International Journal of Heat and Mass Transfer 47 (2004) 2841–2847

www.elsevier.com/locate/ijhmt

An experimental study of enhanced heat transfer

in rectangular PCM thermal storage
Uro
s Stritih *

Faculty of Mechanical Engineering, University of Ljubljana, A
sker
ceva 6, 1000 Ljubljana, Slovenia

Received 14 July 2003

Abstract
The heat-transfer characteristics of a latent-heat storage unit with a finned surface have been experimentally studied
in terms of the solidification and melting processes by comparing them with those of a heat-storage unit with a plain
surface. Paraffin with a melting point of 30
C was used in the investigations because it is appropriate for thermal
storage applications in buildings. Time-based variations of the temperature distributions and heat flux are explained
from the results of observations of the melting and the solidification layers. The dimensionless Nusselt number was
calculated as a function of the Rayleigh number for natural convection in the paraffin for both the melting and the
solidification processes. The effectiveness of the fins was calculated from the quotient of the heat flux with fins and the
heat flux without fins.
 2004 Elsevier Ltd. All rights reserved.

1. Introduction Telkes [4] published the idea of using PCMs in walls;

however, these are better known as Trombe walls.
Thermal storage systems designed for the heating and Bordeau [5] tested a passive solar collector containing
cooling of buildings (HVAC applications) and hot-water CaCl2 Æ 6H2 O. He found that an 8.1-cm PCM wall has a
preparation are becoming increasingly important due to better thermal accumulation than a 40-cm concrete wall.
limited fossil-fuel reserves and the need to protect the An interesting possibility in building applications is the
environment. Thermal storage for HVAC applications impregnation of PCMs into porous construction mate-
can involve storage at various temperatures associated rials, e.g. plasterboard, to increase their thermal mass
with heating and cooling processes [1]. High-temperature [6–8].
storage is typically associated with solar energy, waste- PCMs have also been developed to store ‘‘coolness’’
heat utilization or heat-pump applications, whereas cool for air-conditioning applications: the ‘‘cold’’ is collected
storage is associated with air-conditioning, refrigeration, and stored in the PCM during the night, and then the
or cryogenic-temperature processes. PCM is used to cool the interior of a building during the
A technology that can be used to store large amounts hottest hours of the day. This concept is known as free-
of heat or cold in a defined volume has been the subject cooling [9–12].
of research for a long time. One of the possibilities is to Since the thermal conduction of PCMs is low, several
use a phase-change material (PCM) as the thermal methods exist to enhance the transfer of the heat. The
storage media. Phase-change materials were first used in use of finned tubes with different configurations has been
British trains to prevent them from becoming too cold. proposed by various researchers, these include: Abhat
The first reports of a PCM described in the literature et al. [13], Marcos [14], Sadusuke and Naokatsu [15],
were applications for heating and cooling in buildings, Costa et al. [16], Padmanabhan and Krishna Murthy
by Telkes in 1975 [2], and Lane in 1986 [3]. In 1978 [17], Velraj et al. [18,19] and Ismail et al. [20].
Several other heat-transfer enhancement techniques
have also been reported. Siegel [21] studied improve-
*
Tel.: +386-1-4771-312; fax: +386-1-2518-567. ments in the solidification rate for molten salt containing
E-mail address: [email protected] (U. Stritih). a dispersion of high-conductivity particles. Another
0017-9310/$ - see front matter  2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijheatmasstransfer.2004.02.001
2842 U. Stritih / International Journal of Heat and Mass Transfer 47 (2004) 2841–2847

Nomenclature

a temperature conductivity (m2 /s) T temperature (K)

