See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/305926336

Temperature Patterns of Air Crossflow with

Hollow-Cone Spray Evaporation

Article in AIAA Journal August 2016

DOI: 10.2514/1.J054338

CITATIONS READS

2 114

2 authors:

Haibin Zhang Bo Feng Bai

Xi'an Jiaotong University Xi'an Jiaotong University
23 PUBLICATIONS 88 CITATIONS 207 PUBLICATIONS 1,049 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Flow rate measurement of natural gas/wet gas View project

Development of scientific and technological potential for high-performance droplet-based cooling

systems View project

All content following this page was uploaded by Bo Feng Bai on 13 August 2016.

The user has requested enhancement of the downloaded file.

 AIAA JOURNAL

Temperature Patterns of Air Crossflow with Hollow-Cone

Spray Evaporation

Haibin Zhang
Xian Jiaotong University, 710049 Xian, Peoples Republic of China
and
Bofeng Bai
State Key Laboratory of Multiphase Flow in Power Engineering, Xian Jiaotong University,
710049 Xian, Peoples Republic of China
DOI: 10.2514/1.J054338
This work focused on the heat transfer of the mixing process resulting from a hollow-cone spray evaporation in
high-temperature gas crossflow in a confined mixing space. The study was carried out in a rectangular duct with a
Downloaded by XI'AN JIAOTONG UNIVERSITY on August 6, 2016 | http://arc.aiaa.org | DOI: 10.2514/1.J054338

cross-sectional area of 95 95 mm. The temperatures on four cross sections (45 measuring points on half of a cross
section) of crossflow were measured using thermocouples. The temperature pattern of gas crossflow caused by the
evaporation of hollow-cone spray was obtained, and the effects of factors such as crossflow velocity, crossflow
temperature, spray droplet size, and spray injection angle on the temperature pattern were investigated. The main
findings of this study were as follows: 1) The polydispersed hollow-cone spray droplets and the droplet-wall
impingement made the heat transfer on the cross section of the flowfield inherently nonuniform. 2) The mixing
flowfield fell into five regions characterized by different heat transfer mechanisms. 3) The temperature pattern of
crossflow was significantly affected by the spray injection angle. 4) A good mixing between hollow-cone spray and
high-temperature crossflow could only be achieved with a proper spray penetration depth.

Nomenclature g = dynamic viscosity of crossflow, Pa s

A, B = coefficients of correlation of Nusselt number = spatial uniformity index
BT = heat transfer number d = density of liquid, kgm3
Cp;f = specific heat of droplet film, J
kg K g = density of gas flow, kgm3
D = length of side of square duct, mm d = surface tension of liquid, Nm
D32 = Sauter mean diameter, m
d = diameter of droplet, m Subscripts
H = height above the bottom wall, m
g = air crossflow
hfg = enthalpy of vaporization, Jkg
i = measurement point
J = dimensionless criteria number
M_l = mass flow rate of the liquid spray, kgs
m_l = liquid flow rate, mls
Nu = Nusselt number I. Introduction
Pr
Red
T diff
=
=
=
Prandtl number
Reynolds number of droplet
difference of temperature drop, K
T HE mixing of liquid spray (jet) in gas crossflows is an important
topic in the areas of propulsion, combustion, cooling, and the
pharmaceutical industry. For example, in propulsion engineering,
T drop = temperature of droplet, K some engines (e.g., hydroreactive metal fuel engines [1,2]) use liquid
Tg = initial crossflow temperature, K coolant (also as working substance addition) injection in the mixing
T mix
g = temperature of crossflow in the mixed flowfield, K chamber, in which the spray evaporates in a high-temperature fuel gas
Ug = crossflow velocity, ms crossflow discharged from the combustion chamber. The mixing
Urel = relative velocity between crossflow and droplet, ms chamber is generally small (e.g., less than 100 mm in diameter), and
Vd = initial droplet velocity, ms this requires the mixing to be well accomplished as soon as possible
V In = droplet velocity immediately before impingement, ms after the liquid is atomized to the hollow-cone spray droplets by the
We = Weber number pressure-swirl nozzles. It is therefore important to study the heat
= spray injection angle, deg transfer characteristics of this gas-droplet two-phase flow produced
P = spray pressure drop over nozzle, MPa by the injection of hollow-cone spray into high-temperature cross-
T = temperature drop, K flow. There are, however, very few studies available in the literature
T = temperature-drop average of a cross section concerning this area of mixing, and the vast majority of spray/
= spray angle, deg crossflow-related studies only considers the spray characteristics
f = thermal conductivity of droplet film, W
m K produced by the sprayer (or jet) in crossflows at ambient temperature
in a relatively large space (in which the influence of walls can be
ignored). This paper aims to reveal the heat transfer characteristics
Received 11 March 2015; revision received 9 October 2015; accepted for between the hollow-cone spray and the high-temperature crossflow
publication 18 November 2015; published online 5 August 2016. Copyright in a small mixing space, and it aims to extend the present knowledge
2015 by the American Institute of Aeronautics and Astronautics, Inc. All in the area of spray/crossflow mixing.
rights reserved. Copies of this paper may be made for personal and internal Among the previous relevant studies of liquid spray (or jets) in
use, on condition that the copier pay the per-copy fee to the Copyright crossflows, the most widely studied forms of sprayer or jet are round
Clearance Center (CCC). All requests for copying and permission to reprint
should be submitted to CCC at www.copyright.com; employ the ISSN 0001- liquid jets [39], air-blast or air-assisted sprays [1014], plat-fan
1452 (print) or 1533-385X (online) to initiate your request. sprays [15,16], and effervescent sprays [1720]. In those works, the
*Associate Professor, School of Chemical Engineering and Technology; authors provided insight into the spray characteristics, droplet
hb-zhang@mail.xjtu.edu.cn. dispersion, large-scale vortex structures in crossflow, and the

