International Journal of Heat and Mass Transfer 73 (2014) 110–117

Contents lists available at ScienceDirect

International Journal of Heat and Mass Transfer

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

Instantaneous heat transfer for large drops levitating over a hot surface
Tadeusz Orzechowski ⇑, Sylwia Wciślik
Kielce University of Technology, Environmental Engineering Dep., Heat Engineering Div., Aleja Tysia˛clecia Państwa Polskiego 7, 25-314 Kielce, Poland

a r t i c l e i n f o a b s t r a c t

Article history: The paper deals with the process of evaporation of large water drops with the initial mass of 1 g deposited
Received 29 June 2013 on a hot surface, the temperature of which is higher than Leidenfrost point. The behavior of water drops
Received in revised form 18 November 2013 was examined at the test stand, at which three independent measurement paths were available, namely
Accepted 4 February 2014
those of instantaneous mass measurements, temperature recording and sequential recording of the ther-
Available online 26 February 2014
mal field of the drop upper surface. Thus obtained sets of drop mass, drop temperature, and its area size,
for pre-defined temperature of the heating cylinder having a great thermal capacity, were used to com-
Keywords:
pute instantaneous values of the heat transfer coefficient. The methodology of investigations was dis-
Water drop
Leidenfrost regime
cussed in detail. Measurement uncertainties were analysed using the total differential method. On the
Mass flux basis of thermographic images, recorded with a thermovision camera, of the drop upper surfaces, sub-
Heat transfer stantial thermal diversity of drops was found. The difference between the maximum and minimum tem-
Infrared mapping peratures periodically amounts to above 9 °C, and standard deviation from the area of their upper
surfaces amounts even to 2 °C. Measured instantaneous values of the heat transfer coefficient were
approximated with a power function, dependent on the heating surface temperature and a momentary
drop size. This relation was selected in accordance with the developed approximation procedure, at
the imposed condition of the minimum of the mean square error. Consequently, a constant value of expo-
nent and a dependence on the heating wall temperature were obtained. It was shown that values of mea-
sured and approximated heat transfer coefficients are contained in the interval defined by the value of
the relative error ranging from 13% to +9%. An approximation of the perpendicular projection of the
drop area in the form of polynomial is proposed. Under these assumptions, an analytical solution to
the energy balance equation is given. Exemplary computations provided in the study indicate a very good
effectiveness of the proposed method.
Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Nuclear power plants consume large amounts of water, even

A demand for modern thermal devices that are characterised by
small size and high cooling or heating power resulted in a mark-
edly increased interest in heat transfer enhancement. The most
effective ways of approaching this issue is to apply phase change
phenomena. Advancement in such technologies came as a spin-
off from space exploration and nuclear power plants, a rapid devel-
opment of which started in the middle of the twentieth century.
Despite intensive research into the subject, sufficiently accurate
schemes necessary to design efficient cooling systems have not
been developed. It is also true about very large scale integration
electronic devices [1], in which in addition to the dispersion of high
density heat fluxes, a constant temperature and the lowest possible
noise level are also required [2].
⇑ Corresponding author.
E-mail addresses: todek@tu.kielce.pl (T. Orzechowski), sylwiazw@tu.kielce.pl
(S. Wciślik).
twice as much as conventional power stations of the same capacity
[3]. That creates a demand for innovative cooling technologies
based on two phase mixtures. In study [4], it was found that the
droplets Sauter mean diameter could decrease when the liquid
flows through the grid spacer, which results in an increase in inter-
facial heat transfer surface area. The change in various flow condi-
tions, namely vapor injection velocity, heater temperature, droplet
size, and droplet flow rate inside a heated rod bundle was experi-
mentally investigated in [5].
Understanding the phenomenon of evaporating droplet is very
important for such devices as diesel engines [6,7], liquid–fuel rock-
et engines, aircraft jet engines and industrial furnaces. Improve-
ments in efficiency and power output as well as reliability and
durability depend on efficient cooling systems. From a variety of
cooling technologies discussed in [8,9], droplet injection to maxi-
mise the cooling surface [10] may be effectively applied.
Two-phase water droplets-gas flows are also found in high tem-
perature gas flow cooling [11], heat recovery in ventilation [12], air

