Вы находитесь на странице: 1из 5

PIERS ONLINE, VOL. 2, NO.

6, 2006 667

Properties and Numerical Simulation of CaCl2 .6H2 O Phase


Change
I. Behunek, T. Bachorec, and P. Fiala
Department of Theoretical and Experimental Electrical Engineering
Faculty of Electrical Engineering and Communication
Brno University of Technology, Kolejni 2906/4, 612 00 Brno, Czech Republic

Abstract— This paper deals with the application of phase change materials (PCM) for thermal
management of integrated circuits as a viable alternative to active forced convection cooling sys-
tems. The paper presents an analytical description of melting and freezing process, an analytical
solution of CaCl2 .6H2 O phase change in 1D and the results of a numerical model of phase change.
The numerical model joins thermal field, air flow with respects to material phase change. This
numerical model was obtained by means combined of the finite element method (FEM) and finite
volume method (FVM) applied as main method in ANSYS software.
DOI: 10.2529/PIERS060901084138

1. INTRODUCTION
Phase change materials can store large amounts of heat without undergoing significant temperature
changes because of their high latent heat of fusion. We know different PCM — inorganic, organic,
and combinations of amorphous and crystallic substances, clathrates etc. The stored energy is
given by the calorimetric equation

ZTm ZTe
Q = ρV ∆hm + V ρcdT + V ρcdT (1)
T0 Tm

where ρ is density, V volume, c specific heat, ∆hm enthalpy, Q heat, and Tm temperature. The
PCM may be used for active or passive electronic cooling applications with high power at the
package level.

ribs

PCM

processor

Figure 1: Example of processor cooler with phase change material.

2. PROPERTIES OF CACL2 .6H2 O


Big latent heat, good thermal conductivity and inflammability are the main advantages of inorganic
materials. But they cause corrosion and suffer from loss the of water. Incongruent melting and
supercooling are the biggest problem of their exploitation. During melting and freezing there are
precipitations of other phases, which do not take part in next process of charging and discharging.
Poor nucleation, slow rate of crystal growth or high rate of heat removal may be the reason for
supercooling. Impurities have a strong influence on the cooling curves as well.
PIERS ONLINE, VOL. 2, NO. 6, 2006 668

40 40

38
35
36

34 30

32
25

°C
°C

30
20
28

26 15

24
10
22

20 5
0:00 0:21 0:43 1:04 1:26 1:48 2:09 0:00 0:36 1:12 1:48 2:24 3:00 3:36 4:12

timeas[hours]
[hodiny] as [hours]
time [hodiny]

Figure 2: Phase change of CaCl2 .6H2 O (right — melting, left — freezing).

In the Fig. 2 there is exhibited the phase change of CaCl2 .6H2 O during heating and cooling.
Dashed lines would show theoretical behaviour if melting and freezing were at constant temperature
Tm — case of pure crystallic substance. Impurity and methodology of measuring (probe always is
only in small amount of hexahydrate) are the causes of variances. During solidification the super-
cooling occurred because of weak nucleation. Crystallization was initiated due to a solid particle
of PCM which we put to the measured sample. Otherwise there would not be any crystallization.
We can use plastics, mild steel or copper for enclosures. Aluminium or stainless steel are not suit-
able. Sometimes we can see temperature fluctuation above Tm during solidification. We found an
explanation in the binary diagram.

150

100
o
C

50

40 50 60 70
water % CaCl 2
Figure 3: Binary diagram of CaCl2 .6H2 O.

Modification is possible to do by additives. From number of potential candidates Ba(OH)2 ,


BaCO3 , and Sr(OH)2 were chosen. They seemed to be feasible. When we used Ba(OH)2 and
Sr(OH)2 at 1% part by weight there was no supercooling. We could increase stability of the
equilibrium condition with addition of KCl (2 wt%) and NaCl. NaCl is weak soluble in CaCl2 .6H2 O,
therefore part by weight is only about 0.5%. Disadvantage is that melting point decreases about
3◦ C at 26–27◦ C. The melting point of pure CaCl2 .6H2 O is 29.6◦ C. Due to availability on the
market and price we chose for modification BaCO3 . We obtained the best results for 1.2 wt%. In
Fig. 1 we can see that supercooling is 3–4◦ C but then crystallization started spontaneously and
temperature increased at 28–29◦ C. It is obvious that nucleation is slower in comparison with pure
PIERS ONLINE, VOL. 2, NO. 6, 2006 669

crystallic matter. Supercooling is not considered as a disadvantage. If we use CaCl2 .6H2 O in the
heat accumulator we will be able to store energy at lower temperature (about 26◦ C) and suppress
heat losses. Next disadvantage of BaCO3 is that it carbonates because of atmospheric CO and
CO2 . This means the loss of properties. In our case BaCO3 will be isolated from surrounding
environment.
3. ANALYTICAL DESCRIPTION AND SOLUTION OF PHASE CHANGE
The analytical solution is exact but we consider several simplified assumptions. The most important
of them is that we can solve solidification of PCM only in one dimensional body. The 1D body
could be semifinite plane, cylinder or sphere.

Figure 4: Solidification of semi-infinite plate of PCM.

