Renewable and Sustainable Energy Reviews 78 (2017) 466494

Contents lists available at ScienceDirect

Renewable and Sustainable Energy Reviews

journal homepage: www.elsevier.com/locate/rser

A critical review on bubble dynamics parameters inuencing boiling heat MARK

transfer

Rajiva Lochan Mohanty , Mihir Kumar Das
School of Mechanical Sciences, Indian Institute of Technology Bhubaneswar, Odisha, India 752050.

A R T I C L E I N F O A BS T RAC T

Keywords: The rate of heat transfer during boiling is governed by various bubble dynamics parameters such as bubble
Boiling heat ux departure diameter, active nucleation site density, bubble waiting period, bubble growth period, bubble growth
Boiling heat transfer coecient rate and bubble departure frequency. The study of bubble dynamics during boiling of liquids over a heated
Bubble departure diameter surface is a complex process due to non-linear growth of bubbles. Many studies on bubble dynamics is carried
Active nucleation site density
out by both experimentally and numerically. These studies are carried to propose various empirical and semi-
Bubble waiting period
Bubble growth rate
empirical correlations for determination of bubble dynamics parameters. In the present paper, a comprehensive
Bubble growth period review is carried out to compile various correlations proposed for determination of bubble dynamics
Bubble departure frequency parameters. The correlation for determination of boiling heat ux or boiling heat transfer coecient based
on these bubble dynamics parameters are reported. This is done to identify important bubble dynamics
parameters aecting boiling heat transfer process. Further, factors aecting bubble dynamics parameters such
as eect of thermo-physical properties, heat ux, liquid sub-cooling, wall superheat, contact angle, gravity,
cavity spacing and pressure are also given to get an insight into the correlation proposed for determinations of
bubble dynamics parameters. The present review article proposes the importance of development of generalized
boiling heat transfer correlation using bubble dynamics parameters.

1. Introduction
Boiling is an incredible and challenging area for carrying out
research activities to fulll the requirements of various industrial/
research sectors such as thermal and nuclear power plants, refrigera-
tion and air conditioning units, rocket motors, electronics cooling, food
processing and other aide industries. In boiling heat removes from the
heated surface by means of vapor bubbles generated due to change of
phase. Since vapor bubbles are accountable for degree of dissipation of
heat by boiling heat transfer, it is signicant to understand the
mechanism behind the formation, growth and detachment from the
boiling surface. The detachment of the vapor bubble from the surface is
inuenced by the various forces acting on growing bubble. Out of
various forces acting on the vapor bubble during its growth, the major
force responsible for attachment of vapor bubble to the heated surface
is surface tension whereas the buoyancy force is accountable for
detachment of the vapor bubble.
During initial growth stage of bubble the surface tension force
dominants the buoyancy force. However, with increase in size of the
vapor bubble, buoyancy force gradually supersede the surface tension
force. When the magnitude of buoyancy force becomes much higher
than the surface tension force, the vapor bubble detach from the heated
surface. So, the formation, growth and detachment of the vapor bubble
and the rate of heat transfer thereof requires the knowledge of bubble
dynamics parameters. These bubble dynamics parameters are nuclea-
tion site density, bubble departure diameter, bubble waiting period,
bubble growth period and bubble departure frequency. The determina-
tion of boiling heat transfer coecient can be done by using either
empirical or semi-empirical correlations or correlations developed
using bubble dynamics parameters.
Many empirical correlations for determination of boiling heat
transfer coecient are available in the literature [14]. However,
almost all these correlations predict own experimental results reason-
ably but fails to predict the experimental data of other investigators [5
10]. On the other hand, the correlations developed using bubble
dynamics parameters are based on fundamentals of boiling heat
transfer. It is, therefore, expected that such correlations can predict
experimental data of dierent investigators with minimum errors. This
is important to note that the accuracy of such generalized boiling heat
transfer correlations are entirely dependent on the correlations devel-
oped to predict the bubble dynamics parameters. This is because the
correlations developed for prediction of bubble dynamics parameters

Corresponding author.
E-mail addresses: rajivamohanty@gmail.com (R.L. Mohanty), mihirdas@iitbbs.ac.in (M.K. Das).

http://dx.doi.org/10.1016/j.rser.2017.04.092
Received 20 January 2016; Received in revised form 31 March 2017; Accepted 27 April 2017
Available online 05 May 2017
1364-0321/ 2017 Elsevier Ltd. All rights reserved.
R.L. Mohanty, M.K. Das Renewable and Sustainable Energy Reviews 78 (2017) 466494

Nomenclature tl bubble lift-o time (s)

tw bubble waiting period (s)
Ar Archimedes number [(g/l2)( / l g)3/2 ] tg bubble growth period (s)
A area of heated surface (m2) t+ non-dimensional time

G Average mass velocity of bubble ( 6 Dd3 v f n s) Vb bubble velocity (m/s)
Bo Bond number (gD2d / ) V volume (m3)
Bom Modied Boiling number(q/Ghlv) Vd bubble departure volume
Ca Capillary number (v U/ cos ) V+ non-dimensional volume=(V /Vd)
CD drag coecient + non-dimensional density dierence
Ce empirical constant =20/3
Cs surface uid combination factor Greek symbols
Cp specic heat at constant pressure (J/kg K)
Dd bubble departure diameter (m) contact angle ()
D bubble diameter (m) density (kg/m3)
Dc cavity diameter (m) dierence
Dd,dry diameter of dry area under bubble (m) dimensionless surface roughness parameter
D1 diameter of tube (m) dynamic viscosity (kg/m s)
Ds diameter of heating surface half cone angle ()
Df fractal dimension kinematic viscosity (m2/s)
dD/dt bubble growth rate mean half cone angle ()
f bubble departure frequency (1/s) parameter in Eq. (13)
f (c) bubble volume factor statistical parameter
g gravitational acceleration (m/s2) T Taylor most dangerous wave length
gc gravitational correction factor surface-liquid interaction parameter
h heat transfer coecient (W/m2 K) surface tension (N/m)
hlv latent heat of vaporization (J/kg) thermal boundary layer thickness (m)
Ja Jakob number ( l Cpl (TwTsat)/ v hlv ) thermal diusivity (kl / l Cpl ) (m2/s)
J latent heat removal by per bubble (J/kg ) volumetric expansion coecient (1/K)
k thermal conductivity (W/m K)
m empirical constant Superscripts
Nu Nusselt number (hD/k)
N number of nucleation sites + non-dimensional
ns active nucleation site density (sites/m2)
na average cavity density (sites/m2) Subscripts
n +s non-dimensional active nucleation site density
P pressure (MPa) c cavity
Pr Prandtl number (l Cpl /k l ) l liquid phase
q heat ux (W/m2) max maximum
R gas constant (J/kg K) me microlayer evaporation
Ra surface roughness (m ) min minimum
Rc cavity radius (m) nc natural convection
R c+ non-dimensional critical cavity radius b boiling
r bubble radius (m) r reduced property
rb radius of the liquid microlayer under bubble (m) sat saturation condition
rt instantaneous bubble radius (m) tc transient conduction
r+ non-dimensional bubble radius s heating surface
s standard deviation v vapor phase
T temperature (K) w wall
t time (s) bulk

are again inuenced by various parameters. Owing to the above facts,
initially a review is carried out on existing boiling heat transfer
correlations to predict boiling heat transfer coecient and boiling heat
ux based on bubble dynamics parameters. These correlations are used
to predict experimental data of other investigators to nd its appro-
priateness as a generalized correlation. The bubble dynamics para-
meters addressed in these correlations are then identied. Based on
these information, a comprehensive review is carried out on these
bubble dynamics parameters and factors aecting them. This include
review of existing correlations, their comparison in predicting experi-
mental data of dierent investigators etc. Thus, this review may help
the researchers to develop generalized boiling heat transfer correla-
tions based on bubble dynamics parameters which may be applicable
for wide range of operating parameters.
Fig. 1 shows inter-functional dependency of bubble dynamics
parameters in the determination of boiling heat transfer coecient.
The ow chart is developed by considering the two independent factors
i.e. wall superheat and force balance on the vapor bubble during
growth and detachment from the heated surface. The other bubble
dynamics parameters are correlated with one another with a view to
predict the boiling heat transfer coecient. This ow chart gives the
basic and root level idea about the interconnections of bubble dynamics
parameters with boiling heat ux or boiling heat transfer coecient.
2. Boiling heat transfer coecient and boiling heat ux
The important objective in boiling process is to determine the
boiling heat transfer coecient or boiling heat ux. The dierent

467
R.L. Mohanty, M.K. Das Renewable and Sustainable Energy Reviews 78 (2017) 466494

Fig. 1. Flow chart showing the interdependency of various parameters for developing the semi-analytical model.

models of bubble heat transfer during nucleate pool boing based on the boiling heat transfer. Therefore, development of a better generalized
mechanisms of enhanced convection, transient conduction, microlayer correlation for boiling heat transfer coecient with minimum predic-
evaporation and contact line is reviewed by Kim [11]. Therefore, tion error is essential which can be used comfortably in industry
numerous correlations are available in the literature for determination involving boiling heat transfer processes. Following few correlations
of boiling heat transfer coecient based on experimental results. These are available in the literature for the determination of boiling heat
correlations are either empirical, semi-empirical or theoretical in transfer coecient in terms of bubble dynamics parameters besides the
nature. Most of the empirical or semi-empirical correlations are other properties.
developed considering the parameters such as heat ux, pressure, McNelly [cf.9] proposed a correlation of pool boiling heat
geometrical dimensions, thermo-physical properties and surface-liquid transfer coecient for pure substances including bubble departure
combination factor in terms of dimensional or dimensionless numbers. diameter, heat ux, system pressure and thermo-physical proper-
Though these correlations predict own experimental data with reason- ties of uids.
able accuracy but they fails to predict experimental data of other
k qD 0.69 PD 0.31 0.33 Cpl l 0.69
investigators. However, it is a well-known fact that the amount of heat h = 0.225 l d d l v
removed during the boiling process is basically due to the formation of Dd hlv l l v kl (1)
vapor bubbles on the heating surface. So, the boiling heat transfer
Stephan and Preusser [12] obtained an empirical correlation of
coecient is inuenced by bubble departure diameter, active nuclea-
Nusselt number considering bubble departure diameter and thermal
tion site density, bubble departure frequency, bubble growth and
diusivity during pool boiling of pure substances. This developed
waiting period. Hence, a better correlation of boiling heat transfer
correlation helps to determine the boiling heat transfer coecient.
coecient must include the bubble dynamics parameters which is able
to describe the actual phenomena of boiling process. Therefore, k qDd 0.67 v 0.156 hlv D2d 0.371 l2 l 0.35 l Cpl 0.16
discussions about the correlations developed using bubble dynamics h = 0.1 l 2
Dd kl Tsat l l Dd kl (2)
parameters are presented below.
Stephan and Abdelsalam [13] developed semi-empirical correla-
2.1. Boiling heat transfer coecient tions of boiling heat transfer coecient based on 5000 existing
experimental data of various uids such as water, hydrocarbons,
Boiling heat transfer coecient values for various surface liquid cryogenic uids and refrigerants for natural convection boiling heat
combinations are necessary for better design of heat transfer equip- transfer. They applied the methods of regression analysis to the above
ment. The costlier experiments for the determination of heat transfer experimental data for developing the correlations for hydrocarbons,
values can be avoided by forming semi-empirical and empirical refrigerants and other uids as follows.
correlations from the experimental data or using fundamentals of For hydrocarbons:

468
R.L. Mohanty, M.K. Das Renewable and Sustainable Energy Reviews 78 (2017) 466494

k 0.5 qDd l v 4.33 hlv D2d 0.248 0.674 0.371 2 0.350

0.297
hlv D2
0.67
kl qDd
h = 0.0546 l v 2 h = 0.74 v 2 d l l
Dd l kl Tsat l l Dd kl Tsat l
l
Dd
(3)
l v 1.73 R P 0.133
For refrigerants: ( )0.1 a (Pr )0.5
l (9)
k qDd 0.745 v 0.581 vl 0.53
h = 207 l kw w Cpw
Dd kl Tsat l l (4) where = kC
l l pl
For other uids: This correlation predicted both own experimental data and litera-
ture data within an error of 16%.
k qDd 0.674 v 0.297 hlv Dd2 0.371 l2 l l v 1.73
0.35
h = 0.23 l 2 Fazel and Mahboobpour [18] conducted a pool boiling experiment
Dd kl Tsat l l Dd l on a rod heater by taking working uid as water/monoethylene glycol
(5) binary solution of various concentrations and heat ux at saturation
temperature at atmospheric condition. They proposed a model to
where
predict the boiling heat transfer coecient by using a model given by
2 0.5 Judd and Hwang [19]. For developing this model, they assumed that
Dd = 0.0146 Fritz[14] uniform temperature is maintained on the boiling surface and used the
g(l v ) (6)
basic Newtons cooling law for measuring the boiling heat transfer.
The above correlations predicted the experimental data of boiling They divided the heat transfer area into two distinct regions in the
heat transfer coecient with in an error of 10.57% and 22.3% for presence of bubbles: (a) area aected by bubbles (Ab) and (b) natural
refrigerants and all other uids respectively. From the above correla- convection heat transfer area (Anc) are shown in Fig. 2. Again they
tions it is concluded that the heat transfer coecient depends upon assumed that bubbles is in spherical size and area aected by bubbles is
bubble departure diameter and heat ux besides the thermo-physical equal to the bubbles projected area. By using the Newtons law of
properties of uid. cooling for individual region, they proposed a correlation of boiling
Jung et al. [15] conducted an experiment on a horizontal smooth heat transfer coecient:
tube of 19 mm outside diameter during nucleate pool boiling of
halogenated refrigerants such as CFC123, CFC11, HCFC142b, h = hnc+( k l l f hnc)(ns) D2d
4 (10)
HFC134a, CFC12, HCFC22, HFC125 and HFC32 at the liquid pool
temperature of 7C. They took the data in the order of decreasing heat where hnc is the natural convection heat transfer coecient and also
ux from 80 kW/m2 to 10 kW/m2 with an interval of 10 kW/m2 . They used the expression for heat transfer coecient aected by bubble
proposed a correlation of boiling heat transfer coecient based on their given by Mikic and Rohsenow [20].
own experimental data in terms of bubble departure diameter, heat Most of the above discussed correlations except the correlation
ux, thermal diusivity, liquid saturation temperature and thermo- proposed by Fazel and Mahboobpour [18] are developed by consider-
physical properties of uids: ing the parameters such as thermo-physical properties of uid,
pressure and bubble departure diameter.
k qDd c 0.33 v 0.25 A comparison of correlations proposed by McNelly [cf.9], Stephan
h = 10 l Pr 0.1 (1Tr )1.4 l v l
Dd kl Tsat l l (7) and Preusser [12] and Stephan and Abdelsalam [13] to predict the
experimental data of Hamzekhani et al. [21], Jamialahmadi et al. [22],
v 0.309 Lakhera et al. [23] and Bhaumik et al. [24] during pool boiling of water
where c = 0.855 Pr 0.437
l are carried out and shown in Figs. 3(a), (b), and 4(a), (b) respectively.
This correlation predicted the experimental data within an error of All the experimental data are at 1 atmospheric pressure except the data
7% for all above halogenated refrigerants. of Jamialahmadi et al. [22] which is at 110 kPa. It is found that the
In an another experiment, Jung et al. [16] investigated nucleate correlation of McNelly [cf.9] predicted all the experimental data within
pool boiling of ammable refrigerants such as propylene (R1270), en error band of 20% 40%. Similarly, the correlation of Stephan
propane (R290), isobutane (R600a), butane (R600), dimethylether and Preusser [12] predicted the experimental data within an error band
(RE170), and dimethylether (RE170). They developed a correlation of of 20% 50%. The correlation of Stephan and Abdelsalam [13]
boiling heat transfer coecient taking heat ux, bubble departure predicted the experimental data of Hamzekhani et al. [21],
diameter and reduced pressure besides the thermo-physical properties Jamialahmadi et al. [22] with an error of 10% and 20% but the
of refrigerants: error band increases signicantly up to 40% and 50% while
k qDd 0.835 Pr1.33 0.33 predicting the experimental data of Lakhera et al. [23] and Bhaumik
h = 41.4 l [log (Pr )]1.52 l v et al. [24] respectively .
Dd kl Tsat 10
l (8) Further, the experimental data of Bhaumik et al. [24] and Das et al.
[7] during pool boing of benzene and isopropanol at 1 atmospheric
This correlation predicted the experimental data within an error of
pressure are used for comparison of above three correlations as shown
5.3% for all above ammable refrigerants.
in Fig. 5(a) and (b). It is also found that the error band increases from
Rao and Balakrishnan [17] performed an experiment on aluminium
block of diameter 28.9 mm during pool boiling of acetone-isopropanol-
water and acetone-methyl ethyl ketone at atmospheric pressure and
standard gravity conditions. They prepared the binary mixtures of
isopropanol-water and metyl ethyl ketone with concentration of each
component of 0.5 mol fraction whereas ternary mixtures with the
concentration range of acetone from 0.1 to 0.9 keeping the concentra-
tions of other two components equal. They proposed a correlation of
boiling heat transfer coecient for pure component including bubble
departure diameter, heat ux, thermal diusivity, surface tension,
dimensionless surface-liquid interaction parameter, pressure, surface
roughness besides the thermo-physical properties of uids: Fig. 2. Heat transfer allocated area [18].

