You are on page 1of 30

Universitas Sriwijaya

Chapter 11
Temperature Measurement

11.1 Introduction
Temperature is a very widely measured and frequently controlled variable used in
numerous indrustial applications. In general, chemical reactions in the industrial processes
and products are temperature dependent and the desired quality of a product is possible
only if the temperature is accurately measured and maintained. Further, it forms an
important governing parameter in the thermodynamic and heat transfer operations like
steam raising,gas turbines in power generations and also in numerous propulsion system.
In addition, in the heat treatment of steel and alumunium alloys, temperature measurement
and control plays a crucial role in incorporating the desired material properties in the
finished heat-treated products. The other areas where measurement and control of
temperature is essential are : plastic manufacture, nuclear reactor components, milk and
dairy products, plant furnace and molten metals, heating and air-conditioning systems,
space shuttle components, blades of gas turbines, etc.
A number of definitions of temperature have been proposed. In a layman’s
language one could define this as the degree of hotness or coldness of a body or an
environment measured on a definite scale. Another simplified definition of temperature is
based on its equivalence to a driving force or potential that caused the flow of energy as
heat. Thus, we can define temperature as a condition of a body by virtue of which heat is
transferred to or from other bodies. Further, in the kinetic theory of gases and in statistical
thermodynamics, it is shown that temperature is related to the average kinetic energy of
molecules or atoms of which the material is made of. Finally, the definition of temperature
in thermodynamics sense is based on th ideal Carnot-cycle. According to this, it is defined
as quantity whose difference is proportional to the work obtained from a Carnot engine
operating between a hot source and a cold receiver.

Mechanical Engineering

Universitas Sriwijaya

It may be noted that there is a marked difference between the quantities temperature
and heat. Temperature may be defined as ‘degree’of heat whereas heat is taken to mean as
‘quantity’ of heat. For example, a bucket of warm water would melt more ice than a small
spoon of boiling water. The warm water in the bucket obviously contains greater quantity
of heat than that in the spoon containing boiling water. But its temperature is lower than
the boiling water, a fact that is readily apparent if a finger is dipped in both the vessels.
Temperature is a fundamental quantity, much the same way as mass,length and
time.The law that is used in temperature measurement is known as the zeroth law of
thermodynamics.This state that if two bodies are in thermal equilibrium with a third body,
then they are all in thermal equilibrium with each other. In the words, all the three bodies
would have the same temperature. Thus, if one can set up a reproducible means of
establishing a range of temperature and allowing thermal equilibrium to reach in each case.
In the other words, the thermometer is calibrated against a standard and is subsequently
used to read unknown temperatures.

11.2 Temperature Scales
Two temperatures scales in common use are the Fahrenheit and Celcius scales.
These scales are based on a specification of the number of increments between freezing
point and boiling point of water at the standard atmospheric temperature. The Celcius scale
has 100 units between these points, while the Fahrenheit scale has 180 units. The Celcius
scale is currently more in use because of the adoption of metric units. However, the
absolute temperature scale based on the themodynamic ideal Carnot cycle has been
correlated with the Celcius and Fahrenheit scales as follows :
K ( Absolute temperature, Kelvin scale) = 0C + 273.15
where 0C is temperature on Celcius scale.

Mechanical Engineering


Universitas Sriwijaya

R ( Absolute temperature, Rankine scale) = 0F + 459.69


where 0F is the temperature on the Fahrenheit scale.
The zero points on both the scales represent the same physical state and the ratio of
two value is the same, regardless of the absolute scale used, i.e.

[ ]


[ ]



The boiling and freezing points of water at pressure of one atmosphere (101.3
kN/m2) are taken as 1000 and 00 on the Celcius scale and 2120 and 320 on the Fahrenheit
scale. The following relationship between Fahrenheit and Celcius and Rankine scales can
be easily derived :

F = 32 +









With SI units, Kelvin temperature scale (which is also termed as absolute
temperature scale or ‘thermodynamic’ scale) is used in which the unit of temperature is the
Kelvin (K). It may be noted that degree symbol ( 0 ) is not used in the scale.

11.3 International Practical Temperature Scale
To enable the accurate calibration of a wide range of temperature in terms of the
Kelvin scale, the International Practical Temperature Scale (IPTS-68) has been devised.
This lists 11 primary ‘fixed’ points which can be reproduced accurately. Some typical
values are

Mechanical Engineering

Freezing point of silver 1235. Triple point of oxygen 54.259. Apart from the primary standard points. Triple point of equilibrium hydrogen (equilibrium between solid. radiation methods are used to extrapolate the scale.01 6. Boiling point of equilibrium hydrogen 20. Resistance thermometers are used for this purpose for temperatures below 6300C and thermocouples between 630 and 10640C.43 Primary ‘fixed points 1.18 . Boiling point of water 373.962 273.188 -182.15 100 7.73 419.2.93 9. Some typical values of these points are given in Table 11. and vapour hydrogen) 5.28 .252.58 961.87 3. there are 31secondary points on the International Practical Temperature Scale which forms the convenient working standard for the workshop calibration of the temperature measuring devices. Triple point of water (equilibrium between solid.789 4.218. liquid. Freezing point of zinc 692.1 Typical Values of Primary ‘Fixed’ Points Temperature (K) Temperature (0C) 13.16 0. Freezing point of gold 1337. Boiling point of oxygen 90.58 1064.361 .34 2.Universitas Sriwijaya Table 11. a means of interpolation between the fixed points is required. liquid and vapour phases of water) In order to establish a scale completely.58 8. Mechanical Engineering .

