- Email: [email protected]
Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs)
Applied Energy xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Applied Energy journal homepage: www.elsevier.com/locate/apenergy
Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs) q Xiaoqin Sun a,b, Quan Zhang c, Mario A. Medina b,⇑, Kyoung Ok Lee b a b c
School of Energy and Power Engineering, Changsha University of Science and Technology, Changsha 410114, China Dept. of Civil, Environmental & Architectural Engineering, University of Kansas, Lawrence 66045, USA College of Civil Engineering, Hunan University, Changsha 410082, China
h i g h l i g h t s Heat transfer rate was enhanced by natural convection of PCMs. A mean heat transfer enhancement factor of 1.2 was observed during melting. The time for melting was reduced by approximately 45% because of natural convection.
a r t i c l e
i n f o
Article history: Received 29 September 2014 Received in revised form 14 February 2015 Accepted 16 March 2015 Available online xxxx Keywords: Natural convection Phase change materials (PCMs) Heat transfer Melting process
a b s t r a c t Natural convection is one of the major factors that affect phase transition processes of solid–liquid phase change materials (PCMs). To optimize PCM-based latent thermal energy storage systems (TESS), a better understanding of the heat transfer interactions during these transitions is needed. In this paper the heat transfer rate enhancement caused by natural convection of PCMs undergoing melting is quantified based on experimental observations. For this, a heat transfer enhancement factor and an effective heat transfer coefficient were developed. Differential scanning calorimetry (DSC) tests were run to measure latent heats of fusion and phase transition temperatures of the PCMs. It was found that the experimentallyobtained temperature ranges required for complete melting exceeded those produced by the DSC tests. The reason for this stemmed from the natural convection of the molten PCM. The effective heat transfer coefficients when natural convection was accounted for were greater than when only heat conduction was considered. The increases in effective heat transfer coefficient were 12% and 30% percent for vertical and horizontal heat transfer paths, respectively. The existence of natural convection reduced the time required for complete melting by approximately 45% for vertical heat transfer. However, the melting process time was longer than the solidification process under the same conditions for a horizontal heat transfer path. The reason for this was attributed to the widening of the temperature range required for melting. Ó 2015 Published by Elsevier Ltd.
1. Introduction Thermal energy storage systems (TESS) are used to regulate the time-of-use mismatch between energy supply and consumption
q This article is based on a short proceedings paper in Energy Procedia Volume 161 (2014). It has been substantially modified and extended, and has been subject to the normal peer review and revision process of the journal. This paper is included in the Special Issue of ICAE2014 edited by Prof. J Yan, Prof. DJ Lee, Prof. SK Chou, and Prof. U Desideri. ⇑ Corresponding author at: 1530 W 15th Street, 2150 Learned Hall, University of Kansas, Lawrence, KS 66045, USA. Tel.: +1 785 864 3604. E-mail address: [email protected] (M.A. Medina).
[1]. Among TESS, which include sensible energy storage [2], latent energy storage [3], and chemical energy storage [4], latent energy storage systems that use solid–liquid phase change materials (PCMs) are particularly attractive. The advantages of using solid– liquid PCMs are [5]: (1) smaller weight and volume of material are required for a given amount of energy storage, compared to conventional sensible heat energy storage systems; (2) heat energy is stored at a constant or nearly constant temperature, which corresponds to the phase transition temperature of the PCM. Heat energy is stored in phase changing substances when these melt and it is recovered when these solidify. Baran and Sari [6] studied a mixture of palmitic and stearic acids as PCM within a tube-and-shell heat exchanger for solar
http://dx.doi.org/10.1016/j.apenergy.2015.03.078 0306-2619/Ó 2015 Published by Elsevier Ltd.
Please cite this article in press as: Sun X et al. Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs). Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.03.078
X. Sun et al. / Applied Energy xxx (2015) xxx–xxx
space heating applications. They concluded that the melting and solidification of PCMs was highly influenced by natural convection in the case of melting and by conduction in the case of solidification. Kaizawa et al. [7] analyzed the thermal behavior, during the phase transition processes, of solid–liquid PCMs within a direct heat exchanger. A heat transfer fluid flowed within the PCM to improve its heat transfer rate by agitating the liquid PCM. Therefore, the heat convection, instead of heat conduction, dominated the heat transfer process. Duan and Naterer [8] investigated the feasibility and effectiveness of two different PCM designs used in electric vehicles. They found that the natural convection within the melted PCM enhanced heat transfer rate to and from the PCMs. Longeon et al. [9] experimentally and numerically studied the influence of the natural convective heat transfer mechanism during phase transition processes. They concluded that the natural convection enhanced the melting process, whereas the liquid PCM behind the solidification front remained almost isothermal during the solidification process. Based on this research, a top injection for melting and a bottom one for solidification were recommended for a coaxial thermal energy storage unit. In summary, there is a close relationship between natural convection in molten-state PCMs and the enhancement of heat transfer to and from the PCM during phase transitions, especially during the melting process. Kozak et al. [10] found that close-contact melting in a vertical double pipe concentric storage unit with circumferentially finned inner tube significantly enhanced the melting process. Initiating close-contact melting was recommended to obtain higher heat transfer rates and shorter melting times. Pal and Joshi [11] studied a side heated tall enclosure of aspect ratio 10. They developed correlations of heat transfer rate via Nusselt number, Nu, and melting fraction, via a parameter e, and pointed out that melting was notably affected by natural convection during the later stages of the process. Jany and Bejan [12] separated the melting process with natural convection in an enclosure heated from the side into four regimes. Correlations for heat transfer rate and melting front location during each regime were theoretically and experimentally developed to explain the development of natural convection during melting. Ho and Viskanta [13] studied the phase transitions of n-octadecane in a rectangular cavity with varying aspect ratios under isothermal wall temperature boundary conditions. It was found that natural convection controlled the melting shape, melting rate, and heat transfer during melting. The melting rate sharply decreased as the aspect ratio of the cavity increased. It was later reported [14] that laminar flow persisted throughout the melting process when confined to a small rectangular test cell (Length: 6.35 cm, Width: 3.81 cm, Height: 8.89 cm). Zhang and Bejan [15] studied the melting process of n-octadecane in the presence of natural convection in a larger enclosure (Length: 56 cm, Width: 14.6 cm, Height: 73.7 cm) heated at a constant rate from the side. A turbulent flow, instead of laminar flow, was found near the heated plate as a result of a higher Rayleigh number. Because different conclusions were reported for the phase transitions in the presence of natural convection, a better understanding of the substances used as PCM, as well as the physical phenomena that take place during phase transition, are needed. This paper presents an experimental approach used to quantify the heat transfer enhancement via the development of a heat transfer enhancement factor and an effective heat transfer coefficient. The PCM used in this work was a commercially-available paraffin-based PCM with a phase transition temperature range between 26 and 28 °C. Differential Scanning Calorimeter (DSC) tests were conducted to obtain phase transition temperatures and latent heats of fusion of this PCM. A chamber was designed and fabricated to create environments at several ambient air temperatures in which a PCM-filled insulated metal container
was placed where the PCM was forced to undergo energy storage and energy release processes. Pertinent temperatures (i.e., PCM, ambient air, PCM container and its insulated surfaces) were measured using type-T thermocouples. Heat fluxes under several configurations were also measured. 2. Differential Scanning Calorimeter (DSC) tests A DSC set up was used to measure the PCM’s melting and solidification temperatures and latent heat of fusion. A baseline calibration, which involved heating an empty pan throughout the entire temperature range was performed. Calibration results are shown in Fig. 1. Results of two calibration runs agreed well with each other. For the DSC testing, the sample was encapsulated in a sealed aluminum pan with a mass of 13.2 mg. The test was conducted in the temperature range of 0–50 °C with a temperature ramp of 5 °C/ min using, as per operator’s manual, a constant stream of nitrogen of 50 ml/min at atmospheric pressure. These tests results had an uncertainty of 1–2% [16]. Reproducibility of these tests was achieved [16]. The PCM was Octadecane paraffin. The details of this PCM are shown in Table 1 [17]. DSC test results for melting and solidification processes are shown in Figs. 2 and 3, respectively. The melting and solidification temperatures of the PCM were determined via extrapolation of the onset of temperature peaks. Latent heats of fusion were evaluated via integrating the heat flow curves [6]. Fig. 2 shows the PCM melting process curve. Two peaks were found during this process. The first one appeared at the temperature of 5.7 °C and its corresponding heat of fusion was 22.5 kJ/kg. The second one appeared at the temperature of 32.4 °C and its corresponding heat of fusion was 129.8 kJ/kg. The first peak corresponded to a solid–solid phase transition [18]. The total estimated heat of fusion was 152.3 kJ/kg when the PCM was heated from 0 to 50 °C. This value was close to the one provided by the manufacturer; however, the functional heat of fusion of this material was 129.8 kJ/kg, which corresponded to an extrapolated melting temperature of 27.5 °C, which was 0.5 °C lower than the melting temperature provided by the manufacturer. Fig. 3 shows the PCM solidification process curve. In this case only one peak occurred, which indicated that there was no solid– solid phase transition during this process. During solidification, solid–solid phase transition did not occur. One explanation set forth is the presence of supercooling, which shifts the phase transitions to lower temperatures. Solidification of the sampled PCM started when its temperature was 26.6 °C. The heat flow reached its ‘‘peak’’ at a temperature of 22.6 °C. The estimated heat of solidification was 116.0 kJ/kg. Compared with the results in Fig. 2, the
0.036
Calibration 1 Calibration 2
0.034
Specific heat (J/ oC )
2
0.032 0.030 0.028 0.026 0.024 0.022 0
10
20
30
40
50
o
Temperature ( C) Fig. 1. Baseline calibrations for DSC tests.
Please cite this article in press as: Sun X et al. Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs). Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.03.078
3
X. Sun et al. / Applied Energy xxx (2015) xxx–xxx Table 1 Manufacturer property data of the PCM. Type
Approximate melting point (°C)
Approximate solidification point (°C)
Latent heat of fusion (kJ/kg)
Density (kg/m3) Solid
Liquid
Paraffin
28
26
147
870
750
Conductivity (W/(m °C)) 0.2
30
Heat Flow (mW)
Size: 13.2 mg Method: Ramp 25
o
32.4 C
20 15
o
5.7 C 10
o
3.9 C 22.5 J/g
5 0
o
27.5 C 129.8 J/g 0
10
20
30
40
50
o
Fig. 4. Graphical schematic of the experimental setup.