469
R.L. Mohanty, M.K. Das
Fig. 3. Predicted boiling heat transfer coecient verses experimental boiling heat
transfer coecient: (a) Water data of Hamzekhani et al. [21] (b) Water data of
Jamialahmadi et al. [22].
30% to 40% when predicting the experimental data of benzene and
isopropanol using correlation of McNelly [cf.9] and Stephan and
Preusser [12]. However, the correlation of Stephan and Abdelsalam
[13] predicted the experimental data with reasonable accuracy.
From the above comparisons, it is clear that the correlation which is
used to predict the experimental data of one investigation within
reasonable accuracy is not applicable to other experimental data with
the same accuracy. Though the correlations of Stephan and Abdelsalam
[13] predicted almost all experimental data with reasonable accuracy
but the correlation are dierent for dierent uids. Thus, the predic-
tion error varies signicantly from data of one investigator to other
having either same or dierent surface-uid combination. Therefore,
there is a need to develop generalized correlation with minimal
prediction error. The error band may be reduced by developing a
correlation of boiling heat transfer coecient considering the bubble
dynamics parameters such as bubble departure diameter, active
nucleation site density and bubble departure frequency beside ther-
mo-physical properties of uid. The interdependency relations between
bubble dynamics parameters and boiling heat transfer coecient as
470
Renewable and Sustainable Energy Reviews 78 (2017) 466494
Fig. 4. Predicted boiling heat transfer coecient verses experimental boiling heat
transfer coecient: (a) Water data of Lakhera et al. [23] (b) Water data of Bhaumik
et al. [24].
shown in Fig. 1 may be helpful for developing a new correlation of
boiling heat transfer coecient for wide range of data with minimum
predicted error.
2.2. Boiling heat ux
There are many models for boiling heat ux developed by various
researchers. Basically, the boiling heat ux is inuenced by dierent
regions which is created at the time of boiling on the solid surfaces. The
regions dier from each other by means of formation of vapor bubbles
on the heated solid surface. The regions which are developed during the
boiling are the region of single phase heat transfer i.e. natural
convection, the region aected by vapor bubble formation and the
region in between vapor bubble and heating surface. Han and Grith
[25] expressed the total heat ux in terms of natural convection heat
transfer and transient heat conduction due to thermal boundary
reformation. Han and Grith [25] model is modied by Mikic and
Rohsenow [20] by considering the eect of heating characteristics.
R.L. Mohanty, M.K. Das Renewable and Sustainable Energy Reviews 78 (2017) 466494

D = 2Dd , (iii) qnc=q c and (iv) (n s)avg = n s , they proposed an expression of

heat ux during bubble formation in a single active nucleation site as:

Qb A n
= q q c 1 s D2b
ns ns A (11)

where D is the diameter of the area aected by the bubble and q c is

the natural convection heat ux.
Benjamin and Balakrishnan [28] developed a mechanistic model for
boiling heat ux during nucleate pool boiling of pure liquids at low to
moderate heat uxes. They took the following modes of heat transfer
mechanisms during the model development.

(i) Heat transferred by evaporation microlayer (qme).

(ii) Heat transferred by transient conduction during thermal bound-
ary reformation (qtc).
(iii) Heat transferred by turbulent natural convection (qnc).

The total boiling heat ux is given based on the above three heat
uxes as
(qme t g) + (qtct w)
qtotal = +(qnc)
(t g+t w) (12)

where
2 0.27 3
qme = B Ar Ja (l ) 2 t g l n s
10 (13)

kl Cpl
qtc = 2 l (n sa)(TwTsat)
tw (14)

qnc = h[1 (n s)a](TwTsat) (15)

kw w Cpw Dd,dry 2
where = k C , = 1 ( D ) -
l l pl
B = constant=1.55 for water,CCL 4 and n hexane -
=1/1.55 for n pentane and aceton
They used own developed nucleation site density correlation for
measuring the active nucleation site density which is valid only in the
low to moderate heat ux.
Yu and Cheng [29] developed a fractal model for nucleate pool
boiling heat transfer based on fractal distribution of sites of nucleation
sites on boiling surfaces. They modied the heat ux due to boiling,
natural convection and micro-layer evaporation underneath the bub-
Fig. 5. Predicted boiling heat transfer coecient verses experimental boiling heat bles given by Mikic and Rohsenow [20] and Dhir [30]. The expression
transfer coecient: (a) Benzene data of Bhaumik et al. [24] (b) Isopropanol data of of total nucleate boiling heat ux as follows:
Das et al. [7].
qtotal = qb + qme + qnc or,
2Df T T D Df +1
They proposed the total heat ux model in terms of heat ux due to qtotal = Cb l w sat c,max 1
c,max
D1
transient conduction and heat ux due to natural convection where Df +1 3 Tw T Dc,min
transient conduction is expressed in terms of bubble departure
3 Dc,max T T Df Dc,max Df 1
diameter, active nucleation site density, bubble departure frequency, + C me w
1
wall superheat and thermo-physical properties of uid. Judd and 16 l Tw Tsat Df 1 Dc,min
Hwang [19] modied Mikic and Rohsenow [20] heat model by Dc,min 2Df
including micro layer evaporation at the base of the bubble. The micro + 1 K hnc (TwT)
Dc,max (16)
layer evaporation is expressed in terms of vapor volume, active
nucleation site density, liquid density, latent heat of vaporization and
where Cb = K kl l cpl Dd2 (TwTsat) and K is a proportional constant for
bubble departure frequency. Paul and Abdel-Khalik [26] expressed
boiling heat ux by considering heat ux due to phase change, natural bubble departure diameter, C me = (1.55)2 10 Ar 0.27Ja(l )3/2 l hlv , hnc is
convection and forced convection. the average heat transfer coecient for natural convection and
2
1 Dc,max
Barthau [27] conducted pool boiling experiment on horizontal ln
2 Dc,min

copper cylinder taking R114 as working uid at pressures of 1.50, Df = Dc,max
and Dc, max is the average value over all the maximum
1.19 and 2.47 bar and heat uxes up to 7 10 3W/m2 . They divided the ln
Dc,min
total heat ow rate (Q) into two dierent regions i.e. heat ow by active cavities.
natural convection in the presence of bubbles (Q nc) and heat ow by They compared the total heat ux calculated from fractal model
bubble formation(Q b). By considering the assumptions: (i) excess with the experimental data of Wang and Dhir [31] for contact angle of
temperature is measured on the total surface of the test tube, (ii) 90 and 35 and found excellent agreement.

471
R.L. Mohanty, M.K. Das Renewable and Sustainable Energy Reviews 78 (2017) 466494

Sateesh et al. [32] proposed a modied model for pool boiling heat where Na,total the total number of nucleation sites, K is the proportional
transfer in which the eects of bubbles sliding on the vertical heating constant for bubble diameter of inuence and a is the area of inuence
surface is considered. They took four dierent heat transfer mechan- equal to D2d
isms such as latent heat transfer due to microlayer evaporation, They used Hsus [35] model for minimum and maximum cavity
transient conduction due to thermal boundary reformation, natural diameter, the relation between growth period and waiting period as
convection and heat transfer due to sliding bubbles on the vertical proposed by Van Stralen et al. [36] as
heated surface. The total heat ux in terms of modied or corrected t w = 3t g (29)
values is given as:
And the waiting period expression proposed by Han and Grith
qtotal = (qm m m m
me+q tc )+(qmes+q tcs)+q nc (17)
[25] as
where
9 (Tw T)R c 2
1 tw =
qme = Dd3 v hlv n s f 4l (Tw Tsat)(1 + 2 /R cv hlv) (30)
6 (18)
The term (1 + 2 /R cv hlv) in waiting period expression is neglected
1
k l Cpl 2 D2 by using Wang and Dhir [31] cavity radius value (R c=1.1 27.7m ) for
qtc = 3.6 l (TwTsat) n s d t w f pool boiling of saturated water at atmospheric pressure on a copper
t w 4 (19)
surface. By putting all the values in total heat ux equation and got
1
(D3l D3d) v hlv n s f 4l Tw Tsat 2
2
qmes = Df 1+2
6 (20) qtotal = J Dc,max (Na,total) Df 1
Df +2 3 Tw T
1
k l C pl 2 tl Df T Tsat 2 1 1+1
qtcs = 3.6 l
t w
(TwTsat) n st w f t d
d(t)Ub (t)dt
(21)
+ 2a
Df +1
k l l C pl w
Tw T
Dc,max (Na,total) Df 1

where Dc,min 2Df

+ hnc 1 K .(TwT)
Dc,max
k Ja 0.5 (31)
d(t) = 1.55(Ar )0.135 l t
l Cpl (22) Form the above equation, it is concluded that the information of
wall superheat, fractal dimension, contact angle, total number of active
1
tl A 2 nucleation sites, maximum and minimum cavity diameter and thermo-
t d
Ub (t)dt =
N (23) physical properties of uids are required for measurement of total heat
ux during nucleate boiling. This model is not introduced any
qnc = hncAnc (TwTsat) (24) empirical constant and predicted available experimental data reason-
ably.
From the above expression, it is concluded that the information of
Gerardi et al. [37] conducted experiment on a thin electrically
bubble departure diameter, bubble lift-o diameter during sliding,
heated indium-tin-oxide surface during pool boiling of water at atmo-
active nucleation site density, bubble departure frequency and waiting
spheric pressure. They varied the heat ux in discrete steps up to the
period besides the thermo-physical properties of uids are required for
critical heat ux. They determined the heat ux using RPI heat ux
measurement of total heat ux during nucleate boiling. This model
partitioning model given by Kurul and Podowski [38]. This model is
predicted the experimental data of water, refrigerant R134a and
primarily developed for ow boiling which uses Browing [39] scheme
propane within an error of 25%.
for calculating various boiling heat transfer mechanism individually.
Chu and Yu [33] developed a new model for nucleate boiling heat
According to this model, the total boiling heat ux is the sum of latent
transfer of pure liquid based on the fractal distribution of nucleation
heat ux due to evaporation to form the bubbles, quenching heat ux
sites on the heat ting surface at low to high uxes including critical heat
due to reformation of thermal boundary layer and convection heat ux
ux. For developing the model, they assumed that the total heat ux in
not inuenced by vapor bubble:
nucleate boiling is inuenced by the following three mechanisms:
qtotal = q e + q q + q c (32)
(i) Latent heat by bubbles (qme) due to liquid evaporation as
suggested by Moore and Mesler [34]. The latent heat ux can be expressed as:
(ii) Transient conduction (qtc) due to thermal boundary reformation as
Na,total

proposed by Han and Grith [25]. qe = v hlv [ fn D3b,n]

6A n=1 (33)
(iii) Natural convection heat transfer (qnc) as proposed by Yu and
Cheng [29]. They used Han and Grith [25] correlation for quenching heat ux
as:
The total boiling heat ux is given as:
Na,total
2kl (Tw Tsat)
qtotal = (qme . t g + qtc . t w) f + qnc (25) qq = [D2b,n, ( t w, n fn )]
A l n=1 (34)
where
For measuring the convection heat ux they used the correlation
16l Tw Tsat 2
2 1+2
Df given by McAdams [40] as:
qme =J Dc,max (Na,total) Df 1
Df +2 3 Tw T (26) Na,total

q c = 1 (D b,n)2 hturb (TwTsat)
8 Df T Tsat 2 1+1 4A
k l l Cpl w D1 (35)
c,max (Na,total)
qtc = a Df 1 n=1
3 Df +1 Tw T (27)
where n is the nucleation site number, Na,total is the total number of
Dc,min 2Df nucleation sites and hturb is the turbulent natural convection.
qnc = hnc 1 K . (TwT) They incorporated the measured bubble dynamics parameters such
Dc,max (28) as active nucleation site density, bubble departure diameter and bubble

472
R.L. Mohanty, M.K. Das
departure frequency in the above RPI heat ux partitioning model and
found that the model worked reasonably good. They suggested that the
quenching heat ux (q q ) is the dominant contribution to nucleate
boiling heat transfer.
A comparison of heat ux models proposed by Yu and Cheng [29],
Sateesh et al. [32] and Chu and Yu [33] to predict the experimental
data of Wang and Dhir [31] during pool boiling of water on vertical
surface at a contact angle of 90 are carried out and shown in Fig. 6(a).
From the comparison, it is clear that about 80% of experimental data
forecasted through Yu and Cheng [29] and Sateesh et al. [32] falls in
the range of 25% while Chu and Yu [33] model under predict the
experimental data with respect to the given error band.
Fig. 6(b) shows the prediction of experimental boiling heat ux
value using heat ux model given by Sateesh et al. [32]. The experi-
mental data of Wang and Dhir [31], Barthu and Hahne [41], Luke and
Goreno [42] during pool boiling of water on vertical surface,
refrigerant (R134a) on a horizontal tube surface and organic uid
(Propane) on horizontal tube surface are taken for comparisons. The
Fig. 6. (a) Experiment boiling heat ux verses predicted boiling heat ux by dierent
heat ux models (b) Predicted boiling heat ux verses experimental boiling heat ux with
dierent heat ux data.
473
Renewable and Sustainable Energy Reviews 78 (2017) 466494
heat ux model proposed by Sateesh et al. [32] predicted the three sets
of experimental data within an error 25%.
From the above study, it can be seen that most of the correlations for
determination of boiling heat ux is developed decade ago. However, till
now researcher are validating experimental results using either their own
developed or existing correlations. Therefore, a better correlation of boiling
heat ux model needs to be developed in a generalized form which can
predict the wide range of experimental data with reasonable accuracy.
The overviews of these correlation clearly indicates that the deter-
minations of boiling heat ux requires the information about bubble
dynamics parameters besides the thermo-physical properties of uids.
Therefore, the following section narrates dierent form of correlations
available in literature for bubble dynamics parameters and their
comparisons with experimental data etc. in a meticulous manner.
3. Bubble dynamics parameters
3.1. Bubble departure diameter
The bubble departure diameter is an important parameter for the
determination of boiling heat transfer coecient. It is the nal or
equivalent diameter of the vapor bubble after it departs from the
heated surface during boiling. The bubble departure diameter can be
determined either by force balance or experimentation. So, in order to
measure the bubble departure diameter by experimentation, a high
speed camera is required for taking the images or videos of vapor
bubble during boiling. Then the images are analyzed with the help of
image processing software to calculate equivalent diameter of the vapor
bubble. This measurement is possible in nucleate boiling region at low
to moderate heat ux. But at higher heat ux region, the neighboring
bubbles collide with each other and making a larger size which is not an
exact size of vapor bubble departure diameter during boiling. Further,
making force balance on the growing vapor bubble before detachment
from the heating surface, vapor bubble departure diameter can also be
calculated. There are dierent forces acting on the vapor bubble such as
buoyancy force, pressure force, liquid inertia force, vapor inertia force,
drag force, surface tension force. The buoyancy force and pressure
force are positive forces which tend to push the vapor bubble on to the
heating surface whereas inertia force, surface tension force and drag
force are negative forces which always tend to pull the vapor bubble
from the heating surface. So various correlations are developed to
determine the bubble departure diameter with or without any experi-
mentation are discussed below in details.
Fritz [14] proposed a bubble departure diameter (Dd) correlation as
Eq. (6) for pure liquid and also liquid mixture which is frequently and
widely used by researchers with or without modication.
2
Dd = 0.0146
g(l v )
where = 35 for mixtures and 45 for water
This correlation is treated as the basic form for the bubble
departure diameter which is based on the static force balance on a
hemispherical attached bubble on the surface. From this correlations, it
is clear that bubble departure diameter depends upon thermo-physical
properties of the uid, surface tension and contact angle. Various
correlations for the determination of bubble departure diameter during
boiling developed by several researchers for dierent applications are
shown below in Table 1. All the proposed correlations are either simple
with a limited range of applicability or complex in nature due to the
involvement of geometrical constrains.
Based on their experimental observation, Staniszewski [43] mod-
ied Firtz [14] correlation considering the bubble growth velocity
because he found that faster growing bubble having larger diameter at
the time of departure. Borinshansky and Fokin [44] modied Fritz
correlation [14] and proposed a correlation of bubble departure
R.L. Mohanty, M.K. Das Renewable and Sustainable Energy Reviews 78 (2017) 466494

Table 1
Correlations for the prediction of bubble departure diameter.

