lable at ScienceDirect

Applied Thermal Engineering 63 (2014) 11e22

Contents lists avai

Applied Thermal Engineering

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

Experimental analysis of the influence of microcapsule mass fractionon the thermal and rheological behavior of a PCM slurry

Mónica Delgado*, Ana Lázaro, Conchita Peñalosa, Belen ZalbaAragón Institute for Engineering Research (I3A), Thermal Engineering and Energy Systems Group, University of Zaragoza, Agustín Betancourt Building,C/María de Luna 3, 50018 Zaragoza, Spain

h i g h l i g h t s

� A slurry with 14, 20 and 30% PCM microcapsule concentrations has been analyzed.� Enthalpy and thermal conductivity depending on temperature have been obtained.� Rheological properties have been evaluated.� A 20% PCM microcapsule concentration is the most suitable for heat transfer.� Some problems of rupture in the PCM microcapsules have been found.

a r t i c l e i n f o

Article history:Received 22 March 2013Accepted 6 October 2013Available online 24 October 2013

Keywords:Microencapsulated PCM slurryConvective heat transferThermal energy storageThermophysical propertiesRheological propertiesPhysical stability

* Corresponding author. Tel.: þ34 976761000x5258E-mail addresses: monica@unizar.es (M. Del

(A. Lázaro), conchita.penalosa@unizar.es (C. Peñalosa)

1359-4311/$ e see front matter � 2013 Elsevier Ltd.http://dx.doi.org/10.1016/j.applthermaleng.2013.10.011

a b s t r a c t

A microencapsulated PCM (Phase Change Material) slurry with three different PCM mass fractions (14, 20and 30%) has been analyzed. The present study investigates the influence of the PCM microcapsule massfraction on the thermal and rheological characterization. Specifically, the EnthalpyeTemperature curves,the Thermal ConductivityeTemperature curves and the ViscosityeShear rate curves have been deter-mined. In addition, the physical stability under thermo-mechanical cycles has also been studied, as wellas the heat transfer phenomenon and the fluid mechanics. The results have shown an enhancement inthe heat transfer phenomenon, the slurry with 20% PCM microcapsules being the best option for use as aheat transfer fluid. Rupture of the PCM microcapsules was only observed for the slurry with 30% PCMmicrocapsules, after being pumped during three weeks and having experienced 10,000 solidificationemelting cycles.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Thermal energy storage based on the use of solideliquid phasechange of materials (PCM) is of increasing practical interest. Suchinterest is motivated by the considerable thermal energy storagedensity per unit volume of PCMmaterials in a reduced temperaturerange, and also by the constant incorporation of newmaterials withvery different properties and phase change temperature intervals.The interest in phase change materials is evident when consideringthermal energy storage systems with PCM which have beendeveloped for different applications [1].

Recently, a new technique has been proposed for the use ofphase change materials in thermal storage systems, heat ex-changers and thermal control systems. This new technique consists

; fax: þ34 976762616.gado), ana.lazaro@unizar.es, bzalba@unizar.es (B. Zalba).

All rights reserved.

of forming a two-phase fluid by combining a fluid such as water anda phase change material such as paraffin, resulting in a latent heatstorage fluid. Inaba [2] has classified thermal fluids, describing theirmain characteristics and applications. These latent thermal fluidsinclude the following five types: 1) ice slurries; 2) phase changematerial microemulsions; 3) microencapsulated PCM slurries; 4)clathrate hydrate PCM slurries and 5) shape-stabilized PCM slurries(ssPCM slurries). The present experimental study is focused onmicroencapsulated PCM slurries.

These new fluids offer many advantages and can be used eitheras thermal storage materials or heat transfer fluids [3] due to 1)their high storage capacity during phase change, 2) the possibilityof using the same medium either to transport or to store energy, asthese slurries are pumpable (thus reducing heat transfer losses), 3)heat transfer at an approximately constant temperature, 4) a highheat transfer rate due to the elevated ratio surface/volume, 5) lowerpumping power, as a consequence of the reduction in mass flowdue to the higher heat capacity, and 6) a better heat exchange than

Nomenclature

Bi biot numberf fraction of heat lossesh enthalpy (kJ/kg)G0 elastic module (Pa)G00 loss module (Pa)I current (A)k multiplier fitting coefficient (s)m exponent fitting coefficient (�)m$

mass flow rate (kg/s)Q$

heating power absorbed or transported by the fluid(W)

T temperature (�C)W$

pumping power (W)DU voltage (V)1-D one dimensional

Greek symbolsa convective heat transfer coefficient (W/(m2$K))g$

shear rate (1/s)h energy ratiom steady shear viscosity (Pa$s)

Subscriptsin inlet of the heat transfer sectionm1 melting beginningm2 melting completeout outlet of the heat transfer section0 low shear ratesN high shear rates

AbbreviationCOP coefficient of PerformancemPCM microencapsulated phase change materialPCM phase change materialRTO operation temperatures range

M. Delgado et al. / Applied Thermal Engineering 63 (2014) 11e2212

conventional heat transfer fluids, due to the decrease in fluidtemperature as a consequence of the higher heat capacity.Furthermore, these novel fluids have a more advantageous thermalenergy storage density than conventional systems of sensible heatstorage in water, and can compete with macroencapsulated PCMtanks. Besides, the response time may be shorter using these PCMemulsions or mPCM slurries as storage material than with macro-encapsulated PCM. The tanks will be simpler as there is no need tomacroencapsulate, and conventional tanks can be used.

Huang et al. [4] listed the recommendable properties for a PCMslurry utilized in cold storage and distribution systems. Specifically,the phase transformation range should match the designed oper-ating temperature range. There should be an absence of subcooling,a narrow phase temperature range, a high heat transfer rate, and alow pressure drop in pumping systems. Besides, the slurry shouldbe stable during long-term storage and have reversible freezing/melting cycles under thermalemechanical loads.

Among the different applications that appear on literature, it isfound the utilization of these PCM slurries as thermal storage ma-terials and heat transfer fluids in chilled ceiling [5,6]. In the firstwork, their authors present the simulation results of a combinedsystem of chilled ceiling and storage tank with a PCM slurry. ThisPCMslurrywas cooled andstored in the tankduring thenight,whichresulted in electricity peak shaving, taking advantage of thenocturnal tariff and of a higher COP of themachine due to operationduring lower environmental temperatures. During working hours,the PCM slurry flowed from the tank to the chilled ceiling, meltingthe PCM and releasing the latent heat. In the work of Griffiths andEames, they quantified the reduction of the mass flow flowingthrough a chilled ceiling in a roomwhenworkingwith a PCM slurry.Themass flow is reduced from0.7 L/s down to 0.25 L/s. Furthermoreit could absorb energy at a constant temperature, avoiding in-crements in the panel surface temperature when internal gainsincreased. Another well-known application was carried out at theNarita Airport in Tokyo by Shibutani [7]. The change of refrigerantsduet to environmental reasons resulted in lower cooling power andthe chiller was non-capable to absorb the demand peaks at specifictimes of the day. This problemwas solved through the installation ofa tank filled with a PCM slurry. Pollerberg and Dötsch [8] also pro-posed the use of PCM emulsions for cooling supply networks,allowing for the reduction of the pumping power and pipe di-mensions, with lower operation and investment costs.

