experimental investigation of air-atomized spray with aqueous polymer additive for high heat flux...

International Journal of Heat and Mass Transfer 72 (2014) 362–377

Contents lists available at ScienceDirect

International Journal of Heat and Mass Transfer

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

Experimental investigation of air-atomized spray with aqueous polymeradditive for high heat flux applications

0017-9310/$ - see front matter � 2014 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.ijheatmasstransfer.2014.01.024

⇑ Corresponding author. Tel.: +91 3222 283942 (office).E-mail address: [email protected] (S. Chakraborty).

Satya V. Ravikumar a, Jay M. Jha a, A.M. Tiara a, Surjya K. Pal b, Sudipto Chakraborty a,⇑a Department of Chemical Engineering, IIT Kharagpur, 721302, Indiab Department of Mechanical Engineering, IIT Kharagpur, 721302, India

a r t i c l e i n f o

Article history:Received 5 October 2013Received in revised form 17 December 2013Accepted 11 January 2014Available online 1 February 2014

Keywords:Atomized spray coolingAqueous polymer coolantHeat transfer enhancementTransition boilingUltrafast cooling

a b s t r a c t

The experimental investigation of using a water based polymer additive to enhance the spray cooling per-formance of hot steel plate, has been carried out in the current research. This is essentially important toproduce high-strength steel on the run-out table (ROT) of a Hot Strip Mill. The ultra-high heat flux coolingsystem has been developed using an air-atomized spray, containing dissolved polyvinylpyrrolidone (PVP)in water at different concentration ranges between 10 and 150 ppm and compared with the cooling per-formance of pure water. To understand the heat transfer mechanism of aqueous polymer solutions, thephysical properties such as surface tension, viscosity and thermal conductivity were measured. The cool-ing experiments were conducted using an AISI 304 stainless steel plate of 6 mm thickness initially kept ata temperature above 900 �C, where the Leidenfrost effect is predominant. In these experiments, the tran-sient temperature data during cooling has been measured with three subsurface thermocouples and thistime-temperature history has been used to estimate the surface temperature and the surface heat fluxhistories using a commercial inverse heat conduction software namely, INTEMP. The results explain thatthe polymer solution has a significant effect on the enhancement of surface heat flux, critical heat flux, aswell as the cooling rate of the test plate. It was observed that an increase in the polymer concentrationincreases the heat transfer rate up to an optimal concentration; after which it results in a reduction in therate. A maximum cooling rate of 253 �C/s was obtained with a critical heat flux of 4.212 MW/m2, whichcan be termed as the higher range of an ‘Ultrafast cooling’ process. Overall, the aqueous polymer solutioncan serve as a better heat transfer fluid for high heat flux applications.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Atomized spray cooling is an important technique that can beused for high heat flux applications starting from electronics, nu-clear fuel rods, space systems, and various metal quenching pro-cesses. This method can also be an alternative to conventionallaminar jet impingement cooling in the run-out table (ROT) of aHot Strip Mill (HSM) in steel processing plant. The mechanicalproperties of the steel strip is conditioned and altered only negligi-bly in the hot rolling mill of steel industries. However these prop-erties can be greatly modified by controlling the cooling rate ofsteel between a temperature range of 900–600 �C [1,2]. Steel forhigh-strength applications necessitates multiphase microstruc-tures such as ferrite–bainite, ferrite–martensite and pearlite–mar-tensite, etc. [3,4], which can be processed by applying cooling ratesgreater than 150 �C/s. The conventional means of laminar coolingtechnology equipped with ROT cannot provide these multiphasemicrostructures, because the acquired cooling rate is limited to a

lower extent, usually in the range of 30–80 �C/s. Moreover, jetimpingement cooling can be adversely influenced by film boiling,and the heat transfer rate is limited to the stagnation point, whichcauses unacceptable distortion and high residual stresses.Although the water spray cooling can provide higher cooling rates,it encounters the Leidenfrost phenomenon [5–7]. Due to this alayer of vapor film builds up on the hot surface as a result of a highthermal gradient between the solid and liquid phases, which re-stricts the opportunity of solid–liquid contact resulting a low heattransfer rate at high surface temperatures [8]. Under such circum-stances, air-atomized spray cooling appears to be a promisingalternative for ROT applications. Here, compressed air used foratomization sweeps the partially evaporated droplets from the sur-face, which results in the breakdown of the vapor film and the highmomentum droplets gained by a higher value of air pressure cantouch the hot surface; therefore, the film boiling heat transfergreatly reduces [9–11]. Since the temperature of the plate is muchhigher than the coolant temperature, the current system falls un-der high heat flux application where different boiling heat transferregimes exist during the cooling. The characteristic heat transfermechanisms associated with the air-atomized spray cooling of

http://crossmark.crossref.org/dialog/?doi=10.1016/j.ijheatmasstransfer.2014.01.024&domain=pdf

http://dx.doi.org/10.1016/j.ijheatmasstransfer.2014.01.024

mailto:[email protected]

http://dx.doi.org/10.1016/j.ijheatmasstransfer.2014.01.024

http://www.sciencedirect.com/science/journal/00179310

http://www.elsevier.com/locate/ijhmt

Nomenclature

B strip thickness, mmCp specific heat, J/kg Kd0 orifice diameter, mk thermal conductivity, W/m KL strip length, mmt time, sT temperature, �Cx direction along the length of the plate, mmy direction along the thickness of the plate, mmUA velocity of air, m/s

Greek symbolsh contact angle, degq density, g/ccr surface tension, mN/m

k wave length, cml viscosity, mPa s

AbbreviationsAFR air–fluid ratioCHF critical heat flux, MW/m2

CPC critical polymer concentration, ppmPVP polyvinylpyrrolidoneROT run-out tableSMD sauter mean diameter, lmT/C thermocoupleUFC ultrafast cooling

S.V. Ravikumar et al. / International Journal of Heat and Mass Transfer 72 (2014) 362–377 363

high temperature steel plates are transition boiling at high surfacetemperatures, nucleate boiling heat transfer and finally singlephase forced convection at low surface temperatures [12]. More-over, on the other hand, no film boiling heat transfer exists inair-atomized spray cooling which was the dominant factor in thecase of jet impingement cooling [11]. In order to create the desiredcooling rate for a particular application, it is essentially importantto control the phenomena of boiling heat transfer during the cool-ing process. Therefore, it is an area of overwhelming pragmatic sig-nificance as well as of considerable fundamental importance. Theexperimental approach employed in the present research can serveto control the high heat flux systems existing in various industries.

Considerable experimental and theoretical work on the coolingof a hot metal plate using different processes such as jet, spray, air-atomized spray, etc., have been carried out by different researchers[13–18]. However, only a few studies have reported the coolingperformance of an air-atomized spray impingement on a high-temperature metal surface in conjunction with an inverse heatconduction technique to estimate the surface boiling characteris-tics. For example, De Oliveira et al. [13] reported the air-atomizedspray cooling of an AISI 304 stainless steel plate over a wide rangeof water flow rates and at an initial surface temperatures up to850 �C. They found that heat transfer coefficient is strongly depen-dent on those variables. They also concluded that air-atomizedspray cooling has an advantage over conventional water spraysas it is able to achieve higher heat transfer coefficients under sim-ilar operating conditions. They used the commercial inverse heatconduction solver, namely INTEMP to estimate the surface temper-ature and heat flux values using the measured internal tempera-ture data by the thermocouples. Al-Ahmadi and Yao [19] studiedthe effect of water mass flux and air pressure for an air-mist spray,cooling a stainless steel plate initially kept at 900 �C. They con-cluded that water mass flux has a greater effect on the heat trans-fer rate than air pressure. From the investigation of [11], it wasobserved that water mass flux is an important parameter in atom-ized spray cooling for a metal surface of temperature 600 �C. Bhat-tacharya et al. [20] presented a theoretical analysis of sprayevaporative cooling to achieve higher heat transfer rates in termsof the size of water droplets impinging on the hot surface. It wasreported that droplets of lower size have a higher heat transfercapability due to rapid evaporation from the hot surface.

The boiling heat transfer behavior of the coolant fluid (water) ishighly influenced by its thermo-physical properties [21]; severalworks are reported in the literature for additive based heat transferenhancement in different cooling processes [22–26]. However,some studies are limited to pool boiling heat transfer [27–30].