Bi Biot number(–) V volume (m3 )
cp specific heat at constant pressure (J/kg K) x coordinate (m)
D distance between fins (m) X dimensionless coordinate (–)
Fo Fourier Number (–) y coordinate (m)
h specific enthalpy (J/kg) Y dimensionless coordinate (–)
l length (m) z coordinate (m)
L paraffin depth (m)
m mass (kg) Greek symbols
q specific heat (J/kg) a thermal convection coefficient (W/m2 K)
qt fusion heat (J/kg) k thermal conductivity (W/m K)
Q heat (J) q density (kg/m3 )
Ste Stefan number (–) / heat flux (W)
t time (s)

method was to embed the PCM in a metal-matrix 2. Experimental setup

structure [22–26]. And the use of thin aluminium plates
filled with a PCM was developed by Bauer and Wirtz An experimental setup for the analysis of heat
[27]. transfer was installed in the laboratory.
Mehling and co-workers [28–31] and Py et al. [32] Rectangular thermal storage is used when it is pos-
proposed a graphite compound material, where the sible to transfer heat via a water heat exchanger on one
PCM is embedded inside a graphite matrix. The main wall of the thermal storage, as shown in Fig. 1. The heat
advantage of such a material is the increased heat con- storage (1) has dimensions of 650 · 500 · 120 mm. On
ductivity in the PCM without much of a reduction in the one wall there is a heat exchanger (2) where heat can be
energy storage, but other advantages include a decrease transferred from the water to the PCM (melting) or vice
in the subcooling of salt hydrates and a decrease in the versa (solidification). The outlet (3) and the inlet (4) for
volume change of the paraffins. the water are connected with pipes. The water flows
There are a lot of PCM applications where rectan- from an aqueduct (7) through a rotameter (6) into the
gular thermal storage is needed. One of them is a solar electric heater (5) or directly into the experimental device
PCM storage wall for ventilating buildings [33]. The (2). The thermal storage is insulated with polystyrene
experiments described in the rest of this paper will ex- that has a thermal conductivity of k ¼ 0:038 W/m K.
plain the physical process of heat transfer in this rect- By measuring the volume flow of the water together
angular storage. with the inlet and outlet temperatures it is possible to

Fig. 1. Rectangular thermal storage with experimental setup.

 U. Stritih / International Journal of Heat and Mass Transfer 47 (2004) 2841–2847 2843

analyse the heat in to (and out of) the heat store. The / ¼ mc
_ p DT ð1Þ
signals from the thermocouples marked in Fig. 1 with
AðxÞ (8) are fed to the ADAM module (9) and stored in a The measurement uncertainty of the heat flux is as fol-
PC (10). lows:
"

 2

 2
The experiments were conducted in such a way that o/ o/
M/ ¼ Mm þ Mcp
the melting and solidification were carried out once with om ocp
fins and once without fins, in order to determine the

 2 #1=2
enhancement effect of the fins. RUBITHERM RT 30 o/
þ MDT ð2Þ
paraffin was used for the tests. Fig. 2 shows a DSC oDT
analysis of this paraffin.
The fins for the heat-transfer enhancement were made where
of steel with a thermal conductivity k ¼ 20 W/m K. The
o/ o/ o/
32 fins were rectangular, and had the following dimen- ¼ cp DT ; ¼ m DT ; ¼ mcp ð3Þ
sions: height 0.5 m, length 0.12 m, thickness 1 mm. om ocp oDT
A panel with fins is presented in Fig. 3: By combining Eq. (3) with (2) we obtain:
The temperatures were measured with type-K ther-
mocouples (NiCr–Ni). These thermocouples were thin M/ ¼ ððcp DTMm Þ2 þ ðm DTMcp Þ2 þ ðmcp MDT ÞÞ2 ð4Þ
(0.6 · 1.0 mm) because the temperatures were measured
in the paraffin. The measurement uncertainty, as de- and dividing Eq. (4) by Eq. (1) gives us:
scribed by the standard IEC 584-2, was ±0.5
C. "
2

2 #1=2
M/ Mm M cp 2 MDT
The temperatures of the inlet and the outlet water ¼ þ þ ð5Þ
were also measured. Using this data the heat flux was / m cp DT
calculated with the following equation: M/ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
¼ 0:01252 þ 0:0052 ¼ 0:0135 ð6Þ
/

The measurement uncertainty of the testing line was

DSC diagram for PCM ±1.35%.
25000
Specific heat (J/KgK)

20000 3. Results and discussion

15000
The experiments were made for the following reasons:
10000

5000 • to determine the temperature distribution in the heat

storage,
0
0 10 20 30 40 50 60 • to test the changing interface position between the so-
Temperature (C) lid and the liquid,
• to analyse the process of heat transfer in the heat
Fig. 2. DSC diagram for analysed PCM Rubitherm RT 30. storage,
• to determine the process of charging and discharging
of the heat storage.