Professor; bfbai@mail.xjtu.edu.cn (Corresponding Author). influence of various factors [e.g., liquid injection angle, liquid-to-air
Article in Advance / 1
 2 Article in Advance
Fig. 1 Schematic diagram of the mixing system for a hollow-cone spray in high-temperature crossflow.
momentum ratio (in round liquid jet), or gas-to-liquid flow rate ratio
(in effervescent spray)] on these aspects. Comparatively,
Downloaded by XI'AN JIAOTONG UNIVERSITY on August 6, 2016 | http://arc.aiaa.org | DOI: 10.2514/1.J054338
investigations on the mixing of the hollow-cone spray in crossflow
are relatively scarce. The relevant studies in the literature include
those of Ghosh and Hunt [21], Lynch et al. [22], Prakash et al. [23],
Deshpande et al. [24], Lee et al. [25], Gao et al. [26], and previous
works by Bai et al. [27] and Zhang et al. [2830]. The studies by
Ghosh and Hunt [21], Lynch et al. [22], and Lee et al. [25] only
focused on the spray characteristics. Deshpande et al. [24] and Gao
et al. [26] numerically studied the mixing of hollow-cone spray in
crossflow by using simplified atomization models; the characteristics
of droplet dispersion were investigated, and the counter-rotating
vortex pair (CVP) structure was discovered. In previous investi-
gations by Bai et al. [27] and Zhang et al. [2830], the mixing of
hollow-cone spray in crossflow in a small space was studied
experimentally. The droplet dispersion, the formation of large-scale
vortices, and their interactions were studied. The results showed that,
in a small mixing space, the droplet-wall impingement was inevitable
and the large-scale vortices such as CVPs and coherent structures,
which might exert great influence on the droplet dispersion, were
induced in the confined crossflow. However, in all of the afore-
mentioned studies, the spray/crossflow mixings were carried out at
ambient temperature. Therefore, it is important to improve the
knowledge on the mixing of hollow-cone spray in a confined
crossflow at an elevated temperature.
The main objective of this work is to characterize the heat transfer
during the mixing of a hollow-cone spray and a confined crossflow at
high-temperature conditions. The study is conducted in a rectangular
duct, where the airflow is preheated by ac power supplies. Subse-
quently, the temperatures on several cross sections of the duct are
measured using thermocouples. Special emphasis is given to the
temperature pattern of the gas crossflow caused by the evaporation of
the hollow-cone spray and to the effects of different factors on the
temperature pattern. Furthermore, the degrees of the spatial uni-
formity of the temperature drops on the cross sections are evaluated
and the way for mixing enhancement is discussed. This paper is
expected to give an insight into the complex mixing process and
provide useful data for further numerical and theoretical works.
II. Materials and Methods
A. Experimental System
A schematic diagram of the experimental setup of the hollow-cone
spray in a high-temperature crossflow is shown in Fig. 1. The setup
consists of an air supply system, a liquid supply system, an electric
forced-air heating system, a test section, and a temperature
measurement system. A centrifugal blower is used to generate the
desired airflow in the test channel. The air enters the channel under
atmospheric conditions via a flow meter, a heater, and a rectifier. The
heater consists of several finned electric heating tubes and is heated
by ac power supplies, with a maximum air heating power of up to
120 kW. Downstream of the heater, the air enters the rectifier, which
consists of a wire mesh and honeycomb to ensure the uniformity and
stability of the airflow. The crossflow turbulence is so weak that it
/ ZHANG AND BAI
makes no significant addition to the spray turbulence. The rectified
air flows through the steady flow tube and then into the test channel.
To minimize the heat loss, the high-quality insulating material is
applied to the outer tube wall of the heating system, the rectifier,
the steady flow tube, and the test section. The liquid supply system
consists of a pressurized liquid tank and a flow regulation system. The
high-pressure nitrogen gas is admitted to the liquid tank to propel
the liquid into the rig. The pressure of the nitrogen gas is controlled
by the pressure regulator and remains constant during the test. An
acetone is used as the working liquid because of its low boiling point
(329.4 K). The discharged acetone from the liquid tank passes via the
flow meter and then flows through the regulating valves and into the
nozzle. A needle valve located ahead of the nozzle is used to regulate
the acetone flow rate precisely. The mixing between the spray and
the crossflow happens in the test section.
Figure 2 shows a sketch of the test section, which is a long, square
duct made of stainless steel (304) with a cross section of 0.095
0.095 m and a length of 0.73 m, in correspondence with the practical
engine structure. The nozzle is located 206 mm downstream from the
inlet of the test section. To measure the temperature and the droplet
distributions, the forepart of the test section is replaceable and two
different sections are designed (see Fig. 2). Section 1 is used for the
temperature measurement; whereas section 2 is used for the droplet
visualization, and its one side is slotted. The cross sections of the
flowfield are lit up by the laser sheet, and then the initial spray
droplets are captured by a charge-coupled device (CCD) camera.
The coordinate axes used to orient the measurements are centered
at the nozzle orifice. The three mutually orthogonal directions (X, Y,
and Z) are aligned as shown in Fig. 2. To evaluate the heat transfer
between the spray and the crossflow, a temperature-measuring device
is designed to obtain the temperature distributions on four different
cross sections (xD  2.15, 3.23, 4.30, and 5.39). The temperature is
gauged by 45 NiCr-NiSi K-type thermocouples with 0.3 mm.
Since the mixing flowfield is bilaterally symmetrical, the thermo-
couples are uniformly located on a half-side of the measured cross
sections. A detailed view of the positioning of the thermocouples is
provided in Fig. 3.
B. Spray Facilities
Modified commercially available nozzles (KB-series hollow-cone
spray nozzle, H. Ikeuchi & Company, Ltd.) are used. In this type of
nozzle, the spring is arranged in the cylindrical swirl chamber and
four guide channels are made on the orifice, which contribute to the
generation of the extremely fine spray droplets under a relatively low
spray pressure [28]. In our experiment, the polytetrafluoroethylene
(PTFE) closer and orifice are replaced with the stainless-steel ones
in order to avoid melting of the PTFE. The nozzle spray angle
(  80 deg) and the injection angle are illustrated in Fig. 4.
A simple heat exchanger is installed upstream of the spray nozzle
to keep the temperature of the discharged acetone constant. The
cooling water temperature is 283 K, and the temperature of the
discharged acetone is maintained at 293 K. The spray nozzle is fixed
on the test section via a pedestal, which also has the function for the
adjustment of the nozzle injection angle (see Fig. 5).
 Article in Advance / ZHANG AND BAI 3

Fig. 2 Sketch of the test section of the mixing system.