Although there are numerous advantages in the use of PCMslurries, there is a lack of technical experience. The main issuesencountered when using a thermal storage material are subcoolingand the unstable processes presented by slurries. When using heattransfer fluids, a higher heat transfer rate compared to water is ofinterest but existing studies have not obtained clear conclusions.

Delgado et al. [9] presented a review of the different parametersinfluencing the objective magnitudes in PCM slurries. Specifically,the different effects that the mass fraction has on the heat transferphenomenon are listed in a table. From this compilation, it wasconcluded that increasing the PCMmass fraction in suspension hadtwo opposite effects. On the one hand, this increase means adecrease in the Stefan number and therefore an improvement inthe convective heat transfer coefficient. On the other hand, theincrease also means an increase in the viscosity and a decrease inthe thermal conductivity, resulting in a worse heat transferphenomenon.

A previous work by the present authors described an experi-mental installation for studying the heat transfer phenomenon andfluid mechanics [10]. In this installation, a PCM slurry with a 10%PCM microcapsule mass fraction was analyzed, pending the anal-ysis of the influence of higher PCM microcapsule mass fractions onthe heat transfer and on the flow characteristics. The present studyaims to complete this previous work and to include thermophysicaland rheological characterization. The paper compiles and analyzesthe results of heat transfer and fluid mechanics from a detailedmethodology for a mPCM slurry with three different PCM micro-capsule mass fractions: 14%, 20% and 30%, to analyze which PCMmicrocapsule mass fraction is suitable to be used as heat transferfluid. The thermophysical and rheological properties have beenobtained and the measurement methodology in the case of work-ing with PCM slurries described thoroughly. This characterizationhas allowed to understand the heat transfer process. Physical sta-bility has also been analyzed. Table 1 compiles the different char-acteristics investigated in this paper, together with theexperimental devices used for their characterization.

2. Materials and properties

The studied mPCM slurry consists of microcapsules of paraffincoated by a polymer and dispersed in water through detergents.The mass fraction of the PCM microcapsules in the three mPCM

Fig. 1. EnthalpyeTemperature curves for the slurries with 14, 20 and 30% PCMmicrocapsule mass fraction.

Table 1Compilation table of the characteristics analyzed in the experimental work.

Characteristic to analyze Equipment/method Section

Properties EnthalpyeTemperaturecurves

T-history method 2.1

Thermal conductivityeTemperature curves

Laser Flash method,DSC

2.2

ViscosityeShear ratecurves

Control stressrheometer

2.3

Fluid mechanics Pressure drop Experimentalinstallation builtfor that purpose

3.2

Heat transfer Heat transfer (forcedconvective heat transfer)

Experimentalinstallation builtfor that purpose

3.2

Physical stability Physical stability underthermo-mechanical loads

Environmental SEM 4

M. Delgado et al. / Applied Thermal Engineering 63 (2014) 11e22 13

slurries is 14%, 20% and 30%, respectively. The particle size distri-bution is approximately 1e20 mm according to the manufacturer’sdata. This mPCM slurry has been purchased in the commercialmarket.

2.1. EnthalpyeTemperature curves

The phase change temperature range and the phase changeenthalpy as a function of the temperature were obtained using theT-history method [11e13]. When a material is characterized, thesample must be representative of the material in question. In thiscase, the PCM slurry is composed of different substances. The vol-ume of the sample should be at least a few cm3 or more if possible[13] to ensure that it has the correct chemical and physicalcomposition representative of the bulk material. For this reason, aninstallation using the T-history method was used to determine theEnthalpyeTemperature curve during the phase change of themPCM slurry. This method is based on comparing the temperatureevolution of the PCM and a reference substance during cooling andheating against the ambient temperature of a chamber. Thisreference substance should be a substance with well-known ther-mal properties. The basic aspects of this methodology are:

� 1-D heat transfer in radial direction (samples contained in cy-lindrical containers)

� The systems formed by the container and the water (this is thereference substance) and the PCM respectively are lumpedcapacitance systems. The temperature of the substance and ofthe tube is uniform at all times (Bi << 1)

� Heat transfer from the containers of PCM and of the referencesubstance to the chamber air takes place by natural convection.

In this manner, the methodology proposed by Zhang et al. [11]consists of recording the ambient temperature and the tempera-ture of the sample and of the reference substance, contained inidentical containers. Once obtained the TemperatureeTime curvesof both substances, these data can be processed to estimate theEnthalpyeTemperature curves. Fig. 1 shows the EnthalpyeTem-perature curve for the melting and solidification of the mPCMslurries, obtained by the T-history method. The phase changetemperature range of the mPCM slurry in the three cases wasapproximately 21e24 �C and the phase change enthalpy for thisrange was 15.3 kJ/kg, 21.1 kJ/kg and 28 kJ/kg for the 14%, 20% and30% mPCM slurry, respectively. The phase change enthalpy isespecially low in the slurry with a 14% mass fraction, since there isonly a 14% PCM microcapsules. The remaining 86% would beformed by water and other substances. Furthermore each micro-capsule is formed by the phase change material and by the

polymeric shell, reducing the effective fraction of PCM in suspen-sion. It is not available the core fraction for this material, however itis known that the core usually constitutes between 20 and 95% ofthe total mass [9]. For this reason, the phase change enthalpy isquite low with very low PCM microcapsule mass fractions and asharp rise in the enthalpy is not noticeable, due to the low contentof PCM in the compound. Hysteresis and subcooling phenomenawere not observed when comparing the solidification and meltingcurves.

2.2. Thermal conductivityeTemperature curves

Thermal diffusivity measurements have been made with a LaserFlash device from Netzsch, model LFA 457 MicroFlash. The LaserFlash method was initially designed for measurements in solids,where the thickness of the sample is known, standards are notnecessary and the property is measured in the transient response.This method is indirect, since the thermal conductivity property isobtained from themeasurement of other properties, in this case thethermal diffusivity, the density and heat capacity values.

The density measurements have been obtained from the mea-surement of the sample mass using a Mettler Toledo precisionbalance (accuracy 1 mg) and from the volume measurement of thesample in a calibrated test-tube of 10 mL at room temperature(standard deviation: 0.021 mL). This value has been taken as aconstant value in the temperature range of the test. For the heatcapacity measurement, a differential scanning calorimeter (DSC)from Netzsch, model DSC 200 F3 Maia, has been used, with anaccuracy of 1% in the heat capacity measurement.