http://dx.doi.org/10.1016/j.ijheatmasstransfer.2014.02.008
0017-9310/Ó 2014 Elsevier Ltd. All rights reserved.
 T. Orzechowski, S. Wciślik / International Journal of Heat and Mass Transfer 73 (2014) 110–117
Nomenclature
A area of drop projection on hot surface
a polynomial coefficient
b coefficient of the Maclaurin series
Ch coefficient defined in the text
cp specific heat
hfg enthalpy of vaporisation
M mass flux
m mass
n exponent
T temperature
t time
humidifying and drying in evaporation chambers of air condition-
ing systems [13], humidification process in spray towers [14], draft
cooling towers [15], evaporation in fire [16], and two-phase closed
systems in heat pipes where liquid and gas phases are close to the
saturation equilibrium state [17] and heat transfer limitation is
reached [18,19].
In droplet evaporation discussed above, an interaction with so-
lid surface is a common issue. At sufficiently high temperatures,
vapor created on the droplet surface nearest to the wall produces
a lubrication layer, from which the droplet is suspended. That is
known as the Leidenfrost effect [20]. At surface temperatures
above the Leidenfrost point, the droplet levitates on a thin vapor
layer through which heat is transferred. The Leidenfrost tempera-
ture may be determined from a droplet evaporation curve, where
the droplet lifetime is plotted versus wall superheat. At that tem-
perature, the vapor layer prevents any significant contact between
the droplet and the surface, and the droplet evaporation time
reaches its maximum. In the literature, different values for that
point are given. In [21], large variations in the Leidenfrost temper-
ature for water are discussed. It is found that the differences result
from droplet size and its mass, method of deposition, subcooling,
the heating surface thermal properties, its condition, ambient pres-
sure, and liquid purity. The complexity of the phenomenon there-
fore requires detailed investigations into the impact of separate
factors on the droplet evaporation.
Wettability of the surface is one of the most important param-
eters affecting the process of heat transfer. In [22,23], a change in
evaporation characteristics of water droplets on hydrophobic and
hydrophilic surfaces is discussed. The evaporation is strongly
dependent on convection. The air motion modifies the mass trans-
fer and thermal parameters of the boundary layer with the corre-
sponding change in the droplet shape [24]. The coupling between
the droplet shape and the applied electric field may produce inter-
nal droplet flow, which decreases heat transfer [25]. Evolution of
the thermal and internal flow coexisting with evaporation is also
discussed in [26]. It is concluded that the evaporation rate is not
a constant value. It depends on the energy balance of latent and
sensible heat. An interaction with the hot surface may change
the evaporation significantly. When a droplet is deposited onto a
hot surface, heat is transferred from the solid and consumed to
heat up a fluid and to change phase from liquid to gas. At a suffi-
ciently high surface temperature, a stable vapor layer is created
and heat transfer is reduced. The dependence of the rate of evapo-
ration on droplet shape and size are studied numerically in [27,28].
The evaporation rates of small droplets of different diameters lo-
cated on a horizontal surface are studied in [29], where a faster
evaporation of small droplets is investigated experimentally.
Vaporisation and chemical conversion are simulated for different
ambient conditions in [30].
111
Greek symbols
a heat transfer coefficient
d relative error
h temperature above saturation
Subscripts
0 initial
d drop
s saturation
w heating surface
Evaporation of the droplets is studied in many areas of technol-
ogy. Various investigations conducted into the subject focus on the
physics of the phenomenon, which depends strongly on the liquid
and its components, thermal and flow conditions, interaction with
the surface, etc. Although the problem has been thoroughly ana-
lysed, the mechanisms of evaporation are still not fully clear.
The present paper examines the behavior of a large single drop
levitating over a hot surface, unsteady mass of the drop, and heat
transfer. For the sake of the experiment, it was necessary to devel-
op specific methodology based on measurement uncertainty. The
outcome of the experimental investigations is provided in the form
of changes in the drop weight, on the basis of which instantaneous
heat transfer coefficient is determined.
2. Test facility
Fig. 1 shows a diagram of the research stand, the main element
of which is a cooper heating cylinder, 3.5 cm in diameter. It upper
surface is shaped like a bowl having a very large radius of curva-
ture. Support for the heating system, additionally installed, allows
its independent levelling, which facilitates stable drop depositing.
300 W heater was wrapped on the cylinder lateral side. The
desirable temperature of the system was obtained by setting the
voltage from the autotransformer directly connected to the mains.
The voltmeter in the electrical circuit was used to roughly measure
the voltage supplied. The parameter that was precisely controlled
was the heating surface temperature. For temperature control, a
K-type coat thermocouple, 0.5 mm in diameter, was soldered in
the opening drilled in the bottom. The weld was located just be-
neath the surface and the thermocouple leads were connected to
Fig. 1. Diagram of the test apparatus: 1 – droplet of water, 2 – copper cylinder, 3 –
wrapped heater, 4 – thermocouple, 5 – electronic scales, 6 – infrared camera,
7 – digital camera, 8 – mirror, 9 – A/D signal processing system, 10 – computer, 11 –
electric power supply unit (autotransformer).
112 T. Orzechowski, S. Wciślik / International Journal of Heat and Mass Transfer 73 (2014) 110–117
the AC card, from which a signal was directly sent to the data
acquisition station and recorded on the computer disc.
The complete heating apparatus was placed on special-purpose
electronic scales that have the maximum load of 500 g, sensitivity
of 0.001 g and accuracy of 0.01 g. It is equipped with RS-232C
interface for remote control and measurement data transmission.
The data are collected in RAM, and on completion of the measure-
ment, they are recorded on the disc. Under stationary or slow-vary-
