Experimental Thermal and Fluid Science 110 (2020) 109916

Contents lists available at ScienceDirect

Experimental Thermal and Fluid Science

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

An experimental investigation into the spread and heat transfer dynamics of T

a train of two concentric impinging droplets over a heated surface
⁎
Ganesh Guggillaa,b, Ramesh Narayanaswamyb, Arvind Pattamattaa,
a
Indian Institute of Technology Madras, India
b
Curtin University, Australia

A R T I C LE I N FO A B S T R A C T

Keywords: Extensive studies of two concentric droplets consecutively impinging over a thin heated foil surface are carried
Spread dynamics out to compare the spread and heat transfer dynamics of a single drop, and drop-on-drop configurations using
Droplet heat transfer high speed imaging and infrared thermography. Millimeter-sized deionized water droplets (2.80 ± 0.04 mm)
Concentric droplet impact are impinged upon a heated Inconel surface (thickness of 25 μ m) from a fixed height corresponding to a Weber
Drop-on-drop
number (We) of 50 ± 2 and Reynolds number (Re) of 3180 ± 90 with a flow rate of 20 droplets per minute.
Surface temperature is chosen as a parameter, and is varied from 22 °C (non-heated) to 175 °C. Temperature and
heat flux distributions associated with droplet-surface interactions are obtained, and the outcomes of the process
are measured in terms of spread diameter, droplet input heat transfer, dynamic contact angle, and surface mean
temperature. A decline in the droplet heat transfer for drop-on-drop impingement is observed for all tempera-
tures investigated in the present work. This is attributed to the surface pre-cooling by the initial droplet and also
to the reduced surface area-to-volume ratio i.e., increased spreading film thickness. High heat transfer rates are
observed around the three-phase contact line region, especially during the receding phase of the droplet, for both
configurations, confirming the significance of contact line evaporation in droplet-hot wall interactions.
Theoretical models predicting the maximum spread factor and corresponding input heat transfer into the droplet
are identified from the literature, and found to be in good agreement with present experimental observations.

1. Introduction influence of thermophysical properties, and the observed regimes of

evaporation.
Various natural and industrial processes requires the knowledge of In the case of droplet impingement over hot surfaces, the process
droplet interactions with surfaces. The underlying physics of these involves mass, momentum and heat transfer interactions, and thereby
droplet systems is complex, and has triggered many experimental and requires additional efforts for better understanding of the phenomenon.
numerical investigations in the past decades. Applications such as ink A comprehensive review of studies concerning the fluid mechanics and
jet printing, spray coating, and tablet encapsulation requires the study heat transfer mechanisms of liquid drop impact on a heated wall is
of droplet dynamics over adiabatic (non-heated) surfaces. Droplet in- presented by Liang and Mudawar [7]. Significant contributions were
teraction with heated walls is the topic of interest in processes such as made in the literature in understanding the interfacial behaviour of
metal quenching, spray cooling, fuel-air interaction in internal com- droplet from the moment of impact over heated surfaces. It has been
bustion engines, power plant engineering and refrigeration. observed that heat transfer in droplet impingement over a hot surface is
From earlier studies conducted on droplet impingement over adia- strongly dependent on the magnitude of wall temperature relative to
batic surfaces, the droplet impact scenario can be classified into three the liquid’s saturation temperature. Factors such as droplet diameter,
types based on the nature of target, i.e., solid wall, liquid film and deep impact velocity, physical properties of the liquid, nature of the sur-
liquid pool. Extensive reviews on these subtopics have been provided rounding gas, and wall characteristics can also influence the overall
by Prosperetti and Oguz [1], Rein [2], Yarin [3], Marengo et al. [4], process. Four distinct regimes were identified based on the evaporation
Moreira et al. [5], and Josserand and Thoroddsen [6]. They have lifetime of a single drop at different wall temperatures as film eva-
summarized several aspects associated with the hydrodynamics of the poration, nucleate boiling, transition boiling and film boiling [8,9].
impingement process i.e., nature of impact, surface wettability, Efforts were made to quantify the impact dynamics and heat transfer

⁎
Corresponding author.
E-mail address: [email protected] (A. Pattamatta).

https://doi.org/10.1016/j.expthermflusci.2019.109916
Received 29 March 2019; Received in revised form 3 September 2019; Accepted 3 September 2019
Available online 11 September 2019
0894-1777/ © 2019 Elsevier Inc. All rights reserved.
G. Guggilla, et al. Experimental Thermal and Fluid Science 110 (2020) 109916

Nomenclature dt time interval, ms

T∞ ambient temperature, °C
Ae effective area, m2 Tsat saturation temperature, °C
Ae∗ dimensionless effective area U impact velocity of droplet, m/s
c specific heat capacity of the heater, J/kg K V volume of the droplet, m3
d instantaneous spread diameter, mm
DPM droplets per minute Non-dimensional quantities
D impacting droplet diameter, mm
Ds sessile droplet diameter, mm Bo Bond number, ρl gD 2 /4σ
g acceleration due to gravity, m/s2 Ja Jakob number, c pl ΔT / hlv
hs height of the sessile droplet, mm Pr Prandtl number, μl cpl/ kl
hmax height of the droplet at maximum spread, mm Re Reynolds number, ρl UD / μl
hlv latent heat of vaporization, J/kg We Weber number, ρl DU 2/ σ
m mass of liquid droplet, kg
NETD Noise Equivalent Temperature Difference Greek letters
Qcond net conduction heat transfer, W
Qconv convective heat transfer, W ∊ effectiveness ratio
Qdrop droplet input heat transfer, W θ three-phase contact angle, degree
Qgen generated heat, W σsd standard deviation
Qrad radiation heat transfer, W τ non-dimensional time, tU / D
Qstored stored heat, W
Q∗ effectiveness or cooling efficiency Subscripts
Qe∗ dimensionless evaporation heat transfer (mass)
S∗ spread factor, (d/ D) f final
∗
Smax maximum spread factor, (dmax / D) i initial
T temperature of the surface, °C l liquid
T∗ dimensionless temperature max maximum
t time, ms

behaviour in those regimes in order to characterize the droplet-hot wall foot print from the underside of the impact surface. It was reported that
interactions. Bernardin et al. [10,11] revealed that wall temperature the onset of the nucleate boiling depends on contact temperature, and
and impact Weber number are the two most influencing parameters observed the regime when contact temperature exceeds the liquid’s
governing the impingement process over heated surfaces. Impingement saturation temperature. Studies on the effect of the dissolved gases and
studies were carried out for low and high Weber numbers and extensive salts [24], surface thickness [25], surfactants [32], nano fluids [33],
maps concerning the impact and heat transfer were provided. They droplet size and physical properties [34] on the incipience of bubble
have also studied the effect of surface roughness and found that surface nucleation are available. Predictions of critical heat flux temperatures
features can influence the observed boiling regimes. Using advanced [10,11,35] were also reported in the literature for different liquids in-
diagnostic tools such as high speed imaging [12–14], interferometry cluding water, and correlations provided [36,37] for corresponding
and total internal reflection techniques [15,16], attempts were made to maximum heat transfer rate.
quantify the droplet boiling regimes based on the observations of hy- For liquid-solid interface temperatures at or above certain tem-
drodynamic behaviour during impact. perature, named as the Leidenfrost temperature, the liquid in the im-
Film evaporation takes place when the wall temperature is below mediate vicinity of the wall is instantaneously converted to vapour
the liquid’s saturation temperature, and even when the wall is super- upon contact, and forms a continuous insulating vapour layer between
heated but insufficient to initiate bubble nucleation inside the drop the liquid and the wall [38,39]. In literature, this temperature is
upon contact with the surface [7]. It is observed that, in film eva- identified as the lowest wall temperature of the film boiling regime and
poration regime, droplet heat transfer is affected by temperature var- has been studied in relation to sessile drop over hot surface termed as
iations inside the droplet, wall heat flux and droplet evaporation rate static Leidenfrost temperature [12,33,40]. While, for impinging dro-
[17–21]. Chandra et al. [22] investigated the effect of contact angle on plets, this temperature is termed as dynamic Leidenfrost temperature
droplet evaporation rate by experimental investigation. They have used where rebound of the droplet from the surface can be observed
a surfactant to reduce the contact angle resulting in higher evaporation [33,39,41]. Influence of pressure, wall roughness, gravity and surface
rates. Pasandideh-Fard et al. [23] presented a numerical model and tension on static Leidenfrost temperature [40,42–44], and correlations
carried out simulations revealing that impact velocity has a minor in- [45,46] concerning the precise prediction of dynamic Leidenfrost
fluence on the overall droplet heat transfer. From these works, it is temperatures in terms of saturation temperature, static Leidenfrost
identified that wall temperature is lowest at the impact point and in- temperature and impact Weber number are also available. In a recent
crease in the radial direction toward the edge of the droplet. In- work [47,48], it is shown that microscale droplets with low impacting
vestigators [24,25] also found that the evaporation rate is highest at the velocities can find themselves in a Leidenfrost-type regime (levitating
three-phase contact line and several numerical predictions over the substrate) at substrate temperatures not only far below the
[17,20,26–30] have confirmed these observations. Leidenfrost temperature but even below the saturation temperature. In
Nucleate boiling regime is the region extended from the point of addition, using levitating microdroplets as tracers it is shown that
bubble nucleation, which will take place when wall temperature is evaporation rate has a maximum at the three-phase contact line, con-
above the saturation temperature, to the critical heat flux point which firming the results of other studies [24,25].
corresponds to shortest droplet evaporation time. Tarrozzi et al. [31] The above described studies are related to a single droplet im-
demonstrated a non-intrusive optical method to measure liquid-solid pingement over a hot target surface. Consequently, in order to under-
contact temperature where an infrared camera was used to capture the stand the cooling mechanisms such as spray cooling, basic processes

