EXPERIMENTAL EVALUATION OF THE HEAT FLUXES THROUGH THE WALLS OF A DOMESTIC REFRIGERATOR

Transcription

1 Proceedings of ENCIT 214 Copyright 214 by ABCM November 1-13, 214, Belém, PA, Brazil EXPERIMENTAL EVALUATION OF THE HEAT FLUXES THROUGH THE WALLS OF A DOMESTIC REFRIGERATOR Susan Thiessen, susan@polo.ufsc.br Fernando T. Knabben, fernandok@polo.ufsc.br Cláudio Melo, melo@polo.ufsc.br POLO Research Laboratories for Emerging Technologies in Cooling and Thermophysics Federal University of Santa Catarina, Department of Mechanical Engineering , Florianópolis, SC, Brazil Abstract. In this study, the heat transfer paths through the walls of a top-mount refrigerator were experimentally identified. The refrigerator was instrumented with 55 tangential heat flux meters, strategically distributed over its entire outer surface. Each sensor was carefully calibrated using testing apparatus designed specifically for this purpose. The thermal load of the product was firstly quantified through a reverse heat leakage test. It was found that the heat transfer rates as measured by the heat flux sensors differed by only 8.8% from the power released by the electric heaters. It was also found that the heat transfer rate was highest on the side walls of the fresh-food compartment while the heat flux was highest on the top surface and also at the part of the rear wall attached to the mullion. The thermal load was also measured with the compressor running in the on-off mode. It was verified that a drop in the thermal load is expected when the level of insulation provided by the back edges of the refrigerator and also by the rear wall of the mullion in contact with the supply air duct are improved. The methodology adopted proved to be a convenient tool for gaining a better understanding of the way in which heat is transferred to the interior of the refrigerated compartments. Keywords: insulation, heat flux meter, domestic refrigerator, thermal load 1. INTRODUCTION In order to stimulate efficiency improvements, most governments periodically establish new goals for energy consumption and also new energetic labels for domestic refrigerators. As a consequence, manufacturers are constantly being encouraged to improve the performance of their products, which is strongly dependent on the thermal insulation of the cabinet walls (DOE, 21). After the Montreal Protocol (1987), CFC-11, a typical blowing agent at that time, started to be gradually replaced by HCFC-141b. However, although it has a minor effect on the ozone layer, this substance has a high global warming potential. For this reason it was phased-out in 21 and 214 in Canada and the USA, respectively. The use of HFC-245fa could represent an alternative, but this substance was phased-out in 212 in the European countries and will be phase-out in 216 in the USA. Currently, the dominant blowing agent is cyclopentane, which has approximately a 1% higher thermal conductivity when compared with its predecessors (Melo and Silva, 21). Due to the impossibility of reducing significantly the thermal conductivity of the polyurethane foam and also of increasing the insulation thickness, the vacuum insulation panels (VIPs) become a promising alternative for energy savings (Vineyard et al., 1998 and Tao et al., 24). These panels are composed of a silica or fiberglass core, evacuated at pressures lower than 1 Pa and enclosed in a metalized multi-layered laminate film (Fricke et al., 26). The thermal conductivity of these panels is between 3 and 5 mw/mk, in contrast with the typical average value of 2 mw/(mk) for polyurethane foams. The use of vacuum panels has always been limited by their high cost. The new energy consumption policies, however, have encouraged some manufacturers to start using this technology on a regular basis. However, important questions still need to be answered: what is the optimum coverage area and where should the panels be placed? Experimental results reported in the literature indicate that the traditional method of quantifying the insulation level of the refrigerated compartments, the so-called reverse heat leakage (RHL) test (Gonçalves et al., 2 and Sim and Ha, 211), may not be capable of properly capturing the variation in the insulation level provided by distinct VIP coverage areas. Melo et al. (23), for example, evaluated the thermal insulation of two bottom-mount refrigerators: one with and another without VIPs. The panels covered 22% of the refrigerator area, 56% being of the freezer and only 3% of the fresh-food compartment. The overall thermal conductance (UA) of each sample was measured, and no appreciable difference between them was observed. This behavior was attributed to the redistribution of the heat fluxes to the regions of lower thermal resistance (i.e., regions without VIPs). A discretized evaluation of the heat flux distribution through the cabinet walls is therefore necessary. Computational tools such as computer fluid dynamics (CFD) are used on a regular basis by some manufactures, but their widespread use is limited by the requirement of advanced modeling skills and also by the high computational cost. A more practical and less time consuming alternative is the use of an experimental procedure based on heat flux meters (HFM) strategically distributed over the refrigerator walls.