ing conditions, the scales reading is only transmitted when three
successive readings in the device internal system are identical. This
function can be switched off, and if that is the case, due to technical
issues related to analog-to-digital signal conversion, the signals
can be recorded under non-stationary conditions at the maximum
frequency of 10 Hz.
To assure thermal uniformity of the heating cylinder, its upper
surface was carefully coated in paint with the admissible operating
temperature of over 500 °C. The emissivity coefficient for the paint
was determined in the initial calibration investigations. That was
done as follows, after conditioning the system for several minutes
for constant, pre-set electric power, the paint image was recorded
with a thermal camera. The surface emissivity was selected in such
a way so that so that thermocouple reading would coincide with
mean temperature reading for a circular area, approx. 1 cm in
diameter, which covered the point of thermocouple connection.
The thermovision camera software allows 0.01 step variation in
the coefficient. The value selected in this manner amounted to
0.87 and was constant in the temperature range of 100 to 450 °C.
The quality of paint coating and the thermal uniformity of the lat-
ter were assessed on the basis of the analysis of a selected surface
area, for which the computed value of standard deviation did not
exceed 0.2 K.
In the investigations, VarioCAMÒ hr head thermovision camera
by Jenoptik company was used. The camera operates in infrared
long wavelength (LW) 7.514 lm spectral range. 640  480 pixel
microbolometer matrix (Focal Plane Array) is the component that
converts a part of infrared radiation emitted by a body to an elec-
tric signal. The detector is thermoelectrically stabilised using Pel-
tier phenomenon, due to which it is independent of ambient
thermal conditions. The analog signal conversion system of the
camera makes it possible to record measured temperature fields
in real time with the frequency of 50 Hz for full image recording.
The system is equipped with standard lens, Field of View
(30  23)0 and Instantaneous Field of View 0.8 mrad. Bolometric
sensors do not need cooling and operate correctly under standard
ambient conditions. At 30 °C, their thermal resolution is below
0.08 K. Measurement accuracy, declared by the manufacturer is
±1.5 K for the temperature range 0–100 °C, and ±2% outside this
range. It should be noted that the values given above refer to con-
ditions, under which it is possible to only roughly state, and intro-
duce to the system appropriate corrections that account for
infrared radiation components. Those result from optical absorp-
tion and scattering that occurs between the camera and the surface
observed. That also refers to the emission of radiation from neigh-
boring objects, especially when their temperature is substantially
higher from that of the investigated object. The magnitude of er-
rors is also related to how accurately the coefficient of the self-
emission was chosen. The value of the coefficient can vary with
the temperature of object measured, or of its parts.
Under laboratory conditions, after the measuring system has
been individually calibrated, the accuracy stated by the manufac-
turer can be significantly enhanced [31]. In such investigations,
however, it is more often desirable to know temperature change
or difference with respect to point or area of reference. Then, it
possible to approach the accuracy that is close to the device sensi-
tivity level, which is feasible, although extremely difficult to
achieve in practice.
The thermovision camera, described above, was also used to re-
cord images of a drop deposited on the heating cylinder surface. A
mirror of polished stainless steel was made for that sake. The mir-
ror was positioned in such a way that a reflected image of the pro-
jection of the drop upper surface could possibly accurately
represent the real shape of the drop. The necessary angle of incli-
nation of the mirror was determined by comparing the lengths of
the two perpendicular diameters of the cylinder front end. The
thermovision camera was located at 30 cm distance from the
measurement stand. That allowed reducing the side components
of infrared radiation to the camera. Having made such an assump-
tion, the mirror reflection coefficient was determined in the cali-
bration procedure conducted in the manner identical to that for
the paint emissivity coefficient.
3. Procedure
The experimental setup provides three independent measure-
ment paths, namely those of the scales, temperature measurement
and the camera. The first one is intended to record the mass loss in
a drop deposited on the hot horizontal surface of the heating cylin-
der. The surface temperature is controlled by a thermocouple, and
additionally, by the thermovision camera. On the basis of a re-
corded sequence of thermographic images, the surface tempera-
ture field of the drop is also observed. On these bases the area of
the drop perpendicular projection onto the heating surface is
computed.
In drop mass measurement, power is supplied to the heater,
which is connected to the mains using an earthed wire. The wire
is suspended in such a manner so that it does not generate weigh-
ing errors resulting from elastic action. The mass measurement
system was switched on at least 12 h in advance in order to elim-
inate the interaction of the power supply wire and the thermocou-
ple compensation wire. Power supply was switched on at least two
hours prior to measurements, which assured thermal stabilisation
of the whole system. Before starting measurements, the scales and
surface temperature readings were observed for approx. 10 min.
The duration of observations was always over twice as long as
the expected time of the drop evaporation. The measurements
started if both parameters, i.e. the scales and thermocouple read-
ings, remained stable.
A large drop, having a mass of over 1 g, was deposited on the
heated surface of the cylinder. The analysis of the measurement re-
sults always started after a while, at the instant when the drop
mass was 1 g. In this way, it was possible to eliminate the period
Fig. 2. Drop mass loss as a function of time at the heating surface mean
temperature, Tw = 392.5 °C.
 T. Orzechowski, S. Wciślik / International Journal of Heat and Mass Transfer 73 (2014) 110–117 113

