CHAPTER XI. TESTING OF ICE REFRIGERATORS. (Part 1)

CHAPTER XI. TESTING OF ICE REFRIGERATORS. Constant Temperature Room. — A constant temperature room is necessary to accurately test the heat leakage of a refrigerator. Electrical thermostats and heaters are of con- siderable value for tests of this kind. It is easily possible to maintain a room temperature which will vary not more than 1° F. Fig. 209 shows an arrangement which has proven very satisfactory for a constant temperature rocjm. The electrical heaters are screened so that radiant heat ^vill not pass directh^ from the heaters to the outside surface of the refrigerator being tested. It is always difficult to measure radiant heat. AN'ith high room temperature, the heaters must be on a greater l)ercentage of the time, therefore the heat exchange by radia- tion would increase greatlj' unless the heaters are screened. An asbestos curtain is used for this purpose. There is a circulation of air as indicated by the arrows, insuring a uni- form temperatvu-e in dififerent parts of the constant tempera- ture room. Fig. 210 shows another arrangement for a constant tem- ])erature room using a double wall. The heaters are placed between the walls and the warm air circulates under the floor, over the ceiling and along the walls. This method reduces variation in radiation and convection, due to using the heaters in order to operate at a high room temperature. In order to obtain accurate results, it is best to use as little electrical heat as possible and yet keep the temperature of the test room constant. In this way the heat losses due 399 400 HOUSEHOLD REFRIGERATION to radiation and abnormal convections are reduced to a mini- mum. It is also desirable to control the humidity of the constant temperature room. This factor is not as important on a heat insulation test as with an ice melting test. Ice Melting Method. — -A simple method of measuring the heat leakage is by the ice melting method. The refrigerator -^ -^ -*• ELECTRIC sv THERMOSTAT, } / \ VITC H AND ^ f \ FUSES FOB. HEATERS 1 ^ 1 \ i=x > — -ELECTRIC \ HEATERS REFPUQERATOR ^ ASBESTOS \ — BAFFLE \ CURTAIN / ; V ■^f^-^ ^v y -»- -^ r \ 1 NSULATVON FTC;. 209.- CONSTANT TE.M I'KRATU k K TESTINC ROOMS. must be in use at least 24 hours in order to have the lining and insulation cooled to about the same temperature as they will be during the test. The author has found that a cabinet insulated with three inches of corkboard required three days to establish a temperature eciuilibrium in the walls. A weighed block of ice is then placed in the ice compart ment, noting the shape of the block so that on a subsequent test a similar shape can be used. Then after a certain period, say 24 to 48 hours, the ice block is weighed again to detect the amount of ice melted. Suppose the pounds of ice melted per 24 hours to be W . Then the heat leakage for the cabinet H TESTING OF ICE REFRIGERATORS 401 would be 144xn' in B.t.u. per 24 hours. The heat leakage h, per sq. ft. per degree F. per day would be : 144 X W . Sq. ft. mean area X (room temp. — average cabinet temp.) Temperatures should be taken of the coldest and warmest part of the food compartment. It is very important to have — ^- ^ — , - 1 i ^ — >. . A^ 1 THE«MOST>CT SWITCH AND .1 FOSES FOR HEATERS "' - ASBESTOS-^ LIMED — • ELECTRIC - ► HEATERS REMOVABLE 1 REFP-IGERATOC^ 1 1 ^J ooetK -^ f A INSULATION m. — FIG. 210.— CONSTANT TEMPERATURE TESTING ROOMS. a certain amount of dishes and food on the shelves in makin.u a test if the actual service conditions are desired. Service conditions can be closely duplicated by having plates of pota- toes in 10 or 20-pound units. Two or three times a day a certain number of cold units are removed from the cabinet and the same number of warm units (at room temp.) are used to replace them. Some of the more important variable factors entering into a heat leakage test by ice melting are as follows : 1. Constantly changing weight, surface and form of ice cake. 2. Circulation is afifected by size and position of ice cake. 3. The water from the melting ice may assist in cooling cabinet. 