Seediscussions,stats,andauthorprofilesforthispublicationat:http://www.researchgate.net/publication/256718687

AnexperimentalstudyonthesprayandthermalcharacteristicsofR134atwo-phaseflashingspray

ARTICLEinINTERNATIONALJOURNALOFHEATANDMASSTRANSFER·JULY2012

ImpactFactor:2.38·DOI:10.1016/j.ijheatmasstransfer.2012.04.021

CITATIONS

7

READS

28

5AUTHORS,INCLUDING:

Wei-TaoWu

CarnegieMellonUniversity

12PUBLICATIONS21CITATIONS

SEEPROFILE

BinChen

Xi'anJiaotongUniversity

40PUBLICATIONS102CITATIONS

SEEPROFILE

Guo-XiangWang

UniversityofAkron

94PUBLICATIONS789CITATIONS

SEEPROFILE

LiejinGuo

Xi'anJiaotongUniversity

322PUBLICATIONS6,633CITATIONS

SEEPROFILE

Availablefrom:BinChen

Retrievedon:23October2015

International Journal of Heat and Mass Transfer 55 (2012) 4460–4468

Contents lists available at SciVerse ScienceDirect

International Journal of Heat and Mass Transfer

journal homepage: www.elsevier .com/locate / i jhmt

An experimental study on the spray and thermal characteristicsof R134a two-phase flashing spray

Zhou Zhifu a, Wu Weitao a, Chen Bin a,⇑, Wang Guoxiang a,b,⇑, Guo Liejin a

a State Key Laboratory of Multiphase Flow in Power Engineering, Xi’an Jiaotong University, Xi’an 710049, Chinab Department of Mechanical Engineering, University of Akron, Akron, OH 44325-3903, USA

a r t i c l e i n f o

Article history:Received 31 October 2011Received in revised form 10 April 2012Accepted 10 April 2012Available online 9 May 2012

Keywords:Flashing sprayR134aDroplet dynamicsThermal characteristics

0017-9310/$ - see front matter � 2012 Elsevier Ltd. Ahttp://dx.doi.org/10.1016/j.ijheatmasstransfer.2012.04

⇑ Corresponding authors at: State Key LaboratoryEngineering, Xi’an Jiaotong University, Xi’an 710049, Cfax: +86 29 82669033 (C. Bin), tel.: +1 330 972 738Guoxiang).

E-mail addresses: chenbin@mail.xjtu.edu.cn (C(W. Guoxiang).

a b s t r a c t

Flashing spray of volatile liquids is a common phenomenon observed in many industrial applicationssuch as fuel injection of engines, accidental release of flammable and toxic pressure-liquefied gases, fail-ure of a vessel or pipe in the form of a small hole in chemical industry, and cryogenic spray cooling inlaser dermatology, etc. In flashing spray, the volatile liquid is depressurized rapidly at the exit of a nozzle(or a hole in a vessel) and becomes superheated. Such superheated liquid (in the form of either a jet ordroplets) will lead to explosive atomization with fine droplet and a short spray distance. This paper pre-sents an experimental investigation to the spray and thermal characteristics of flashing spray using cryo-gen R134a. A photographic study of the spray is firstly conducted to visualize the spray formation andthe dynamic characteristics of the spray. Afterwards, the spray characteristics are measured by the phaseDoppler Particle Analyzer (PDPA). The distributions of the diameter reveals the dramatic dynamic vari-ation of the liquid droplets due to explosive atomization of large droplets in the region near the exit ofnozzle, while the self-similar velocity profiles are fitted by two empirical correlations to describe thenon-dimensional axial and radial velocities, respectively. The temperature field within the spray is mea-sured by a small thermocouple. The temperature measurements provide detailed quantitative informa-tion of both radial and axial temperature distributions of droplets within the spray. These experimentalresults provide deep understanding into the whole characteristics of two-phase flashing spray of volatileliquids.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Flashing spray occurs when a high-pressure liquid is injectedinto low pressure environment to make the liquid superheated,characterized by explosive atomization of superheated liquid togenerate fine droplets and accompanied by strong evaporation ofthese droplets, leading to extremely low droplets temperature.Flashing spray finds many industrial applications, for example,flash-spray internal combustion engine [1], distillation of saltwater by flashing evaporation [2], cryogen spray cooling in lasertreatment of dermatology [3,4], etc. The explosive flashing spraymay also take place during the accidental release of flammableand toxic pressure-liquefied gases in chemical or nuclear industry,when failure of a vessel or pipe in the form of a small hole will re-sult in formation of a flashing jet containing a mixture of liquiddroplets and vapor [5]. Flashing spray is also relevant for the

ll rights reserved..021

of Multiphase Flow in Powerhina. Tel.: +86 29 82667326;

9; fax: +1 330 972 6027 (W.

. Bin), gwang@uakron.edu

aerospace engineering, where high-pressure fluids expand intonear vacuum during engine start-up will lead to high superheatstate as well as flash atomization and vaporization [6].

With a low boiling point (�26.1 �C at the atmospheric pressure)and high volatility, R134a has been widely used for flashing sprayin many industrial applications as a non-toxic and ozone-friendlyrefrigerant. One important case is the cryogen spray cooling thatis successfully used in laser dermatology to prevent burning injuryof skin during surgery [3,4]. Saturated R134a at room temperaturein a storage tank is injected into the atmospheric environmentthrough a special designed nozzle. Flashing atomization and strongevaporation will result in a low temperature spray which providesefficient cooling to skin [7–10]. Spray cooling with R134a also hasbeen used in the metal foundries, cooling of microelectronics andchiller in air-conditioning systems to remove high heat fluxes[11–13]. In the field of industrial safety, Yildiz et al. [14,15] usedR134a as model fluid to simulate the accidental release of pres-sured liquid.