For obtaining the thermal diffusivity, one of the sample surfaceswas heated in a homogeneous manner using a laser pulse, wherethe voltage and transmission filter were controlled. Using thismethod, the heat absorbed in the surface is transferred through thesample and an increase in temperature is produced in the rearsurface. This increase is measured over time from a liquid nitrogen-cooled InSb photocell. A mathematical model describes the tem-perature rise versus time signal, and this was fitted to the experi-mental data using a non-linear regression algorithm.

Few studies can be found in the literature that measure liquidswith the Laser Flash method [14e16], and only one involves a PCMchanging its phase, specifically NaNO3 [17] .Tomeasure the thermaldiffusivity in liquids, a special sample holder is necessary to containthe mPCM slurries. The sample is introduced between two layers ofa material whose properties are well-known. The thickness and

M. Delgado et al. / Applied Thermal Engineering 63 (2014) 11e2214

distances between the two layers are also perfectly known. In thisway, the material can be evaluated as a three-layer compound,where the unknown factor is the intermediate layer.

The presence of the sample holder for liquids may disturb theprocess of heat conduction during the transient heating of thesample. Coquard and Panel [16] analyzed the influence of differentparameters or phenomena on the results of the liquids. These au-thors considered that all the materials of the sample holder wereopaque to the infrared radiative heat. They ignored the naturalconvection in the samples, considering the heat transfer purelyconductive. They also assumed the heat transfer of the externalsurfaces with the environment occurred by convection and radia-tion. Consequently, the total external heat transfer was consideredas an only coefficient h.

To estimate the uncertainty, Coquard and Panel made a reviewof the parameters that may cause errors. They concluded that thecorrect determination of the sample thickness and a rigorous fillingup of the sample holder were key parameters. An air fraction in thesample of about 1.25% meant errors up to 15.4%, since this air layerwould work as a thermal barrier. They also observed that infraredradiation could not propagate in water. Thus the hypothesis of noradiative exchange did not mean an error when measurementswere made in materials with a sufficient amount of water.

The sample holder for liquids supplied by Netzsch is made ofPt90Rh10, whose thermal conductivity is 38W/(m K). This is a veryhigh value compared to the thermal conductivity of the liquids tobe measured (in the range from 0.15 to 0.6 W/(m K)). However, thesample holder from Netzsch provides a side space between thebase and the lid. The thermal resistance of this air space is higherthan the thermal resistance of the liquid to be measured, mini-mizing the heat transfer through the sample holder. The samephenomenon occurs with the upper contact. However, given thatthe joint is not under pressure, the thermal resistance is higherthan the thermal resistance of the fluid. It can be said that the in-fluence of the sample holder on the results has been minimized.

Additionally, the empty sample holder has been tested and theresponse was compared to the response with the sample holderfilled with water. A comparison of the results of both tests is shownin Fig. 2. The response is much longer when the sample holder isempty. When selecting the time range for the software to make thecalculation, it is important that the data acquisition time should beshort to avoid the contribution of the sample holder.

Regarding the measurement of the thermal diffusivity duringthe phase change from solid to liquid, since this is a transientmethod there may be great changes in the thermal properties

Fig. 2. Detector signal of the Laser Flash with the sample holder empty or filled withwater.

during the phase change. However, the mathematical model con-siders constant properties. It was thus decided to make the mea-surements in the single-phase states, solid and liquid.

There are few standards for liquids in comparison to solids.Therefore, three different liquids were measured prior to the testswhose thermal diffusivity or conductivity is known: distilled water,hexadecane and glycerin. These liquids have thermal diffusivityvalues within the range of mPCM slurries for the temperatureranges of the application of these fluids.

To measure solid samples with the Laser Flash equipment, avacuum was first created and then an inert atmosphere of N2.However, when this procedure was carried out for liquids, thevacuum and the pressure reduction in the equipment chambercaused the water to evaporate when reaching the vapor pressure.This was checked by weighing the sample before and after thevacuum. Finally, the vacuum was omitted and a longer time wasgiven for the creation of the N2 atmosphere.

The external surfaces of the sample holder were coated withgraphite to increase the amount of energy absorbed and to guar-antee that all the parts of the sample had the same absorption. Thetemperature of the front surface can reach very high values, so it isimportant to know the upper limit of this temperature to avoidtransition phases of the tested material. In the case of water, it isnecessary to avoid evaporation.

From the previous study by Coquard and Panel [16], it wasknown that a complete filling up of the sample holder was crucial,as well as the correct determination of the thickness of the liquidsample. The sample thickness was obtained from the measure-ments of the thickness of the sample holder executed by a caliber.In order to guarantee the complete filling up of the sample holder,the volume of the liquid sample holder was calculated from thegeometrical data and the amount of sample was controlled by amicropipette.

Taking all these considerations into account, thermal conduc-tivity values were obtained for the three liquids tested: water,hexadecane and glycerin. The values are shown in Fig. 3. Thesevalues are the average value of five pulses executed both for ther-mal diffusivity and temperature, together the standard deviation ofthese measurements. In the case of distilled water, the results showa maximum error of 7.87% and for hexadecane 4.31%. In the case ofglycerin, higher errors were obtained, up to 15.38%. The referencevalues for water, hexadecane and glycerin were taken from thefollowing references, respectively [18e20]. Although due to the

Fig. 3. Values obtained of thermal conductivity of liquids in comparison to the refer-ence values.

M. Delgado et al. / Applied Thermal Engineering 63 (2014) 11e22 15

graph scale is not very perceptible, these obtained values show anincreasing monotonous variation with temperature.

The thermal diffusivity values were obtained from the threelayer model provided by the equipment software. Fig. 4 shows thethermal conductivity values for the mPCM slurries with PCMmicrocapsule mass fractions of 14%, 20% and 30%. The measure-ments taken at 20 �C are not considered very reliable, since even avery small increase in the temperature due to the laser pulse willcause the heat capacity to change abruptly (phase change regionbetween 20 and 24 �C) and this methodology may be not validgiven that the heat capacity is considered constant in the calcula-tion. So if the attention is paid to the values at 25 and at 30 �C, it canbe observed that thermal conductivity of PCM slurries increasesslightly when temperature rises. It must be pointed out that theincrease of the PCM microcapsule mass fraction entails a decreaseof the thermal conductivity. This behavior was expected, as thethermal conductivity of paraffin is lower than of the water. Spe-cifically the slurry for the PCMmicrocapsule mass fraction of 14, 20and 30% have experienced a 24, 32 and 39% reduction in compar-ison to water respectively at a temperature of 30 �C.

2.3. Rheological characterization

To complete the present study, the ViscosityeShear rate curveshave been obtained with a control stress rheometer from TA In-struments, model AR-G2. There are several works in the literaturewhere ViscosityeShear rate curves are presented. However, the au-thorsdonotdescribe ingreatdetail theprocedureused toobtain them.