2
G. Guggilla, et al. Experimental Thermal and Fluid Science 110 (2020) 109916

such as drop-on-drop impact and multiple droplet interference have to impact is considered as two separate configurations i.e., single droplet
be studied. Bernardin and Mudawar [49] presented an empirical ap- and drop-on-drop impact. The temporal variation of droplet deforma-
proach to determine film boiling heat transfer of a spray from extra- tion in terms of spread diameter, dynamic contact angle and heat
polation of the heat transfer characteristics of an isolated droplet transfer rate are used and compared for these configurations.
stream. They found that interference resulting from a drop impinging
on top of another spreading drop or with an offset between the droplets,
minimizes effective liquid-solid contact area and corresponding heat 2. Experimental methodology
transfer rate, in contrast to isolated drops. Fujimoto et al. [50] studied
the successive impact of drops over heated surfaces and presented the Experimental apparatus consists of image acquisition system, dro-
discussion of hydrodynamics for both normal and oblique impacts on plet generating unit and heater surface arrangement powered by a high
the walls. Breitenbach et al. [51] developed a model for heat transfer capacity DC supply. The schematic of experimental set up is shown in
rate into a single drop impacting onto a hot solid wall and then used it Fig. 1. A microfluidic pressure pump (Dolomite, Mitos P-pump) con-
to estimate the average heat transfer coefficient for spray cooling in the nected to an external air compressor, is used to generate the desired
film boiling regime. Minamikawa et al. [52] numerically studied suc- rate of droplets at the needle tip and are made to fall under gravity to
cessive impact of two drops on a heated wall and found that the mor- achieve the required impact conditions. A trial set of 30 droplets is
phology in film boiling regime is strongly dependent on vertical spacing considered for diameter calculation and the generated droplet size is
between the drops. Guggilla et al. [53] used a phase-change numerical found to be 2.80 ± 0.04 mm.
model and studied the drop-on-drop impact over heated surfaces in film Image acquisition system consists of a high-speed camera (Photron
evaporation regime. The effect of non-dimensional numbers on eva- fastcam SA3 120K) running at 10000 FPS (frames per second) with a
poration dynamics of drop-on-drop collision and theoretical model to spatial resolution of 20 μ m/pixel. Shadow photography technique is
evaluate the numerical findings was developed. Batzdorf et al. [54] adopted for imaging the droplets using a LED light source with a dif-
developed a numerical model and simulated simultaneous collisions of fusion screen. Factory calibrated high-performance infrared camera
two drops with a solid substrate. (FLIR X6540sc) is used to capture the thermal foot print (temperature
From the previous studies, it can be observed that the impact dy- distribution) of the droplet on the surface. With a frame rate of 1000
namics and heat transfer mechanism involved in multiple droplet col- FPS and a spatial resolution of 136 μ m/pixel, the infrared camera is
lisions are not fully known. There is a need to assess various config- triggered simultaneously along with high-speed camera. The post-pro-
urations of these droplet collisions and its interference over heated cessing of images is carried out using Matlab Image processing tool box
surfaces for different boiling regimes. Comparison with an isolated and an open source java based image processing program, ImageJ [55].
droplet impact and theoretical models estimating the dynamics of the An annealed Inconel 600 alloy foil of thickness 25 μ m is used as the
process will provide more insights in understanding the physical pro- target surface, sandwiched between copper bus bars on either side, and
cess of spray cooling. The present work is aimed at studying the spread fixed to a wooden base. The surface is polished, and the surface
and heat transfer dynamics of a consecutive impingement train of two roughness measurement, Ra , using stylus probe profilometer is within
water droplets. High-speed photography and infrared thermographic the range of 0.15–0.30 μ m. DC power supply (BK Precision 1900, 1-16
techniques are employed to capture the post impingement events as- VDC, 60 A) is provided through the copper bus bars to maintain the
sociated with the process. surface at different temperatures using power supply controls. To im-
A thin Inconel 600 foil has been used as the target surface and prove the response of the infrared camera imaging of the surface, a thin
temperature is chosen as the parameter, varied from ambient tem- layer of high heat-resistant black paint is applied underneath the sur-
perature of 22 °C to 175 °C, and found to be within the film evaporation face. The emissivity of the paint was measured using an emissometer (D
regime. From the instant of impact, the droplets are found to undertake & S Emissometer, Model AE) and found to be 0.82. The dimensions of the
a series of spreading and receding phases until it achieves an equili- foil surface is about 45 mm × 40 mm × 0.025 mm. Droplet impinge-
brium and evaporates as a spherical liquid cap [7]. In the present work, ment experiments were carried out at an ambient temperature of 22 °C
the impact dynamics of droplet initial stage i.e., spreading and receding and a relative humidity of about 50%.
phases are captured and studied in detail. The event of consecutive

Fig. 1. Schematic showing the experimental apparatus used in the present study.

3
G. Guggilla, et al. Experimental Thermal and Fluid Science 110 (2020) 109916

2.1. Impingement configuration

A train of two water droplets are consecutively impacted on to the

foil surface. The event is captured through the high-speed camera from
the side view while the temperature variation of the surface, upon in-
teraction with the droplet, is acquired from the bottom of the surface
using the thermal camera. The surface is hydrophilic, and contact angle
measurements are made using Holmarc′s contact angle meter. The static
contact angle, quasi-static advancing and receding angles over the non-
heated surface are 72 ± 1°, 83 ± 4° and 13 ± 1° respectively. The
impingement scenarios are presented in Fig. 2 where both the sche-
matic diagram and high speed images are provided.
The time interval between the drops (δtD ) is approximately 3 s i.e.,
the flow rate is about 20 droplets per minute (DPM). With this flow
rate, the leading droplet that impinges the foil surface will become
Fig. 3. Schematic showing the temporal change of spread diameter during the
sessile, before the trailing droplet impacts on the sessile droplet. Thus
impact.
the configuration can be treated as a drop-on-drop impact.
Fig. 3 is a schematic that demonstrates the temporal change of
droplet spread diameter upon impact with the surface. The first droplet,
when impacted, oscillates on the surface for a while, and will remain
sessile upon which the second droplet is impinged resulting in the
spreading and receding phases, as shown in the Fig. 3.

2.2. Image post-processing

Information regarding hydrodynamics such as droplet initial dia-

meter (volume), spread diameter, and dynamic contact angle are
measured using the side-view images of the impingement process. Fig. 4. Steps involved in image post-processing: (a) Grayscale (b) Binary (c)
Standard procedures of image conversion i.e., conversion of grey to Region recognition.
binary image followed by edge and region recognition, are im-
plemented, and data is retrieved using resources available in Matlab
and ImageJ post-processing toolbox. The resulting image after post-
processing is shown in Fig. 4.

2.2.1. Droplet volume (diameter) calculation

High-speed images obtained from experiments are used for the
calculation of droplet volume and diameter. Assuming an axi-sym-
metric droplet, the volume of the droplet is calculated [56] by summing
up the cylindrical slices of unit pixel height as
π 3
Volume , V = Zp ∑ di2 Fig. 5. Droplet volume calculation.
4 (1)
where di , the diameter of each cylindrical strip in the droplet image
various contact angles and the effect of contact line velocity and tem-
given as (x i, max − x i, min ) as shown in Fig. 5, and Zp is the resolution of the
perature on contact angle. Measurement of this dynamically changing
image measured in meter /pixels
angle will enhance the understanding of the key aspects associated with
Then diameter of the droplet can be obtained as
the spread and evaporation dynamics of the present work.
6V 1/3 In the present work, the three-phase contact angle is calculated
Diameter , D = ⎡ ⎤
⎣π ⎦ (2) using the side-view images of droplet impingement. During the impact
process, the observed profiles of the droplet are complex, and standard
methods of curve fitting for obtaining the droplet profile is mathema-
2.2.2. Dynamic contact angle
tically tedious and complicated. Instead, as shown in Fig. 6, at least five
The wetting characteristics of a surface for an impinging liquid can
points on the droplet profile near the contact line region are considered.
be represented using the contact angle in the three-phase contact re-
Upon analysis, a second-order polynomial fits well with the selected
gion. Former studies on droplet-wall interactions [26,57,58] discussed

Fig. 2. Impingement configurations considered in the present work.