2 Proceedings of ENCIT 214 Copyright 214 by ABCM November 1-13, 214, Belém, PA, Brazil Therefore, the focus of this study was to map the heat transfer paths through the walls of a top-mount refrigerator insulated with polyurethane only. To this end the refrigerator was carefully instrumented with 55 tangential-type heat flux meters and submitted to reverse and direct heat leakage tests. The results of this study will establish a baseline for future work focused on the same refrigerator model but with different VIP coverage areas. 2. EXPERIMENTAL WORK 2.1 Heat flux meters The tangential heat flux meters used herein were manufactured by the Porous Media and Thermophysical Properties Laboratory (LMPT) of the Federal University of Santa Catarina (UFSC). These 45x45mm sensors are characterized by their low thickness (around 45μm), high sensitivity (2μV/W/m 2 ) and low time constant. They are extremely sensitive to electrical noise from external sources (e.g. the compressor), and therefore were individually connected to the data acquisition system with shielded cables. In addition, the sensors were painted white, the same color as the refrigerator outer shell, in order to minimize any radiative effects. Each sensor was also properly calibrated using a purpose-built apparatus. The transducers were calibrated according to the auxiliary-transducer method proposed by Orlande et al. (211). The apparatus is comprised of an upper aluminum plate, a thermal insulating plate, a bottom aluminum plate (base), a resistor (heater) and a fan (see Fig. 1). The function of the metallic upper plate is to compress the insulating plate against the base plate where the transducer to be calibrated is positioned. The base plate temperature is kept uniform by the fan airflow. A voltage source induces an electric current across the resistor, and thus provides the desired heat flux. The resistor is of the skin heater type, made of constantan, and its electrical resistance does not vary significantly within the desired temperature range. The resistor is also the same size as the transducer, to avoid lateral heat losses, which, according to Güths and Nicolau (1998), can reach almost 3% of the total heat flux. Ammeter Insulation Voltage source q aux q HFM aux Heater HFM Isothermal plate Data acquisition Fan Figure 1. Heat flux meter calibration apparatus. The transducer sensitivity, S, is obtained from the relation between the imposed heat flux and the transducer response voltage (S = q/v). The heat flux, q, is given by the difference between the total heat flux dissipated by the resistor (q T = Ri 2 /A) and the heat flux lost through the insulating plate, which is measured by the auxiliary transducer (q aux = S auxv aux), for which the sensitivity should be known. To determine the auxiliary sensitivity an interactive procedure is required. This procedure is divided into three steps. First, using the apparatus described above, two HFMs (A and B) are placed in position (one above and one below the resistor, respectively). The data are registered and an assumption is made for HFM A (which plays the role of an auxiliary): for example, 3% of the total heat flux is lost through the insulating plate, and thus it is possible to estimate S A. The second step is to use this approximation to estimate the sensitivity of the other HFM. The third step is to change the positions between the HFMs, so that HFM B is now the auxiliary. With the new data and the estimated S b it is possible to determine the new S A. Lastly, S A is used to determine an updated S b value and this iterative process finishes when the difference between two consecutively calculated sensitivity values is lower than Refrigerator The experiments were carried out with a 429-liter top-mount refrigerator, running with 46g of the refrigerant HC- 6a. The airflow is supplied by an axial fan installed in the upper part of a tube-and-fin evaporator. A thermostatic damper is used to balance the airflow between the refrigerated compartments and thus control the fresh-food compartment temperature. The freezer compartment temperature, in turn, is controlled by a thermostat that switches the compressor on and off. The condenser is a natural draft wire-and-tube heat exchanger. The air temperatures at different positions inside the cabinet were monitored by 28 T-type thermocouples, with a maximum uncertainty of ±.2 C. Reverse refrigerator off under steady-state conditions and direct heat leakage (DHL) refrigerator on under periodic cyclic conditions tests were carried out. The refrigerator was positioned inside a climate-controlled chamber