Downloaded by XI'AN JIAOTONG UNIVERSITY on August 6, 2016 | http://arc.aiaa.org | DOI: 10.2514/1.J054338

C. Test Procedures and Uncertainty Analysis

Fig. 3 Temperature-measuring device and positions of thermocouples
(unit: millimeters).
Fig. 4 Schematic of the spray angle and the spray injection angle of the
nozzle.
Our experiments were performed at the ambient pressure. The
initial atomized droplet diameter (Sauter mean diameter D32 ) of
the spray is measured by using a Malvern Mastersizer particle size
analyzer before the experiment at ambient temperature, and the
detailed experimental conditions are shown in Table 1. At first, we
regulated the flow volumetric flow rate of the air and the pressure of
the acetone under the predetermined operating condition, opened the
valve of the cooling water, and heated the air to the desired tem-
perature. Then, we opened the spray controlling valve when the air
temperature was stable. Finally, we measured the flowfield temper-
atures on the different cross sections with the different injection
angles when the flowfield became steady. Simultaneously, we
collected the experimental data with a data acquisition system
(National Instruments). A CCD camera captured the droplet distri-
butions on the cross sections of the flowfield.
The uncertainties of measurement depend on the experimental
conditions and the measuring instruments. The uncertainties of
variables are given in Table 2. The measuring error of pressure is
estimated through the precision of the pressure transmitter and the
data acquisition board, which are 0.01 and 0.02%, respectively. The
uncertainties in the temperature measurements are estimated with the
error of the thermocouples and the error of the data acquisition board,
and the maximum error of the thermocouples is 1.5C. The mean gas
flow rate is measured with the vortex shedding flow meter with a
variation of less than 2%. Therefore, the uncertainty in the crossflow
velocity that depends on the accuracy of the volume flow measured
with the vortex flow meter, and on the precision of the temperature
measured by the thermocouple, is less than 3%. The mass flow of the
acetone is measured with an electromagnetic flow meter with an
uncertainty of 1%. Keller pressure sensors with an uncertainty of
0.01% are used to precisely control the pressure drop over the nozzle.
To validate the experimental results, we conduct two independent
experiments under each of the testing conditions.

III. Results and Discussions

A. Temperature Distribution Pattern on Cross Section
By using the cubic spline interpolation method, the continuous
temperature distribution in the measuring region (half of a cross
section) is calculated based on the temperatures measured on discrete
points. Since the flowfield is symmetric, the temperature distribution
on the whole cross section is obtained and the temperature contour
map of the cross section is produced by employing MATLAB. In the
experiments, it is found that most of the droplets evaporate com-
pletely ahead of the first temperature measurement cross section and,
at the last (forth) temperature measurement cross section, the droplets
are scarce. Hence, we can consider that the evaporation of the spray is
accomplished in the temperature measurement domain. The typical
Fig. 5 Heat exchanger and the pedestal of the nozzle. droplet distributions on different cross sections of the mixing
 4 Article in Advance / ZHANG AND BAI

Table 1 Experimental conditions

Case P, MPa m
_ l , mls D32 , m V d , ms T g , K Ug , ms , deg
1 0.4 1.92 47 18.5 473 10 60, 75, 90, 105, 120
2 0.4 1.92 47 18.5 573 10 60, 75, 90, 105, 120
3 0.4 1.92 47 18.5 573 20 60, 75, 90, 105, 120
4 0.4 1.92 47 18.5 573 30 60, 75, 90, 105, 120
5 0.4 1.92 47 18.5 573 40 60, 75, 90, 105, 120
6 0.7 2.57 46 21.3 573 40 60, 75, 90, 105, 120
7 0.4 3.12 65 18.7 573 10 90
8 0.4 3.12 65 18.7 573 20 90
9 0.4 3.12 65 18.7 573 30 90

Table 2 Uncertainties of variables The reasons for this unique temperature pattern are the poly-
dispersed hollow-cone spray droplets and the spray-wall impinge-
Parameters Standard uncertainty ment process. For a hollow-cone spray, the droplet concentration is
Pressure 0.02 MPa large near the nozzle and decreases rapidly as the droplets disperse.
Crossflow temperature 1.34 K
Downloaded by XI'AN JIAOTONG UNIVERSITY on August 6, 2016 | http://arc.aiaa.org | DOI: 10.2514/1.J054338

Crossflow velocity 0.7 ms
Liquid flow rate 0.03 mls
Spray injection angle 2.1 deg
flowfield near the nozzle are shown in Fig. 6. From the droplet
distributions, we can see that the spray features a conical distribution
at first; then, as the mixing develops, the conical spray structure
disappears gradually and the droplet distribution tends to be uniform,
accompanied by the dispersion and evaporation of the droplets. In
addition, a high concentration of droplets is observed near the
sidewall. The temperature distributions are measured on the cross
sections located relatively far downstream (xD > 2) of the spray
(see Fig. 7). It can be seen clearly that the temperature distribution on
the cross sections is evidently nonuniform. Large temperature drop
occurs in the upper middle part of the cross sections, where the
droplet concentration is high, and in the middle part near the side-
walls, where the droplet-wall impingement takes place. The mod-
erate temperature-drop areas are mainly located in the middle and
lower parts of the cross section. Apart from the areas close to the top
corners, two small low-temperature-drop areas (high-temperature
spots) are seen in the lower part of the cross section. This temperature
pattern can persist for a relatively long distance along the mixing
flowfield. As the droplet evaporation and vapor convection continue,
the areas for lower- and higher-temperature regions shrink and the
temperature differences among different regions diminish gradually.
Thus, in the upper middle part of the mixing flowfield, the fierce
evaporation of droplets is caused by the interaction between the hot
gas flow and the high concentration of droplets, and hence large
temperature drops result in this area; but, beyond this region, the
temperature drops of the crossflow decrease with decreasing droplet
concentration. Since the mixing space is small, the droplet im-
pingement on the sidewalls is inevitable. When the spray impinges on
the hot wall, a thin liquid film will form on the wall surface; thereafter,
the succeeding droplets impinge on the liquid film [31]. The blowing-
upward boiling vapor caused by the hot wall and the droplet-film
impingement will make the film or the impinging droplets break up
into small splashing droplets. Hence, due to the droplet impinge-
ment and the interaction between gas flow and the resultant high
concentration of droplets, the intense heat transfer happens in this
area and a large temperature drop is thus achieved. In our previous
studies on the cold state mixing of the hollow-cone spray with a
confined crossflow [28,29], under the combined effects of shear
stress on the gas flow exerted by the high-speed spray droplets and the
spray-wall impingement, a counter-rotating vortex pair (CVP) is
formed in the lower part of the flowfield where small droplets can
be entrained by the rotary gas flow. In this study of hollow-cone spray
mixing with a high-temperature crossflow, two obvious high-
temperature spots are observed in the lower part of the cross section
of the flowfield (see Fig. 7). This indicates that the CVP is induced
and the high-temperature spot is generated in the vortex core region
where the convection is relatively weak. Apart from the vortex core
regions, the heat transfer is promoted in the lower part of the flowfield

Fig 6 Spatial droplet distributions on different cross sections near the nozzle (case 2).