In this work, rotational tests have been carried out. These testsentail applying a torque (or stress) and measuring the strain, inorder to obtain viscosity values. The ViscosityeShear rate curveshave been obtained through a shear sweep from 0.001 to 1000 s�1.For this purpose, a stress is applied to the sample. The measure-ment of viscosity is carried out when the material has reached thestate of steady flow. The stress is increased logarithmically and theprocess is repeated, providing the flow viscosity curve. For thedefinition of the steady flow state, a condition has been proposedthat the variation of the stress for three consecutive points shouldbe lower than 5%. When this condition is reached, it is consideredthat the steady flow state has been reached.

In the works found in the literature describing part of themeasurement procedure, a cone tends to be used as the geometry[21,22]. However, in this case a stainless steel plate of 40 mm hasbeen chosen because of the size of the PCM microcapsules in

Fig. 4. Values obtained of thermal conductivity for the mPCM slurry with PCMmicrocapsule mass fractions of 14, 20 and 30%.

suspension. These PCM microcapsules have a diameter range from1 to 20 mm, according to the manufacturer’s data. It is importantthat in the rotational tests, the particles can flow without anyproblem. A gap (or the truncated gap in cones) must be 10 timeshigher than the size of the particles in suspension. For this reasonthe plate was chosen, where the gap can be controlled. In therotational tests with the plate, there is no constant shear rate alongthe radius of the geometry (unlike the cone). Therefore, a correction(included in the software of the rheometer) must be applied.

It must be mentioned that this geometry allows the use of the“solvent trap” accessory. With this accessory, a saturated atmo-sphere of humidity is created, avoiding the drying of the sample.For the temperature control of the sample, a Peltier plate was used.The Peltier plate guarantees that the plate is at the set temperature.If the set temperature is much higher or much lower than the roomtemperature, temperature gradients in the sample will be able totake place. In this case, the Peltier plate is considered the appro-priate temperature controller, since the temperatures of the test aretemperatures close to the room temperature, and because the“solvent trap” accessory can be used with this configuration.

Before obtaining the viscosity measurements, a time sweep(oscillatory test) was done to study the possible drying of thesample, and to determine in this way the maximum time of thetests (to avoid this phenomenon and to avoid the obtaining ofincorrect measurements). In the time sweep, the elastic (G0) and theloss (G00) components of the mPCM slurry are measured over timefor a set frequency and strain (within the viscoelastic region). Thefact that the sample dries would cause an increase in the elasticcomponent (G0) and an increase in the loss component (G00), as theresulting sample would have a higher PCM mass fraction as thewater evaporates.

The time sweep was done without the solvent trap (without awater-saturated atmosphere). As the slurries are complex fluids, apre-shear of 100 1/s was carried out for 1 min in order tocompletely destroy the structure of the sample. In this way, ifduring loading the sample structure is partially broken, the samplewould be completely broken. It was observed that to recover thestructure required at least 600 s and that from 1200 s the samplestarts to evaporate. This time (1200� 600¼ 600 s) is insufficient tocarry out the test.When the geometry rose, it was observed that thesample was dried at the edges. The visual results agreed with theresults obtained with the rheometer. For this reason, the use of thesolvent trap with water was necessary.

Regarding the ViscosityeShear rate curves at 27 �C shown inFig. 5, the curves for the slurries with 14% and 20% PCM

Fig. 5. Comparison of the ViscosityeShear rate curves according to the PCM micro-capsule mass fractions in suspension.

M. Delgado et al. / Applied Thermal Engineering 63 (2014) 11e2216

microcapsule mass fractions are very similar to each other. Theslurry with a 14% PCM microcapsule mass fraction has a viscosityapproximately 5 times higher than water, the 20% slurry 6 timeshigher and the 30% slurry 18 times higher. It is observed that whenincreasing the shear rate, the viscosity decreases down to theNewtonian plateau, where the viscosity remains constant. Thisshear-thinning behavior has also been found in the majority of theworks in this field found in the literature [21e24]. Other worksshow a Newtonian behavior when working with PCM concentra-tions lower than 25% [25,26]. This shear thinning behavior can beexplained through the spatial layout of the microcapsules in sus-pension. When a slurry is stable and at rest, the particles aredispersed randomly in the continuous phase. When the slurry issheared, there is no cooperative motion between themicrocapsulesso that these microcapsules move in the flow direction and there-fore the viscosity is high. However, when the slurry is sheared athigh shear rates, the microcapsules start to move from theirrandom layout to a situationwhere they start to form layers. In thisway, the average distance between microcapsules decreases in theflow direction and increases in the direction perpendicular to theflow. This change of the spatial layout makes the motion is mucheasier and decreases the viscosity [27].

It is also observed for the slurry with 14% and 20% PCM micro-capsule mass fractions that the flow curve starts to increase atapproximately 200 s�1. Rotational tests with liquids of very lowviscosity at very high shear rates may cause secondary flows,increasing the apparent viscosity [27].

The “Best ViscosityeShear rate” tool of the rheometer softwarewas used, which provided the behavior equation that relates theviscosity to the shear rate that best fits with the measured values.The equation that best predicts the shape of the flow curve for thethree mPCM slurries is the Carreau model [28]. This model isdefined according to Equation (1):

m� mNm0 � mN

¼ 1�1þ

�k$g

$�2�m=2 (1)

where m0 and mN refer to the asymptotic values of viscosity at verylow and very high shear rates respectively, k is a constant param-eter with the dimension of time andm is a dimensionless constant.Table 2 shows the values of these parameters.

3. Pressure drop and heat transfer analysis

To study both flow and heat transfer characteristics of thesemPCM slurries (basically to measure the convective heat transferunder constant heat flux and the pressure drop flowing through acircular tube), a flow loop was designed and built. This installationappears in Fig. 6 with the label of each device. In order to obtainthese convective coefficients, it is necessary to record the heat fluxabsorbed by the heat transfer section of the figure, the fluid tem-perature and the wall temperature at several locations along thistube. This experimental setup consists of a thermostatic bathwherethe mPCM slurry is prepared to the set temperature and pumped tothe flow loop. The mass flow is controlled and measured respec-tively by control valves and by a Coriolis flow meter. In the heat

Table 2Parameters according to the Carreau model [28].

PCM microcapsules concentration h0 (Pa$s) hN (Pa$s) k (s) m

14% 13.80 4.89$10�3 288.80 1.0220% 10.88 6.45$10�3 315.40 1.0630% 6.45 18.32$10�3 82.05 1.03

transfer section, heating wires supply uniformly-distributed heatflux. Finally, the mPCM slurry returns to the thermostatic bath forcooling.