of the initial stabilisation of the measurement system, related to

drop deposition.
The experiment was conducted for a drop of water under ambi-
ent pressure. The frequency of the scales readings and the drop
surface thermogram registration was always the same and equaled
2 Hz. For each surface temperature, at least 30 measurements were
taken, which allowed conducting a statistical analysis at appropri-
ately high level of confidence.
A typical set of measurements is shown in Fig. 2, in which a
change in the drop mass over time is shown for the pre-set temper-
ature of the heating surface. The mean value of the latter was
Tw = 392.5 °C.
Because of a growing relative error related to the accuracy of
the scales used in the experiment, the measurements ended when
the drop was approx. 0.2 g, i.e., twenty times greater than the
scales accuracy stated by the manufacturer (see Fig. 2). The dura-
tion of all measurements series was assumed to be 200 s.
During the drop evaporation, the heating surface temperature
also changed. Using recorded thermocouple readings, standard
deviation was computed. For the heating surface temperature
Tw = 392.5 °C, it equals 2.35 K in the 250 s time range considered.
The mass of the copper cylinder, including the heater, was over
360 g and it was much greater than the mass of water drop (1 g).
Additionally, during the drop evaporation, a continuous power
supply is delivered to the heater. Taking into account a large ther-
mal capacity of the heating system and a relatively low value of the
film boiling heat flux, it is supposed that fluctuations and non-uni-
formities in the heating surface temperature result mainly from
random rotational movements of the drop, related to a change in
its shape (Fig. 4).
For the set shown in Fig. 2, measurement instantaneous mean
values were determined. The mean curve thus obtained (broken
line in Fig. 2), was further employed to compute the heat transfer
coefficient. With respect to the mean value, instantaneous stan-
dard deviation (as a function of drop mass loss) was computed,
which is shown in Fig. 3.
A gradual increase in standard deviation value that can be ob-
served in the figure results from different evaporation in individual
measurement series. The value of standard deviation is stabilised
at roughly half of the drop initial mass. Outside the range consid-
ered, i.e. for drop mass below 0.2 g, standard deviation value in-
creases rapidly.

4. Visualisation
Parallel to mass loss measurements with the thermovision cam-
era, the drop surface was observed. It was noted that the thermal
field on its upper surface was greatly diversified. That results from
non-symmetrical vapor flow from under the bottom surface. The
Fig. 4. Thermovision images showing the drop projection onto the heating surface
for temperature of the heating cylinder, Tw = 392.5 °C.
vapor, swilling the drop sides causes an observable evolution of
its shape, which is illustrated in Fig. 4.
For each of the presented thermograms, minimum, maximum,
mean values of temperature and standard deviation were given.
To eliminate disturbances related to a directional emissivity when
observations were conducted at large angles, all values were deter-
mined within the area that covered approx. 80% of the whole area
of the drop (Fig. 4a).
The investigations that were conducted contradict assumptions
of the symmetry. In a droplet evaporation, turbulent convection
movements are observed, which cause mixing of liquid portions
that have different temperatures. That results in both the fluctua-
tions of the mean temperature and uncontrolled changes in geom-
etry. These phenomena are illustrated in thermograms in Fig. 4.

5. Methodology

The instantaneous value of the heat transfer coefficient that is

Fig. 3. Standard deviation error from the set presented in Fig. 2. sought is closely related to the amount of the evaporated drop
114 T. Orzechowski, S. Wciślik / International Journal of Heat and Mass Transfer 73 (2014) 110–117

mass. For high values of the heating surface temperature Tw com-

pared with the drop mean temperature Td, it can be assumed that
heat transfer results from convection and radiation between the
heating surface and the drop. Both phenomena can be described
by the value of heat transfer coefficient of two components: con-
vection and radiation.
The heat taken up by the drop bottom surface A is partially used
to heat up the evaporated liquid to the saturation temperature Ts
and to cause a transition from liquid to gaseous phase, which can
be expressed by the following equation:

dm dm
aðT w  T d ÞA ¼ hfg  cp ðT s  T d Þ ð1Þ
dt dt
where dm/dt is the flux of evaporated mass, hfg and cp are phase
change enthalpy and specific heat, respectively.
In order to determine the instantaneous value of heat transfer Fig. 6. Difference between computed and approximated drop mass flux as a
coefficient a, it is necessary to give all the quantities found in Eq. function of time at mean heating surface temperature Tw = 392.5 °C.