3 Proceedings of ENCIT 214 Copyright 214 by ABCM November 1-13, 214, Belém, PA, Brazil with the air temperature and relative humidity kept at 18.1ºC and 5% during the RHL test, and at 25.1ºC and 5% during the DHL test. During the RHL test the freezer and fresh-food compartment temperatures were kept, respectively, at 61.1ºC and 47.9ºC by several PID-driven electrical heaters strategically placed within the compartments. During these tests the thermostat was deactivated while the damper was fixed at a predefined position. Also, the fan was kept on to homogenize the air temperatures. The heaters and the fan power were measured by a power transducer with an uncertainty of ±.1% of the full scale (Yokogawa WT-23). During the direct heat leakage tests, the heaters were removed and the thermostat reactivated. The average temperatures of the freezer and fresh-food compartments were -21.6ºC and 6.1ºC, respectively. Figure 2 illustrates the positioning of the heat flux meters over the refrigerator walls and also the adopted control surfaces. The heat flux distribution over the doors, top and rear walls were assumed to be symmetric. Top 3. RESULTS Doors Sides Rear Bottom 3.1 Reverse heat leakage test Figure 2. Positioning of heat flux meters. A reverse heat leakage test was carried out to check the measurements of the heat flux meters and also to map the regions of higher heat flux. The total power of the heaters (freezer + fresh-food compartment) was 65.W. The total heat transfer rate as measured by the heat flux meters was 59.3W, which represents a deviation of only 8.8%. Moreover, it should be noted that the rate of the heat transfer through the door gaskets was not measured. In order to simplify the data analysis, the refrigerator was discretized into micro-regions, as shown in Figure 3. It can be seen that the side walls (23.8%), rear wall (15.9%) and door (15.6%) of the fresh-food compartment are the regions where the heat transfer rate was highest. The heat flux, as expected, was more intense at the top wall, since this surface is submitted to a higher airflow. It can also be observed that the heat flux in the rear region of the mullion (horizontal wall which separates the freezer and the fresh-food compartments) is of the same order of magnitude as that observed for the top wall, since in this region the air streams from the freezer and fresh-food compartments merge. For the lateral wall of the fresh-food compartment (Fig. 4), the percentage contribution of the shelf regions (N13 and N14) is higher in terms of the heat transfer rate, although the heat flux is more intense in the fast cooling regions (N11 and N12 on the left side and N29 and N3 on the right side). This occurs because the areas corresponding to N13 and N14 are larger and the areas associated with N11, N12, N29 and N3 are submitted to a higher temperature difference. The rear and door regions of the fresh-food compartment have practically the same percentage contribution in terms of heat transfer rate (Fig. 3). In addition, the average heat fluxes in each of these micro-regions are very similar. In the rear region (Fig. 5) not only the highest heat transfer rate but also the highest heat flux were observed in the area corresponding to transducer N28, located on the surface adjacent to the supply air duct of the fresh-food compartment. In the case of the fresh-food compartment door (Fig. 6) it is clear that the heat exchange is more intense in the areas associated with transducers A3 and A4, related to the shelves. The heat flux, however, is higher in the region corresponding to transducer N2, located at the height of the fast cooling drawer, repeating the same behavior observed for the lateral wall of the fresh-food compartment.

4 Heat flux [W/m 2 ] Heat flux [W/m 2 ] Proceedings of ENCIT 214 Copyright 214 by ABCM November 1-13, 214, Belém, PA, Brazil 25 2 Thermal load distribution Freezer door Freshfood door Freezer side Mullion side Freshfood side Freezer rear Mullion rear Freshfood rear Bottom Rate [W] Flux [W/m²] Percentage Top Figure 3. Heat transfer rate, heat flux and percentage of heat transfer rate for each micro-region. 9% 8% 6% 1% 6% N11 N % 22% 28% N13 N14 N15 N16 to N2 N29 N3 Heat flux [W/m 2 ] N11 N12 N13 N14 N15 N16 N17 N18 N19 N2 N29 N3 Figure 4. Lateral wall of fresh-food compartment: heat transfer rate (left) and heat flux (right). 32% 22% A % 27% 7% 9% A9 A11 A12 M2 and M3 N A7 A9 A11 A12 M2 M3 N28 Figure 5. Rear wall of fresh-food compartment: heat transfer rate (left) and heat flux (right). 18% 6% 25% A % 19% 3% A4 A5 N1 N A3 A4 A5 N1 N2 M4 Figure 6. Door of fresh-food compartment: heat transfer rate (left) and heat flux (right).