The heat transfer section consists of a 10 mm copper tube,1.82 m in length. Eleven type T thermocouples measure the walltemperature along the circular tube and two Pt100 sensors mea-sure the fluid temperature at the inlet and at the outlet of the heattransfer section. The temperatures of the fluid along the tube werecalculated. Ten isolated nichromewires, connected in parallel, werecoiled around the copper tube to heat the section. These heatingwires were connected to the 230 V AC power supply through aphase angle electronic regulator, which allows the heating powerprovided to the heat transfer section to be varied. The maximumheating power was 3600 W. The heat flux provided to the heattransfer section must guarantee complete phase change of thedispersed PCM, so that the difference between the inlet and theoutlet temperatures is noticeable and higher than the Pt100 un-certainty. The current and the voltage were measured by anammeter and a voltmeter, respectively. Heat losses were minimal,approximately 3%, and therefore the heat transfer section was notthermally isolated. The heat losses were taken into account duringdata processing. A pressure differential transducer measures thepressure drop in the heat transfer section. All measured data wererecorded by a HP-34970A Data Logger. The error introduced by thedata logger is negligible with regard to the other measurementdevices.

A previous article of the authors gives more specific details ofthe devices about the main characteristic of the equipments andtheir measurement range [10]. This work also explains the valida-tion process of the experimental setup. This validation wasaccomplished by testing the setup with water and comparing theresults with theoretical values. Pressure drop, heat flux and walltemperature measurements were successfully validated. Somedifferences were observed between the experimental wall tem-peratures and those obtained by the analytical solution. In thesecases, an empirical correction model was established. This empir-ical model of correction corrects such deviations, obtaining in thisway an average error in the measurement of the wall temperatureof 0.24 �C. In this manner, the uncertainty in the measurement ofthe internal forced convective coefficient from the experimentalinstallation (taking into account that it will be different for eachlocal position and that it will depend on the measurement condi-tions, is around 5e10%.

3.1. First measurements

Prior to the flowand heat transfer tests, the energy balancemustbe verified to guarantee that microcapsules do not adhere to anypart of the experimental installation, since non-stable and non-homogeneous mPCM slurries can increase the danger of obstruc-tion [29]. When calculating the energy balance (Equation (2)),thermal equilibrium between the microcapsules and water isassumed.

ð1� f Þ$DU$I ¼ m$$ðh½Tout� � h½Tin�Þ (2)

The verification tests were performed for different mass flowsand heat fluxes. Fig. 7 shows the EnthalpyeTemperature curveobtained through the energy balance equation (along with itsfitting to a sigmoid curve) and its comparison with the EnthalpyeTemperature curve previously obtained with the T-history instal-lation. If the curves for the three PCM microcapsule mass fractionsare compared to their respective heT curves obtained with the T-history installation, it can be observed that the curve for the 30%PCM microcapsule mass fraction moves slightly towards higher

Fig. 6. Picture of the experimental setup.

M. Delgado et al. / Applied Thermal Engineering 63 (2014) 11e22 17

temperatures. In addition, this phenomenon is accentuated whenincreasing the mass flow. On the other hand, for the 14% PCMmicrocapsule mass fraction, the curve is displaced towards lowertemperatures in comparison to the curve obtained with the T-his-tory installation. For the 20% PCMmicrocapsule mass fraction, bothcurves are almost the same. Three possible causes are suggested forthese differences.

Firstly, in view of the curve for the 30% PCM microcapsule massfraction, it was thought that a fraction of the microcapsules wasdeposited in different components of the installation, since ac-cording to the curve obtained in the experimental installation, theenthalpy was lower for the same fluid temperature. In order tocheck this hypothesis, the components more susceptible to depo-sition were dismantled. However, no deposition was observedeither in the elbows or in the valves. Having dismissed thisconjecture, it was considered that the differences could arise fromthe experimental installation itself (possibly inappropriate fordetermining the heT curves). The slurry temperature for eachsection shows a temperature profile and according to this meth-odology an average temperature values is taken. These differencescould also arise from the hypothesis that the thermal equilibriumbetween the PCMmicrocapsules and the water was incorrect, sincethere might be a heat transfer process between the PCM particlesand the water. Regarding this last approach, Diaconu [30] studiednumerically this heat transfer phenomenon between PCM micro-capsules and water. In his results, he observed that the watertemperature and the microcapsule temperature were very closeeach other, due to the high surface of heat exchange. The biggerdifferences appeared during melting and solidification, causing ahysteresis phenomenon. This hysteresis was influenced by thecapsule diameter and by the heat transfer coefficient between thePCM capsules and the water.

3.2. Evaluation as a thermal storage and heat transfer fluidcompared to water

As in the previous study [10], the evaluation of the PCM slurrywith different PCM microcapsule mass fractions began with themeasurement of the pressure drop in the heat transfer section.Fig. 8 shows that when the PCM microcapsule mass fraction isincreased to 30%, the pressure drop increases significantly, whereasan increase in the PCMmicrocapsule mass fraction from 14% to 20%hardly had any appreciable effect on the pressure drop. These re-sults correspond with the previous viscosity measurements. Theslurry with a 14% PCM microcapsule mass fraction and the slurrywith a 20% PCM microcapsule mass fraction showed a similar vis-cosity, while the viscosity of the slurry with a 30% PCM microcap-sule mass fraction was three times higher. From these pressuredrop values and the phase change enthalpy, the transported

Thermal Energy vs. Pumping power compared to water could beevaluated. The following energy ratio of improvement was defined(Equation (3)) for this comparison:

h ¼

�Q$

mPCMslurry

W$

mPCMslurry

��

Q$

water

W$

water

� (3)

Fig. 9 shows this ratio obtained for different average fluid ve-locities while Fig. 10 shows the Pumping power vs. Transportedthermal energy ratio. It can be seen in Fig. 9 that if the PCMmicrocapsule mass fraction is increased from 14 to 20%, the velocityfrom which the energy ratio of improvement is equal to 1 de-creases. In contrast, increasing the PCMmicrocapsule mass fractionto 30% led to in an increase in this velocity, a consequence of thesharp increase in the pressure drop observed in Fig. 8. In this case,for velocities lower than 1 m/s the negative effect of the increase inthe viscosity with the mass fraction is greater than the improve-ment of the thermal energy that can be transported in comparisonto water for the same velocity. In Fig. 10, it is observed that for thesame transported thermal energy, the pumping power decreases incomparison to water.

To analyze the heat transfer characteristics, the mass flow andthe heating power of the heat transfer section were varied and thewall temperatures were measured. The tests were carried out un-der the boundary condition of constant heat flux, the heat transfersection was a fully hydrodynamic developed section and the flowwas laminar, with mass flows from 20 to 50 kg/h. The correctionmodel was applied to the measured wall temperatures, and thesevalues were then compared to the calculated values for the case ofwater (obtained through Kays correlation [31]). Dependence of themass flow and the operation temperature range on the decrease inthewall temperaturewas also studied. As in the previous paper, thesame parameter, “operation temperature range” was defined forthe analysis, according to Equation (4). This parameter shows if thephase change is in accordance with the heat transfer section. Aparameter equal to 1 would mean that the mPCM slurry startsmelting just as it enters the heat transfer section, and leaves theheat transfer section only when the PCM microcapsules arecompletely melted. A parameter under 1 would mean that the PCMmicrocapsules did not completely melt in the heat transfer section,and a parameter above 1 means that both liquid and phase changeregions coexist in the heat transfer section. The parameter wasdefined taking into account the phase change and liquid regions,attributing in this way different phenomena to each region.