Fig 7 Temperature distributions on different cross sections of the mixing flowfield (case 2).
 Article in Advance
Fig. 8 Temperature distribution pattern on cross sections of mixing
Downloaded by XI'AN JIAOTONG UNIVERSITY on August 6, 2016 | http://arc.aiaa.org | DOI: 10.2514/1.J054338
flowfield.
because of the entrained cooled gas flow (from the droplet-wall
impingement region). In the area close to the top corner, the droplets
are scarce, and thus the temperature drop is low. Based on the
aforementioned analysis and the features of temperature distribution,
the flowfield of hollow-cone spray mixing with a confined hot
crossflow can be divided into five regions (see Fig. 8), namely, the
corner region (I), spray region (II), droplet-wall impingement region
(III), mainstream region (IV), and bottom CVP region (V). Clearly,
the largest temperature drops occur in the spray region (II) and
droplet-wall impingement region (III), followed by mainstream
region (IV), and then the bottom CVP region (V) and corner
region (I).
From the preceding discussion, we know that the temperature
pattern of the hollow-cone spray/crossflow mixing flowfield is
primarily determined by the evaporation and dispersion character-
istics of spray droplets, and the distribution of five regions as well as
the temperature drop on them should be strongly influenced by the
spray and crossflow conditions. Thus, the influences of different
parameters on the temperature pattern such as the crossflow velocity,
crossflow temperature, and liquid flow rate are investigated. First, a
simply droplet evaporation model that will be used in the analyses is
given here. According to the studies on droplet evaporation [32,33],
the evaporation rate of a spherical droplet in high-temperature gas
flow can be expressed in terms of heat transfer by
f liquid flow rate leads to an increase in the number and the velocity of
_  d
m Nu ln
1  BT  (1)
Cp;f
where the heat transfer number BT and the Nusselt number Nu are
given, respectively, by
Cp;f
T g T drop 
BT  (2)
hfg
Nu  A  B Re12
d Pr
13 (3)
where the coefficients A and B vary in different models, and the
Reynolds number Red, is given by
g Urel d
Red  (4)
g 0.4 MPa are shown in Figs. 10a and 10b (cases 8 and 9 in Table 1). In
The temperature patterns on cross sections along the flowfield
under different spray and crossflow conditions are given in Figs. 9a
9d. As can be seen in these figures, clear differences of temperature
drop in five regions are observed. From Figs. 9a and 9b, we can see
the influence of crossflow velocity on temperature drop on the cross
/ ZHANG AND BAI 5
sections. We find that the increase in crossflow velocity exerts great
influence on the temperature drop in the spray region (II) and the
bottom CVP region (V). Based on the droplet evaporation model, we
know that, when crossflow velocity increases, the initial relative
velocity Urel becomes larger, and thus a higher droplet evaporation
rate will be expected. However, on the other hand, when the crossflow
velocity increases, the spray penetration depth will decrease due to
the increased crossflow momentum; thereby, the droplet dispersion in
the transverse direction becomes weaker. Hence, a smaller tem-
perature drop is observed in the corner region (I) and a smaller spray
region (II) is observed in the upper part of the flowfield in the case
with higher crossflow velocity (see in Fig. 9b). In addition, because of
the higher evaporation rate and the lower spray penetration depth
caused by the increase of crossflow velocity, fewer droplets can reach
the bottom region of the flowfield, and thus the temperature drop
becomes smaller in this region.
The influences of the crossflow temperature on temperature drop
on the cross sections can be seen in case 2 (see Fig. 7) and case 1(see
Fig. 9c), in which the crossflow temperatures are 573 and 473 K,
respectively, whereas the crossflow velocity (Ug  10 ms) and the
spray conditions are same. From Eqs. (1) and (2), we know that, when
the gas flow temperature decreases, the droplet evaporation rate will
decrease as a result of a smaller heat transfer number BT ; thus, the
droplets can disperse further and the spray penetration depth
increases. Therefore, for a lower crossflow temperature condition, a
smaller temperature drop in the spray region (II) is produced and the
size of this region shrinks more quickly; an intensified droplet-wall
impingement process is triggered with more droplets involved, and
thus a larger droplet-wall impingement region is caused; and the heat
transfer in the lower part of the flowfield is enhanced obviously as a
consequence of an increase in spray penetration depth. In addition,
we can find that the decrease in crossflow temperature can also
produce a weak heat transfer in the top corner region (I) (see case 1).
For a given hollow-cone spray, the characteristics of initial spray
droplets such as the droplet size, the droplet concentration, and the
droplet speed have a clear relation to the liquid flow rate. It is hard to
vary one parameter while keeping the others constant. Hence, the
influence of the liquid flow rate on the temperature drop on the cross
sections is analyzed qualitatively here. Figure 9d shows the
temperature distributions on cross sections of the flowfield under the
condition of Ug  40 ms, T g  573 K, and m _ l  2.57 (case 6 in
Table 1). Comparing case 6 with case 5, in which the crossflow
parameters are the same and the initial spray droplets sizes D32 are
similar, we can see that when the liquid flow rate increases, except for
the top corner region (I): an obvious increase of temperature drops on
the cross sections is obtained. This is because the increase in the
initial spray droplet. The droplet concentration increases can increase
the evaporation amount. Base on the droplet evaporation model
[Eqs. (14)], the droplet velocity increases can enlarge the relative
velocity Ure ) between droplet and gas flow; thereby, a larger Nusselt
number N u is produced and the evaporation of the droplet is
promoted, which all contribute to the heat transfer between the spray
and the gas flow, and thus a larger temperature drop is obtained. In
addition, the increase in droplet number and the droplet velocity will
intensify the droplet-wall impingement process and promote the
formation of the CVP, resultantly enhancing the droplet evaporation
in the droplet-wall impingement region (III) and the convective heat
transfer in the lower part of the flowfield.
To better understand the influence of spray parameters on the
temperature pattern, another nozzle with larger orifice that can
produce a large size of droplets under a low spray pressure, is
investigated. The temperature distributions on cross sections of the
flowfield under two crossflow velocity conditions with P 
case 3 (see Fig. 9a) and case 8 (see Fig. 10a), the crossflow conditions
and the spray pressure drop in both cases are same and the initial
droplet velocities are similar; hence, the influence of the spray drop-
let size on the temperature pattern can be discussed qualitatively. We
can find that, in the case with larger droplet size (case 8), compared
with the small droplet size case (case 3), an obvious increase in
 6 Article in Advance / ZHANG AND BAI
Downloaded by XI'AN JIAOTONG UNIVERSITY on August 6, 2016 | http://arc.aiaa.org | DOI: 10.2514/1.J054338

Fig. 9 Temperature distributions on cross sections under different conditions.