Temperature ( C) Fig. 2. DSC melting process curve for the selected PCM.
0 o
Heat Flow (mW)
-5
26.6 C 116.0J/g
-10 -15 -20 -25 -30
Size: 13.2 mg Method: Ramp
o
22.6 C -35
0
10
20
30
40
50
Temperature ( oC) Fig. 3. DSC solidification process curve for the selected PCM.
solidification process occurred at a temperature that was 0.9 °C lower than the melting temperature. The supercooling for a pure substance could be eliminated if the heating and cooling processes were carried out at extremely slow, but unrealistic heating and cooling rates [19]. 3. Experimental setup and calibration tests An environment simulating system was designed and fabricated to artificially create several surrounding (ambient) air temperatures to drive the phase transition processes (Fig. 4). The system included an insulated enclosed chamber, a portable air conditioner, a data acquisition unit and other ancillary components. A photograph of the setup is shown in Fig. 5. The insulated chamber was 1.2 m 1.2 m 1.2 m. Its enclosure consisted of a 12.7 mm gypsum wallboard, five 1.27 cm polystyrene foam insulating layers, and a 20.5 mm oriented strand board (OSB), an engineered wood particle board [20]. Inside the chamber, eight 200-W incandescent light bulbs were installed and used as heat sources to heat
up the air inside the chamber. These light bulbs were connected to rheostats and digital timers that were programmed to simulate a forced heat load on the outside of the PCM holding container. Two 80 mm 80 mm fans were placed inside the chamber for uniform air mixing and temperature distribution within the chamber. A 2.5 kW portable air conditioner was connected to the chamber through a circular duct. The cold air came into the chamber across its top enclosing surface. This experimental set up was located in an air-conditioned laboratory where the air temperature was kept at approximately 22–24 °C all year round. The PCM was enclosed in a cubic aluminum container with dimensions of 20 cm on each side. To restrict the heat flow to and from the aluminum container, it was insulated with foam with a thermal resistance of 8.63 m2 °C/W. To measure the temperature distribution within the PCM, 20 Type-T thermocouples (T/C) were installed inside the container using a structural frame. Each external surface temperature of the PCM-holding container as well as each insulation outer surface temperature was measured by four and two thermocouples, respectively. Six heat flux meters (HFMs) were attached to the insulation outer surfaces to monitor the heat fluxes across each surface. The thermocouples were arranged in seven layers, from top to bottom. Their locations are shown in Fig. 6. In addition, to monitor the air temperatures inside the chamber, five thermocouples were installed at various locations. All thermocouples were shielded with aluminum tape to eliminate the effects of radiation on the temperature measurements. The thermocouples and HFMs descriptions and their reading errors are given in Table 2. Before the experiments, the PCM-holding container was filled with 6.06 kg of PCM. Two calibration tests were conducted. Each container surface was insulated for Test 1. The insulation of the bottom surface was removed for Test 2. Fig. 7 shows the PCM-holding container outer surface temperature and heat flux with time during the calibration tests. The PCM-holding container was suspended at the center of the chamber as shown in Fig. 5(a). The temperatures of each container surface as well as the surrounding air during Test 1 are shown in Fig. 7(a). The surface temperatures were kept between 28.3 and 29.3 °C for 19 h when the temperature of the air surrounding the container (herein referred to as ‘‘ambient’’ temperature) was maintained at approximately 37 °C. The surface
Please cite this article in press as: Sun X et al. Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs). Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.03.078
4
X. Sun et al. / Applied Energy xxx (2015) xxx–xxx
Cold air entrance
Insulated chamber Cold air inlet duct
Heatsources
Computer A/C
PCMcontainer Heatfluxmeters Thermocouples
(a) Interior view
(b) Exterior view
Fig. 5. Experimental setup.
PCM Container
Side 3
Insulation Layer 1
T/Cs
Layer 2
5 cm
Layer 3 5 cm Layer 4
Side 4
Side 2
5 cm Layer 5 5 cm
HFMs
Layer 6 Layer 7
5 cm 5 cm Side 1
20 cm
(a) PCM container-Front view
(c) Layer 1 and 7-Top view
Frame 5 cm 5 cm
5 cm 5 cm 5 cm 5 cm
5 cm 5 cm (c) Layers 2, 4 and 6-Top view
(d) Layers 3 and 5-Top view
Fig. 6. Thermocouple and HFM locations within the PCM-holding container.
Table 2 T/C and HFMs description and their accuracies. Sensor
Type
Size (T W L) (mm)
Range
Error
T/Cs HFMs
T type Flat solid plate
1.4 2.3 0.6 38 38
100 °C 6.3 104 W/m2
0.6 °C 2%
temperatures did not increase past 30 °C regardless of the ambient air temperature. This suggested that the thermal resistance of the insulation was sufficient to significantly reduce any heat transfer to and from the PCM from and to its surroundings. However, these stable surface temperatures might also be a result caused by the PCM melting process, which occurred at the temperature range of 26 and 28 °C. Therefore, one more calibration test (Test 2) was carried out. During this test, the insulation attached to the bottom surface of the PCM-holding container (Layer 7) was removed to allow a heat transfer path between the PCM and the ambient air. Fig. 7(b) shows the variation in heat fluxes with time through each container surface during the PCM melting processes. Heat fluxes through the bottom surface of the PCM-holding container (Layer 6) were approximately 30 W/m2 when the ambient air
temperature was 37 °C while the heat fluxes through all the other surfaces were negative. These negative values were explained by the fact that the insulation around the PCM-holding container was lined with aluminum foil with a reflectance of about 0.95. This forced a significant amount of heat transfer to be reflected at the insulated surfaces [21]. Consequently, the insulation outer surface temperatures were lower than the ambient air temperature, resulting in negative heat fluxes. Based on the above calibration tests, it was concluded that the insulation of the PCM-holding container significantly reduced the heat transfer to and from the PCM effectively. 4. Energy release and storage results To understand the phase transition characteristics of the PCM, twelve experiments were performed at different ambient air temperatures. It was pre-established that when all PCM temperatures had surpassed the PCM’s melting point, the melting process was complete. After the melting process was completed, the solidification period was initiated directly by reducing the ambient air temperature to a temperature below the solidification point.
Please cite this article in press as: Sun X et al. Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs). Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.03.078
5
X. Sun et al. / Applied Energy xxx (2015) xxx–xxx Table 3 Experimental conditions for energy storage and energy release processes.
38 Ambient
Name Mode
29.5
Layer 2 Layer 6
o
Temperature ( C)
36
34
32
29.0
Side 2 Side 4
S-1
28.5
Side 3 Side 1
S-2
Energy storage Energy storage Energy storage Energy release Energy release Energy release
S-3
28.0 4
30
5
6
7
8
R-1 R-2
28 0
4
8
12
16
20
R-3
Time (h)
Heat source
Cold source
Ambient air temperature (°C) Vertical heat transfer
Horizontal heat transfer
Low power Medium power High power -
–
35.5
44.1
–
36.7
49.2
–
46.0
64.8
A/C
13.1
9.3
A/C
16.5
11.2
A/C
20.1
13.2
Low power Medium power
(a) PCM-holding container outer surfaces temperatures in Test 1 50
Ambient air
40
PCM
Layer 6
Layer 6
Side 3
-2 20
-3
Side 4 Layer 1
-4 10
-5 -6
0
Temperature ( o C)
Heat flux (W/m 2 )
-1
4
5
Side 1 Side 2 6 7
R-3
R-2
R-1
30
40
30
20
8
S-1 10
-10 0
4
8
12
16
20
0
92
S-3
S-2 184
276
368
460
Time (h)
Time (h)
(b) Heat fluxes during Test 2
(a) Ambient air and PCM mean temperature variations
Fig. 7. Temperatures and heat fluxes variations during the calibration tests.
4.1. Energy storage and release in the vertical heat transfer direction
PCM mean temperature ( oC)
Temperature data were collected at an interval of 10 s but analyzed every 15 min. The PCM-holding container’s bottom surface and Side 1 were selected as heat transfer surfaces for vertical and horizontal heat transfer, respectively. The energy release process in the vertical heat transfer direction through the container’s bottom surface (Layer 6) during solidification was initiated by introducing liquid PCM at 40 °C. The energy storage process in the horizontal heat transfer direction through Side 1 during melting was initiated by introducing solid PCM at 24 °C. All of the experimental conditions are described in Table 3. The power input for the heat sources was adjusted into three levels by two rheostats, namely low, medium, and high.
40
35
30
o
29 C
25
o
24 C S-1 Tair =34.70 C S-2 Tair =36.74oC S-3 T =46.08oC air o
20
15 0
10
20
30
40
50
60
70
80
Time (h) (b) PCM mean temperature variation during melting Fig. 8. Temperature variations in the vertical heat transfer direction.