Recently, the enhancement of heat transfer rate in jet and air-atomized spray impingement cooling of hot metal surfaces withdissolved surface active agents in pure water have been carriedout by the authors [25,31,32]. There exists an optimal solution con-centration of surfactant for higher heat transfer enhancement. Amaximum cooling rate of 214 �C/s has been achieved with AISI304 stainless steel (initial temperature 900 �C) using non-ionic sur-factant added air-atomized water spray [33]. It is concluded thataddition of surfactant decreases the surface tension and hencehigher wettability of the surface occurs with minimal solid–liquidcontact angle, which increases the heat removal rate from the sur-face. However, foamability of the surfactant type and its concentra-tion also play a key role in the motion of wetting fronts on the hotsurface. Moreover, the influence of viscosity of coolant on the heattransfer rate while working with surfactant added water was alsoexplored in detail. The results are in substantial agreement withthose from the previous studies [26,29,34–37]. Apart from usingsurfactants, studies have also been conducted using polymericadditives on the enhancement of heat transfer [37]. Kotchaphakdeeand Williams [38] did the pioneering work identifying the poolboiling heat transfer enhancement with different polymeric addi-tives in water. They reported that surface tension alone is not thedominant factor in heat transfer enhancement, whereas, molecularweight and the concentration of the polymer, and the viscosity ofthe solution also play a crucial role. Therefore, it is important tounderstand the effect of polymer additives on the physical proper-ties of the coolant for heat transfer enhancement. In general, mostpolymeric solutions do not show significant effect on surface ten-sion; whereas polymer of higher concentrations dissolved in watercause a significant increase in viscosity. However, some water-sol-uble polymers like hydroxyethyl cellulose (HEC) and polyvinylpyr-rolidone (PVP) decrease the surface tension with increasingconcentration similar to that of surfactants [38–40]. Zhang andManglik [41] experimentally studied the decrease in surface ten-sion of water by HEC polymeric additive. Similar to the critical mi-celle concentration of surfactant, a critical polymer concentrationis also relevant in polymeric surfactant additive solutions.

It must be pointed out from the literature that selection of apolymeric additive that can serve like a surfactant is better for heattransfer enhancement when working with polymeric additivesolutions. Moreover, working at low concentration of polymers isadvantageous for heat transfer enhancement as non-Newtonianfluidic shear-thinning behavior will arise at high concentrations.It is also found from the literature that polymeric additives of low-er concentration enhance the nucleate boiling heat transfer rate.However, most of the previously published work on polymeric

364 S.V. Ravikumar et al. / International Journal of Heat and Mass Transfer 72 (2014) 362–377

additive heat transfer enhancement is confined to pool boiling oflow temperature surfaces [38,42,43]. There is hardly any work re-ported on the heat transfer enhancement from high temperaturesurfaces (above 900 �C) using polymeric additive solutions in air-atomized spray cooling process. Hence, this work can provide thefundamental mechanisms involved in the enhancement of heattransfer from a high temperature stainless steel plate of 900 �C un-der transition and nucleate boiling heat transfer conditions, whileworking with polymeric additive water coolant in air-atomizedspray impingement.

In the present paper, the experimental results of heat flux andcooling rate measurements in an air-atomized spray cooling sys-tem with polymeric additive (PVP) in pure water as the workingfluid are reported for an AISI 304 stainless steel plate at 900 �C.The plate has an area of 100 mm2 and a thickness of 6 mm. A com-mercial inverse heat conduction solver has been used to estimatethe transient surface heat flux and temperatures in conjunctionwith the measured temperature histories by thermocouples duringcooling. The influence of polymer concentration on the surface heatflux removal rate is discussed. The flow rates of air and water usedin the entire study are 10 lpm and 30 Normal m3/h, which havebeen optimized for high cooling rate in the authors’ earlier studies[12]. Apart from the heat transfer study, the effect of polymer onthermo-physical properties such as surface tension, viscosity andthermal conductivity of the coolant have been measured. The tech-nique of enhancement of air-atomized spray cooling heat transferby polymeric surfactant additives can then be useful to optimizecooling process in ROT for designing the steel of desired micro-structure and mechanical properties for a particular application.

2. Experimental facility

2.1. Test setup

Fig. 1(a), which mainly consists of the nozzle and test surface;the coolant (pure water or water with polymer additive) and airsupply systems, along with the data acquisition system. The testplate used in the study is stainless steel of grade AISI 304 with athickness of 6 mm. The surface of the plate facing the sprayimpingement has a square shape with a 100 mm side length. Theschematic representation of the test plate is shown in Fig. 1(b).During the experiment, the plate is initially heated to a tempera-ture of 1050 �C in a muffle furnace. In order to measure the tran-sient temperature during cooling, three ‘K’ type thermocouplesare inserted in the steel plate at different locations parallel to thetest surface. All the thermocouples are placed 3 mm beneath thetest surface through 3 mm diameter holes drilled up to 50 mm inthe plate. The methodology of this thermocouple installation iswell in agreement with the study of [44] to avoid the uncertaintycaused by the hole diameter. Since the heat extracted from thetop surface, while the other surfaces are nearly adiabatic, the heattransfer rate is predominately in the thickness direction than theother two directions on the surface. This effect is more pronouncedin the case of ultrafast cooling which is a rapid quenching process.Therefore, the current system has been modeled as a two-dimen-sional transient heat conduction problem to calculate the surfaceheat flux and temperatures.

A full cone twin fluid nozzle (Lechler, Germany, Model:170.801) is used in the current air-atomized spray cooling study.The nozzle can provide uniform spray distribution over the wholecircular impact area on the surface with a cone angle of 32 deg. Allthe experiments have been conducted at a fixed nozzle-to-surfacedistance of 60 mm with an air and coolant flow rates of 30 Nor-mal m3/h and 10 L/min, respectively, optimized by the authors ear-lier study [12].

In order to provide the required air and coolant flow to the noz-zle for spray formation, a compressor and pump have been used.The flow rates of air and coolants are adjusted by the valves whichare connected to the rotameters for flow measurement. A pressureregulator is mounted to supply the desired pressure for atomiza-tion of coolant into a fine spray.

A data acquisition system consisting of NI cDAQ-9174 USB chas-sis along with an NI 9211 thermocouple input module has beenused to measure the thermocouple input signals with a rate of10samples/s during cooling.

2.2. Atomized spray cooling procedure

After heating the test plate to 1050 �C in the muffle furnace, it isloaded on to the ceramic brick arrangement in the experimentalapparatus with the test surface in a face up position and the nozzleis placed above for cooling. The ceramic brick contains a groove tothermally insulate all other sides except the top quench surface.

After transferring the hot plate from the muffle furnace to theceramic brick slot, the air and coolant solenoid valves are openedand the air-atomized spray from the nozzle impinges on the hotsurface. The complete experimental design for this study has beenshown in Table 1.

In these experiments, the coolants are prepared by dissolvingdifferent concentrations of polymeric additive in the base fluidi.e., pure water. During each test, the coolant temperature wasmaintained at 30 �C. The polymeric compound used in this studyis polyvivylpyrrolidone (PVP)-K30, having an average molecularweight of �40,000 g/mol procured from Sigma–Aldrich (India).The base fluid used is normal filtered water having a surface ten-sion of 73 mN/m (298 K) and a viscosity of 0.91 cP. The solutionsare prepared by vigorous mixing with a high-speed stirrer to en-sure uniform dispersion of the polymer additive in water. The char-acterization of these coolants is done by measuring the physicalproperties such as surface tension, viscosity and the thermal con-ductivities. The polymer concentrations taken in the study are atthe lower range of 0–150 ppm, as higher concentration causes alarge increase in the viscosity of the coolant due to non-Newtonianfluidic behavior [37].

3. Characterization of coolant

3.1. Surface tension

The surface tension of the coolant was measured using Tensi-ometer [Krüss (Germany), model: K100]. The method used to mea-sure the surface tension is Wilhelmy plate, which is a thin plateconsists of few centimeters in area made up of platinum. Beforestarting of each measurement, the platinum plate was burnt tored-hot conditions using Bunsen burner to roughen the surfacefor complete wetting. The sample vessels were moved at a verylow speed of about 200 lm/s during the measurements. The sur-face tension values taken by this procedure are highly accurateand reproducible. Fig. 2(a) shows the effect of concentration ofthe polymeric additive on the surface tension. It can be seen thatthe addition of polymer (PVP) substantially decreases the surfacetension of water, which indicates surfactant like behavior of PVP.Therefore, some studies classify polyvinylpyrrolidone (PVP) as apolymer surfactant of non-ionic type [39]. It is observed that an in-crease in PVP concentration decreases the surface tension substan-tially up to a concentration of 110 ppm. Thereafter, a furtherincrease in PVP concentration causes a rise in the surface tension,due to the effect of critical polymer concentration (CPC). The sur-face tension results can be compared with the results obtainedfor the Tween 20 surfactant in the authors’ earlier study [33] and

Fig. 1. Schematic diagram of (a) air-atomized spray cooling system, and (b) details of test plate.