Fig. 10 Temperature distributions on cross sections for cases 8 and 9.

 Article in Advance
temperature drop is obtained over the whole cross section. The main
reason for this should be the increase in the evaporation rate with
increasing droplet size [see Eq. (1)]. In addition, as the droplet size
increases, the droplets can penetrate further into the mainstream and a
fiercer droplet-wall impingement will be induced due to the increased
droplet momentum. Hence, larger temperature drops are achieved in
these two regions due to the promoted droplet evaporation. In the
lower part of the section, a stronger CVP structure should also be
generated, and thus a larger temperature drop is observed due to the
enhanced convection heat transfer. Figure 10b shows the measured
results under a larger crossflow velocity (compare with the case in
Fig. 10a) condition. From Figs. 10a and 10b, we can find that, as was
observed in Figs. 9a and 9b, the crossflow velocity increases will
make the temperature drops in the mainstream region (IV) and
bottom CVP region (V) decrease obviously. This is because the
crossflow momentum increases with increasing velocity; as a result,
fewer droplets can disperse into the mainstream owing to the
decreased spray penetration depth; the droplet-wall impingement
Downloaded by XI'AN JIAOTONG UNIVERSITY on August 6, 2016 | http://arc.aiaa.org | DOI: 10.2514/1.J054338
becomes weaker, and the large momentum gas flow is not conducive
to the formation of the CVP.
The comparisons of the temperature drops along the central line of
the measured cross sections under different conditions are shown in
Fig. 11. As can be seen, from top to bottom, the temperature drop on
the central line shows a rapid increase first and peaks in the spray
region (II); after that, it decreases gradually in the mainstream region
(IV) while the area near the bottom region experiences a slight
increase due to the enhanced convection heat transfer caused by the
bottom CVP. The increases in the crossflow temperature and in the
liquid flow rate both contribute to the increase of temperature drop in
the spray region (II) (see Figs. 11b and 11d). It should be noticed that,
in the condition of large droplet size and low crossflow velocity, a
significant increase of temperature drop is observed near the bottom
of the section due to the strong CVP structure; as the vortices develop,
the convection heat transfer is promoted and along the four measured
cross-sections continued increase in the temperature-drop in this
region is thus obtained (see Fig. 11e).
B. Influences of Spray Injection Angle
It has been found that, for the liquid spray in crossflow, the spray
injection angle has a great influence on the spray and the droplet
dispersion [8,28,29]. Hence, it should also have great influence on
droplet evaporation. Therefore, the effects of the injection angle on
the temperature pattern are investigated in this section. The temper-
ature and droplet distributions on the measured cross sections for
 60 deg (the spray is inclined forward) and  120 deg (the
spray is inclined backward) are shown in Fig. 12. We can see that, for
both inclined spray cases, the five-region temperature pattern is also
observed on different cross sections. Compared with the case of
 90 deg (the case in Fig. 9c), the location of the spray region (II)
moves upward and the temperature drops in mainstream region (IV)
decrease remarkably; the location of the droplet-wall impingement
region (III) moves upward, and the temperature drops in this region
reduce obviously for  60 deg while increasing slightly for
 120 deg. In addition, in the case of  60 deg, the spray
region (II) becomes flatter and the area of the mainstream region (IV)
enlarges. The reason for these aforementioned phenomena is that,
compared with  90 deg, when the spray is injected into the
crossflow with an inclination angle, the locations of the initial spray
droplets move upward and the spray penetration depth reduces; thus,
the spray region (II) and the droplet-wall impingement region (III)
move upward and, in the mainstream region (IV), lower temperature
drops result due to the reduced concentration of droplets. For an
inclined forward spray (  60 deg), the relative velocity Urel
between the crossflow and the droplets increased; thus, the
evaporation rate of the droplets becomes larger and, as a result, a
weaker droplet-wall impingement is generated due to the fact that
fewer and smaller droplets will impinge on the wall after the
intensified vaporization in the spray region (II). In contrast, for the
inclined backward spray (  120 deg), an enhanced droplet-wall
impingement is induced due to the decreased droplet evaporation rate
/ ZHANG AND BAI 7
caused by the reduced relative velocity Urel between the crossflow
and the droplets. In addition, compared with the cases of  90 deg
and  120 deg, in the case of  60 deg, the upward-moved
spray region (II) and the enhanced droplet vaporization in this region
will make the area of the mainstream region (IV) enlarge and the
temperature drops in it decrease. Furthermore, for the cases of  60
and 120 deg, the downward velocity component of the droplets on
the leeward side of the spray and the upwind side of the spray
increase, respectively, which will intensify the interactions between
these droplets and the crossflow and contribute to the formation of the
bottom CVP; thus, the heat transfer near the bottom wall is enhanced.
The temperature distributions on the cross sections of the mixing
flowfield under different conditions with  60 and 120 deg are
shown in Fig. 13. Based on Figs. 12a and 13a1, we can find that, as
was discovered in the case of  90 deg, the increase in crossflow
temperature can enhance the droplet vaporization in the spray region
(II); thereby, the number and size of droplets in the mainstream region
(IV) reduce, and lower temperature drops in this region result (see
Figs. 12a, 13a1, and 13b1); the increase in crossflow velocity will
inhibit the droplet dispersion in the transverse direction and reduce
the spray penetration depth; thus, the spray region (II) shrinks and
temperature drops in the mainstream region (IV) decrease. In
addition, the droplet-wall impingement becomes weaker due to the
increased gas flow momentum (see Figs. 12a, 13a2, and 13b2).
C. Uniformity of Temperature-Drop Distribution on Cross Sections
For a good mixing of spray in crossflow, the heat transfer between
the droplets and the gas flow should be sufficient and a good
uniformity of temperature drops in the flowfield should be achieved
in a short mixing distance. In this section, the uniformities of the
temperature drops on the cross sections under different conditions are
analyzed based on the spatial uniformity index . Here, the spatial
uniformity index is defined based on all the temperature drops
measured on the cross section using the variance method:
Pn
S2
1n
T i T2
1 1 i1
(5)
T 2 T 2
with
T i 
T 0g i
T mix
g i (6)
1Xn
T  T i (7)
n 1
where T i is the temperature drop at the measurement point i; T
is the temperature drop average of the cross section; and n is the
number of measurement points on the cross section. When the
temperature drops are uniform on the cross section, T 1  T 2 
 T n  T, and  1. It follows that the greater the value of ,
the closer the temperature drop is to a uniform distribution.
The spatial uniformity indices of the temperature drop on the cross
sections under different conditions with  90 deg are given in
Fig. 14. As can be seen, the spatial uniformity index increases along
the mixing flowfield for all measured conditions, and the increase
indicates that the mixing is improved along with the convection and
conduction. The results in cases 1 and 2 indicate that, under the same
spray condition, a worse mixing would be obtained as a result of a
reduction in the crossflow temperature, which can lead to a more
intense spray-wall impaction and a larger area of droplet-wall
impingement region (III). The results in cases 25 show that the
mixing becomes worse as a result of an increase in the crossflow
velocity, which can produce a smaller area of the spray region (II) and
a weaker heat transfer in the bottom CVP region (V). The results in
cases 69 show that the mixing is promoted by an increase in the
liquid flow rate and the spray droplet size, which can increase the
concentration of droplets and droplet momentum, and thereby
enhance the heat transfer in the spray region (II) and the bottom CVP
region (V). A comparison of the spatial uniformity indices on the last
 8 Article in Advance / ZHANG AND BAI