Mechanical Engineering .674 7.592 271.87 630. Non-electrical methods 2. change in the physical properties.15 0 4. Melting point of copper 1357.Melting point of tungsten 3660 3387 Secondary points 11.652 327.674 -74. Melting point of lead 600.Universitas Sriwijaya Table 11. Freezing point of mercury 234. Temperature measurement methods can be broadly classified as follows : 1. Sublimation point of carbon dioxide 194. Melting point of bismuth 544. 2.442 5. Electrical methods 3. and 3.476 2.52 660. change in the chemical properties.502 6. Melting point of alumunium 933.6 1084. change in the physical state.Melting point of platinum 2045 1772 11.5 10.4 Measurement Of Temperature Temperature cannot be measured directly but must be measured by observing the effect that temperature variation causes on the measuring device. Boiling point of pure sulphur 717. Radiation Methods 11.5 Non-Electrical Methods The non-electrical methods of temperature measurement can be based on any one of the following principles: 1. Equilibrum between ice and water (ice point) 273. Melting point of antimony 903.74 8.288 -38.37 9.824 444.862 3.2 Typical Values of Secondary Points Temperature (K) Temperature (0C) 1.

Mechanical Engineering . therefore they are commonly used in areas where there is a risk of explosion. e. namely the simple thermal expansion due to the change in temperature. 11.1 Bimetallic thermometer Fig.5. Further. we have to look for a refersible chemical properties.11. further.1(a)).g. for using the change in chemical properties. we have to look for a reversible chemical process in order to obtain a repeatable/reproducible scale with respect to change in temperature. provide display of temperature in petrol-storage tanks. welded or riveted together so that relative motion between them is prevented. In fact. most non-electrical methods of temperature measurement are based on the change in a physical property. although the magnitude of expansion in such thermometers is small. Hence these are not suitable for measuring a ranges of temperatures.Universitas Sriwijaya As mentioned before. etc.1 Bimetalllic thermometer This type of thermometer also employs the principle of solid expansion and consists of a ‘bimetal’ strip usually in the form of a cantilever beam (fig. having different coefficients of thermal expansion. yet they produce enough power for direct operation of the mechanical indicating devices. This comprises strips of two metals. such processes are very few in nature and thus this principle of temperature measurement is hardly used. 11. the temperatures at which a number of pure substances change their physical states are used for the calibration of temperature scales. But these devices give one particular value of a unique temperature corresponding to the change in the state of the substance. Since these devices have no electrical connections.

allowing the bimetal element ( enclosed in a protective sheath) to be submerged in a hot substance without the indicator it self being subjected to excessive temperatures. The length of the stem may be up to about 0. but the sensitivity is imprived by using a longer strip in a helical from as shown in fig. Inaccuracies of the order of ± 0.0% of full scale deflection are expected in bimetallic thermometer of high accuracies. This is an iron-nicel alloy containing 36% nicel. whereas nickel alloys are used when higher temperatures have to be measured.2. 11. Fig.6 m. Its coefficientof thermal expansion is around 1/20 th of the ordinary metals. Invar is commonly employed as the low expansion metal.Universitas Sriwijaya An in crease in temperature causes the deflection of the free end of the strip as shown in fig.5 to ± 7. the bimetal expands and the helical bimetal rotates at its free and. In response to the temperature change. bimetal elements find a wide range of applications in the combined sensing and Mechanical Engineering . assuming that metal A has the higher coefficient of expansion. 11. The deflection with the temperature is nearly linear. Bimetallic thermometers are usually employed in the range of -30 to 550ºC.1(b). depending mainly on the coefficient of linear thermal expansion.2 Bimetal helix thermometer One end of the helix is anchored to the casing and the other end which is free is conveniently connected to the pointer which sweep over a circular dial graduated in degrees of temperature. In addition to temperature indication. 11. Brass is used as high expansion material for the measurement of low temperatures. A plain bimetallic strip is somewhat insensitive. thus turning the stem and pointer to a new position on the dial.