Operation temperatures range½RTO� ¼ h½Tout� � h½Tm1�h½Tm2� � h½Tm1�

(4)

Fig. 7. EnthalpyeTemperature curves obtained through the energy balance to the heattransfer section.

Fig. 8. Pressure drop measurements in the heat transfer section for 14, 20 and 30%PCM microcapsule mass fractions.

Fig. 9. Energy ratio of improvement vs. Fluid average velocity for different PCMmicrocapsule mass fractions: 14, 20 and 30%.

M. Delgado et al. / Applied Thermal Engineering 63 (2014) 11e2218

From the measurement of the wall temperature, the convectivecoefficients were obtained for the laminar region. The resultsshowed a significant decrease in the wall temperature compared towater for the three different PCM microcapsule mass fractions,resulting in a better cooling performance. This decrease in the wall

temperature was higher when the “operation temperature range”was higher. In the case of the heat transfer coefficients by convec-tion, the axex curve for the slurry with PCM microcapsule massfractions of 14 and 30% for the mass flow of 50 kg/h was very closeto the curve for water. The slurry with a mass fraction of 20%showed better results. In spite of there being almost no improve-ment in the convective heat transfer coefficient for the mass frac-tions of 14% and 30%, the wall temperature is lower in comparisontowater due to the decrease in the temperature of themPCM slurry,as a consequence of its higher heat capacity.

The tests were planned so that the RTO tested for the slurry withthe 30% PCM microcapsule mass fraction was the same as that forthe tests with the slurries with 14% and 20% PCM microcapsulemass fractions (RTO ¼ 0.75 and RTO ¼ 1.1), always calculated ac-cording to the EnthalpyeTemperature curve obtained with the T-history installation. However, in the analysis of the test results thephenomenon explained in Section 3.1 was observed. The heT curveobtained from the energy balance was displaced for the slurry withthe 30% PCM microcapsule mass fraction. In this way, the outlettemperature of the fluid was determined with the RTO value, theinlet temperature of the fluid and the values of enthalpy from the T-history curves. The heat flux was controlled to reach this outlet

Fig. 10. Pumping power vs. transported thermal energy for different PCM microcap-sule mass fractions: 14, 20 and 30%. Fig. 11. Average decrease of the wall temperature in comparison to water according to

the PCM microcapsule mass fraction.

Fig. 12. Average improvement of the convective heat transfer coefficient in compari-son to water according to the PCM microcapsule mass fraction.

M. Delgado et al. / Applied Thermal Engineering 63 (2014) 11e22 19

temperature. In this manner the RTO parameter took the values0.68 and 0.89 when calculated from the h-T curves obtainedthrough the energy balance (whereas for the slurry with 14% and20% PCM microcapsule mass fraction, the RTO was 0.75 and 1.1). Inother words, the phase change was not completed in either of thetwo tests.

Figs. 11 and 12 show the average decrease in the wall temper-ature (in Celsius scale) and the average increase in the convectiveheat transfer coefficient with regard to the water for the three PCMmicrocapsule mass fractions of 14%, 20% and 30%.

It can be observed that when increasing the PCM microcapsulemass fraction up to 30%, the decrease in the wall temperature andthe increase in the convective heat transfer coefficient are lowerthan for the slurry with the 20% PCM microcapsule mass fractionbut slightly higher than that with the 14% PCM microcapsule massfraction. The decrease in the wall temperature for the 30% massfraction compared to the 20% mass fraction is around 10% for themass flow of 20 kg/h and around 38% for the mass flow of 50 kg/h.This decrease may be lower than that with the 20% PCM micro-capsule mass fraction because the RTO parameter for this case isslightly lower, as a consequence of the heT curve being slightlydisplaced with the temperature.

As mentioned in the previous paper [10], it is observed thatwhen the “operation temperature range” parameter fits with thephase change temperatures (RTO ¼ 1 or higher), the higher is thedecrease in thewall temperature. It is also observed in this case thathigher mass flow or higher velocities lead to a lesser improvementin the convective heat transfer coefficient. This can be explained bythe patterns of the thermally developing flow. At higher mass flowrates, the length of the thermally developing region is greater. Thismeans that for a given position, the fraction of melted PCM mi-crocapsules in that section is lower, resulting in a worse heattransfer from the wall to the core region. This phenomenon wouldalso occur at higher PCM mass fractions.

The decrease in the improvement of the heat transfer coefficientby convection when the PCM microcapsule mass fraction increasesfrom 20 to 30% is around 30% for the mass flow of 20 kg/h, andaround 70% for the mass flow of 50 kg/h. This reduction may also bedue to the noticeable viscosity increase, as observed in Fig. 5,decreasing the degree of turbulence and therefore worsening theheat transfer phenomenon. Another cause could be the lowereffective phase change enthalpy, since the heT curve is displacedand the phase change is not completed. The decrease in thermal

conductivity deteriorates the heat transfer to the core region of theflow. In this case the decrease in the thermal conductivity when thePCMmicrocapsule mass fraction increases from 20 to 30% is around10%, and around 40% lower than water.

In view of these results, it can be stated that the slurry with a20% PCM microcapsule mass fraction is the most effective slurryregarding the heat transfer by convection, and taking into accountthat for the slurry with a 30% PCM microcapsule mass fraction, alower fraction of PCM was melted.

4. Physical stability

During the tests with the slurry with a 30% PCM microcapsulemass fraction, the thermostatic bath and pump were turned offduring one day, stopping the flow of the slurry through theinstallation loop. Without agitation, the PCM microcapsules insuspension separated from the water and caused the clogging ofone of the balanced valves. These samples were analyzed with amicroscope.

To study a possible rupture of the PCM microcapsules in sus-pension, samples of the slurries were observed with a Philips XL30

Fig. 13. Sample not thermally and mechanically cycled observed with an environ-mental SEM. PCM microcapsule mass fraction ¼ 30%.

M. Delgado et al. / Applied Thermal Engineering 63 (2014) 11e2220

environmental SEM (Scanning Electronic Microscope) after beingthermally cycled (melting and solidification cycles) and pumped.Preparation of the sample is unnecessary for environmental SEMsand samples with water content can be observed. This is possiblebecause with the environmental SEM, there is a gas in the chamberof the sample enabling the examination of samples which would bedifficult to observe in a conventional SEM for different reasons (forexample, they are not conductive, they are not compatible with thehigh vacuum of a conventional SEM, or they need difficult prepa-ration steps). When the gas present in the chamber is water vapor,then wet samples can be observed, even samples in solutionswithout the necessity of previous preparation. In addition to savingtime and preparation material, the use of other devices, that maygive a non-real observation of the sample, is not necessary.