(1), namely Tw, Td, A and dm/dt. All the remaining ones, i.e., Ts, hfg
and cp, are taken for water from thermodynamic tables.
For computations, constant values of the heating surface tem- values and those approximated as a function of time are presented
perature and that of the drop were assumed. Tw is an arithmetic in Fig. 6. Standard deviation, which is 1.41  104 g/s, is also shown
mean of the thermocouple readings. Td is a corresponding arithme- in this figure. This value is used further onto assess the error of the
tic mean of mean values of the drop upper surface, computed for method.
the area shown in Fig. 4. A was replaced with the drop projection Substituting thus obtained quantities for (1), an instantaneous
onto the heating surface, which was computed every time on the value of heat transfer coefficient is determined. The values of the
basis of appropriate thermogram [32]. coefficient as a function of the drop projection A on the heating
The last quantity necessary for computing, from Eq. (1), the surface for constant temperature Tw = 392.5 °C are shown in
sought heat transfer coefficient, is the mass flux dm/dt. Fig. 7.
Current measurement of the drop mass was taken at 2 Hz fre- Heat transfer coefficient is measured in an indirect way. Its
quency for 250 s. It produced as set of 500 points for a single series. value is computed after all the quantities in the equation in which
Such a great number of points makes it possible to apply numerical it is found are determined. In this case, it is Eq. (1), the solution to
differentiation combined with smoothing, which is available in which, due to measured quantities, can be written in the following
many commercial computational packages. Such a procedure was form:
applied to the average curve, determined from all measurement
 
series, of drop mass change (Fig. 2). After the curve smoothing dm
and differentiation, local values of mass flux of evaporated drop a ¼ f T w ; T d ; A; ð2Þ
dt
were obtained, which is shown in Fig. 5.
At the assumption that the mass flux under discussion can be de- The equation provides the basis for uncertainty analysis, which was
scribed by a holomorfic function, the former can be presented in the conducted with the total differential method, using the following
form of a power series. For accurate representation of such a dependence:
function, it is necessary to account for indefinitely many terms.  
@f @f @f @f dm
Due to the fact that the experimental values of the function are bur- Da ¼ DT w þ DT d þ DA þ dm
D ð3Þ
dened with many errors, it is justifiable to choose the polynomial of @T w @T d @A @ dt dt
least possible degree, then, the approximation error is the rest of
such a series. For data in Fig. 5, approximation with fourth degree Heat transfer coefficient takes on different instantaneous
power series was assumed, the determination coefficient of which values that are burdened with variable error, which is shown in
is very high and equal to 0.998. Differences between experimental Fig. 8.

Fig. 5. Drop mass flux, calculated in accordance with mean mass loss (Fig. 2), and Fig. 7. Heat transfer coefficient as a function of water drop area for the heating
its approximation with fourth degree polynomial as a function of time. surface temperature, Tw = 392.5 °C.
 T. Orzechowski, S. Wciślik / International Journal of Heat and Mass Transfer 73 (2014) 110–117 115

where n is a constant exponent, and coefficient Ch = C(h) is different

for individual measurement curves, and varies with the excess of
heating surface temperature over the saturation temperature
h = Tw  Ts.
The procedure of exponent n and coefficient Ch selection for
data in Fig. 9 was performed at the condition of the minimum of
the mean square error. In this way, the dependence described with
formula (5) was obtained, in which A is given in m2, and the result
is heat transfer coefficient expressed in kW/m2 K.

A0:33
a ¼ C h An ¼ ðkW=m2 KÞ ð5Þ
76:3  0:00039h2
Values of a computed on the basis of formula (5) are shown in Fig. 9.
Differences between values of the heat transfer coefficient deter-
mined on the basis of experimental data and their equivalents com-
Fig. 8. Instantaneous values of heat transfer coefficient and its absolute error
puted in accordance with approximation dependence (5) can be
(Tw = 392.5 °C).
assessed on the basis of the relative error d:
aexperiment  apredicted
d¼ 100% ð6Þ
6. Discussion of the results aexperiment
Measurements of water drop evaporation were conducted in which is 9% for values exceeding those measured ones, and 13% for
accordance with the methodology described above, at ambient heat transfer coefficient values lower than the experimental ones,
pressure for mean values of the heating surface temperature: which can be seen in Fig. 10.
297, 310.5, 337.5, 372, 392.5, and 404.5 °C. Instantaneous values The energy balance (1) is a nonlinear equation, in which a time-
of heat transfer coefficient were computed for each series. They dependent area A of the drop projection on the heating surface is
are presented in Fig. 9 as a function of the drop projection on the found (see Fig. 4). In order to calculate the quantity of heat ab-
heating surface. sorbed by the drop, it is necessary to provide instantaneous values
At the beginning of evaporation of large-sized drops, for each of A. This quantity is also required to determine the drop mass
heating surface temperature, this coefficient takes on the lowest change over time, which, taking into account the relation (5), can
value, which then grows gradually with a decrease in drop size. be calculated from the equation:
A probable reason for such distribution could be related to the time  nþ1  nþ1
C h ðT w  T d ÞAnþ1
0 A A dm
of the gaseous phase flow below the drop surface. The vapor gen-  ¼ M0 ¼ ð7Þ
cp ðT s  T d Þ þ hfg A0 A0 dt
erated on the drop bottom surface is, at the same time, a carrier of
the heat received from the heating surface. The heat, in accordance where: M0 is the vaporised mass flow, and A0 – the surface area.
with balance (1), is used to continuously generate vapor. It should Both values are given for the initial moment, i.e., for time t = 0 s.
be noted that, the amount of radiation heat transfer is low when Assuming that the area is described by the holomorphic func-
compared with that through convection, and it does not exceed tion, its value can be represented as a power series of the form:
5%. Vapor flowing under the drop is heated, and thus the temper- A
ature of vapor flowing from underneath the drop is the higher, the ¼ 1 þ a1 t þ a2 t 2 þ a3 t 3 þ    ð8Þ
Ao
longer it remains in direct contact with the heating surface. For lar-
ger drops, the contact time is longer, so the mean temperature is in which the unknown coefficients ai can be calculated by means of
higher, which results in lower values of heat transfer coefficient, the approximation of the experimental data using typical regression
when compared with smaller-sized drops. procedure.
Experimental data indicate exponential dependence of heat Substituting the above into (7), the equation is obtained, the
transfer coefficient on the surface. For that reason, the following solution of which can be found using the Maclaurin series expan-
form of the function was assumed for further considerations: sion of non-linear term of the left-hand side of the Eq. (7). As a re-
sult we receive:
a ¼ C h An ð4Þ