Authors Correlations Eq. No.

Staniszewski [43] 2 1/2 dD (36)

Dd = 0.0071 g( ) 1 + 0.435
l v dt
Borinshansky and Fokin [44] Dd C C2 1/2 (37)
= 2R + +1
2Rf f (2Rf ) 2

6 0.4 h 1/2
C = g l v h , Rf = 0.0104 g( )
l v l v lv l v
Ruckenstein [45] 3 2 2 g0.5 ( )0.5 2 1/2 (38)
Dd = l l 3/2 l v Ja 4/3 g( )
l v
Cole and Shulman [46] 3 (39)
g 1/2 dD 2
Dd = 0.0208 g( c ) 1 + 0.0025
l v dt

Hatton and Hall [47] 3 2 CD l
v
(40)
Dd4 g(l v ) + Dd gc Dc sin 64 l (Ja)2
l
+ = gc Dc2
8 12 6

Cole [48] 2 1/2 (41)
Dd = 0.04Ja g( )
l v
Cole and Rohsenow [49] 2g 1/2 (42)
Dd = CJa5/4 g( c )
l v
C = 1.5 104 for water and 4.65 104 for others
Van Stralen and Zijl [50] Ja2 2 1/3 2 0.5 (43)
1/4
Dd =2.63 g l 1 +
3Ja
Golorin et al. [51] 1 (44)
0 . 0099 15.6l 3
Dd = + [0.6l Ja]
g(l v ) g(l v )

Kutateladze and Gogonin [52] Dd = 0.5(1 + 10 5C), for C < 0.06 (45)
Ja g ( ) 3/2
1

C = l l 2 v g( )
Pr l

l v

Kacamastafaogullari [54] 1/2 0.9 (46)

Dd = 2.64 105 g( ) l v
l v l
Goreno et al. [55] Dd = c(Ja 4l 4 /g)1/3 [1 + (1 + 2/3Ja )1/2 ]4/3 (47)
Stephan [56] Ja 2 100000 2 1/2
0.5 (48)
Dd = 0.25 1 + g( )
Pr Ar l v
Zeng et al. [57] (49)
3 n/(2n)
Dd = 2 (K2/n/g)[(3/2)Ce n2 + n(n 1)]
4
K and n are determined empirically
Yang et al. [59] l (Cpl Tsat)0.5Prl3/5 (50)
Dd = 3.0557 103 v hlv
2
where = Ja0.3, = c2 / m 1 [ f (c)] 3 ,
1 2
= 1 + 2 (/6Ja ) 3 + /6Ja

3 2 1 3
f (c) 1 4 [1 1c2 ] + 4 [1 1c2 ] ,
c = Rb/Rt,m = 1.4
Lee et al. [60] Dd = (50 27 Ja l / )
2 (51)
Jamialahmadi et al. [22] Dd = (96.75 + 0.01425(q)/ln(q))1 (52)
Kim and Kim [61] 1/2 (53)
Dd = 0.1649 g( ) Ja0.7
l v
Fazel and Shafaee [62] q 1/3 1/2 (54)
Dd = 40 v h / cos g( )
lv v l v
1/3 l
1/2 1/3
Phan et al. [63] 1/2 (55)
Dd = (6 3/2 ) l 1 tan 1/6 g( )
v v l v
(continued on next page)

474
R.L. Mohanty, M.K. Das
Table 1 (continued)
Authors Correlations
Nam et al. [64]
Dd = 24(sin )2 /(2 + 3 cos (cos )3 ) g(
Lamas et al. [65] Cpl T 0.148 l 0.024 0.027
Dd = 0.0027 l h
v lv v l v
Hamzenkhani et al. [21]
vVb 0.25 l Cpl T 0.775 gl
Dd = g ( cos ) h 2
v lv
Suszko and El-Genk [67] Dd = 234 + 81 t g For smooth surface
Dd = 206 + 48 t g For rough surface
Bovard et al. [68] f c4 1/2
Dd = c0 c1 + Jac2 Cac3 g( )
s l v
c0 = 17.952177, c1 = 0.0172742, c2 = 1.285607, c3 = 0.661205,
c4 = 0.025346
diameter by in terms of density ratio and heat transfer coecient. The
coecient of this correlation is modied by Gong [9] taking 0.6 instead
of 6 which is able to predict the experimental data with reasonable
accuracy. Ruckenstein [45], Van Stralen and Zijl [50] and Goreno
et al. [55] modied the original Fritz correlation [14] by including the
thermal diusivity of the solution and non-dimensionless Jakob
number. Goreno et al. [55] modied Van Stralen and Zijl [50]
correlation with a new constant for pure liquids and mixture and this
model is substantially used at high heat ux when the inertia becomes
predominant. Cole and Rohsenow [49] and Cole [48] also modied the
fritz correlation [14] involving Jacob number and replacing the contact
angle with a new constant. Cole and Rohsenow [49] correlation worked
reasonably well for a large variety of ordinary liquids for saturated
boiling whereas Cole [48] correlation worked well for pure liquids and
mixtures besides pure liquids. Cole and Shulman [46] proposed a
correlation by including bubble growth rate in Fritz correlation.
During the detailed study of bubble departure diameter, Hatton and
Hall [47] concluded that bubble diameter is strongly inuenced by
Cavity size and pressure. They proposed a correlation by including the
eect of cavity diameter, thermal diusivity and Jakob number.
Golorin et al. [51] developed a correlation in terms of thermal
diusivity and Jakob number for pure liquids. Kutateladze and
Gogonin [52] derived a correlation in terms of liquid viscosity, Jakob
number and Prandtl number. Based on the available departure
diameter data, Jensen and Memmel [53] modied the correlation of
Kutateladze and Gogonin [52] by changing the constant value to 0.117
instead of 0.5. Kacamastafaogullari [54] modied the Firtz correlation
by including the density ratio. Again the Fritz correlation [14] is
modied by Stephan [56] by including three dimensionless numbers
such as Jakob number, Prandtl number (Pr ) and Archimedes
number(Ar ). This modied correlation showed improvements in pre-
dicting the bubble departure diameter compared to Fritz correlation for
pure liquids and mixtures but is not accurate for electrolyte solutions.
Zeng et al. [57] developed a model of bubble departure diameter by
balancing the forces acting on the growth vapor bubble during boiling
which is similar form used by Klausner et al. [58] for ow boiling. They
assumed that the dominant forces leading to bubble detachment form
the surface would be the unsteady growth force and buoyancy force. A
necessary input to the model is the vapor bubble growth rate. This model
is tested over the pressure ranging from 0.02 to 2.8 bar, Jakob number
ranging from 4 to 869 and gravity ranging from 1 to 0.014 respectively.
This model predicted the experimental data of bubble departure
diameter with relative error of 15%, 26% and 16% for atmospheric
pressure, elevated pressure and reduced gravity respectively.
Renewable and Sustainable Energy Reviews 78 (2017) 466494
Eq. No.
1/2 (56)
)
l v
(57)
e g( )
(59)
1.5
0.05
l g
(60)
(61)
(62)
Yang et al. [59] developed an empirical correlation of bubble departure
diameter including the eect of superheat for water and organic liquids but
not considering the gravity eect. They introduced dimensionless form
characteristic length scale and time scale for describing the bubble dynamic
growth process during their correlation development. Related to the
developed correlation, the eect of superheat on bubble departure diameter
is represented by the eect of superheat on . They observed that the
variation of with Ja ranging from 1 to 1000 is very small.
Lee et al. [60] performed experiment during nucleate pool boiling of
R11 and R113 on a constant wall temperature surface. They carried out a
dimension analysis of bubble growth and departure in characteristic time
and length scale. They obtained time scale from the ratio of latent heat and
the conduction heat transfer rate through the interface. They also derived
the characteristic velocity scale from Rayleigh equation of motion. By
combining characteristic time and velocity scale, they estimated character-
istic length scale. Based on the characteristic time and length scale they
obtained an expression of bubble departure in terms of Jakob number,
thermal diusivity, liquid density and surface tension. They predicted the
experimental data within an average error of 30%.
Jamialahmadi et al. [22] conducted nucleate pool boiling experiment on
outside of a cylindrical stainless test heater at heat ux and bulk
temperature ranging from 10 to 350 kW/m2 and 100 to 110 C respec-
tively. They took electrolyte aqueous solutions such as NaCl, KNO3 and
Na2SO4 as working uids and measured bubble departure diameter. They
also proposed an empirical correlation for bubble departure diameter for
electrolyte solutions by considering the boiling heat ux. They did not
consider the electrolyte concentration and the system pressure in the
development of the correlation. This model predicted the experimental of
bubble departure diameter with an absolute mean error of 5%.
Kim and Kim [61] proposed a correlation for bubble departure
diameter based on the experimental data of other investigators during
sub-cooled, saturated and superheated nucleate pool boiling of dier-
ent uids at atmospheric and sub-atmospheric pressures. They corre-
lated the bubble departure diameter with Jakob number with thermo-
physical properties of uid. The correlation predicted the experimental
data within an error of 20%.
Fazel and Shafaee [62] conducted nucleate pool boiling experi-
ments on a horizontal rod heater with electrolyte aqueous solutions
such as NaCl, KNO3 and Na2SO4 as working uids and measured
bubble dynamics parameters. They found that the bubble departure
diameter increases with increasing boiling heat ux and electrolyte
concentration. They also propped a new model of bubble departure
diameter for electrolyte solutions in nucleate boiling by taking wide
range of heat uxes and concentrations by incorporating the boiling
475
R.L. Mohanty, M.K. Das Renewable and Sustainable Energy Reviews 78 (2017) 466494

heat ux and dynamic viscosity of vapor. This model predicted the a heat ux of 0.5 W/cm2 in saturated condition. They varied surface
experimental data of bubble departure diameter with an error of 2%, roughness for rough surface about 0.211.79 m and 0.039 m for
10% and 5% for NaCl, KNO3 and Na2SO4 respectively. smooth surface. They derived two correlations of bubble departure
Phan et al. [63] also modied fritz correlation and proposed a new diameter in terms of bubble growth period for both smooth and rough
correlation to predict bubble departure diameter for pure liquids and surface. The correlations predicted the experimental data within an
organic uids based on the concept of macro and micro- contact angles error of 15% and 8% for smooth and rough surface respectively.
with the following assumptions. (i) Based on the concept of macro and Bovard et al. [68] carried out experimental investigation on
micro-contact angles, the maximum bubble volume is determined by horizontal cylindrical heater during pool boiling of pure liquids such
force balance. (ii) The bubble gains its volume by evaporation of the as water, ethanol and acetone in heat ux range of 10 100kW/m2 .
liquid micro-layer and the maximum bubble volume is determined by They also considered the heater rod materials like aluminium, stainless
using the conservation of mass. (iii) The mass transfer is assumed to be steel 316A, copper and brass with dierent surface roughness (0.03
negligible during the stage of liquid rewetting. They have correlated the 0.43 m) for copper and aluminium. They derived a non-dimensional
bubble departure diameter with contact angle and thermo-physical correlation for bubble departure diameter using Buckinghams
properties of uid. This model predicted the experimental data of theorem considering the dierent forces acting on the growing bubble
bubble departure diameter within error of 30% at low superheat and in terms of non-dimensional parameters such as bond number,
low sub-cooling and also valid for the values of contact angle varies Capillary number, Jakob number and ratio of thermal diusivity. The
from 0 to 90. However, the proposed correlation unable to predict the error band for the prediction of experimental data lies within 15%.
bubble departure diameter at high superheat and high sub-cooling.
Nam et al. [64] performed an experiment by making an isolated
microcavity on a superhydrophilic surface prepared by forming CuO
nanostructures on a silicon substrate on the superhydrophilic surface.
Their result showed that the bubble departure diameter of water
becomes 2.5 times smaller and the growth period 4 times shorter than
on a silicon substrate. They proposed a model for bubble departure
diameter by balancing the buoyancy and surface tension force acting on
the bubble neglecting the other forces such as viscous drag and liquid
inertia force. Since they have conducted experiment on a super-
hydrophilic surface the importance of wettability on bubble dynamics
has come into sight. Therefore, the contact angle of Fritz correlation
[14] is replaced by the term 24(sin )2 /(2 + 3 cos (cos )3 ) .
Lamas et al. [65] numerically investigated the bubble departure
diameters of water, R143a and R-123 at dierent operating pressures
and contact angles ranging from 0.1 to 10 bar and 10 80respectively.
They proposed a correlation of bubble departure diameter using
Buckinghams theorem in terms of four dimensionless parameters such
as contact angle, Bond number, Jakob number and density ratio. The
model ts the numerical data with absolute average error of 15.9%.
Hamzenkhani et al. [21] performed an experiment during saturated
pool boiling of water, ethanol and various binary mixtures such as
ethanol/water, NaCl/water and Na2SO4/water over a wide range of
concentration. They proposed a correlation of bubble departure
diameter by using Keshock and Siegel [66] force balance model:

Dd = 3 /g (58)

The above correlation predicts the experimental data and proposed

models available in the literature with an error of more than 60%
because of the following reasons.
(i) The number of nucleation sites increased at high heat uxes as
result increases the probability of a single vapor bubble from more than
one adjacent sites. (ii) Bubble departure diameter is signicantly
aected by lateral motion of liquid which increases at higher heat ux.
Due to above reasons, they derived the following expression in terms of
non-dimensional parameters such as Bond number (Bo), Capillary
number (Ca), Jakob number (Ja) and Archimedes number (Ar) based
on the thermo-physical properties of uids. They used Buckinghams
theorem to develop these non-dimensional parameters. All the bubble
departure diameter data is tted in the derived expression with the
tting parameters c1=0.25, c 2=0.775 and c3=0.05.
They concluded that the vapor bubble departure diameter increases
with increase in heat ux and electrolyte concentration because of high
interfacial tension while decreases with increasing ethanol mass
fraction. This model predicted the experimental data of bubble
departure diameter within an average error of 7%. Fig. 7. Experimental verses predicted bubble departure diameter by dierent correla-
Suszko and El-Genk [67] conducted experiment on smooth and tions (a) data of water by Hamzekhani et al. [21] (b) data of water by Jamialahmadi et al.
rough copper surface during nucleate pool boiling of PF-5060 liquid at [22].