Interface position s = s(t) Temperature distribution T = T(t,x)


0, 03 40 x [m]
0, 0275
35 0, 001
0, 025
0, 005
0, 0225 30
0, 01
0, 02
tem perat ure [˚C]

25 0, 015
0, 0175
0, 02
s [m ]

0, 015 20
0, 025
0, 0125
15 0, 03
0, 01
0, 04
0, 0075 10
0, 05
0, 005
5
0, 0025

0 0
0 430 860 1290 1720 2150 2580 3010 3440 3870 4300 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000
time [s ] time [s ]

Figure 5: Interface position and temperature distribution in CaCl2 .6H2 O.

We consider semi-infinite mass of liquid PCM at initial temperature T0 which has been cooled
by a sudden drop of surface temperature Tp = 0◦ C. This temperature is constant during the whole
process of solidification. The simplified assumptions are
• body is semi-infinite plane, heat flux is one-dimensional in the x-axis, plane is an interface
between the solid and the liquid, there is an ideal contact on the interface, temperature
of surface is constant (Tp = 0◦ C), crystallization of PCM is at constant temperature Tm ,
thermophysical properties of the solid and the liquid are different but they do not depend on
temperature, there is no natural convection in the liquid.
Initial and boundary conditions:
• initial temperature T0 for x ≥ 0 at time 0 s
• temperature equals Tm on the interface between the solid and the liquid (x = s)
x=s∧t>0 ⇒ Ts = Tl = Tm = const (2)
PIERS ONLINE, VOL. 2, NO. 6, 2006 670

46

44
42
40

38

temperature [°C]
36

34

32

30
28

26

24
22

20
0:00 0:19 0:38 0:57 1:17 1:36 1:55 2:15 2:34 2:53 3:13 3:32 3:51

time [hours]

measuring simulation

Figure 6: Phase change of CaCl2 .6H2 O (the comparison between measuring and simulation).

• evolved latent heat during a motion of interface (volume element ds, area 1 m2 , time 1 s)

ds
dQ∆hm = ∆hm ρl 1 (3)
dt

• position of interface is function of time



s = s(t) = 2ε as t, (4)

this dependence is called the parabolic law of solidification.

• boundary condition for phase change


µ ¶ µ ¶
∂Ts ∂Tl ds
λs = λl + ∆hm ρl (5)
∂x x=s ∂x x=s dt

for

x → ∞ ∧ t > 0 ⇒ Tl = T0 = const (6)


x = 0 ∧ t ≥ 0 ⇒ Tp = Ts (x = 0) = 0◦ C (7)

If we solve fourier relations of heat conduction with conditions above for the solid and the liquid,
we get the following equations which allow calculating temperatures in solid, liquid PCM and the
location of interface. The results are in the Fig. 5 which are related to Eqs. (8), (9), (10).
2 r − as ε2 √
e−ε λl as T0 − Tm e al ∆hm ρ1 εas π
− ³ q ´= (8)
erf(ε) λs al Tm erfc ε as λs Tm
al

We get ε = 0.3514 for CaCl2 .6H2 O. Temperatures in the solid and liquid part of body are
µ ¶
Tm x
Ts = erf √ (9)
erf(ε) 2 as t
µ ¶
T0 − Tm x
Tl = T0 + ³ q ´ erf √ (10)
erfc ε as 2 al t
al
PIERS ONLINE, VOL. 2, NO. 6, 2006 671

4. NUMERICAL SOLUTION
In Fig. 6 there is a result of numerical model. We can see time dependence of CaCl2 .6H2 O tempera-
ture in the package. We can compare results obtained by numerical simulation with the measuring.
Differences between simulation and measuring are caused due to inaccuracy of model with respect
to reality. We used tabular values of pure CaCl2 .6H2 O but in pipes there is modified hexahydrate
with 1.2% of BaCO3 . We would need to know temperature dependence of thermal conductivity,
specific heat, and density during phase change exactly.
5. CONCLUSION
We showed measured properties of CaCl2 .6H2 O and did its modification in order to avoid supercool-
ing. There is the demonstration in the article that we can apply theory of solidification of metals
to PCM. We presented analytical description and solution phase change in 1D. In 3D we had to
exploit numerical method. The numerical model was solved by means of combined FEM/FVM
as main software in ANSYS program. If we compare the results of simulation with experimental
measuring, we see quite good congruence.
ACKNOWLEDGMENT
The paper was prepared within the framework of the research plan No. MSM 0021630516 of the
Ministry of Education, Youth and Sports of the Czech Republic.
REFERENCES
1. Ansys User’s Manual, Svanson Analysys System, Inc., Huston, USA, 2003.
2. Běhunek, I. “Properties of inorganic PCM,” Honeywell EMI Conference and Competition 2005,
Brno, Ing. Zdeněk Novotný CSc., Ondráčkova 105, ISBN 80-214-2942-9, Brno, 2005.
3. Lane, G. A., Solar Heat Storage: Latent Heat Materials, Volume II: Technology, CRC Press,
Inc., ISBN 0-8493-6586-4, Boca Raton, Florida, USA, 1986.

Вам также может понравиться