The importance of these applications has motivated research onthe flashing spray. Brown and York [16] firstly used the photographtechnology to investigate the flashing spray pattern using water.After that, several other visualization studies were also performed

Nomenclature

d inner diameter of the nozzle (mm)L length of the nozzle (mm)z spray axial distance from the nozzle exit (mm)Z non-dimensional axial distanceR spray radial distance from the centerline (mm)D10 droplet arithmetical mean diameter (lm)D32 droplet Sauter mean diameter (lm)U axial velocity (m/s)V radial velocity (m/s)T droplet temperature (�C)

Subscriptsmax maximummin minimums saturatedMTD minimum temperature distanceCSD Cold Spray DistanceSTL Spray Thermal LengthSTW spray thermal widthSCW spray cold width

Z. Zhifu et al. / International Journal of Heat and Mass Transfer 55 (2012) 4460–4468 4461

using water. Miyatake et al. [17] and Peter et al. [18] studied theeffect of the liquid superheat and the flow rate on the flashingspray. Reitz [19] and Park and Lee [20] investigated the mecha-nisms of the flashing sprays. Additionally, Allen [21,22] used theMalvern technique and LDV system to measure the droplet diam-eter and the axial velocity distributions of two-phase flashing pro-pane jets. Yildiz et al. [14,15] and Aguilar et al. [23,24] conductedexperiments to investigate the two-phase flashing R134a jets usingthe PDPA system.

There are only a few reports on the thermal characteristics ofR134a sprays [23–25]. Aguilar and his co-workers [23,24] foundan exponential decay of the average temperature of steady stateR134a droplets along the spray distance by using a large thermo-couple with a bead diameter of approximately 0.3 mm. Yildiz etal. [25] also used thermocouples to measure the centerline dropletstemperature of R134a sprays from nozzles of 1–3 mm. Their datashows similar variation as reported by Aguilar et al. [23,24]. Miya-take et al. [17] proposed an empirical correlation for the variationof spray temperature for superheated water spray. In a review arti-cle, Polanco et al. [5] mentioned a minimum spray temperaturealong the centerline of the flashing spray and defined a ‘‘minimumtemperature distance’’. They reported a dimensionless minimumtemperature distance of 150–170 (normalized by the nozzle diam-eters) for propane spray, similar to that by Yildiz et al. [25] forR134a spray.

Although above studies provided some useful information onthe flashing spray, the results have not given a comprehensiveinvestigation on the R134a two-phase flashing spray. It is apparentthat more accurate experiments are required to provide a better in-sight into the mechanisms of the R134a flashing spray. The motiva-tion of present work is aimed at conducting a full experimentalstudy on both spray and thermal characteristics of the R134atwo-phase flashing spray. The flashing spray pattern and the distri-bution of droplet diameter, velocity and temperature have beeninvestigated by high-speed camera, PDPA and thermocouple.Empirical correlations based on the experiment results have beenproposed to describe the droplets spray and thermal behaviors.These data should shed important light to the spray developmentand would be useful for future spray design.

2. Experiment system and procedures

2.1. Spray system

Fig. 1a shows a schematic of the experimental system for flash-ing spray study. The system consists of a commercial pressure vesselfor storage of R134a (Dupont), a three-dimensional translationalelectric positioner (WN105TA300M by Beijing Winner OpticsInstruments Co., China) with resolution of 8 lm, a solenoid electricvalve (B2021SBTTO24DVC by Gems, US) which can open or close

within 5 ms, and a specially-designed nozzle installed on the posi-tioner. The nozzle is made of a stainless steel tube of length of63.5 mm and inner diameter of 0.81 mm, which resembles that ofcommercial nozzles used for cryogen spray cooling in conjunctionwith dermatological laser treatments. A standard high-pressurehose connects the cryogen vessel to the valve, while the nozzlewas fit tightly into the opening of the solenoid valve. The internalstructure of the solenoid valve includes four 90� bends and two sud-den contraction sections, as shown in Fig. 1b. A micro-scale flowme-ter (931-06xx by Gems, US) is located in the middle of the high-pressure hose to monitor the flow rate of the spray. The experimen-tal rig and PDPA system were shown in Fig. 1c and d.

2.2. Imaging system

A high-speed video camera (MotionXtra HG-100, US) with shut-ter speed of 997 ls is used to take photographs of the spray. A PLS-SXE300 Xe lamp with high power provides illumination for thehigh-speed camera at far distance for large view, while a white lampwith low power near the nozzle for close view. The camera and thelamps are positioned in the same horizontal plane, with the cameraviewing perpendicularly to the spray axis. All photos are taken atthe speed of 1000 fps and the same resolution of 1504 � 1128 pix-els. The camera is placed either 1900 or 500 mm from the spray axis.At a far distance, the camera can catch the view of the entire spray;while at a close distance, the camera can take a photo of the spraynear the nozzle exit. The two distances give the fields of view ofabout 190 � 140 and 50 � 37.5 mm2, respectively.