Heat transfer in latent-heat storage is very complex.

The process is non-stationary, there is also a phase-
change problem, and the process is non-homogeneous
because of the fins. Time-dependent charging and dis-
charging of the process is the most important for solar-
energy applications.
The results are presented in two parts: first, the re-
sults of the melting; and second, the results of the
solidification.

3.1. Melting

The melting of the paraffin without fins is presented

Fig. 3. Fins for heat-transfer enhancement analysis. first. The measurements began with the temperature of
2844 U. Stritih / International Journal of Heat and Mass Transfer 47 (2004) 2841–2847

80

70

Temperature (C)
60

50
A1

40 A2
A3
30 A6
A7
20

10
0 50 100 1 50 2 00 250 300 350

Time (min)

Fig. 4. Melting of the paraffin without fins.

350 2000
1800
300
Phase interface (mm)

ana l. 1600
y = 0 ,022x 1 ,7471
250 eksp. 1400
Heat flux (W)

R2 = 0, 99 77
eksp. (fun kc.) 1200 Heat flux
200
Heat flux (funct.)
1000
150 800

10 0 600
400 y = 3 387,8x -0 ,390 8
50 R2 = 0, 7694
200

0 0
0 50 100 150 200 250 300 0 50 1 00 150 200 250 300 350

Time (min) Time (min)

Fig. 5. Phase-change interface position. Fig. 6. Heat flux at heating without fins.

thermal storage at room temperature. Next, hot water at data, a correlation for the heat transfer in the heat
65
C was fed into the heat exchanger. Temperatures storage can be written as:
were measured at five main points, and the results are
presented in Fig. 4. From the results we can see that the Nu ¼ 8  1012 Ra1:0392 ð7Þ
phase transformation takes place at points where the
line is no longer smooth. where the following equations have been used [34].
Fig. 5 shows the comparison of the measured phase-
change interface position with the analytical results al
Nu ¼ ð8Þ
(conduction). From the results we can see that the data k
can only be compared at the beginning of the process. Rt
/ /ðtÞ dt
The reason for this is natural convection, which in- a¼ ¼ 0 ð9Þ
AðTw  Tt Þt AðTw  Tt Þ
creases the heat transfer.
The measured heat flux into the heat storage can be gl3 ðTw  Tt Þ m
calculated from the inlet and outlet temperatures of the Ra ¼ Gr
Pr ¼
ð10Þ
m2 a
water, and is presented in Fig. 6. From the integration of
this heat flow we could obtain a value for the heat that The melting of the paraffin with and without fins under
was stored in the paraffin. the same conditions is presented in Fig. 7, which shows
Heat transfer is described with the Nusselt number, the temperature variations.
which is a function of various parameters. During nat- The heat flux as a function of time was also measured
ural convection the Rayleigh number (Ra ¼ Gr
Pr) is for the melting of the paraffin with fins. To compare the
the main parameter. On the basis of the experimental heat transfer with and without fins a quotient of both
 U. Stritih / International Journal of Heat and Mass Transfer 47 (2004) 2841–2847 2845

70 3.2. Solidification
60
The solidification of the paraffin without the fins is
Temperature (C)

50 presented in Fig. 9. The experiment was made with the

cooling of the melted paraffin from Fig. 4. Cold water at
40 A1
A2
18
C was fed through the storage and the result of the
30
A3 temperature change of the paraffin shows us the inter-
20 A6 face position.
A7 Solidification is a process where natural convection
10 can be neglected in comparison with melting. This can
0 50 100 150 200 250 300
Time (min) be seen from Fig. 10, where the measured data are
compared with the analytical data (conduction).
Fig. 7. Melting of the paraffin with fins. The comparison is good, and the conclusion could be
that solidification can be modelled only by conduction
(k). Similar results were also obtained by other authors,
and these can be found in the literature. This can also
2
confirm the dimensionless equation (14):
1.8
1.6 Nu ¼ 7  1013 Ra0:9364 ð14Þ
1.4
A comparison of Eq. (14) with Eq. (7) shows that nat-
1.2
ural convection in the case of melting is more than 10
1
ηf

times higher than during solidification.