Table 1Experimental design.

Parameter/Test substances Operating condition

Initial plate temperature >900 �CCoolant temperature 30 �CSpray cone angle 32 degSurface-to-nozzle distance 60 mmFlow rate of air 30 Normal m3/hFlow rate of water 10 L/minAir pressure 0.8 MPaWater pressure 0.4 MPaConcentration of polymer additive 0, 10, 30, 50, 70, 90, 110, 130, 150 ppm


the reduction of surface tension with the addition of polymericadditive is highly appreciable. This is because the molecular weightof the polymer surfactant (PVP) is about 33 times higher than thatof Tween 20 surfactant. This is in accordance with the study of[45,46] that a surfactant of higher molecular weight can have high-er order in the depression of surface tension than a surfactant oflower molecular weight, as is reviewed by Cheng et al. [37].

It should be pointed out that the decrease in surface tension ofthe liquid is responsible for higher wettability of solid surface

along with the minimization of the solid–liquid contact angle,which can be described by the Young’s equation. Therefore, in thisstudy it is important to know how the solid–liquid contact angle ischanging with the polymer concentration in water. A Goniometer(Rame-Hart instrument Co. USA, model no. 190-F2) has been usedto measure the contact angle between the solid surface and cool-ant. Before starting the measurement, proper care has been takento eliminate the manual error by cleaning the test surface with ace-tone and dried for a few minutes in air. The measurement was re-peated thrice and the average value of the contact angle has beenreported. Fig. 2(b) shows the effect of polymer additive concentra-tion on the solid–liquid contact angle. It can be seen that an in-crease in polymer concentration decreases the solid–liquidcontact angle (or wetting angle) thereby increasing the wettabilityof the surface. On the other hand, a decrease in contact angle pro-motes quick spreadability of coolant on the solid surface, whichcan explained by the change in solid–liquid contact angle withtime and coolant type, as shown in Fig. 3. Overall, it is concludedthat lowering of the surface tension facilitates the higher solid–li-quid contact with higher surface area occupied by the individualdroplet, which may cause the droplet evaporation to start at afaster rate. As a result, the heat removal rate from the impinged

Fig. 2. Effect of PVP concentration on (a) surface tension, (b) solid–liquid contact angle.

Fig. 3. Images of contact angle relaxation on the solid surface while working with different coolants.

Fig. 4. Effects of polymeric additive concentration on viscosity of solutions at roomtemperature.


surface may be high. Moreover, according to the cavitations theory[47], the energy required to rupture a liquid in tension is directlyproportional to the surface tension. Therefore, lower the surfacetension, easier will be the atomization of droplets into finer size.From the theoretical analysis by Bhattachraya et al. [20], it hasbeen found that if the droplet size impinging on the hot substrateis small, it is easy to attain an ultrafast cooling rate due to highevaporation rate. In addition, a droplet of smaller size facilitateslower vapor layer thickness, when impinged on high temperaturesurface.

3.2. Viscosity

As pointed out in the literature that viscosity also plays a majorrole in the enhancement of convective heat transfer rate [48], theauthors have measured the zero-shear viscosity of the coolant inthe entire concentration range of polymer additive by using vis-cometer (Model: DV2T, Brookfield Viscometer, USA). All the mea-surements have been repeated thrice and the average value ofviscosity has been reported in Fig. 4. It can be seen that an increasein polymeric additive concentration causes a marginal increase inthe viscosity of the coolant.

Fig. 5. Effects of polymeric additive concentration on thermal conductivity ofpolymer solutions at room temperature.


3.3. Thermal conductivity

Fig. 5 represents the thermal conductivity of coolant with re-spect to the concentration of polymeric additive (PVP) at roomtemperature. These measurements are performed using KD2 PROThermal properties analyzer manufactured by DECAGON DevicesInc., USA. It can be observed from the results that the addition ofpolymeric additives of very low concentration (<70 ppm) causesno significant alteration in the thermal conductivity of water; how-ever, a minor decreasing trend is found at higher PVP concentra-tions used in the subject field.

4. Application of commercial inverse heat conduction solver(INTEMP)

4.1. Experimental data reduction

Based on the cooling characteristics of the experiments, theconduction heat transfer through the steel plate has been modeledas a two dimensional transient heat conduction problem. A planarmodel of the actual plate with true dimensions has been taken forthe thermal model and the governing energy equation can be writ-ten as,

@

@xk@T@x

� �þ @

@yk@T@y

� �¼ qSCp

@T@t; 0 < x < ðL ¼ 100 mmÞ;

0 < y < ðB ¼ 6 mmÞ; t > 0 ð1Þ

where k is the thermal conductivity of steel and it is assumed thatthe material is isotropic, Cp is the specific heat and qS is the densityof steel.

In order to compute the surface heat flux and surfacetemperatures on the quench side of the hot plate, a commercial in-verse solver called INTEMP, developed by Trujillo and Busby [49], isapplied. There are some other surface temperature measurementtechniques like infrared thermography that is employed by the sci-entific community; however, it is not applicable for the currentsystem due to the difficulty in conducting measurements undersurface-water interaction environment, and also the vapor formedduring the cooling process disrupts the temperature measurement.As a result, the surface temperature and heat fluxes are obtainedusing inverse heat conduction technique, which uses the thermo-couple readings from inside of the plate.

The commercial inverse solver used in this study (INTEMP) re-quires the input data of heat transfer model in the form of a freefield format consisting of a geometrical field (nodes and elements),material properties and the measured time-temperature histories.Fig. 6 shows the computational domain with three thermocoupleholes used for inverse analysis. The boundary conditions employedare also shown in the figure; moreover, care has been taken toneglect the heat losses from all the sides of the plate except forthe quench surface using a ceramic insulator. The entire domainconsists of 3528 nodes with 3340 elements discretized with four-node rectangular elements of step sizes Dx = 0.0003 m andDy = 0.0006 m. Grid independency has also been verified for thegenerated mesh. The temperature dependent material propertiesof the AISI 304 steel have been considered from the study of[15]. The basic procedure in solving inverse heat conduction solu-tion by INTEMP is to predict surface heat flux and the surface tem-peratures. It initially assumes the surface heat flux as a knownboundary condition and thereafter all nodal temperatures are cal-culated as a standard conduction problem. The assumed surfaceheat fluxes are then modified by a nonlinear optimization tech-nique to minimize the error between the measured (input fromthermocouple) and the predicted nodal temperatures at the samelocations. In essence, the thermocouple data is not used to calcu-late the temperatures at that location in the computational do-main. It is rather used to find the correct surface heat flux values,which minimize the error. INTEMP uses the Crank–Nicolson for-mulation to solve the heat conduction model because of its higheraccuracy, and the error in estimated fluxes is minimized by usingthe least square method with an added regularization term. Oncethe heat flux reaches the optimum value, INTEMP predicts the cor-rect temperature at each nodal point. Moreover, in order to elimi-nate the noisy flux, INTEMP minimizes the least square error to avalue below 5 percent using an optimal smoothing parameter byL-curve analysis. This curve represents a plot of the mean squareerror versus the norm of the estimated heat fluxes. The detaileddescription about the INTEMP software and L-curve analysis arepresented in the literature [49,50].

As shown in Fig. 6, three thermocouples were employed in thepresent study to record the temperature data; therefore, based oneach thermocouple location INTEMP is designed to predict oneheat flux on the surface. As a result, the surface has been dividedinto three zones, and the heat flux in each zone is assumed to beuniform. Furthermore, from Fig. 6 it is evident that there are twotypes of flow regions found on the surface: a region dominatedby the direct spray impingement and two other regions dominatedby the horizontal flow of coolant. Therefore, based on the spraydiameter on the impingement surface, it is categorized into threedifferent heat transfer zones extending from 0 to 35 mm (q1), 35to 65 mm (q2) and 65 to 100 mm (q3). This is a similar kind of treat-ment of heat transfer zones as anticipated for the existence of fivedifferent zones by Zumbrunnen et al. [51] for the jet impingementcooling.

4.2. Validation of INTEMP

In this study, to check the accuracy of the nodal temperaturespredicted by INTEMP, an attempt has been made to compare thefinal predicted temperatures with the measured temperature his-tories at internal thermocouple locations of the plate. Moreover,for better validation of the surface temperature calculations by IN-TEMP, an air cooling experiment was conducted on the hot steelplate and the surface temperatures were measured with infraredthermography (FLIR, Model: A320, USA). Fig. 7(a) and (b) showsthe validation of INTEMP software, in which Fig. 7(a) expressesthe comparison of temperature histories obtained by INTEMP atthermocouple location 2 and the actual thermocouple history used,

Fig. 6. 2D-thermal model consists of three thermocouple holes used for data reduction.