Liquid-in-glass thermometers have notable qualities like low cost simplicity in use. However. down to -130°C for pentane. Precission thermometers are sometimes marked for partial or total immersion and also for horizontal or vertical orientation.2 Liquid-in-Glass Thermometer The liquid-in-glass thermometer is one of the most commont temperature measuring devices. aneroid barometers and as balance wheel compensators in some watches. Movement of the strip has sufficient force to actuate the control switches employed in domestic oven. The calibration of the thermometer should be occasionally checked against the ice point to take into account the aging effects. It contains a big bulb attached to a very fine capillary. It is not Mechanical Engineering . However. when increased accuracy is required.005°C. However. mainly of the on-off type (thermostats). The lower temperature limit is -37. electric irons.8°C for mercury. its main disadvantage is its inability to measure rapidly changing temperatures due to its relatively higher thermal inertia.5. Both liquid and glass expand on heating and their differential expansion is used to indicate the temperature. The range of the thermometer is limited to 5-6°C with an accuracy of 0. there by increasing its boiling point and range. car winker lamp and the refrigerators. these devices are also used as compensating element for the ambient temperatures in the pressure thermpmeters. The precision of the thermometer depends on the care used in calibration.1°C. The bimetallic strip has the advantage of being self-generating type instrument with low cost practically no maintenance expenses and stable operation over extended period of time. The higher temperature range is 340°C (boiling point of mercury is 357°C) but this range may be extended to 560°C by filling the space above mercury with CO2 or N2 at high pressure. In addition. Intermediate points are marked by inter-polation. portability and convenient visual indication without the useof any external power. A typical instrument is checked and marked from two to five reference temperatures. their use is limited ro certain laboratory applications.Universitas Sriwijaya control elements in temperature control sisyem. 11. a Beckmann range thermometer can be used. The accuracy of these thermometers does not exceed 0.

Universitas Sriwijaya preferred in industrial application because of its fragility and its lack of adaptability to remote indication. It has a relativity large metal bulb.3 Pressure Thermometers Pressure thermometer is based on the principle of fluid expansion due to an increase in the pressure in a given volume of the temperature measuring system.5. easy-to-read thermometer that maybe read remotely by connecting the bulb to a Bourdon gauge or any other pressure measuring device by means of a capillary tube as illustrated in Fig. 11. It is one of the most economical.3 A schematic diagram of pressure thermometer The entire assembly the bulb. The bulb of the thermometer may be filled with either a liquid (usually mercury) or gas or a liquid-vapour mixture and depending upon the type of fluid. it introduces time lag in the measurement of dynamic signals because of relatively high heat capacityof the bulb. This results in a robust. 11.(often stainless steel) instead of glass. versatile and widely used devices in industrial temperature measurements. the thermometer is termed as mercury-in-steel thermometer or constant volume gas thermometer respectively. capillary and gauge is calibrated is directly on the basis of pressure change corresponding to the temperature change. Further. 11. Mechanical Engineering .3 Fig.

self-operated types. Further. As a total expansion of mercury is dependent not only on the bulb temperature but also on the temperatures of the capillary tube and Bourdon tube. stable in operation and accurate to 10 on. with no maintenance expenses. the response of these instruments can be increased by using a small bulb connected to an electrical type of pressure sensor connected throught a short length of capillary tube. This inevitably increase the thermal capacity and hence the thermal lag. the system is subject to ambient temperature errors. then the Bourdon gauge experiences an increase in preassure equal to ρgh. Pressure thermometers are often connected with long capillary (connecting) tubes for the remote measurements. The volumes of the capillary and Bourdon tube are made as small as the transmission distance and the required size of display could allow. Ambient temperature error may be reduced by suitable compensation techniques. This increase in level introduces an error in the indicator. of the order of 100 m. The bulb size is therefore a matter of compromise. One ambient temperature change on this compensating system cancels the corresponding on the measuring system. Further. Mechanical Engineering . A reduction of the error is obtain by making the sensing-bulb volume considerably greater than that of the capillary and Bourdon tubes. This error is contanfor any specified relative position of the bulb and display and may the removed by means of the zero adjustments of the indicating mechanism.Universitas Sriwijaya Fluid expansion thermometers are low in cost. Sufficient power is available to operate a recording pen if required. Another potential cause of error is the change in pressure head which is introduced by any change in relative levels of the bulb and the display. Mercury-in-steel Thermometer The mercury-in-steel thermometer has a near-linear scale. rugged in construction. The temperature range over which mercury-in-steel thermometer may be used is -25 to 5500 C when the mercury is filled under pressure in the steel bulb. the magnitude of these error depends on the ratio of the volume of the mercury in the capillary at the measured temperature to the volume of the bulb. If the bulb is raised by a height h from the calibration elevation.