Fig. 8 shows the temperature variations of the PCM, the ambient air, and the PCM-holding container’s bottom surface (Layer 6) during the vertical heat transfer direction experiments for various boundary conditions. PCM temperature was the average of the reading of the thermocouples located inside the PCM. The energy storage and release processes were divided into three stages [12]. The primary stage consisted of sensible heat storage or heat release, where the PCM temperature was lower than its solidification temperature or higher than its melting temperature. The PCM was kept in a homogenous solid or liquid state during this stage, which ended when the PCM temperature reached its melting or solidification temperatures. The secondary stage was the latent
heat storage or heat release stage, where the PCM temperature was within the phase transition temperature range. This stage continued until all the PCM had either melted or solidified. The last stage was also a sensible heat storage or heat release stage where the PCM temperature was either at a higher temperature than its melting temperature or at a lower temperature than its solidification temperature. During this stage, the PCM was in a single phase as either liquid or solid. For the energy storage processes, during the primary stages, sensible heat was transferred from the ambient air to the solid
Please cite this article in press as: Sun X et al. Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs). Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.03.078
6
X. Sun et al. / Applied Energy xxx (2015) xxx–xxx
PCM. The temperature of Layer 6 increased with the increase of the ambient air temperature. Because of the large temperature difference between the ambient air and the PCM, it was observed that PCM temperatures increased rapidly until each primary stage was completed. The secondary stages consistently ended when the PCM temperature increased at a faster rate. The temperature of Layer 6 followed the ambient air temperature during the secondary stages. The amplitude of the temperature of Layer 6 was smaller than that of the ambient air temperature because of the inertia of the melting PCM inside the container. During the third stages, both the PCM temperature and Layer 6 surface temperatures significantly increased. PCM temperature variations during energy storage are shown in Fig. 8(b). The temperature range for achieving complete melting was 24–29 °C, which was larger than the temperature range that was provided by the manufacturer. One possible reason for this was the motion of the molten PCM. For the energy release process, during the primary stages, the surface temperature of Layer 6 coincided with the PCM temperature. This indicated that the thermal resistance of the container wall could be neglected. That is, the surface temperature of Layer 6 represented the PCM temperature closer to the surface of the container. During the secondary stages, the PCM temperature was maintained between 26 and 28 °C; while Layer 6 surface temperature decreased. This happened because the PCM released not only its latent heat but also sensible heat [22]. The PCM temperature coincided with its DSC-measured melting temperature of 27.5 °C and its solidification temperature of 26.6 °C. A small discrepancy between the DSC values and the experimental results existed because the amount of PCM used during the experiments (6.06 kg) was much larger than the amount used for the DSC tests (13.2 mg) [23]. The ambient air temperature was kept almost constant for each test except for R-2. The energy release process during the R-2 test was divided into two steps at the point where the ambient air temperature dropped from 17.6 to 15.5 °C. This change might have been caused by a voltage decrease, which reduced the heat generated by the light bulbs, resulting in the temperature decrease of the ambient air. During the tertiary stages, both PCM and Layer 6 temperatures decreased to the point indicative that sensible heat was being released. Fig. 9 shows the heat fluxes and the temperature differences between the ambient air and Layer 6 over time during the vertical heat transfer direction experiments. When the PCM absorbed energy from the ambient air (i.e., energy storage process), the heat flux was positive. Otherwise, the heat flux was negative. It was observed that the heat flux, q, varied linearly with this temperature difference, as depicted in Eq. (1).
20 15
75
R-3
R-2
R-1
Tair -TLayer6 ( oC)
100
Temperature difference Heat flux
10
50
5
25
0
0
-5
-25
-10
-50
-15
S-1
S-3 -75
S-2
-20 0
92
184
276
368
-100 460
Time (h) Fig. 9. Heat flux variations during the vertical heat transfer direction experiments.
q¼
T air T surf Rair
ð1Þ
where Tair was the ambient air temperature, °C, Tsurf was the heat transfer surface temperature, which was the temperature of Layer 6, °C, and Rair was the thermal resistance of the ambient air, (m2 °C/W). In the primary stages, during the energy storage process, because of the large temperature difference between the ambient air and Layer 6, the value of the heat flux was large. Sensible heat was stored during this stage. The heat flux increased with the increase of the temperature difference. During the energy release process at this stage, the heat flux decreased over time with the reduction of the temperature difference. During the secondary stages, the change of rate of the temperature difference decreased because of the reduction in the change of rate of the temperature of Layer 6. During the third stages, with the melting or solidifying processes completed, the energy storage and energy release patterns changed back to the patterns observed during the sensible heat storage and heat release periods. The heat flux decreased with the decreasing temperature difference between the ambient air and the Layer 6. 4.2. Energy storage and release in the horizontal heat transfer direction Fig. 10 shows the temperature variations of the PCM, the ambient air, and the PCM-holding container Side 1 during the horizontal heat transfer direction experiments for various boundary conditions. As stated above, heat transfer processes through Side 1 were also divided into three stages, namely sensible heat storage or heat release stage, latent heat storage or heat release stage, and sensible heat storage or heat release stage. During the first and third stages, the temperature variations of the PCM and the surface of Side 1 followed the same pattern as those of the temperatures of the PCM and the surface of Layer 6 during the experiments of heat transfer processes in the vertical direction, respectively. However, it was observed that the temperature range required for completing the melting process changed from 24–30 °C to 24–32.5 °C during the secondary stages. This is shown in Fig. 10(b). It was inferred that this difference in temperatures was caused by the natural convection of the molten PCM because as the PCM melting process progressed, liquid PCM moved to the top of the container while the solid PCM descended to the bottom of the container as a result of the liquid PCM having a lower density and viscosity than the solid PCM. Similarly, natural convection increased with ambient air temperature as did the temperature range for completing the melting process. The solidification temperature range was maintained at 26–28 °C. This was the case because natural convection did not play a role during the solidification of the PCM. Two inflection points in each test were found on the PCM temperature curve, which illustrated when the phase transition process started and ended. This was true also for the heat transfer processes in the vertical direction, in which the phase transition process involved both latent heat and sensible heat energy storage and release. Fig. 11 shows the heat fluxes, temperature difference between ambient air and Side 1 during the horizontal heat transfer direction experiments. Heat flux was positive for the melting processes and negative for the solidification processes, which varied with the temperature difference between the ambient air and the surface of Side 1. The temperature difference increased during the primary stages, was maintained almost constant during the secondary stages, and decreased during thereafter, all during the energy storage processes. The rate of temperature difference decreased during the secondary stages, which indicated that a solidification process was occurring. The heat flux decreased over time with temperature
Please cite this article in press as: Sun X et al. Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs). Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.03.078
7
Temperature ( oC)
X. Sun et al. / Applied Energy xxx (2015) xxx–xxx
70
Ambient air
60
R-1
PCM
5. Discussions
Side 1
When considering the above mentioned heat transfer scenarios, natural convection of the molten PCM caused by buoyancy effect should be considered. A mean PCM thermal resistance, RPCM, was proposed to analyze this convective effect. According to the previous analysis, the thermal resistance of the PCM-holding container could be ignored. Based on energy conservation, the mean PCM thermal resistance was calculated by Eq. (2).
R-3
R-2
50 40 30 20 10 0
S-1 0
S-3
S-2 86
T surf T PCM T air T surf ¼ RPCM Rair
172
258
344
430
Time (h) (a) Ambient air and PCM mean temperature variations
where TPCM was the mean PCM temperature, °C, and RPCM was the thermal resistance of the PCM, (m2 °C)/W. According to the concept of Biot number, an adjusted Biot number, Bi0 , was defined as the ratio of the internal thermal resistance of the PCM to its external thermal resistance. In equation form
40
PCM mean temperature ( oC)
0
Bi ¼ 35
25
ð3Þ
o
30 C
where Rext was the external thermal resistance. In the context of this paper, it represented the ambient air thermal resistance (m2 °C)/W. Rayleigh number, Ra, which is the ratio of buoyancy forces and thermal and momentum diffusivities, was used to study the effects of natural convection in the liquid PCM. In equation form
o
24 C
20 S-1 Tair =43.54 C S-2 Tair =49.71oC S-3 Tair =64.51oC o
15
0
10
20
Ra ¼ Gr Pr ¼
30
40
50
Time (h)
150
Temperature difference Heat flux
75
0
0
-25
-75 S-1
S-3
S-2
-50 0
86
172
Heat flux (W/m2 )
R-3
R-2
R-1
258
m
2
3
3
Pr ¼
gbðT air T PCM Þl
ð4Þ
ma
3
Gr ¼
Fig. 10. Temperature variations in the horizontal heat transfer direction.
25
gbðT air T PCM Þl
60
(b) PCM mean temperature variation during melting
50
8 > < > 1 PCM thermal resistance dominated ¼ ¼ 1 PCM thermal resistance dominated > : < 1 External thermal resistance dominated
o
31 C
30
10
T air -T Face1 ( o C)
RPCM Rext
o
32.5 C
ð2Þ
344
-150 430
Time (h) Fig. 11. Heat flux variations during the horizontal heat transfer direction experiments.
difference during the energy release processes. The different variations of heat flux during energy storage and energy release processes were caused by the temperature difference variations, as depicted in Eq. (1). In this case, Tsurf represented the surface temperature of Side 1. The difference between heat flux and temperature difference curves during the energy storage processes was caused by the decrease of the PCM’s thermal resistance, which was induced by natural convection of the molten PCM.
gbðT air T PCM Þl
ð5Þ
m
2
where Gr was Grashof number, which represented the ratio of the buoyancy forces to the viscous forces acting on the fluid, Pr was Prantl number, which describes the relationship between momentum diffusivity and thermal diffusivity, g was gravitational acceleration, m/s2; b was the coefficient of volume expansion, 1/°C; l was the characteristic length of the PCM container, m; m was the PCM kinematic viscosity, m2/s and a was the PCM thermal diffusivity, m2/s. All the PCM parameters used in this equation were from Ref. [24]. These are shown in Table 4. Fig. 12 shows the adjusted Biot number, Bi0 , as a function of Rayleigh number, Ra. The temperature differences in Bi0 and Ra were the mean values when PCM was undergoing phase transitions. For all tests, the Bi0 number was less than 1, which indicated that the external thermal resistance was dominant during the heat transfer process. Any design of heat exchanger used for the heat transfer between PCM and air should emphasize the reduction of the air thermal resistance [25]. For energy release processes with solidifying PCM, Bi0 decreased with increasing Ra number. The observed decrease in Bi0 may be caused by an increase in the thermal conductivity of the PCM as a result of the decreasing PCM temperature [26] and a relative weak natural convection [14]. The thermal conductivity of nonmetallic solid material (i.e., paraffin) was mainly the result of the lattice vibrational wave effect [27]. With the increase of Table 4 PCM parameters used in Eqs. (4) and (5). g (m/s2)
b (1/°C)
l (m)
9.8
0.0008
0.2
v (m2/s)
a (m2/s)
25 °C
30 °C
25 °C
30 °C
5.7 106
5.0 106
2.1 107
9.0 108
Please cite this article in press as: Sun X et al. Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs). Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.03.078
8
X. Sun et al. / Applied Energy xxx (2015) xxx–xxx
0.40
90
Vertical heat storage Vertical heat release Horizontal heat storage Horizontal heat release
0.35
Vertical heat storage Vertical heat release Horizontal heat storage Horizontal heat release
80 70
0.30
Time (h)
Bi '
60
f = 1.33
f = 1.12
A
50
melting temperature of 26~28 C
40
B
30
0.25
C
20 10
0.20 6
12
18
24
36 X10
30
D
0
5
5
Ra
temperature, the collision frequency of phonons increased, which blocked the heat flow. For energy storage processes, Bi0 was reduced when the buoyancy forces increased. As a result of the increase of the temperature difference between the ambient air and the PCM, the increased buoyancy forces contributed to the natural convection. In the horizontal direction, thermal resistances were smaller during the heat storage process. This indicated that natural convection was more influential in promoting the heat transfer in this direction. The heat transfer enhancement, which represented the increase in heat transfer through a liquid PCM layer as a result of convection relative to conduction across the same liquid PCM layer, caused by natural convection, was evaluated by means of a heat transfer enhancement factor, f. This factor was defined as the ratio of the heat transfer coefficient with natural convection effect to that without natural convection under same conditions. In equation form
f ¼ /nc =/ ¼ R=Rnc
ð6Þ 2
where / was the heat transfer coefficient, W/(m °C), /nc was the effective heat transfer coefficient considering natural convection, W/(m2 °C), and R was the thermal resistance, (m2 °C)/W. The mean values of the enhancement factor in the vertical and horizontal heat transfer directions were 1.12 and 1.30, respectively, which indicated that the vertical and horizontal heat transfer was enhanced by 12% and 30% because of the natural convection, respectively. Fig. 13 shows the heat fluxes, calculated using Eq. (7), during phase transitions.