4(L' HOUSEHOLD REFRIGERATION The instruments required for a test are thermometers, preferably of the recording type. If regular glass stem mer- cury thermometers are used, it is advisable to place them in a small flask tilled with oil. The flask should have a cork with the thermometer held in place in a small hole through the cork. This eliminates the error of reading a rapidly rising temperature when the door of the cabinet is opened. Electrical Heater Method. — The electrical heater has ad- vantages o\er the ice-melting method of testing the heat leak- age of a refrigerator cabinet. The circulation is not changed by a different shape of the ice cake, and the rate of heat supply may be kept quite uniform and may be measured accurately even without opening the doors of the cabinet under test. Suppose an ice test indicates that a cabinet would be used with an average food compartment temperature of 45° F. in a 75° F. room. The temperature differential through the wall is therefore 30° F. To conduct a heat test, say in an 80° F. constant temperature room, a heating element is placed in the cabinet so that it will have the same wall differential temperature of 30° F. Therefore, the food compartment tem- perature is maintained at 80° F. plus 30° F. or 110° F. If the electrical heating element requires 20 watts in order to maintain this 30° F. temperature difference through the walls of the cabinet, then the total heat leakage in B.t.u. per 24 hours=20X 24X3.416= 1640. (1 watt hour of electrical energy is equi\alent to 3.416 B.t.u.) The heat leakage is usually rated by the number of B.t.u. lost per square foot per degree temperature difference per day. .Suppose the average surface of the inside and outside walls to be 20 sq. ft., then 1640 =r2.73 B.t.u. heat leakage per sq. ft. per degree F. per 24 hours. 30 X 20 Sources of Heat Losses in Refrigerators.— The following pertains to a test on an ice refrigerator to determine rate of ice melting due to heat loss of insulation, opening doors, and changing food. TESTING OF ICE REFRIGERATORS 4UJ The object of this test was to determine the relative amount of ice melted by the three principal heat losses which occur in an average household refrigerator. These are: 1. Heat transfer through the insulated walls. 2. Opening doors allowing warm air to enter, and cold air to drop (tut of the refrigerator. 3. Changing food or placing in the refrigerator, food and dishes 111 be cooled. The refrigerator was a top icer witli panel construction throughout and had the following specifications: Inches Inches Inches Height Depth Width Outside Cunipartmenl 60 21 29 Food Compartment 28 15^ 22J/2 Ice Compartment 16^ 16^ 20^ Food Compartment Door Opening 26 .. 20K' Ice Compartment Door Opening 14 • .. 20J^ 'olume Food Compartment 5.7 cubic feet 'olume Ice Compartment 3.2 cubic feet Total Inside Surface 28.9 square feet I'otal Outside Surface 40.0 square feet The in,>-.ulation consisted of the following: 1. Oak case. 2. J/2-inch mineral wool. 3. 2 air spaces. 4. Layers insulating ])aper. 5. 5^-inch spruce wall. 6. Porcelain on steel lining. The rated ice capacity of the refrigerator was 120 pounds and the net weight of the refrigerator was 280 pounds. The flue opening was 1^ inches wide on both sides of the ice com- ])artment and extended the total dei)th of the compartment. There was a 2-inch air space under the ice shelf. The test was conducted in a constant temperature room. The room had double walls on all six sides. The effect of heat transfer by radiation from the electric heaters used to maintain a constant temjjerattire was eliininated l)y placing the heaters between the double walls of the room. The oper- ation of these heaters was controlled by a thermostat. The humidity of the room was controlled by an electric humido- stat. It was found necessary to maintain constant condition of temperature and humidit}- for several days before an accurate 404 HOUSEHOLD REFRIGERATION reading could be obtained of the amount of ice melted for each particular test condition. Throughout the entire test, the room temperature was maintained at 75° F.. while the relative humidity was main- tained at 40 per cent. The quantity of ice, as well as the ice surface exposed. wa> kept as nearly uniform as possible throughout the test. The refrigerator was first oi)erate(l without any food changes or door opening process. The food compartment was empty so that the heat losses were due entirely to the heat transfer through the insulation of the cabinet. Of course, a very small percentage of this heat h^ss was caused by cooling and dehumidifying warm air which leaked into the