476
R.L. Mohanty, M.K. Das Renewable and Sustainable Energy Reviews 78 (2017) 466494

A comparison of correlations proposed by Firtz [14], Hamzekhani

et al. [21], Cole [48], Cole and Rohsenow [49], Borinshansky and Fokin
[44], Van Stralen and Zijil [50], Lee et al. [60], Nam et al. [64], Fazel
and Shafaee [62], and Phan et al. [63] to predict the experimental data
of Hamzekhani et al.[21], and Jamialahmadi et al. [22] during pool
boiling of water is carried out and shown in Fig. 7(a) and (b)
respectively. The above mentioned correlations are also used to predict
the experimental data of Hamzekhani et al. [21], Fazel and
Mahboobpour [18] and McHale and Garimella [69], during pool
boiling of ethanol, pure monoethyleneglycol and FC-77 as shown in
Figs. 8(a), (b) and 9(a) respectively.
The correlation of Borinshansky and Fokin [44], Stephan [56],
Fazel and Shafaee [62] and Hamzekhani et al. [21] predicted the
experimental data of water within an error band of 30% while
80 90% of data forecast through the correlation of Cole [48] and
Lee et al. [60] in the same error band as shown in Fig. 7(a). The
correlation of Cole and Rohsenow [49] over predicts whereas Van
Stralen and Zijil [50] under predicts the experimental data at the set
error band. The correlation proposed by Fazel and Shafaee [62] is able
to predict about 90% of experimental data of Jamialahmadi et al. [22]
within an error band of 30% as shown in Fig. 7(b). However, the
correlation of Van Stralen and Zijil [50] under predicts whereas Cole
[48], Cole and Rohsenow [49], Stephan [56] and Hamzekhani et al.
[21] over predicts the data at same error band.
The similar results are found for experimental data of ethanol as
shown in Fig. 8(a). But the correlation of Lee et al. [60] unable to
predict the experimental data of ethanol within an error band of 30%.
The correlation of Fazel and Shafaee [62] and Hamzekhani et al. [21]
are able to predict the experimental data of pure monoethyleneglycol
within 30% as shown in Fig. 8(b). In the meantime, the correlation of
Cole and Rohsenow [49] over predicts while Van Stralen and Zijil [50]
and Lee et al. [60] under predicts the experimental data of pure
monoethyleneglycol [18] at the same error band.
The correlation of Stephan [56], Fazel and Shafaee [62] and Lee
et al. [60] correlations are able to predict the experimental data of
McHale and Garimella [69] within an error band of 30% whereas the
correlation of Hamzekhani et al. [21] slightly under predicts the
experimental data as shown in Fig. 9(a). The correlation proposed by
Cole and Rohsenow [49] over predicts whereas Van Stralen and Zijil
[50] under predicts the experimental data at the same error band. In all
the cases, the correlation proposed by Fritz [14], Nam et al. [64] and
Phan et al. [63] are giving constant values of bubble departure
diameter because these correlations are based on constant properties
such as static contact angle, density dierence, acceleration due to
gravity and surface tension. All the set of experimental data for NaCl, Fig. 8. Experimental verses predicted bubble departure diameter by dierent correla-
tions (a) data of ethanol by Hamzekhani et al. [21] (b) data of pure monoethyleneglycol
KNO3 and Na2SO4 given by Fazel and Shafaee [62] are predicted by the
by Fazel and Mahboobpour [18].
dierent correlations as presented in Fig. 9(b). Out of all the correla-
tions, Jamialahmadi et al. [12] correlation is able to predict the data of
NaCl, KNO3 and Na2SO4 with an error of 25%, 14% and 13% Yang et al. [59] are not used to predict the experimental data. Because
respectively whereas 2%, 10% and 5% errors are given by the the information of growth rate is required for the correlation [43,46]
correlation of Fazel and Shafaee [62] for respective solutions. whereas the correlation [47] requires the information of both growth
A comparison of correlation proposed by Ruckenstein [45] and rate and cavity size. The correlation of Goreno et al. [55] requires the
Jamialahmadi et al. [22] to predict the experimental bubble departure value of proportionality constant which is varied from experiment to
diameter given by Hamzekhani et al. [21] for water and ethanol, experiment whereas the information of bubble base radius and
Jamialahmadi et al. [22] for water, Fazel and Mahboobpour [18] for instantaneous bubble radius are also required for the correlation
pure monoethyleneglycol and McHale and Garimella [69] for FC-77 proposed by Yang et al. [59]. Therefore, due to these lack of informa-
during pool boiling are carried out and shown in Fig. 10(a) and (b) tion the above correlations are not taken into consideration for
respectively. The correlation proposed by Ruckenstein [45] is unable to comparisons of the experimental data.
predict all sets of experimental data within an error of 60%. However, The overviews of these correlations clearly indicates that the
similarly, the correlation given by Jamialahmadi et al. [22] also unable prediction of bubble departure diameter requires the information
to predict the experimental data within an error of 30% as it does about the thermo-physical properties of the uids, surface tension,
follow the experimental trends of bubble departure diameter because in and surface wettability in terms of contact angle, heat ux, wall
this correlation the bubble departure diameter is inversely proportional superheat and thermal diusivity. A correlation of bubble departure
to the heat ux. diameter may be developed by considering pressure, surface rough-
The other correlations such as Staniszewski [43], Cole and Shulman ness, surface-liquid properties and cavity radius along with thermo-
[46], Hatton and Hall [47], Goreno et al. [55], Zeng et al. [57] and physical properties of uid which may lead to the prediction of the

477
R.L. Mohanty, M.K. Das
Fig. 9. Experimental verses predicted bubble departure diameter by dierent correla-
tions (a) data of peruorinated hydrocarbon (FC-77) by McHale and Garimella [69] (b)
data of electrolyte solutions by Fazel and Shafaee [62].
results with minimum error and also applied to all surface-uid
combination for saturated boiling of liquids.
3.2. Active nucleation site density
Nucleation is a process in which vapor bubbles are formed at the
surface liquid interface during boiling. In general, there are two types
of nucleation processes such as homogeneous nucleation and hetero-
geneous nucleation. The formation of a vapor-liquid interface in
superheated liquid in the absence of pre-existing gas, vapor nuclei
and heating surface refers to the homogeneous nucleation. Whereas the
heterogeneous nucleation refers to the formation of a vapor-liquid
interface in the presence of gas or vapor on the heating surface. So,
referring to the heterogeneous nucleation, the nucleation site density is
dened as the number of cavities present on the heating surface from
which vapor bubbles start growing per unit heating surface area. There
are dierent shapes and sizes of cavities present on the heating surface
depending upon the type of manufacturing process. During hetero-
geneous boiling for a given heat ux, there exist a minimum cavity
radius above which all the cavities are capable of generating vapor
478
Renewable and Sustainable Energy Reviews 78 (2017) 466494
Fig. 10. Experimental verses predicted bubble departure diameter for all sets of uids
(a) Ruckenstein [45] correlation (b) Jamialahmadi et al. [22] correlation.
bubble, known as active nucleation site. So, the number of cavities
capable of forming vapor bubble per unit heating surface area is called
as active nucleation site density.
The determination of active nucleation site density along with
bubble departure diameter may resulted in determination of boiling
heat ux or boiling heat transfer coecient. Several correlations are
developed by various researchers for the determination of nucleation
site density under dierent sets of experimental conditions as shown in
Table 2. A brief discussion about these correlations are given below.
Gaertner and Westwater [70] conducted pool boiling experiments
on a horizontal at copper surface with aqueous solution of nickel salts
containing 20% solids as working uid at atmospheric pressure. They
varied the heat ux and wall superheat ranging from 24.2 to
1687.7 kW/m2 and 8.1to 102.67 C . They found that the nucleation
site density is proportional to square of heat ux.
Micki and Rohsenow [20] rst correlated the active nucleation site
density with cavity diameter present on the boiling surface and
expressed the active nucleation site density as a function of cavity
diameter available on commercial surfaces with an exponent. They
R.L. Mohanty, M.K. Das Renewable and Sustainable Energy Reviews 78 (2017) 466494

Table 2
Correlations for the prediction of nucleation site density.

Autors Correlations Eq. No.

Gaertner and Westwater [70] ns q2.1 (63)

Micki and Rohsenow [20] ns [Dc,max /Dc]m (64)
Dc = 4Tsat /v hlv (Tw Tsat), m = 6.5
Bier et al. [72] ln ns = ln(nmax)[1 (Dc /Dc,max)m ] (65)
For Freon-115 and Freon-11, the value of m is 0.42 and 0.26 for copper surface.
Cornwell and Brown [73] ns (Tw Tsat)4.5 (66)
Paul and Abdel-Khalik [26] ns = 1.027 103q + 15.74 (67)
Kocamustafaogullari and Ishii [74] n+ +
s = f ( )R c
+(4.4) (68)
n+ 2
s = nsD d ,
f (+) = 2.157 107+(3.2) (1 + 0.0049+ )4.13
+ (l v )/v , R+
c = 2R c/Dd
Yang and Kim [75] /2 1 Rs (69)
ns = ns exp{()2 /(2s2)} d exp(R c)dR c
0 2s Rc
Wang and Dhir [31] ns = 5.0 10 5 (1
cos )D6.0 (70)
c
Benjamin and Balakrishnan [77] 1 kl l Cpl (71)
ns = 218.8(Pr )1.63 0.4 (Tw Tsat)3
kww Cpw

= (kww C pw /k l l C pl )1/2 ,
= 14.5 4.5(Ra P/)+0.4(Ra P/)2
Sakashita and Kumada [78] m (72)
ns = Cs [(Ja R c)3/10 (1/R c)]

used Hsu and Graham model [71] for calculating the cavity diameter
which is depends upon the wall superheat.
Bier et al. [72] conducted pool boiling experiment on horizontal
surface and proposed a correlation of active nucleation site density as a
function of the ratio of cavity diameter to maximum cavity diameter
present on the boiling surface. Cornwell and Brown [73] performed
boiling experiment on a copper surface varying the surface conditions
from smooth to rough by taking water as working substance at
atmospheric pressure. They found that active nucleation site density
is proportional to the wall superheat with a power of 4.5. They also
related the total nucleation sites present on the surface with cavity
diameter and found that total nucleation sites site is inversely propor-
tional to the square of cavity diameter.
Paul and Abdel-Khalik [26] conducted experiments on pool boiling
of statured water at atmospheric pressure along an electrically heated
horizontal platinum wire. They determined the active nucleation site
density using high speed photography. They expressed the nucleation
site density as linear function of boiling heat ux.
Kocamustafaogullari and Ishii [74] proposed a correlation of active
nucleation site density in pool boiling based on the experimental data
available in the literature. This correlation also applied to few forced
convection nucleate boiling data. They related active nucleation site
density in dimensionless form as a function of dimensionless minimum
cavity size and density ratio. They used Hsu and Graham model [71]
and Fritz [14] correlation for measuring the cavity radius present on
the surface and the bubble departure diameter respectively.
Yang and Kim [75] derived a correlation for nucleation site density
taking the Poisson distribution for cavity size and the normal distribu-
tion for cone angle of the cavities. They also used Bankos [76] criteria
to measure the number of active nucleation sites. By combining the
probability distribution functions and Bankos [76] criteria, they
expressed the nucleation site density with the average of nucleation
site density presented on the surface. The statistical parameters used in
the expression depends on the surface preparation procedure and
material of the surface.
Wang and Dhir [31] conducted an experiment during pool boiling
of water on vertical copper surfaces for contact angle of 18, 35 and
90 respectively at atmospheric pressure. They have correlated active
nucleation site density with contact angle and cavity diameter. This
correlation indicates that wettability has a strong inuence on active
nucleation site density. This correlation is valid for contact angle rage
from 18 to 90 and the values of cavity diameter smaller than 5.8 m.
Benjamin and Balakrishnan [77] performed an experiment during
nucleate pool boiling of saturated pure liquids such as distilled water,
carbon tetrachloride, n-hexane and acetone respectively at low to
moderate heat uxes. They used the test specimen of stainless steel
and aluminium with dierent roughness. They proposed a correlation
of active nucleation site density in terms of non-dimensional surface
roughness parameters, Prandtl number, wall superheat and a surface
liquid interaction parameter. This correlation predicted the experi-
mental data of active nucleation site density within an error of 30%.
Sakashita and Kumada [78] determined the value of active nuclea-
tion site density as function of cavity radius and non-dimensional
Jakob number express in terms of surface temperature. The constant Cs
is measured from the experimental data and the cavity size is
determined by using Hsu and Graham model [71].
Basu et al. [79] performed sub-cooled ow boiling experiments on
at plate copper surface and a nine-rod (zircalloy-4) bundle. They used
CCD camera for taking the high speed motion pictures of heater surface
during boiling. From these pictures they measured the active nuclea-
tion site density. The experimental result matched well with Wang and
Dhirs correlation [31] for wall superheat higher than 15C but at lower
superheat Wang and Dhirs correlation is unable to predict the result.
So, based on their experimental data, they proposed two correlations
for nucleation site density for two range of wall superheat:
n s = 0.34 10 4 (1 cos ) T2.0 , TONB <T < 15C (73)
n s = 3.4 101 (1 cos ) T5.3, 15C T (74)
From the above correlations, it is found that nucleation site density
depends on static contact angle and wall superheat but not depend on
ow rate and liquid sub-cooling.
Hibiki and Ishii [80] formulated one-dimensional interfacial area
transport equation for ow boiling by considering the homogeneous
and heterogeneous bulk liquid nucleation, wall cavity nucleation,
bubble collapse, bubble breakup, and coalescence, and the bubble
expansion or shrinkage. Based on the derived interfacial area transport
equation, they realized the signicance of the active nucleation site
density and proposed a theoretical model for active nucleation site
density. They also used the basic concept of Yang and Kim [75] and
considering the four physically sound functions such as cavity number
distribution function of the cavity radius, the number of cavity in a
certain range of the cavity radius, the cavity number distribution
function of the half cone angle and the number of cavity in a certain

479
R.L. Mohanty, M.K. Das Renewable and Sustainable Energy Reviews 78 (2017) 466494

range of the half cone angle. The developed model for active nucleation superheat, contact angle and statistical parameters besides the thermo-
site density is given as: physical properties of uid. The few forms of correlations given by
Micki and Rohsenow [20], Bier et al. [72] are not possible to predict the
2
n s = n s 1 exp 2 exp f(+ ) 1 experimental value because of unknown parameters such as maximum
8 R c (75) cavity size present on the surface and average nucleation site density
corresponds to zero cavity size present on the surface. Besides these,
where n s = 4.72 10 5sites/m2 , = 0.722rad , = 2.50 106 m
some of the correlations given by Gaertner and Westwater [70], Micki
2{1 + (v / l )}/Pl and Rohsenow [20] and Cornwell and Brown [73] are function
Rc =
exp{hlv (Tw Tsat)/RTw Tsat} 1 (76) relationship with the heat ux and cavity size ratio. So it is dicult
to nd out as the corresponding proportional constant as it is varied
f(+ ) = 0.01064 + 0.48246+ 0.22712+2 + 0.05468+3 , + according to surface characteristics. Yang and Kim [75] correlation is
not used to predict the experimental data as it requires the addition
= log(*), * =
v information about mean value of cone half angle and statistical
parameter. But the correlations given by Paul and Abdel-Khalik [26],
Range of parameters for validating the correlation:
Kocamustafaogullari and Ishii [74], Benjamin and Balakrishnan [77],
kg kg Wang and Dhir [31], Basu et al. [79] and Hibiki and Ishii [80] are
0 2 < G < 886 2 , 0.101MPa < P < 19.8MPa, 5 < < 90,
ms ms based on well-known parameters. So, these correlations are taken into
1.00 10 4
sites
< n s < 1.51 1010
sites consideration for predicting the dierent experimental data.
m2 m2
A comparison of correlations proposed by Paul and Abdel-Khalik
From the above model it is concluded that the nucleation site
density is a function of critical cavity size and the contact angle. This
model predicted the experimental data of active nucleation site density
within an error of 28.5%.
Based on the statistical property in boiling, Xiao et al. [81]
proposed a theoretical model for nucleation site density. They have
obtained the cumulative size distribution of active nucleation site
density based on Wang and Dhirs correlation [31]. They expressed
the total number of nucleation sites from the minimum active cavity to
the maximum active cavity as:
R c,min 6
n s,total = 7.8125 1029 (1 cos )R c,min6 1
R c,max (77)
If the distribution of the active nucleation site satises the prob-
ability theory, the above equation is reduced as:
n s,total = 7.8125 1029 (1 cos )R c,min6 (78)

T 2
Ts 4C3
R c,max,R c,min = 1 1 sat

C1 Tw Tw Tw
(79)
where
2Tsat (1+cos ) k
= , C1 = , C3 = 1 + cos , = l
v hlv sin hnc

g(Tw T) 2 1/3
hnc = 0.14 l Cpl l
Turbulent region
l (80)

g(T T) l3
1/4
hnc = 0.54 l Cpl w
Laminar region
Al (81)
where Tw is the value of (Tw T) and Tsat is the value of (Tsat T).
From the above model it is concluded that the active nucleation site
density is the function of contact angle and minimum cavity radius.
This model does not include any empirical constant and predicted the
experimental data reasonably available in literature.
Suszko and El-Genk [67] used an expression for calculating the
active nucleation site density in terms of own experimental data of
bubble departure diameter, bubble departure frequency, the fraction
increase in wetted surface area and Jakob number as follows:
6q
ns = Ja
D3d v hlv f (1 + Ar ) (82)
The various forms of correlations for active nucleation site density
which are listed in Table 2 are based on surface characteristics (such as Fig. 11. Experimental verses predicted active nucleation site density (a) by Paul and
surface roughness, maximum and minimum cavity size), heat ux, wall Abdel-Khalik [26] correlation (b) by Kocamustafaogullari and Ishii [74] correlation.

