Document mpDvxZqkqEbzZezBpV2BeaOQQ

1100 CHAPTER 51 1954 Guide D = number of degrees on the thermometer scale which are not immersed. ti = temperature indicated on the thermometer, Fahrenheit. * = temperature of the non-immersed mercury column, Fahrenheit. 0.00009 = difference in the coefficient of expansion of the mercury and glass. When a thermometer is used in a liquid or in air or gas near room tem perature, the effects of radiation can often be ignored, but when the tem perature of hot air or gas is desired, means are usually provided for mini mizing the effect of radiation.7 These include bright metallic shields around the thermometer bulb, and the use of aspirated thermometers in which a stream of the air or gas is drawn at considerable speed across the bulb and increases the influence of the gas temperature on the thermom eter indication. In any case, to prevent errors in temperature measure ments, there should be ample circulation so that the thermometer will indicate a true temperature of the medium under observation, and ample time should be allowed for the thermometer to reach the same temperature as the medium. In reading a thermometer the eye should be at the same level as the top of the liquid to avoid parallax. Industrial-type thermometers are available for permanent installation in pipes or ducts. These instruments are fitted with metal guards to prevent breakage, and are useful for many purposes. However, the con siderable heat capacity and conductance of the guards or shields prevent such thermometers from following closely the fluctuations in. a medium of varying temperature. Thermocouples When two wires made of dissimilar metals are joined by soldering, weld ing or merely by twisting, a thermocouple or thermo-junction is formed and an electromotive force, which depends upon the temperature of the junc tion, is found to exist between the wires. When the wires are joined at two points a thermocouple circuit is formed, and if one junction is kept at a temperature different from the other, an electric current flows through the circuit due to the difference in emf developed by the two junctions. This phenomenon is employed for temperature measurements in thermocouple systems, one junction being ordinarily kept at a constant temperature, as in an ice bath, while the other junction is placed at a point at which it is desired to observe the temperature. In practice it is desirable to utilise emf to indicate temperature because, at small or zero current flow, the resistance of the circuit is unimportant. A high resistance millivolt meter is useful in some cases but the potentiometer yields better results. In the potentiometer the electromotive force generated by the thermocouples is balanced against an electromotive force from the battery so that observa tions are made with no flow of current through the thermocouple circuit. A conventional arrangement is illustrated in Fig. 1. The thermocouple leads A-B are so connected that their polarity opposes that of battery CIf the position of E on the graduated slide wire rheostat DF is adjusted until galvanometer G shows no current flowing, resistance DE will indicate directly the voltage generated by the thermocouple. In order to calibrate the instrument, switch S is thrown over to the standard eell circuit while rheostat R is adjusted so that the galvanometer shows zero current. Bat tery C then exerts the known voltage of the standard cell at DH. ` The choice of materials for thermocouple wires is determined by the range of temperature to be measured. Up to about 600 F base metals such as iron-constantan or, preferably from the corrosion standpoint, copp; constantan are satisfactory and develop a relatively large emf of 40 to 60 Instruments and Measurements 1101 microvolts per degree. Chromel-alumel couples are useful in the flue gas temperature range, while platinum (platinum-rhodium) couples are used for higher temperatures. Impurities make large differences in the perform ance of thermocouple wires and for this reason calibration of samples from each spool of wire is essential for precise work. Data on wire can usually be obtained from the manufacturer. The act of adjusting rheostat D-F (Fig. 1) for zero current flow is known as balancing the potentiometer. Automatic self-balancing instruments of both the indicating and recording types are on the market. They usu ally contain an automatically compensating cold junction to avoid the vise of an ice bath, and special thermocouple wire is furnished with them from the factory. With a suitable potentiometer, small wires serve as well for thermo couples as large ones, and the fineness of the wires is limited only by consid- Fig. 1. Basic Circuit and Connections por Thermocouple and Potentiometer Fig. 2. Typical Resistance Thermometer Circuit and Connections eration of mechanical strength and convenience in handling. Small couples respond more promptly to changes in temperature and are less affected by radiation than large ones. For use in heated air or gases, thermocouples are often shielded, as are thermometers, and aspirated thermocouples are sometimes used. An arrangement has been described for avoiding error due to radiation. It involves several thermocouples of different sizes, the true temperature being estimated by extrapolation of readings to zero diameter.8 By the use of thermocouples, temperatures at remote points may be indi cated or recorded on conveniently located instruments; average tempera ture may be readily obtained by connecting several couples in parallel or in series; and temperatures may be obtained within thin materials, narrow spaces, or otherwise inaccessible locations. Thermocouples in series with every alternate junction maintained at a common temperature will give an emf which, divided by the number of couples to give the average emf9 per couple, may be used to find the aver age temperature. Thermocouples in parallel, having the similar metals of a number of couples connected together and run to a common cold junction, will cause an indication on a potentiometer which is the true emf only if the electrical resistances of the parallel junctions are the same.9,10