cabinet, replacing cold drier air A\hich leaked out. and a small loss due to heat transfer by radiation which could not easily be measured. The food change tot was then conducted, the food which, in this case, was ])otatoes, l^eing changed three times each day. The potatoes were remo'ed at the temi)erature of the food comi>artment, while the potatoes ])laced in the box at each change were at room temperature. The food change consisted of remoxing a china plate weighing 2.4 pounds holding 8.6 pounds of potatoes, and then placing a similar cfuantity of plate and potatoes in the food compartment. Finally, a door opening test was conducted in conjunction with the food changing test. This approximated the average liousehold service condition, indicating the difference between a laboratory test and actual household service conditions. During this test, the relative humidity of the room was maintained at 40 per cent, while the relative humidity in the lower part of the food compartment of the refrigerator varied from 62 to 68 per cent. The results of the ice-making tests indicated that 93 per cent of the ice was melted, due to heat transmitted through the insulation, 4 per cent was required for cooling the food at the rate of 33 pounds per day, and that 3 per cent was lost in the opening of the doors which occupied one minute per liour, or a total of ten minutes during the test. These losses are shown graphically by Fig. 211. The foregoing data, to- TESTING OF ICE REFRIGERATORS 405 gether with Fig. 211, illustrate the great importance of having a refrigerator efificientl}' insulated. IT) O -J INSULATION COOLING FOOD OPENINQ DOORS FIG. 211.— COMP.\RISOX OF REFRIGERATOR HEAT LOSSES. Effect of Room Humidity. — The following test was on an ice refrigerator to determine the eftect of room humidity on the rate of ice melting. The refrigerator used in this experi- ment was the top icer described in the previous report. The test was conducted in a constant temperature room in which the humidity could be regulated and controlled very closely. The room temperature was maintained at 75° F. during the entire test which lasted 22 days. During this test the food storage space in the refrigerator contained only a recording thermometer and a recording humidostat. The quantity of ice as well as the amount of ice surface was maintained as nearly ccmstant as possible. The iiom lemperatuif Degrees F. 75 75 Food rompartment 65 65 Room 40 75 406 HOUSEHOLD REFRIGERATION following results were obtained with two different conditions of room humidity : Per cent Relative Humidity Ice melte<t per day, pounds 17.75 22.56 This test shows that the rate of ice melting was increased about 27 per cent, simply by changing the relative humidity of a 75 degree room from 40 to 75 per cent. This difference would be greater in actual service conditions as the doors are opened more frequently and sometimes not closed tightly, thus greatly increasing the amount of air leakage. It is therefore very important in refrigerator tests to know the rehitive Inimidit}- Ix^th of the air inside the refrigerator and of the room in which the refrigerator is located. Room or refrigerator environment air is constantly leak- ing into tlie upper part of a refrigerator, replacing the cold air leaking out of the lower part. This warm air circulates and is cooled to the food compartment temperature by coming in close contact with the ice or cooling element. Heat must be absorbed, either by melting ice or evaporating the liquid re- frigerant, to counteract the following heat losses : Heat losses due to air leakage or refrigerator ventilation. 1. To cool incoming dry air. 2. To cool moisture of incoming air. 3. To condense jiart of moisture of incoming air. 4. To freeze the condensed moisture. In a mechanical refriger- ator the surface temperature of the cooling element is usually below 32° F. 5. To cool the frozen moisture. It is then readily understood that with a nearly constant supply of warm room air entering the refrigerator, it will re- quire more heat to cool and dehumidify to the same dryness the 75 per cent humiditx air than it -wotdd to cf)ol and dehu- midify the 40 per cent humidity air. Humidity Diagram for Room and Refrigerator. — Fig. 212 shows the relation between room and food compartment hu- midities. This test was made in a constant temperature room held at 86° F. TESTING OF ICE REFRIGERATORS 407 The mechanical refrigerator maintained a temperature ot 42° F. in the lower and 50° F. at the top of the food compart- FIG. 212.