Fig. 11 Temperature-drop distributions along central line of cross sections.

Downloaded by XI'AN JIAOTONG UNIVERSITY on August 6, 2016 | http://arc.aiaa.org | DOI: 10.2514/1.J054338

Fig. 12 Temperature and droplet distributions on different cross sections for
60 and 120 deg (case 1).
 Article in Advance
Downloaded by XI'AN JIAOTONG UNIVERSITY on August 6, 2016 | http://arc.aiaa.org | DOI: 10.2514/1.J054338
Fig. 13 Temperature distributions on cross sections with different spray injection angles under different conditions.
Fig. 14 Uniformity index of temperature-drop distributions on
different cross sections for
90 deg.
/ ZHANG AND BAI 9
measurement cross section (xD  5.39) for five injection angles
is shown in Fig. 15. We can see that, in the case of T g  573 K, the
mixing uniformity increases as the injection angle gets closer to
vertical injection (  90 deg), which can enhance the convection
heat transfer in the bottom CVP region (V). However, in the case of
T g  473 K, the maximum uniformity occurs when the spray is
inclined forward (  75 deg), which can increase the area of the
spray region (II) and weaken the spray-wall impaction. These results
lead to a conclusion that the mixing uniformity can be improved by
promoting the heat transfer in the spray region (II) and the bottom
CVP region (V), and they can become worse when an intense spray-
wall impingement occurs.
The underlying cause of the phenomena in Figs. 14 and 15 is
related to the spray penetration depth. For a low spray penetration
depth, the heat transfer in the lower part of the flowfield [bottom CVP
region (V)] is weak due to both the low droplet concentration and the
weak convection, and thus a poor mixing results. In contrast, for a
large spray penetration depth, a poor mixing will also be obtained, but
it will be due to the intense spray-wall impingement. Hence, in the
case of T g  473 K (e.g., case 1), a poor mixing is produced because
 10 Article in Advance
Downloaded by XI'AN JIAOTONG UNIVERSITY on August 6, 2016 | http://arc.aiaa.org | DOI: 10.2514/1.J054338
Fig. 15 Uniformity index of temperature-drop distributions for
different injection angles.
of the large spray penetration depth (the evaporation rate of droplets is
small), which can lead to an intense spray-wall impaction; and a
better mixing can be obtained by decreasing the spray injection angle
(forward), which can reduce the spray penetration depth, and thereby
weaken the spray-wall impaction, and enhance the heat transfer in the
spray region (II). In the case of T g  573 K (case 29), the spray
penetration depth becomes small due to the increased evaporation
rate of droplet; thus, the increase in droplet momentum and decrease
in crossflow velocity, as well as the injection angle getting closer to
vertical injection (  90 deg), can improve the mixing as a conse-
quence of an increase in the spray penetration depth, which can
promote the heat transfer in the bottom CVP region (V). Therefore,
we can conclude that, for the mixing of hollow-cone spray in high-
temperature crossflow, a good mixing can be achieved with a proper
spray penetration depth.
It is still difficult to determine the optimum penetration depth of a
hollow-cone spray mixing with a hot crossflow in a confined space, in
which the mixing is largely affected by droplet vaporization and
droplet-wall impingement. The optimum penetration depth will
undoubtedly vary with crossflow conditions and the geometries of
mixing space. Zhang et al. [29] previously proposed a criteria number
J, which could well evaluate the distribution pattern of droplets of
hollow-cone spray in crossflow in a rectangular duct at ambient
temperature. The expression of J is given by
d V 2In D32 M_l
J (8)
d g Ug D2
where M _ l is the mass flow rate of the liquid spray in kilograms per
second. J considers the initial state of the spray droplets and the liquid
flow rate, as well as the diameter of the mixing duct; and results show
that the droplet distributions are similar for approximate J values.
However, this criteria number does not consider the droplet
vaporization and the spray injection angle, which undoubtedly will
exert great influence on droplet dispersion and droplet-wall impinge-
ment in a hot mixing process. A good criteria number for the mixing
of hollow-cone spray in hot crossflow in a confined space should at
least include three important factors: 1) the evaporation rate of drop-
lets, which is related to the heat transfer number BT and the Nusselt
number N u ; 2) the droplet-wall impingement, which is related to the
Weber number We of the droplet immediately before impingement
and the wall temperature; and 3) the droplet dispersion, which is
related to the spray angle , the spray injection angle , the geometry
of the mixing space, and the droplet size and velocity. These three
factors are not independent, and they affect each other. More experi-
mental and theoretical work is still needed for a better understanding
of the mixing process.
Though insufficient data were obtained to statistically evaluate the
mixing between the hollow-cone spray and hot crossflow in a
/ ZHANG AND BAI
confined space, this study gives an insight into the complex mixing