Fig. 7. Validation of (a) estimated internal temperature by INTEMP and measured internal temperature by thermocouple 2, (b) measured surface temperature by infraredthermograph and estimated surface temperature by INTEMP.


and Fig. 7(b) depicts the comparison of estimated surface temper-ature by using INTEMP and infrared thermography. The experi-mental conditions used for this test were 30 Normal m3/h airflow rate and 60 mm nozzle-to-surface distance in the air coolingexperiment of hot plate surface. It can be observed from Fig. 7 thatthe results are in good agreement with each other and closely pre-dict the pulse in estimated temperatures. Therefore, it can be con-cluded that INTEMP can accurately predict the surface heat fluxand temperatures for the current heat transfer model used in thisstudy.

4.3. Uncertainty analysis

In the current study, an effort has been made to minimize theuncertainties in the experimental and computational results tomaximize the accuracy of the research outcome. While there areseveral methods to determine the experimental uncertainties, inthis research we have used an uncertainty methodology by ASME

test code PTC 19.8-1983 [52]. The two major sources of error inthe current work are: (1) error arising from fault in the measuringinstrument (bias) and, (2) error due to lack of preciseness in defi-nition of measured quantity (random or precision error). The mainuncertainties in the results are the ‘K’ type thermocouples whichare used to measure the real time temperature data during exper-imentation. The bias in temperature measurement is ±2.2 �C with azero precision error. There are also some errors in the values oftemperature recorded by the data card through the thermocoupleextension wire. A specially calibrated thermocouple wire is used inthis study to record the instantaneous readings measured by thethermocouple and effort has been made to restrict the uncertaintydue to thermocouple wire within ±2.4 �C. The uncertainty due tolack of contact between the thermocouple tip and the plate hasbeen avoided by using a highly conductive material in thermocou-ple holes to fill the isolated air gaps. The bias due to thermocouplelocation is 1%. A data acquisition system (NI-9211) manufacturedby National Instruments Co., USA) is used in this study for the


temperature recording. As reported by the manufacturer, the offsetin the temperature measurement is +0.7 �C and the sensitivity is<0.07 �C.

An inverse heat conduction analysis software called INTEMP hasbeen used to calculate the surface temperature and surface heatflux. It eliminates the noise present in the boundary values by opti-mal smoothing parameter using L-curve methodology which re-ports the uncertainties in the output values. Based on the valueof optimal smoothing parameter, the precision error in surface heatflux estimation is 3%. All the experiments have been conductedthrice and the standard deviation in cooling rate estimation hasbeen found to be 1.38%.

In this study the thermo-physical properties such as surfacetension, viscosity and thermal conductivity have been measuredto gain the basis data related to heat transfer enhancement inpolymer added water coolants as compared to pure water. Theuncertainty associated with surface tension, viscosity and thermalconductivity are based on measured data obtained from a cali-brated tensiometer, viscometer and thermal property analyzer.From the calibrated instruments, the accuracy in the surface ten-sion, viscosity and thermal conductivity is ±0.2%, ±0.3% and±1.82%, respectively.

5. Results and discussion

5.1. Visual analysis

A video camera (Sony, Model No: DCR-SX63, Japan) with a sam-pling frequency of 30 frames per second has been employed to re-cord the cooling experiments. These videos are further processedinto individual frames for analyzing the progression of coolingzone on the test surface. For example, Fig. 8(a) and (b) shows theseries of video images extracted with in a specific time interval,in which Fig. 8(a) is for pure water as coolant and Fig. 8(b) for poly-mer added water coolant (10 ppm) experiments. It can be said thatthe hot plate is bright red in color with a temperature of around900 �C before the spray cooling starts. At an instant when the sprayimpinges on the hot surface, the stagnant zone turns dark and ap-pears as a disc, as shown in Fig. 8 at a time of 0.02 s. Shortly afterthe cooling proceeds, the dark zone quickly spreads radially out-wards. This dark zone signifies that the coolant is in direct contactwith the hot surface; therefore, a very high value of heat flux isachieved within a short period of time and also significant temper-ature drop on the surface can be observed. It can be said that con-siderable evaporation of spray droplets has been observed andmany tiny steam bubbles can be seen growing from the peripheryof the dark zone. The visible periphery of the dark zone is called thewetting front. During the progression of cooling, these bubblescoalesce and prevent the motion of the wetting front outwardsalong the radial direction. It has been noticed by Abdalrahmanet al. [53] that the speed of wetting front is responsible for thequench ability of the hot surface. It can be compared fromFig. 8(a) and (b) that the motion of wetting front is higher in caseof polymer added water coolant than that of pure water. The quan-titative measurements of wetting front propagation on the basis ofvideo images shown in Fig. 8 are also reported in Fig. 9. The reasonfor this result is that the addition of polymer decreases the surfacetension due to adsorption at the air-water interface which resultsin a decrease in contact angle; hence, the droplets will spread onthe surface at a faster rate with a higher evaporation rate. It shouldbe noted that higher wettability could be suitably expressed by alower contact angle between a solid surface and impinged liquiddroplets. Moreover, in surfactant or polymer added water, the ad-sorbed surfactant molecules (here polymer acts as a surfactant)prevent the bubble coalescence by rupturing the thin liquid films

trapped between the bubbles; therefore, the spreadability of wet-ting front will be higher. This statement is well in agreement withthe study of [54].

In the experimental study, the concentration range of polymeradditive used is 0–150 ppm; therefore, it is important to calculatethe wetting front velocity for indicating how these coolants spreadto contribute effective heat removal from the hot surface. Thus, thewetting front velocity is calculated on the basis of the video framesusing image analysis software. Fig. 10 depicts the wetting frontvelocity against the polymer additive concentration in water cool-ant. It can be seen from this figure that the wetting front velocityincreases monotonically with an increase polymer additive con-centration. However, it increases sharply in concentration rangesof 0–10 ppm and 10–90 ppm; thereafter increases intermediatelyup to 110 ppm and then it changes relatively slowly or shows nochange up to 150 ppm. The possible reason for the wettabilityimpedance at higher polymer concentration is due to critical poly-mer concentration of the PVP where molecular agglomerationtakes place. It results in gradual increase in surface tension and vis-cosity of the coolant as shown in Figs. 2 and 4, which causes anobstacle against the wettability of liquid over the dry region.

5.2. Experimental and inverse heat conduction analysis

During the cooling experiments, the transient internal temper-ature histories have been measured at three different locationsusing three subsurface thermocouples placed across the plate asdescribed earlier. This measured temperature data further hasbeen subsequently used to calculate the surface temperature andheat flux histories using inverse heat conduction software INTEMP,as discussed in Section 4. In air-atomized polymer added waterspray cooling, the most important aspect affecting the heat transferrate of the hot surface is polymer additive concentration whichwas varied from 0 to 150 ppm. Fig. 11 presents the measured timehistories of each thermocouple readings (internal temperatures)from experiments for four different cases: pure water, low concen-tration (10 ppm), medium concentration (70 ppm), and high con-centration (110 ppm). Similarly, Figs. 12 and 13 show thecalculated surface temperature and heat flux corresponding tothe thermocouple location, both obtained from inverse heat con-duction analysis by INTEMP at three different locations acrossthe plate. The cooling data reported here is from a temperatureabove 950 �C; thermocouple 2 (T/C 2) is at the stagnant locationof the spray impingement, thermocouple 3 (T/C 3) is nearer tothe stagnant point than thermocouple 1 (T/C 1) as clearly indicatedin Fig. 1(b).

It can be seen from Fig. 11 that the temperature of the plate isfairly uniform before spray hits the hot surface and all the temper-ature curves coincide well, which indicates that proper thermalsoaking has taken place inside the furnace. As soon as the spray im-pinges on the hot surface, a remarkable and sudden fall in temper-ature at all locations has been observed, and it gradually decreasesas the cooling proceeds. However, the rate of temperature drop isfastest at thermocouple location 2, while it is faster in location 3than that of location 1. The reason for this result is that wettingfront takes a particular period of time to cover the locations awayfrom the stagnant zone; therefore, the heat transfer rate decreaseswith the locations away from the stagnant zone of the sprayimpingement. It can be seen that the rate of temperature drop atall three locations of the plate is higher in case of polymer additivewater than that of pure water taken alone and also this effect in-creases with increase in polymer additive concentration as ob-served from Fig. 11(b)–(d).