Fig. 13 shows a sample of the non-thermalemechanical cycledslurry. The shape of the microcapsules is spherical, although certaincavities can be observed on the surface. The samplewith a 30% PCMmicrocapsule mass fraction after undergoing thermal and me-chanical cycles in the experimental installation during three weeks(having experienced approximately 10,000 solidificationemeltingcycles) is shown in Fig. 14. The joins between the PCM microcap-sules can be seen. To dismiss a possible optical effect resulting fromthe opacity of water to electrons, the sample was dehydrated dur-ing the observation, decreasing the pressure in the microscope

Fig. 14. Sample cycled during three weeks (10,000 meltingesolidification cycles)observed with an environmental SEM (dehydration process). PCM microcapsule massfraction ¼ 30%.

chamber. Even with a pressure of 2 Torr, the effect appeared.Therefore, the fact that the microcapsules join each other as aconsequence of their possible rupture can be confirmed. Thesample with a 20% PCM microcapsule mass fraction but havingundergone thermalemechanical cycles during two weeks was alsoobserved (having experienced approximately 7000 solidificationemelting cycles). Its imagewas very similar to the non-cycled sampleshown in Fig. 13. In this case, the microcapsules had not broken. It ispossible that in the case of the sample with a 30% PCM microcap-sule mass fraction, the rupture and subsequent joining of micro-capsules to each other has caused the EnthalpyeTemperature to bedisplaced, because of the higher effective particle diameter. Thisdecreases the heat transfer area and may cause hysteresis.

At the beginning, this analysis of the possible rupture of the PCMmicrocapsules was not considered in the frame of this work, so thenumber of thermo-mechanical cycles had not been set and itdepended on the period of time of the tests. The best option toestablish a correct comparison between the different samples hadbeen subject to the different slurries at a same number of cycles.However this microscopy analysis was considered when analyzingthe slurry with 30% PCM microcapsule mass fraction, due to theblockage of the installation. When this fact happened, the otherslurries with lower PCM microcapsule mass fraction had beenalready tested. It is expected that with higher PCM microcapsulesmass fraction, there is more contact and friction among them andthe rupture may take place more easily. It is evident that prior totheir use in a specific application, their stability should be studiedfor the predicted number of cycles in the useful life of the system.

It is expected that with higher PCMmicrocapsules mass fraction,easier the rupture as more contact and friction among them maytake place.

5. Results compilation

Table 3 shows the results compilation together with the mainconclusions and findings of this work.

6. Conclusions

A PCM slurry with three different mass fractions (14%, 20% and30%) has been characterized. Specifically, the Enthalpyetempera-tures have been obtained. The methodology for obtaining mea-surements of thermal diffusivity using Laser Flash equipment hasbeen fully described. The geometrical determination of the sampleholder and its correct filling are crucial for obtaining appropriatemeasurements. In the samemanner, themethodology for obtainingthe ViscosityeShear rate curves using a control stress rheometerhas been described. The flow curves for the three slurries are fittedto the Carreau model.

In the evaluation of the paraffin slurry as a new heat transferfluid compared to water, it has been observed that when increasingthe mass fraction from 14% to 20%, the velocity from which theðQ$

slurry=W$

slurryÞ=ðQ$

water=W$

waterÞ ratio is higher than 1 decreases.However, when the mass fraction was increased to 30%, this ve-locity rose as the viscosity increased sharply.

With regard to the heat transfer phenomenon, the convectiveheat transfer coefficient for the three mass fractions represented animprovement in comparison towater. The slurry with the 20%massfraction gave the best results, for the conditions of heat flux andvelocity analyzed in the frame of this work. If the attention is paidto the transported thermal energy in relation to the pumping po-wer, the slurry with 30% PCMmicrocapsule mass fractionwould bethe best option. However the improvement of the slurry with the30% PCM mass fraction was less than that with the 20% massfraction regarding the internal convective heat transfer

Table 3Results compilation.

Characteristic to analyze Mass fraction % Main conclusions and findings for the PCMslurry analyzed

14% 20% 30%

Properties Phase change enthalpy (kJ/kg) in thetemperature range 21e24 �C

15.3 21.1 28.0 Improvement of the stored energy density,up to 2 times higher than water.No hysteresis or subcooling observed.

Thermal conductivity at 30 �C(W/(m$K))

0.46 0.41 0.37 Methodology proposed to measure thermaldiffusivity with a Laser Flash equipment.A 20% PCM microcapsule mass fractionmeans a decrease of 30% in comparison towater.

Viscosity at the Newtonian plateau(mPa$s)

4.9 6.4 18.3 Methodology proposed to measure viscositywith a control stress rheometer.Non-Newtonian behavior. Viscosity 3 timeshigher when increasing the PCM massfraction from 20% to 30%.

Fluid mechanics Pressure drop (in comparisonto water)

Differences little significant, especiallywhen increasing the turbulence degree

Between 2 and 3times higher

Although the pressure drop increases, theenergy ratio of improvement is higher than1 from velocities of 0.58, 0.39 and 0.98 forthe PCM microcapsule mass fraction of 14,20 and 30% respectively.

Heat transfer Heat transfer (forced convectiveheat transfer)-average improvementin comparison to water (massflow ¼ 20 kg/h and RTO ¼ 1.1)

28% 47% 35% Detailed description to obtain heat transfercoefficients by internal forced convection.The mPCM slurry with a 20% PCMmicrocapsules is the suitable mass fractionto be used as heat transfer fluid.

Physical stability Physical stability underthermo-mechanical loads

Broken microcapsules after having been pumped during 3weeks (10,000 meltingesolidification cycles)

Special care should be taken to avoid therupture of the PCM microcapsules. Thickercapsules will be more resistant, but on thecontrary the heat capacity will get worse.

M. Delgado et al. / Applied Thermal Engineering 63 (2014) 11e22 21

phenomenon in the analyzed range, probably as a consequence ofan incomplete phase change, an increase in the viscosity and adecrease in the thermal conductivity.

The slurries were observedwith an environmental SEM after thepumping (thermal and mechanical loads). In the case of the slurrywith 30% PCM microcapsules, the rupture of the microcapsulestogether with their subsequent joining together was observed afterthree weeks of pumping (having experienced approximately10,000 solidificationemelting cycles). When the experimentalprocess was stopped and the slurry was at rest, the same slurryblocked the control valve. However, the microparticles in the slurrywith 20% mass fraction were not ruptured after having beenpumped during two weeks (approximately 7000 solidificationemelting cycles).

It can be concluded that these slurries may be a suitable heattransfer and storage fluid. However, before any technical applica-tion, the obtaining of more resistant microcapsules should bestudied.

Acknowledgements

The authors would like to thank the Spanish Government forpartially funding this work within the framework of research pro-jects (MICINN-FEDER): ENE2008-06687-CO2/CON, ENE2011-28269-C03-01 and ENE2011-22722. Mónica Delgado is especiallygrateful to the Vice Deanship for Research of the University ofZaragoza for her grant.

