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 laymans
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 degreeof 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

(11.1)

Universitas Sriwijaya

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

(11.2)

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.

[ ]
T2
T1

Rankine

[ ]
T2
T1

Kelvin

(11.3)
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 :
0

F = 32 +

R=

9
5

9
5

(11.4)

(11.5)

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

Universitas Sriwijaya

Table 11.1

Typical Values of Primary Fixed Points

Temperature (K)

Temperature (0C)

13.18

- 259.34

2. Boiling point of equilibrium hydrogen

20.28

- 252.87

3. Triple point of oxygen

54.361

- 218.789

4. Boiling point of oxygen

90.188

-182.962

273.16

0.01

6. Boiling point of water

373.15

100

7. Freezing point of zinc

692.73

419.58

8. Freezing point of silver

1235.58

961.93

9. Freezing point of gold

1337.58

1064.43

Primary fixed points

1. Triple point of equilibrium hydrogen (equilibrium between solid,
liquid, and vapour hydrogen)

5. Triple point of water (equilibrium between solid, liquid and

vapour phases of water)

In order to establish a scale completely, a means of interpolation between the fixed

points is required. Resistance thermometers are used for this purpose for temperatures
below 6300C and thermocouples between 630 and 10640C, radiation methods are used to
extrapolate the scale.
Apart from the primary standard points, 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. Some typical values of
these points are given in Table 11.2.

Mechanical Engineering

Universitas Sriwijaya

Table 11.2 Typical Values of Secondary Points

Temperature (K)

Temperature (0C)

1. Sublimation point of carbon dioxide

194.674

-74.476

2. Freezing point of mercury

234.288

-38.862

3. Equilibrum between ice and water (ice point)

273.15

4. Melting point of bismuth

544.592

271.442

5. Melting point of lead

600.652

327.502

6. Boiling point of pure sulphur

717.824

444.674

7. Melting point of antimony

903.87

630.74

8. Melting point of alumunium

933.52

660.37

9. Melting point of copper

1357.6

1084.5

10.Melting point of platinum

2045

1772

11.Melting point of tungsten

3660

3387

Secondary points

11.4 Measurement Of Temperature

Temperature cannot be measured directly but must be measured by observing the
effect that temperature variation causes on the measuring device. Temperature
measurement methods can be broadly classified as follows :
1. Non-electrical methods
2. Electrical methods
3. 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. change in the physical state,
2. change in the chemical properties, and
3. change in the physical properties.

Mechanical Engineering

Universitas Sriwijaya

As mentioned before, 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. Hence these are not suitable for measuring a ranges of
temperatures. Further, for using the change in chemical properties, we have to look for a
refersible 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. However,
such processes are very few in nature and thus this principle of temperature measurement
is hardly used. In fact, most non-electrical methods of temperature measurement are based
on the change in a physical property, namely the simple thermal expansion due to the
change in temperature. Since these devices have no electrical connections, therefore they
are commonly used in areas where there is a risk of explosion, e.g.to provide display of
temperature in petrol-storage tanks, etc. further, although the magnitude of expansion in
such thermometers is small, yet they produce enough power for direct operation of the
mechanical indicating devices.
11.5.1 Bimetallic thermometer

Fig. 11.1 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.11.1(a)). This comprises strips of two metals, having different coefficients of
thermal expansion, welded or riveted together so that relative motion between them
is prevented.

Mechanical Engineering

Universitas Sriwijaya

An in crease in temperature causes the deflection of the free end of the strip as
shown in fig. 11.1(b), assuming that metal A has the higher coefficient of
expansion. The deflection with the temperature is nearly linear, depending mainly
on the coefficient of linear thermal expansion. Invar is commonly employed as the
low expansion metal. This is an iron-nicel alloy containing 36% nicel. Its
coefficientof thermal expansion is around 1/20 th of the ordinary metals. Brass is
used as high expansion material for the measurement of low temperatures, whereas
nickel alloys are used when higher temperatures have to be measured. A plain
bimetallic strip is somewhat insensitive, but the sensitivity is imprived by using a
longer strip in a helical from as shown in fig. 11.2.