As discussed earlier, Fig. 12 shows the corresponding transientsurface temperature results obtained from the inverse heat con-duction. By comparing the results of the temperature profiles in

Fig. 8. Visual analysis of the cooling experiments conducted with (a) pure water, (b) polymer added water.


Fig. 11 with Fig. 12, the response in a surface temperature drop andmagnitude of fall at each location exhibits a faster rate than theinternal temperature profiles. This is due to contact heat transferbetween the coolant and the tested surface.

Fig. 13 shows the variation in calculated surface heat flux withcooling time corresponding to the location of each thermocouple,obtained by the inverse heat conduction analysis with INTEMPsoftware. It can be seen from the figure that the trends of all theheat flux curves are identical and initially showing very small heatflux value closer to zero. After the spray hits the hot surface, theheat flux rises monotonically and reaches a maximum value,known as critical heat flux, due to the onset of transition boiling.Thereafter, the heat flux decreases with further increase in coolingtime and this regime is known as nucleate boiling. Finally when the

surface attains low temperature, a single phase forced convectionregime has been observed. As observed from Fig. 13, the thermo-couple 1 is the farthest from stagnant location (thermocouple 2)than thermocouple 3. Therefore, it can be found that the most sig-nificant heat transfer is experienced at the direct spray impinge-ment zone and it decreases with the locations away from thestagnant zone. The critical heat flux also decreases with the loca-tions away from the spray impingement zone. In the boiling heattransfer with pure water, during cooling of very hot surfaces, filmboiling prevails at high surface temperatures, which separates thesolid surface by a continuous stable vapor layer causing weakeningof the heat exchange intensity to the coolant. As the temperaturefalls, transition boiling ensues which is an intermittent contact ofsolid surface and liquid coolant; thereafter, nucleate boiling is

Fig. 9. Comparisons of the wetting front moment during cooling with pure waterand polymer added water coolants. The surface area of plate used in theexperimental study is 100 mm � 100 mm.

Fig. 10. Effect of polymer additive concentration on wetting front velocity.


followed by single-phase convection at low surface temperatureswhich possesses high cooling efficiency due to direct heatexchange between solid surface and liquid coolant. However, air-atomized spray cooling has an advantage over other cooling tech-niques in that it is not affected by the film boiling heat transfermechanism. Here, the finer droplets of water with higher momen-tum touch the hot surface, get deformed increasing the contactarea; the conduction heat transfer takes place through the contactarea (common interface), hence the water temperature inside thedroplet rises and it starts evaporating. Before the droplet evapo-rates completely, they are swept away from the plate surface byhigh velocity superposed air flow. Hence, the formation of stablevapor film is prevented as there is no occurrence of water poolon the hot surface unlike the other cooling techniques. Thereforein the current research, at high surface temperatures transitionboiling prevails; as the temperature decreases, nucleate boiling en-sues and finally single phase convection follows at low surfacetemperatures. In case of transition boiling, film boiling and nucle-ate boiling will coexist on the hot surface which makes the vaporlayer unstable for exchanging heat between solid surface and thecooling water. In this regard, vapor layer thickness plays a domi-nant role for high heat flux intensity. Moreover, the contact area

of the water droplets is also important for easier evaporation bytaking higher heat energy from the solid surface.

Fig. 13(b) depicts results similar to those in Fig. 13(a), but withthe difference that the coolant is water added with polymer of10 ppm concentration. The trends of all the curves in the figureare qualitatively similar to those of pure water coolant with a var-iation in quantitative values. It is obvious that the addition of poly-mer enhances the heat transfer rate as discussed in Section 5.1.This can be compared from the heat flux curves in Fig. 13(a) and(b) that the increase in critical heat flux value with polymer addi-tive water varies between 8.3% (T/C 1) and 35.9% (T/C 2) comparedwith that of pure water over the surface. This indicates that moreeffective heat transfer can be achieved on the entire plate surfaceusing polymer added water. The reason is that the vapor layer inthe transition boiling regime is more unstable due to higher rateof nucleus formation by polymer added water spray and this resultin quick movement of coolant to the downstream spatial locationswith a higher bubble coalescence rate. This is well in agreementwith the study of [55] that lower surface tension creates betterconditions for nucleate boiling. The visual observation presentedin Figs. 8 and 9 is an example of the faster movement of polymeradded water to the downstream locations of the test surface thanthat of pure water. In the case of polymer added water, the de-crease in surface tension causes the atomization of the dropletsinto finer size and these droplets can gain higher momentum bycompressed air, which can easily rupture the vapor layer apart,and the droplet can touch a wider contact surface area formed bya decrease in surface tension of water. In addition, when the drop-let of polymer added water touches the plate surface, the waterevaporates without any delay, and the polymer molecules get re-tained on the vapor layer because of its higher melting point(PVP of �200 �C) which disrupts the vapor layer by allowing thesubsequent droplets to touch the solid surface. This results in aquick shift from transition to nucleate boiling with a higher heatenergy removal in transition boiling regime while working withpolymer added water where nucleate boiling is dominant, asshown in Fig. 13(b). Furthermore, the role of surface tension inthe breaking up of vapor film and heat transfer augmentation can-not be ignored in the context of high temperature boiling transi-tion. Decrease in surface tension by the addition of polymer(PVP) could be another possible reason of increase in transitionboiling heat transfer because it is the governing parameter behindthe vapor film wavelength of instability, as first explained by Zuber[56] to be governed by the Taylor and Helmholtz instabilities.Therefore, the hydrodynamic instability length scale can be ex-pressed interms of surface tension (r) and densities of liquid (qL)and gas (qG)as follows:

k ¼ rgðqL � qGÞ

� �1=2

ð2Þ

The calculated interfacial wavelength scale for normal water isaround 2 mm and it is reduced with a reduction in surface tensionas it is directly proportional to the square root of a surface tensionvalue. Overall, the above mentioned underlying mechanisms willcoexist which accounts for the experimental observation of heattransfer enhancement during working with polymer added watercoolants.

Similarly, Fig. 13(c) and (d) represents the heat transfer behav-ior like in Fig. 13(a), but the only variation is the concentration ofpolymer additive in water. From these results, it can be said thatcooling capacity can be enhanced by the polymer concentration.The heat transfer rate at the spray impingement and parallel flowzones exhibit an increase due to an increase in the polymer addi-tive concentration. This is obvious when comparing the heat fluxcurves in Fig. 13(c) and (d) with Fig. 13(b). As can be seen from

Fig. 11. Transient internal temperature distribution, (a) pure water, (b) polymer added water of 10 ppm, (c) polymer added water of 70 ppm, (d) polymer added water of110 ppm.


the figures, the critical heat flux value at each location increaseswith an increase in concentration of polymer additive. More pre-cisely, at the stagnant zone, the critical heat flux value with110 ppm concentration of polymer additive water is higher thanabout 12.6% of the critical heat flux of 10 ppm polymer additiveconcentrations shown in Fig. 13(d); whereas, the CHF at 70 ppmpolymer additive concentrations is 4.3% higher than that of criticalheat flux of Fig. 13(b). The reason for the increase in heat transferrate from the entire surface is that the decrease in surface tensionof water to several orders with an increase in polymer additiveconcentration plays an important role in size, evaporation rateand contact area of the droplets, and the coalescence rate of bubbleon the quench surface. The increase in wetting front velocityshown in Fig. 10 resembles the highest heat transfer rate in poly-mer added coolant.

Overall, it can be noted from Figs. 11–13 that in all the coolingexperiments especially the direct spray impingement zone contrib-utes a maximum amount of heat transfer rate than that attainedfrom parallel flow zones. Therefore, then boiling in this zone hasbeen considered particularly for analyzing the experimental resultsin later sections.

5.2.1. Boiling curveFig. 14 shows the effect of polymer additive concentration on a

boiling curve at the impingement zone. This curve represents thechange in heat flux of the surface with temperature. For industrial

applications, particularly for heat treatment of steels consists of acooling process designed to alter the phase microstructure for aspecialized mechanical property application with the rate of tem-perature fall between 900 to 600 �C and 900 to 200 �C being espe-cially important. Therefore, in Fig. 14, the cooling results from 900to 200 �C temperature has been reported. It can be seen from thefigure that all the curves start from about 900 �C at the bottomright corner of the plot. As the cooling proceeds, they move tothe left in a qualitatively similar manner while taking differentpaths for different polymer concentrations. It can be observed thatall the cases follow two major boiling heat transfer regimes, suchas transition boiling at the early stages of cooling followed bynucleate boiling at a lower surface temperature. This is in accor-dance with the results of [57] that there is a change of regimesoccurring in the boiling curve. At high surface temperatures, theheat is transferred through the discontinuous vapor layer by forcedconvection and monotonically attains a high value called criticalheat flux. Then, complete nucleate boiling starts on the surfaceand the bubbles grow in size due to coalescence and initiate a con-tinuous thin water film on the hot surface with a decrease in sur-face temperature. As a result, the heat flux decreases sharply with areduction in surface temperature which evident from the boilingcurve.

