DPM) via User-defined functions
(UDF). One specific case of the simulation has been performed and validated with experimental results. Both
meet a good agreement in terms of the evaporation rate of droplets and the averaged outlet temperature of
vapor and liquid residue. The simulation revealed the non-uniformed characteristics of the vapor phase (i.e.,
distributions of temperature and velocity), as well as the thermal non-equilibrium between the two phases.

2. MODELING AND CASE SETUP

In this section, a multi-component spray flash evaporation model is proposed with two-way coupling between
discrete phase (droplets) and continuous gaseous phase (vapor) by the Lagrangian-Eulerian modeling. The
modeling equations are complied and solved via FLUENT DPM and its associated UDFs. The geometric
structure in our case is shown in Figure 1. The spray flash desalination happens within an isolated chamber
that has three openings: the vapor outlet (top), the liquid outlet (bottom), as well as the spray inlet (top-center).
From the center hole of the top side, a steady salty water spray with constant flow rate, size distribution, and
the initial temperature is jetted into the domain. Due to the huge volume ratio between water and vapor (around
3300:1), we assume the volumetric effect from the liquid phase will be ignored. Once the droplets reach the

2
 TFEC-2020-36274

chamber wall or if the droplet temperature is sufficiently close to the equilibrium temperature corresponding
to the vacuum pressure, the droplets will be assumed as immediately escaped from the chamber. The liquid
outlet in CFD simulation is treated as a part of “wall” for the continuous phase (vapor) since the liquid outlet
pipe in the experimental system does not allow any vapor escape, which is always filled up with saturated
water. Since all droplets are treated as points in the CFD simulation, there are no volumetric effects of droplets
on vapor flow nor volumetric accumulation of water residue. All chamber walls are thermally insulated.

Fig 1. Geometric structure of the CFD simulation

2.1 Lagrangian Model of Individual Droplets

In this case, the discrete phase (droplets) is strongly coupled with the continuous gas phase, and the evaporative
source (vapor) constitutes the entire continuous gas phase. As each iteration of the computation, FLUENT
firstly calculates the continuous phase, then solves the discrete droplet phase of the Lagrangian model. The
results will be used to feedback and update the source terms to the flow domain for the next iteration. Based
on this method, we will have two coupled groups of governing equations: one set is for the discrete phase
(droplets); the other set is for the continuous gas phase (vapor).

For model closure of the discrete phase particle (droplet), Four variables of each droplet are considered as the
velocity(ud), the temperature(Tp), the mass(mp), and the molar fraction of NaCl(Nw). Four corresponding
equations are user-defined as the momentum equation eq(1), the energy equation eq(2), the evaporation
equation eq(3), as well as the species equation (3-b).
For a droplet settling by gravity in a depressurized vapor flow, the motion of the droplet is governed by the
momentum balance of relevant forces as gravity and drag force, expressed by
d  md u d  
 md g  C D Ad v u v  u d (u v  u d ) (1)
dt 2
Since the droplets are under a single gas phase environment (pure vapor) with 100% relative humidity, the
boiling process is assumed there only has mass transfer between droplets and continuous gas. Therefore the
energy equation is expressed by
dTd dm
md cd  hAd (T  Td )  d h fg (2)
dt dt

3
 TFEC-2020-36274

Under the active vacuuming operation, the evaporation rate of the droplet is controlled by the pressure
difference between the saturated vapor pressure at the droplet temperature and the ambient pressure controlled
by vacuuming. Hence, the evaporation rate could be determined from the modified Hertz-Kundsen relation
[4,15]:
Mv
m v   d d 2  Psat  P0  (3)
2 RTd
where α is the evaporation coefficient, and it needs to be experimentally determined [16]. Mv is the molecular
weight of the vapor, R is the universal gas constant, Psat is the saturated vapor pressure at Td, and P0 is ambient
pressure. Moreover, according to Raoult’s law, the saturated pressure of the volatile component needs to be
corrected by below equation for the calculation of evaporation rate:
Psat  Nw Pd (3-a)
Where Nw is the mole fraction of volatile component (water), and can be determined as:
1
Nw 
MWw Cs (3-b)
1 ( )
MWs 1  Cs
Where MWw is the molar weight of water, and MWs is the molar weight of NaCl. Cs is the local salinity on
the evaporating droplet surface.