Fig. 11.2 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 response to the temperature change, the bimetal expands
and the helical bimetal rotates at its free and, thus turning the stem and pointer to a
new position on the dial. The length of the stem may be up to about 0.6 m, 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.
Bimetallic thermometers are usually employed in the range of -30 to 550C.
Inaccuracies of the order of 0.5 to 7.0% of full scale deflection are expected in
bimetallic thermometer of high accuracies. In addition to temperature indication,
bimetal elements find a wide range of applications in the combined sensing and
Mechanical Engineering

Universitas Sriwijaya

control elements in temperature control sisyem, mainly of the on-off type

(thermostats). Movement of the strip has sufficient force to actuate the control
switches employed in domestic oven, electric irons, car winker lamp and the
refrigerators. In addition, these devices are also used as compensating element for
the ambient temperatures in the pressure thermpmeters, aneroid barometers and as
balance wheel compensators in some watches.
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. However, its main disadvantage is its inability to
measure rapidly changing temperatures due to its relatively higher thermal inertia.

11.5.2 Liquid-in-Glass Thermometer

The liquid-in-glass thermometer is one of the most commont temperature
measuring devices. Both liquid and glass expand on heating and their differential
expansion is used to indicate the temperature. The lower temperature limit is
-37,8C for mercury, down to -130C for pentane. The higher temperature range is
340C (boiling point of mercury is 357C) but this range may be extended to 560C
by filling the space above mercury with CO2 or N2 at high pressure, there by
increasing its boiling point and range. The precision of the thermometer depends on
the care used in calibration. A typical instrument is checked and marked from two
to five reference temperatures. Intermediate points are marked by inter-polation.
The calibration of the thermometer should be occasionally checked against the ice
point to take into account the aging effects. Precission thermometers are sometimes
marked for partial or total immersion and also for horizontal or vertical orientation.
The accuracy of these thermometers does not exceed 0.1C. However, when
increased accuracy is required, a Beckmann range thermometer can be used. It
contains a big bulb attached to a very fine capillary. The range of the thermometer
is limited to 5-6C with an accuracy of 0.005C.
Liquid-in-glass thermometers have notable qualities like low cost simplicity
in use, portability and convenient visual indication without the useof any external
power. However, their use is limited ro certain laboratory applications. It is not
Mechanical Engineering

Universitas Sriwijaya

preferred in industrial application because of its fragility and its lack of adaptability
to remote indication. Further, it introduces time lag in the measurement of dynamic
signals because of relatively high heat capacityof the bulb.

11.5.3 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. It
is one of the most economical, versatile and widely used devices in industrial
temperature measurements. It has a relativity large metal bulb.(often stainless steel)
instead of glass. This results in a robust, 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. 11.3

Fig. 11.3 A schematic diagram of pressure thermometer

The entire assembly the bulb, capillary and gauge is calibrated is directly on
the basis of pressure change corresponding to the temperature change. 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, the thermometer is
termed as mercury-in-steel thermometer or constant volume gas thermometer
respectively.

Mechanical Engineering

Universitas Sriwijaya

Fluid expansion thermometers are low in cost, self-operated types, rugged

in construction, with no maintenance expenses, stable in operation and accurate to
10 on. Further, 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.

Mercury-in-steel Thermometer
The mercury-in-steel thermometer has a near-linear scale. Sufficient power is
available to operate a recording pen if required. 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, the system is subject to ambient temperature errors. Further, 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. A reduction of the error is obtain by
making the sensing-bulb volume considerably greater than that of the capillary and
Bourdon tubes. This inevitably increase the thermal capacity and hence the thermal lag.
The bulb size is therefore a matter of compromise. The volumes of the capillary and
Bourdon tube are made as small as the transmission distance and the required size of
display could allow. Pressure thermometers are often connected with long capillary
(connecting) tubes for the remote measurements, of the order of 100 m.
Ambient temperature error may be reduced by suitable compensation techniques.
One ambient temperature change on this compensating system cancels the corresponding
on the measuring system.
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. If the bulb is raised by a height
h from the calibration elevation, then the Bourdon gauge experiences an increase in
preassure equal to gh. This increase in level introduces an error in the indicator. 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.
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.

Mechanical Engineering

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. However, the volume of the system, i.e. that of bulb,
capillary and Bourdon tube, does not remain constant and increases due to increase in
temperature.
Gas filled system operate over a range -130 to 540 0 C with linear ranges asa large
as 5000 C. However, 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. The accuracy of these instruments is the order of 1% at
lower ranges, i.e. uo to 3000C and 2% above this range.