2.3. PDPA measurement system

A Phase Doppler Particle Analyzer (PDPA by TSI, US) is used tosimultaneously measure the velocity and diameter distribution ofthe droplets in the spray of R134a. The PDPA generates four inter-facing laser beams with a power of 0.8 W of two wavelengths,514.5 nm (channel one) and 488 nm (channel two). Two beams fo-cus on a probe volume, typically smaller than 1 mm3. When drop-lets go through the probe volume, these beams are interfaced and aDoppler signal with a frequency shift proportional to the dropletvelocity is generated. The diameter and axial and the radial veloc-ity of droplets can be measured simultaneously. The phase differ-ence between the signals collected by adjacent detectors isproportional to the droplet diameter. Before taking the measure-ments, the optimum values of the PDPA parameters have to be se-lected including the diameter range (0–1000 lm), velocity range(0–100 m/s) and the laser power. For each measurement, the sprayduration lasts 10 s. Repeatability tests have been conducted toevaluate the confidence on the experimental results. It has beenshown that the dispersion did not exceed ±3% in the droplet meanvelocities (U and V) and ±7% in the droplet mean diameters (3% in

P-1

R134aContainer

Valve

Flowmeter

ElectricValve

ElectricPositioner

Nozzle

DAQ

P

CCD

PDPA

zR

Centerline

(a)

Unit:mm

aaaaaR134a

Nozzle

Valve

0.81

3.088888

1.58

63.5

(b)

(c) (d)

Fig. 1. (a) Schematic of the experimental system, (b) Schematic of the internal structure of the solenoid valve, (c) photograph of experimental test rig, (d) PDPA system.

4462 Z. Zhifu et al. / International Journal of Heat and Mass Transfer 55 (2012) 4460–4468

arithmetic mean diameter D10 and 7% in Sauter mean diameterD32). Here D10 presents the arithmetic mean diameter of dropletsas calculated on the diameters from the total number of dropsand D32 denotes the diameter of that drop whose volume/surfaceratio value is the same as the arithmetic mean of volume/surfacevalues on the total number of drops in the spray.

2.4. Droplets temperature measurement

A small thermocouple probe has been developed in this study tosystematically measure the temperature variation of liquid drop-lets in the R134a spray. The thermocouple probe is made of a stan-dard T-type thermocouple (Omega, US) with the bead diameter ofabout 100 lm and is placed into the spray supported by a fine rigidstick. Its ASTM standard wire error is less than ±1 �C for the mea-surement range, while the signal is acquired at 100 kHz and con-verted to the temperature data using a DAQ board (NI: M-6251,US) and Labview control soft system. The relative space betweenthe thermocouple bead and the nozzle tip is controlled by thethree-dimensional electric positioner. Only one thermocoupleprobe is used during each measurement to avoid the disturbanceof the thermocouple probe to the spray. All experimental condi-tions were kept as steady as possible so a steady state spray canbe maintained during the entire time period of measurement. Foreach location, the measurement time is 2 s to ensure the steadymeasurement of the thermocouple.

2.5. Experiment procedures

Experiments have been carried out successively to record pho-tographs by high-speed camera and measure the diameter, velocityand temperature of droplets by PDPA and thermocouple,respectively.

PDPA measurements are firstly performed on the spray center-line with the axial interval (Dz) of 10 mm from the nozzle exit tothe axial distance of 200 mm. Then the PDPA measurements are ta-ken at four different cross-sections (z = 50, 90, 130 and 170 mm)with the radial interval (DR) of 1 mm or 2 mm from the center tothe periphery of the spray as shown in Fig. 1a.

Similarly with the PDPA measurements, the temperature mea-surements are also firstly performed on the spray centerline withthe axial interval (Dz) of 10 mm from the nozzle exit to the dis-tance where the droplets temperature is high than its saturationpoint at the atmospheric pressure. Then the droplets temperatureat eleven cross-sections (z = 5,10,20, . . . ,100 mm) is measuredfrom the center to the periphery of the spray until the temperaturereached its saturation point with the radial interval (DR) of0.5 mm.

In all the experiments, the room temperature is kept at about25 �C, and the relative humidity is about 20%. The R134a cryogencontained in the vessel is kept at the saturation state. For steadyspray of R134a, the volume flow rate of liquid R134a is3.15 � 10�6 m3/s from the micro-scale flowmeter experimental

Z. Zhifu et al. / International Journal of Heat and Mass Transfer 55 (2012) 4460–4468 4463

result. The relative error is less than 2% with our calibration usingwater.

3. Experimental results and discussion

3.1. High speed camera observations

Fig. 2a shows a typical photo of the R134a spray from the high-speed camera (1000 fps) with a fairly strong white light, whichillustrates the typical shape of the R134a spray. In the photo, thecentral white region is resulted by strong scattering of light fromdense droplets, while the light-colored periphery is due to smalland dilute droplets and/or the R134a vapor. In general, the sprayis characterized by a dense, jet-like region near the nozzle exit, fol-lowed by an unsteady spray that may last from 80 to 200 mm. Thespray becomes unstable after 60 mm, indicating an increasedentrainment of air. From Fig. 2a, one can see that there is a largeradial expansion of the spray at the nozzle exit, but the radial sizequickly becomes relative stable interface between the spray andthe surrounding air.

The expansion and quick formation of the spray after nozzleexit can be seen more clearly in Fig. 2b, which shows a close viewof the spray near the exit of the nozzle. In this figure, more detailedfeature of the spray can be observed. First of all, a white core regiondue to strong scattering of the light within the spray near the exitof the nozzle suggests dense droplets in the area and this regiondiminishes quickly within 25 mm. Compared to the small size ofthe nozzle, the spray near the nozzle exist shows significant expan-sion in the radial direction, giving a spray angle larger than 90�. It isfound that the present wide spray angle spray is very differentfrom those of non-volatile liquids such as water [19,20]. In the caseof R134a, the liquid is at a saturated state in the tank with a highpressure (0.7 MPa) at the room temperature. When it flows outof the valve, the liquid experience a lower pressure and becomessuperheated. Bubbles are then quickly formed within the long tubeof the nozzle and a two-phase mixture instead of single phase

Fig. 2. Photos of steady state spray of R134a, (a) overview and (b) close view nearthe nozzle.

liquid flows out of the nozzle [26]. The two-phase flow in the tubeof the nozzle should be at a pressure higher than the ambient pres-sure. When the high-pressure bubble-containing liquid exits thenozzle, the liquid is quickly atomized and experienced significantexpansion in the radial direction. Such a flashing-like atomizationis typical for volatile liquids like R134a.