480
R.L. Mohanty, M.K. Das
Fig. 12. Experimental verses predicted active nucleation site density (a) by Basu et al.
[79] correlation (b) by Hibiki and Ishii [80] correlation.
[26] and Kocamustafaogullari and Ishii [74], Basu et al. [79] and Hibiki
and Ishii [80] to predict the experimental data of active nucleation site
density for water at dierent contact angles of 90, 35 and 18 given by
Wang and Dhir [31] are carried out and shown in Figs. 11(a), (b), and
12(a), (b) respectively. Form the comparison graphs, it is found that
the correlation of Kocamustafaogullari and Ishii [74], Basu et al. [79]
and Hibiki and Ishii [80] over predicts the experimental data while the
correlation proposed by Paul and Abdel-Khalik [26] under predicts as
the error band is set to 60%.
Similarly, the experimental data of active nucleation site density for
water and monoethyleneglycol given by Benjamin and Balakrishnan
[77] and Fazel and Mahboobpour [18 ] are tested in the correlation
proposed by Paul and Abdel-Khalik [26], Wang and Dhir [31], Basu
et al. [79] and Hibiki and Ishii [80] and shown in Fig. 13(a) and (b)
respectively. It is observed that the correlations of Paul and Abdel-
Khalik [26], Wang and Dhir [31], Basu et al. [79] and Hibiki and Ishii
[80] are unable to predicted both the experimental data within an error
of 60%. In both the cases, all the correlations over predicts the
experimental data except Paul and Abdel-Khalik [26] under predicts
481
Renewable and Sustainable Energy Reviews 78 (2017) 466494
Fig. 13. Experimental verses predicted active nucleation site density by dierent
correlations (a) data of water by Benjamin and Balakrishnan [77] (b) data of pure
monoethyleneglycol by Fazel and Mahboobpour [18].
the experimental data of Benjamin and Balakrishnan [77].
The experimental data of FC-77 given by McHale and Garimella
[69] at 1 atmospheric pressure is used to for comparisons of dierent
correlations as shown in Fig. 14(a). Out of the ve correlations, only
Hibiki and Ishii [80] correlation is able to predict the experimental data
within an error of 60% where as Benjamin and Balakrishnan [77]
correlation is slightly under predicts the experimental data. The
correlations given by Paul and Abdel-Khalik [26] and Basu et al. [79]
under predicts where as Kocamustafaogullari and Ishii [74] correlation
over predicts the experimental data.
In Fig. 14(b) the correlations of Sakashita and Kumada [78] and
Hibik and Isii [80] are used to predict the experimental data of water at
dierent pressure given by Sakashita [82]. Both the correlations are
able to predict the experimental data within an error band of 60%
whereas the average r.m.s error for Sakashita and Kumada [78] and
Hibiki and Ishii [80] correlations are 40.3% and 54.1% respectively.
The discussion on above correlations clearly shows that the
determination of active nucleation site density requires the information
about the thermo-physical properties of the uids, surface tension, and
R.L. Mohanty, M.K. Das Renewable and Sustainable Energy Reviews 78 (2017) 466494

Table 3
Correlations for the prediction of bubble growth rate.

Authors Correlations Eq. No.

Firtz and D = 2.256Ja l t (84)

Ende
[84]
Plesset and D = 3.908Ja l t (85)
Zwick
[85]
Forster and D = 3.544Ja l t (86)
Zuber
[86]
Labuntsov 1 (87)
D = 6.928Ja 2 l t
et al. [87]
Cole and 3 (88)
D = 5Ja 4 l t
Shulman
[88]
Mikic and
t 0.5 (89)
Tw T t

0.5
Rohseno- D = 3.908 1 1+ w w Ja l t
Tw Tsat t t
w [89]
Van Stralen t 0.5 (90)
and D = 3.908b exp Ja l t
tg
Sluyter
b=constant bubble growth parameter
[90]
For hemispherical bubble, b 0.794 and for sphere
b 0.794
Copper [91] D = 3.96
Ja
l t (91)
Pr 0.5

9 (Tw T)R c 2
tw =
4l (Tw Ts)[1+(2/R cv hlv)]

From the above waiting period relation, it is concluded that for a

Fig. 14. Experimental verses predicted active nucleation site density (a) data of FC-77
by McHale and Garimella [69] (b) data of water at dierent pressure by H. Sakashita
[82].
surface wettability in terms of contact angle, heat ux, wall superheat
and cavity mouth diameter. Further, it can also be seen that inspite of
best eort these correlations are unable to predict experimental data
accurately. The above correlations of active nucleation site density
which are listed in Table 3 are the function of either one or maximum
two parameters. It may be suggested that expressing active nucleation
site density as a function of more than two parameters may lead to
better correlation with minimum prediction error.
3.3. Bubble waiting period
Another important parameter for bubble dynamics parameter
which aects the rate of heat transfer from the surface during nucleate
boiling is bubble waiting period. It is dened as the time interval
between a departing bubble from a nucleating cavity to initiation of
next bubble from the same cavity.
Han and Grith [25] proposed an expression for the waiting period
based on the criterion of the bubble nucleation and potential ow
theory as the heating time required for the establish the thermal layer
thickness equivalent with 3/2 times of the cavity radius. By simplifying
the temperature prole equation, they obtained an explicit form for the
waiting period as in Eq. (30):
given cavity radius, the bubble waiting period decreases with increase
in wall superheat.
For a given cavity size, with increasing wall temperature the waiting
time becomes shorter. However, for a given wall temperature, there is a
range of cavity sizes for which the waiting time is nite given by Hsu
and Graham [83]. The experimental the data of cavity radius were is
unavailable. So in order to reduce the diculties for measuring the
bubble waiting time, they developed the minimum bubble waiting time
by putting the uid temperature line and bubble equilibrium tempera-
ture curve tangent to each other as follows:
144(Tw T)Ts2 2
t w,min =
l v2 hlv
2
(Tw Ts)4 (83)
Van Stralen et al. [36] established a relation between bubble
waiting period and bubble growth period for pure liquids as in Eq. (29).
t w = 3t g
They found that bubble waiting period is three times longer than
bubble growth period in the same nucleation cavity.
3.4. Bubble growth rate
The bubble growth rate is also a major parameter which aects the
bubble dynamics parameters. It is dened as the change in bubble size
with respect to time. It is also correlated with bubble departure
diameter given by Staniszewski [43] and Cole and Shulman [46]. The
evaluation of forces such as liquid inertia and drag force acting on the
growing bubble requires knowledge of the bubble growth rate. The
dierent models of bubble growth during nucleate boiling at dierent
conditions are given in Table 3. The bubble growth rate is proportional
to Jakob number, square root of thermal diusivity and square root of
time are found in the correlations proposed by [8486]. Labuntsov
et al. [87] introduced square root of Jakob number whereas Cole and
3
Shulman [88] included 4 th power of Jakob number in their bubble
growth model keeping other two parameters remain same. Sakashita
[82] modied the constant value of correlation proposed by [87] and

482
R.L. Mohanty, M.K. Das Renewable and Sustainable Energy Reviews 78 (2017) 466494

taken 4.898 which is able to predict the bubble growth rate with good departure radius. The model predicts good accuracy when Stefan
accuracy. The bubble waiting period, growth period and Prandtl number is varied within 0.19 to 0.3. The expression is given below:
number are introduced in the bubble growth model which are proposed 1
1 t 3 i
1.67
by Mikic and Rohsenow [89], Van Stralen and Sluyter [90] and Copper
D = 212Dd 0.037
v ste1.15Fo3
[91] respectively. t g f (103)
Mikic et al. [92] proposed an analytical expression of bubble growth
rate in pure liquids by considering the shape of bubble growing form where,
heating surface is spherical. They also assumed that the relationship i = density ratio of tested liqiud, p = density ratio of propane
between the vapor pressure and the temperature is expressed by the
linearized Clausius-Clapeyron equation and kept the vapor density Lesage et al. [96] proposed a mathematical model for bubble
remains constant. growth during nucleate boiling in terms of cavity diameter, bubble
The bubble growth rate is given as: waiting period and Jakob number. They solved the interfacial energy-
mass balance for truncated spherical shape bubble and also assumed
2 +
r+ = [(t +1)2/3 (t+)3/2 1] that the temperature gradient is uniform at the interface. The model is
3 (92)
expressed as:
where
D+ D2 D2c Dc = 7.817Ja t ( t t + 3t w+ 3t w ) (104)
A+
r+ = r
(B+ )2 (93)
3.5. Bubble growth period
2 hlv (Tw Tsat) 1/2
A+ = v

3 l Tsat (94) Bubble growth period is an indication of amount of heat removed
from the heating surface and therefore aects bubble dynamics
12 1/2
B+ = l Ja signicantly. It is dened as the time interval between the vapor
(95) bubble start growing from a cavity until the bubble departs from the
same cavity on the heating surface. In other words it is the time taken
A+ 2
t+ = + t for a bubble to grow from its initial to its departure size before it is
B (96) detached from the heated surface.
Miyatake and Tanaka [93] proposed a bubble growth rate equation Zuber [97] proposed a correlation for bubble growth period in non
in pure liquids in terms of dimensionless bubble radius and dimen- uniform temperature elds by including the bubble departure dia-
sionless time as follows: meter, thermal diusivity of the liquid phase and non-dimensional
Jacob number. The bubble growth period is given as:
2 t+
r+ = 1 + exp [(t ++1)1/2 ] [(t++1)3/2 (t+)3/2 1] D2d
3 3 (97) tg =
16b 2 (Ja )2 l (105)
where
where b is a constant whose magnitude varied between 1 and 3 .
A+ 2 From the above correlation it is found that bubble growth period
r+ = + 2 r
(B ) P0 (98) varied proportionally with square of bubble departure diameter.
By using Plesset and Zwick's [85] bubble growth rate expression,
2 P0 1/2 Hatton and Hall [47] proposed a correlation of bubble growth period
A+ =
3 l T (99) by considering the parameters of Zuber [97] excluding the Jacob
number and including the cavity radius.
12 1/2 1/2
B+ = l Ja l (v hlv)2 Dd R c
2
Tsat (100) tg =
3 8kl Tsat (106)
A+ 2
6R
A+t 2
This correlation indicated that bubble growth period is the function
t+ = + t +c 1 exp
B A 6R c
dependency of square of bubble departure diameter and nucleation
(101)
cavity radius.
where P0 is the initial pressure dierence between interior. Lee et al. [60] evaluated the dimensional bubble growth period
The above expression shows that bubble growth period requires the approximately 60 based on dierent boundary conditions and working
information of nucleation cavity radius, bubble growth delay period uids. Using this dimensionless growth period, they proposed a
(the time at interaction of r = R c ) besides the thermo-physical proper- correlation of bubble growth period as follows:
ties of uids.
Dd
Zhao et al. [94] proposed a correlation for bubble growth consider- t g = 67.5Jal
(107)
ing hemispherical shape of growing bubble situated on the heated
surface. They solved the energy-mass balance at the interface of bubble In this correlation the bubble growth period is proportional to the
and microlayer during their development. The expression is given bubble departure diameter and Jakob number in terms of surface
below: temperature.
4k(Tw Tsat) The summery of the above discussed correlations for bubble waiting
D= t period as well as bubble growth period clearly shows that the
l hlv 0.64Prl (102)
information of nucleation cavity radius and bubble departure diameter
Haustein et al. [95] developed a model for bubble growth in the are necessary for the determination of bubble waiting period and
radial convection dominated region (RCD model) during nucleate growth period. However, the information of bubble waiting period,
boiling based on experimental parametric study. During the model bubble growth period, cavity size and wall superheat along with
development, they considered density ratio, Stefan number and Fourier thermo-physical properties of uids are required for prediction of
number which is in the form of bubble growth period and bubble bubble growth rate.

483
R.L. Mohanty, M.K. Das Renewable and Sustainable Energy Reviews 78 (2017) 466494

3.6. Bubble departure frequency
The bubble departure frequency is another key bubble dynamics
parameter for the determination of boiling heat transfer coecient. In
general, it totally depends on bubble waiting and bubble growth period.
It is dened as the reciprocal of the time period of two nucleation
during nucleate boiling. Experimentally, it is measured by counting the
total number of bubbles that emerged from a cavity during 1 scale of
recording time. The frequency of vapor bubble emission at a single
cavity during boiling may be written as:
f = 1/(t w+t g)
The bubble departure frequency depends upon wall superheat,
thermo-physical properties of uid, phase contact angle, cavity size and
interaction between neighboring bubbles. A summary of existing
correlations are discussed below. Jakob and Linke [98] correlated the
bubble departure frequency in terms of bubble velocity, bubble
departure diameter, bubble waiting period and bubble growth period.
Jakob and Fritz [99] found that the product of bubble departure
frequency and departure diameter is constant during pool boiling of
water and liquid nitrogen. Jakob [100] again modied the product of
bubble departure frequency and departure diameter by including the
thermo-physical properties of working uid and surface tension. He is
assumed that bubble growth period and bubble waiting period are
equal. Peebles and Garber [101] modied Jakob [100] correlation by
including the bubble waiting period and growth period. McFadden and
Grassman [102] given that the product of bubble departure frequency
and square root of bubble departure frequency is a constant quantity.
Zuber [103] proposed a correlation which is the modication version of
Jakob [100] by taken into consideration the bubble waiting period is
equal to the bubble growth period.
Hatton and Hall [47] derived an expression where the bubble
departure frequency and square of bubble departure diameter is
proportional to liquid thermal diusivity. Cole [cf.111] obtained a
(108)
relation between bubble departure diameter and departure frequency
in terms of gravity and density ratio in hydrodynamic region by force
balance of buoyancy and drag forces. But surface tension is replaced by
a constant of (4/3) and the power of bubble departure diameter with a
value of 0.5 based on the experimental data. The mathematical forms of
the presented correlations are mentioned in Table 4.
Based on the experimental data Ivey [104] suggested the three
dierent correlations in three dierent regions such as hydrodynamic
region, transition region and thermodynamic region.
(i) In the hydrodynamic region two forces such as buoyancy and
1 1
drag forces were acting on the growing bubble: f Dd2 = 0.9g2 . (ii) The

Table 4
Correlations for the prediction of bubble departure frequency.

Autors Correlations Eq. No.

Jakob and Linke [98] tg (109)

f Dd = Vb
(tg + tw)
Jakob and Fritz [99] f Dd = 0.078 (110)
Jakob [100] g( ) 1/4 (111)
f Dd = l v
l2
Peebles and Garber [101] tg g(l v ) 1/4 (112)
f Dd = 1.18
(tg + tw) l2
McFadden and Grassman [102] f Dd 0.5 = 1.75 (113)
Zuber [103] 1.18 g(l v ) 1/4 (114)
f Dd =
2 l2
Hatton and Hall [47] f Dd 2 = 284.7l (115)
Cole [cf.111] 1 (116)
4g(l v ) 2
f Dd 0.5 = 3
l
Ivey [104] 1 1 (117)
(a) fDd2 = 0.9g 2
3 1 (118)
(b) f Dd 4 = 0.44g 4
Mikic and Rohsenow [89] 1 1 (119)
1 4 tg 2 tg 2
f 2 Dd = Ja 3 l + 1 + 1
tg + tw tg + tw

Malenkov [105] Vb q (120)
f Dd = 1
Vb v hlv
1
D g( ) 2 2
Vb = d l v +
2(l + v ) Dd (l + v )

Katto and Yokoya [106] 1 3 (121)

3 2 q 5 4(l v ) 5
f= T
4l hlv g (l v )
T = 2 [3 /(l v )]0.5 , = 11/16
Stephan [107] 1 (122)
1 g
4 2

f Dd = D +
2 d l gDd

Kumada and Sakashita [108] 5 (123)

g ( ) 9 1
f = 0.215 l v /(l Ds3) 9
l
Sakashita and Ono [109] 2
1 (124)
g ( ) 3 3
g (l v ) 2 4 0.25
f = 0.6 l v l l
l l 3

Miglani et al. [110] f + = 28.89Bom + 116.8 (125)

Bom = q /Ghlv , G = Dd3 v fns
6
Hamzekhani et al. [111] 0.25g0.75 q 0.44 0.25g0.75Dd 0.88 (126)
f = 0.015
0.25 0.25g0.75 0.75 0.5