2.2 Eulerian Model of Vapor Flow

For model closure of the gas phase, the number of total independent variables should equal the number of total
independent equations. As illustrated in Table 1, we have nine variables, which are density (ρ), velocity (uv:
u, v, w), pressure (p), temperature (T), Turbulent viscosity (μT), turbulence kinematic energy (k) and turbulence
dissipation (ε). Meanwhile, we have the same amount of independent equations, which are continuity,
momentum (3: u, v, w), energy, equation of state, turbulent viscosity, turbulence kinematic energy, and
turbulence dissipation. The functions group is closed. Detailed major governing equations of the gas phase are
summarized in Table 1.

Table 1. Major governing equations of gas phase


Continuity    (uv )  Sm
t

Momentum  uv     uvuv   p       T  uv   g  Fvd
t

Energy  cpTv     cpuvTv      K  KT  Tv 
t

Equation of state p  RT

u v

k2
Turbulent viscosity T   C

  T  
Turbulence k   k       kU         k   Gk  Gb    YM
t  k  
  T   2 
Turbulence ε         U             C2  C1 C3 Gb
t     k  v k

To solve this coupling problem, the source terms of the continuity equation and momentum equation are also
needed to be determined. For the continuity equation, the source term Sm of a polydispersed spray, which
represents vapor generation rate per volume, can be expressed by:

4
 TFEC-2020-36274

N
1  N j 
Sm =    m vij  (4)
V cell  j 1 i 1 
where subscripts i and j denote the ith droplet of jth size group (or on jth trajectory), Nj is the total number of
droplets of jth size-group within Vcell, N is the total number of droplet size groups, and mv is the evaporation
rate of an individual droplet that is previously solved from the Lagrangian model. Nj depends on the local
number density of droplets of group j, nj, by
N j  Vcell nj (5)
For the momentum equation, the Fvd is the droplet-vapor momentum transfer per volume, which is also coupled
with the discrete phase (polydisperse spray), can be expressed by:
N
1  N j 
Fvd =    Fij  (6)
Vcell  j 1 i 1 
where Fij , the force coupled with each single droplet, is given by:
d  md u d 
Fij = (7)
dt

2.3 Case setup

Based on the modeling, a FLUENT DPM simulation was set up with UDFs. The solver description boundary
definitions are summarized in Table 1. As one necessary initial condition, droplets sizes distribution and the
average velocity of the spray are experimentally determined via a laser-scanning added optical measurement
system [16]. The calibrated inlet spray, as well as geometric information of the evaporator, are summarized in
Table 2. To validate the numerical results with the experimental results, the operating conditions of the CFD
are defined the same as experiments. The temperature depended polynomial relations of physical properties
used in models and FLUENT are summarized as Table 3.

To justify the mesh cell model, the mesh sensitivity analysis was conducted on three different mesh densities
to ensure the results are independent of the grid, corresponding to the total cell number is 0.28 million, 0.58
million, and 1.33 million, respectively. A ten-seconds operating was simulated for these mesh models and it
was found the results will not change significantly between 0.58 million and 1.33 million with the value
differences range from 0.14% to 1.2% for curial parameters such as average outlet temperature of two phases,
the average temperature of full-field vapor phase, and overall evaporation rate of droplets. To reduce the
computational cost, the 0.58 million-case is adopted for the simulation.

Table 1. Setup details of CFD simulation

Solver description
Steady/unsteady Transient/unsteady particle tracking
Dimension 3-D
Turbulence model Re-Normalized Group (RNG) k-ε model
Discrete phase Droplets with UDF evaporating laws and coupled heat-mass solution; interacted
model with continuous phase
Boundary conditions
Spray inlet Mass flow inlet (zero-in); DPM inlet
Steam outlet Pressure outlet; DPM escape
Water outlet Mass flow outlet (zero-out); DPM escape
wall No heat flux; DPM escape