Universitas Sriwijaya Constant Volume Thermometer The constants volume thermometer uses an inert gas (usually nitrogen) in place of mercury and the principle of its working is the increase is pressure of the gas with increase in temperature at constant volume. capillary and Bourdon tube. Vapour Pressure Thermometer The system is vapour pressure thermometer is filled partly with liquid and partly with vapour of the same liquid so that these is a liquid-vapour interface in the bulb. that of bulb. pentane. chlorobenzene.e. ethyl alcohol. the volume of the system. The liquid-vapour system does not have any error as long as a free liquid surface exist in the sensing bulb. However. The general usefulness of the vapour pressure thermometers is restricted due to the limited number of liquids providing suitable mostl saturation vapour pressure ranges. toluene. there is only one saturation pressure corresponding to a given temperature. i. methyl chloride.e. The accuracy of these instruments is the order of ± 1% at lower ranges. ether. However. etc. does not remain constant and increases due to increase in temperature. These include mostly hydrocarbon type of liquids like ethane. Mechanical Engineering . The scale range is usually of the order of 1000C and accuracy is up to ± 1 % of the differential range. This is because such a system follows one of the Dalton’s laws of partial pressure which states that if both liquid and vapour present. i. ethyl chloride. the temperature is roughly a logarithmic function of the temperature (log p = a – b/T) and therefore the scale of the vapour pressure thermometers is noticeably non-linear. its disadvantages over the liquid filled system is that the pressure developed for a given temperature change is smaller and further ambient temperature compensation is more difficult. uo to 3000C and ± 2% above this range. Further. Gas filled system operate over a range -130 to 540 0 C with linear ranges asa large as 5000 C. acetone.

the change in resistance of various materials. Metallic resistace thermometers are very suitable for both laboratory and Mechanical Engineering . In general.6. There are two main electrical methods for measuring temperature.6 Electrical Methods Electrical methods are in general preferred for the measurement of temperature as they furnish a signal can be easily detected.e. Further.4 illustrates the typical variation of specific resistance of the metals (platinum for example) and the NTC thermistor.93 0C. Because it provides extremely reproducible output. However for the measurement of lower temperature up to 600 0C. Thermo-resistive type i. forms the basis of this important sensing technique. Thermo-sensitive resistors having such negative temperature characteristic are commonly known as NTC thermistors. RTD sensor is made of nickel.which varies in reproducible manner wit temperature. conductors (metals) and semiconductors.. They are : 1. the resistance of the highly conducting materials (metals) increases with increase in temperature and the coils of such materials are called metallic resistance thermometers. Whereas.e. Figure 11. Metallic Resistance Thermometers or Resistance-Temperature Detectors (RTDs) Metals such as platinum copper.1 Electrical Rsistance Thermometers In resistance thermometers.81 K to 961. amplified or used for control purposes. The materials in actual use fall in two classes namely. 11.Universitas Sriwijaya 11. the sensing element is fabricated using pure copper wire. Platinum is a very widely used sensor and its operating range is from 4 K to 1064 0C. tungsten and nickel exhibit small increases in resistance as the temperature rises because they have a positive temperature coefficient of resistance. for the ranges of temperature below 300 0C. emf generating transducers. variable resistance transducers and 2. Thermo-electric type i. the resistance of semiconductor materials generally (not always) decreases with increases in temperature. it is used in establishing International Practical Temperature Scale from 13..

RTD sensored’s may be fabricated by depositing thin the films of platinum. These thin film sensors have advantage of extremely low mass and consequently more rapid thermal response. However.5. This is because of the following : Mechanical Engineering . A typical construction is shown in Fig. thus giving a fast speed of response. The ends of the coils are welded to stiff copper leads that are taken out to be connected in one o the arms of the Wheatstone bridge circuit. relativily higher cost as compared to other temperature sensors ad their proneness to errors caused due to contact resistance. nickel or copper on a ceramic substrate. the limitations of the RTDs are low sensitivity. However. but the temperature sensitive element is usually in the form of a coil of fine wire supported in a stress-free manner. is the most widely used material for metallic resistance element. Fig. In some cases. this arrangement can be used directly in the medium whose temperature is being measured. a protective metal sheat is used to provide rigidity and mechanical strength. 11. shock and accelerations. 11. In addition. where the wire of metal is wound on the grooved hollow insulating ceramic former and covered with protective cement. in spite of its low sensivity and high cost as compared to nickel and copper. in the most applications. Alternatively.5 Construction of a platinum thermometer (PRT) Platinum.Universitas Sriwijaya industrial applications because of their high degree of accuracy as well as long-term stability. they have a wide operating range and have linear characteristic throughout the operating range. Metallic resistance thermometers are constructed in many forms.

0043 0 Nickel = 0.7) becomes: R2 = R1 + R0 α (t2 – t1) Mechanical Engineering . It forms the most easily reproducible type of temperature transducer with a high degree of accuracy. In general. 3. R0 = resistance at 00C t = temperarure relative to 00C. the resistance relationship of most metal over a wide range of temperature of temperatures is given by the quadratic relationship: R = R0 [ 1 + aT + bT2 ] where (11.7) where α = the temperature coefficient of resistance of material in (Ω/Ω/0C or 0C-1.01oC up to 500oC and ±0.Universitas Sriwijaya 1. 2. over a limited temperature range around 0C (273K). The temperature-resistance characteristics of pure platinum are well defined stable over a wide range of temperature. Eq. (11. The accuracy attainable with PRT is ± 0.6) R = resistance at absolute temperature T R0 = resistance at 00C a dan b = experimentally determined constans. However . It has high resistance to chemical attack and contamination ensuring long-term stability. the following linear relationship can be applied. Rt = R0 ( 1 + α t ) (11.0068 0 Platinum = 0. Some typical values of α are: Copper = 0.039 C-1 C-1 0 C-1 If a change temperature from t1 to t2 is considered.1oC up to 1200oC.