4
G. Guggilla, et al. Experimental Thermal and Fluid Science 110 (2020) 109916

Fig. 6. Contact angle measurement.

data points, and the tangent of the polynomial at the three-phase and the overall droplet heat transfer during the impact, expressed as
contact point is calculated as the dynamic contact angle. effectiveness (Q∗) , is not significantly affected by the filtering proce-
dure, as given in Fig. 7(d).

2.3. Infrared image post-processing

2.3.1. Droplet input heat transfer calculation
The infrared camera used in the present study is factory calibrated, The droplet input heat transfer is one of the important parameters
and the uncertainty associated with temperature measurement is ± 1 K. required for understanding the droplet-hot wall interactions and the
It is noticed that the recorded raw images are prone to noise, and is ongoing cooling process. The temperature variation of the surface is
estimated in terms of the noise equivalent temperature difference obtained from the bottom of the surface via infrared images. An energy
(NETD) value of thermography system. For the given temperature range balance is applied at every pixel element of the surface, as shown in
used in the study, the NETD values are within the acceptable range of Fig. 8 to calculate the heat transfer into the droplet.
60–200 mK. However, it is shown that the heat conduction term used in The energy balance applied to the pixel element results in
the heat transfer analysis is sensitive to the spatial signal noise of the
Qstored = Qgen + Qcond − Qrad − Qconv − Qdrop (3)
input temperature field [56] and extensive filtering is required to re-
duce the noise. Time and spatial averaging are applied to the tem-
where droplet input heat transfer is represented as Qdrop
perature field, and it is followed by the application of Matlab provided
Thus,
Gaussian filter (σsd = 2 ). The detailed description of the method can be
found in reference [56]. Fig. 7(a) and (b) shows the raw and filtered Qdrop = Qgen + Qcond − Qrad − Qconv − Qstored (4)
heat flux image calculated during single droplet impact over a surface
temperature of 154 °C, and at a time instant, t = 15 ms. From 7(c), it is and droplet input heat flux qdrop is obtained, using the length of the pixel
visible that the non-physical noise in heat flux distribution is reduced, element Lp , as

Fig. 7. (a) Raw image (T = 154 °C and t = 15 ms) (b) Filtered image (c) Droplet input heat flux distribution along the centreline X-X (d) Effectiveness.

5
G. Guggilla, et al. Experimental Thermal and Fluid Science 110 (2020) 109916

Fig. 8. Heat transfer calculation: energy balance at a pixel element.

Qdrop 120 °C. Fig. 13 shows the temporal variation of spread factor during the
qdrop =
Lp2 (5) impingement. To validate the accurate variation of spread dynamics
over a heated wall, the surface temperature and impact conditions
It is noted that Qstored represents the change in energy of the surface due should be exactly maintained. However, inspite of the differences in the
to cooling, and Qgen being heat generated due to DC supply. While target surface (Inconel versus Stainless steel), the results shown in
Qcond , Qrad and Qconv are net conduction heat transfer along the surface, Fig. 13 show similar trends confirming the validity of the present ex-
radiation and convection heat transfers underneath the surface re- perimental set up for droplet impingement studies over heated target
spectively. Further details for obtaining each term in the energy balance surfaces.
is provided in the appendix.
Using the above energy balance, the contributions of heat transfer
quantities towards the droplet input heat transfer is compared. Two 3. Present experimental investigation
instants, one each in the spreading and receding phase, are selected and
the percentage of heat transfer quantities is calculated against the During the present investigation, a train of two water droplets of
magnitude of droplet input heat transfer at the impact point (pixel). diameter 2.8 mm is impacted, with a velocity of 1.138 m/s, onto a thin
Fig. 9 shows the selected points which are marked over the temporal Inconel surface maintained at a constant temperature. The thermo-
change of spread factor for the droplet impingement over the surface at physical properties of the deionized water and Inconel surface are listed
a temperature of 154 °C. Figs. 10 and 11 present the comparison of in Table 1 and Table 2 respectively. The impact conditions corresponds
these quantities during the advancing and receding phases as a per- to a Weber number of 50 and Reynolds number of 3180 with a constant
centage of the droplet heat transfer. It is evident that Qgen and Qcond are flow rate of 20 droplets per minute (DPM). The surface temperature is
significant quantities compared to Qrad and Qconv in contributing to the the parameter and varies from 22 °C (non-heated) to 175 °C. At every
droplet input heat transfer. temperature, the images of single drop and drop-on-drop impacts are
Even though the heat loss by convection and radiation seems neg- recorded separately and analysis is carried out. Here the focus is to
ligible in receding phase, it is important for the accurate estimation of analyze the spread and heat transfer characteristics at the instant of
droplet heat transfer in the spreading phase. Thus, in the present work, impact where effective cooling of the surface will take place. The time
all the above described heat transfer quantities will be included for scale of impingement is of order; time t = 45 ms corresponds to a non-
droplet heat transfer calculations. dimensional time, τ = 18 for each configuration. The spread dynamics
is photographed using a high-speed camera, and the temperature re-
sponse during the impact is recorded from the underside of the surface
2.4. Experimental methodology: validation cases
using infrared thermography.
At each chosen temperature, three sets of data is recorded (n = 3),
The present experimental methodology is validated using previously
and average values are used to represent the data. The experimental
published studies available in the literature. Two cases: drop-on-drop
uncertainties associated with different parameters are presented in the
impingement over a non-heated surface, and a single droplet impact
Table 3. Here ΔX and Δx are used to represent the absolute and relative
over a heated surface are carried out. The spreading parameter i.e.,
spread factor is calculated and compared with experimental results.

2.4.1. Drop-on-drop impact over a non-heated surface

For the present study, the generation of multiple droplets to achieve
the drop-on-drop configuration is crucial. Wakefield et al. [59] carried
out drop-on-drop impingement studies over a non-heated Teflon surface
with the Weber number as a parameter. A case with Weber number of 2
is considered for the validation, and the results are compared in terms
of the spread factor. Fig. 12 shows the results from the present ex-
periments compared with Wakefield et al. [59]. The variation of spread
factor with time was found to be in agreement within ± 10%, thus
validating the experimental methodology followed in the present work.

2.4.2. Single droplet impact over a heated surface

Pasandideh-Fard et al. [23] studied the cooling effectiveness of a
single droplet over a heated surface. A single water droplet is impacted
over a stainless steel surface maintained at a constant temperature of
120 °C with an impact Weber number of 47. In the present set up, a thin
Inconel surface is used instead of stainless steel, and maintained at Fig. 9. Single droplet impact over the target surface (T = 154 °C).

6
G. Guggilla, et al. Experimental Thermal and Fluid Science 110 (2020) 109916

Table 1
Thermo-physical properties of the deionized water used in the present
study, at 1 atm and ambient temperature of 22 °C.
Properties Value

Saturation temperature, Tsat , ° C 100

Density, ρl , kg/m3 998
Dynamic viscosity, μ , Ns/m2 0.001
Surface tension, σ , N/m 0.0725
Specific heat capacity, cp , kJ/kg K 4.18
Latent heat of vaporization, hlv , kJ/kg 2260

Fig. 10. Comparison in advancing phase.

Table 2
Thermo-physical properties of the Inconel 600 alloy used in the present study.
Properties Value

Density, ρ , kg/m3 8470

Thermal conductivity, ks , W/mK 14.8
Electrical resistivity, ρs , Ohm-m 103·10−8
Specific heat capacity, c , kJ/kg K 444
Temperature coefficient of resistance, αs , K−1 12·10−5

uncertainties respectively.

Fig. 11. Comparison in receding phase.