5
 TFEC-2020-36274

Table 2. Spray calibration and geometric information

Parametric Value Unit
Spray pattern Full-cone N/A
Spray angle 59 degree
Spray fluid Saltwater N/A
Droplets diameter distribution
κ=4.432 and δ=1.099 mm
(Rosin-Rammler)
Average spray velocity 3.3m/s m/s
Mass flow rate 0.0441 Kg/s
Nozzle arrangement one at center N/A
Numbers of nozzle 1 N/A
Chamber height 0.5 m
Chamber diameter 0.3 m
Vapor outlet diameter 0.01905 m
Liquid outlet diameter 0.0127 m
Chamber material Stainless steel with insulation N/A
Thermal resistance of chamber 2.16 K·m2/W

Table 3. Temperature dependency of phase properties

equations remarks
hfg=aT2+bT+c a=-1.794;b=-2256;c=2.5e6
ρv= aT3+bT2+cT+d a=1.063e-6;b=-7.363e-5;c=0.002929;d=-0.01492
uv=aT+b a=4.031e-8;b=7.882e-6
cl= aT4+bT3+cT2+dT1+e a=1.815e-6;b=-4.754e-4;c=0.052;d=-2.139;e=4207
cpv= aT2+bT+c a=0.01609;b=0.07276;c=1859

4. RESULTS AND DISCUSSION

In this section, a typical case is numerically simulated with an operating pressure of 40kpa and 367K inlet
temperature of the spray, and all other operating conditions are followed as Table 2. The transient solving
approach has been used to run the simulation and until the critical variables become time non-sensitive (300
seconds in this case), such as distributions of temperatures, velocities of both two phases, as well as yield rate.
As for validations, the computed evaporation rates of the droplet (in terms of droplet’s temperature) via CFD
are compared with experimental measurements [16] in Figure 2. Except for the salty water spray, the pure
water spray without NaCl is also simulated under the same operating condition to present the salinity effect on
decreasing the evaporation rate.

Since the specific evaporation completeness of each droplet will be affected by its size, to effectively validate
the result with experimental data, the droplet in CFD is sampled whose diameter equals Sauter Mean Diameter
and located on the vertical central path for the comparison. The red squares are the data measured from the
experiments. In our case, since the thermal probe only has ten sensors and limited length, the last reading stops
around 0.26m, the sampling of numerical results is also cut off at 0.29m. However, comparing the last value
of the probe and the temperature at the water outlet, a similar reading indicates the thermal probe still
successfully covers most parts of the evaporation process. In addition, the simulated overall evaporation rate,
and the averaged outlet temperature of vapor and water, are also compared with experimental measurements

6
 TFEC-2020-36274

for the case of salty water spray, which are summarized in Table 4. The result shows the numerical estimation
of the overall thermal performance meets a good agreement with experiments.

Fig 2. Comparison of droplets temperature along with the vertical travel distance

Table 4. Experimental validation of overall thermal characteristics

CFD experiment
Evaporation rate 4.13 kg/h 4.05kg/h
Outlet temperature
356.8K 356.0K
of the vapor
Outlet temperature
349.9K 348.6K
of the water

Fig 3. droplets temperature along the trajectory Fig 4. droplets mass along the trajectory

7
 TFEC-2020-36274

Based on the size distribution function mentioned in Table 2, the injected droplets of the spray are separated
into ten groups according to different sizes in each parcel. However, to remove the size influences on the
evaporation rate, the droplet whose diameter equals the median diameter (1.03mm) has been sampled from the
results and presented in Figure 2 to show the temperature changing trend. We can see most droplets could
achieve saturated temperature (350K) before hitting the wall. In fact, in this case, the Sauter mean diameter
d32 is only about 0.3mm, which means most droplets have enough time to reach the saturated status within the
travel. Figure 3 shows the droplet mass change along with the traveling trajectory. As expected, both the mass
change and temperature change are representing the evaporation process and follow the same trend as
expected.

Figure 3 shows the mass fraction of NaCl of the same sampled droplets. No surprise, the mass fraction of NaCl
increased along with the evaporation process going on due to the continued mass decrease of liquid water.
Worth noting, figures 1-3 all indicate the flash evaporation is very fast that it only needs around 1/5 lengths of
the entire height of the chamber to complete the most process. A cut-off reference based on the evaporative
completeness could be discussed in the future to optimize the design of the evaporator.