484
R.L. Mohanty, M.K. Das
transition region in which the buoyancy force, drag force and surface
1 1
tension force were dominant bubble: f Dd2 = 0.44g2 . (iii)
Thermodynamic region in which bubble growth is dominant:
f D2d = constant .
From the experimental observation, it is clear that the relationship
Vb = f Dd is approximated valid over the range of bubble diameters:
0.02 > Dd > 3cm . The relationship (Vb/ f Dd) is not incorporated into
accurate theoretical analysis of the boiling phenomenon because
experimental pattern shows an approximate relationship.
Mikic and Rohsenow [89] proposed a correlation in which the
product of root-frequency and bubble departure diameter is equal to a
term containing the ratio of growth period to bubbling period, thermal
diusivity of the liquid phase and a non-dimensional Jakob number in
terms of surface temperature.
Usually, the bubble departure frequency is dierent at hydrody-
namic region and thermodynamic region found during boiling.
Malenkov [105] proposed a new form of correlation for the entire
Fig. 15. Experimental verses predicted fDd (a) Hamzekhani et al. [111] data (b) Fazel and Mahboobpour [18] data (c) McHale and Garimella [69] data.
485
Renewable and Sustainable Energy Reviews 78 (2017) 466494
region in which the product of departure frequency and bubble
departure diameter equal to the quantities contains bubble velocity,
heat ux and thermo-physical properties of working uid. Katto and
Yokoya [106] developed a correlation of bubble departure frequency
for both isolated and coalesced bubble regimes using Taylor most
dangerous wave length, heat ux and thermo-physical properties of
uid such as surface tension, liquid-vapor density and latent heat of
vaporization. This correlation also contains a constant value of 11/16.
Stephan [107] obtained a product of bubble departure frequency and
diameter in terms of bubble departure diameter, acceleration due to
gravity, surface tension and liquid density.
Kumada and Sakashita [108] developed a correlation of bubble
departure frequency for coalesced bubble regime based on density
ratio, kinematic viscosity and size of the heating surface. Sakashita and
Ono [109] proposed a correlation of bubble departure frequency for
high heat ux during saturated pool boiling of water at high pressure
on a rectangular plate. The expression contains the information of
density ratio, kinematic viscosity and surface tension of the tested uid.
R.L. Mohanty, M.K. Das
So, after putting all the information, the correlation give a constant
value due to constant uid properties. Miglani et al. [110] conducted
experiment on inconel foil during pool boiling of refrigerant R-134a at
high pressures with heat ux varied from 2.5 to 12 kW/m2 . Based on
their experimental results, they proposed a dimensionless correlation
of bubble generation frequency in terms of modied boiling number
using regression analysis.
Hamzekhani et al. [111] developed a correlation of bubble depar-
ture frequency by taking own pool boiling experimental data of water
and NaCl solution at 1 atmospheric pressure under low to moderate
heat ux. They carried out a dimensionless analysis using
Buckinghams theorem on the inuence parameters such as bubble
departure diameter, bubble departure frequency, surface tension,
acceleration due to gravity, vapor-liquid density dierence and heat
ux and obtained three -terms. They obtained three -terms which
are present in Eq. 126 with the tting parameters such as constant
0.015 and exponent 0.44 and 0.88 respectively. They found that the
predicted bubble departure frequency lie an average error of 10% and
4% for NaCl solutions and water.
All the correlations for bubble departure frequency which are listed
in Table 4 are combined with bubble departure diameter except the
correlation of Kumada and Sakashita [108], Sakashita and Ono [109],
Fig. 16. Experimental verses predicted fDd0.5 (a) Hamzekhani et al. [111] data (b) Fazel and Mahboobpour [18] data (c) McHale and Garimella [69] data.
486
Renewable and Sustainable Energy Reviews 78 (2017) 466494
Miglani et al. [110] and Hamzekhani et al. [111]. From the experi-
mental data, the increase trend of both the bubble departure frequency
and diameter are found with respect to heat ux. But from the
comparison point of view, the bubble departure frequency cannot be
predicted directly from the available correlations except the correla-
tions [108111] because almost all the correlations are given along
with bubble departure diameter. Though the bubble departure fre-
quency is predicted from the existing correlations by dividing the
respective magnitude of bubble departure diameter but unable to
capture the experimental trend as the magnitude of departure
frequency decreases with increase in bubble departure diameter.
Therefore, for comparing the order of experimental data, all the
correlations are categorized into total ve cases such as frequency-
diameter product ( f Dd), frequency-square root diameter
product( f D0.5
d ), square root frequency-diameter product ( f
0.5D )
d , fre-
quency-square diameter product ( f D2d) and frequency (f).
A comparison of correlations ( f Dd ) proposed by Jakob and Linke
[98], Jakob and Fritz [99], Jakob [100], Peebles and Garber [101],
Zuber [103], Malenkov [105] and Stephan [107] to predict order of the
experimental data of water, pure monoethyleneglycol and FC-72 given
by Hamzekhani et al. [111], Fazel and Mahboobpour [18] and McHale
and Garimella [69] are carried out and shown in Fig. 15(a), (b) and (c)
R.L. Mohanty, M.K. Das Renewable and Sustainable Energy Reviews 78 (2017) 466494

respectively. From all the graph, it is found that only the correlation of
Jakob and Linke [98] able to predict the experimental trend but the
accuracy does not lie within an error of 30%. The order of experi-
mental data predicted by Malenkov [105] decreases as the product of
departure frequency and departure diameter is inversely proportional
to the heat ux. The correlation of Jakob and Fritz [99], Jakob [100],
Peebles and Garber [101], Zuber [103] and Stephan [107] forecast
constant magnitude of f Dd due to the involvement of constant uid
properties.
Similarly, the correlation given by Cole [cf.111], McFadden and
Grassman [102] and Ivey(a) [104] are tested to predict the above three
set of experimental data in the form of f D0.5
d and shown in Fig. 16(a),
(b) and (c) respectively. It is found that all the experimental data
predicted through the above four correlations remain constant and also
unable to capture the experimental trend due the absent of thermal
potential.
A comparison of correlations proposed by Mikic and Rohsenow
[89] and Hatton and Hall [47] to predict the order of experimental data
of Hamzekhani et al. [111], Fazel and Mahboobpour [18] and McHale
and Garimella [69] in the form of f 0.5 Dd and fDd2 are carried out and
shown in Fig. 17(a) and (b) respectively. From the Fig. 17(a), it is
observed that the correlation of Mikic and Rohsenow [89] over predicts
the experimental data McHale and Garimella [48] while under predicts
the data of Hamzekhani et al. [78] and Fazel and Mahboobpour [18] as
the error band set to 30%. It is also seen that the order of magnitude
of experimental data is comparable with large deviation from the set
error band. Similarly, from the Fig. 17(b), it is concluded that the
correlation proposed by Hatton and Hall [47] predicts the constant
magnitude of all set of experimental data because of constant uid
properties i.e. thermal diusivity. So, this correlation is unable to
forecast the trend seen in experiment.
Again the correlation proposed by Hamzekhani et al. [111] is used
to predict experimental data of bubble departure frequency for water,
monoethyleneglycol and FC-77 given by Hamzekhani et al. [111], Fazel
and Mahboobpour [18] and McHale and Garimella [69] as presented in
Fig. 18(a), (b) and (c) respectively. From Fig. 18(a), it is clear that the
correlation of Hamzekhani et al. [111] able to predict their own
experimental data within an error of 10%. However, this correlation
is signicantly fails to predict the experimental data of Fazel and
Mahboobpour [18] and McHale and Garimella [69] within an error of
30%. Both the experimental data [18,69] are over predicted by this
correlation.
From the above point of discussion, it is found that in all the
existing correlations, the bubble departure frequency is correlated with Fig. 17. (a) Experimental verses predicted f 0.5 Dd (b) Experimental verses predicted
the bubble departure diameter in various forms. So, the prediction of f D2d .
bubble departure frequency is extremely dicult than the bubble
departure diameter and vice-versa. The bubble departure frequency 4. Factors aecting the bubble dynamics parameters
is strongly inuenced by various regimes and also a property depen-
dent parameter. A conclusion is drawn from the above study that the The eect of various factors such as thermo-physical properties,
knowledge of bubble departure diameter, bubble waiting period, heat ux, wall superheat and liquid sub-cooling, contact angle, gravity
bubble growth period, surface tension, heat ux and thermo-physical level, cavity spacing, pressure and surface roughness on bubble
properties uids are necessary for the determination of bubble dynamics parameter with a view to predict the boiling heat transfer
departure frequency. Although few correlations available which can coecient or boiling heat ux are studied in this section. In order to
be able to compare the order magnitude of the product of departure develop a better correlation for boiling heat transfer coecient the
frequency and diameter based on experimental data but not the actual eect of these factors are considered. Therefore, this section present a
magnitude as well as the trends bubble departure frequency found in short discussion over various factors which aects the bubble dynamics
experiments. Though the correlation of Hamzekhani et al. [111] able to parameters in a concise manner.
predict their own experimental data with reasonable accuracy but fails
to predict the experimental data of other investigators with that
accuracy. Therefore, a better correlation of bubble departure frequency 4.1. Eect of thermo-physical properties
may be developed by considering the parameters surface roughness,
cavity size, system pressure and dierent solid-liquid interaction From the previous section, it is found that, almost all the correla-
parameters along with thermal potential, bubble departure diameter tions of bubble dynamics parameters specially bubble departure
and thermo-physical properties of uid. It may be suggested that the diameter, bubble growth period, bubble waiting period and bubble
non-dimensionless analysis of bubble departure frequency may lead to departure frequency are inuenced by thermo-physical properties of
better correlation with less deviation. uid. Gong et al. [9] experimentally informed that lager departure

487
R.L. Mohanty, M.K. Das
Fig. 18. Experimental verses predicted bubble departure frequency (a) Hamzekhani et al. [111] data (b) Fazel and Mahboobpour [18] data (c) McHale and Garimella [69] data.
diameter and smaller departure frequency obtained during pool boiling
of pure ethane than pure isobutene. However, with increase in
concentration of binary mixture, the bubble diameter and frequency
are varied signicantly. Fazel and Shafaee [62] experimentally ob-
served that the bubble departure diameter becomes smaller, medium
and larger for KNO3 solution, Na2SO4 solution and NaCl solution
respectively. This may be due to the change in thermo-physical
properties of uids. The bubble departure diameter found to bigger,
medium and smaller in water, R134a and R-123 because of change in
surface tension studied by Lamas et al [65]. Hamzekhani et al. [21], it
is reported that smaller bubble departure diameter is obtained in
ethanol than water due to low interfacial tension. In another work of
Hamzekhani et al. [111], it is reported that the bubble departure
diameter increases and the bubble departure frequency decreases with
increase in mass fraction of NaCl solution. The minimum bubble
departure size and maximum departure frequency found in water than
NaCl solution. Due to high interfacial tension in NaCl solution, the
bubble departure size becomes larger. Bovard et al. [68] also observed
488
Renewable and Sustainable Energy Reviews 78 (2017) 466494
similar result for water and ethanol. However, they found that the
departure diameter of methanol and acetone lies in between water and
ethanol. This departure size is found due to change in thermo physical
properties of uid which inuences the growth process of the vapor
bubble.
4.2. Eect of heat ux
Generally the bubble departure diameter, bubble departure fre-
quency and active nucleation site densities increases with increase in
heat ux. The departure diameter increases with increase in heat ux
investigated by various researchers for dierent uids
[9,18,21,22,62,68,112]. Gong et al. [9] also found that the departure
diameter and frequency of binary mixture also follow the increase trend
with respect to heat ux while the departure frequency rst increases
and then decreases for mixtures at ethane concentration of 0.7 and 0.9.
The similar increased trend is also found in case departure frequency
and active nucleation site density [18,112]. McHale and Garimella [69]
R.L. Mohanty, M.K. Das Renewable and Sustainable Energy Reviews 78 (2017) 466494

found that with increase in heat ux, the terminal velocity is increased 4.4. Eect of contact angle
due to the increased buoyancy from larger bubble volumes and also
increases departure frequency. The active nucleation site density Contact angle is the angle formed by a liquid at the three phase
increases with increase in heat ux [31,77]. Gong and Cheng [113] boundary where liquid, gas and solid intersect which gives the
found that the bubble waiting period decreases with increase in heat quantitative measure of wetting of a solid by liquid. Depending upon
ux. With increase in heat ux, the bubble departure diameter the movement of three phase boundary, the contact angle is divided
decreases and the bubble departure frequency increases [19,114]. into two types such as static contact angle and dynamic contact angle.
Chien et al. [115] found that with increase in heat ux, the bubble When the three phase is moving, the dynamic contact angle can be
growth period and bubble departure diameter decreases while the measured and is referred as advancing and receding contact angles
bubble frequency and nucleation site density increases. Pascual et al. (maximum and minimum value of static contact angle). Abarajith and
[116] reported that with increasing heat ux, the average bubble Dhir [126] investigated the eects of contact angle on the growth and
departure frequency increases and then decreases. Gao et al. [117] departure of a single bubble on a horizontal heated surface during pool
found that with increase in heat ux, the bubble growth period boiling of water and PF5060 under normal gravity. The departure
becomes shorter. Siedel et al. [118] observed that at low heat uxes, diameter becomes larger for water with contact angle of 35 while
the bubble departure diameters are similar for both smooth and rough smaller for PF5060 with contact angle of 10. Mukherjee and Kandlikar
surface. However, at high uxes, the active site densities and ebullition [127] used the numerical model of Mukherjee and Dhir [128] to
frequencies are similar for both the surfaces while the departure simulate the single vapor bubble growth on a heated wall with dynamic
diameter becomes larger for smooth surface. Hetsroni et al. [119] contact angle during nucleate pool boiling. When the dynamic contact
conducted experiment on at plate during pool boiling of water and angle is taken as a function of contact line velocity, the bubble
non-ionic 600 ppm Alkyl (8 to 16) surfactant with constant heat ux of departure time longer in case of constant advancing contact angle of
10 kW/m2 and 50 kW/m2 at atmospheric pressure. They observed that 61 with receding contact angle of 48 while shorter for static contact
the shape, life time and volume of bubble growth is equivalent to that angle of54. The eect of increase in advancing and decrease in
water at heat ux of 10 kW/m2. But at heat ux of 50 kW/m2, the receding contact angle is to increase in bubble departure diameter
bubble departure diameter of water is increased while decreased trend with increase in bubble departure period. The bubble initially grows
of bubble departure diameter is found in surfactant with shorter life with a spherical shape but gradually turns into a hemispherical shape
time. due to eect of high advancing contact angle. Fig. 20(a) and (b) shows
the bubble growth rate corresponding to advancing contact angle 90
and 54 and receding contact angle 54 and 20. Dhir at al. [121]
4.3. Eect of wall superheat and liquid sub-cooling numerically observed the departure diameter decreases linearly as the
contact angle is decreased which is consistent with the correlation of
McHale and Garimella [69] experimentally observed the eect of
increase in wall superheat is to increase bubble departure diameter,
bubble departure frequency and nucleation site density during pool
boiling of peruorinated hydrocarbon (FC-77) on smooth and rough
surfaces having roughness values (R a)0.03 and 5.89 m . Gong and Cheng
[113] also reported the similar results for bubble departure diameter
while the bubble growth period decreases with increase in wall super
heat. Zuber [120] suggested that departure diameter increases with
increase in wall superheat as the superheated thermal boundary layer
becomes thicker with increasing superheat levels. The departure
diameter is independent or only weakly depended on the wall super-
heat [49,55]. Numerical simulation results of Dhir et al. [121] showed
good agreement with experimental data of Qiu et al. [122] with respect
to time dependent, growth rate and bubble departure diameter. The
eect of increase in wall superheat is to increase the growth rate and
bubble departure diameter but decrease the growth period. The
departure diameter increase due to increased liquid inertia with
increased growth rate. The eect of increase in sub-cooling is to
increase the growth period and decrease the bubble departure dia-
meter. Fig. 19(a) and (b) shows the eect of wall superheat and liquid
sub-cooling on a growing bubble for saturated water at 1 atm pressure
and T = 8K .
Siedel et al. [118] found that with increasing the wall superheat, the
bubble growth time is signicantly decreases and the bubble frequency
increases whereas the departure volume become independent of wall
superheat. They concluded that the bubble growth rates follow a non-
dimensional law as V+ = t+0.6 for t+ > 0.2 and V+ = 2 t+ for t+ < 0.2 ,
where V+ = V/Vd and t+ = t/t g .
Hutter et al. [123] found that the departure diameter and bubble
waiting period increases with increase in wall superheat while the
bubble departure frequency becomes constant for high wall superheats.
Zhang et al. [124] observed that the bubble departure diameter
increases with increase in wall superheat at both isolated and coalesced
bubble regimes. The eect of increasing sub-cooling is to decrease the
departure (coalesced) bubble size and departure frequency during pool Fig. 19. Comparison of numerical simulations with experimental data: (a) Eect of wall
boiling is investigated by Coulibaly et al. [125]. superheat and (b) Eect of liquid sub-cooling [121].

489
R.L. Mohanty, M.K. Das Renewable and Sustainable Energy Reviews 78 (2017) 466494

[129] numerically observed the eect of gravity on the bubble

departure diameter and release frequency using the lattice Boltzmann
approach. They found that the bubble departure diameter and bubble
release frequency are proportional to g1/2 and g3/4 are shown in
Fig. 21(a) and (b) respectively. Gong and Cheng [113] also observed
similar results for bubble departure and bubble departure frequency.
The bubble nucleation and lift-o becomes faster in normal gravity
whereas in micro-gravity the bigger size bubbles are formed on the
heated surface even at low heat ux reported by Colin et al. [132].

4.6. Eect of cavity spacing

Zhang and Shoji [133] conducted pool boiling experiments in

distilled water with creating two cylindrical cavities on the heated
surface. They found that bubble departure diameter and bubble
departure frequency increased with increase in heat ux for S/Dd ratio
of 1.5. The bubble departure frequency increased when S/Dd ratio is
varied between 1.5 to 2 and dropped to a S/Dd ratio of 3, where S/Dd is
the ratio of the inter-cavity spacing and the bubble departure diameter.
Nimkar et al. [134] prepared three surfaces with cavity spacing
0.5 mm, 0.75 mm and 1 mm for creating 225 cavities, 100 cavities
and 64 cavities respectively. They found that active nucleation site
density increased with 0.5 mm spacing while the bubble departure
diameter and bubble departure frequency are not inuenced with
cavity spacing. Dhir et al. [121] numerically observed the variation
non-dimensional bubble departure diameter and departure time with
cavity spacing when the bubbles are placed on the of an equilateral

Fig. 20. Bubble growth (a) Advancing contact angle 90, receding contact angle 54 and
(b) Advancing contact angle 54 , receding contact angle 20 [127].