process between hollow-cone spray and hot crossflow in a confined
space by providing the temperature pattern of the mixing flowfield.
The experimental results can give a good reference for further experi-
mental and theoretical studies, and it is anticipated to contribute to
developing numerical models of the mixing process.
IV. Conclusions
The heat transfer between the hollow-cone spray and the gas
crossflow in a rectangular duct is investigated experimentally. The
temperature pattern of crossflow caused by the evaporation of
the polydispersed spray droplets is obtained, and the underlying
formation mechanism is analyzed. The effects of different factors
on the temperature pattern are discussed. The uniformities of the
temperature drops on the measurement cross sections under different
conditions are evaluated, and a comparison among them is made. The
conclusions are as follows:
1) In a confined mixing space, the heat transfer between the
hollow-cone spray and crossflow is significantly affected by the walls
due to the droplet-wall impingement. The polydispersed hollow-cone
spray droplets and the droplet-wall impingement make the heat
transfer on the cross section of the flowfield inherently nonuniform.
2) According to the heat transfer mechanisms, the cross section of
the mixing flowfield can be divided into five regions, namely, the
corner region (I), the spray region (II), the droplet-wall impacting
region (III), the mainstream region (IV), and the bottom CVP region
(V). Larger temperature drops occur in the spray region (II) and
droplet-wall impingement region (III), followed by the mainstream
region (IV), and then the bottom CVP region (V) and corner region
(I). The bottom CVP contributes to the convection heat transfer in the
lower part of the flowfield. It can be promoted by increasing droplet
momentum, which can be weakened by increasing the crossflow
velocity and crossflow temperature.
3) The spray injection angle has significant influence on the
temperature pattern of crossflow. As the spray injection angle is
decreased (spray forward), the heat transfer is enhanced in the spray
region (II) but weakened in the droplet-wall impingement region (III)
and bottom CVP region (V). As the spray injection angle is increased
(spray backward), the heat transfer is promoted in thje droplet-wall
impacting region (III) but weakened in the bottom CVP region (V).
4) The spray penetration depth is a determining factor for the
mixing between hollow-cone spray and high-temperature crossflow.
For the case of weak spray penetration depth, the mixing can be
improved with the increase in droplet momentum and decrease in
crossflow velocity, as well as the injection angle getting closer to
vertical injection (  90 deg ), as a result of an enhancement in the
convection (heat transfer) in the bottom CVP region (V) due to the
increased spray penetration depth. For the case of large spray
momentum where intense spray-wall impaction occurs, the
uniformity can be improved by decreasing the spray injection angle
(spray forward) as a consequence of the weakened spray-wall
impaction.
Acknowledgments
Financial support from the National Nature Science Foundation of
China under contract no. 51306146 and the China National Founds
for Distinguished Young Scientists under contract no. 51425603 are
highly appreciated.
References
[1] Tian, W. P., Cai, T. M., Lu, H. J., and Gao, P., Thermodynamic
Calculation for Water Ramjet, Journal of Solid Rocket Technology,
Vol. 29, No. 2, 2006, pp. 9598.
[2] Yang, Y. J., He, M. G., and Xu, H. D., Thermodynamic Cycle
Performance for Water Ramjet Motors, Journal of Propulsion
Technology, Vol. 30, No. 4, 2009, pp. 474478.
[3] Wu, P. K., Kirkendall, K. A., Fuller, R. P., and Nejad, A. S., Breakup
Processes of Liquid Jets in Subsonic Crossflows, Journal of Propulsion
and Power, Vol. 13, No. 1, 1997, pp. 6473.
doi:10.2514/2.5151
 Article in Advance
[4] Song, J., Cain, C. C., and Lee, J. G., Liquid Jets in Subsonic Air
Crossflow at Elevated Pressure, Journal of Engineering for Gas
Turbines and Power, Vol. 137, No. 4, 2015, Paper 041502.
doi:10.1115/1.4028565
[5] Wang, X. H., Huang, Y., Wang, S. L., and Zhi, L. L., Bag Breakup of
Turbulent Liquid Jets in Crossflows, AIAA Journal, Vol. 50, No. 6,
2012, pp. 13601366.
doi:10.2514/1.J051451
[6] Tambe, S. B., Liquid Jets in Subsonic Crossflow, M.S. Thesis, Univ. of
Cincinnati, Cincinnati, OH, 2004.
[7] Sallam, K. A., Ng, C. L., Sankarakrishnan, R., Aalburg, C., and Lee, K.,
Breakup of Turbulent and Non-Turbulent Liquid Jets in Gaseous
Crossflows, AIAA Paper 2006-1517, 2006.
[8] Almeida, H., Sousa, J. M. M., and Costa, M., Effect of the Liquid
Injection Angle on the Atomization of Liquid Jets in Subsonic
Crossflows, Atomization and Sprays, Vol. 24, No. 1, 2014, pp. 8196.
doi:10.1615/AtomizSpr.2013008310
[9] Costa, M., Melo, M. J., Sousa, J. M. M., and Levy, Y., Spray
Characteristics of Angled Liquid Injection into Subsonic Crossflows,
AIAA Journal, Vol. 44, No. 3, 2006, pp. 646653.
Downloaded by XI'AN JIAOTONG UNIVERSITY on August 6, 2016 | http://arc.aiaa.org | DOI: 10.2514/1.J054338
doi:10.2514/1.10887
[10] Lozano, A., Barreras, F., Hauke, G., and Dopazo, C., Longitudinal
Instabilities in an Air-Blasted Liquid Sheet, Journal of Fluid