— RELATINE HUMIDITY IN RKFKIGERATOR. ment. The refrigerator contained only the recording instru- ments. The average brine tank temperature was 20° F. The test was started with a warm refrigerator so that the temperature and humidity of both room and food compart- ment were equalized. 408 HOUSEHOLD REFRIGERATION Calorimeter Testing. — There is a great difference of opin- ion as to the proper method of determining" the exact or com- parative rating of household and small commercial compressor units. From tests compiled by various manufacturers of house- hold machines, the simplest and by far the most practical for every day usage, was found to be actual measurement by vol- ume of the refrigerant circulated, usually in a calibrated drum with sight glass, located directly under the condenser. By using the pressure on this drum at the beginning and end of the test, the mean pressure is obtained for determining the exact density of the liquid from authoritive tables on refrigerants. CoTid<inse.r CalihrateJ ■Drums ' FIC.. _'lo. After the actual pounds of refrigerant circulated has been found, the net available B.t.u. per pound of refrigerant can be determined from the table. This value multiplied by the pounds circulated gives the total B.t.u. of refrigerating work done in the evaporator. As numerous tests haAC demonstrated, superheating of the suction gas to the compressor has very little effect on com- pressor capacity under normal operating conditions, so that the volume of gas per pound of liquid can be obtained from the refrigerant table for the suction pressure noted near the compressor. This volume per pound multiplied by the pounds circulated will gi^'c the x^olume of gas handled by the compressor. TESTING OF ICE REFRIGERATORS 409 If the volume of gas handled by the compressor is divided by the piston displacement of the compressor for the same interval of time, the actual efficiency of the compressor is obtained. It has been found that the error between this de- termination method and the use of a calorimeter or B.t.u. measurement box around the evaporator is practically neg- ligible, and is much simpler. As a general rule, the determination which is most desired by all users of refrigeration is how much power must be paid for at the end of each month for holding the refrigerator at a temperature which best conserves food. The problem then becomes, how many kilowatts or their equivalent B.t.u.s must be paid for in power consumption for a definite number of B.t.u.s actually abstracted from the refrigerator? If a unit B.t.u. per hour abstraction is used as a basis, the machine requiring the minimum B.t.u. input for this work would be the most efficient overall, which, in the end, is the determining factor, provided the machines rate equal mechan- ically in construction. This value has been given the name "Performance Factor" and from the consumer's viewpoint is the most important, if other considerations such as appear- ance, size, arrangement, ice cubes, etc., are basically equal. A unit of comparison sometimes used is B.t.u. per Watt hour. From the manufacturer's standpoint, however, the prob- lem is somewhat different, as it is the individual parts and their efficiencies which affect him, so that each piece of equip- ment, compressor, evaporator, condenser or motor must be brought up to its maximum efficiency, which in turn will auto- matically take care of the overall efficiency or "Performance Factor." Looking at it from this angle the testing must include such factors as discharge pressure, compressor r.p.m. and size, size and type evaporator, size condenser and quantity of air to be blown over it, type drive, type motor and size, arrange- ment of parts, noise, vibration, lubrication, control, type suc- tion and discharge valves, kind of refrigerant, etc. Tests are subject to a great many variations which cause duplication of determined data to often be a difficult problem. In tests on a refrigerator load some of the variables that occur 410 HOUSEHOLD REFRIGERATION wcnild be: door leakaj^^e, air circulation, insulation efficiency, cooling chamber shape and size, location of coolinjj unit with its shape, size and arrang"ement, and the important factor of