0.8
0.6
0.4
A1
0.2
35 A2
0 33 A3
0.00E+00 1.00E-04 2.00E-04 3.00E-04 4.00E-04 A6
31
Temperature (C)

Ste*Fo A7
29
27
Fig. 8. Fin effectiveness at melting of the paraffin.
25
23
21
heat fluxes was made. This is referred to as the fin 19
effectiveness (Eq. (11)): 17
15
q_ fin 0 10 20 30 40 50 60 70
gf ¼ ð11Þ Time (h)
q_ without fin
Fig. 9. Temperature variations at solidification of the paraffin.
This ratio is a function of time and is presented in Fig. 8
in dimensionless form (s): where dimensionless time is
written as a multiplication of the Stefan and Fourier
140
numbers, as in Eq. (12):
y = 7,3308x0,6124
120 R2 = 0,9932
cp ðTs  Tt Þ at
Phase interface (mm)

s¼ ¼ Ste
Fo ð12Þ
qt H2 100

80
The approximation of the function is written in the anal.
eksp.
following equation (R2 ¼ 0:87): 60
eksp. (funkc.)
7 2 40
gf ¼ 1  10
ðSte  FoÞ  1719:4  ðSte  FoÞ þ 0:7813
20
ð13Þ
0
0 20 40 60 80 100
From the results it can be seen that for a low Fourier
Time (h)
number the ratio is less than 1. The reason for this is that
natural convection, which is dominant because of the Fig. 10. Phase-change interface position at solidification with-
melting, is reduced because of the fins. out fins.
2846 U. Stritih / International Journal of Heat and Mass Transfer 47 (2004) 2841–2847

40 gives us an equation for the temperature distribution in

terms of place and time. Since ideal cases are not often
35 found in the nature, numerical methods that provide us
Temperature (C)

A1
A2 with temperature and velocity fields can also be used.
30 A3
However, because of the complexity of systems, experi-
A6
A7
ments are usually one of the best ways in which we can
25
obtain a measure of the real performance of a process.
20
We analysed the heat-transfer enhancement in a
phase-change material in the Laboratory for Heating,
15 Sanitary and Solar Technology at the Faculty of
0 10 20 30 40 50 60
Mechanical Engineering in Ljubljana. The heat storage
Time (h)
with a phase-change material was made with a heat
Fig. 11. Solidification of the paraffin with fins. exchanger where the heat could be transferred from the
heat-transfer media (water) to the phase-change mate-
4
rial.
3, 5
Measurements were made on the melting and solid-
3
ification of paraffin, and the time-dependent storage of
2, 5
heat in the PCM with and without fins for heat-transfer
enhancement was analysed. The results were presented
ηf

2
1, 5 in terms of temperature and heat flux as a function of
1 time.
0, 5 A new correlation between the dimensionless num-
0 bers (Nusselt number as a function of Rayleigh number)
0.00E+00 4.00E-04 8.00E-04 1.20E-03 1.60E-03 was made. A comparison of the equations for melting
Ste*Fo and freezing shows that natural convection is present
during melting and increases the heat transfer. During
Fig. 12. Fin effectiveness at solidification with fins.
solidification, conduction is the dominant form of heat
transfer. We can conclude that heat storage (melting) is
not a problem during thermal storage applications, and
The solidification of the paraffin with the fins is
that the extraction of heat (solidification) can be effec-
presented in Fig. 11. The temperature variation shows a
tively enhanced with fins.
typical solidification process. The lines are very steep at
The fin effectiveness was also investigated. It was
the beginning, after which there is a change in the incline
found that for a low Fourier number the ratio is less
because the heat is extracted as latent heat and not as
than 1. The reason for this is that natural convection,
sensible heat (DT ).
which is dominant during melting, is reduced because of
The fin effectiveness, which is presented in Fig. 12,
the fins. The fin effectiveness during solidification shows
shows that the value increases up to 3.06 and then de-
that the value is increasing up to 3.06 and is then
creases. The same results were also obtained by other
decreasing. This shows that heat transfer during solidi-
authors [35]. It is obvious that the heat transfer during
fication is greater if the fins are included. In our
solidification is greater if fins are included. In our case a
case a 40% reduction in the solidification time was cal-
40% reduction in the solidification time was calculated.
culated.
The functional dependence of the fin efficiency is
presented with Eq. (15) (R2 ¼ 0:787):
gf ¼ 2  106  ðSte  FoÞ2 þ 3000:4  ðSte  FoÞ þ 1:0567 Acknowledgements
ð15Þ
The authors appreciate the financial support pro-
vided by the Ministry of Education, Science and Sport
of the Republic of Slovenia.
4. Conclusion