Fig. 9. Values of heat transfer coefficient, computed on the basis of measurements

and approximated with power function (4) for constant exponent n = 0.33, as a Fig. 10. Boundary values of the heat transfer coefficient determined on the basis of
function of the drop projection on the heating surface. measurements and computed in accordance with dependence (5).
116 T. Orzechowski, S. Wciślik / International Journal of Heat and Mass Transfer 73 (2014) 110–117
dm
M 0 ð1 þ b1  t þ b2  t 2 þ b3  t3 þ   Þ ¼
dt
where the coefficients bi depend on the exponent n and the polyno-
mial coefficients ai defined in (8):
b1 ¼ ðn þ 1Þa1
1
b2 ¼ ðn þ 1Þ n a21
2
1
b3 ¼ ðn þ 1Þnðn  1Þ a31 þ ðn þ 1Þ n a1 a2 þ ðn þ 1Þ a3
6
Integration of the Eq. (9) gives the following expression:
 
1 1 1
m0  m ¼ M 0 t þ b1  t2 þ b2  t 3 þ b3  t 4 þ   
2 3 4
where, according the initial condition at t = 0 s, the mass m = m0 is
known.
Registered thermographic sequences (see Fig. 4) are used to cal-
culate the instantaneous values of the area of the drop orthogonal
projection for different temperatures of the heating surfaces, as
shown in Fig. 11. The experimental data are approximated by
third-degree polynomials, for which the coefficients of determina-
tion for all measurement series where higher than 0.98. All instan-
taneous values of the area, given in the relation to the initial value,
are raised to the power of n + 1, which is found in the equation. As
a result, it is possible to compare, in a single chart, those values
with the values obtained from the expansion of this expression
in the Maclaurin series (see Fig. 11).
The obtained solution (11) is compared with the experimental
data. That results in obtaining drop mass change versus time.
Fig. 12 presents the computational results for the case discussed.
Fig. 11. The relative change in the drop projection (A/A0)n+1 for different temper-
atures of the heating surface: values measured in the experiment, fitted by the
third-order polynomial, and calculated after expending in the Maclaurin series.
Fig. 12. Experimental and calculated results of the drop mass change versus time.
ð9Þ
ð10Þ
ð11Þ
Fig. 13. The relative errors in mass calculated in accordance with the analytical
solution (11).
A thin dashed line in this figure represents selected extreme mea-
surement series for the recorded drop mass loss versus time,
whereas a scatter line denotes the average from all measurement
series. Both dashed lines indicate experimental limits. As can be
seen in Fig. 12, a very good congruence between measurement
data and theoretical computations is obtained.
The analytical solution compliance with the experimental data
can be assessed on the basis of calculated relative error between
measured and computed values Eq. (11). For the data shown in
Fig. 12, the relative errors are presented in Fig. 13.
7. Conclusions
It is difficult to provide a theoretical description of evaporation
of a liquid drop deposited on a hot surface, the temperature of
which is higher than Leidenfrost point. That is caused by simulta-
neously occurring processes of heat and mass transfer in a system
in which a drop is freely floating on a gaseous cushion. Composed
gas flow in the boundary layer inducts intensive, convective move-
ments in the near-surface zone. Those lead to the mixing inside the
drop of liquid portions that have different temperatures. That re-
sults in mean temperature fluctuations and uncontrolled changes
in geometry, which can be seen in thermograms in Fig. 4. Drop
upper surface thermograms recorded with the thermovision cam-
era show great thermal variation of drops. Differences between the
maximum and minimum temperature values are occasionally
above 9 °C, and standard deviation from the area covering 80%
is substantial and amounts to 2 °C.
The drop behavior was examined at the stand which comprised
three independent measurement paths, namely instantaneous
measurement of the weight of evaporating drop, temperature mea-
surement and sequential recording of the thermal field of the drop
upper surface. The data sets on drop mass, temperature and surface
size for pre-set temperature of the heating cylinder having vast
thermal capacity were used to compute instantaneous values of
the heat transfer coefficient. The heat taken over by the drop bot-
