Thermocouples

Back

Thermocouples are the most widely used temperature sensors in industry due to their low cost, simplicity, robustness, size and useable temperature range.

A thermocouple consists of two wires made of dissimilar metals and welded at one end. This junction is called the hot or measuring junction. The other junction is called the cold or reference junction and is connected to the output device (voltmeter, temperature indicator).

These wire pairs or thermoelements can be composed of noble metals - such as platinum or rhodium - or base metals, such as copper, iron or nickel-copper alloy.

An electromotive force (emf) of typically a few millivolts is generated by thermal gradients along the length of the exposed conductors.

  The thermocouple emf is a function of the difference in temperature between the measurement junction and the reference junction.
 
The reference or cold junction is maintained at constant temperature, usually at 0C in an ice bath. Instead of an ice-bath an electronic reference or compensation can be used.

Intermediate extension or compensating cables are often used between the thermocouple and the measuring output device. Extension cables use the actual thermocouple materials. Compensating cables use completely different alloys that happen to exhibit very similar thermoelectric properties to the thermocouple

The electromotive force E is a function of the temperature gradient as follows:

where S(T, x) is the Seebeck coefficient. For homogeneous thermoelements, S(T,x) = S(T).

The correspondence between emf and temperature is available for the most commonly used thermocouples in reference tables (for thermocouple types R, S, B, J, T, E, K, N).

The standardised thermocouples are identified by a letter. The positive element of the thermocouple is always quoted first. Written standards for these thermocouples give (i) temperature versus emf in a given temperature range, (ii) compensating cables, connectors and associated standard colours and (iii) tolerances.

   Both conductors are separated by an electrical insulator and often inserted in a metallic or non-metallic protecting tube

Medium temperature thermocouples

Type +ve wire -ve wire Sensitivity(V/C) Range of use (C) Forbidden atmosphere Notes for users
E Ni-Cr Cu-Ni 78 -270 to 870 reducing, vacuum Has the highest sensitivity of common thermocouples; well suited for low temperatures but, like type K, can be unstable and have hysteresis between 300C and 600C
J Fe Cu-Ni 55 -40 to 800 - Very sensitive but becomes unstable at temperatures above 400C
K Ni-Cr Ni-Alumel 41 -270 to 1270 reducing, vacuum The most commonly used thermocouple; high sensitivity, easy to use, cheap but unstable and can exhibit hysteresis in the range 300C - 600C
N Ni-Cr-Si Ni-Si 38 -270 to 1300 reducing Similar to type K but more stable oxidation resistant at high temperatures; cheaper than noble metal thermocouples
T Cu Cu-Ni 51 -200 to 370 - Has good long-term stability; accurate in a very restricted temperature range; often used in environmental and food applications. Its use in air is restricted to 370C

High temperature thermocouples

Thermocouples for use at high temperatures have been developed, particularly for nuclear and space technology applications, for use to 2000 C and more. High temperature thermocouples are usually made of noble metals (platinum, rhodium, iridium) and their alloys, and refractory metals with very high melting points - mainly tungsten, rhenium, molybdenum, niobium, tantalum, and their alloys.

Type +ve wire -ve wire Sensitivity(V/C) Range of use (C) Forbidden atmosphere Notes for users
B Pt30Rh Pt 9 50 to 1700 reducing
- stable in air or oxidising atmospheres
- can sustain short periods in vacuum
- sensitivity poorer than base metal thermocouples
- due to a local minimum in its thermoelectric emf, type B gives the same output at 0C and 42C, hence inappropriate below 50C
- low sensitivity and high cost makes it unsuitable for general purpose
- very stable, hence used as standard thermocouples in temperature calibration
R Pt13Rh Pt 12 0 to 1600 reducing
- stable in air or oxidising atmospheres
- can sustain short periods in vacuum
- sensitivity poorer than base metal thermocouples
- low sensitivity and high cost makes it unsuitable for general purpose
- very stable, hence used as standard thermocouples in temperature calibration
S Pt10Rh Pt 10 0 to 1600 reducing
- stable in air or oxidising atmospheres
- can sustain short periods in vacuum
- sensitivity poorer than base metal thermocouples
- low sensitivity and high cost makes it unsuitable for general purpose
- very stable, hence used as standard thermocouples in temperature calibration
W Tungsten W26Re 20 0 to 2600 oxidising
- can be used up to 2700 C
- a reference table of temperature v emf is available (ASTM E 988-96 standard)
W3 W3Re W25Re 20 0 to 2600 oxidising
- can be used up to 2700 C
- a reference table of temperature v emf is available (ASTM E 988-96 standard)
W5 W5Re W26Re 20 0 to 2600 oxidising
- can be used up to 2700 C
- Ir40Re Iridium 5 0 to 2100 reducing
- one of the few thermocouples that can be used up to 2100 C in air
- suitable for continuous use in vacuum or in inert atmosphere
- iridium is rare so this thermocouple is very expensive
- its thermoelectrical voltage is low
- Platinum Palladium 20 0 to 1500 reducing
- this thermocouple has recently gained much interest in National Metrology Institutes for its very high accuracy
- main use between 660 C and up to 1500 C
- Mo Niobium 12 400 to 1700 oxidising
- mainly used in the nuclear industry because of its resistance to neutron bombardment must not be used in an oxidising atmosphere

Thermocouple advantages and disadvantages

  • Thermocouples allow measurement of temperatures higher than that possible with resistance devices (RTDs) like the platinum resistance thermometer. Their operating range is far wider: compare -200 to 650C for platinum probes with -200 to more than 2000 C with refractory thermocouples.
  • On the other hand, thermocouples are less accurate and less stable than RTDs and more subject to drift over time. They need to be checked and calibrated frequently.
  • Inhomogeneities in thermocouple wires induce parasitic emf in regions where temperature gradients exist. This leads to temperature measurement errors which are a major limitation in using thermocouples for highest accuracy temperature measurement.

References

  1. T D McGee (editor), Principles and methods of temperature measurement, John Wiley & Sons, ISBN 0 471-62767-4
  2. J Scholz and T Ricolfi (editors), Thermal sensors, VCH, vol 4, ISBN 3-527-26770-0
  3. G Bonnier, E Devin (eds), Couples thermomelectriques - Caractristiques et Mesures, Technique de l'Ingnieur R 2 590[4]
  4. A Tong, Improving the accuracy of temperature measurements, Sensor Review, vol 21, n3, 2001
  5. P R N Childs, J R Greenwood, C A Long, Review of temperature measurement, Review of scientific instruments, vol 71, n8 2000
  6.  J Yoder, Making contact with temperature, Flow Research, 2000
  7.  Manual on the use of thermocouples in temperature measurement - ASTM STP 470B


    Back