Melting and solidification are physical processes that References

are used in different areas: in the food industry (the
freezing of food), in metallurgy and in heat-storage [1] ASHRAE HANDBOOK, HVAC Applications SI Edition,
processes. 1999.
There are many different ways to investigate these [2] M. Telkes, Thermal storage for solar heating and cooling,
processes. The oldest is the analytical principle, which in: Proceedings of the Workshop on Solar Energy Storage
 U. Stritih / International Journal of Heat and Mass Transfer 47 (2004) 2841–2847 2847

Subsystems for the Heating and Cooling of Buildings, around a vertical axially finned isothermal cylinder, Appl.
Charlottesville, Virginia, USA, 1975. Thermal Eng. 21 (2001) 53–77.
[3] G.A. Lane, Solar Heat Storage: Latent Heat Material, vol. [21] R. Siegel, Solidification of low conductivity material
II, Technology, CRC Press, Florida, 1986. containing dispersed high conductivity particles, Int. Heat
[4] M. Telkes, Trombe wall with phase change storage Mass Transfer 20 (1977) 1087–1089.
material, in: Proceedings of the 2nd National Passive Solar [22] M. Kamimoto, Y. Abe, S. Sawata, T. Tani, T. Ozawa,
Conference, Philadelphia, PA, 1978. Latent heat storage unit using form-stable high density
[5] L.E. Bordeau, Study of two passive solar systems contain- polyethylene for solar thermal applications, in: Proceedings
ing phase-change material for thermal storage, in: Pro- of the International Symposium on Thermal Application
ceedings of the 5th National Passive Solar Conference, of Solar Energy, Hakone, Kanagawa, Japan, 1985.
1980. [23] C.J. Hoogendoorn, G.C.J. Bart, Performance and model-
[6] M.M. Shapiro, D. Feldman, D. Hawes, D. Banu, PCM ling of latent heat stores, Solar Energy 48 (1992) 53–
thermal storage in wallboard, in: Proceedings of the 12th 58.
Passive Solar Conference, Portland, Oregon, USA, 1987. [24] M.A. Khan, P.K. Rohatgi, Numerical solution to a moving
[7] I.O. Salyer, A.K. Sircar, R.P. Charloff, Advanced phase- boundary problem in a composite medium, Numer. Heat
change materials for passive solar storage applications, in: Transfer 25 (1994) 209–221.
Proceedings of the Solar Building Conference, Silver [25] X. Tong, J. Khan, M.R. Amin, Enhancement of heat
Spring, Washington, D.C., USA, 1985. transfer by inserting a metal matrix into a phase change
[8] K. Peippo, P. Kauranen, P.D. Lund, A multicomponent material, Numer. Heat Transfer, Part A 30 (1996) 125–
PCM wall optimized for passive solar heating, Energy 141.
Buildings 17 (1991) 259–270. [26] X. Tong, J.A. Khan, M.R. Amin, Enhancement of heat
[9] B. Zalba, Free-cooling, an application of PCMs in TES, transfer by inserting a metal matrix into a phase change
Ph.D. Thesis, University of Zaragoza, Spain, 2002. material, Numer. Heat Transfer, Part A 30 (1996) 125–
[10] S.M. Vakilaltojjar, W. Saman, Analysis and modeling of a 141.
phase change storage system for air conditioning applica- [27] C.A. Bauer, R.A. Wirtz, Thermal characteristics of a
tions, Appl. Thermal Eng. 21 (2001) 249–263. compact, passive thermal energy storage device, in: Pro-