6 Cable Compensation arrangements for platinum resistance thermometer Mechanical Engineering . For more precise result. Leads of same length appropriate to the situation are normally required and any resistance change there in due to any cause. that the lead resistance be kept as low as possible relative to the element resistance. 11. Fig. It is desirable. However.6). These can be eliminated by utilising ac excitation or by manually varying the polarity of he dc supply.Universitas Sriwijaya Rearranging gives: t2 = t1 + R 2−R1 α R0 (11. therefore. including temperature. may be credited to the thermometer element. 11. particular attention must be given to the manner in which the thermometer is connected into the bridge. it is ssential that the thermo-electric emf’s do not affect the system. In addition.8) The variation of resistance of the sensing element is normally measured using some form of electrical bridge circuit which may employ either the deflection mode of operation or the null (manually or automatically balanced ) mode. either the Siemen’s three wire lead arrangement or Callender’s four wire lead arrangement may be employed (Fig. some modifications may be employed for providing the lead compensation. Further.

7 Range of thermistor forms Thermistors are fabricated from the semiconducting materials which include the oxides of copper. thin chip or wafer to a large sizes rod as illustrated in Fig. thin disc.7. manganese. cobalt. 2. resulting in a dense cermic body with the recuired resistancetemperature characteristics. These oxides are blended in a suitable proportion and compressed into desired shapes from powders and hea treated to recrystallise them. nickel. 11. The sizes can range from extremely small bead. Mechanical Engineering . lithium and titanium. They are available in a greater variety of shapes and sizes having cold resistance ranging from afew ohms to mega ohms. thermistors respond negatively to temperature and their coefficient of resistance is of the order of 10 times higher than that of platinum or copper. ability to withstand electrical and mechanical stresses. a large temperature coefficient which makes the thermistors an extremely sensitive device. Unlike metals. thus enabling accuracy of measurement up ± 0. Fig. Thermistors have the following adventages for temperature measurement: 1. 11.Universitas Sriwijaya Semiconductor Resistance Sensors (Thermistors) Thermistors (shortened form of the words: thermal resistor) is a thermally sensitive variable resistor made of ceramic-like semiconducting materials.01 oC with proper calibration.

The temperature-resistance characteristics of thermistor is of exponential type and is given by: [ ( )] β R = R0 exp where 1 1 − T T0 (11.045 K-1 R The value of dR/dT R for platinum is 0.0039 K-1 . fairly good operating range which lies between – 100 and 300oC. fairly cost and easy adaptability to the available resistance bridge circuits. 4.6.10) Assuming we get . 11.2 Thermo-electric Sensors Mechanical Engineering . and 5.Universitas Sriwijaya 3. β = 4000 K and T = 298 K dR/dT = -0. the disadventages are a highly non-linear resistance-temperature characteristics and problems of self-heating effects which necessitate the use of much lower current levels than those with metallic sensors.9) R0 is the resistance at the reference temperature T0 (Kelvin) R is the resistance at the measured temperature T (Kelvin) β is the experimentally determine constant for the given thermistor material The values of β usally lie between 3000 and 4400 K depending on the formulation or grade.9) we can obtain the temperature coefficient of resistance as : dR/dT R = β 2 T (11. (11. the high sensitivity and the availability in extremely small sizes (of the sizes of a pin head) enable a fast speed of thermal response. Thus. indicating that the thermistor is at least 10 times more sensitive than the platinum resistance element. these devices are extremely useful for dynamic temperature measurement. Using Eq. However.

It consists of two wire of different metals twisted and brazed or welded together with each wire covered with insulation which may be either.the overall relation between emf E or the flow of current has been found that the magnitude of E depends upon the materials as well as the temperature T1 and T2. then an infinite resistance voltmeter detects the electromotive Force E. or if a low resistance ammeter is connected. say A and B.8 Basic thermo-electric circuit Mechanical Engineering .Universitas Sriwijaya The most common electrical method of temperature measurement uses the thermo-electric sensor. The construction of a thermocouple is quite simple. such as ice point which is very commonly used in Mineral (magnesium oxide) insulation for normal duty or 2. when two conductors of dissimiliar metals. respectively.8. of the two junctions is usually maintained at some known temperature. 11. The thermocouple is a temperature transducer that develops an emf which is a function of the temperature between hot junction and cold junction. The measured emf E then indicates the temperature difference relative to the reference temperature. for convenience of measurements and standardisation. Experimentally. Fig. 1. are joined together to form a loop (thermocouple) and two unequal temperatures T1 And T2 are interposed at two junctions J1 and J2. a current flow I is measured. also known as the thermocouple (TC). Ceramic insulation for heavy duty The basic principle of temperature measurement using a thermo-electric sensor was discoveredby Seebeck in 1821 and is illustrated in fig.