vertical heat storage vertical heat release
2
Absolute heat flux O W/m P
120 100 80 60 40 20
horizontal heat storage horizontal heat release
0 6
12
18
15
20
25
30
35
40 X10
5
Ra
Fig. 12. Adjusted Biot number vs. Rayleigh number during energy release and energy storage processes.
140
10
24
30
36 X10
Fig. 14. Phase transition time vs. Rayleigh number during energy release and energy storage processes.
q¼
fk Ra ma 1 ðT surf T PCM Þ ¼ 0 l 1 þ Bi gbl3 Rair
/nc ¼
fk l
ð7Þ
ð8Þ
The PCM thermal conductivity was given by k. For energy storage with natural convection, the heat flux was simplified by using an effective heat transfer coefficient as long as the heat transfer enhancement factor was given. When f = 1, heat conduction dominated the heat transfer. The abscissa represented the Ra, which was calculated in the same manner as Ra in Fig. 12. The heat flux was the absolute value during its corresponding energy release or storage processes. In all cases the absolute heat flux increased with increasing Ra, as predicted by Eq. (7). For the heat transfer in the same direction and with the same Ra, the heat flux during the energy storage process with melting PCM was larger than during the energy release process. This was the case because of natural convection. Fig. 14 shows the phase transition times for melting and solidification for the vertical and horizontal heat transfer directions. When comparing the time required for absorbing the latent heat of fusion by the PCM during the melting process with the time required for releasing the latent heat during the solidification process in the vertical heat transfer direction with the same Ra, it was observed that the time required for complete melting was shorter than the time required for complete solidification. For example, for heat transfer in the vertical direction under the same conditions (A and B), the time required for absorbing latent heat was 34.1 h, while the time required for releasing the heat was 62.1 h. The reason for this acceleration in the time during the melting process was the result of natural convection. The same was true for horizontal heat transfer under the same conditions (C and D). Curve D represented the horizontal heat transfer during melting for the temperature range of 26–28 °C, which was the same temperature range for solidification. The solid line represented the same melting process, but for a wider melting range, which was the actual measured range. From the data at points C and D, the phase transition time was reduced by 43% from 22.8 h, which also indicated the presence of natural convection. 6. Conclusions
5
Ra Fig. 13. Absolute heat flux vs. Rayleigh number during energy release and energy storage processes.
This paper presented an experimental approach used to quantify heat transfer rate increase by natural convection of PCMs undergoing phase transitions. A differential scanning calorimeter (DSC) test was carried out to obtain the PCM thermophysical parameters,
Please cite this article in press as: Sun X et al. Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs). Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.03.078
X. Sun et al. / Applied Energy xxx (2015) xxx–xxx
such as melting temperature, solidification temperature, and latent heats of fusion. There was a discrepancy between the temperature range for completing phase transition between the DSC test results and the experimental values. The reason for this was related to the amount of PCM used in each test. The temperature range for completing the melting process was widened as a result of natural convection of the molten PCM before the melting processes ended, changing from 26–28 °C to 24–32.5 °C. As the melting process progressed, the liquid PCM moved to the top of the PCM-holding container while the solid PCM descended to the bottom of the container. This occurred because the liquid PCM had lower density and viscosity than the solid PCM. Mean heat transfer enhancement factors of 1.12 and 1.30 were observed and attributed to natural convection during phase transitions for vertical and horizontal heat transfer directions, respectively. When comparing the time required for absorbing the latent heat of fusion by the PCM during the melting process with the time required for releasing the latent heat during the solidification process in the vertical direction, natural convection reduced the time required for complete melting by approximately 45%. However, the melting process took longer than the solidification process under the same conditions for the heat transfer in the horizontal direction because of the widening of the temperature range required for complete melting. Acknowledgements This work was supported by the University of Kansas, China Scholarship Council, International Cooperation Project (2015DFA61170), National 863 Project (SS2012AA052507) and Hunan Province Projects 2013WK2001. References [1] Sun X, Zhang Q, Medina M, Liu Y, Liao S. A study on the use of phase change materials (PCMs) in combination with a natural cold source for space cooling in telecommunications base stations (TBSs) in China. Appl Energy 2014;117:95–103. [2] Miró L, Navarro M, Suresh P, Gil A, Fernández A, Cabeza L. Experimental characterization of a solid industrial by-product as material for high temperature sensible thermal energy storage (TES). Appl Energy 2014;113:1261–8. [3] Casano G, Piva S. Experimental and numerical investigation of the steady periodic solid–liquid phase-change heat transfer. Int J Heat Mass Transfer 2002;45(20):4181–90. [4] Shao H, Nagel T, Roßkopf C, Linder M, Wörner A, Kolditz O. Non-equilibrium thermo-chemical heat storage in porous media: Part 2-A 1D computational model for a calcium hydroxide reaction system. Energy 2013;60:271–82.
9
[5] Agyenim F, Hewitt N, Eames P, Smyth M. A review of materials, heat transfer and phase change problem formulation for latent heat thermal energy storage systems (LHTESS). Renew Sust Energy Rev 2010;14(2):615–28. [6] Baran G, Sari A. Phase change and heat transfer characteristics of a eutectic mixture of palmitic and stearic acids as PCM in a latent heat storage system. Energy Convers Manage 2003;44(20):3227–46. [7] Kaizawa A, Kamano H, Kawai A, Jozuka T, Senda T, Maruoka N, et al. Thermal and flow behaviors in heat transportation container using phase change material. Energy Convers Manage 2008;49(4):698–706. [8] Duan X, Naterer G. Heat transfer in phase change materials for thermal management of electric vehicle battery modules. Int J Heat Mass Transfer 2010;53(23–24):5176–82. [9] Longeon M, Soupart A, Fourmigué J, Bruch A, Marty P. Experimental and numerical study of annular PCM storage in the presence of natural convection. Appl Energy 2013;112:175–84. [10] Kozak Y, Rozenfeld T, Ziskind G. Close-contact melting in vertical annular enclosures with a non-isothermal base: theoretical modeling and application to thermal storage. Int J Heat Mass Transf 2014;72:114–27. [11] Pal D, Joshi YK. Melting in a side heated tall enclosure by a uniformly dissipating heat source. Int J Heat Mass Transf 2001;44:375–87. [12] Jany P, Bejan A. Scaling theory of melting with natural convection in an enclosure. Int J Heat Mass Transf 1988;31(6):1221–35. [13] Ho CJ, Viskanta R. Inward solid–liquid phase-change heat transfer in a rectangular cavity with conducting vertical walls. Int J Heat Mass Transf 1984;27:1055–65. [14] Gau C, Viskanta R. Melting and solidification of a pure metal on a vertical wall. Transactions of the ASME. J Heat Transfer 1986;108:174–81. [15] Zhang Z, Bejan A. Melting in an enclosure heated at constant rate. Int J Heat Mass Transf 1989;32(6):1063–76. [16] Gmelin E, Sarge StM. Calibration of differential scanning calorimeters. Pure Appl Chem 1995;67(11):1789–800. [17] Yuan F. A comprehensive study of phase change materials (PCMs) for building walls applications (Thesis). University of Kansas; 2009. [18] Powers J, Craig R, Peyton F. Calorimetric analysis of commercial and dental waxes. J Dent Res 1969;48(6):1165–70. [19] Callister W. Materials science and engineering: an introduction. 7th ed. Manhattan: John Wiley & Sons, Inc; 2007. [20] Jin X, Medina M, Zhang X. On the importance of the location of PCMs in building walls for enhanced thermal performance. Appl Energy 2013;106:72–8. [21] Sheridan W, Ragland R. Pad including heat sink and thermal insulation areas. United States Patent 5011743, issued April 30, 1991. [22] Kozak Y, Abramzon B, Ziskind G. Experimental and numerical investigation of a hybrid PCM-air heat sink. Appl Therm Eng 2013;59:142–52. [23] Zhang Y, Jiang Y, Jiang Y. A simple method, the T-history method, of determining the heat of fusion, specific heat and thermal conductivity of phase-change materials. Meas Sci Technol 1999;10:201–5. [24] Ho C. Solid–liquid phase change heat transfer in enclosures (Dissertation). Purdue University; 1982. [25] Tao Y, He Y, Qu Z. Numerical study on performance of molten salt phase change thermal energy storage system with enhanced tubes. Sol Energy 2012;86(5):1155–63. [26] Puchkovska GO, Danchuk VD, Kravchuk AP, Kukielski JI. About the nature of phase transition in pure n-paraffin crystals. J Mol Struct 2004;704:119–23. [27] Çengel Y, Ghajar A. Heat and mass transfer: fundamentals and Applications. 4th ed. New York: McGraw-Hill; 2009.