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. The
liquid-vapour system does not have any error as long as a free liquid surface exist in the
sensing bulb. This is because such a system follows one of the Daltons laws of partial
pressure which states that if both liquid and vapour present, there is only one saturation
pressure corresponding to a given temperature.
The general usefulness of the vapour pressure thermometers is restricted due to the limited
number of liquids providing suitable mostl saturation vapour pressure ranges. These
include mostly hydrocarbon type of liquids like ethane, ethyl alcohol, ethyl chloride,
methyl chloride, chlorobenzene, toluene, pentane, ether, acetone, etc. The scale range is
usually of the order of 1000C and accuracy is up to 1 % of the differential range. Further,
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.

Mechanical Engineering

Universitas Sriwijaya

11.6 Electrical Methods

Electrical methods are in general preferred for the measurement of temperature as
they furnish a signal can be easily detected, amplified or used for control purposes. There
are two main electrical methods for measuring temperature. They are :
1. Thermo-resistive type i.e., variable resistance transducers and
2. Thermo-electric type i.e., emf generating transducers.

11.6.1 Electrical Rsistance Thermometers

In resistance thermometers, the change in resistance

of various

materials,which varies in reproducible manner wit temperature, forms the basis of

this important sensing technique. The materials in actual use fall in two classes
namely, conductors (metals) and semiconductors. In general, 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, the
resistance of semiconductor materials generally (not always) decreases with
increases in temperature. Thermo-sensitive resistors having such negative
temperature characteristic are commonly known as NTC thermistors. Figure 11.4
illustrates the typical variation of specific resistance of the metals (platinum for
example) and the NTC thermistor.

Metallic Resistance Thermometers or Resistance-Temperature Detectors (RTDs)

Metals such as platinum copper, tungsten and nickel exhibit small increases in
resistance as the temperature rises because they have a positive temperature coefficient of
resistance. Platinum is a very widely used sensor and its operating range is from 4 K to
1064 0C. Because it provides extremely reproducible output, it is used in establishing
International Practical Temperature Scale from 13.81 K to 961.93 0C. However for the
measurement of lower temperature up to 600 0C, RTD sensor is made of nickel. Further,
for the ranges of temperature below 300 0C, the sensing element is fabricated using pure
copper wire. Metallic resistace thermometers are very suitable for both laboratory and

Mechanical Engineering

Universitas Sriwijaya

industrial applications because of their high degree of accuracy as well as long-term

stability. In addition, they have a wide operating range and have linear characteristic
throughout the operating range. However, the limitations of the RTDs are low sensitivity,
relativily higher cost as compared to other temperature sensors ad their proneness to errors
caused due to contact resistance, shock and accelerations.
Metallic resistance thermometers are constructed in many forms, but the
temperature sensitive element is usually in the form of a coil of fine wire supported in a
stress-free manner. A typical construction is shown in Fig. 11.5, where the wire of metal is
wound on the grooved hollow insulating ceramic former and covered with protective
cement. 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. In some cases, this
arrangement can be used directly in the medium whose temperature is being measured,
thus giving a fast speed of response. However, in the most applications, a protective metal
sheat is used to provide rigidity and mechanical strength. Alternatively, RTD sensoreds
may be fabricated by depositing thin the films of platinum, nickel or copper on a ceramic
substrate. These thin film sensors have advantage of extremely low mass and consequently
more rapid thermal response.

Fig. 11.5 Construction of a platinum thermometer (PRT)

Platinum, in spite of its low sensivity and high cost as compared to nickel and
copper, is the most widely used material for metallic resistance element. This is because of
the following :

Mechanical Engineering

Universitas Sriwijaya

1. The temperature-resistance characteristics of pure platinum are well defined

stable over a wide range of temperature.
2. It has high resistance to chemical attack and contamination ensuring long-term
stability.
3. It forms the most easily reproducible type of temperature transducer with a high
degree of accuracy. The accuracy attainable with PRT is 0.01oC up to 500oC and
0.1oC up to 1200oC.
In general, 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

(11.6)

= resistance at absolute temperature T

R0

= resistance at 00C

a dan b

= experimentally determined constans.