Fritz [14] while the growth period varies nonlinearly with contact
angle. The eect of increase in contact angle is to increase in microlayer
evaporation rate due to increase in bubble base area and interfacial
area. With increasing the contact angle, the bubble growth rate
increases and the bubble release frequency increases exponentially
with the contact angle is studied by Hazi and Markus [129]. Lamas
et al. [65] numerically found that the bubble departure diameter goes
on increasing with increase in contact angle. The active nucleation site
density increases with increase in contact angle Wang and Dhir [31].
Gong and Cheng [113] numerically observed that the bubble growth
period increases and bubble waiting period decreases with increase in
contact angle while little eect on the bubble departure diameter. Gong
et al. [130] numerically observed that the bubble departure diameter
and bubble departure frequency increases with increase in static
contact angle.

4.5. Eect of gravity level

Abarajith et al. [131] investigated the eects of gravity on bubble

departure diameter and departure time during pool boiling of water
and PF5060 under the reduced gravity conditions. Dhir et al. [121]
numerically studied for water with a contact angle of 54, the bubble
departure diameter is Dd ~ g0.5 and the growth period of the bubble
becomes t g ~ g0.93 where as for PF5060 with a contact angle of 10, the
bubble departure diameter is Dd ~ g0.42 and the growth period of the
bubble becomes t g ~ g0.82 . Finally, they concluded that there is a very
weak nonlinear eect of contact angle on the eect of gravity on both Fig. 21. (a) Bubble departure diameter as a function of gravity and (b) Bubble departure
the bubble departure diameter and growth period. Hazi and Markus frequency as a function of gravity [192].

490
R.L. Mohanty, M.K. Das Renewable and Sustainable Energy Reviews 78 (2017) 466494

triangle at a gravity of 0.01 ge , where ge is the earth normal gravity. The

eect of increase in cavity spacing is to decrease in departure bubble
diameter and increase in departure time. Hutter et al. [123] observed
that the cavity depth has no eect on bubble departure diameter and
bubble growth period. Gong et al. [130] numerically found that the
bubble departure diameter increases with increase in cavity depth up to
a certain value beyond that there is a little eect on departure diameter
whereas the bubble departure frequency is suddenly jump when the
depth of a hydrophilic cavity reaches a critical value. They also
observed that the bubble departure diameter and bubble departure
frequency increases with cavity width after it reaches the critical value.
Bon et al. [135] experimentally found that the boiling heat transfer
decreases with increasing in cavity spacing while increase in cavity
depth enhances active nucleation site densities because of improved
vapor retention.

4.7. Eect of pressure

Labuntsov et al. [87] and Akiyama et al. [136] experimentally

observed the bubble growth rate on at and cylindrical surface during
pool boiling of water at high pressure. They found that bubble growth
rate decreases with increase in pressure. The bubble growth at high
pressure is due to the conduction near the bubble base as assumed by
Labuntsov et al. [87] or the evaporation at triple contact line formed at
the bubble base. Sakashita [82] investigated the bubble growth rate for
horizontal and vertical surface during the pool boiling of water at high
pressure. With increasing in pressure, the bubble growth rate becomes
small and the detachment radius decreases for horizontal heating
surface is shown in Fig. 22(a). For vertical heating surface, the bubble
grows in a certain size, then departs and slides. The eect of increasing
in pressure is to increase in nucleation site density at a relatively low
heat ux (q = 5 10 4W/m2 ) for the vertical heating surface as shown in
Fig. 22(b). The nucleation site density is increased in proportional to
Fig. 22. (a) Primary bubble radius at dierent pressures and (b) Boiling behaviors at
the 1.5th power of pressure. Gong et al. [9] experimentally found that
dierent pressures [82].
with increase in pressures, the departure diameter and departure
frequency decreases during pool boiling of pure ethane and isobutane.
study is required for complete understanding of the above factors on
Lamas et al [65] also observed that the departure diameters of water,
bubble dynamics parameters which may be helpful for the prediction of
R134a and R-123 are decreased with increased in pressure. Hutter
boiling heat transfer. The variety of solid-liquid combinations during
et al. [123] reported that the bubble departure frequency slightly
boiling process need to be studied for complete understanding of
decreases with increase in pressure whereas there is no inuence of
boiling phenomena.
pressure on departure diameter and bubble waiting period. The
average bubble departure frequency increases and the departure
diameter deceases with increase in pressure observed by Miglani 5. Conclusion
et al. [110].
This review paper is primed with an aim to help investigators to
4.8. Eect of surface roughness develop generalized boiling heat transfer correlations based on various
bubble dynamics parameters for prediction of boiling heat transfer
The nucleation site density declines as the surface roughness coecient or boiling heat ux during nucleate boiling of liquids over
increases and then increases with surface roughness at a given heat heated surfaces. In order to identify the bubble dynamics parameters,
ux [77,137]. The bubble departure diameter becomes larger in various correlations proposed by dierent investigators to predict
polished surface than rough surface at a given heat ux whereas the boiling heat transfer and boiling heat ux based on bubble dynamics
departure frequency and active nucleation sited density are higher in parameters are examined. The assessment of these correlations for
rough surface than polished surface as reported by McHale and prediction of boiling heat transfer coecient and boiling heat ux
Garimella [69]. The similar results of bubble departure diameter for reveals that bubble departure diameter, active nucleation site density
smooth and rough surface is also observed by Siedel et al. [118]. They and departure frequency are major bubble dynamics parameters
also reported that with increase in surface roughness, the nucleation aecting nucleate boiling heat transfer process. Besides, other para-
site densities on the boiling surfaces increases. Suszko and El-Genk meters such as bubble growth period, waiting period and growth rate
[67] experimentally observed that the active nucleation site density and are also pointed with a view to help in the development of correlations
departure frequency becomes higher while departure diameter become for bubble departure diameter and bubble departure frequency. The
smaller for rough surface than smooth surface. Bovard et al. [68] found correlations proposed by respective investigators found to predict their
that the bubble departure diameter decreases with increase in surface own experimental data with reasonable accuracy. However, when these
roughness for dierent heated surface. correlations are tested against experimental data of other investigators
The overviews of the above study is that the bubble dynamics with same or dierent surface uid combination, their prediction
parameters such as bubble departure diameter, active nucleation site accuracy decreases signicantly. This may be due to the fact that
density and bubble departure frequency are strongly inuenced by the expressions for bubble dynamics parameters used in these correlations
various discussed experimental and numerical parameters. Still more are not accurately determined. Because, the correlations proposed for