4. Results and discussion

When the droplet comes in contact with a hot surface, heat transfer
takes place which results in the cooling of the surface. The temperature
of the droplet increases with time; evaporation ensuing across the li-
quid-gas interface affects the droplet spread diameter. Thus, it is im-
portant to study the effect of surface temperature on both the spread
and the heat transfer characteristics. Figs. 14 and 15 show the spread
behaviour of single and drop-on-drop configurations, respectively over
the surface with a pre-impact surface temperature of 154 °C. The pre-
sent arrangement of hot surface, using Joule heating, resulted in
slightly non-uniform pre-impact surface temperature. Here, the spatial
mean temperature (maximum deviation of ± 3 °C is observed at
T = 154 °C) is represented as the surface temperature. Also, to realise
the temperature contours during drop-on-drop impact, the change in
temperature (ΔT ) for each pixel, is calculated as the difference of the
initial temperature to the instantaneous temperature. The temperature
contours, the corresponding change in temperature (Δ T), and droplet
Fig. 12. Drop-on-drop impact over a non-heated surface. input heat flux (qdrop ) are also presented. A considerable amount of heat
transfer, termed as effective heat transfer, is observed to occur during
the initial stage of droplet interaction with the surface in both the
configurations. From Figs. 14(d) and 15(d), it can be noted that the
significant heat transfer is during the initial spreading phase whereas
the peak value is detected at the maximum spread of the droplet.
However the heat transfer associated with single droplet impact is
prominent compared to that of drop-on-drop configuration. This is
because of the low pre-impacting surface temperatures for drop-on-
drop scenario as given in Fig. 15(b), due to the presence of the initial
droplet on the target surface, thereby resulting in lower heat transfer
rates. The subsequent sections of this paper describe the spread hy-
drodynamics in terms of the spread factor and surface wetting i.e.,
contact angle. Detailed description of heat transfer characteristics are
also provided.

4.1. Spread hydrodynamics

Upon impact, the leading droplet performs a series of advancing and

Fig. 13. Single droplet impact over a heated surface (T = 120 °C).
receding phases by dissipating the impact energy and attains a sessile
droplet state. Consecutively, the second droplet, which impinges on to
the sessile droplet, will coalesce for specific instant followed by the
spreading and receding phases. Thus, for a single droplet impact, the

7
G. Guggilla, et al. Experimental Thermal and Fluid Science 110 (2020) 109916

Table 3
The experimental uncertainties associated with different parameters used in the study. Here Δ X represents the absolute uncertainty where as Δ x stands for the
relative uncertainty.
Parameter Uncertainty

Temperature Δ X = ±1 K
Generated volumetric heat flux [56] qgen
″ = Qgen/ Vs ″ = 20·106 W/m3 at T = 50 ° C)
Δxmax = 11% (qgen
Weber number, We Δ X = ± 2 (We = 50 )
Reynolds number, Re Δ X = ± 90 (Re = 3180 )
Droplet diameter, D Δ X = ± 0.04 mm (D = 2.8 mm )
Droplet impact velocity, U Δ X = ± 0.0171 m/s (U = 1.138 m/s )
Dynamic contact angle, θ Δxmax = 36% (θ = 62° at T = 175 ° C, Single droplet impact) Δxmin = 1.2% (θ = 81° at T = 175 ° C, Drop-on-drop impact)
Spread factor, S ∗ Δxmax = 12% (S ∗ = 0.43° at T = 175 ° C, Single droplet impact) Δxmin = 2% (S ∗ = 2.63 at T = 175 ° C, Single droplet impact)

initial cycle consists of two phases; advancing and receding. Whereas, to the impacting droplet diameter (D), is more for the single droplet
in a drop-on-drop impingement, three stages, namely; coalescing, ad- case. The net spread factor has reduced during the drop-on-drop im-
vancing and receding, are identified during the initial cycle. A non- pingement due to the high energy dissipation resulted from the droplet
dimensional quantity called spread factor, S ∗, is defined as the ratio of coalescence.
spread diameter at an instant (d ) to the pre-impact droplet diameter Observations revealed that the dynamics of spread is coupled with
(D) . The temporal variation of spread factor during single droplet and droplet heat transfer. Especially, the maximum spread factor will dic-
drop-on-drop impact, with identified phases at various surface tem- tate the extent of heat transfer over the surface. So, in order to analyse
peratures, is plotted, as shown in Fig. 16. For both the configurations, heat transfer rate, the maximum spread factor for the initial and second
the temperature effect on the spread is evident from the first cycle of cycle of the post-impingement is considered. It is noted that, in the
spreading. Also, there is a notable reduction in spread factor with present context, a cycle refers to a sequence of spreading and receding
temperature in subsequent cycles for both the single droplet as well as phases. Fig. 18(a) and (b) shows the comparison of maximum spread
drop-on-drop impingement configuration. The comparison of spread factor during the first and second cycles which convey that the initial
factor during single droplet and drop-on-drop impact at a surface cycle’s maximum spread factor has a weak dependence on the surface
temperature of 154 °C is obtained to understand the effect of config- temperature, whereas it decreases with temperature during the second
uration on hydrodynamics, as shown in Fig. 17. Due to the interference cycle and the effect is significant for both configurations during the
of droplets during the impact, the cycle of spreading and receding is second cycle.
delayed, for drop-on-drop impingement, which resulted in longer initial Likewise, another important parameter related to hydrodynamics is
cycle time. The cycle time of single droplet impact is about t = 18 ms the contact angle and its variation during both impingement config-
(τ ∼ 7.5) and drop-on-drop impingement is about t = 24 ms (τ ∼ 9.5) urations. The three-phase contact angle is known to vary with velocity
where coalescing phase is about t = 1 ms (τ ∼ 0.5). The presence of two [57] and increase with the surface temperature [26,58]. It will affect
droplets resulted in a higher spread factor for the drop-on-drop con- the spread of the droplet, and therefore, the heat transfer rate. The
figuration. However, the net spread factor (δS ∗) at a given instant of variation of the dynamic contact angle with time is obtained for the
time, which is defined as the ratio of change in spread diameter (d − Ds ) present configurations to ascertain the effect of temperature, as shown

Fig. 14. Single droplet impingement over the foil surface (T = 154 °C): (a) Side-view image (b) Foil surface temperature after impingement (c) Change in tem-
perature (d) Heat flux distribution.

8
G. Guggilla, et al. Experimental Thermal and Fluid Science 110 (2020) 109916

Fig. 15. Drop-on-drop impingement over the foil surface (T = 154 °C): (a) Side-view image (b) Foil surface temperature after impingement (c) Change in temperature
(d) Heat flux distribution.

Fig. 16. Spread factor versus time.

in Fig. 19. As soon as the droplet impacts the surface, a high contact uncertainty of the presented data. Additionally, sessile droplet contact
angle is observed as it is under the influence of impacting velocity, and angle (Static contact angle) variation with surface temperature is in-
the value declines during the receding phase [57]. In the present study, spected and given in Fig. 20, and for the temperatures used in the
the captured contact angle variation exhibits a similar behaviour during present work, there is only a minor increase in contact angle with
both impingement configurations, as shown in Fig. 19(a) and (b). surface temperature. Previous studies [26,58] reported a strong effect
For a single droplet impact, the contact angle is increased till it of temperature on contact angle which is not so evident in the present
reaches the maximum spread (advancing phase end) and decreases to a work. The difference in volatility of the liquid, and surface conditions
minimum angle at the end of receding phase which is given in are attributed to this behaviour.
Fig. 19(a). Meanwhile, for drop-on-drop impingement, as presented in
19(b), the trend is similar to single droplet impact, additionally exhibits
a constant angle during the coalescing stage. In the present study, the 4.2. Heat transfer characteristics
effect of temperature on dynamic contact angle is found to be weak.
During the single droplet impingement, a slight increase in dynamic Furthermore, to understand the heat transfer into the droplet, an
contact angle is observed for the heated case (T = 175 °C) compared to average quantity of heat transfer is calculated over an effective area in
non-heated case (T = 22 °C) in subsequent stages of spreading as shown which a significant amount of heat transfer takes place. The effective
in Fig. 19(a). However, the increase is marginal and within the area is identified using Canny edge detection technique, implemented
in Matlab image post-processing toolbox, applied to a heat flux image

9
G. Guggilla, et al. Experimental Thermal and Fluid Science 110 (2020) 109916

Fig. 17. Comparison of single and drop-on-drop impact over the surface (T = 154 °C).