However , over a limited temperature range around 0C (273K), the following linear
relationship can be applied.
Rt = R0 ( 1 + t )

(11.7)

where = the temperature coefficient of resistance of material in (//0C or 0C-1.

R0 = resistance at 00C
t = temperarure relative to 00C.
Some typical values of are:
Copper

= 0.0043

Nickel

= 0.0068

Platinum = 0.039

C-1
C-1

C-1

If a change temperature from t1 to t2 is considered, Eq. (11.7) becomes:

R2 = R1 + R0 (t2 t1)

Mechanical Engineering

Universitas Sriwijaya

Rearranging gives:

t2 = t1 +

R 2R1
R0

(11.8)

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. However, particular
attention must be given to the manner in which the thermometer is connected into the
bridge. Leads of same length appropriate to the situation are normally required and any
resistance change there in due to any cause, including temperature, may be credited to the
thermometer element. It is desirable, therefore, that the lead resistance be kept as low as
possible relative to the element resistance. In addition, some modifications may be
employed for providing the lead compensation. For more precise result, either the
Siemens three wire lead arrangement or Callenders four wire lead arrangement may be
employed (Fig. 11.6). Further, it is ssential that the thermo-electric emfs do not affect the
system. These can be eliminated by utilising ac excitation or by manually varying the
polarity of he dc supply.

Fig. 11.6 Cable Compensation arrangements for platinum resistance thermometer

Mechanical Engineering

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. They are available in a
greater variety of shapes and sizes having cold resistance ranging from afew ohms to mega
ohms. The sizes can range from extremely small bead, thin disc, thin chip or wafer to a
large sizes rod as illustrated in Fig. 11.7. Unlike metals, thermistors respond negatively to
temperature and their coefficient of resistance is of the order of 10 times higher than that of
platinum or copper.

Fig. 11.7 Range of thermistor forms

Thermistors are fabricated from the semiconducting materials which include the
oxides of copper, manganese, nickel, cobalt, lithium and titanium. These oxides are
blended in a suitable proportion and compressed into desired shapes from powders and hea
treated to recrystallise them, resulting in a dense cermic body with the recuired resistancetemperature characteristics.
Thermistors have the following adventages for temperature measurement:
1. a large temperature coefficient which makes the thermistors an extremely
sensitive device, thus enabling accuracy of measurement up 0.01 oC with
proper calibration,
2. ability to withstand electrical and mechanical stresses,

Mechanical Engineering

Universitas Sriwijaya

3. fairly good operating range which lies between 100 and 300oC,
4. fairly cost and easy adaptability to the available resistance bridge circuits, and
5. 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, these devices are extremely useful for dynamic temperature measurement.
However, 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.
The temperature-resistance characteristics of thermistor is of exponential type and
is given by:

[ ( )]

R = R0 exp
where

1 1

T T0

(11.9)

R0 is the resistance at the reference temperature T0 (Kelvin)

R is the resistance at the measured temperature T (Kelvin)
is the experimentally determine constant for the given thermistor material
The values of usally lie between 3000 and 4400 K depending on the formulation

or grade. Using Eq. (11.9) we can obtain the temperature coefficient of resistance as :
dR/dT
R

2
T

(11.10)
Assuming
we get ,

= 4000 K and T = 298 K

dR/dT
= -0.045 K-1
R

The value of

dR/dT
R

for platinum is 0.0039 K-1 , indicating that the thermistor is

at least 10 times more sensitive than the platinum resistance element.

11.6.2 Thermo-electric Sensors

Mechanical Engineering

Universitas Sriwijaya

The most common electrical method of temperature measurement uses the

thermo-electric sensor, also known as the thermocouple (TC). The thermocouple is
a temperature transducer that develops an emf which is a function of the
temperature between hot junction and cold junction. The construction of a
thermocouple is quite simple. It consists of two wire of different metals twisted and
brazed or welded together with each wire covered with insulation which may be
either.
1. Mineral (magnesium oxide) insulation for normal duty or
2. 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. 11.8. when two
conductors of dissimiliar metals, say A and B, are joined together to form a loop
(thermocouple) and two unequal temperatures T1 And T2 are interposed at two
junctions J1 and J2, respectively, then an infinite resistance voltmeter detects the
electromotive Force E, or if a low resistance ammeter is connected, a current flow I
is measured. Experimentally,is has been found that the magnitude of E depends
upon the materials as well as the temperature T1 and T2.now,the overall relation
between emf E or the flow of current I. for convenience of measurements and
standardisation,one of the two junctions is usually maintained at some known
temperature. The measured emf E then indicates the temperature difference relative
to the reference temperature, such as ice point which is very commonly used in
practice.