if the thermocouple voltage is measured by means of potentiometer. of junctions J1 and J2 respectively are slightly altered if the thermo-electric current is allowed to flow in the circuit.Universitas Sriwijaya It may be noted that temperature T1 and T2. no current flows and Peltier heating and cooling are not present. This phenomenon is termed Peltier effect.In addition.9). Heat is generated at the cold junction and is absorbed from the hot junction thereby heating the cold junction slightly and cooling the hot junction slightly. the Thomson effect may be neglected in practical thermo-electrics circuits and potentiometric voltage measurements are not susceptible to this error as there is no current flow in the circuit. Again. the junction emf may be slightly altered if a temperature gradient exists along either or both the materials. Law of Intermediate Temperatures This states that the emf generated in thermocouple with junctions at temperatures T1 and T3 is equal to the sun of the emf’s generated by similar thermocouples. Further.9 Law of intermediate temperatures Mechanical Engineering . one acting between T1 and T2 and the other between T2 and T3 when T2 lies between T1 and T2 (Fig. 11. This is known as Thomson effect. 11. The actual application of thermocouples to the measuresment requires consideration of the laws of thermo-electricity. these heating and cooling effects are proportional to the current and are fortunately quite negligible in a thermocouple circuit which is practically a millivolt range circuit. Fig.

it is preferable to employ the system shown in Fig. then the correction required for the observation would be the emf produced by the thermocouple between 0 and 20°C .10 Law of intermediate metals It may be noted extensions wires are needed when the measuring instrument is to be placed at a considerable distance from the reference junction. (b). For example. 11. This ensures maximum accuracy in the thermocouples operation.10 (a)].11 (b) to keep the copper. Maximum accuracy is obtained when the leads are of the same material as the thermocouple element [Fig. Therefore. The emf generated remains unaltered if the two new junctions B-C and C-A are the same temperature. 11.Universitas Sriwijaya This law is useful in practice because it helps in giving a suitable correction in case reference junction temperature (which is usually an ice bath at 0°C is employed. 10. a small inaccuracy is still possible if the binding post of the instrument is made of say copper and the two binding posts are at different temperatures. if a themocouple is calibrated for a reference junction temperature of 0°C and used with junction temperature of say 20°C.iron and copper-costantan junctions in the thermos flask at 0°C and provide binding posts copper. then three junction are formed as shown in fig. However this approach is not economical while using exspensive thermocouple materials. If a third wire in introduced. 11. Fig. 11 11 (a)]. 11. Law of Intermediate Metals The basic thermocouple loop consist of two dissimiliar metals A and B [ Fig. Further. Mechanical Engineering .

etc. Rase-metal thermocouple.11) Thermocouple can be broadly classified in two categories: 1. cheaper and have nearly linear characteristics. Linearity characteristics. Mechanical Engineering .Universitas Sriwijaya Fig. Base-metal thermocouples use the combination of pure metals and alloys of iron. 11. It may be noted that the relationship between thermo-electric emf and the difference between hot and cold junctions temperatures is approximately of the parabolic form: E=aT + bT2 (11. Ability to withstand the temperature at which they are used. 2. Base-metal thermocouples.11 Schematics of Thermocouple circuits with and without extension leads in a typical iron-constant an thermocouple circuit Thermocouples Materials The choice of the materials for thermocouples in governed by the following factors: 1. copper and nickel and are used for temperature up to 1450 K. These are most commonly used in practice as they are more sensitive . and 2. and 3. Qhich ensures maintenance of the precise thermoelectric properties with continuous use. Immunity from contamination /oxidation.

Table 11.3 Characteristics of Some Thermocouples Approximate Useful Approximate sensitivity Temperature accuracy Thermocouples S. K Chromel-Alumel 40-55 -200 to +1300 ± 0.Universitas Sriwijaya Their chief limitation is the lower operating range because of their low melting point and vulnerability to oxidation. When connected in parallel.75 3. This is commonly known as thermopile. rhodium and molybdenum for temperatures up to 3000°C.3. S Platinum-Platinum/10% Rhodium 5-12 0 to +1400 ± 0.25 6. J Iron-Constantan 45-55 -180 to +850 ± 0.No. rare metal thermocouples use a combinations of pure metals and alloys of platinum for temperatures up to 1600°C and tungsten. E Cheromel-Constantan 55-80 -180 to +850 ± 0. alumel = nickel/alumunium For special purposes where high sensitivityis needed. Type material 0 0 in (μV/ C range( C) (%) 1.5 5. Rhenium-Tungston/20% W5 Rhenium * Constantan = copper/nickel.75 4. The following are the advantage of the TC sensors: Mechanical Engineering . R Platinum-Platinum/13% Rhodium 5-12 0 to +1600 ± 0. B 5-12 +100 to +1800 ± 0. T Copper-Constantan 20-60 -180 to +440 ± 0.15 Platinum/30% Rhodium-Platinum/6% Rhodium Tungston/5% 8.25 7.25 5-12 0 to +3000 ± 0. a group of thermocouples will give a reading that is the numerical average of the individual thermocouple being the same. are given in table 11. thermocouples may be attached in series. Typical thermocouples with their temperature ranges and other salient operating characteristics. The output is then the numerical sum of the voltages expected from each of the single couples. On the other hand. chromel = nickel/chromium.75 2.