It should be mentioned here that the typical film boiling regime,which is characterized by a declining heat flux with a decrease insurface temperature at the initial stage of cooling, is not obtained

Fig. 12. Transient surface temperature data, (a) pure water, (b) polymer added water of 10 ppm, (c) polymer added water of 70 ppm, (d) polymer added water of 110 ppm.


in the boiling curves for air-atomized spray cooling as depicted inFig. 14. This is well in agreement with the findings related to air-atomized spray cooling that the superposed air used for atomiza-tion helps to sweep away the partially evaporated droplets withoutallowing them to accumulate on the hot surface forming liquidfilms [11,13].

By referring the results in Fig. 14, early shift from transition tonucleate boiling with a rise in critical heat flux value occurs whileworking with different concentrations of polymer added watercoolants. This conveys that the most effective boiling heat transfermechanism exists using polymer additive water coolant than thatof pure water. It can be seen that the peak-and-troughs in boilingcurve occurs early up to a particular concentration of polymeradditive in water, beyond which a further increase in concentra-tion causes a shrink in the curve to lower heat flux values. It resem-bles that too low or too high concentration of polymer additivemay not be suitable for higher cooling performances. Therefore,optimum concentration is observed for polymer additive used inthis study. This is in accordance with the results of [48,58]. Chenget al. [37] reported the state-of-the-art review on boiling phenom-ena with surfactants and polymeric additives and also shown thatthe enhancement in boiling heat transfer occurs up to a certain va-lue of surfactant or polymer and then decreases relatively. The rea-son for the occurrence of optimum concentration for betterenhancement is that a decrease in surface tension due to increasein concentration is more pronounced up to 110 ppm beyond which

critical polymer concentration (CPC) exists as shown in Fig. 2(a).Moreover, by referring to the viscosity and thermal conductivityresults shown in Figs. 4 and 5, the viscosity rise and thermal con-ductivity fall at a higher polymer concentration makes the heattransfer enhancement lower. At concentrations below 110 ppm,no significant changes in the viscosity and thermal conductivityare found; therefore, surface tension only plays the dominant rolein the heat transfer enhancement. As discussed earlier, by loweringthe surface tension of liquid in contact with the solid surface, thedroplet size is decreased, and the contact area of droplet (seeFig. 3), bubble nucleation and coalescence rates are also increased.The study of [20] reveals that droplets of lower size facilitate fasterevaporation such that the cooling intensity of steel strip increases.In order to quantify the effect of surface tension and viscosity ofwater on droplet atomization and to calculate the sauter meandiameter (SMD), the following correlation has been proposed bythe Semião et al. [59].

SMD ¼ 1:58� 103 rqAU2

Ad0

" #0:5

d0r

lLUA

� �0:55 qA

qL

� �1þ 1

AFR

� �0:5

þ 166lL

qLd0UA

� �1:1 rqAd0U2

A

" #0:2

d0qA

qL

� �0:35

1þ 1AFR

� ��0:48

ð3Þ

where qL and qA are the densities of the liquid and air, respectively,r is the surface tension, lL is the viscosity of liquid, UA is the air

Fig. 13. Transient surface heat flux data, (a) pure water, (b) polymer added water of 10 ppm, (c) polymer added water of 70 ppm, (d) polymer added water of 110 ppm.

Fig. 14. Heat flux as a function of surface temperature for different polymeradditive water coolants.


velocity, d0 (i.e., 0.005 m) is the orifice diameter and finally AFR isthe ratio of air and water. Using the above correlation, the effectof polymer additive concentration in water on the SMD of theatomized droplets has been calculated and plotted in Fig. 15. The

uncertainty in the calculated SMD is ±0.51%, which arises fromthe error in the liquid properties measured by calibrated instru-ments. It can be seen from Fig. 15 that a decrease in surface tensionby the addition of polymer added water reduces the droplet size toone fourth of the droplet diameter of pure water. On the other hand,very small increase in viscosity has an insignificant effect in increas-ing in the size of the droplets.

5.2.2. Critical heat flux and temperaturesTo understand the effect of polymer additive on heat transfer

enhancement from the test surface, the critical heat flux and criti-cal temperature values at the stagnant surface location has beenpresented in Fig. 16. In the figure, the primary y-axis shows the ef-fect of polymer concentration on the critical heat flux, whereas sec-ondary y-axis corresponds to critical surface temperature. As canbe derived from these figures, the critical heat flux and critical sur-face temperature increase by using different concentrations ofpolymer additive coolants. At a concentration of 110 ppm of poly-mer additive, the critical heat flux at stagnant location is as high as4.212 MW/m2, which is 53.67% higher than that of pure water atthe same location. The critical surface temperature at this polymerconcentration is 569 �C which corresponds to a thinner vapor layeron the plate surface as compared to pure water (374 �C). This iswell in agreement with the study of [60] that the vapor layer whichis formed by the droplet impingement on the hot surface is propor-tional to the initial droplet diameter. Therefore, the droplets of

Fig. 15. Effect of surface tension and viscosity modified by polymer additiveconcentration on droplet diameter.

Fig. 16. Effect of polymer additive concentration on critical heat flux andcorresponding surface temperatures.

Fig. 17. Variation in ultrafast cooling rate with the concentration of polymeradditive in water.


lower size facilitates the thinner vapor layer on the hot surface. Thethinner the vapor layer, the higher will be the heat transfer rate athigh surface temperatures due to the easier rupturing of the vaporlayer thereby allowing the spray droplets to come in direct contactwith the hot surface. Which creates a bubble nucleation environ-ment, so a nucleate boiling is also responsible for higher heat fluxvalues. Although polymer concentration above 110 ppm enhancesthe heat transfer rate compared to that of pure water, but the de-gree of enhancement is low as the reduction in surface tension at110 ppm concentration is maximum.

5.3. Ultrafast cooling rate for ROT application

Since the microstructure and mechanical properties of steelafter being subjected to hot rolling processes significantly dependson the cooling rate between 900 and 600 �C temperatures on therunout table (ROT) of Hot Strip Mill in the steel industry. Hencethis temperature range has been considered to define the coolingrates for the current experimental study and has been reportedin Fig. 17 as a function of the polymer concentration in water. Anenhancement factor, which is defined as the ratio between thecooling rate with additives and that of pure water, has been intro-duced and reported in the secondary y-axis of Fig. 17. It is observed

from the figure that air-atomized spray cooling with pure watercoolant has an ability to give the ultrafast cooling rate of 167 �C/s, which is in accordance with the previous results by the authors[33,61]. In the current study, even an addition of 10 ppm of poly-mer additive to water has increased the cooling rate to 195 �C/s,which increases further with an increase in polymer concentrationup to 110 ppm and attains a maximum value of 253 �C/s. Thereaf-ter, the cooling rate shows a downward trend with an increase inconcentration of polymer additive up to 150 ppm. The reason forthis behavior is the change in thickness of the vapor layer on theplate surface with a change in polymer additive concentration inwater. Therefore, the results indicate that the optimal cooling ratefor polymer added water has been attained at 110 ppm concentra-tion and that the obtained ultra fast cooling rate is as high as253 �C/s, which is 51.5% higher than that of pure water at the sameoperating conditions.

The authors have recently carried out an extensive research onthe enhancement of cooling performances of pure water usingdifferent types of surfactant additives such as sodiumdodecyl sul-phate of anionic, cetramonium bromide of cationic and polysor-bate: 20 (Tween 20) of non-ionic for the current experimentalset-up under the same operating conditions [33]. It is observedthat among the surfactants used, non-ionic type viz. Tween 20can have better ultrafast cooling capability, as high as 214 �C/s,on the hot AISI 304 steel plate at that optimum concentration.The present results can be compared with the results of non-ionicsurfactant case, the cooling rate at optimum concentration ofwater based polymer (PVP) being more than 18.2% that of non-io-nic surfactant (Tween 20). This might be due to better coolantcharacteristics (-or properties) of polymer (PVP) additive as com-pared to surfactant (Tween 20).