tom surface is partially used to heat up the evaporated liquid to
saturation temperature and to change phase from liquid to gas-
eous. For the assumptions made above, the energy balance was
written in the form of a differential equation. The latter, after
determining all the quantities that occur in it, was employed to
compute instantaneous values of the heat transfer coefficient.
Following the methodology described in Section 4, investiga-
tions into a large water drop having the initial mass 1 g were con-
ducted for six different temperatures of the heating surface (297,
310.5, 337.5, 372, 392.5, and 404.5 °C). On the basis of measured
mass loss, drop temperature and the drop perpendicular projection
 T. Orzechowski, S. Wciślik / International Journal of Heat and Mass Transfer 73 (2014) 110–117
on the heating surface, instantaneous values of the heat transfer
coefficient were computed. The cumulative results are shown in
Fig. 9. The results were given for drops, the mass of which ranged
from approx. 1 g to 0.2 g. A lower limit was adopted due to increas-
ing unreliability of measurements, which mainly resulted from the
scales accuracy and camera resolution.
The results presented in Fig. 9 show growing, with the heating
surface temperature, difference between initial (for large drops)
and final (for small drops) values of heat transfer coefficient. At
Tw = 297 °C, e.g., the coefficient equals 0.273 kW/m2 K for mass
m  1 g, whereas at the end for m  0.2 g, the coefficient value is
0.371 kW/m2 K. For Tw = 404.5 °C, the respective values are
0.464 kW/m2 K and 0.837 kW/m2 K (for m  0.1 and Tw = 404.5 °C
a = 1.14 kW/m2 K). That constitutes an increase by approx. 36%
for lower temperature, and as much as by 80% for higher temper-
ature at the same mass loss.
Measurement results were approximated with power function
a = ChAn. The temperature-dependent function coefficient and
exponent were selected by following the optimisation procedure
while imposing the condition of the minimum of the mean square
error. The result is given in the form of a constant value n = 0.33
and a dependence of coefficient Ch on temperature, which is ex-
pressed with formula (5). Fig. 10 shows that values of measured
and approximated heat transfer coefficients are contained in the
range defined by the value of relative error, namely from 13%
to +9%.
The methodology for computing instantaneous values of the
heat transfer coefficient proposed in the study provides a tool that
is precise enough to accurately represent the drop evaporation pro-
cess (Fig. 10).
The area of the drop orthogonal projection onto a heating area,
measured on the basis of thermal images, is approximated by the
third-order polynomial. With these assumptions, the analytical
solution of the energy balance equation is obtained. It describes
the drop mass change over time (Eq. (11)). The accuracy of this ap-
proach was assessed by comparing the calculated and measured
mass loss, as shown in Fig. 12. The relative error between these val-
ues does not exceed 2% for drops of more than 0.2 g. With the mass
loss of the drops, the error is growing rapidly to exceed 10% for the
mass below 0.1 g (see Fig. 13). That confirms the requested range
of applicability of the Eq. (5) describing the relationship between
heat transfer coefficient and the orthogonal projection of a drop
onto the heating surface at the pre-set temperature.
References
[1] M.A. Chan, Ch.R. Yap, K.Ch. Ng, Modeling and testing of an advanced compact
two-phase cooler for electronics cooling, Int. J. Heat Mass Transfer 52 (2009)
3456–3463.
[2] J. Choi, M. Jeong, J. Yoo, M. Seo, A new CPU cooler design based on an active
cooling heat sink combined with heat pipes, Appl. Therm. Eng. 44 (2012) 50–
56.
[3] I. Khamis, K.C. Kavvadias, Trends and challenges toward efficient water
management in nuclear power plants, Nucl. Eng. Des. 248 (2012) 48–54.
[4] F.B. Cheung, S.M. Bajorek, Dynamics of droplet breakup through a grid spacer
in a rod bundle, Nucl. Eng. Des. 241 (2011) 236–244.
117
[5] H.K. Cho, K.Y. Choi, S. Cho, C.-H. Song, Experimental observation of the droplet
size change across a wet grid spacer in a 6  6 rod bundle, Nucl. Eng. Des. 241
(2011) 4649–4656.
[6] R.-H. Chen, S.-L. Chiu, T.-H. Lin, On the collision behaviors of a diesel drop
impinging on a hot surface, Exp. Therm. Fluid Sci. 32 (2007) 587–595.
[7] Y.M. Arifin, T. Furuhata, M. Saito, M. Arai, Diesel and bio-diesel fuel deposits on
a hot surface, Fuel 87 (2008) 1601–1609.
[8] Y. Wang, Y. Li, S. Weng, M. Su, Experimental investigation on humidifying
performance of counter flow spray saturator for humid air turbine cycle,
Energy Convers. Manage. 48 (2007) 756–763.
[9] A.M. Al-Ibrahim, A. Varnham, A review of inlet air-cooling technologies for
enhancing the performance of combustion turbines in Saudi Arabia, Appl.
Therm. Eng. 30 (2010) 1879–1888.
[10] R. Kouhikamali, H. Hesami, A. Ghavamian, Convective heat transfer in a
mixture of cooling water and superheated steam, Int. J. Therm. Sci. 60 (2012)
205–211.
[11] H. Nomuraa, M. Takahashi, Y. Ujiie, H. Hara, Observation of droplet motion
during flame spread on three-fuel-droplet array with a pendulum suspender,
Proc. Combust. Inst. 30 (2005) 1991–1999.
[12] C. Balocco, P. Lio, Assessing ventilation system performance in isolation rooms,
Energy Build. 43 (2011) 246–252.
[13] C. Maqua, G. Castanet, F. Grisch, F. Lemoine, T. Kristyadi, S.S. Sazhin,
Monodisperse droplet heating and evaporation: experimental study and
modeling, Int. J. Heat Mass Transfer 51 (2008) 3932–3945.
[14] M. Lemouari, M. Boumaza, A. Kaabi, Experimental analysis of heat and mass
transfer phenomena in a direct contact evaporative cooling tower, Energy
Convers. Manage. 50 (2009) 1610–1617.
[15] S.P. Fisenko, A.A. Brin, A.I. Petruchik, Evaporative cooling of water in a
mechanical draft cooling tower, Int. J. Heat Mass Transfer 47 (2004) 165–177.
[16] H.-Z. Yu, Froude-modeling-based general scaling relationships for fire
suppression by water sprays, Fire Saf. J. 47 (2012) 1–7.
[17] S. Cioulachtjian, S. Launay, S. Boddaert, M. Lallemand, Experimental
investigation of water drop evaporation under moist air or saturated vapour
conditions, Int. J. Therm. Sci. 49 (2010) 859–866.
[18] P. Nemec, A. Čaja, M. Malcho, Mathematical model for heat transfer limitations
of heat pipe, Math. Comput. Modell. 57 (2013) 126–136.
[19] G. Canti, G.P. Celata, M. Cumo, M. Furrer, L. Saraceno, Effect of droplets
generation in the boiling limit conditions: heat pipe visualization, Int. J. Therm.
Sci. 52 (2012) 1–7.
[20] N. Nagai, S. Nishio, Leidenfrost temperature on an extremely smooth surface,
Exp. Therm. Fluid Sci. 12 (1996) 373–379.
[21] J.D. Bernardyn, I. Mudawar, The Leidenfrost point: experimental study and
assessment of existing models, J. Heat Transfer Trans. ASME 121 (1999) 894–
903.
[22] D.H. Shin, S.H. Lee, J.-Y. Jung, J.Y. Yoo, Evaporating characteristics of sessile
droplet on hydrophobic and hydrophilic surfaces, Microelectron. Eng. 86
(2009) 1350–1353.
[23] C.Y. Lee, B.J. Zhang, J. Park, K.J. Kim, Water droplet evaporation on Cu-based
hydrophobic surfaces with nano and micro-structures, Int. J. Heat Mass
Transfer 55 (2012) 2151–2159.
[24] L. Bin, R. Bennacer, A. Bouvet, Evaporation of methanol droplet on the Teflon
surface under different air velocities, Appl. Therm. Eng. 31 (2011) 3792–3798.
[25] H. Oprins, J. Danneels, B. Van Ham, B. Vandevelde, M. Baelmans, Convection
heat transfer in electrostatic actuated liquid droplets for electronics cooling,
Microelectron. J. 39 (2008) 966–974.
[26] G. Lu, Y.-Y. Duan, X.-D. Wang, D.-J. Lee, Internal flow in evaporating droplet on
heated solid surface, Int. J. Heat Mass Transfer 54 (2011) 4437–4447.
[27] M. Birouk, M.M. Abou Al-Sood, Droplet evaporation in a turbulent high-
pressure freestream – a numerical study, Int. J. Therm. Sci. 49 (2010) 264–271.
[28] D. Chatzikyriakou, S.P. Walker, G.F. Hewitt, C. Narayanan, D. Lakehal,
Comparison of measured and modelled droplet–hot wall interactions, Appl.
Therm. Eng. 29 (2009) 1398–1405.
[29] E.F. Crafton, W.Z. Black, Heat transfer and evaporation rates of small liquid
droplets on heated horizontal surfaces, Int. J. Heat Mass Transfer 47 (2004)
1187–1200.
[30] B. Pathak, S. Basu, R. Kumar, Heat and mass transfer and chemical
transformation in a cerium nitrate droplet, Int. J. Heat Mass Transfer 63
(2013) 301–312.
[31] J. Bodzenta, A. Kaźmierczak, T. Kruczek, Analysis of thermograms based on FFT
algorithm, J. Phys. (IV) 129 (2005) 201–206.
[32] T. Orzechowski, S. Wciślik, Analysis of the droplet film boiling heat transfer
under ambient pressure, Energy Convers. Manage. 76 (2013) 918–924.