Fig 5. NaCl mass fraction along the trajectory Fig 6. the contour of vapor source

Figures 6 and 7 show the temperature gradient caused by the spray evaporation and the vapor source to the gas
phase and temperature gradient caused by the spray evaporation in the median plane of the chamber. Agreed
with the results of the discrete phase, Figure 6 also clearly indicates that most vaporization happens at the
upper center zone of the chamber. Figure 7 shows the temperature distribution within the chamber. It indicates
the spray flash desalination of our case generated a thermal non-equilibrium in the full-field domain of the
vapor. Specifically, vapor was continuously introduced from the injecting spray, the hottest vapor generated
and accumulated around the top zone of the chamber, and the fresh vapor from downstream becomes cooler
and cooler until evaporation vanished. Around the exit, hotter vapor and cooler vapor get mixed and then
finally extracted out through the steam outlet.

Figure 8 shows the velocity contour of the vapor in the median plane of the chamber. The result indicates, due
to the downward movement of the droplets, the generated vapor also comes with an initial downward velocity
before it moves upwards by the vacuum extraction. Near the vapor outlet, there is much-enhanced vapor

8
 TFEC-2020-36274

suction due to the small cross-section area of the vapor outlet. This strong vapor suction region is luckily away
from the conical spray coverage zone marked by droplet trajectories. Otherwise, a sufficiently strong sucking
flow could directly entrain some small droplets or droplets of weak inertia into the vapor extraction tube. The
velocity scale in this figure is limited by 2.5 m/s to try presenting the details of the velocity distribution in most
weak-flowing areas. The strong-flowing jetting and sucking area are represented as the red zone in the contour,
and vapor velocities in there could go much higher than 2.5m/s (especially near the vapor outlet).

Fig 7. the contour of vapor temperature Fig 8. the contour of vapor velocity magnitude

In addition, the velocity contour also indicates the vapor will go through a mixing process before escaped out
of the chamber. And in theory, such a mixing process could reduce the level of thermal non-equilibrium of the
full-field domain by prompt the convection of the gaseous phase. How strong the mix could depend on the
operational conditions, such as the injected velocity of the spray and the pressure level of the vacuuming.
Either a higher evaporation rate or a higher spray velocity will lead to a stronger turbulent field. This unique
phenomenon worth further study in the future. In addition, as mentioned, the sucking flow pattern near the
steam outlet could be a concern for the industrial application. A safe range of outflow velocity to avoid
extracting small droplets should be considered in the future.

5. CONCLUSIONS
This paper presented a three-dimensional CFD study of simulating spray flash desalination in an isolated
evaporator with active vapor extraction. A polydispersed multi-component spray flash evaporation model is
proposed with two-way coupling between discrete phase and continuous gaseous phase by the Lagrangian-
Eulerian modeling. The modeling equations are complied and solved via FLUENT DPM and its associated
UDFs. A typical desalination process within a cylindrical evaporator has been simulated under vacuuming at
40kPa operating pressure with a single full-cone spray at an inlet temperature of 367K and NaCl mass fraction
of 3.5%. The spray is polydispersed and follows the Rosin-Rammler size distribution with an average size of
1 mm. Both simulations and experiments suggest the spray flash desalination process happens very fast, that
most droplets achieve near-equilibrium status very shortly. Two simulation results with different salinity of

9
 TFEC-2020-36274

the spray suggest the salt fraction brings negative impacts on evaporating efficiency and evaporating rate. The
simulation results revealed the vapor characteristics of the thermal non-equilibrium in temperature
distributions of the vapor, as well as the non-uniformity in the velocity and mass distributions. The motion of
droplets and vacuuming will also promote the vapor convection within the chamber, which could result in
reducing the thermal non-equilibrium level of full-field.