6. Conclusions

In this study, the experimental results concerned with air-atomized spray cooling of a stainless steel plate at temperatureabove 900 �C, cooled by polymer added water, covering the transi-tion and nucleate boiling regime have been presented. The poly-mer additive used was polyvinylpyrrolidone (PVP), which can actlike a surfactant. Tests have been carried out at different concen-trations of polymer additive ranging between 10 and 150 ppm.The cooling intensity obtained by polymer added water coolantsdepicts a large enhancement when compared with the results ac-quired using pure water. In all, the hot surface cooled by polymeradded water coolant has significantly higher heat fluxes than by


using pure water depending on the polymer, and the optimumconcentration is sought for higher heat transfer rate. The physicalproperties of the coolant used play a key role in the enhancementof heat transfer rate from the heated surface.

The following conclusions can be drawn from the results ofpresent experimental investigations:

1. In general, film boiling prevents during the cooling of surfacesat high surface temperatures; it causes a large reduction inthe heat transfer rate. However, in the case of air-atomizedspray cooling, film boiling heat transfer mechanism does notappear as high flow rate of air used for atomization helps inbreakdown of vapor film by sweeping the partially evaporateddroplets from the hot surface before they accumulate on thetest surface; hence, the formation of a stable vapor film isprevented.

2. While working with polymer added water coolants, the heattransfer rate increases with an increase in polymer concentra-tion, particularly in transition boiling heat transfer regime.Quick shift from transition boiling to nucleate boiling occurswith high heat transfer rate in the former. This is due todecrease in surface tension by the addition of different concen-tration ranges of polymer in water. Surface tension plays a dom-inant role in the vapor film break up and heat transferaugmentation because it is the governing parameter behindthe vapor film wave length of instability. Moreover, a decreasein surface tension atomizes the droplets into finer sizes, whichcan easily evaporate from the surface due to a large contact areawith the solid surface.

3. In transition boiling, the molecules of polyvinylpyrrolidone(PVP) dispersed in spray droplets helps in disrupting the vaporfilm and allow them to touch the solid surface due its highmelting point as compared to the boiling point of water. There-after, these molecules settle over the hot surface and enhancethe nucleation sites for better nucleate boiling environment intransition boiling regime.

4. Cooling intensity is highly enhanced by the polymer additive upto an optimum concentration value. Further increase in poly-mer concentration does not enhance the heat transfer signifi-cantly, thereby making it ineffective as a cooling medium. Theoptimum polymer concentration for the system under study is110 ppm. The decrease in heat transfer enhancement beyondthis is that due to the attainment of critical polymer concentra-tion (CPC) value for PVP.

5. The critical heat flux value increases with an increase in poly-mer additive concentration, a maximum of 4.212 MW/m2

occurs at 110 ppm of polymer added water which is 53.67%higher than that of pure water at the same surface location.The critical heat flux for polymer coolants exist at much highertemperatures as compared to pure water.

6. For application in steel industries to design the required metal-lurgical microstructures for enhanced mechanical properties ofsteel products, the cooling rate between austenitization tem-perature (rolling temperature in Hot Strip Mill, which is approx-imately 900 �C) and coiling temperature (600 �C) is calculated.A maximum cooling rate of 253 �C/s is found at an optimumconcentration of polymer additive (110 ppm). Hence, theenhancement in cooling rate by using polymer additive showsa maximum of 51.5% as compared to pure water coolant. There-fore, the cooling rates obtained while working with polymeradded water coolants are in the higher range of an ultrafastcooling process. The ultrafast cooling rate at an optimum con-centration of water-based polymer (PVP) is more than 18.2%that of non-ionic surfactant (Tween 20) studied in the authors’earlier work using the current experimental set-up under sameoperating conditions.

References

[1] S.V. Ravikumar, J.M. Jha, S.S. Mohapatra, S.K. Pal, S. Chakraborty, Influence ofultrafast cooling on microstructure and mechanical properties of steel, SteelRes. Int. 84 (11) (2013) 1157–1170.

[2] S.D. Cox, S.J. Hardy, D.J. Parker, Influence of runout table operation setup on hotstrip quality, subject to initial strip condition: heat transfer issues, IronmakingSteelmaking 28 (5) (2001) 363–372.

[3] J.C. Herman, Impact of new rolling and cooling technologies onthermomechanically processed steels, Ironmaking Steelmaking 28 (2) (2001)159–163.

[4] A. Lucas, P. Simon, G. Bourdon, J.C. Herman, P. Riche, J. Neutjens, P. Harlet,Metallurgical aspects of ultra fast cooling in front of the down-coiler, Steel Res.Int. 75 (2) (2004) 139–146.

[5] J. Wendelstorf, K.H. Spitzer, R. Wendelstorf, Spray water cooling heat transferat high temperatures and liquid mass fluxes, Int. J. Heat Mass Transfer 51 (19–20) (2008) 4902–4910.

[6] J. Wendelstorf, R. Wendelstorf, K.H. Spitzer, Spray cooling heat transfer andcalculation of water impact density for cooling of steel sheet materials byinverse process modelling, Steel Res. Int. 80 (9) (2009) 639–644.

[7] U. Alam, J. Krol, E. Specht, J. Schmidt, Enhancement and local regulation ofmetal quenching using atomized sprays, J. ASTM Int. 5 (10) (2008).

[8] G.G. Nasr, R.A. Sharief, A.J. Yule, High pressure spray cooling of a movingsurface, J. Heat Transfer Trans. ASME 128 (8) (2006) 752–760.

[9] J. Fukai, Y. Shiiba, O. Miyatake, Theoretical study of droplet impingement on asolid surface below the Leidenfrost temperature, Int. J. Heat Mass Transfer 40(10) (1997) 2490–2492.

[10] J.D. Bernardin, I. Mudawar, Film boiling heat transfer of droplet streams andsprays, Int. J. Heat Mass Transfer 40 (11) (1997) 2579–2593.

[11] F. Puschmann, E. Specht, Transient measurement of heat transfer in metalquenching with atomized sprays, Exp. Therm. Fluid Sci. 28 (6) (2004) 607–615.

[12] S.S. Mohapatra, S.V. Ravikumar, S.K. Pal, S. Chakraborty, Ultra fast cooling of ahot steel plate by using high mass flux air atomized spray, Steel Res. Int. 84 (3)(2013) 229–236.

[13] M. De Oliveira, J. Ward, D.R. Garwood, R.A. Wallis, Quenching of aerospaceforgings from high temperatures using air-assisted, atomized water sprays, J.Mater. Eng. Perform. 11 (1) (2002) 80–85.

[14] M.S.A. Oliveira, A.C.M. Sousa, Neural network analysis of experimental data forair/water spray cooling, J. Mater. Process. Technol. 113 (1–3) (2001) 439–445.

[15] H. Wang, W. Yu, Q. Cai, Experimental study of heat transfer coefficient on hotsteel plate during water jet impingement cooling, J. Mater. Process. Technol.212 (9) (2012) 1825–1831.

[16] H. Bostanci, D.P. Rini, J.P. Kizito, V. Singh, S. Seal, L.C. Chow, High heat fluxspray cooling with ammonia: investigation of enhanced surfaces for CHF, Int. J.Heat Mass Transfer 55 (13–14) (2012) 3849–3856.

[17] S.-J. Chen, A.A. Tseng, Spray and jet cooling in steel rolling, Int. J. Heat FluidFlow 13 (4) (1992) 358–369.

[18] M.J. Cho, B.G. Thomas, P.J. Lee, Three-dimensional numerical study ofimpinging water jets in runout table cooling processes, Metall. Mater. Trans.B 39 (4) (2008) 593–602.

[19] H.M. Al-Ahmadi, S.C. Yao, Spray cooling of high temperature metals using highmass flux industrial nozzles, Exp. Heat Transfer 21 (1) (2008) 38–54.

[20] P. Bhattacharya, A.N. Samanta, S. Chakraborty, Spray evaporative cooling toachieve ultra fast cooling in runout table, Int. J. Therm. Sci. 48 (9) (2009) 1741–1747.

[21] W. Jia, H.H. Qiu, Experimental investigation of droplet dynamics and heattransfer in spray cooling, Exp. Therm. Fluid Sci. 27 (7) (2003) 829–838.

[22] M. Gradeck, A. Ouattara, D. Maillet, P. Gardin, M. Lebouché, Heat transferassociated to a hot surface quenched by a jet of oil-in-water emulsion, Exp.Therm. Fluid Sci. 35 (5) (2011) 841–847.

[23] S. Mitra, S.K. Saha, S. Chakraborty, S. Das, Study on boiling heat transfer ofwater–TiO2 and water–MWCNT nanofluids based laminar jet impingement onheated steel surface, Appl. Therm. Eng. 37 (2012) 353–359.

[24] Q. Cui, S. Chandra, S. McCahan, The effect of dissolving salts in water spraysused for quenching a hot surface: Part 2 – Spray cooling, J. Heat Transfer Trans.ASME 125 (2) (2003) 333–338.