[11] S.M. Vakilaltojjar, W. Saman, Domestic heating and ceedings of the 2000 ASME IMECE, Orlando, Florida,
cooling with thermal storage, in: Proceedings of Terra- USA, 2000.
stock, Stuttgart, Germany, 2000. [28] P. Satzger, B. Exka, F. Ziegler, Matrix-heat-exchanger for
[12] F. Bruno, W. Saman, Testing of a PCM energy storage a latent-heat cold-storage, in: Proceedings of Megastock
system for space heating, in: Proceedings of the World ’98, Sapporo, Japan, 1998.
Renewable Energy Congress VII, Cologne, Germany, [29] H. Mehling, S. Hiebler, F. Ziegler, Latent heat storage
2002. using a PCM-graphite composite material: advantages and
[13] A. Abhat, S. Aboul-Enein, N. Malatidis, Heat of fusion potential applications, in: Proceedings of the 4th Work-
storage systems for solar heating applications, in: C. Den shop of IEA ECES IA Annex 10, Bendiktbeuern, Ger-
Quden (Ed.), Thermal Storage of Solar Energy, Martinus many, 1999.
Nijhoff, 1981. [30] H. Mehling, S. Hiebler, F. Ziegler, Latent heat storage
[14] V.H. Marcos, Investigation of latent heat thermal energy using a PCM-graphite composite material, in: Proceedings
storage system, Solar Wind Technol. 7 (2/3) (1990) 197– of Terrastock 2000––8th International Conference on
202. Thermal Energy Storage, Stuttgart, Germany, 2000, pp.
[15] I. Sadasuke, M. Naokatsu, Heat transfer enhancement by 375–380.
fins in latent heat thermal energy storage devices, Solar [31] L.F. Cabeza, H. Mehling, S. Hiebler, F. Ziegler, Heat
Eng., ASME (1991) 223–228. transfer enhancement in water when used as PCM in
[16] M. Costa, D. Buddhi, A. Oliva, Numerical simulation of thermal energy storage, Appl. Thermal Eng. 22 (2002)
latent heat thermal energy storage system with enhanced 1141–1151.
heat conduction, Energy Convers. Manage. 39 (3/4) (1998) [32] X. Py, R. Olives, S. Mauran, Paraffin/porous-graphite-
319–330. matrix composite as a high and constant power thermal
[17] P.V. Padmanabhan, M.V. Krishna Murthy, Outward storage material, Int. J. Heat Mass Transfer 44 (2001)
phase change in a cylindrical annulus with axial fins on 2727–2737.
the inner tube, Int. J. Heat Mass Transfer 29 (1986) 1855– [33] U. Stritih, P. Novak, Solar energy storage in phase change
1868. material, The Second ISES Europe Solar Congress, Euro-
[18] R. Velraj, R.V. Seeniraj, B. Hafner, C. Faber, K. Schwar- sun 98, Slovenia, 1998.
zer, Experimental analysis and numerical modelling of [34] D. Streit, Free convection on horizontal pipe with stone/
inward solidification on a finned vertical tube for a latent water filling with making and melting of ice, Doctoral
heat storage unit, Solar Energy 60 (1997) 281–290. dissertation, mentor E. Hahne, University of Stuttgart,
[19] R. Velraj, R.V. Seeniraj, B. Hafner, C. Faber, K. Schwar- 1996.
zer, Heat transfer enhancement in a latent heat storage [35] K. Sasaguchi, H. Imura, Heat transfer characteristics of a
system, Solar Energy 65 (1999) 171–180. latent heat storage unit with a finned tube 1st report:
[20] K.A.R. Ismail, C.L.F. Alves, M.S. Modesto, Numerical experimental studies on solidification and melting pro-
and experimental study on the solidification of PCM cesses, Bull. JSME 29 (255) (1986).