t
[60] as shown in Fig. 21. A dimensionless effective area Ae∗ is used to ∫0 (Qdrop) dt
compare the present impingement configurations. This is calculated as Q∗ =
m (cp (Tsat − T∞) + hlv ) (8)
the ratio of the surface area with effective heat transfer to the cross-
sectional area of the impacting droplet. Fig. 23 shows the variation of effectiveness (Q∗) with time for both
single and drop-on-drop impingement at different temperatures.
4Ae
Ae∗ = A comparison of both the configurations at a given surface tem-
πD 2 (6)
perature, as shown in Fig. 24, reveals that the droplet input heat
where Ae is the area where effective heat transfer is observed. In the transfer is lower for the drop-on-drop impact compared to single dro-
present work, dimensionless effective area Ae∗ provides a quantitative plet case. This is because of the reduction in surface mean temperature
measurement of area being cooled during the impingement and it can as a result of initial droplet (sessile) interaction with the surface. Also,
∗
also be observed that the maximum spread factor Smax during the im- the previous work using numerical modelling [53] revealed that there is
pact can be approximated from the effective area as rapid decline in heat transfer rate due to the increased film thickness
during the drop-on-drop impingement.
∗ In order to interpret the surface cooling during the impingement,
Smax ∼ ((Ae∗ )max )0.5 (7)
the surface temperature change with time is determined. The surface
Fig. 22 shows that the droplet heat transfer is enhanced with an temperature change upon impact is plotted by tracking the temperature
increase in the surface temperature and this trend is similar for both of the impact point, and termed as centre temperature as shown in
single and drop-on-drop configurations. A maximum in droplet heat Fig. 25. The impact point is always the lowest temperature over the
transfer rate is realised at the end of the first advancing phase for all surface during the impingement [23]. The change in surface tempera-
surface temperatures and confirms that most of the surface cooling is ture is rapid for the case of single droplet impingement compared to
takes place during the initial cycle of the droplet impact. A di- drop-on-drop impact, and follows a similar trend for all surface tem-
mensionless input heat transfer, termed as effectiveness or cooling ef- perature cases considered in the study. However, it is observed that the
ficiency (Q∗) , is introduced to estimate the overall heat transfer per effective area where considerable heat transfer occur, is improved
droplet. It is defined as the ratio of the time integral of droplet input during the drop-on-drop impingement as shown in Fig. 26.
heat transfer to the total heat required for the droplet evaporation. In addition, a mean surface temperature is required to represent the
overall surface cooling, and is calculated considering the effective area.

Fig. 18. Maximum spread factor with surface temperature: Single drop and drop-on-drop impact.

10
G. Guggilla, et al. Experimental Thermal and Fluid Science 110 (2020) 109916

Fig. 19. Dynamic contact angle versus time.

temperature, T ∗ is defined and given as,

Ti − Tf
T∗ =
Ti − T∞ (9)

where Ti , Tf are initial and final surface temperatures respectively and

T∞ being the ambient temperature, in order to compare the two con-
figurations considered in the present study.
Fig. 29(a) and (b) shows the distribution of dimensionless tem-
perature at the instant of maximum spread during the single droplet
and drop-on-drop impact over the surface with a temperature of 154 °C
respectively. The comparison of the dimensionless temperature along
the identified centreline is given in Fig. 29(c). For the single droplet,
dimesnionless temperature (T ∗) of about 0.3 is observed in the inter-
acted area. Whereas, in the case of drop-on-drop impact configuration
due to the presence of sessile droplet, the cooling effect has reduced,
Fig. 20. Static contact angle versus temperature. with a T∗ value of 0.1 in most of the spreading region. However the
surface cooling has improved (T ∗ ∼ 0.3) in the peripheral of the droplet
spread. Thus, the investigation confirms that there is always a decline in
Fig. 27 represents the mean surface temperature variation with time at
cooling effect by the trailing droplet during drop-on-drop impingement.
different temperatures. As given in Fig. 28, the comparison reveals that
To quantify the heat transfer characteristics of the impingement
the overall cooling is significant for the first (single) droplet impact
configurations considered in the study, an effectiveness ratio (∊) is used
compared to the drop-on-drop impact configuration. Nevertheless, as
which is defined as the ratio of dimensionless heat input during the
shown in Fig. 26, it should be noted that the area being cooled, is im-
drop-on-drop impact to that of a single droplet impact.
proved during the drop-on-drop impact.
The present study investigates two configurations of droplet im- (Q∗)drop − on − drop
pingement: single droplet as well as the drop-on-drop. The pre-im- ∊=
(Q∗)single (10)
pacting surface temperatures are different for drop-on-drop impact,
compared to that of single droplet impact. Therefore, a dimensionless It provides a better understanding of each droplet performance (during

Fig. 21. Effective area recognition to calculate the average surface heat transfer rate.

11
G. Guggilla, et al. Experimental Thermal and Fluid Science 110 (2020) 109916

Fig. 22. Droplet input heat transfer versus time.

consecutive droplet impact) in cooling the surface at different wall superimposed onto the infrared temperature and heat flux images. It is
superheats, and the effect of droplet coalescence on spread and heat observed that the surface temperature increases in the radial direction
transfer characteristics during the impingement. Fig. 30 shows the ef- from the center of the droplet (impact point). For both configurations,
fectiveness ratio for different surface temperatures. The ratio is found to the maximum heat flux value is recorded in the vicinity of three-phase
be nearly constant around a value of 0.62 for all observed temperatures. contact line as shown in Fig. 31 and is found to be significant in re-
It can be inferred that the heat transfer for a trailing droplet is always ceding phase. Low film thickness near the contact line region is at-
lower compared to a leading droplet during the drop-on-drop config- tributed to the observed high heat transfer rates. For the case of drop-
uration. The pre-cooling of the surface caused by the initial droplet, on-drop impingement, Fig. 31(b) also unveils that there is an effective
reduces the surface mean temperature, and thereby decreases the heat heat transfer in the annulus portion i.e., the region of change in spread
removal rate of the trailing droplet. It is worth noted that the magni- and thereby extends the area being cooled. These observations will be
tude of reduction in heat transfer could be influenced by the droplet used in further sections to develop a model for estimating droplet heat
flow rate, which controls the surface mean temperature. Also, the transfer during impingement.
boiling regimes such as nucleate boiling with rigorous bubbles, and film
boiling, can determine the outcomes of drop-on-drop impingement
phenomenon. In the present work, the flow rate was constant at 20 4.4. Analytical modelling
droplets per minute (DPM), and the adopted surface temperatures are
not adequate to initiate the bubbles (of nucleate boiling) in the droplet. 4.4.1. Maximum spread
Further investigations are needed to analyze these parameter effects on Earlier studies [23,58,59] modelled the maximum spread theoreti-
the spread and evaporation dynamics. cally using the energy conservation principle. Two instances during the
droplet impingement are considered i.e., pre-impact state and the in-
stant of maximum spread. The associated kinetic, potential and surface
4.3. Three-phase contact line region: temperature and heat flux distribution energies are taken into consideration to estimate the maximum spread
factor. The theoretical models proposed in the literature are adopted in
Fig. 31 shows the temperature and heat flux distribution of the the present work in order to validate the present experimental ob-
target surface during single droplet and drop-on-drop impact at an in- servations. Batzdorf [61] implemented an analytical model for evalu-
stant. The spread diameter estimated from the high speed image is ating the maximum spread during the single droplet impact over a hot

Fig. 23. Effectiveness versus time.

12
G. Guggilla, et al. Experimental Thermal and Fluid Science 110 (2020) 109916

Fig. 24. Comparison of single and drop-on-drop impact over the target surface (T = 154 °C).

surface. A schematic of the droplet system with the initial and final
states considered are presented in Fig. 32. Using energy balance it is
shown [61] that the maximum spread can be calculated from following
equation.

∗2 9a We ∗4
We + 4Bo + 12 − 3(1 − cos (θmax )) Smax = Smax
2 Re (1 − Qe∗) (11)

where θmax and Qe∗

are contact angle at the instant of maximum spread
and dimensionless evaporated mass, respectively. The dimensionless
evaporated mass (Qe∗) is given as
me
Qe∗ =
msingle (12)

‘me ’ and ‘msingle ’ are the cumulative evaporated mass and pre-impacting
droplet mass.
A similar approach was applied to the drop-on-drop impingement
over a hot surface by Guggilla et al. [53] as shown in Fig. 33. In this Fig. 26. Comparison of effective area for the target surface (T = 154 °C).
case, the maximum spread factor is derived as
16c 3Bo 8c 3
∗5
ASmax ∗3
+ BSmax ∗
+ CSmax +D=0 (13) C = −(We + 4Bo + ∗2
+ 3Sin∗2 (1 − cosθ1) + ∗ + 12)
3Sin Sin (16)
where
18 We 1 D = 8(1 + c 3)(1 − Qe∗) (17)
A= a
4 Re (1 + c 3)(1 − Qe∗) (14)
and
B = 3(1 − cosθ2) (15)

Fig. 25. Target surface center temperature versus time.

13
G. Guggilla, et al. Experimental Thermal and Fluid Science 110 (2020) 109916

Fig. 27. Surface mean temperature versus time.