Please cite this article in press as: Sun X et al. Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs). Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.03.078
Contents lists available at ScienceDirect
Applied Energy journal homepage: www.elsevier.com/locate/apenergy
Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs) q Xiaoqin Sun a,b, Quan Zhang c, Mario A. Medina b,⇑, Kyoung Ok Lee b a b c
School of Energy and Power Engineering, Changsha University of Science and Technology, Changsha 410114, China Dept. of Civil, Environmental & Architectural Engineering, University of Kansas, Lawrence 66045, USA College of Civil Engineering, Hunan University, Changsha 410082, China
h i g h l i g h t s Heat transfer rate was enhanced by natural convection of PCMs. A mean heat transfer enhancement factor of 1.2 was observed during melting. The time for melting was reduced by approximately 45% because of natural convection.
a r t i c l e
i n f o
Article history: Received 29 September 2014 Received in revised form 14 February 2015 Accepted 16 March 2015 Available online xxxx Keywords: Natural convection Phase change materials (PCMs) Heat transfer Melting process
a b s t r a c t Natural convection is one of the major factors that affect phase transition processes of solid–liquid phase change materials (PCMs). To optimize PCM-based latent thermal energy storage systems (TESS), a better understanding of the heat transfer interactions during these transitions is needed. In this paper the heat transfer rate enhancement caused by natural convection of PCMs undergoing melting is quantified based on experimental observations. For this, a heat transfer enhancement factor and an effective heat transfer coefficient were developed. Differential scanning calorimetry (DSC) tests were run to measure latent heats of fusion and phase transition temperatures of the PCMs. It was found that the experimentallyobtained temperature ranges required for complete melting exceeded those produced by the DSC tests. The reason for this stemmed from the natural convection of the molten PCM. The effective heat transfer coefficients when natural convection was accounted for were greater than when only heat conduction was considered. The increases in effective heat transfer coefficient were 12% and 30% percent for vertical and horizontal heat transfer paths, respectively. The existence of natural convection reduced the time required for complete melting by approximately 45% for vertical heat transfer. However, the melting process time was longer than the solidification process under the same conditions for a horizontal heat transfer path. The reason for this was attributed to the widening of the temperature range required for melting. Ó 2015 Published by Elsevier Ltd.
1. Introduction Thermal energy storage systems (TESS) are used to regulate the time-of-use mismatch between energy supply and consumption
q This article is based on a short proceedings paper in Energy Procedia Volume 161 (2014). It has been substantially modified and extended, and has been subject to the normal peer review and revision process of the journal. This paper is included in the Special Issue of ICAE2014 edited by Prof. J Yan, Prof. DJ Lee, Prof. SK Chou, and Prof. U Desideri. ⇑ Corresponding author at: 1530 W 15th Street, 2150 Learned Hall, University of Kansas, Lawrence, KS 66045, USA. Tel.: +1 785 864 3604. E-mail address: [email protected] (M.A. Medina).
[1]. Among TESS, which include sensible energy storage [2], latent energy storage [3], and chemical energy storage [4], latent energy storage systems that use solid–liquid phase change materials (PCMs) are particularly attractive. The advantages of using solid– liquid PCMs are [5]: (1) smaller weight and volume of material are required for a given amount of energy storage, compared to conventional sensible heat energy storage systems; (2) heat energy is stored at a constant or nearly constant temperature, which corresponds to the phase transition temperature of the PCM. Heat energy is stored in phase changing substances when these melt and it is recovered when these solidify. Baran and Sari [6] studied a mixture of palmitic and stearic acids as PCM within a tube-and-shell heat exchanger for solar
http://dx.doi.org/10.1016/j.apenergy.2015.03.078 0306-2619/Ó 2015 Published by Elsevier Ltd.
Please cite this article in press as: Sun X et al. Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs). Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.03.078
X. Sun et al. / Applied Energy xxx (2015) xxx–xxx
space heating applications. They concluded that the melting and solidification of PCMs was highly influenced by natural convection in the case of melting and by conduction in the case of solidification. Kaizawa et al. [7] analyzed the thermal behavior, during the phase transition processes, of solid–liquid PCMs within a direct heat exchanger. A heat transfer fluid flowed within the PCM to improve its heat transfer rate by agitating the liquid PCM. Therefore, the heat convection, instead of heat conduction, dominated the heat transfer process. Duan and Naterer [8] investigated the feasibility and effectiveness of two different PCM designs used in electric vehicles. They found that the natural convection within the melted PCM enhanced heat transfer rate to and from the PCMs. Longeon et al. [9] experimentally and numerically studied the influence of the natural convective heat transfer mechanism during phase transition processes. They concluded that the natural convection enhanced the melting process, whereas the liquid PCM behind the solidification front remained almost isothermal during the solidification process. Based on this research, a top injection for melting and a bottom one for solidification were recommended for a coaxial thermal energy storage unit. In summary, there is a close relationship between natural convection in molten-state PCMs and the enhancement of heat transfer to and from the PCM during phase transitions, especially during the melting process. Kozak et al. [10] found that close-contact melting in a vertical double pipe concentric storage unit with circumferentially finned inner tube significantly enhanced the melting process. Initiating close-contact melting was recommended to obtain higher heat transfer rates and shorter melting times. Pal and Joshi [11] studied a side heated tall enclosure of aspect ratio 10. They developed correlations of heat transfer rate via Nusselt number, Nu, and melting fraction, via a parameter e, and pointed out that melting was notably affected by natural convection during the later stages of the process. Jany and Bejan [12] separated the melting process with natural convection in an enclosure heated from the side into four regimes. Correlations for heat transfer rate and melting front location during each regime were theoretically and experimentally developed to explain the development of natural convection during melting. Ho and Viskanta [13] studied the phase transitions of n-octadecane in a rectangular cavity with varying aspect ratios under isothermal wall temperature boundary conditions. It was found that natural convection controlled the melting shape, melting rate, and heat transfer during melting. The melting rate sharply decreased as the aspect ratio of the cavity increased. It was later reported [14] that laminar flow persisted throughout the melting process when confined to a small rectangular test cell (Length: 6.35 cm, Width: 3.81 cm, Height: 8.89 cm). Zhang and Bejan [15] studied the melting process of n-octadecane in the presence of natural convection in a larger enclosure (Length: 56 cm, Width: 14.6 cm, Height: 73.7 cm) heated at a constant rate from the side. A turbulent flow, instead of laminar flow, was found near the heated plate as a result of a higher Rayleigh number. Because different conclusions were reported for the phase transitions in the presence of natural convection, a better understanding of the substances used as PCM, as well as the physical phenomena that take place during phase transition, are needed. This paper presents an experimental approach used to quantify the heat transfer enhancement via the development of a heat transfer enhancement factor and an effective heat transfer coefficient. The PCM used in this work was a commercially-available paraffin-based PCM with a phase transition temperature range between 26 and 28 °C. Differential Scanning Calorimeter (DSC) tests were conducted to obtain phase transition temperatures and latent heats of fusion of this PCM. A chamber was designed and fabricated to create environments at several ambient air temperatures in which a PCM-filled insulated metal container
was placed where the PCM was forced to undergo energy storage and energy release processes. Pertinent temperatures (i.e., PCM, ambient air, PCM container and its insulated surfaces) were measured using type-T thermocouples. Heat fluxes under several configurations were also measured. 2. Differential Scanning Calorimeter (DSC) tests A DSC set up was used to measure the PCM’s melting and solidification temperatures and latent heat of fusion. A baseline calibration, which involved heating an empty pan throughout the entire temperature range was performed. Calibration results are shown in Fig. 1. Results of two calibration runs agreed well with each other. For the DSC testing, the sample was encapsulated in a sealed aluminum pan with a mass of 13.2 mg. The test was conducted in the temperature range of 0–50 °C with a temperature ramp of 5 °C/ min using, as per operator’s manual, a constant stream of nitrogen of 50 ml/min at atmospheric pressure. These tests results had an uncertainty of 1–2% [16]. Reproducibility of these tests was achieved [16]. The PCM was Octadecane paraffin. The details of this PCM are shown in Table 1 [17]. DSC test results for melting and solidification processes are shown in Figs. 2 and 3, respectively. The melting and solidification temperatures of the PCM were determined via extrapolation of the onset of temperature peaks. Latent heats of fusion were evaluated via integrating the heat flow curves [6]. Fig. 2 shows the PCM melting process curve. Two peaks were found during this process. The first one appeared at the temperature of 5.7 °C and its corresponding heat of fusion was 22.5 kJ/kg. The second one appeared at the temperature of 32.4 °C and its corresponding heat of fusion was 129.8 kJ/kg. The first peak corresponded to a solid–solid phase transition [18]. The total estimated heat of fusion was 152.3 kJ/kg when the PCM was heated from 0 to 50 °C. This value was close to the one provided by the manufacturer; however, the functional heat of fusion of this material was 129.8 kJ/kg, which corresponded to an extrapolated melting temperature of 27.5 °C, which was 0.5 °C lower than the melting temperature provided by the manufacturer. Fig. 3 shows the PCM solidification process curve. In this case only one peak occurred, which indicated that there was no solid– solid phase transition during this process. During solidification, solid–solid phase transition did not occur. One explanation set forth is the presence of supercooling, which shifts the phase transitions to lower temperatures. Solidification of the sampled PCM started when its temperature was 26.6 °C. The heat flow reached its ‘‘peak’’ at a temperature of 22.6 °C. The estimated heat of solidification was 116.0 kJ/kg. Compared with the results in Fig. 2, the
0.036
Calibration 1 Calibration 2
0.034
Specific heat (J/ oC )
2
0.032 0.030 0.028 0.026 0.024 0.022 0
10
20
30
40
50
o
Temperature ( C) Fig. 1. Baseline calibrations for DSC tests.
Please cite this article in press as: Sun X et al. Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs). Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.03.078
3
X. Sun et al. / Applied Energy xxx (2015) xxx–xxx Table 1 Manufacturer property data of the PCM. Type
Approximate melting point (°C)
Approximate solidification point (°C)
Latent heat of fusion (kJ/kg)
Density (kg/m3) Solid
Liquid
Paraffin
28
26
147
870
750
Conductivity (W/(m °C)) 0.2
30
Heat Flow (mW)
Size: 13.2 mg Method: Ramp 25
o
32.4 C
20 15
o
5.7 C 10
o
3.9 C 22.5 J/g
5 0
o
27.5 C 129.8 J/g 0
10
20
30
40
50
o
Fig. 4. Graphical schematic of the experimental setup.