Fig. 11.8 Basic thermo-electric circuit

Mechanical Engineering

Universitas Sriwijaya

It may be noted that temperature T1 and T2, of

junctions J1 and J2

respectively are slightly altered if the thermo-electric current is allowed to flow in

the circuit. 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. This phenomenon is termed Peltier effect.if the thermocouple voltage is
measured by means of potentiometer, no current flows and Peltier heating and
cooling are not present. Further, 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.In addition, the junction emf may be slightly
altered if a temperature gradient exists along either or both the materials. This is
known as Thomson effect. Again, 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.
The actual application of thermocouples to the measuresment requires
consideration of the laws of thermo-electricity.

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 emfs generated by similar thermocouples, one acting
between T1 and T2 and the other between T2 and T3 when T2 lies between T1 and T2 (Fig.
11.9).

Fig. 11.9 Law of intermediate temperatures

Mechanical Engineering

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 0C is employed. For
example, if a themocouple is calibrated for a reference junction temperature of 0C and
used with junction temperature of say 20C, then the correction required for the
observation would be the emf produced by the thermocouple between 0 and 20C .
Law of Intermediate Metals
The basic thermocouple loop consist of two dissimiliar metals A and B [ Fig. 11.10
(a)]. If a third wire in introduced, then three junction are formed as shown in fig. 11. 10.
(b). The emf generated remains unaltered if the two new junctions B-C and C-A are the
same temperature.

Fig. 11.10 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. Maximum accuracy is
obtained when the leads are of the same material as the thermocouple element [Fig. 11 11
(a)]. However this approach is not economical while using exspensive thermocouple
materials. Further, 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. Therefore, it
is preferable to employ the system shown in Fig. 11.11 (b) to keep the copper;iron and
copper-costantan junctions in the thermos flask at 0C and provide binding posts copper.
This ensures maximum accuracy in the thermocouples operation.

Mechanical Engineering

Universitas Sriwijaya

Fig. 11.11 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. Ability to withstand the temperature at which they are used,
2. Immunity from contamination /oxidation, etc. Qhich ensures maintenance of
the precise thermoelectric properties with continuous use, and
3. Linearity characteristics.
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

(11.11)

Thermocouple can be broadly classified in two categories:

1. Base-metal thermocouples, and
2. Rase-metal thermocouple.

Base-metal thermocouples use the combination of pure metals and alloys of iron,
copper and nickel and are used for temperature up to 1450 K. These are most commonly
used in practice as they are more sensitive , cheaper and have nearly linear characteristics.
Mechanical Engineering

Universitas Sriwijaya

Their chief limitation is the lower operating range because of their low melting point and
vulnerability to oxidation. On the other hand, rare metal thermocouples use a combinations
of pure metals and alloys of platinum for temperatures up to 1600C and tungsten, rhodium
and molybdenum for temperatures up to 3000C.
Typical thermocouples with their temperature ranges and other salient operating
characteristics, are given in table 11.3.
Table 11.3 Characteristics of Some Thermocouples
Approximate

Useful

Approximate

sensitivity

Temperature

accuracy

Thermocouples
S.No.

Type
material

in (V/ C

range( C)

(%)

1.

Copper-Constantan

20-60

-180 to +440

0.75

2.

Iron-Constantan

45-55

-180 to +850

0.75

3.

Chromel-Alumel

40-55

-200 to +1300

0.75

4.

Cheromel-Constantan

55-80

-180 to +850

0.5

5.

Platinum-Platinum/10% Rhodium

5-12

0 to +1400

0.25

6.

Platinum-Platinum/13% Rhodium

5-12

0 to +1600

0.25

7.

5-12

+100 to +1800

0.25

5-12

0 to +3000

0.15

Platinum/30%

Rhodium-Platinum/6%

Rhodium
Tungston/5%

8.

