Printer Friendly

Cooling characteristics of new refrigeration system without heat transfer fluid.


In recent years, conventional refrigerators have used CFCs, HFCs, and HCFCs as refrigerants. However, this has caused concern because of their high global warming potentials. In contrast, magnetic refrigeration is a remarkable technology that does not use these refrigerants. In contrast with conventional refrigeration, magnetic refrigeration is commonly used for extremely low temperatures. However, various research and development efforts have been carried out in recent years to develop magnetic refrigerators that can be operated near room temperature for general air conditioning purposes.

A magnetic refrigerator makes use of the magnetocaloric effect. This is a temperature change phenomenon that occurs when a magnetic field changes. It corresponds to the behavior of a gas when the pressure changes. In a vapor cycle system, the temperature is increased by adiabatic compression and decreases during adiabatic expansion. In a magnetic refrigeration cycle, magnetization corresponds to adiabatic compression and demagnetization corresponds to adiabatic expansion. Magnetic refrigeration uses this phenomenon to implement a refrigeration cycle through the flow of a heat transfer fluid into a magnetocaloric material, which is capable of producing a temperature change.

Room-temperature magnetic refrigeration has been researched to develop practical applications. In the 1990s, Pecharsky (1) discovered a magnetocaloric material that has a large magnetocaloric effect near room temperature. Furthermore, a magnetocaloric material in combination with a heat storage role called an active magnetic regenerator was developed in the United States and studies examined the application of a liquid hydrogen product (2,3,4,5). In 1998, Zimm, from the Astronautics company, developed a magnetic refrigerator using a superconducting magnet as the magnetic field source to verify the possibility of operating close to room temperature (6). This magnetic refrigerator uses Gd as a magnetocaloric material and water as a heat transfer fluid.

However, these magnetic refrigerators use a heat transfer fluid and system to circulate this transfer fluid. This increases the size of the machine. In order to overcome this potential problem, we have devised a new magnetic refrigeration model that eliminates the heat transfer fluid (7,8). In this model, the heat is transported from the cold side to the hot side through well-controlled thermal switches. In this research, we verified the thermal transport property of our devised model experimentally.


1.1 Magnetocaloric effect

A magnetocaloric material exhibits a strong magnetocaloric effect near the Curie temperature, when the phase changes from paramagnetic to ferromagnetic. The temperature difference of the magnetocaloric effect is a significant factor in determining the heat transfer performance.

A magnetic refrigerator uses the magnetocaloric effect that appears when a paramagnetic material is magnetized and demagnetized. The molecules and atoms in the magnetic material have magnetic moments, and increasing the external magnetic fields causes these magnetic moments to align. Therefore, the magnetic entropy decreases, and the temperature of the magnetic material rises. On the other hand, when the magnetic field is removed, the temperature of the magnetic material decreases because the magnetic moments become confused. A magnetic refrigerator uses this exothermic process and endothermic process as a refrigeration cycle.

1.2 Magnetic material

It is preferable for the magnetic material that is used in a magnetic refrigerator to have a Curie point close to room temperature, along with a large magnetocaloric effect. Therefore, we use Gd, which has a Curie point of 20[degrees]C and a large magnetocaloric effect. Figure 1 shows the experimental temperature change results when an external magnetic field is applied to Gd slab used in this research. The Gd slab is 30 mm (1.18 inch) in length, 20 mm (0.78 inch) in height, and 7.4 mm (0.30 inch) in width. The temperature is measured using a copper-constantan thermocouple. The change in the magnetic flux density in this experiment is between 0.75 T and 0.19 T, implying a magnetic field change of 0.56 T. The adiabatic temperature change is 1.2[degrees]C near 20[degrees]C, which is the Curie point of Gd.


1.3 Magnetic refrigeration cycle

Figure 2 shows the basic refrigeration principle of a magnetic refrigeration cycle with thermal switches. As an example, three magnetic slabs are aligned in Fig. 2, and permanent magnets are used to alternately magnetize and demagnetize the magnetocaloric material. The magnetocaloric slabs are composed of magnetocaloric materials and high thermal conductivity materials. The cycle for any given magnetocaloric slab (for instance, the center magnetocaloric slab in Fig. 2) consists of the following four processes: (1) magnetization, (2) heat transfer, (3) demagnetization, and (4) heat transfer. The ON, OFF mechanism of the thermal switch is implemented by the alternating contact with the contiguous magnetocaloric slabs. In addition, Fig. 3 shows the external magnetic field change in one cycle and the switching timing of a thermal switch, with the time on the horizontal axis, for the center magnetocaloric slabs of Fig. 2.

(1) Magnetization: In the magnetization process, the external magnetic field magnetizes the center magnetocaloric materials. The temperature of the magnetocaloric slabs rise, causing the magnetocaloric effect. At this time, the contiguous slabs are demagnetized and the temperature decreases.

(2) Heat transfer: The right side thermal switch turns ON, and the heat generated by magnetization is transferred to the right side magnetocaloric slabs that experienced a temperature decrease owing to the demagnetization. At this time, the left side thermal switch remains OFF.