5 Proceedings of ENCIT 214 Copyright 214 by ABCM November 1-13, 214, Belém, PA, Brazil In Figure 7 it can be observed that although the transducer A16 contributes more than half of the heat transfer through the freezer top, the corresponding heat flux is not as significant when compared to the heat flux measured by transducer M7. This difference is due to the fact that this transducer is located at the freezer door hinge, which is made of a metallic material and is in direct contact with the outer surface of the refrigerator. For this reason, this component acts as a fin, conducting heat to the environment. 4% 11% 1% N % 11% N34 N38 A16 M7 Heat flux [W/m 2 ] N33 N34 N38 A16 M7 3.2 Direct heat leakage test Figure 7. Top of freezer compartment: heat transfer rate (left) and heat flux (right). In order to evaluate the heat transfer paths in a more realistic way, a direct heat leakage (DHL) test was carried out. This test is called direct because the compressor is turned on and thus the heat flows from the outside to the inside of the cabinet. The purpose of this test is to identify the heat transfer paths that are induced by warm sources that only appear when the system is running, such as the compressor, the condenser and the anti-sweating loop. As shown in Figs. 8 and 9, it was observed that the heat fluxes through the doors of the freezer (A2, M6, N3, N39,) and fresh-food (A3, A4, A5, N1, N2, M4) compartments are practically constant, but the heat flux through the fresh-food door is slightly lower due to the lower temperature difference. 5 A2 M6 N3 N39 5 A3 A4 A5 M4 N1 N Figure 8. Door of freezer compartment. Figure 9. Door of fresh-food compartment. On the other hand, a transient heat flux behavior sharp drop when the compressor shuts off and linear rise until the beginning of the next cycle was identified for the freezer (N4, N5, N6, N7, N8) and mullion (N9, N1) side walls, as shown in Figs. 1 and 11. This happens because the fan is switched off right after the compressor stops, which considerably reduces the convection heat transfer coefficient. The heat fluxes at the top part of the side wall of the fresh-food compartment (N11, N12, N29, N3) also present this behavior (Fig. 12). At the bottom of the side walls of the fresh-food compartment (Fig. 13) in the regions of the shelves (N13, N14) and vegetable drawer (N15, N16) the heat flux is slightly cyclic and even negative, implying that the wall temperature in this region is higher than the room temperature. This behavior is captured by the other transducers (N17, N18, N19, N2, N21) at the edge of the side wall of the fresh-food compartment (Fig. 14). The negative heat flux is not only caused by the anti-sweating loop but also by the condenser and the compressor, which are warm components. It is also observed that the heat flux is more intense at the edge of the side wall of the fresh-food compartment, a finding that was not observed during the RHL test. At the rear of the freezer compartment (A6, M1, N22, N23, N24, N37) and of the mullion (N25, N26, N27), Figs. 15 and 16, respectively, the heat flux is considerably higher, due to the greater temperature difference and higher convective heat transfer coefficient. The cycling condition due to the on/off compressor operation is also clearly observed in these regions. In the DHL test, the heat fluxes measured by sensors N25 and N26 - the first one in the

6 Proceedings of ENCIT 214 Copyright 214 by ABCM November 1-13, 214, Belém, PA, Brazil region of the incidence of the airflow from the fresh-food return duct and the second one in the region of the incidence of the airflow from the freezer return duct - are intense, a finding that was also not observed during the RHL test. 5 4 N4 N5 N6 N7 N8 5 4 N9 N Figure 1. Side wall of the freezer compartment. Figure 11. Side wall of the mullion compartment. 5 4 N11 N12 N29 N3 5 4 N13 N14 N15 N Figure 12. Top of side wall of the fresh-food compartment. Figure 13. Bottom of side wall of the fresh-food compartment. 5 N17 N18 N19 N2 N21 5 A6 M1 N22 N23 N24 N Figure 14. Edge of side wall of the fresh-food compartment Figure 15. Rear wall of the freezer compartment Figure 17 shows that the rear wall of the fresh-food compartment (A7, A9, A11, A12, M2, M3, N28) also has a cyclic behavior but of less intensity, except in the case of transducer M3, which registered a higher heat exchange than the other transducers while the compressor was running. This behavior can be attributed to the fact that this transducer is located at the junction between the vertical and horizontal surfaces of the compressor compartment. Also, transducers A7, N28 and A12 located between the wall of the fresh-food compartment and the condenser, exhibited similar heat fluxes, but these were slightly lower than those of the freezer rear wall transducers N24, N37, A6 (despite the thicker freezer walls, these transducers are subjected to a greater temperature difference). The heat flux at the top wall of the freezer compartment (A16, M7, N33, N34, N38) has the same intensity as that of the freezer door, with the exception of transducer M7, which presented a considerable variation in the heat flux, which is predominantly negative since the surface temperature is higher than the ambient temperature. This occurs because this transducer is positioned on the freezer door hinge, in direct contact with the metallic surface which heats up due to the presence of the anti-sweating loop.