3.2. Spray characteristics by PDPA measurements

3.2.1. Droplets density by PDPA count numberFirstly, we examine the variations of the droplets number mea-

sured by PDPA along the spray axial distance (z) and radial distance(R). As shown in Fig. 3a, the centerline droplets number is no morethan 1000 near the exit of the nozzle (z < 30 mm), which indicatesthat there exists very dense droplets or incompletely broken liquidligaments near the nozzle exit. A sudden increase of the dropletsnumber is observed near z = 50 mm and the maximum value(about 35,000) emerges at z = 100 mm. In this region, the spray be-comes more atomized resulting in the increment of the dropletsnumber as the spray distance increases. After reaching the maxi-mum value, the droplets number reduces with the increasementof the spray distance due to the complete evaporation of the smalldroplets. Overall, the variation of centerline droplets number canbe correlated by a Gaussian function as following,

N ¼ 953:66þ 2:33� 106

54:62ffiffiffiffiffiffiffiffiffip=2

p exp �2z� 103:84

54:62

� �2 !

ð1Þ

where N denotes the droplets number measured by PDPA in 10 sspurt duration and z is the spray axial distance.

Fig. 3b illustrates the radial variations of the droplets number atfour spray distances, z = 50, 90, 130 and 170 mm. It is noticed thatthe radial variation of the droplets number at z = 50 mm shows amaximum value at about R = 5 mm, which indicates again that adense cloud of droplets exists in the central region near the nozzleexit. The radial reduction after the maximum shows similar trendas that in the case of centerline variation, i.e., the reduction of thedroplets number reflects the decrease in the droplet density alongthe radial direction. At z = 90 mm, the maximum point shifts toR = 1 mm. However, for the two cases of z = 130 and 170 mm, themaximum points locate at the centerline R = 0 mm. In all threecases, one can find a continuous reduction of the droplets numberoutward, as expected for a spray of a volatile fluid. Fig. 3b also im-plies that the spray becomes fairly uniform in droplet density along

0 40 80 120 160 2000

10

20

30

40

50(b)

Cou

nt n

umbe

r, N(× 1

03 )

z (mm)

Fitting curve

(a)

0 5 10 15 20 25 30

R (mm)

Axial distance 50 mm 90 mm 130 mm 170 mm

Fig. 3. Variation of the droplets number, (a) along the spray central axis and (b)along the radial distance at four spray axial distances.

4464 Z. Zhifu et al. / International Journal of Heat and Mass Transfer 55 (2012) 4460–4468

the radial directions at z = 170 mm where the number changes lit-tle with R.

3.2.2. Droplets diameter distributionTo ensure the validity of the statistic analysis, such as the Sauter

mean diameter (D32) and the mean velocities (U and V), we onlytake into account the points where the measured droplet numberis more than 1000 and the distributions follow the normaldistribution.

Fig. 4 presents the variations of the D32 along the axial distance(z) and radial distance (R). As shown in Fig. 4a, D32 shows a quickreduction within the first 50 mm along the centerline. One expla-nation of such a quick reduction in the droplet size is related tothe atomization mechanism of superheated cryogen out of the noz-zle. The high temperature liquid droplets out of the nozzle findthemselves in a superheated state and would be explosively atom-ized quickly into smaller droplets to reduce the non-equilibriumdriving force. This may explain why the droplets in R134a sprayare fairly small. In the region of the spray distances from 50 to130 mm, there is only very small change in D32, which remainsas 12 lm. This suggests that when the spray distance reaches50 mm, the droplets reach a quite uniform diameter distribution.A slightly increase of D32 in the later region can be observed dueto the complete evaporation of the very small droplets. The impactof the complete evaporation of these small droplets on the reduc-tion of total surface area is greater than that of the total volume,which leads to the slightly increasement of D32 since it is the ratioof the total volume to the total surface area of all the droplets.

Fig. 4b shows the radial variations of D32 at four spray dis-tances: z = 50, 90, 130, and 170 mm. It can be seen from the figurethat D32 changes little along the radial distance on the spray cross-sections near the nozzle of z = 50 and 90 mm. As increasing thespray distance to 130 and 170 mm, D32 also changes little in thecentral region of the spray cross-sections while the slightlyincreasement of D32 attributes to the measurement uncertaintyas mentioned in Section 2.3. At the periphery, D32 has a quickreduction along the radial distance due to the strong evaporationof the volatile droplets.

It is noticed that the data of droplet diameters in Fig. 4 is com-parable with the experimental data of Aguilar et al. [23], but ismuch smaller than that of R134a spray by Yildiz et al. [15]. Wecarefully examine those experimental conditions and find thatthe main reason is the different states of liquid R134a kept in thecontainer. In Aguilar’s and our experiments, the liquid R134a aresaturated and the bubbles were formed inside the nozzle when

0 50 100 150 2005

10

15

20

25Centerline

D32

(µm

)

z (mm)

(a)

0 5 10 15 20 25 30

(b) Axial distance 50 mm 90 mm 130 mm 170 mm

R (mm)

Fig. 4. Variation of the droplets Sauter mean diameter (D32), (a) along the praycentral axis and (b) along the radial distance at four spray axial distances.

the liquid R134a are released from the nozzle, which will lead tothe small droplets outside the nozzle. However, the liquid R134ais subcooled by increasing the pressure of the container whilekeeping the temperature constant in Yildiz’s experiment, wherethe bubbles were formed outside of the nozzle.

3.2.3. Droplets velocity measurement resultsThe axial and radial mean velocities (U and V) along the center-

line are given in Fig. 5 as a function of spray axial distance (z). Asexpected, the axial velocity is significantly larger than the radialvelocity at all spray distances. Overall, the radial velocity of thedroplets is relatively small (less than 10 m/s), which indicate asmall diverging spray at the radial direction compared with thatalong the axial direction as observed in Fig. 2a. It is interesting tosee in Fig. 5 that the spray shows a maximum droplet axial velocityat the spray distance of 40 mm. The velocities of atomized dropletsquickly increase to a maximum value of nearly 55 m/s, and thengradually decrease along the spray direction due to the continuousdrag force acting on the droplets. The acceleration of droplets nearthe nozzle exit can be explained by explosive atomization of super-heated liquid in the spray. The unbroken liquid with high pressurein the center region propels the atomized droplets outward. As thedroplets fly further away, the spray becomes better atomized andthe unbroken liquid disappears at the axial distance of 40 mm(as can be seen in Fig. 4) where the maximum axial velocity exists.Then the drag force begins to dominate the droplets motion. Theacceleration of the droplets near the nozzle exit was also observedby Vu et al. [24] and Yildiz et al. [15] in the R134a flashing sprayand by Allen [21] in the propane flashing spray. Vu et al. [24] sug-gest that the acceleration of droplets near the nozzle exit is attrib-uted to the high velocity difference between the liquid and gasphases. However, Reitz’s photographic study of the water flashingspray shows that the acceleration is caused by the expelling fromthe unbroken liquid jet [19], which is consistent with ourexplanation.

Fig. 6 presents the variations of droplets axial velocity and ra-dial mean velocity as a function of the radial distance at four spraydistances: z = 50, 90, 130 and 170 mm. From this figure, one cansee that the largest axial and radial velocities locate at the centerof the spray independent of the spray distance. As the radial dis-tance increases, the velocities begin to reduce. However, thespeeds of the decrease in the velocities are different for differentspray distances. The smaller the spray distance is, the faster thevelocities will decrease along the radial direction. In other words,the droplets velocities become more uniform as the spray sectionis far away from the exit of the nozzle.

0 40 80 120 160 2000

10

20

30

40

50

60 Axial Velocity Radial Velocity

Vel

ocity

(m

/s)

Axial distance, z (mm)

Fig. 5. Variation of axial and radial mean velocities of centerline droplets along thespray distance.

0 5 10 15 20 25 300

2

4

6 (b)

U(m

/s)

Radial distance, R (mm)

0

10

20

30

40

50

(a)Axial distance

50 mm 90 mm 130 mm 170 mm

V(m

/s)

Fig. 6. Variations of droplet velocities, (a) axial velocity and (b) radial velocity alongthe radial direction at four spray distances.

Z. Zhifu et al. / International Journal of Heat and Mass Transfer 55 (2012) 4460–4468 4465

Fig. 7 shows the variations of the non-dimensional velocitiesnormalized by their maximum value as a function of the non-dimensional radial distance normalized by the spray distance. Asshown clearly in this figure, the non-dimensional velocities showself-similar profile and follow the Gaussian distribution, whichcan be correlated by two Gaussian functions as given in Eqs. (2)and (3), where Umax and Vmax are the maximum axial velocityand radial velocity on each spray cross-section, R and z are radialdistance and spray axial distance, respectively.

UUmax

¼ 1:386� 1:333 exp �12

R=z� 0:1750:11

� �2 !

ð2Þ

VVmax

¼ 1:153� 0:998 exp �12

R=z� 0:1790:094

� �2 !

ð3Þ

0.00 0.04 0.08 0.12 0.16 0.200.0

0.2

0.4

0.6

0.8

1.0

R/z

50 mm 90 mm 130 mm 170 mm Fitting curve

0.0

0.2

0.4

0.6

0.8

1.0

(a)

U/U

max

V/V

max

(b)

Fig. 7. Variation of the non-dimensional velocities, (a) axial velocity and (b) radialvelocity as a function of the non-dimensional radial distance.

3.3. Spray thermal characterization

3.3.1. Average temperature of dropletsPlease note that it is the steady state temperature value that

we reported in this section. It is understood that the temperaturevalue should represent an average temperature of the dropletsat the given location since both the droplet size and temperaturewill experience a statistical distribution around their meanvalues.

Fig. 8 presents typical readings from the thermocouple probe atthree locations, where the centerline distances are 10, 30, and80 mm from the nozzle exit. As can be seen from Fig. 8, all of thespray temperature readings reached a steady value no more than1 s. However, there still exist some slight fluctuations around thesteady value. Therefore, the average value of 100 temperature dataduring 1.999–2 s is used as the droplet average temperature toeliminate the fluctuations.

3.3.2. Characteristics of centerline droplets temperatureThe variation of the centerline droplets average temperature

(T) of a given R134a spray has been reported in several cases inthe literature [5,23–25]. Nevertheless, Fig. 9 is the first one toshow a complete picture from the nozzle exit to the end of thespray. As shown in the figure, the centerline temperature initiallyshows an almost exponential decay with the fastest drop in thetemperature taking place near the nozzle exit. The temperaturedrop slows down as spray develops further for the droplets’ rapidevaporation and the insufficiency of the convective heat transferfrom the surrounding. This exponential decay of droplets averagetemperature can be described by an exponential function as givenin Eq. (4),

T � Tmin

Ts � Tmin¼ exp �0:22

zd

� �ð4Þ

where Tmin and Ts are the droplets minimum temperature and sat-urated temperature at the atmospheric pressure. Eventually, thecenterline temperature reaches a minimum value of �59.2 �C atz = 140 mm where the latent heat for evaporation equals to the con-vective heat transfer. From z = 140 mm to z = 150 mm, the temper-ature fluctuates near the minimum temperature. At z = 150 mm,however, one can see a sudden large increasement in the centerlinetemperature. At z = 180 mm, the centerline temperature rises to theboiling point of R134a, indicating a complete evaporation of alldroplets. Afterwards, the spray becomes a mixture of R134a vaporand air. By carefully examining our high-speed camera photo in

0.0 0.4 0.8 1.2 1.6 2.0-60

-40

-20

0

20

80 mm

30 mm

-55.8oC

-46.1oCTem

pera

ture

, T (

o C)

Time, t (s)

Taver

=-34.5oCz =10 mm

Fig. 8. Typical droplets temperature readings obtained during measurements.(Three centerline locations: 10, 30 and 80 mm from the nozzle exit).

0 20 40 60 80 100 120-12

-8

-4

0

4

8

12-60-50-40-30 -60-50-40-30 -60-50-40-30

-60-50-40-30-60-50-40-30

Rad

ial d

ista

nce,

R (m

m)

T (oC)

RSCW

RSTW

0 50 100 150 200

-60

-50

-40

-30

-20

Tsat

=-26.07oC

zSTL

zCSD

zMTD

Tem

pera

ture

(o C

)

Axial distance, z (mm)

Tsat

=-26.07oC

Fitting curve

Fig. 9. Variation of the centerline droplets temperature of R134a spray as a functionof the spray distance (z) from the exit of the nozzle.

4466 Z. Zhifu et al. / International Journal of Heat and Mass Transfer 55 (2012) 4460–4468

Fig. 2a, we find that the light intensity transits from white to dimpink1 at the same spray distance, confirming the transition from adroplet spray to a gas mixture.

Several interesting thermal characteristics of an R134a spraycan be observed from Fig. 9. For example, we find that the droplettemperature near the exit, e.g. at z = 5 mm, is already well belowthe boiling point of R134a at the atmospheric pressure, which sug-gests that liquid superheating and flashing boiling takes placewithin the nozzle tube or very near the exit region of the nozzle(z < 5 mm). The thermal measurement is in consistent with ourhigh-speed camera observation as well as those by Vu and Aguilar[26]. Spray formation takes place in the region near the exit of thenozzle and the droplets become thermal equilibrium with the sur-rounding pressure in a very short distance.

Another interesting feature is the minimum spray temperatureand the distance to reach this minimum temperature. The presentminimum temperature of �59.2 �C is almost identical to �60 �Cmeasured by Yildiz et al. [25] who used larger nozzle sizes (1–4 mm in diameter). It can be easily understood that the minimumtemperature corresponds to the situation where the heating of thedroplet by the surrounding gas is balanced by evaporation at thedroplet surface. Before reaching the minimum temperature, the la-tent heat carried away from the droplet by evaporation is largerthan the heat transferred into the droplet from warm air, whileafter reaching the minimum temperature, the droplet evaporationslows down and the convective heating becomes dominated. Thelatter can explain the fast increasement of the droplet temperaturein the region between z = 150 and z = 180 mm.

Polanco et al. [5] proposed a concept of ‘‘minimum temperaturedistance’’ which ‘‘defines the end of the boiling and nucleation pro-cess, and the start of the mechanical processes.’’ It is not clear,however, what is the meaning of ‘‘the mechanical processes’’ tothe present authors. Instead, we can define a similar ‘‘minimumtemperature distance’’ (MTD) as the spray distance when the drop-let temperature reaches minimum. In present case, the minimumtemperature distance is zMTD = 140 mm, or non-dimensionalizedvalue of Z = z/d = 170, where d is the inner diameter of the nozzle.It is noticed that these data fall in the range of the dimensionless‘‘Minimum Spray Distance’’ of 150–170 as reported in the litera-ture [5]. Fig. 9 also shows that there are other two characteristicthermal lengths in this spray, i.e. the spray distance at which sud-den heating of droplets takes place and the spray distance wherethe average temperature of droplets reaches the boiling point.We can name the former as the ‘‘Cold Spray Distance’’ (CSD) and

1 For interpretation of color in Fig.2, the reader is referred to the web version of thisarticle.

the latter as the ‘‘Spray Thermal Length’’ (STL). As indicated bythe name itself, the ‘‘Cold Spray Distance’’ refers to the spray dis-tance where the spray maintains a cold temperature around�60 �C while the ‘‘Spray Thermal Length’’ refers to a total spraylength where liquid droplets exist. For the present R134a spray,the dimensionless CSD is about ZCSD = 185 and dimensionless STLis about ZSTL = 235.

3.3.3. Temperature profile of the whole sprayFig. 10 shows the radial temperature profiles in the R134a spray

at five spray distances from 10 to 100 mm. As one can see, there isa strong dynamic variation of the droplet temperature in both ra-dial and axial directions. In the region near the exit of the nozzle(z < 50 mm), the radial temperature at a given spray distanceshows a ‘‘W’’ shape with high temperatures at the central region.Because of the very limited air entrainment in the central regionnear the nozzle, there will be a core region where liquid dropletsexperience a high vapor concentration in the spray and fairly weaksurface evaporation. Therefore, such a core region is made of densedroplets or some unbroken liquid with a fairly high temperaturenear the boiling point (�26.2 �C) of the R134a at the atmosphericpressure. Along the radial direction from the centerline, the droplettemperature decreases firstly to a minimum value and then in-creases again. As the spray expands quickly along the radial direc-tion at the exit of the nozzle, the liquid droplets lose much energydue to the strong evaporation and insufficient convective heatingfrom the surrounding, which leads to the reduction of temperature.As the droplets fly further away from the center, they quickly dis-perse as air entrainment penetrates deeper into the spray core, asdemonstrated by our high-speed camera photo in Fig. 2b (whitecore region). On the other hand, at the outer periphery of the spray,droplets experience a warm air with little R134a vapor, convectiveheating will dominate droplet evaporation, leading to the increas-ement in the droplet temperature again.