Temperature ( C) Fig. 2. DSC melting process curve for the selected PCM.
0 o
Heat Flow (mW)
-5
26.6 C 116.0J/g
-10 -15 -20 -25 -30
Size: 13.2 mg Method: Ramp
o
22.6 C -35
0
10
20
30
40
50
Temperature ( oC) Fig. 3. DSC solidification process curve for the selected PCM.
solidification process occurred at a temperature that was 0.9 °C lower than the melting temperature. The supercooling for a pure substance could be eliminated if the heating and cooling processes were carried out at extremely slow, but unrealistic heating and cooling rates [19]. 3. Experimental setup and calibration tests An environment simulating system was designed and fabricated to artificially create several surrounding (ambient) air temperatures to drive the phase transition processes (Fig. 4). The system included an insulated enclosed chamber, a portable air conditioner, a data acquisition unit and other ancillary components. A photograph of the setup is shown in Fig. 5. The insulated chamber was 1.2 m 1.2 m 1.2 m. Its enclosure consisted of a 12.7 mm gypsum wallboard, five 1.27 cm polystyrene foam insulating layers, and a 20.5 mm oriented strand board (OSB), an engineered wood particle board [20]. Inside the chamber, eight 200-W incandescent light bulbs were installed and used as heat sources to heat
up the air inside the chamber. These light bulbs were connected to rheostats and digital timers that were programmed to simulate a forced heat load on the outside of the PCM holding container. Two 80 mm 80 mm fans were placed inside the chamber for uniform air mixing and temperature distribution within the chamber. A 2.5 kW portable air conditioner was connected to the chamber through a circular duct. The cold air came into the chamber across its top enclosing surface. This experimental set up was located in an air-conditioned laboratory where the air temperature was kept at approximately 22–24 °C all year round. The PCM was enclosed in a cubic aluminum container with dimensions of 20 cm on each side. To restrict the heat flow to and from the aluminum container, it was insulated with foam with a thermal resistance of 8.63 m2 °C/W. To measure the temperature distribution within the PCM, 20 Type-T thermocouples (T/C) were installed inside the container using a structural frame. Each external surface temperature of the PCM-holding container as well as each insulation outer surface temperature was measured by four and two thermocouples, respectively. Six heat flux meters (HFMs) were attached to the insulation outer surfaces to monitor the heat fluxes across each surface. The thermocouples were arranged in seven layers, from top to bottom. Their locations are shown in Fig. 6. In addition, to monitor the air temperatures inside the chamber, five thermocouples were installed at various locations. All thermocouples were shielded with aluminum tape to eliminate the effects of radiation on the temperature measurements. The thermocouples and HFMs descriptions and their reading errors are given in Table 2. Before the experiments, the PCM-holding container was filled with 6.06 kg of PCM. Two calibration tests were conducted. Each container surface was insulated for Test 1. The insulation of the bottom surface was removed for Test 2. Fig. 7 shows the PCM-holding container outer surface temperature and heat flux with time during the calibration tests. The PCM-holding container was suspended at the center of the chamber as shown in Fig. 5(a). The temperatures of each container surface as well as the surrounding air during Test 1 are shown in Fig. 7(a). The surface temperatures were kept between 28.3 and 29.3 °C for 19 h when the temperature of the air surrounding the container (herein referred to as ‘‘ambient’’ temperature) was maintained at approximately 37 °C. The surface
Please cite this article in press as: Sun X et al. Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs). Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.03.078
4
X. Sun et al. / Applied Energy xxx (2015) xxx–xxx
Cold air entrance
Insulated chamber Cold air inlet duct
Heatsources
Computer A/C
PCMcontainer Heatfluxmeters Thermocouples
(a) Interior view
(b) Exterior view
Fig. 5. Experimental setup.
PCM Container
Side 3
Insulation Layer 1
T/Cs
Layer 2
5 cm
Layer 3 5 cm Layer 4
Side 4
Side 2
5 cm Layer 5 5 cm
HFMs
Layer 6 Layer 7
5 cm 5 cm Side 1
20 cm
(a) PCM container-Front view
(c) Layer 1 and 7-Top view
Frame 5 cm 5 cm
5 cm 5 cm 5 cm 5 cm
5 cm 5 cm (c) Layers 2, 4 and 6-Top view
(d) Layers 3 and 5-Top view
Fig. 6. Thermocouple and HFM locations within the PCM-holding container.
Table 2 T/C and HFMs description and their accuracies. Sensor
Type
Size (T W L) (mm)
Range
Error
T/Cs HFMs
T type Flat solid plate
1.4 2.3 0.6 38 38
100 °C 6.3 104 W/m2
0.6 °C 2%
temperatures did not increase past 30 °C regardless of the ambient air temperature. This suggested that the thermal resistance of the insulation was sufficient to significantly reduce any heat transfer to and from the PCM from and to its surroundings. However, these stable surface temperatures might also be a result caused by the PCM melting process, which occurred at the temperature range of 26 and 28 °C. Therefore, one more calibration test (Test 2) was carried out. During this test, the insulation attached to the bottom surface of the PCM-holding container (Layer 7) was removed to allow a heat transfer path between the PCM and the ambient air. Fig. 7(b) shows the variation in heat fluxes with time through each container surface during the PCM melting processes. Heat fluxes through the bottom surface of the PCM-holding container (Layer 6) were approximately 30 W/m2 when the ambient air
temperature was 37 °C while the heat fluxes through all the other surfaces were negative. These negative values were explained by the fact that the insulation around the PCM-holding container was lined with aluminum foil with a reflectance of about 0.95. This forced a significant amount of heat transfer to be reflected at the insulated surfaces [21]. Consequently, the insulation outer surface temperatures were lower than the ambient air temperature, resulting in negative heat fluxes. Based on the above calibration tests, it was concluded that the insulation of the PCM-holding container significantly reduced the heat transfer to and from the PCM effectively. 4. Energy release and storage results To understand the phase transition characteristics of the PCM, twelve experiments were performed at different ambient air temperatures. It was pre-established that when all PCM temperatures had surpassed the PCM’s melting point, the melting process was complete. After the melting process was completed, the solidification period was initiated directly by reducing the ambient air temperature to a temperature below the solidification point.
Please cite this article in press as: Sun X et al. Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs). Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.03.078
5
X. Sun et al. / Applied Energy xxx (2015) xxx–xxx Table 3 Experimental conditions for energy storage and energy release processes.
38 Ambient
Name Mode
29.5
Layer 2 Layer 6
o
Temperature ( C)
36
34
32
29.0
Side 2 Side 4
S-1
28.5
Side 3 Side 1
S-2
Energy storage Energy storage Energy storage Energy release Energy release Energy release
S-3
28.0 4
30
5
6
7
8
R-1 R-2
28 0
4
8
12
16
20
R-3
Time (h)
Heat source
Cold source
Ambient air temperature (°C) Vertical heat transfer
Horizontal heat transfer
Low power Medium power High power -
–
35.5
44.1
–
36.7
49.2
–
46.0
64.8
A/C
13.1
9.3
A/C
16.5
11.2
A/C
20.1
13.2
Low power Medium power
(a) PCM-holding container outer surfaces temperatures in Test 1 50
Ambient air
40
PCM
Layer 6
Layer 6
Side 3
-2 20
-3
Side 4 Layer 1
-4 10
-5 -6
0
Temperature ( o C)
Heat flux (W/m 2 )
-1
4
5
Side 1 Side 2 6 7
R-3
R-2
R-1
30
40
30
20
8
S-1 10
-10 0
4
8
12
16
20
0
92
S-3
S-2 184
276
368
460
Time (h)
Time (h)
(b) Heat fluxes during Test 2
(a) Ambient air and PCM mean temperature variations
Fig. 7. Temperatures and heat fluxes variations during the calibration tests.
4.1. Energy storage and release in the vertical heat transfer direction
PCM mean temperature ( oC)
Temperature data were collected at an interval of 10 s but analyzed every 15 min. The PCM-holding container’s bottom surface and Side 1 were selected as heat transfer surfaces for vertical and horizontal heat transfer, respectively. The energy release process in the vertical heat transfer direction through the container’s bottom surface (Layer 6) during solidification was initiated by introducing liquid PCM at 40 °C. The energy storage process in the horizontal heat transfer direction through Side 1 during melting was initiated by introducing solid PCM at 24 °C. All of the experimental conditions are described in Table 3. The power input for the heat sources was adjusted into three levels by two rheostats, namely low, medium, and high.
40
35
30
o
29 C
25
o
24 C S-1 Tair =34.70 C S-2 Tair =36.74oC S-3 T =46.08oC air o
20
15 0
10
20
30
40
50
60
70
80
Time (h) (b) PCM mean temperature variation during melting Fig. 8. Temperature variations in the vertical heat transfer direction.