491
R.L. Mohanty, M.K. Das
determination of bubble dynamics parameters are either based on their
own experimental data or a correlation proposed by other investigator.
Therefore, these correlations fails to make themselves a generalized
correlation. Based on these facts, an inclusive review is carried out on
existing correlations for prediction of bubble dynamics parameters.
The conclusions made from this review is reported below:
The dierent form of correlations available for bubble departure
diameter, active nucleation site density and bubble departure fre-
quency in the respective sections are compared with various existing
experimental data available in the literature. The results conclude that
either two or three correlations of bubble departure diameter are able
to predict one of the experimental data within an average error of
25% but fails to predict the experimental data of other investigators.
From the comparison of correlations for active nucleation site density,
it can be seen that the correlations are unable to forecast the data of
experiments within a minimum error of 60% except correlations
proposed by Sakashita-Kumada [78] and Hibiki and Ishii [80]. Almost
all the correlations proposed for prediction of bubble departure
frequency are in conjunction with the departure diameter. Therefore,
product of departure frequency and departure diameter in dierent
form is compared with experimental observations. From the compar-
ison results, it is clear that the correlations for bubble departure
frequency found to forecast the order of magnitude of experimental
data within an maximum error of 30%.
The above summary provides a clear vision that the correlations for
bubble dynamics parameters available in the literature are highly
specic to their own experimental data and therefore, when used to
predict experimental data of other investigators, it fails to do so. This
may be the due to the following facts:
The bubble departure diameter and departure frequency depends
upon various inuencing parameters such as thermo-physical proper-
ties of uid, contact angle, acceleration due to gravity, vapor bubble
velocity, bubble growth rate, bubble growth period, bubble waiting
period, cavity size, surface roughness, heat ux or wall superheat, and
system pressure. Since, the correlations of departure diameter and
frequency are developed by taking only few of the parameters men-
tioned above, their applicability as a generalized correlation does not
work.
Similarly, the active nucleation site densities are highly prejudiced
with the inuencing parameters such as cavity size (maximum or
minimum), cavity angle, surface roughness, contact angle, heat ux,
wall superheat and system pressure. Therefore, the generalization of
active nucleation site density correlations become dicult.
Therefore, it is extremely important that correlations for bubble
departure diameter should be developed taking maximum number of
inuencing parameters. Such correlations may able to predict experi-
mental data of various investigators with minimum error. Once the
correlations for bubble dynamics parameters are developed with
minimum error, the same parameters can be properly correlated
among themselves to develop a generalized correlation for the predic-
tion of boiling heat transfer coecient or boiling heat ux.
References
[1] Pioro IL, Rohsenow W, Doerer SS. Nucleate pool boiling heat transfer. II:
assessment of prediction methods. Int J Heat Mass Transf 2004;47:504557.
[2] Rao GV, Balakrishnan AR. Heat transfer in nucleate pool boiling of multi-
component mixtures. Exp Therm Fluid Sci 2004;29:87103.
[3] Taboas F, Valles M, Bourouis M, Coronas A. Pool boiling of ammonia/water and
its pure components: comparision of experimental data in the literature with the
predictions of standard correlations. Int J Refrig 2007;30:77888.
[4] Spindler K. Overview and discussion on pool boiling heat transfer data and
correlations of ammonia. Int J Refrig 2010;33:1292306.
[5] Sarma PK, Srinivas V, Sharma KV, Subrahmanyam T, Kakac S. A correlation to
predict heat transfer coecient in nucleate boiling on cylindrical heating elements.
Int J Therm Sci 2008;47:34754.
[6] Fazel A, Ali S. Experimental investigation in pool boiling heat transfer of pure/
binary mixtures and heat transfer correlations. Iran J Chem Chem Eng
2008;27(3):13550.
[7] Das MK, Gupta SC, Agarwal VK. Dimensionless local and average boiling heat
492
Renewable and Sustainable Energy Reviews 78 (2017) 466494
transfer correlation for saturated liquids. Heat Transf Res 2010;41(5):53158.
[8] Sathyabhama A, Ashok Babu TP. Experimental investigation in pool boiling heat
transfer of ammonia/water mixture and heat transfer correlations. Int J Heat
Fluid Flow 2011;32:71929.
[9] Gong M, Wu Y, Ding L, Cheng K, Wu J. Visualization study on nucleate pool
boiling of ethane, isobutene and their binary mixtures. Exp Therm Fluid Sci
2013;51:16473.
[10] Zajaczkowski B, Halon T, Krolicki Z. Experimental verication of heat transfer
coecient for nucleate boiling at sub-atmospheric pressure and small heat uxes.
Heat Mass Transf 2016;52:20515.
[11] Kim J. Review of nucleate pool boiling bubble heat transfer mechanisms. Int J
Multiph Flow 2009;35:106776.
[12] Stephan K, Preusser P. Wrmebergang und maximale wrmestromdichte beim
behiterienden binrer und ternrer ussigkeitsgemische. Chem Ing Tech 1979,
[51:37 (Synopse MS 649/79)].
[13] Stephan K, Abdelsalam M. Heat transfer correlations for natural convection
boiling. Int J Heat Mass Transf 1980;23:7387.
[14] Fritz W. Maximum volume of vapor bubbles. Phys Z. 1935;36:37984.
[15] Jung D, Kim Y, Ko Y, Song K. Nucleate boiling heat transfer coecients of pure
halogenated refrigerants. Int J Refrig 2003;26:2408.
[16] Jung D, Lee H, Bae D, Oho S. Nucleate boiling heat transfer coecients of
ammable refrigerants. Int J Refrig 2004;27:40914.
[17] Rao GV, Balakrishnan AR. Heat transfer in nucleate pool boiling of multi-
component mixtures. Exp Therm Fluid Sci 2004;29:87103.
[18] Fazel SAA, Mahboobpour M. Pool boiling heat transfer in monoethyleneglycol
aqueous solutions. Exp Therm Fluid Sci 2013;48:17783.
[19] Judd RL, Hwang KS. A comprehensive model for nucleate pool boiling heat
transfer including microlayer evaporation. J Heat Transf 1976;98(4):6239.
[20] Mikic BB, Rohsenow WM. A new correlation of pool boiling data including the
eect of heat surface characteristics. ASME J Heat Transf 1969;91:24550.
[21] Hamzekhani S, Falahieh MM, Akbari A. Bubble departure diameter in nucleate
pool boiling at saturation: pure liquids and binary mixtures. Int J Refrig
2014;46:508.
[22] Jamialahmadi M, Helalizadeh A, Muller-Steinhagen H. Pool boiling heat transfer
to electrolyte solutions. Int J Heat Mass Transf 2004;47:72942.
[23] Lakhera VJ, Gupta A, Kumar R. Enhanced boiling outside 83 plain and coated
tube bundles. Heat Transf Eng 2012;33(9):82834.
[24] Bhaumik S, Agarwal VK, Gupta SC. A generalized correlation of nucleate pool
boiling of liquids. Ind J Chem Technol 2004;11:71925.
[25] Han CY, Grith P. The mechanism of heat transfer in nucleate pool boiling parts
I and II. Int J Heat Mass Transf 1965;8:887913.
[26] Paul DD, Abdel-Khalik SI. A statistical analysis of saturated nucleate boiling along
a heat wire. Int J Heat Mass Transf 1983;26:50919.
[27] Barthu G. Active nucleation site density and pool boiling heat transfer-an
experimental study. Int J Heat Mass Transf 1992;35:2718.
[28] Benjamin RJ, Balakrishnan AR. Nucleate pool boiling heat transfer of pure liquids
at low to moderate heat uxes. Int J Heat Mass Transf 1996;39:2495504.
[29] Yu BM, Cheng P. A fractal model for nucleate pool boiling heat transfer. ASME J
Heat Transf 2002;124:111724.
[30] Dhir VK. Review, nucleate and transition boiling heat transfer under pool and
external ow conditions. Int J Heat Fluid Flow 1991;12(4):290313.
[31] Wang CH, Dhir VK. Eect of surface wettability on active nucleation site density
during pool boiling of water on a vertical surface. ASME J Heat Transf
1993;115:65969.
[32] Sateesh G, Das SK, Balakrishnan AR. Analysis of pool boiling heat transfer: eect
of bubbles sliding on the heating surface. Int J Heat Mass Transf
2005;48:154353.
[33] Chu H, Yu B. A new comprehensive model for nucleate pool boiling heat transfer
of pure liquid at low to high heat uxes including CHF. Int J Heat Mass Transf
2009;52:420310.
[34] Moore FD, Mesler RB. The measurement of rapid surface temperature uctuations
during nucleate boiling of water. AIChE J 1961;7:6204.
[35] Hsu YY. On the size range of active nucleation cavities on a heating surface. ASME
J Heat Transf 1962;84:20715.
[36] Van Stralen SJD, Sohal MS, Cole R, Sluyter WM. Bubble growth rates in pure and
binary systems: combined eect of relaxation and evaporation microlayers. Int J
Heat Mass Transf 1975;18:45367.
[37] Gerardi C, Buongiorno J, Hu L, Mckrell T. Study of bubble growth in water pool
boiling through synchronized, infrared thermometry and high-speed video. Int J
Heat Mass Transf 2010;53:418592.
[38] Kurul N, Podowski MZ. Multidimensional eects in forced convection subcooled
boiling. In: Proceedings of the 9th international heat transfer conference at
Jerusalem. Israel; 1990. p. 2125.
[39] Bowring, RW. Physical model based on bubble detachment and calculation of
steam voidage in the subcooled region of a heated channel. OECD Halden Reactor
Project Report HPR-10; 1962.
[40] McAdams WH. Heat transmission, third ed. New York: McGraw-Hill; 1945. p.
180.
[41] Barthau G, Hahne E. Nucleation site density and heat transfer in nucleate pool
boiling of refrigerant R134a in a wide pressure range. In: Proceedings of the 3rd
European thermal sciences conference. 2000;2. p. 731736.
[42] Luke A, Goreno D. Heat transfer and size distribution of active nucleation sites in
boiling propane outside a tube. Int J Therm Sci 2000;39:91930.
[43] Staniszewski, BE. Nucleate boiling bubble growth and departure. Technical report
no. 16, Division of sponsored research, Massachusetts Institute of Technology:
Cambridge, MA; 1959.
[44] Borinshansky VM, Fokin FS. Heat transfer and hydrodynamics in steam gen-
erators. Tr TsKTI 1963;62.
[45] Ruckenstein R. Recent trends in boiling heat and mass transfer. Appl Mech Rev
1964;17:66372.
[46] Cole R, Shulman HL. Bubble departure diameters at subatmospheric pressures.
R.L. Mohanty, M.K. Das
AIChE Chem Symp Ser 1966;62(64):616.
[47] Hatton P, Hall IS, Photographic study of boiling on prepared surfaces. In:
Proceedings of the 3rd international heat transfer conference, AIChE. 1966;4(2).
p. 2437.
[48] Cole R. Bubble frequency and departure volumes at subatmospheric pressures.
AIChE J 1967;13:77983.
[49] Cole R, Rohsenow WM. Correlation of bubble departure diameters for boiling of
saturated liquids. Chem Eng Prog Symp Ser 1969;65(92):2113.
[50] Van Stralen SJD, Zijl W. Fundamental developments in bubble dynamics. In:
Proceedings of the 6th international heat transfer conference. 1978;(6). p. 429
450.
[51] Golorin VS, Kolchugin BA, Zakharova EA. Investigation of the mechanism of
nucleate boiling of ethyl alcohol and benzene by means of high speed motion-
picture photography. Heat Transf Sov Res 1978;10:7998.
[52] Kutateladze SS, Gogonin II. Growth rate and detachment diameter of a vapor
bubble in free convection boiling of a saturated liquids. High Temp
1979;17:66771.
[53] Jensen MK, Memmel GJ. Evaluation of bubble departure diameter correlations.
Prog Eighth Int Heat Transf Conf 1986;4:190712.
[54] Kocamustafaogullari G. Pressure dependence of bubble departure diameter for
water. Int Commun Heat Mass Transf 1983;10:5019.
[55] Goreno D, Knabe V, Beiling V. Bubble density on surfaces with nucleate boiling-
Its inuences on heat transfer. In: Proceedings of the 8th international heat
transfer conference. 1986; (4). p. 19952000.
[56] Stephan K. Saturated pool boiling and sub-cooled ow boiling of mixtures (Ph.D.
thesis). New Zealand: University of Auckland; 1992.
[57] Zeng LZ, Klausner JF, Mei R. A unied model for the prediction of bubble
detachment diameters in boiling systems- I. Pool boiling. Int J Heat Mass Transf
1993;36(9):226170.
[58] Klausner JF, Mei R, Bernhard DM, Zeng LZ. Vapor bubble departure in forced
convection boiling. Int J Heat Mass Transf 1993;36:65162.
[59] Yang C, Wu Y, Yuan X, Ma C. Study on bubble dynamics for pool nucleate boiling.
Int J Heat Mass Transf 2000;43:2038.
[60] Lee HC, Oh BD, Bae SW, Kim HM. Single bubble growth in saturated pool boiling
on a constant wall temperature surface. Int J Multiph Flow 2003;29:185774.
[61] Kim J, Kim MH. On the departure behaviours of bubble at nucleate pool boiling. J
Multiph Flow 2006;32:126986.
[62] Fazel SAA, Shafaee SB. Bubble dynamics for nucleate poll boiling of electrolyte
solutions. ASME J Heat Transf 2010;132(2):8150217.
[63] Phan HT, Caney N, Marty P, Colasson S, Gavillet J. A model to predict the eect of
contact angle on the bubble departure diameter during heterogeneous boiling. Int
Commun Heat Mass Transf 2010;37:9649.
[64] Nam Y, Aktinol E, Dhir VK, Ju YS. Single bubble dynamics on a superhydrophilic
surface with articial nucleation sites. Int J Heat Mass Transf 2011;54:15727.
[65] Lamas MI, Jabardo JMS, Arce A, Farinas P. Numerical analysis of the bubble
detachment diameter in nucleate boiling. In: Proceedings of the 6th European
Therm Sci Conf, J Phy: Conf Series 395, 2012.
[66] Keshock EG, Siegel R. Forces acting on bubbles in nucleate boiling under normal
and reduced gravity conditions. NASA TN D-2299; 1964.
[67] Suszko A, El-Genk MS. Saturation boiling of PF-5060 on rough Cu surfaces:
bubbles transient growth, departure diameter and departure frequency. Int J Heat
Mass Transf 2015;91:36373.
[68] Bovard S, Asadinia H, Hosseini G, Fazel SAA. Investigation and experimental
analysis of the bubble departure diameter in pure liquids on horizontal cylindrical
heater. Heat Mass Transf 2017;53:1199210.
[69] McHale JP, Garimell SV. Bubble nucleation characteristics in pool boiling of a
wetting liquid on smooth and rough surfaces. Int J Multiph Flow 2010;36:24960.
[70] Gaertner RF, Westwater JW. Population of active sites in nucleate boiling heat
transfer. Chem Eng Prog Symp Ser 1960;56:3948.
[71] Hsu YY, Graham RW. An analytical and experimental study of the thermal
boundary layer & ebullition cycle in nucleate boiling. NASA TND 1961;594.
[72] Bier K, Goreno D, Salem M, Tanes Y. Pool boiling heat transfer and size of active
nucleation centers for horizontal plates with dierent surface roughness. In:
Proceedings of the 6th heat transfer conference. Toronto; 1978(1). p.151156.
[73] Cornwell K, Brown RD. Boiling surface topography. In: Proceedings of the 6th
international heat transfer conference. Toronto; 1978(1). p. 157161.
[74] Kocamustafaogullari G, Ishii M. Interfacial area and nucleation site density in
boiling systems. Int J Heat Mass Transf 1983;26(9):137787.
[75] Yang SR, Kim RH. A mathematical model of pool boiling nucleation site density in
terms of surface characteristics. Int J Heat Mass Transf 1988;31:112735.
[76] Banko SG. Entrapment of gas in spreading of liquid over a rough surface. AIChE
J 1958;4:246.
[77] Benjamin R, Balakrishnan A. Nucleate site density in pool boiling of saturated
pure liquids: eect of surface microroughness and surface and liquid physical
properties. Exp Therm Fluid Sci 1997;15:3242.
[78] Sakashita H, Kumada T. Method for predicting boiling curves of saturated pool
boiling. Int J Heat Mass Transf 2001;44:67382.
[79] Basu N, Warrier GR, Dhir VK. Onset of nucleate boiling and active nucleation site
density during ow boiling. ASME J Heat Transf 2002;124:71727.
[80] Hibiki T, Ishii M. Active nucleation site density in boiling systems. Int J Heat Mass
Transf 2003;46:2587601.
[81] Xiao B, Jiang G, Zheng D, Chen L, Liu B. Calculation of active nucleation site
density in boiling systems. Res J Appl Sci Eng Technol 2013;6(4):58792.
[82] Sakashita H. Bubble growth rates and nucleation site densities in saturated pool
boiling of water at high pressures. J Nucl Sci Technol 2011;48(7):73443.
[83] Hsu YY, Graham RW. Transport processes in boiling and two-phase systems.
Chaps 5 and 6, 2. New York: Hemisphere; 1976.
[84] Firtz W, Ende W. Verdampfungsvorgang kinematographischen aufnahmen und
dampbasen. J Appl Phys 1936;25:391.
[85] Plesset MS, Zwick SA. The growth of vapor bubbles in superheated liquid. J Appl
Phys 1954;25:493500.
493
Renewable and Sustainable Energy Reviews 78 (2017) 466494
[86] Forster HK, Zuber N. Growth of a vapour bubble in a superheated liquid. J Appl
Phys 1954;25:4748.
[87] Labuntsov DA, Kolchugin BA, Golovin VS, Zakharova EA, Vladimirova LN. Study
of the growth of bubbles during boiling of saturated water within a wide range of
pressures by means of high-speed moving pictures. UCS 1964;536.423(1):4049.
[88] Cole R, Shulman HL. Bubble growth rate at high Jakob number. Int J Heat Mass
Transf 1966;9:137790.
[89] Mikic BB, Rohsenow WM. Bubble growth rates in non-uniform temperature eld.
Prog Heat Mass Transf 1969;2:28393.
[90] Van Stralen SJD, Sluyter WM. Local temperature uctuations in saturated pool
boiling of pure liquids and binary mixtures. Int J Heat Mass Transf
1969;12:18798.
[91] Copper MG. The microlayer and bubble growth in nucleate pool boiling. Int J Heat
Mass Transf 1969;12:91533.
[92] Mikic BB, Rohsenow WM, Grith P. On bubble growth rates. Int J Heat Mass
Transf 1970;13:64766.
[93] Miyatake O, Tanaka I. Bubble growth in uniformly superheated water at reduced
pressures, Part 1: numerical analysis and derivation of a simplied expression.
Trans JSME Ser B 1982;48:35563.
[94] Zhao YH, Masuoka T, Tsuruta T. Unied theoretical prediction of fully developed
mucleate boiling and critical heat ux based on a dynamic microlayer model. Int J
Heat Mass Transf 2002;45:318997.
[95] Haustein HD, Gany A, Dietze GF, Elias E, Kneer R. The dynamics of bubble
growth at medium-high superheat: boiling in an innite medium and on a wall. J
Heat Transf 2013;135:0715018.
[96] Lesage F, Siedel S, Cotton JS, Robinson AJ. A mathematical model for predicting
bubble growth for low Bond and Jakob number nucleate boiling. Chem Eng Sci
2014;112:3546.
[97] Zuber N. The dynamics of vapor bubbles in nonuniform temperature elds. Int J
Heat Mass Transf 1961;2:8398.
[98] Jakob M, Linke W. Heat transfer from a horizontal plate. Forsch Geb Ing
1933;4:7581.
[99] Jakob M, Fritz W. Versuche uber den verdampfungsvorgang. Forsch auf dem Geb
Des Ing 1935;2:43547.
[100] Jakob M. Heat Transfer, Chapter 29, 1. New York: Wiley and Sons; 1949.
[101] Peebles FN, Garber HJ. Study on motion of gas bubbles in liquids. Chem Eng Prog
1953;49:8897.
[102] McFadden PW, Grassman P. The relation between bubble frequency and diameter
during nucleation pool boiling. Int J Heat Mass Transf 1962;5:16973.
[103] Zuber N. Nucleate boiling. The region of isolated bubbles and the similarity with
natural convection. Int J Heat Mass Transf 1963;6:5378.
[104] Ivey HJ. Relationships between bubble frequency, departure diameter and rise
velocity in nucleate boiling. Int J Heat Mass Transf 1967;10:102340.
[105] Malenkov IG. Detachment frequency as a function of size of vapor bubbles
(transl.). Inzh Fiz Zh 1971;20(2):988.
[106] Katto Y, Yokoya S. Behavior of a vapour mass in saturated nucleate and transition
pool boiling. Heat Transf Jpn Resour 1976;5:4565.
[107] Stephan K. Heat transfer in condensation and boiling. New York: Springer; 1992.
[108] Kumada T, Sakashita H. Pool boiling heat transfer-I. measurement and semi-
emprical relations of detachment frequencies of coalesced bubbles. Int J Heat
Mass Transf 1995;38(6):96977.
[109] Sakashita H, Ono A. Boiling behaviours and critical heat ux on a horizontal plate
in saturated pool boiling of water at high pressures. Int J Heat Mass Transf
2009;52:74450.
[110] Miglani A, Joo D, Basu S, Kumar R. Nucleation dynamics and pool boiling
characteristics of high pressure refrigerant using thermochromic crystals. Int J
Heat Mass Transf 2013;60:188200.
[111] Hamzekhani S, Falahieh MM, Kamalizadeh MR, Nazari Z. Experimental study on
bubble departure frequency for pool boiling of water/NaCl solutions. Heat Mass
Transf 2015;51:131320.
[112] Tong W, Wu C, Toh KC, Duan F. Experimental study of bubble dynamics in highly
wetting dielectric liquid pool boiling through high speed video. In: Proceedings of
the 15th IEEE ITHERM Conference. 2016.
[113] Gong S, Cheng P. Lattice Boltzmann simulation of periodic bubble nucleation,
growth and departure from a heated surface in pool boiling. Int J Heat Mass
Transf 2013;64:12232.
[114] Nakayama N, Daikoku T, Kuwahara H, Nakajima T. Dynamic model of enhanced
boiling heat transfer on porous surfaces, Part-I and Part II. Trans ASME J Heat
Transf 1980;102:44556.
[115] Chien LH, Webb RL. Measurement of bubble dynamics on an enhanced boiling
surface. Exp Therm Fluid Sci 1998;16:17786.
[116] Pascual CC, Jeter SM, Abdel-Khalik SI. Visualization of boiling bubble dynamics
using a uniformly heated transparent surface. Tech Note Int J Mass Transf
2002;45:6916.
[117] Gao M, Cheng P, Quan X. An experimental investigation on eects of an electric
eld on bubble growth on a small heater in pool boiling. Int J Heat Mass Transf
2013;67:98791.
[118] Siedel S, Cioulachtjian S, Bonjour J. Experimental analysis of bubble growth,
departure and interactions during pool boiling on articial nucleation sites. Exp
Therm Fluid Sci 2008;32:150411.
[119] Hetsroni G, Mosyak A, Pogrebnyak E, Sher I, Segal Z. Bubble growth in saturated
pool boiling in water and surfactant solution. Int J Multiph Flow 2006;32:15982.
[120] Zuber N, Hydrodynamics aspects of boiling heat transfer, US AEC Report AECU
4439, 1959.
[121] Dhir VK, Abarajith HS, Li D. Bubble dynamics and heat transfer during pool and
ow boiling. Heat Transf Eng 2007;28(7):60824.
[122] Qiu DM, Dhir VK, Hasan MM, Chao D, Neumann E, Yee G, Witherow J. Single
Bubble Dynamics during Nucleate Boiling under Microgravity Conditions.
Engineering foundation Conference on microgravity uid physics and heat
transfer. Honolulu, HI; 1999.
[123] Hutter C, Kenning DBR, Seane K, Karayiannis TG, Lin H, Cummins G, Walton
R.L. Mohanty, M.K. Das
AJ. Experimental pool boiling investigations of FC-72 on silicon with articial
cavities and integrated temperature microsensors. Exp Therm Fluid Sci
2010;34:42233.
[124] Zhang X, Chen J, Xiong W, Jin T. Visualization study of nucleate pool boiling of
liquid nitrogen with quasi-steady heat input. Cryogenics 2015;72:1421.
[125] Coulibaly A, Lin X, Bi J, Christopher DM. Bubble coalescence at constant wall
temperatures during pool subcooled nucleate pool boiling. Exp Therm Fluid Sci
2013;44:20918.
[126] Abarajith HS, Dhir VK. Eect of Contact angle on the dynamics of a single bubble
during pool boiling using numerical simulations. Proceeding of the IMECE2002
ASME Int Mech Eng Cong & Exposition. New Orleans, Louisiana; 2002.
[127] Mukhejee A, Kandlikar SG. Numerical study of single bubbles with dynamic
contact angle during nucleate pool boiling. Int J Heat Mass Transf
2007;50:12738.
[128] Mukherjee A, Dhir VK. Study of lateral merger of vapor bubbles during nucleate
pool boiling. J Heat Transf 2004;126:102339.
[129] Hazi G, Markus A. On the bubble departure diameter and released frequency
based on numerical simulation results. Int J Heat Mass Transf 2009;52:147280.
[130] Gong S, Cheng P, Quan X. Two-dimensional mesoscale simulation of saturated
pool boiling from rough surface. Part I: Bubble nucleation in a single cavity at low
494
Renewable and Sustainable Energy Reviews 78 (2017) 466494
superheats. Int J Heat Mass Transf 2016;100:92737.
[131] Abarajith HS, Dhir VK, Son G. Numerical simulation of the dynamics of the
multiple bubble merger during pool boiling under reduced gravity conditions. In:
Proceedings of the Japan USA seminar on multiphase ow. Moriyama, Japan,
2004.
[132] Colin C, Kannengieser O, Bergez W, Lebon M, Sebilleau J, Sagan M, Tanguy S.
Nucleate poll boiling in microgravity: recent progress and future prospects. C R
Mec 2017;345:2134.
[133] Zhang L, Shoji M. Nucleation site interaction in pool boiling on the articial
surface. Int J Heat Mass Transf 2003;46:51322.
[134] Nimkar ND, Bhavnani SH, Jaeger RC. Eect of nucleation site spacing on the pool
boiling characteristics of a statured surface. Int J Heat Mass Transf
2006;49:282939.
[135] Bon B, Klausner J, Mckenna E. An investigation of pool boiling heat transfer on
single crystal surfaces and a dense array of cylindrical cavities. J Heat Transf
2013;135:12150113.
[136] Akiyama M, Tachibana H, Ogawa N. Eects of system pressure on bubble growth
rate. Trans JSME 1969;35269:11726, [in Japanese].
[137] Raad T, Myers JE. Nucleation studies on pool boiling on thin plates using liquid
crystals. AIChE J 1971;17:12601.