They have good accuracy of the order of ± 0. 2. 3.C device like silicon diodes and transistors exhibit a stable and reproducible response to temperature. they can withstands rough handling. 2. 9.e. They cost considerably less as compared to other thermal sensors and further. have the following limitations : 1. The require insulation covering while using them in conducting fluids 3. Output signal is electrical and they can be used for indicating recording or 6.3 Solid State Temperature Sensors Common I. 7.75 % of f. They are quite rugged type. i. When a PN junction is forward biased by a constant current source. 8. emf is independent of length or diameter of the wire.e. 5.e. microprocessor based control system Output signal. 4.d They have excellent stability for a long period of time. however. from -200 to 3000 0C. The output signal. emf requires amplification in most applications.Universitas Sriwijaya 1.2 to ± 0. i.38 x 10-23 J/K) Mechanical Engineering . The TC sensors. They can be conveniently mounted in a variety of temperature measurement situations.6. 11. its governing equation between current and voltage is as follows: VBE = kT q ln Ic I es (11. Inhomogeneity of composition of the thermocouple material and cold working of wires affect the sensitivity of the thermocouple. they require no maintenance. In other words dynamic response of sensor is fairly good. They cover wide range of temperature.e. i. i. Thermocouple bead can be made of small size and consequently with low thermal capacity.12) where VBE = base emitter voltage Ic = collector current Ies = emitter saturation current K = Boltzman constant (1.s.

yhe detection of change in frequency of oscillation of 1 Hz gives a resolution of 0. temperature in the range of -40 to 230oC can be measured precisely and accurately by this method.6 x 10-19 C) T = absolute temperature (K) Generally. Further.4 Quartz Thermometer A piezo-electric crystal provides a highly accurate and sensitive methode of temperature measurement based on the change in its resonant frequency which is directly proportional to the temperature change. the crystal is cut in the form of shear type LC cut.e. In the order words.5% of F. The main avantage of the solid state temperature sensors is heir inherent lnear operating characteristics with excellent accuracy of the order of ± 1oC. (11.S. 11. The associated electronic circuitry of this thermometer consists of frequency counters and digital read-out of the measured frequency.12) is constant and the emitter base voltage i.6. the term within the parenthesis in Eq. The sensitivity of the silicon transistor wihin its useable range of -55 to 150 oC is of the order of -2mV/oC. since the output is electrical. Herein.Universitas Sriwijaya q = electron change (1. The advantages of the quartz thermometer are : 1. in which the change in resonant frequency is highly linear as well as repeatable. they have the capability of μp based control applications. Highly linear output as the linearity error is ± 0. the output of the transducer becomes directly propotational to T whih is the measured input.001oC. The fundamental frequency f0 depends on the thickness of the crystal and can be adjusted so as to give a sensitivity of the order of 1000 Hz for a temperature change of 1oC. In addition they have high levels of output signal which is capable of direct indication without n signal conditioning.. Further. Mechanical Engineering . The disadvantages of these sensors are their limited temperature measuring range and their thermal mass which limits their response characteristics.

thermistor or thermocouple. a non-contacting device for measuring the temperature is most convenient. and (c) infrared (IR) pyrometer. The first is sensitive to all the radiation that enters the instrument and the second only to radiation of a particular wavelength. Long-term stability and reliability.001oC. -40 to 230oC. 3. the IR pyrometers employ the infrared portion of the spectrum by using a thermal detector to measure the temperature on the surface of the body. for bodies that are moving.7 RADIATION METHODS (PYROMETRY) All the temperature measuring device discussed so far i. Secondly. the thermometer may melt at the high temperature. 4.e. Excellent repeatability in the measuring range of -40 to 230oC. Instruments that employ radiation principles fall into three general classes : (a) total radiation pyrometer.Universitas Sriwijaya 2. This means that the thermometer must be capable of with standing this temperature. For temperatures above 650oC. The limitations of the quartz thermometer are : 1. Mechanical Engineering . Further. the heat radiations emitted from the body are of sufficient intensity to be used for the measuring the temperature. Thirdly.. In the case of very hot bodies. 11. Limited measuring range i. etc. Pressure thermometer. require the thermometer to be brought into physical contact with the body whose temperature is to be measured. Piezo-electric crystals have strong cross-sensitivity for pressure changes if they occur simultaneously in the temperature measuring systems. (b) selective (or partial) radiation pyrometers. if the distribution of temperature over the surface of an object is required. a noncontacting device can readily 'scan' the surface. High resolution of the order of 0. 2.e.