[25] S.S. Mohapatra, S.V. Ravikumar, A. Verma, S.K. Pal, S. Chakraborty,Experimental investigation of effect of a surfactant to increase cooling of hotsteel plates by a water jet, J. Heat Transfer Trans. ASME 135 (3) (2013)032101–032107.

[26] S.C. Tinker, M. di Marzo, P. Tartarini, S. Chandra, Y.M. Quiao, Dropwiseevaporative cooling: effect of dissolved gases and surfactants, in: D.P. Lund,E.A. Angell (Eds.), International Conference on Fire Research and Enineering,Boston, 1995, pp. 91–96.

[27] T.J. Hendricks, S. Krishnan, C. Choi, C.H. Chang, B. Paul, Enhancement of pool-boiling heat transfer using nanostructured surfaces on aluminum and copper,Int. J. Heat Mass Transfer 53 (15–16) (2010) 3357–3365.

[28] T. Inoue, Y. Teruya, M. Monde, Enhancement of pool boiling heat transfer inwater and ethanol/water mixtures with surface-active agent, Int. J. Heat MassTransfer 47 (25) (2004) 5555–5563.

[29] V.M. Wasekar, R.M. Manglik, Pool boiling heat transfer in aqueous solutions ofan anionic surfactant, J. Heat Transfer Trans. ASME 122 (4) (2000) 708–715.

[30] W. Cheng, B. Xie, F. Han, H. Chen, An experimental investigation of heattransfer enhancement by addition of high-alcohol surfactant (HAS) anddissolving salt additive (DSA) in spray cooling, Exp. Therm. Fluid Sci. 45(2013) 198–202.

http://refhub.elsevier.com/S0017-9310(14)00052-0/h0005


















































































[31] S.S. Mohapatra, S.V. Ravikumar, S. Andhare, S. Chakraborty, S.K. Pal,Experimental study and optimization of air atomized spray with surfactantadded water to produce high cooling rate, J. Enhanced Heat Transfer 19 (5)(2012) 397–408.

[32] S.S. Mohapatra, J.M. Jha, K. Srinath, S.K. Pal, S. Chakraborty, Enhancement ofcooling rate for a hot steel plate using air-atomized spray with surfactant-added water, Exp. Heat Transfer 27 (1) (2013) 72–90.

[33] S.V. Ravikumar, J.M. Jha, I. Sarkar, S.S. Mohapatra, S.K. Pal, S. Chakraborty,Achievement of ultrafast cooling rate in a hot steel plate by air-atomized spraywith different surfactant additives, Exp. Therm. Fluid Sci. 50 (2013) 79–89.

[34] Y.M. Yang, J.R. Maa, Pool boiling of dilute surfactant solutions, J. Heat TransferTrans. ASME 105 (1) (1983) 190–192.

[35] Y.M. Qiao, S. Chandra, Experiments on adding a surfactant to water dropsboiling on a hot surface, Proc. R. Soc. London, Ser. A Math. Phys. Eng. Sci. 453(1959) (1997) 673–689.

[36] E.R. Morgan, T.E. Dancy, M. Korchynsky, Improved high strength low alloysteels through hot strip mill controlled cooling, Blast Furnace Steel Plant 53(10) (1965) 921–929.

[37] L. Cheng, D. Mewes, A. Luke, Boiling phenomena with surfactants andpolymeric additives: a state-of-the-art review, Int. J. Heat Mass Transfer 50(13–14) (2007) 2744–2771.

[38] P. Kotchaphakdee, M.C. Williams, Enhancement of nucleate pool boiling withpolymeric additives, Int. J. Heat Mass Transfer 13 (5) (1970) 835–848.

[39] Z. Mingzheng, X. Guodong, L. Jian, C. Lei, Z. Lijun, Analysis of factors influencingthermal conductivity and viscosity in different kinds of surfactant solutions,Exp. Therm. Fluid Sci. 36 (2012) 22–29.

[40] J. Águila-Hernández, A. Trejo, B.E. García-Flores, Volumetric and surfacetension behavior of aqueous solutions of polyvinylpyrrolidone in the range(288 to 303) K, J. Chem. Eng. Data 56 (5) (2011) 2371–2378.

[41] J. Zhang, R.M. Manglik, Nucleate pool boiling of aqueous polymer solutions ona cylindrical heater, J. Non-Newtonian Fluid Mech. 125 (2–3) (2005) 185–196.

[42] R.M. Manglik, V.M. Wasekar, J. Zhang, Dynamic and equilibrium surfacetension of aqueous surfactant and polymeric solutions, Exp. Therm. Fluid Sci.25 (1–2) (2001) 55–64.

[43] R.Y.Z. Hu, A.T.A. Wang, J.P. Hartnett, Surface tension measurement of aqueouspolymer solutions, Exp. Therm. Fluid Sci. 4 (6) (1991) 723–729.

[44] D. Li, M.A. Wells, Effect of subsurface thermocouple installation on thediscrepancy of the measured thermal history and predicted surface heat fluxduring a quench operation, Metall. Mater. Trans. B 36 (3) (2005) 343–354.

[45] V.B. Fainerman, D. Möbius, R. Miller, Surfactants: Chemistry, InterfacialProperties, Applications, Elsevier Science, 2002 (p. 678).

[46] K. Robert Lange, Surfactants: A Practical Handbook, Hanser Gardner Pubns,1999 (p. 237).

[47] J.W. Westwater, in: Thomas B. Drew, John W. Hoopes Jr., Theodore Vermeulen(Eds.), Advances in Chemical Engineering, vol. 6, Academic Press, New York,1966, p. 455 ($17.50, AIChE J. 13(4) (1967) 827–828).

[48] D. Klein, G. Hetsroni, A. Mosyak, Heat transfer characteristics of water and APGsurfactant solution in a micro-channel heat sink, Int. J. Multiphase Flow 31 (4)(2005) 393–415.

[49] D.M. Trujillo, H.R. Busby, Practical Inverse Analysis in Engineering, CRC Press,1997 (pp. 235).

[50] D.M. Trujillo, H.R. Busby, Optimal regularization of the inverse-heatconduction problem using the L-curve, Int. J. Numer. Methods Heat FluidFlow 4 (5) (1994) 447–452.

[51] D.A. Zumbrunnen, R. Viskanta, F.P. Incropera, Effect of surface motion onforced convection film boiling heat transfer, J. Heat Transfer Trans. ASME 111(3) (1989) 760–766.

[52] R.B. Abernethy, R.P. Benedict, R.B. Dowdell, ASME measurement uncertainty, J.Fluids Eng. 107 (2) (1985) 161–164.

[53] K.H.M. Abdalrahman, U. Alam, E. Specht, Wetting front tracking during metalquenching using array of jets, ASME Conf. Proc. 2010 (49408) (2010) 475–484.

[54] G. Suryanarayana, P. Ghosh, Adsorption and coalescence in mixed-surfactantsystems: air-water interface, Ind. Eng. Chem. Res. 49 (4) (2010) 1711–1724.

[55] W.-T. Wu, Y.-M. Yang, J.-R. Maa, Enhancement of nucleate boiling heat transferand depression of surface tension by surfactant additives, J. Heat TransferTrans. ASME 117 (2) (1995) 526–529.

[56] N. Zuber, Hydrodynamic Aspects of Boiling Heat Transfer, University ofCalifornia, California, 1959.

[57] N.L. Chester, M.A. Wells, V. Prodanovic, Effect of inclination angle and flow rateon the heat transfer during bottom jet cooling of a steel plate, J. Heat TransferTrans. ASME 134 (12) (2012) 122201–122209.

[58] G. Hetsroni, J.L. Zakin, Z. Lin, A. Mosyak, E.A. Pancallo, R. Rozenblit, The effect ofsurfactants on bubble growth, wall thermal patterns and heat transfer in poolboiling, Int. J. Heat Mass Transfer 44 (2) (2000) 485–497.

[59] V. Semião, P. Andrade, M.D.G. Carvalho, Spray characterization: numericalprediction of sauter mean diameter and droplet size distribution, Fuel 75 (15)(1996) 1707–1714.

[60] A.L. Biance, C. Clanet, D. Quéré, Leidenfrost drops, Phys. Fluids 15 (6) (2003)1632–1637.

[61] S. Ravikumar, J. Jha, S. Mohapatra, A. Sinha, S. Pal, S. Chakraborty, Experimentalstudy of the effect of spray inclination on ultrafast cooling of a hot steel plate,Heat Mass Transfer 49 (10) (2013) 1509–1522.






























































































experimental investigation of air-atomized spray with aqueous polymer additive for high heat flux...

Documents