References

[1] B. Zalba, J.M. Marín, L.F. Cabeza, H. Mehling, Review on thermal energy storagewith phase change: materials, heat transfer analysis and applications, Appl.Therm. Eng. 23 (3) (2003) 251e283.

[2] H. Inaba, New challenge in advanced thermal energy transportation usingfunctionally thermal fluids, Int. J. Therm. Sci. 39 (9e11) (2000) 991e1003.

[3] L. Royon, G. Guiffant, Forced convection heat transfer with slurry of phasechange material in circular ducts: a phenomenological approach, EnergyConvers. Manag. 49 (5) (2008) 928e932.

[4] L. Huang, M. Petermann, C. Doetsch, Evaluation of paraffin/water emulsion asa phase change slurry for cooling applications, Energy 34 (9) (2009) 1145e1155.

[5] X. Wang, J. Niu, Performance of cooled-ceiling operating with MPCM slurry,Energy Convers. Manag. 50 (3) (2009) 583e591.

[6] P.W. Griffiths, P.C. Eames, Performance of chilled ceiling panels using phasechange material slurries as the heat transport medium, Appl. Therm. Eng. 27(10) (2007) 1756e1760.

[7] S. Shibutani, PCM-microcapsule slurry thermal storage system for cooling inNarita Airport, in: 3rd Experts Meeting and Workshop of IEA, Annex 17, Tokyo(Japan), 2002.

[8] C. Pollerberg, C. Dötsch, Phase changing slurries in cooling and cold supplynetworks, in: 10th International Symposium on District Heating and Cooling,Section 8A, 13, Hannover (Germany), 2006.

[9] M. Delgado, A. Lázaro, J. Mazo, B. Zalba, Review on phase change materialemulsions and microencapsulated phase change material slurries: materials,heat transfer studies and applications, Renew. Sustain. Energy Rev. 16 (1)(2012) 253e273.

[10] M. Delgado, A. Lázaro, J. Mazo, J.M. Marín, B. Zalba, Experimental analysis of amicroencapsulated PCM slurry as thermal storage system and as heat transferfluid in laminar flow, Appl. Therm. Eng. 36 (2012) 370e377.

[11] Y. Zhang, Y. Jiang, Y. Jiang, A simple method, the T-history method, ofdetermining the heat of fusion, specific heat and thermal conductivity ofphase-change materials, Meas. Sci. Technol. 10 (3) (1999) 201e205.

[12] J.M. Marín, B. Zalba, L.F. Cabeza, H. Mehling, Determination of enthalpyetemperature curves of phase change materials with the T-history method eimprovement to temperature dependent properties, Meas. Sci. Technol. 14 (2)(2003) 184e189.

[13] A. Lázaro, E. Günther, H. Mehling, S. Hiebler, J.M. Marín, B. Zalba, Verificationof a T-history installation to measure enthalpy versus temperature curves ofphase change materials, Meas. Sci. Technol. 17 (8) (2006) 2168e2174.

[14] W.Q. Jin, T. Nagashima, S. Yoda, X.A. Liang, Z.L. Pan, Thermal diffusivitymeasurements of Li2B4O7 and KNbO3 solution by Laser Flash method, Chin.Phys. Lett. 19 (4) (2002) 569e571.

[15] J. Blumm, A. Lindemann, S. Min, Thermal characterization of liquids and pastesusing the flash technique, Thermochim. Acta 455 (1e2) (2007) 26e29.

[16] R. Coquard, B. Panel, Adaptation of the FLASH method to the measurement ofthe thermal conductivity of liquids or pasty materials, Int. J. Therm. Sci. 48 (4)(2009) 747e760.

[17] T. Bauer, D. Laing, U. Kröner, R. Tamme, Sodium nitrate for high temperaturelatent heat storage, in: The 11th International Conference on Thermal EnergyStorage-Effstock, Lecture Number 4, Stockholm (Sweden), 2009.

[18] S.A. Klein, Engineering Equation Solver, 2012. Academic Commercial V9.215.[19] G.K. Mukhamedzyanov, A.G. Usmanov, A.A. Tarzimanov, Determinations of

the thermal conductivity of liquid saturated hydrocarbons, Izv. Vyssh. Ucheb.Zaved. Neft Gaz 6 (9) (1963) 75e79.

[20] R.H. Perry, D.W. Green, Perry’s Chemical Engineers’ Handbook, Mc Graw Hill,Sidney, Australia, 1997.

M. Delgado et al. / Applied Thermal Engineering 63 (2014) 11e2222

[21] L. Huang, C. Doetsch, C. Pollerberg, Low temperature paraffin phase changeemulsions, Int. J. Refrig. 33 (8) (2010) 1583e1589.

[22] L. Royon, P. Perrot, G. Guiffant, S. Fraoua, Physical properties and thermo-rheological behaviour of a dispersion having cold latent heat-storage material,Energy Convers. Manag. 39 (15) (1998) 1529e1535.

[23] W. Lu, S.A. Tassou, Experimental study of the thermal characteristics of phasechange slurries for active cooling, Appl. Energy 91 (1) (2012) 366e374.

[24] B. Chen, X. Wang, R. Zeng, Y. Zhang, X. Wang, J. Niu, Y. Li, H. Di, An experi-mental study of convective heat transfer with microencapsulated phasechange material suspension: laminar flow in a circular tube under constantheat flux, Exp. Therm. Fluid Sci. 32 (8) (2008) 1638e1646.

[25] X. Wang, J. Niu, Y. Li, X. Wang, B. Chen, R. Zeng, Q. Song, Y. Zhang, Flow andheat transfer behaviors of phase change material slurries in a horizontal cir-cular tube, Int. J. Heat Mass Transf. 50 (13e14) (2007) 2480e2491.

[26] G.H. Zhang, C.Y. Zhao, Thermal and rheological properties ofmicroencapsulatedphase change materials, Renew. Energy 36 (11) (2011) 2959e2966.

[27] H.A. Barnes, A Handbook of Elementary Rheology, Institute of Non-NewtonianFluid Mechanics, University of Wales, Aberystwyth, United Kingdom, 2000.

[28] J.P. Carreau, Rheological equations from molecular network theories, Trans.Soc. Rheol. 16 (1) (1972) 99e127.

[29] S. Gschwander, P. Schossig, Phase Change Slurries as heat transfer and storagefluids for cooling applications, in: The 10th International Conference on Ther-mal Energy Storage-Ecostock, Lecture Number 55, New Jersey (USA), 2006.

[30] B.M. Diaconu, Transient thermal response of a PCS heat storage system, En-ergy Build. 41 (2) (2009) 212e219.

[31] W.M. Kays, Numerical solutions for laminar-flow heat transfer in circulartubes, Trans. ASME 77 (1955) 1265e1272.