NOMENCLATURE

A projected area (m2) C drag coefficient (-)

c specific heat (kJ/kgK) d diameter (m)
F force (N) h heat (J)
M molecular weight (kg/mol) m mass (kg)
N group number/mole fraction (-) n number (-)
P pressure (Pa) R universal gas constant (J/molK)
S Source (-) s salt (-)
T temperature (C) t time (s)
u velocity (m/s) V volume (m3)

Greeks
α evaporative coefficient ε Dissipation/emissivity
ρ density (kg/m3) μ dynamic viscosity (kg/m∙s)
∆ difference

Subscripts
A area a added mass
ch chamber d droplet, drag
e end j trajectory number
i order number l water
m modified phase fg latent heat
p specific heat of gas phase sat saturated condition
v vapor 0 ambient, initial condition

10
 6th Thermal and Fluids Engineering Conference (TFEC)
May 26–28, 2021
Virtual

TFEC-2020-36274
REFERENCES

[1] M. Maria Antony Raj, K. Kalidasa Murugavel, T. Rajaseenivasan, K. Srithar, A review on flash evaporation desalination,
Desalin. Water Treat. 57 (2016) 13462–13471.
[2] O. Miyatake, T. Tomimura, Y. Ide, T. Fujii, An experimental study of spray flash evaporation, Desalination. 36 (1981) 113–
128.
[3] B. Cai, Y. Yin, Y. Zheng, W. Wang, H. Gu, J. Yao, H. Wang, Mathematical study of spray flash evaporation in a spray-
assisted seawater desalination chamber, Desalination. 465 (2019) 25–37.
[4] G. Guo, H. Deng, C. Zhu, Z. Ji, Non-volatile fraction effects in dispersed vacuum spray flash evaporation, in: Int. Mech.
Eng. Congr. Expo., 2020.
[5] W.A. Sirignano, Fluid Dynamics and Tandsport of Droplets and Spray, Cambridge University Press, 2010.
[6] E.L.Cussler, Diffusion Mass Transfer in Fluid System, 2009.
[7] G. Guo, H. Deng, C. Zhu, Z. Ji, Non-equilibrium evaporation in a vacuum spray flash, in: 4th Therm. Fluids Eng. Conf., 4th
thermal and Fluids Engineering Conference, 2019.
[8] G. Guo, C. Zhu, Z. Ji, CFD simulation of isolated spray flash evaporation with active vapor extraction, in: 5th Therm. Fluids
Eng. Conf., Begellhouse, Connecticut, 2020: pp. 269–278.
[9] L.X. Huang, K. Kumar, A.S. Mujumdar, A comparative study of a spray dryer with rotary disc atomizer and pressure nozzle
using computational fluid dynamic simulations, Chem. Eng. Process. Process Intensif. 45 (2006) 461–470.
[10] F. Xuening, C. Lei, D. Yuman, J. Min, F. Jinping, CFD modeling and analysis of brine spray evaporation system integrated
with solar collector, Desalination. 366 (2015) 139–145.
[11] M.H. Sadafi, I. Jahn, K. Hooman, Cooling performance of solid containing water for spray assisted dry cooling towers,
Energy Convers. Manag. 91 (2015) 158–167.
[12] ANSYS Inc., ANSYS FLUENT 19.0 theory guide, ANSYS Inc., Canonsburg, PA, 2019.
[13] B. Zhang, G. Guo, C. Zhu, Z. Ji, C.-H. Lin, Transport and trajectory of cough-induced bimodal aerosol in an air-conditioned
space, Indoor Built Environ. (2020) 1420326X2094116.
[14] B. Zhang, G. Guo, C. Zhu, Z. Ji, Transport of aerosol by coughing in an air-conditioned space, in: 4th Therm. Fluids Eng.
Conf., Begell House Inc., 2019: pp. 1341–1353.
[15] A.H. Persad, C.A. Ward, Expressions for the Evaporation and Condensation Coefficients in the Hertz-Knudsen Relation,
Chem. Rev. 116 (2016) 7727–7767.
[16] G. Guo, H. Deng, C. Zhu, Z. Ji, Modeling and experimental study of non-equilibrium isolated spray flash evaporation, in:
TFEC2019 Conf., 2020: pp. 1–8.

*Corresponding Author: chao.zhu@njit.edu

11

View publication stats