The ‘‘W’’ shape distribution of the radial temperature is mainlyobserved in the region of the first 50 mm spray distance and thecenterline temperature diminishes continuously along the spraydirection as shown in Fig. 9. Fig. 10 shows that the central hightemperature core disappears completely at about z = 50 mm, andone finds a larger spray core region with a uniform temperaturesurrounded by a warm outer region. As the spray flies furtherdownstream, this uniform temperature core region expands

Axial distance, z (mm)

Fig. 10. Radial temperature profiles in R134a spray at five spray distances: z = 10,30, 50, 80 and 100 mm.

0 20 40 60 80 1000

3

6

9

12

RSCW

Rad

ial D

ista

nce,

R (

mm

)

Axial Distance, z (mm)

RSTW

Fig. 11. Variation of the two thermal widths (RSTW and RSCW) as a function of thespray distance.

Z. Zhifu et al. / International Journal of Heat and Mass Transfer 55 (2012) 4460–4468 4467

outward slowly and the value of the temperature also decreasescontinuously.

Two radial thermal lengths can be defined by examining the ra-dial temperature distribution at any given spray distance, the spraythermal width (STW) RSTW and the spray cold width (SCW) RSCW, asindicated by the dash lines in Fig. 10. STW here refers to the radialwidth of the spray at a given spray distance, enclosed by the iso-therm of the boiling point of the refrigerant. Beyond RSTW thereshould be no liquid droplets. SCW refers to the cold region withinthe spray where the droplet temperature is low. For the case ofspray cooling, RSCW represents the overall cooling spot of the spray,while SCW indicates the central core cooling region. Fig. 11 plotsthe variation of such two thermal widths as a function of the spraydistance.

To further visualize the thermal field of the R134a spray, a two-dimensional (z � R) temperature field is plotted based on the mea-sured temperature data within the spray distance of 100 mm andcompared with a high-speed camera photo, as given in Fig. 12.

Fig. 12. Two dimensional (z � R) temperature field (left half) and a comparisonwith the photo image (right half) of the R134a spray.

Some interesting observations can be made from such a plot. Firstof all, fast expansion of the spray near the nozzle is clearly demon-strated if one notices that the nozzle diameter is less than 1 mmwhile the spray width within 1 mm spray distance expands tomore than 3 mm. Such an expansion in the spray formation is fairlyunique for the present R134a spray comparing to superheatedwater spray [19,20]. Near the nozzle exit, the thermal plot clearlyshows a warm core region that is extended to z = 40 mm alongthe central line. The axial distance from z = 40 to z = 50 mm is atransitional stage in which the warm core region has disappearedgradually. Fig. 12 also shows the uniform cold region within thespray in the downstream of the spray axial distance beyond50 mm (z > 50 mm). We can also notice that there is always awarm shell surrounding the central cold region. In addition, themagnified high-speed camera photo clearly demonstrates thedevelopment of spray from an early relatively stable spray to amore unstable one, as indicated by the sudden increase in thespray width near z = 80 mm.

4. Conclusions

Experiments have been conducted to investigate the spray andthermal characteristics of R134a two-phase flashing spray. High-speed camera photos show fast expansion of the spray at the exitof the nozzle, leading to a wide spray angle. The PDPA measure-ment results show that there exist a quick reduction in dropletsdiameter along the centerline and acceleration in the axial velocitynear the nozzle exit due to the explosive atomization of this flash-ing spray. The radial variations of the droplets number, diameterand velocities suggest that the spray shows a poly-disperse charac-teristic close to the nozzle exit and it becomes more uniform in thedownstream. In addition, the radial variations of the non-dimen-sional velocities are self-similar and two empirical equations wereproposed to describe the non-dimensional velocities.

The temperature data shows that the thermal field of the sprayconsists of three regions, a hot core region near the exit of the noz-zle, a cold region with uniform low temperature in the down-stream, and a warm periphery. The hot core region shows a ‘‘W’’shape temperature profile in the radial direction. Two thermalwidths, i.e. the spray thermal width (RSTW) and spray cold width(RSCW), are defined to describe the radial variation of the dropletstemperature. The variation of the centerline droplets temperaturecan be characterized by three axial thermal lengths, minimumtemperature distance (zMTD), Spray Thermal Length (zSTL), and ColdSpray Distance (zCSD). An exponential equation has been developedto describe the temperature decay from the nozzle exit to zCSD.Additionally, quantitative data have been obtained for those char-acteristic lengths.

Acknowledgements

This work was jointly supported by Chang Jiang Scholars Pro-gram of the Ministry of Education of China and Li KaShing Founda-tion of Hong Kong (G.-X. Wang, 2006–2009) as well as theFundamental Research Funds for the Central Universities(2011jdhz35, xjj20100214). Also, we acknowledge Institutionalsupport from the State Key Laboratory of Multiphase Flow inPower Engineering, Xi’an Jiaotong University.