Rhenium-Tungston/20%

W5
Rhenium

Constantan = copper/nickel; chromel = nickel/chromium; alumel = nickel/alumunium

For special purposes where high sensitivityis needed, thermocouples may be

attached in series. The output is then the numerical sum of the voltages expected from each
of the single couples. This is commonly known as thermopile. When connected in parallel,
a group of thermocouples will give a reading that is the numerical average of the individual
thermocouple being the same.
The following are the advantage of the TC sensors:

Mechanical Engineering

Universitas Sriwijaya

1. Thermocouple bead can be made of small size and consequently with low
thermal capacity. In other words dynamic response of sensor is fairly good.
2. They cost considerably less as compared to other thermal sensors and further,
they require no maintenance.
3. They are quite rugged type, i.e. they can withstands rough handling.
4. They cover wide range of temperature, i.e. from -200 to 3000 0C.
5. Output signal is electrical and they can be used for indicating recording or
6.
7.
8.
9.

microprocessor based control system

Output signal, i.e. emf is independent of length or diameter of the wire.
They have good accuracy of the order of 0.2 to 0.75 % of f.s.d
They have excellent stability for a long period of time.
They can be conveniently mounted in a variety of temperature measurement
situations.

The TC sensors, however, have the following limitations :

1. Inhomogeneity of composition of the thermocouple material and cold working
of wires affect the sensitivity of the thermocouple.
2. The require insulation covering while using them in conducting fluids
3. The output signal, i.e. emf requires amplification in most applications.

11.6.3 Solid State Temperature Sensors

Common I.C device like silicon diodes and transistors exhibit a stable and
reproducible response to temperature. When a PN junction is forward biased by a
constant current source, its governing equation between current and voltage is as
follows:

VBE =

kT
q

ln

Ic
I es

(11.12)
where

VBE

= base emitter voltage

Ic

= collector current

Ies

= emitter saturation current

= Boltzman constant (1.38 x 10-23 J/K)

Mechanical Engineering

Universitas Sriwijaya

= electron change (1.6 x 10-19 C)

= absolute temperature (K)

Generally, the term within the parenthesis in Eq. (11.12) is constant and the
emitter base voltage i.e., the output of the transducer becomes directly
propotational to T whih is the measured input. The main avantage of the solid state
temperature sensors is heir inherent lnear operating characteristics with excellent
accuracy of the order of 1oC. In addition they have high levels of output signal
which is capable of direct indication without n signal conditioning. The sensitivity
of the silicon transistor wihin its useable range of -55 to 150 oC is of the order of
-2mV/oC. Further, since the output is electrical, they have the capability of p based
control applications.
The disadvantages of these sensors are their limited temperature measuring
range and their thermal mass which limits their response characteristics.

11.6.4 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. Herein, the crystal is cut in the
form of shear type LC cut, in which the change in resonant frequency is highly
linear as well as repeatable. The associated electronic circuitry of this thermometer
consists of frequency counters and digital read-out of the measured frequency.
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 the order words, yhe detection of change in frequency of
oscillation of 1 Hz gives a resolution of 0.001oC. Further, temperature in the range
of -40 to 230oC can be measured precisely and accurately by this method.
The advantages of the quartz thermometer are :
1. Highly linear output as the linearity error is 0.5% of F.S.

Mechanical Engineering

Universitas Sriwijaya

2. Long-term stability and reliability.

3. High resolution of the order of 0.001oC.
4. Excellent repeatability in the measuring range of -40 to 230oC.
The limitations of the quartz thermometer are :
1. Limited measuring range i.e., -40 to 230oC.
2. Piezo-electric crystals have strong cross-sensitivity for pressure changes
if they occur simultaneously in the temperature measuring systems.

11.7 RADIATION METHODS (PYROMETRY)

All the temperature measuring device discussed so far i.e. Pressure thermometer,
thermistor or thermocouple, etc. require the thermometer to be brought into physical
contact with the body whose temperature is to be measured. This means that the
thermometer must be capable of with standing this temperature. In the case of very hot
bodies, the thermometer may melt at the high temperature. Secondly, for bodies that are
moving, a non-contacting device for measuring the temperature is most convenient.
Thirdly, if the distribution of temperature over the surface of an object is required, a noncontacting device can readily 'scan' the surface.