Mechanics, Vol. 437, June 2001, pp. 143173.
doi:10.1017/S0022112001004268
[11] Liu, K., and Sun, L. Y., Experimental Study of Atomization
Characteristic of Air Blast Atomizer, Applied Mechanics and
Materials, Vols. 300301, Feb. 2013, pp. 185188.
doi:10.4028/www.scientific.net/AMM.300-301.185
[12] Leong, M. Y., Mixing of An Airblast-Atomized Fuel Spray Injected
into a Crossflow of Air, NASA CR-210467. 2000.
[13] Lorenzetto, G. E., and Lefebvre, A. H., Measurements of Drop Size on
a Plain-Jet Airblast Atomizer, AIAA Journal, Vol. 15, No. 7, 1977,
pp. 10061010.
doi:10.2514/3.60742
[14] Eroglu, H., and Chigier, N., Initial Drop Size and Velocity Distributions
for Airblast Coaxial Atomizers, Journal of Fluids Engineering,
Vol. 113, No. 3, 1991, pp. 453459.
doi:10.1115/1.2909517
[15] Miller, P. C. H., Butler Ellis, M. C., and Tuck, C. R., Entrained Air and
Droplet Velocities Produced by Agricultural Flat-Fan Nozzles,
Atomization and Sprays, Vol. 6, No. 6, 1996, pp. 693707.
doi:10.1615/AtomizSpr.v6.i6.40
[16] Phillips, J. C., Miller, P. C. H., and Thomas, N. H., Air Flow and Droplet
Motions Produced by the Interaction of Flat-Fan Spray and Cross
Flows, Atomization and Sprays, Vol. 10, No. 1, 2000, pp. 83103.
doi:10.1615/AtomizSpr.v10.i1.40
[17] Ghaffara, Z. A., Hamidb, A. H. A., Syazwan, M., and Rashidc, F. M., A
Review on Spray Characteristics of Effervescent Atomizer Under
Various Geometrical Parameters and Operating Condition, AIP
Conference Proceedings, Vol. 1440, June 2012, pp. 586590.
doi:10.1063/1.4704266
[18] Huang, X., Wang, X., and Liao, G., Visualization of Two Phase Flow
Inside an Effervescent Atomizer, Journal of Visualization, Vol. 11,
No. 4, 2008, pp. 299308.
[19] Sallam, K. A., Aalburg, C., Faeth, G. M., Lin, K. C., Carter, C. D., and
Jackson, T. A., Primary Breakup of Round Aerated-Liquid Jets in
Supersonic Crossflows, Atomization and Sprays, Vol. 16, No. 6, 2006,
pp. 657672.
doi:10.1615/AtomizSpr.v16.i6.40
[20] Miller, B., Sallam, K. A., Bingabr, M., Lin, K. C., and Carter, C. D.,
Breakup of Aerated-Liquid Jets in Subsonic Crossflows, Journal of
View publication stats
/ ZHANG AND BAI 11
Propulsion and Power, Vol. 24, No. 2, 2008, pp. 253258.
doi:10.2514/1.30390
[21] Ghosh, S., and Hunt, J. C. R., Spray Jets in a Cross-Flow, Journal of
Fluid Mechanics, Vol. 365, June 1998, pp. 109136.
doi:10.1017/S0022112098001190
[22] Lynch, A., Batchelor, R. G., Kiel, B., Miller, J., Gord, J., and Reeder, M.,
Spray Characteristics of a Pressure-Swirl Fuel Injector Subjected to a
Crossflow and a Coflow, Atomization and Sprays, Vol. 21, No. 8, 2011,
pp. 625643.
doi:10.1615/AtomizSpr.2012004012
[23] Prakash, R. S., Gadgil, H., and Raghunandan, B. N., Breakup
Processes of Pressure Swirl Spray in Gaseous Cross-Flow,
International Journal of Multiphase Flow, Vol. 66, Nov. 2014,
pp. 7991.
doi:10.1016/j.ijmultiphaseflow.2014.07.002
[24] Deshpande, S. S., Gao, J., and Trujillo, M. F., Characteristics of Hollow
Cone Sprays in Crossflow, Atomization and Sprays, Vol. 21, No. 4,
2011, pp. 349361.
doi:10.1615/AtomizSpr.2011003586
[25] Lee, S., Kim, W., and Yoon, W., Spray Formation by a Swirl Spray Jet
in Low Speed Cross-Flow, Journal of Mechanical Science and
Technology, Vol. 24, No. 2, 2010, pp. 559568.
doi:10.1007/s12206-009-1222-6
[26] Gao, J., Trujillo, M. F., and Deshpande, S. S., Numerical Simulation of
Hollow-Cone Sprays Interacting with Uniform Crossflow for Gasoline
Direct Injection Engines, SAE International Journal of Engines, Vol. 4,
No. 2, 2011, pp. 22072221.
doi:10.4271/2011-24-0007
[27] Bai, B. F., Zhang, H. B., Liu, L., and Sun, H. J., Experimental Study on
Turbulent Mixing of Spray Droplets in Crossflow, Experimental
Thermal and Fluid Science, Vol. 33, No. 6, 2009, pp. 10121020.
doi:10.1016/j.expthermflusci.2009.05.002
[28] Zhang, H. B., Bai, B. F., Liu, L., Sun, H. J., and Yan, J. J., Droplet
Dispersion Characteristics of the Hollow Cone Sprays in Crossflow,
Experimental Thermal and Fluid Science, Vol. 45, No. 1, 2013,
pp. 2533.
doi:10.1016/j.expthermflusci.2012.09.021
[29] Zhang, H. B., Bai, B. F., Liu, L., Sun, H. J., and Yan, J. J., Mixing of
Hollow-Cone Spray with a Confined Crossflow in Rectangular Duct,
AIAA Journal, Vol. 51, No. 3, 2013, pp. 615622.
doi:10.2514/1.J051925
[30] Zhang, H. B., Bai, B. F., Liu, L., Sun, H. J., and Yan, J. J., Experimental
Study of the Mixing of Two Impinging Pressure-Swirl Sprays in
Crossflow, Experimental Thermal and Fluid Science, Vol. 49,
Sept. 2013, pp. 6774.
doi:10.1016/j.expthermflusci.2013.04.002
[31] Senda, J., Kobayashi, M., Iwashita, S., and Fujimoto, H., Modeling of
Diesel Spray Impinging on Flat Wall, JSME International Journal,
Series B: Fluids and Thermal Engineering, Vol. 29, No. 4, 1996,
pp. 859866.
doi:10.1299/jsmeb.39.859
[32] Sazhin, S. S., Advanced Models of Fuel Droplet Heating and
Evaporation, Progress in Energy and Combustion Science, Vol. 32,
No. 2, 2006, pp. 162214.
doi:10.1016/j.pecs.2005.11.001
[33] Sun, H., Bai, B., and Zhang, H., Comparative Investigation on Droplet
Evaporation Models for Modeling Spray in Cross-Flow, Heat Transfer
Engineering, Vol. 35, Nos. 68, 2014, pp. 664673.
doi:10.1080/01457632.2013.837707
J. P. Drummond
Associate Editor