References

[1] E. Sher, T. Bar-Kohany, A. Rashkovan, Flash-boiling atomization, Prog. EnergyCombust. Sci. 34 (4) (1998) 417–439.

[2] K. Patrick, Takahashi, Project blue revolution, J. Energy Eng. 122 (3) (1996)115–124.

[3] J.S. Nelson, T.E. Milner, B. Anvari, B.S. Tanenbaum, L.O. Svaasand, S. Kimel,Dynamic epidermal cooling in conjunction with laser-induced

4468 Z. Zhifu et al. / International Journal of Heat and Mass Transfer 55 (2012) 4460–4468

photochermolysis of port wine stain blood vessels, Lasers Surg. Med. 19 (2)(1996) 224–229.

[4] K.M. Kelly, J.S. Nelson, Update on the clinical management of port wine stain,Lasers Med. Sci. 15 (4) (2000) 220–226.

[5] G. Polanco, A.E. Holdo, G. Munday, General review of flashing jet studies, J.Hazard. Mater. 173 (1-3) (2010) 2–18.

[6] W.M. Liu, Study on mechanism of heat transfer and flow in the flash processwith liquid was discharged into vacuum, PhD Thesis, Xi’an Jiaoting University,Xi’an, China, 2006.

[7] Z.F. Zhou, H. Xin, B. Chen, G.X. Wang, Evaporation characteristics of a singledroplet in laser treatment of PWS in conjunction with cryogen spray cooling,Chin. J. Lasers 35 (6) (2008) 952–956 (in Chinese).

[8] G. Aguilar, G.X. Wang, J.S. Nelson, Dynamic behavior of cryogen spray cooling:effects of spurt duration and spray distance, Lasers Surg. Med. 32 (2) (2003)152–159.

[9] G. Aguilar, G.X. Wang, J.S. Nelson, Effect of spurt duration on the heat transferdynamics during cryogen spray cooling, Phys. Med. Biol. 48 (14) (2003) 2169–2181.

[10] W. Franco, J. Liu, G.X. Wang, J.S. Nelson, G. Aguilar, Radial and temporalvariations in surface heat transfer during cryogen spray cooling, Phys. Med.Biol. 50 (2) (2005) 387–397.

[11] S.S. Hsieh, T.C. Fan, H.H. Tsai, Spray cooling characteristics of water and R-134a. Part I: nucleate boiling, Int. J. Heat Mass Transfer 47 (26) (2004) 5703–5712.

[12] S.S. Hsieh, T.C. Fan, H.H. Tsai, Spray cooling characteristics of water and R-134a. Part II: transient boiling, Int. J. Heat Mass Transfer 47 (26) (2004) 5713–5724.

[13] S.S. Hsieh, C.H. Tien, R-134a spray dynamics and impingement cooling in thenon-boiling regime, Int. J. Heat Mass Transfer 50 (3–4) (2007) 502–512.

[14] D. Yildiz, P. Rambaud, J. Rambaud, J. Van Beek, Break-up, droplet size andvelocity characterizations of a two-phase flashing R134a jet, in: Proceedings ofthe Fifth International Conference on Multiphase Flow, Yokohama, Japan, 2004.

[15] D. Yildiz, P. Rambaud, J. Rambaud, J. Van Beek, J.M. Buchlin, Evolution of thespray characteristics in superheated liquid jet atomization in function of initialflow conditions, in: Proceedings of the Tenth International Conference onLiquid Atomization and Spray Systems, Kyoto, Japan, ICLASS06-122, 2006.

[16] R. Brown, L. York, Spray formed by flashing liquid jets, AIChE J. 28 (2) (1962)149–153.

[17] O. Miyatake, T. Tomimura, Y. Ide, T. Fujii, An experimental study of sprayevaporation desalination, Desalination 36 (1981) 113–128.

[18] E.M. Peter, A. Takimoto, Y. Hayashi, Flashing and shattering phenomena ofsuperheated liquid jets, JSME Int. J. 37 (2) (1994) 313–322.

[19] R.D. Reitz, A photographic study of flashing–boiling atomization, Aerosol Sci.Technol. 12 (3) (1990) 561–569.

[20] B.S. Park, S.Y. Lee, An experimental investigation of the flash atomizationmechanism, Atomization Sprays 4 (2) (1994) 159–179.

[21] J.T. Allen, Laser based measurements in two-phase flashing propane jets. Partone: velocity profiles, J. Loss Prevention Process Ind. 11 (5) (1998) 291–297.

[22] J.T. Allen, Laser based measurements in two-phase flashing propane jets. Parttwo: droplet size distribution, J. Loss Prevention Process Ind. 11 (5) (1998)299–306.

[23] G. Aguilar, B. Majaron, W. Verkruysse, Y. Zhou, J.S. Nelson, E.J. Lavernia,Theoretical and experimental analysis of droplet diameter, temperature, andevaporation rate evolution in cryogenic sprays, Int. J. Heat Mass Transfer 44(17) (2001) 3201–3211.

[24] H. Vu, O. Garcia-Valladares, G. Aguilar, Vapor/liquid phase interaction in flareflashing sprays used in dermatologic cooling, Int. J. Heat Mass Transfer 51 (23–24) (2008) 5721–5731.

[25] D. Yildiz, P. Rambaud, J. Van Beeck, J.M. Buchlin, Thermal characterization of aR134A two-phase flashing jet, in: ICLASS 2003–9th International Conferenceon Liquid Atomization and Spray Systems, Sorrento, Italy, 2003.

[26] H. Vu, G. Aguilar, High-speed internal nozzle flow visualization of flashing jets,in: Proceedings of the 11th Triennial International Annual Conference onLiquid Atomization and Spray Systems, Vail, Colorado, USA, 2009.