For temperatures above 650oC, the heat radiations emitted from the body are of
sufficient intensity to be used for the measuring the temperature. Instruments that employ
radiation principles fall into three general classes : (a) total radiation pyrometer, (b)
selective (or partial) radiation pyrometers, 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. Further, 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.

Mechanical Engineering

Universitas Sriwijaya

11.7.1 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. The term 'total radiation' includes both visible (light) and invisible
(infrared) radiations. It may be noted that the wave lengths of light in visible range
is form 0.3 to 0.72 m, whereas the infrared radiations are associated with
relatively large wavelengths of 0.72 to 100 m. They require special optical
materials for focussing. Ordinary glass is unsatisfactory, as it absorbs infrared
radiations. In fact, the practical radiation pyrometers are sensitive to a limited
wavelength band of radiant energy, (i.e., from 0.32 to 4m m) although theory
indicates that rhey should be sensitive to the entire spectrum of energy radiated by
the object.

Fig. 11.12 Schematic of Ferys total radiation pyrometer

Figure 11.12 shows a schematic diagram of the Fery's total radiation
pyrometer. It consists of blackened tube T open at one end to recieve the radiations
from the object whose temperature is to be measured. 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. The detector disc is usually of blackened platinum
sheet/foil and is connected to a thermocouple/thermopile junctions or to a
Mechanical Engineering

Universitas Sriwijaya

resistance thermometer bridge circuit. 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.
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), i.e. EA/B is
given by the Stafan-Boltzmann law as follows :
EA/B = C [TA4 TA4]
were

(11.13)

EA/B

= is the energy recieved by the pyrometer in W/m-2

= is a geometrical factor to adjust the relative shapes of the two

bodies

= is the emissivity of the detector disc which varies from 0.05 to 1.0
for the theoritical black body.

= is the Stefan Boltzman constant and its value is 56.7 x 10 -12 kW/
(m2-K4).

TA and TB are the steady state absolute temoerature of the source and
pyrometer detector disc.
Such pyrometers are usually calibrated againts known temperatures in the
range of 700 - 2000oC where thermocouples and resistance thermometers cannot be
employed. However, the errors arise from two sources in actual use. Any filtering
material such as smoke, dust, gases, windows, etc. which were not present in the
calibration will reduce the energt recieved hence cause an unknown error. 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, therefore the error due unknown emissivity is
usually not known. To reduce such un certainties, pyrometers calibrated from time
to time in actual use.
In view of the troubles due to filtering and emissivity, the total radiation
pyrometer is not a very accurate temperature indicator. However, it can be used to
Mechanical Engineering

Universitas Sriwijaya

good advantage in fixed locations where the emissivity and optical paths are well
known and constant. A typical use is a large furnace in metal industries. The signal
is electrical and therefore can be used for control applications.
11.7.2 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. As the temperature increases, the emissive power shifts to shorter
wavelengths. The planck's distribution equation is :
c1

W=

where

c 2/ T

(11.14)

c1

= 3.740 x 10-12 (W-cm2)

c2

= 1.4385 (cm-0C)

= wavelength (cm)

= absolute temperature in (K)

= 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). It is most
accurate of all radiation pyrometers; however, it is limited to temperatures greater
than about 700oC since it requires visual brightness match by a human operator.
This instrument is used realise the International Practical Temperature Scale above
1063oC.
It is abvious from Planck's distribution equation that for a given
wavelength, the radiant intensity (brightness) varies with the temperature. In the

Mechanical Engineering

Universitas Sriwijaya

disappearing filament instrument shown in the Fig. 11.13, an image of the target is
superimposed on the heated filament.

Fig. 11.13 Schematic of the disappearing filament type of optical pyrometer

The tungsten lamp, which is very stable, is previously calibrated so that

when the current trough the filament is known, the brightness temperature of the
filament is also known. A red filter that passes only a narrow band of wavelengths
around 0.65 m is placed between the observer eye and the tungsten lamp and the
target image. The observer controls the lamp current until the filament disappear in
the superimposed target image [Fig. 11.14 (c)].

The temperature calibration is made in terms of the lamp heating current.

Because of the manual null balancing principle, the opticcal pyrometer is not usable
for continuous recording or automatic control application. However, it is more

Mechanical Engineering

Universitas Sriwijaya

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

Mechanical Engineering