whereas the infrared radiations are associated with relatively large wavelengths of 0.32 to 4m μm) although theory indicates that rhey should be sensitive to the entire spectrum of energy radiated by the object.3 to 0. Fig. It consists of blackened tube T open at one end to recieve the radiations from the object whose temperature is to be measured.e. 11.Universitas Sriwijaya 11. The other end of the tube has a sighting aperture in which an adjustable eyepiece is usually fitted. The thermal radiations impinge on the concave mirror whose position can be adjusted suitably by a rack-and-pinion arrangement so as to get proper focussing of the thermal radiations on the detector disc S.12 shows a schematic diagram of the Fery's total radiation pyrometer. Ordinary glass is unsatisfactory.1 Total Radiation Pyrometer The total radiation pyrometer receives a controlled sample of the total radiation of a hot boddy (say a furnace) and focusses it on a temperature sensitive transducer.. (i. The detector disc is usually of blackened platinum sheet/foil and is connected to a thermocouple/thermopile junctions or to a Mechanical Engineering . It may be noted that the wave lengths of light in visible range is form 0. as it absorbs infrared radiations.72 μm.7. They require special optical materials for focussing. from 0. In fact.12 Schematic of Fery’s total radiation pyrometer Figure 11. The term 'total radiation' includes both visible (light) and invisible (infrared) radiations. the practical radiation pyrometers are sensitive to a limited wavelength band of radiant energy.72 to 100 μm.

13) EA/B = is the energy recieved by the pyrometer in W/m-2 C = is a geometrical factor to adjust the relative shapes of the two bodies ε = is the emissivity of the detector disc which varies from 0.2000oC where thermocouples and resistance thermometers cannot be employed. TA and TB are the steady state absolute temoerature of the source and pyrometer detector disc. Secondly. an error may be caused due to a surface having emissivity other than used in the calibration. Since surface emissivities are not known very accurately an d a change occurs with time due to oxidation. the total radiation pyrometer is not a very accurate temperature indicator. dust. gases.0 for the theoritical black body.Universitas Sriwijaya resistance thermometer bridge circuit. The theory underlying the operation of total radiation pyrometers is that the rate of radiation from a body A (the source) to a body B (the pyrometer).7 x 10 -12 kW/ (m2-K4). EA/B is given by the Stafan-Boltzmann law as follows : EA/B = Cεσ [TA4 – TA4] were (11. windows. Such pyrometers are usually calibrated againts known temperatures in the range of 700 . Any filtering material such as smoke. To reduce such un certainties. etc. the errors arise from two sources in actual use. σ = is the Stefan Boltzman constant and its value is 56. which were not present in the calibration will reduce the energt recieved hence cause an unknown error. Leads from the detector are led out of the casing to a meter for measuring the thermoelectric emf or the variation the electric resistance of the platinum foil. However.e. it can be used to Mechanical Engineering . therefore the error due unknown emissivity is usually not known. In view of the troubles due to filtering and emissivity. However. i.05 to 1. pyrometers calibrated from time to time in actual use.

In the Mechanical Engineering .4385 (cm-0C) λ = wavelength (cm) T = absolute temperature in (K) W = energy level associated with wavelength at temperature T (W/cm3) The classical form of this optical pyrometer is the disappearing filament optical pyrometer (or the monochromatic brightness radiation pyrometer).740 x 10-12 (W-cm2) c2 = 1. A typical use is a large furnace in metal industries. As the temperature increases. 11.Universitas Sriwijaya good advantage in fixed locations where the emissivity and optical paths are well known and constant. This instrument is used realise the International Practical Temperature Scale above 1063oC.7.14) c1 = 3. however. The signal is electrical and therefore can be used for control applications. the emissive power shifts to shorter wavelengths. It is most accurate of all radiation pyrometers. It is abvious from Planck's distribution equation that for a given wavelength. The planck's distribution equation is : c1 λ −5 W= where e c 2/ λT −1 (11. it is limited to temperatures greater than about 700oC since it requires visual brightness match by a human operator.2 Selective Radiation Pyrometer The principle of this instrument is based on Planck's law which states that the energy level in the radiations from a hot body are distributed in the different wavelengths. the radiant intensity (brightness) varies with the temperature.

Because of the manual null balancing principle. 11. is previously calibrated so that when the current trough the filament is known.13 Schematic of the disappearing filament type of optical pyrometer The tungsten lamp. The temperature calibration is made in terms of the lamp heating current. A red filter that passes only a narrow band of wavelengths around 0. The observer controls the lamp current until the filament disappear in the superimposed target image [Fig.Universitas Sriwijaya disappearing filament instrument shown in the Fig. 11. the brightness temperature of the filament is also known. However. the opticcal pyrometer is not usable for continuous recording or automatic control application.65 μm is placed between the observer eye and the tungsten lamp and the target image. an image of the target is superimposed on the heated filament. 11.14 (c)]. it is more Mechanical Engineering .13. which is very stable. Fig.

Mechanical Engineering .Universitas Sriwijaya accurate and less subject to large errors than the total radiation pyrometer. its accuracy is better than ± 10oC. The accuracy of such pyrometers is usually ± 5oC.