7 Proceedings of ENCIT 214 Copyright 214 by ABCM November 1-13, 214, Belém, PA, Brazil Finally, the heat fluxes at the bottom wall of the fresh-food compartment are presented in Fig. 19. It was found that the heat flux is low under the vegetable drawer (A15, N36) and quite high above the compressor (A13, A14), as expected. 5 N25 N26 N27 5 A7 A9 A11 A12 M2 M3 N Figure 16. Rear wall of mullion. Figure 17. Rear wall of fresh-food compartment. 5 A16 M7 N33 N34 N38 5 A13 A14 A15 N Figure 18. Top wall of freezer compartment. Figure 19. Bottom wall of fresh-food compartment. 4. CONCLUSIONS In this work the heat transfer paths through the walls of a top-mount refrigerator were mapped. The refrigerator was instrumented with 28 thermocouples and with 55 tangential heat flux meters. The heat flux meters were all previously calibrated with a purpose-built apparatus and connected to the data acquisition system with shielded cables. Reverse and direct heat leakage tests were carried out during the experiments. It was found that the sum of the heat transfer measured by the heat flux sensors is approximately equal to the heat released by the electric heaters during the reverse heat leakage test. The highest heat transfer rate during the RHL test occurred at the side walls of the fresh-food compartment while the heat flux was more intense at the top wall and in the rear wall of the mullion. The DHL test showed that the thermal load of the fresh-food compartment can be decreased by improving the insulation of the back edges of the refrigerator. Furthermore, it was found that the heat flux at the rear wall close to the condenser is very high, particularly near the mullion, and thus, this is another potential area for energy savings. 5. REFERENCES DOE, Office of Energy Efficiency and Renewable Energy, Department of Energy. 21, Energy Conservation Program: Energy Conservation Standards for Residential Refrigerators, Refrigerator-Freezers, and Freezers, EE-28-BT- STD-12, Washington, USA. Fricke, J.; Schwab, H.; Heinemann, U., 26. Vacuum insulation panels exciting thermal properties and most challenging applications. International Journal of Thermophysics, Vol. 27, No. 4, pp Gonçalves J.M., Melo C., Vieira L.A.T., 2. Experimental study of frost-free refrigerators, Part I: Heat transfer through the cabinet walls. In Proceedings of the 1 st National Congress of Mechanical Engineering, Natal, RN, Brazil. (in Portuguese). Güths, S., Nicolau, V.P Instrumentation in thermal sciences. Internal Report, Federal University of Santa Catarina, Florianópolis, SC, Brazil (in Portuguese). Melo, C., Silva, L.W., 21. A perspective on energy savings in household refrigerators. In Proceedings of the Sustainable Refrigeration and Heat Pump Technology Conference, Stockholm, Sweden.

8 Proceedings of ENCIT 214 Copyright 214 by ABCM November 1-13, 214, Belém, PA, Brazil Melo, C., Vieira, Torquato, L.A,. 23. Analysis of the Thermal Insulation of Refrigerators Covered with VIPs. Internal Report, Federal University of Santa Catarina, Florianópolis, SC, Brazil (in Portuguese). Orlande, H.R.B., Fudym, O., Maillet, D. Cotta, R.M., 211. Thermal Measurements and Inverse Techniques. CRC Press, Boca Raton, USA. Sim J.S., Ha J.S., 211. Experimental study of heat transfer characteristics for a refrigerator by using reverse heat loss method, International Communications in Heat and Mass Transfer, Vol. 38, No. 5, pp Tao, W-H., Huang, C-M., Hsu, C-L., Lin, J., 24. Performance study of an energy-efficient display case refrigerator. Chemical Engineering Communications, Vol. 191, No. 4, pp Vineyard E.A., Therese K.S., Kenneth E.W., Kenneth W.C., Superinsulation in refrigerators and freezers. ASHRAE Transactions, Vol. 14, No. 2, pp RESPONSIBILITY NOTICE The authors are solely responsible for the printed material included in this paper.