ρl gD 2 number (We), Reynolds number (Re) and Bond number (Bo). The sur-
Bond number, Bo =
4σlv (18) face temperature effects are also considered in the form of evaporated
mass (Qe∗ ) and obtained contact angles (θ1, θ2 ) at respective tempera-
ρl DU tures. The present impingement scenario corresponds to an impact
Reynolds number, Re =
μ (19) condition with We = 50, Bo = 0.27; and Re = 3180. The theoretical
maximum spread factor at different temperatures are calculated using
ρl DU 2
Weber number, We = Eqs. (11) and (13) for single droplet and drop-on-drop impact respec-
σlv (20) tively. The computed results are compared against experimental values
∗ Dmax as shown in Fig. 34. The implemented theoretical models are found to
Maximum spread factor, Smax = be efficient in capturing the maximum spread values, and agreed well
D (21)
with experimental values within a deviation of 8% at all temperatures.
Ds
Initial spread factor, Sin∗ =
D (22)
4.4.2. Input heat transfer
where θ1 and θ2 are the corresponding contact angles at the initial and It is evident that the bulk of heat transfer takes place during the
final states. Here constant ‘a ’ is taken as 15, in order to approximate the spreading phase, and is accompanied by convection heat transfer,
present experimental observations, and ‘c ’ is the radius ratio of im- which can be modelled using a Nusselt number correlation. Assuming
pacting to sessile droplet (c = 1). In the present work, the liquid used is the spreading droplet as a single impinging jet, Batzdorf et al. [61]
deionized water which is non-volatile and for the surface temperatures developed a theoretical model for estimating the overall heat trans-
used, the total evaporation time of droplet is ranging from 720 s (at ferred during the spreading phase which is proportional to the con-
50 °C) to 100 s (at 175 °C). The time interval between the two con- vective heat transfer, and is reproduced below.
secutive droplets at the considered flow rate of 20 droplets per minute ∗ ∗
(DPM) is around 3 s, and the total evaporated mass during this time is Smax (Smax − 1.1) (1 + 0.005Re 0.55 )0.5 Ja
Q∗ = 3b ∗
τmax
assumed to be negligible for the sessile droplet (equal volume as the (Smax − 0.6) Re 0.5 Pr 0.58 (23)
impacting droplet) in the analytical model given in Eq. (13). The eva-
Here, Q∗ is the effectiveness which is represented as
porated mass during the impingement is calculated from the side view
t
images of the droplet, and is used in Eqs. (11) and (17) to estimate the ∫0 (Qdrop) dt
maximum spread factor. Q∗ =
mhlv (24)
It should be noted that the above correlations are able to capture the
effects of all influential dimensionless parameters such as Weber where all relevant properties are calculated at the film temperature,

Fig. 28. Comparison of single and drop-on-drop impact over the surface with temperature 154 °C.

14
G. Guggilla, et al. Experimental Thermal and Fluid Science 110 (2020) 109916

Fig. 29. Comparison of dimensionless temperature (at T = 154 °C) (a) Single droplet impact (b) Drop-on-drop impact (c) Distribution along the centreline X-X.

Previous studies concerning the droplet impact over the heated

surfaces are considered to validate the proposed correlations and ex-
amine the sensitivity of the constants ‘a ’ and ‘b ’ described in the Eqs.
(11), (13), and (23). Teodari et al. [62] carried out the thermographic
analysis of interfacial heat transfer mechanisms on drop/wall interac-
tions. Single droplets of water and ethanol and a heated stainless steel
surface (25 μm), are utilized. The study examined the effect of the
surface temperature, liquid surface tension, and wettability on heat
transfer processes during a single droplet impact. Jung et al. [60]
conducted heat transfer analysis of droplet collision over superheated
surfaces and detected a dynamic Leidenfrost point based on the droplet
heat transfer. In this work, water droplet impingement is carried out
over the superheated platinum-coated sapphire glass maintained at
temperatures of 176–226 °C. The details of the impingement studies,
used for the present validation, are summarized in Table 4. The max-
imum spread factor, and the corresponding effectiveness, as per Eq.
(24), is calculated using the data from references [60,62] and compared
Fig. 30. Effectiveness versus temperature. with the theoretical values from Eqs. (11) and (23). The constants ‘a ’
and ‘b ’ are chosen such that the theoretical values fit well with the
experimental outcomes.
and the constant ‘b ’ is taken as 0.1 in order to fit the experimental data.
Fig. 36(a) and (b) show the comparison of experimental observa-
The above correlation was used for estimating the heat transfer
tions with theoretical results of maximum spread factor and effective-
during single and drop-on-drop impact over the hot surface. The max-
∗
ness, respectively. In the case of Teodari et al. [62], the experimental
imum spread factor Smax and the corresponding non-dimensional time
conditions (liquid on the heated hydrophilic surface) are similar to the
τmax during the initial cycle which is of order τmax ∼ 2 (t = 5 ms) for
present work. so, the values of the constants a = 15 and b = 0.1 are
single droplet impingement and τmax ∼ 3 (t = 7.5 ms) for drop-on-drop
considered. With these values, the correlations predicted the outcomes
impact, are taken from the experimental observations. Nonetheless, for
for the cases of the water droplet on the stainless steel surface (hy-
drop-on-drop impingement, it is found that the effective heat transfer
drophilic and super-hydrophobic) within the acceptable range. On the
takes place in the annulus region of the initial and post-impact droplet
contrary, significant deviations in the results, are observed during the
spread, as shown in Fig. 31(b). Hence, to obtain an accurate estimation
case of ethanol droplet impact over the heated surface. On the other
of heat transfer, the net spread factor is more relevant and used in the
hand, for the cases of Jung et al. [60], the constants a = 1 and b = 1.4
Eq. (23). Whereas for the single droplet impact, the spread factor and
are found to provide a better approximation for the experimental ob-
the corresponding time values are used. The theoretical results obtained
servations. While the model for the maximum spread factor under-
is found to agree well with the experimental findings as shown in
predicts the results, the effectiveness is observed to be within 25%
Fig. 35. Especially for the drop-on-drop impact, the model is able to
deviation, as given in Fig. 36. The discrepancy with the spread factor
capture the heat transfer rate efficiently using net spread factor. The
prediction is due to the boiling phenomena reported in the droplet. And
maximum deviation in the results are about 20% and can be considered
there is a need to account these effects, which are not included in the
as a good approximation for heat transfer calculations.
present model.

15
G. Guggilla, et al. Experimental Thermal and Fluid Science 110 (2020) 109916

Fig. 31. Post-impact behaviour over the target surface (T = 154 °C; t = 8 ms).

Moreover, in the previous studies of Batzdorf [61], the values scenario is classified as single droplet and drop-on-drop configurations
a = 8/3 and b = 4/3, are adopted for FC-72 droplet collision over over the hot surface and compared for relevant parameters. Outcomes
chromium surface and obtained a good approximation for the cases such as spread factor, droplet input heat transfer, surface temperatures,
studied. Later on, Guggilla et al. [53] extended the study to drop-on- effectiveness or cooling efficiency, and dynamic contact angle are ob-
drop impact of FC-72 and the values of the same order, a = 8/3 and tained and compared. The following conclusions are made from the
b = 3.4 are utilized to estimate the quantities. With these observations, study.
it is determined that the analytical models are efficient in capturing the
spread and heat transfer dynamics for the given constants ‘a ’ and ‘b ’, 1. The effect of temperature on spread dynamics is dominant from the
and these values are sensitive to the nature of the liquid, surface, and initial cycle of spreading for both configurations. However, the
boiling regimes (wall superheat). maximum spread factor trends indicate that the spread factor is
significantly affected by surface temperature during single droplet
5. Summary and conclusions impingement compared to drop-on-drop impact.
2. High heat transfer rates are observed in the vicinity of the three-
The present work provides results of an experimental investigation phase contact line, and input heat transfer rates are strongly influ-
of the spread and heat transfer dynamics of a train of two concentric enced by the surface temperature during single droplet, as well as
impinging droplets over a hot surface. At constant impact conditions drop-on-drop impact over the surface.
(We = 50, Re = 3180), and flow rate of 20 droplets per minute (DPM), 3. Comparison of droplet input heat transfer between the configura-
the behaviour is captured by high-speed imaging and infrared ther- tions confirms that there is a reduction in the trailing droplet heat
mography. Deionized water droplets are impinged over the heated transfer, during drop-on-drop collision, compared to the leading
Inconel surface, and the surface temperature is chosen as a parameter, droplet. The pre-cooling due to sessile droplet(initial) interaction
and varied from 22 °C (non-heated) to 175 °C. The impingement and decrease in surface area-to-volume ratio is attributed to the low

Fig. 32. Single drop impact: Maximum spread.

16
G. Guggilla, et al. Experimental Thermal and Fluid Science 110 (2020) 109916

Fig. 33. Drop-on-drop impact: Maximum Spread.

Fig. 34. Maximum spread factor: Experimental versus theoretical.

Fig. 35. Effectiveness: Experimental versus theoretical.

Table 4
Experimental details of the considered literature cases in the analysis.
Reference Liquid-Surface We Re Surface temperature (° C) a b

Teodari et al. [62] Water on stainless steel (hydrophilic) 22.8 1980 100 15 0.1
Teodari et al. [62] Water on stainless steel (hydrophilic) 22.8 1980 60 15 0.1
Teodari et al. [62] Water on coated stainless steel (superhydrophobic) 22.8 1980 100 15 0.1
Teodari et al. [62] Ethanol on stainless steel (hydrophilic) 50 1221 60 15 0.1
Jung et al. [60] Water on platinum coated sapphire 6.3 1130 176 1 1.4
Jung et al. [60] Water on platinum coated sapphire 6.3 1130 206 1 1.4
Jung et al. [60] Water on platinum coated sapphire 6.3 1130 221 1 1.4
Present experiment Water on Inconel surface (hydrophilic) 50 3180 22–175 15 0.1

17
G. Guggilla, et al. Experimental Thermal and Fluid Science 110 (2020) 109916

Fig. 36. Validation: Experimental versus theoretical.

heat transfer rates observed during the drop-on-drop impact. impingement. This parameter was found to be constant (around
4. The extent of surface area being cooled has increased during drop- 0.62) for all surface temperatures concluding the reduction in heat
on-drop impingement, and the region corresponds to the net spread transfer during drop-on-drop impact.
factor i.e., the annulus portion between the initial and post-impact 7. Relevant analytical models available in literature were identified,
spread is found to provide effective heat transfer during the im- and used to predict the maximum spread factor and heat transfer
pingement. rates during the spreading phase for the present impingement con-
5. The dynamic contact angle variation is provided, for different sur- figurations. The models captured the spread and heat transfer dy-
face temperatures, and the effect of temperature on contact angle is namics with a deviation of 8% and 20%, respectively. The perfor-
weak for both the configurations. Also, there is only a marginal mance of these models are examined with the previous studies for
increase of static contact angle over the heated surface due to the broader validity.
non-volatility of water.
6. To compare the input heat transfer rates among the configurations, Declaration of Competing Interest
an effectiveness ratio is defined as the ratio of dimensionless input
heat transfer during drop-on-drop impact to a single droplet The authors declared that there is no conflict of interest.

Appendix A

A.1. Energy balance applied to a pixel element

The filtered temperature field is used and an energy balance is applied to each pixel to calculate the corresponding droplet heat transfer. The
energy balance applied to the pixel element results in
Qstored = Qgen + Qcond − Qrad − Qconv − Qdrop (25)
where droplet input heat transfer is represented as Qdrop
Thus,
Qdrop = Qgen − Qstored + Qcond − Qrad − Qconv (26)
A continuous DC supply is provided to the surface and is maintained at a constant temperature. Upon droplet impingement, considerable heat
transfer takes place resulting in the cooling of the surface. The generated heat due to the DC supply is calculated as Qgen
I 2RVp
Qgen =
Vs (27)
where I being the supplied current, Vp and Vs are the volumes of considered pixel element and total surface respectively.
Following the reference [56], the heater foil resistance ’R’ is obtained from
ρs Lp (1 + αs (T − T∞))
R=
Ap (28)
where ρs , Lp , Ap (=Lp δ ), αs and δ represents surface electrical resistivity, pixel length, cross-sectional area, temperature coefficient of resistance, and
thickness of the pixel element respectively. The properties of the surface is outlined in Table 2.
The net energy change in the pixel, is termed as stored heat Qstored
ms c (Tt − Tt − 1)
Qstored =
dt (29)
where ms is the mass of the pixel element, c specific heat capacity, Tt and Tt − 1 are the temperatures of the pixel element at a time intervals of t and
t − 1 respectively.
Due to negligible thickness [56,62], the conduction effects perpendicular to the heater surface is minimal compared to other directions.

18
G. Guggilla, et al. Experimental Thermal and Fluid Science 110 (2020) 109916

Thus, the conduction heat transfer along the surface is taken into account and is given as
Qcond = (Qcond )in − (Qcond )out (30)
can be simplified into
ks Ap (Ti + 1, j + Ti − 1, j + Ti, j + 1 + Ti, j − 1 − 4Ti, j )
Qcond =
Lp (31)
where ks is surface thermal conductivity and Ti, j represents the temperature of considered element, and Ti + 1, j, Ti − 1, j, Ti, j + 1, Ti, j − 1 are the temperatures of
neighbouring pixel elements in respective directions.
The bottom side of heater surface is coated black and is maintained at high temperatures. The radiation heat transfer underneath the surface is
considered as
Qrad = σ ∊r Lp2 (Ti, j 4 − T∞4 ) (32)
Also, natural convection currents will form eventually underneath the hot surface which can be calculated as
Qconv = hi Lp2 (Ti, j − T∞) (33)
where natural convectional heat transfer coefficient at a pixel element, hi can be taken from the correlation
hi = 0.27Rai0.25 (34)
and Rai is the Rayleigh number and all the properties are considered at the film temperature Tf
Ti, j + T∞
Tf =
2 (35)

References [23] M. Pasandideh-Fard, S. Aziz, S. Chandra, J. Mostaghimi, Cooling effectiveness of a

water drop impinging on a hot surface, Int. J. Heat Fluid Flow 22 (2) (2001)
201–210.
[1] A. Prosperetti, H.N. Oguz, The impact of drops on liquid surfaces and the under- [24] Q. Cui, S. Chandra, S. McCahan, The effect of dissolving gases or solids in water
water noise of rain, Annu. Rev. Fluid Mech. 25 (1) (1993) 577–602. droplets boiling on a hot surface, J. Heat Transf. 123 (4) (2001) 719–728.
[2] M. Rein, Phenomena of liquid drop impact on solid and liquid surfaces, Fluid Dyn. [25] V. Nakoryakov, S. Misyura, S. Elistratov, The behavior of water droplets on the
Res. 12 (2) (1993) 61–93. heated surface, Int. J. Heat Mass Transf. 55 (23) (2012) 6609–6617.
[3] A. Yarin, Drop impact dynamics: splashing, spreading, receding, bouncing, Annu. [26] S. Herbert, T. Gambaryan-Roisman, P. Stephan, Influence of the governing di-
Rev. Fluid Mech. 38 (1) (2006) 159–192. mensionless parameters on heat transfer during single drop impingement onto a hot
[4] M. Marengo, C. Antonini, I.V. Roisman, C. Tropea, Drop collisions with simple and wall, Colloids Surf., A 432 (2013) 57–63.
complex surfaces, Curr. Opin. Colloid Interface Sci. 16 (4) (2011) 292–302. [27] M. Francois, W. Shyy, Computations of drop dynamics with the immersed boundary
[5] A. Moreira, A. Moita, M. Panão, Advances and challenges in explaining fuel spray method, part 2: Drop impact and heat transfer, Numer. Heat Transf., Part B:
impingement: how much of single droplet impact research is useful? Prog. Energy Fundam. 44 (2) (2003) 119–143.
Combust. Sci. 36 (5) (2010) 554–580. [28] G. Strotos, M. Gavaises, A. Theodorakakos, G. Bergeles, Numerical investigation of
[6] C. Josserand, S. Thoroddsen, Drop impact on a solid surface, Annu. Rev. Fluid the cooling effectiveness of a droplet impinging on a heated surface, Int. J. Heat
Mech. 48 (1) (2016) 365–391. Mass Transf. 51 (19) (2008) 4728–4742.
[7] G. Liang, I. Mudawar, Review of drop impact on heated walls, Int. J. Heat Mass [29] Y. Ge, L.-S. Fan, 3-d modeling of the dynamics and heat transfer characteristics of
Transf. 106 (2017) 103–126. subcooled droplet impact on a surface with film boiling, Int. J. Heat Mass Transf. 49
[8] Y.S. Ko, S.H. Chung, An experiment on the breakup of impinging droplets on a hot (21) (2006) 4231–4249.
surface, Exp. Fluids 21 (2) (1996) 118–123. [30] W. Healy, J. Hartley, S. Abdel-Khalik, On the validity of the adiabatic spreading
[9] J.D. Naber, P.V. Farrell, Hydrodynamics of droplet impingement on a heated sur- assumption in droplet impact cooling, Int. J. Heat Mass Transf. 44 (20) (2001)
face, in: SAE Technical Paper, SAE International, 1993. 3869–3881.
[10] J.D. Bernardin, C.J. Stebbins, I. Mudawar, Mapping of impact and heat transfer [31] L. Tarozzi, A. Muscio, P. Tartarini, Experimental tests of dropwise cooling on in-
regimes of water drops impinging on a polished surface, Int. J. Heat Mass Transf. 40 frared-transparent media, Exp. Thermal Fluid Sci. 31 (8) (2007) 857–865.
(2) (1997) 247–267. [32] Y.M. Qiao, S. Chandra, Experiments on adding a surfactant to water drops boiling
[11] J.D. Bernardin, C.J. Stebbins, I. Mudawar, Effects of surface roughness on water on a hot surface, Proc. Roy. Soc. London. Ser. A: Math. Phys. Eng. Sci. 453 (1959)
droplet impact history and heat transfer regimes, Int. J. Heat Mass Transf. 40 (1) (1997) 673–689.
(1996) 73–88. [33] T. Okawa, K. Nagano, T. Hirano, Boiling heat transfer during single nanofluid drop
[12] A.-B. Wang, C.-H. Lin, C.-C. Cheng, Pattern analysis of a single droplet impinging impacts onto a hot wall, Exp. Thermal Fluid Sci. 36 (2012) 78–85.
onto a heated plate, Heat Transf. Asian Res. 34 (8) (2005) 579–594. [34] T. Xiong, M. Yuen, Evaporation of a liquid droplet on a hot plate, Int. J. Heat Mass
[13] A.-B. Wang, C.-H. Lin, C.-C. Chen, The critical temperature of dry impact for tiny Transf. 34 (7) (1991) 1881–1894.
droplet impinging on a heated surface, Phys. Fluids 12 (6) (2000) 1622–1625. [35] S. Kandlikar, M. Steinke, High speed photographic investigation of liquid-vapor
[14] H.J.J. Staat, T. Tran, B. Geerdink, G. Riboux, C. Sun, J.M. Gordillo, D. Lohse, Phase interface and contact line movement during CHF and transition boiling, Vol. 369,
diagram for droplet impact on superheated surfaces, J. Fluid Mech. 779 (2015) R3. 2001, pp. 323–329.
[15] T. Tran, H.J.J. Staat, A. Prosperetti, C. Sun, D. Lohse, Drop impact on superheated [36] F. McGinnis, J. Holman, Individual droplet heat-transfer rates for splattering on hot
surfaces, Phys. Rev. Lett. 108 (2012) 036101. surfaces, Int. J. Heat Mass Transf. 12 (1) (1969) 95–108.
[16] M.A.V. Limbeek, M. Shirota, P. Sleutel, C. Sun, A. Prosperetti, D. Lohse, Vapour [37] J. Holman, P. Jenkins, F. Sullivan, Experiments on individual droplet heat transfer
cooling of poorly conducting hot substrates increases the dynamic Leidenfrost rates, Int. J. Heat Mass Transf. 15 (8) (1972) 1489–1495.
temperature, Int. J. Heat Mass Transf. 97 (2016) 101–109. [38] J. Bernardin, I. Mudawar, The Leidenfrost point: experimental study and assessment
[17] M. Di Marzo, D.D. Evans, Evaporation of a water droplet deposited on a hot high of existing models, J. Heat Transf. 121 (4) (1999) 894–903.
thermal conductivity surface, J. Heat Transf. 111 (1) (1989) 210–213. [39] J.D. Bernardin, I. Mudawar, A cavity activation and bubble growth model of the
[18] M.D. Marzo, P. Tartarini, Y. Liao, D. Evans, H. Baum, Evaporative cooling due to a leidenfrost point, J. Heat Transf. 124 (5) (2002) 864–874.
gently deposited droplet, Int. J. Heat Mass Transf. 36 (17) (1993) 4133–4139. [40] C. Avedisian, J. Koplik, Leidenfrost boiling of methanol droplets on hot porous/
[19] O.E. Ruiz, W.Z. Black, Evaporation of water droplets placed on a heated horizontal ceramic surfaces, Int. J. Heat Mass Transf. 30 (2) (1987) 379–393.
surface, J. Heat Transf. 124 (5) (2002) 854–863. [41] C. Pedersen, An experimental study of the dynamic behavior and heat transfer
[20] E. Berberović, I.V. Roisman, S. Jakirlić, C. Tropea, Inertia dominated flow and heat characteristics of water droplets impinging upon a heated surface, Int. J. Heat Mass
transfer in liquid drop spreading on a hot substrate, Int. J. Heat Fluid Flow 32 (4) Transf. 13 (2) (1970) 369–381.
(2011) 785–795. [42] F. Celestini, T. Frisch, Y. Pomeau, Room temperature water Leidenfrost droplets,
[21] G. Strotos, M. Gavaises, A. Theodorakakos, G. Bergeles, Numerical investigation on Soft Matter 9 (2013) 9535–9538.
the evaporation of droplets depositing on heated surfaces at low Weber numbers, [43] H.-M. Kwon, J.C. Bird, K.K. Varanasi, Increasing Leidenfrost point using micro-nano
Int. J. Heat Mass Transf. 51 (7) (2008) 1516–1529. hierarchical surface structures, Appl. Phys. Lett. 103 (20) (2013) 201601.
[22] S. Chandra, M. di Marzo, Y. Qiao, P. Tartarini, Effect of liquid-solid contact angle on [44] D. Arnaldo del Cerro, L.G. Marín, G.R.B.E. Römer, B. Pathiraj, D. Lohse, A.J. Huis in
droplet evaporation, Fire Saf. J. 27 (2) (1996) 141–158. ’t Veld, Leidenfrost point reduction on micropatterned metallic surfaces, Langmuir

19
G. Guggilla, et al. Experimental Thermal and Fluid Science 110 (2020) 109916

28 (42) (2012) 15106–15110. J. Heat Mass Transf. 123 (2018) 1050–1067.

[45] G.P. Celata, M. Cumo, A. Mariani, G. Zummo, Visualization of the impact of water [54] S. Batzdorf, J. Breitenbach, C. Schlawitschek, I.V. Roisman, C. Tropea, P. Stephan,
drops on a hot surface: effect of drop velocity and surface inclination, Heat Mass T. Gambaryan-Roisman, Heat transfer during simultaneous impact of two drops
Transf. 42 (10) (2006) 885. onto a hot solid substrate, Int. J. Heat Mass Transf. 113 (2017) 898–907.
[46] V. Bertola, K. Sefiane, Controlling secondary atomization during drop impact on hot [55] C.A. Schneider, W.S. Rasband, K.W. Eliceiri, NIH image to imagej: 25 years of image
surfaces by polymer additives, Phys. Fluids 17 (10) (2005) 108104. analysis, Nat. Methods 9 (7) (2012) 671.
[47] D.V. Zaitsev, D.P. Kirichenko, V.S. Ajaev, O.A. Kabov, Levitation and self-organi- [56] N. Schweizer, Multi-scale investigation of nucleate boiling phenomena in micro-
zation of liquid microdroplets over dry heated substrates, Phys. Rev. Lett. 119 (9) gravity (Ph.D. thesis), Technische Universität, Darmstadt, 2010.
(2017) 094503. [57] Š. Šikalo, H.-D. Wilhelm, I. Roisman, S. Jakirlić, C. Tropea, Dynamic contact angle
[48] O.A. Kabov, D.V. Zaitsev, D.P. Kirichenko, V.S. Ajaev, Interaction of levitating of spreading droplets: experiments and simulations, Phys. Fluids 17 (6) (2005)
microdroplets with moist air flow in the contact line region, Nanoscale Microscale 062103.
Thermophys. Eng. 21 (2) (2017) 60–69. [58] S. Chandra, C. Avedisian, On the collision of a droplet with a solid surface, Proc. R.
[49] J.D. Bernardin, I. Mudawar, Film boiling heat transfer of droplet streams and Soc. Lond. A 432 (1884) (1991) 13–41.
sprays, Int. J. Heat Mass Transf. 40 (11) (1997) 2579–2593. [59] J. Wakefield, C.F. Tilger, M.A. Oehlschlaeger, The interaction of falling and sessile
[50] H. Fujimoto, A.Y. Tong, H. Takuda, Interaction phenomena of two water droplets drops on a hydrophobic surface, Exp. Thermal Fluid Sci. 79 (2016) 36–43.
successively impacting onto a solid surface, Int. J. Therm. Sci. 47 (3) (2008) [60] J. Jung, S. Jeong, H. Kim, Investigation of single-droplet/wall collision heat transfer
229–236. characteristics using infrared thermometry, Int. J. Heat Mass Transf. 92 (2016)
[51] J. Breitenbach, I.V. Roisman, C. Tropea, From drop impact physics to spray cooling 774–783.
models: a critical review, Exp. Fluids 59 (3) (2018) 55. [61] S. Batzdorf, Heat transfer and evaporation during single drop impingement onto a
[52] T. Minamikawa, H. Fujimoto, T. Hama, H. Takuda, Numerical simulation of two superheated wall (Ph.D. thesis), Technische Universität (2015).
droplets impinging successively on a hot solid in the film boiling regime, ISIJ Int. 48 [62] E. Teodori, P. Pontes, A. Moita, A. Moreira, Thermographic analysis of interfacial
(5) (2008) 611–615. heat transfer mechanisms on droplet/wall interactions with high temporal and
[53] G. Guggilla, A. Pattamatta, R. Narayanaswamy, Numerical investigation into the spatial resolution, Exp. Thermal Fluid Sci. 96 (2018) 284–294.
evaporation dynamics of drop-on-drop collisions over heated wetting surfaces, Int.

20