(3) Demagnetization: In the demagnetization process, the external magnetic field is removed and a temperature decrease is caused by the magnetocaloric effect. At this time, the contiguous slabs are magnetized and the temperature rises.

(4) Heat transfer: The left side thermal switch turns ON, and heat transfers from the left side magnetocaloric slabs to the center slabs. In this way, heat is transferred from the low-temperature side to the high-temperature side from (from the left side to the right side in Fig.2).




2.1 Composition of magnetocaloric slab

Figure 4 shows the composition of the magnetocaloric slabs laminated with Gd sheets and oxygen-free copper sheets. The physical properties (9) of the Gd and oxygen-free copper sheets that are used in this study are listed in Table 1 and Table 2, respectively. To improve the heat conduction performance from the low temperature side to the high temperature side (horizontal direction in Fig. 4), 0.6-mm-thick (0.023 inch) Gd and 0.2-mm-thick (0.0078 inch) oxygen-free copper sheets are alternately laminated and integrated using a disffudion bonding method. At both ends, the oxygen-free copper block is integrated by soldering.

The magnetocaloric slab is 30 mm (1.18 inch) in length, 20 mm (0.78 inch) in height, and 7.4 mm (0.30 inch) in width. The four faces that do not contact contiguous components (magnetocaloric slabs, high temperature bath, and low temperature bath) are covered with 1-mm-thick (0.04 inch) MC nylon for insulation. Moreover, a copper-constantan thermocouple with a wire diameter of 0.2 mm (0.0078 inch) is embedded at the center of each magnetocaloric slab and each copper block. In addition, permanent magnets, with a 12-mm (0.47 inch) gap between the magnetic poles, are placed in the outer circumference of the magnetocaloric slab (top and bottom in Fig. 4). The effective thermal conductivity in the heat transfer direction can be given by Eq. (1), and the value is approximately 110 W/m K.

[[lambda].sub.MC] = ([A.sub.Cu][[lambda].sub.Cu] + [A.sub.Gd][[lambda].sub.Gd])/A (1)


2.2 Experimental Apparatus

Figure 5 shows a schematic of the experimental apparatus with the abovementioned magnetocaloric slabs. The system is composed of six magnetocaloric slabs, seven thermal switches, a cold-side bath, a hot-side bath, a ceramic heater, and permanent magnets. A time series of temperature measurements is possible using the thermocouples embedded in the center of each magnetocaloric slab and at both ends of the copper block. The input power value when the ceramic heater switches on is given by the measured values of current and voltage. The permanent magnets are aligned to correspond to every other magnetocaloric slab, reciprocating the horizontal direction. During operation, when a slab contacts the slab on the right side or left side, the thermal switch turns ON/OFF. When the slabs are in contact, the thermal switch is ON. At the other side, because the slabs are not in contact, the thermal switch is OFF. When the slabs are in contact, 28.13 kg/[cm.sup.2] (400.00 lb/[in.sup.2]) of pressure force is applied at the contact surface.


Figure 6 shows the operating mechanism of the experimental apparatus, including the movement of the slabs and permanent magnets, the switching contact surfaces, and repeated magnetization and demagnetization (phases 1 and 3). During the heat transfer period, heat is transferred between the slabs in contact with each other (phases 2 and 4). Repeating phase 1 to phase 4, the heat that is supplied owing to magnetization and demagnetization is transferred from the low-temperature heat bath to the high-temperature heat bath. The complete operation from phase 1 to phase 4 is regarded as one cycle.

Figure 7 shows the magnetic field distribution of the central section between the permanent magnets in Fig. 5. The average magnetic flux in the magnetocaloric slab at magnetization is 0.75 T, at demagnetization it is 0.19 T, and the change in the average magnetic flux is 0.56 T in this research.

In the experiment, the moving time for the magnets is kept constant at 1.1 s, whereas the cycle time change by the heat transfer time varies from 1 s to 30 s. After thermal equilibrium, the temperature difference between the cold-side bath and hot-side bath is defined to [DELTA]T, and the amount of heat transport from the cold-side to the hot-side bath when [DELTA]T = 0 is defined to Qin in this study.




3.1 Heat transfer mechanism

Figure 8 shows the temperature change of the slabs No. 1 and No. 2 in a cycle when the cycle time is 42.2 s and heat load is 0. The horizontal axis shows the time, and the vertical axis is the temperature of the slabs No. 1 solid line and No. 2 dashed line. In this graph, the temperature change of the magnetocaloric slabs is verified for the four processes: magnetization, heat transfer, demagnetization, and heat transfer. In the graph, the temperature of slab No. 1 rises because of magnetization, whereas that of No. 2 decreases because of demagnetization. Therefore, the temperatures of No. 1 and No. 2 become close during the heat transfer period (20 s). When the magnetization and demagnetization are switched, slab No. 1 is demagnetized, and the temperatures of slab No. 1 and the cold-side bath become close during the heat transfer period (20 s). As a result, the temperature difference between the cold-side bath and hot-side bath is increasing as time passes as shown in Fig. 9. We can see from this figure that [DELTA]T of 1.2[degrees]C is obtained when the heat load is 0.