Fig. 8 shows the temperature variations of the PCM, the ambient air, and the PCM-holding container’s bottom surface (Layer 6) during the vertical heat transfer direction experiments for various boundary conditions. PCM temperature was the average of the reading of the thermocouples located inside the PCM. The energy storage and release processes were divided into three stages [12]. The primary stage consisted of sensible heat storage or heat release, where the PCM temperature was lower than its solidification temperature or higher than its melting temperature. The PCM was kept in a homogenous solid or liquid state during this stage, which ended when the PCM temperature reached its melting or solidification temperatures. The secondary stage was the latent
heat storage or heat release stage, where the PCM temperature was within the phase transition temperature range. This stage continued until all the PCM had either melted or solidified. The last stage was also a sensible heat storage or heat release stage where the PCM temperature was either at a higher temperature than its melting temperature or at a lower temperature than its solidification temperature. During this stage, the PCM was in a single phase as either liquid or solid. For the energy storage processes, during the primary stages, sensible heat was transferred from the ambient air to the solid
Please cite this article in press as: Sun X et al. Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs). Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.03.078
6
X. Sun et al. / Applied Energy xxx (2015) xxx–xxx
PCM. The temperature of Layer 6 increased with the increase of the ambient air temperature. Because of the large temperature difference between the ambient air and the PCM, it was observed that PCM temperatures increased rapidly until each primary stage was completed. The secondary stages consistently ended when the PCM temperature increased at a faster rate. The temperature of Layer 6 followed the ambient air temperature during the secondary stages. The amplitude of the temperature of Layer 6 was smaller than that of the ambient air temperature because of the inertia of the melting PCM inside the container. During the third stages, both the PCM temperature and Layer 6 surface temperatures significantly increased. PCM temperature variations during energy storage are shown in Fig. 8(b). The temperature range for achieving complete melting was 24–29 °C, which was larger than the temperature range that was provided by the manufacturer. One possible reason for this was the motion of the molten PCM. For the energy release process, during the primary stages, the surface temperature of Layer 6 coincided with the PCM temperature. This indicated that the thermal resistance of the container wall could be neglected. That is, the surface temperature of Layer 6 represented the PCM temperature closer to the surface of the container. During the secondary stages, the PCM temperature was maintained between 26 and 28 °C; while Layer 6 surface temperature decreased. This happened because the PCM released not only its latent heat but also sensible heat [22]. The PCM temperature coincided with its DSC-measured melting temperature of 27.5 °C and its solidification temperature of 26.6 °C. A small discrepancy between the DSC values and the experimental results existed because the amount of PCM used during the experiments (6.06 kg) was much larger than the amount used for the DSC tests (13.2 mg) [23]. The ambient air temperature was kept almost constant for each test except for R-2. The energy release process during the R-2 test was divided into two steps at the point where the ambient air temperature dropped from 17.6 to 15.5 °C. This change might have been caused by a voltage decrease, which reduced the heat generated by the light bulbs, resulting in the temperature decrease of the ambient air. During the tertiary stages, both PCM and Layer 6 temperatures decreased to the point indicative that sensible heat was being released. Fig. 9 shows the heat fluxes and the temperature differences between the ambient air and Layer 6 over time during the vertical heat transfer direction experiments. When the PCM absorbed energy from the ambient air (i.e., energy storage process), the heat flux was positive. Otherwise, the heat flux was negative. It was observed that the heat flux, q, varied linearly with this temperature difference, as depicted in Eq. (1).
20 15
75
R-3
R-2
R-1
Tair -TLayer6 ( oC)
100
Temperature difference Heat flux
10
50
5
25
0
0
-5
-25
-10
-50
-15
S-1
S-3 -75
S-2
-20 0
92
184
276
368
-100 460
Time (h) Fig. 9. Heat flux variations during the vertical heat transfer direction experiments.
q¼
T air T surf Rair
ð1Þ
where Tair was the ambient air temperature, °C, Tsurf was the heat transfer surface temperature, which was the temperature of Layer 6, °C, and Rair was the thermal resistance of the ambient air, (m2 °C/W). In the primary stages, during the energy storage process, because of the large temperature difference between the ambient air and Layer 6, the value of the heat flux was large. Sensible heat was stored during this stage. The heat flux increased with the increase of the temperature difference. During the energy release process at this stage, the heat flux decreased over time with the reduction of the temperature difference. During the secondary stages, the change of rate of the temperature difference decreased because of the reduction in the change of rate of the temperature of Layer 6. During the third stages, with the melting or solidifying processes completed, the energy storage and energy release patterns changed back to the patterns observed during the sensible heat storage and heat release periods. The heat flux decreased with the decreasing temperature difference between the ambient air and the Layer 6. 4.2. Energy storage and release in the horizontal heat transfer direction Fig. 10 shows the temperature variations of the PCM, the ambient air, and the PCM-holding container Side 1 during the horizontal heat transfer direction experiments for various boundary conditions. As stated above, heat transfer processes through Side 1 were also divided into three stages, namely sensible heat storage or heat release stage, latent heat storage or heat release stage, and sensible heat storage or heat release stage. During the first and third stages, the temperature variations of the PCM and the surface of Side 1 followed the same pattern as those of the temperatures of the PCM and the surface of Layer 6 during the experiments of heat transfer processes in the vertical direction, respectively. However, it was observed that the temperature range required for completing the melting process changed from 24–30 °C to 24–32.5 °C during the secondary stages. This is shown in Fig. 10(b). It was inferred that this difference in temperatures was caused by the natural convection of the molten PCM because as the PCM melting process progressed, liquid PCM moved to the top of the container while the solid PCM descended to the bottom of the container as a result of the liquid PCM having a lower density and viscosity than the solid PCM. Similarly, natural convection increased with ambient air temperature as did the temperature range for completing the melting process. The solidification temperature range was maintained at 26–28 °C. This was the case because natural convection did not play a role during the solidification of the PCM. Two inflection points in each test were found on the PCM temperature curve, which illustrated when the phase transition process started and ended. This was true also for the heat transfer processes in the vertical direction, in which the phase transition process involved both latent heat and sensible heat energy storage and release. Fig. 11 shows the heat fluxes, temperature difference between ambient air and Side 1 during the horizontal heat transfer direction experiments. Heat flux was positive for the melting processes and negative for the solidification processes, which varied with the temperature difference between the ambient air and the surface of Side 1. The temperature difference increased during the primary stages, was maintained almost constant during the secondary stages, and decreased during thereafter, all during the energy storage processes. The rate of temperature difference decreased during the secondary stages, which indicated that a solidification process was occurring. The heat flux decreased over time with temperature
Please cite this article in press as: Sun X et al. Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs). Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.03.078
7
Temperature ( oC)
X. Sun et al. / Applied Energy xxx (2015) xxx–xxx
70
Ambient air
60
R-1
PCM
5. Discussions
Side 1
When considering the above mentioned heat transfer scenarios, natural convection of the molten PCM caused by buoyancy effect should be considered. A mean PCM thermal resistance, RPCM, was proposed to analyze this convective effect. According to the previous analysis, the thermal resistance of the PCM-holding container could be ignored. Based on energy conservation, the mean PCM thermal resistance was calculated by Eq. (2).
R-3
R-2
50 40 30 20 10 0
S-1 0
S-3
S-2 86
T surf T PCM T air T surf ¼ RPCM Rair
172
258
344
430
Time (h) (a) Ambient air and PCM mean temperature variations
where TPCM was the mean PCM temperature, °C, and RPCM was the thermal resistance of the PCM, (m2 °C)/W. According to the concept of Biot number, an adjusted Biot number, Bi0 , was defined as the ratio of the internal thermal resistance of the PCM to its external thermal resistance. In equation form
40
PCM mean temperature ( oC)
0
Bi ¼ 35
25
ð3Þ
o
30 C
where Rext was the external thermal resistance. In the context of this paper, it represented the ambient air thermal resistance (m2 °C)/W. Rayleigh number, Ra, which is the ratio of buoyancy forces and thermal and momentum diffusivities, was used to study the effects of natural convection in the liquid PCM. In equation form
o
24 C
20 S-1 Tair =43.54 C S-2 Tair =49.71oC S-3 Tair =64.51oC o
15
0
10
20
Ra ¼ Gr Pr ¼
30
40
50
Time (h)
150
Temperature difference Heat flux
75
0
0
-25
-75 S-1
S-3
S-2
-50 0
86
172
Heat flux (W/m2 )
R-3
R-2
R-1
258
m
2
3
3
Pr ¼
gbðT air T PCM Þl
ð4Þ
ma
3
Gr ¼
Fig. 10. Temperature variations in the horizontal heat transfer direction.
25
gbðT air T PCM Þl
60
(b) PCM mean temperature variation during melting
50
8 > < > 1 PCM thermal resistance dominated ¼ ¼ 1 PCM thermal resistance dominated > : < 1 External thermal resistance dominated
o
31 C
30
10
T air -T Face1 ( o C)
RPCM Rext
o
32.5 C
ð2Þ
344
-150 430
Time (h) Fig. 11. Heat flux variations during the horizontal heat transfer direction experiments.
difference during the energy release processes. The different variations of heat flux during energy storage and energy release processes were caused by the temperature difference variations, as depicted in Eq. (1). In this case, Tsurf represented the surface temperature of Side 1. The difference between heat flux and temperature difference curves during the energy storage processes was caused by the decrease of the PCM’s thermal resistance, which was induced by natural convection of the molten PCM.
gbðT air T PCM Þl
ð5Þ
m
2
where Gr was Grashof number, which represented the ratio of the buoyancy forces to the viscous forces acting on the fluid, Pr was Prantl number, which describes the relationship between momentum diffusivity and thermal diffusivity, g was gravitational acceleration, m/s2; b was the coefficient of volume expansion, 1/°C; l was the characteristic length of the PCM container, m; m was the PCM kinematic viscosity, m2/s and a was the PCM thermal diffusivity, m2/s. All the PCM parameters used in this equation were from Ref. [24]. These are shown in Table 4. Fig. 12 shows the adjusted Biot number, Bi0 , as a function of Rayleigh number, Ra. The temperature differences in Bi0 and Ra were the mean values when PCM was undergoing phase transitions. For all tests, the Bi0 number was less than 1, which indicated that the external thermal resistance was dominant during the heat transfer process. Any design of heat exchanger used for the heat transfer between PCM and air should emphasize the reduction of the air thermal resistance [25]. For energy release processes with solidifying PCM, Bi0 decreased with increasing Ra number. The observed decrease in Bi0 may be caused by an increase in the thermal conductivity of the PCM as a result of the decreasing PCM temperature [26] and a relative weak natural convection [14]. The thermal conductivity of nonmetallic solid material (i.e., paraffin) was mainly the result of the lattice vibrational wave effect [27]. With the increase of Table 4 PCM parameters used in Eqs. (4) and (5). g (m/s2)
b (1/°C)
l (m)
9.8
0.0008
0.2
v (m2/s)
a (m2/s)
25 °C
30 °C
25 °C
30 °C
5.7 106
5.0 106
2.1 107
9.0 108
Please cite this article in press as: Sun X et al. Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs). Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.03.078
8
X. Sun et al. / Applied Energy xxx (2015) xxx–xxx
0.40
90
Vertical heat storage Vertical heat release Horizontal heat storage Horizontal heat release
0.35
Vertical heat storage Vertical heat release Horizontal heat storage Horizontal heat release
80 70
0.30
Time (h)
Bi '
60
f = 1.33
f = 1.12
A
50
melting temperature of 26~28 C
40
B
30
0.25
C
20 10
0.20 6
12
18
24
36 X10
30
D
0
5
5
Ra
temperature, the collision frequency of phonons increased, which blocked the heat flow. For energy storage processes, Bi0 was reduced when the buoyancy forces increased. As a result of the increase of the temperature difference between the ambient air and the PCM, the increased buoyancy forces contributed to the natural convection. In the horizontal direction, thermal resistances were smaller during the heat storage process. This indicated that natural convection was more influential in promoting the heat transfer in this direction. The heat transfer enhancement, which represented the increase in heat transfer through a liquid PCM layer as a result of convection relative to conduction across the same liquid PCM layer, caused by natural convection, was evaluated by means of a heat transfer enhancement factor, f. This factor was defined as the ratio of the heat transfer coefficient with natural convection effect to that without natural convection under same conditions. In equation form
f ¼ /nc =/ ¼ R=Rnc
ð6Þ 2
where / was the heat transfer coefficient, W/(m °C), /nc was the effective heat transfer coefficient considering natural convection, W/(m2 °C), and R was the thermal resistance, (m2 °C)/W. The mean values of the enhancement factor in the vertical and horizontal heat transfer directions were 1.12 and 1.30, respectively, which indicated that the vertical and horizontal heat transfer was enhanced by 12% and 30% because of the natural convection, respectively. Fig. 13 shows the heat fluxes, calculated using Eq. (7), during phase transitions.
vertical heat storage vertical heat release
2
Absolute heat flux O W/m P
120 100 80 60 40 20
horizontal heat storage horizontal heat release
0 6
12
18
15
20
25
30
35
40 X10
5
Ra
Fig. 12. Adjusted Biot number vs. Rayleigh number during energy release and energy storage processes.
140
10
24
30
36 X10
Fig. 14. Phase transition time vs. Rayleigh number during energy release and energy storage processes.
q¼
fk Ra ma 1 ðT surf T PCM Þ ¼ 0 l 1 þ Bi gbl3 Rair
/nc ¼
fk l
ð7Þ
ð8Þ
The PCM thermal conductivity was given by k. For energy storage with natural convection, the heat flux was simplified by using an effective heat transfer coefficient as long as the heat transfer enhancement factor was given. When f = 1, heat conduction dominated the heat transfer. The abscissa represented the Ra, which was calculated in the same manner as Ra in Fig. 12. The heat flux was the absolute value during its corresponding energy release or storage processes. In all cases the absolute heat flux increased with increasing Ra, as predicted by Eq. (7). For the heat transfer in the same direction and with the same Ra, the heat flux during the energy storage process with melting PCM was larger than during the energy release process. This was the case because of natural convection. Fig. 14 shows the phase transition times for melting and solidification for the vertical and horizontal heat transfer directions. When comparing the time required for absorbing the latent heat of fusion by the PCM during the melting process with the time required for releasing the latent heat during the solidification process in the vertical heat transfer direction with the same Ra, it was observed that the time required for complete melting was shorter than the time required for complete solidification. For example, for heat transfer in the vertical direction under the same conditions (A and B), the time required for absorbing latent heat was 34.1 h, while the time required for releasing the heat was 62.1 h. The reason for this acceleration in the time during the melting process was the result of natural convection. The same was true for horizontal heat transfer under the same conditions (C and D). Curve D represented the horizontal heat transfer during melting for the temperature range of 26–28 °C, which was the same temperature range for solidification. The solid line represented the same melting process, but for a wider melting range, which was the actual measured range. From the data at points C and D, the phase transition time was reduced by 43% from 22.8 h, which also indicated the presence of natural convection. 6. Conclusions
5
Ra Fig. 13. Absolute heat flux vs. Rayleigh number during energy release and energy storage processes.
This paper presented an experimental approach used to quantify heat transfer rate increase by natural convection of PCMs undergoing phase transitions. A differential scanning calorimeter (DSC) test was carried out to obtain the PCM thermophysical parameters,
Please cite this article in press as: Sun X et al. Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs). Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.03.078
X. Sun et al. / Applied Energy xxx (2015) xxx–xxx
such as melting temperature, solidification temperature, and latent heats of fusion. There was a discrepancy between the temperature range for completing phase transition between the DSC test results and the experimental values. The reason for this was related to the amount of PCM used in each test. The temperature range for completing the melting process was widened as a result of natural convection of the molten PCM before the melting processes ended, changing from 26–28 °C to 24–32.5 °C. As the melting process progressed, the liquid PCM moved to the top of the PCM-holding container while the solid PCM descended to the bottom of the container. This occurred because the liquid PCM had lower density and viscosity than the solid PCM. Mean heat transfer enhancement factors of 1.12 and 1.30 were observed and attributed to natural convection during phase transitions for vertical and horizontal heat transfer directions, respectively. When comparing the time required for absorbing the latent heat of fusion by the PCM during the melting process with the time required for releasing the latent heat during the solidification process in the vertical direction, natural convection reduced the time required for complete melting by approximately 45%. However, the melting process took longer than the solidification process under the same conditions for the heat transfer in the horizontal direction because of the widening of the temperature range required for complete melting. Acknowledgements This work was supported by the University of Kansas, China Scholarship Council, International Cooperation Project (2015DFA61170), National 863 Project (SS2012AA052507) and Hunan Province Projects 2013WK2001. References [1] Sun X, Zhang Q, Medina M, Liu Y, Liao S. A study on the use of phase change materials (PCMs) in combination with a natural cold source for space cooling in telecommunications base stations (TBSs) in China. Appl Energy 2014;117:95–103. [2] Miró L, Navarro M, Suresh P, Gil A, Fernández A, Cabeza L. Experimental characterization of a solid industrial by-product as material for high temperature sensible thermal energy storage (TES). Appl Energy 2014;113:1261–8. [3] Casano G, Piva S. Experimental and numerical investigation of the steady periodic solid–liquid phase-change heat transfer. Int J Heat Mass Transfer 2002;45(20):4181–90. [4] Shao H, Nagel T, Roßkopf C, Linder M, Wörner A, Kolditz O. Non-equilibrium thermo-chemical heat storage in porous media: Part 2-A 1D computational model for a calcium hydroxide reaction system. Energy 2013;60:271–82.
9
[5] Agyenim F, Hewitt N, Eames P, Smyth M. A review of materials, heat transfer and phase change problem formulation for latent heat thermal energy storage systems (LHTESS). Renew Sust Energy Rev 2010;14(2):615–28. [6] Baran G, Sari A. Phase change and heat transfer characteristics of a eutectic mixture of palmitic and stearic acids as PCM in a latent heat storage system. Energy Convers Manage 2003;44(20):3227–46. [7] Kaizawa A, Kamano H, Kawai A, Jozuka T, Senda T, Maruoka N, et al. Thermal and flow behaviors in heat transportation container using phase change material. Energy Convers Manage 2008;49(4):698–706. [8] Duan X, Naterer G. Heat transfer in phase change materials for thermal management of electric vehicle battery modules. Int J Heat Mass Transfer 2010;53(23–24):5176–82. [9] Longeon M, Soupart A, Fourmigué J, Bruch A, Marty P. Experimental and numerical study of annular PCM storage in the presence of natural convection. Appl Energy 2013;112:175–84. [10] Kozak Y, Rozenfeld T, Ziskind G. Close-contact melting in vertical annular enclosures with a non-isothermal base: theoretical modeling and application to thermal storage. Int J Heat Mass Transf 2014;72:114–27. [11] Pal D, Joshi YK. Melting in a side heated tall enclosure by a uniformly dissipating heat source. Int J Heat Mass Transf 2001;44:375–87. [12] Jany P, Bejan A. Scaling theory of melting with natural convection in an enclosure. Int J Heat Mass Transf 1988;31(6):1221–35. [13] Ho CJ, Viskanta R. Inward solid–liquid phase-change heat transfer in a rectangular cavity with conducting vertical walls. Int J Heat Mass Transf 1984;27:1055–65. [14] Gau C, Viskanta R. Melting and solidification of a pure metal on a vertical wall. Transactions of the ASME. J Heat Transfer 1986;108:174–81. [15] Zhang Z, Bejan A. Melting in an enclosure heated at constant rate. Int J Heat Mass Transf 1989;32(6):1063–76. [16] Gmelin E, Sarge StM. Calibration of differential scanning calorimeters. Pure Appl Chem 1995;67(11):1789–800. [17] Yuan F. A comprehensive study of phase change materials (PCMs) for building walls applications (Thesis). University of Kansas; 2009. [18] Powers J, Craig R, Peyton F. Calorimetric analysis of commercial and dental waxes. J Dent Res 1969;48(6):1165–70. [19] Callister W. Materials science and engineering: an introduction. 7th ed. Manhattan: John Wiley & Sons, Inc; 2007. [20] Jin X, Medina M, Zhang X. On the importance of the location of PCMs in building walls for enhanced thermal performance. Appl Energy 2013;106:72–8. [21] Sheridan W, Ragland R. Pad including heat sink and thermal insulation areas. United States Patent 5011743, issued April 30, 1991. [22] Kozak Y, Abramzon B, Ziskind G. Experimental and numerical investigation of a hybrid PCM-air heat sink. Appl Therm Eng 2013;59:142–52. [23] Zhang Y, Jiang Y, Jiang Y. A simple method, the T-history method, of determining the heat of fusion, specific heat and thermal conductivity of phase-change materials. Meas Sci Technol 1999;10:201–5. [24] Ho C. Solid–liquid phase change heat transfer in enclosures (Dissertation). Purdue University; 1982. [25] Tao Y, He Y, Qu Z. Numerical study on performance of molten salt phase change thermal energy storage system with enhanced tubes. Sol Energy 2012;86(5):1155–63. [26] Puchkovska GO, Danchuk VD, Kravchuk AP, Kukielski JI. About the nature of phase transition in pure n-paraffin crystals. J Mol Struct 2004;704:119–23. [27] Çengel Y, Ghajar A. Heat and mass transfer: fundamentals and Applications. 4th ed. New York: McGraw-Hill; 2009.
Please cite this article in press as: Sun X et al. Experimental observations on the heat transfer enhancement caused by natural convection during melting of solid–liquid phase change materials (PCMs). Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.03.078