3.2 Relationship between cycle time and heat load

The relationship between the cycle time and the heat load when the temperature difference between the hot side heat bath and the cold side heat bath becomes 0[degrees]C (Qin) obtained in this experiment is shown in Fig. 10. The experimental results are confirmed to have an error of less than 3%. This figure shows that Qin takes a maximum value near a cycle time of 15~25 s. The cause of the decline in Qin is inferred as an adequate heat transfer time which is not provided at an area shorter than the cycle time with the maximum value. On the other hand, when the cycle time becomes longer, the time duration of small temperature difference between adjacent test pieaces becomes longer (for instance, the time period indicated a circle in Fig.8). In this period, the amount of heat to be transported between the adjacent test pieaces does not increase inspite of the lapse of time, because the heat transport is drived by the temperature difference in the present apparatus. This fact results in the decrease in Qin at an area longer than the cycle time showing a maximum value.



In this research, we proposed a new magnetic refrigeration model that does not use a heat transfer fluid, but uses thermal switches to transfer heat from a cold-side bath to a hot-side bath. We fabricated a test machine to understand the operating characteristics experimentally.

(1) We verified that there is a maximum temperature difference between the two heat baths of about 1.2[degrees]C, and it is confirmed that heat can be transferred from the cold-side bath to the hot-side bath using a thermal switch instead of a heat transfer fluid.

(2) There is a cycle time to maximize the amount of heat transfer when AT = 0, which infers that an adequate heat transfer time is not supplied at a shorter area than the cycle time showing a maximum value, while on the other side decreasing in the transferred heat value per unit time at a longer area than the cycle time showing a maximum value.

A        = cross section

h        = heat transfer coefficient

T        = temperature

[lambda] = thermal conductivity

t        = time

H        = magnetic flux


ad       adiabatic

Cu       copper

Gd       gadolinium

MC       magnetocaloric slab


(1) V K. Pecharsky, K. A. Gschneidner: Giant Magnetocaloric Effect in Gd5 (Si2Ge2), Physical Review Letters, 78 (23), 1997.

(2) A. J. DeGregoria, P. J. Claybaker, J. R. Trueblood, R. A. Pax, T. M. Stankey, S. F. Kral, J. A. Barclay: Initial test result of an active magnetic refrigerative refrigerator, Adv. Cryo. Eng., 37, 1992.

(3) G. Geoffrey, P. George, S. John, H. James: Magnetocaloric Refrigeration, Research and development rept, 46, 1987.

(4) J. W Johnson and C. B. Zimm: Performance modeling of a 4K active magnetic regenerative refrigerator, J. Appl. Phys, 79 (5), 1996.

(5) C. B. Zimm, J. W. Johnson, R. W Murphy: Test result on a 50K magnetic refrigerator, Adv. Cryo. Eng, 41, 1996.

(6) C. Zimm, A. Jastrab, A. Sternberg, V Pecharsky, K JR. Gschneidner, M. Osborne, I. Anderson: Description and performance of a near-room temperature magnetic refrigerator, Adv. Cryo. Eng, 43, 1998.

(7) Y Tasaki, H. Takahashi, Y Yasuda, T. Okamura, K. Ito: A study on the fundamental heat transport potential of an in--vehicle magnetic refrigerator, Fifth IIF-IIR International Conference on Magnetic Refrigeration at Room Temperature, Thermag V Grenoble, France, 17-20 September 2012.

(8) U. L. Olsen, C. R. H. Bahl, KEngelbrecht, K K Nielsen, Y Tasaki, H. Takahashi, Y Yasuda: Modeling of In-vehicle Magnetic refrigeration, Proceedings of the fifth IIF-IIR International Conference on Magnetic Refrigeration at Room Temperature: Thermag V. 2012.

(9) G. K. White, M. L. Miges: Thermo physical properties of some Key Solids: International Journal of Thermophysics, 1269-1327, 1997.

Author Y. Tanaka and H. Makino are Master course students and T. Okamura is a professor in the Department of Energy Sciences, Tokyo Institute of Technology, Yokohama, Japan. Y. Tasaki, H. Takahashi, Y. Yasuda are research fellow at Nissan Research Cen ter,
Table 1 Physical properties of gadolinium

Quantity                             Value

Curie temperature                    293 K
Density             7900 kg/[m.sup.3] (0.28 lb/[in.sup.3])
Heat capacity                    230 J/kg x K
Thermal                          10.5 W/m x K

Table 2 Physical properties of oxygen-freecopper

Quantity                         Value

Density         8880 kg/[m.sup.3] (0.31 lb/[in.sup.3])
Heat capacity                386 J/kg x K
Thermal                       398 W/m x K
COPYRIGHT 2013 American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Tanaka, Yoshiki; Makino, Hiroto; Okamura, Tetsuji; Tasaki, Yutaka; Hidekazu, Takahashi; Ito, Kouji
Publication:ASHRAE Transactions
Article Type:Report
Geographic Code:1USA
Date:Jul 1, 2013
Previous Article:Experimental study of effects of solution subcooling and wall superheat on heat transfer of a horizontal-tube, falling-film heat exchanger.
Next Article:Energy implications of in-line filtration in California homes.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters