An Introduction to Emissivity
An introduction to material emissivity and how it affects measurement using an infrared thermometer or radiation thermometer.
Surfaces emit thermal radiation, at different wavelengths and in varying amounts. At any given wavelength and temperature there is a theoretical maximum amount of radiation that can be emitted. A so-called ’perfect’ body would emit the maximum possible amount of radiation at all temperatures and at all frequencies. Such a body is known as a ’blackbody’ and it would have an Emissivity of 1.0. Not surprisingly no such body actually exists.
Real surfaces then are not perfect blackbodies, but emit only a percentage of the radiation of a blackbody. The fraction that they emit is the measure of their Emissivity. For example, if at a given temperature and frequency a surface were to emit half the radiation of a blackbody it’s emissivity would be 0.5.
There are situations where bodies come very close to an emissivity of 1.0. Measuring temperatures in Industrial ovens inside very small spy holes for example can achieve an emissivity very near to 1.0. Most organic materials and dull metal surfaces have an emissivity between 0.85 and 0.98.
Emissivity also varies with frequency, sometimes surfaces emit a lot of energy at one frequency and very little at another.
Radiation Thermometers
Radiation Thermometers detect thermal radiation and convert it into an electrical signal, the signal strength being used to compute the surface temperature. Typically these thermometers use selective filters to detect a narrow range of frequencies, often in the Infrared region of the spectrum, and are commonly known as Infrared Thermometers, (they are also known as pyrometers and a wide range of other names too). For specialist applications it may be wise to choose an Infrared Thermometer with a specific spectral response. General purpose thermometers typically work in the 8 to 14 micron range, a band that avoids interference from airborne water vapour.
The question arises, how do you measure the actual temperature of something when all you know is that the emissivity is less than 1.0? If you measure assuming the emissivity to be 1.0 the reading you get will be lower than the actual temperature. Conversely if the emissivity is too low the result will be a temperature higher than the real temperature. It looks like whatever temperature reading you get is going to be wrong. And just to make things worse the emissivity of real surfaces varies with frequency, getting progressively worse at lower frequencies (longer wavelengths).
Ideally then to get an accurate reading you need to know the emissivity of the surface you want to measure. So all you have to do is get a value from one of the many published lists of emissivity? While it is true that lists do exist, and they are useful as starting point, it’s not as easy as that, because a lot of other things can, and do, mess up the readings.
Things to take into account
It may seem odd, but the colour of the surface generally makes no difference to the emissivity. Nearly all paints have the same emissivity, except for metallic paints containing large amounts of aluminium. Polished surfaces, gold, silver and aluminium in particular, are very difficult to measure accurately.
However, a wide range of other things apart from emissivity per se can affect how accurate the readings are. The following tips will give an idea as to some of the major factors that you may need to consider.
- Make sure that the surface to be measured fills the field of view of the Radiation Thermometer. Many thermometers have a laser sight that shows the extent of the area covered. This may seem obvious, but in many industrial applications it may be difficult to get close to the surface you want to measure.
- The viewing angle. Try to keep the thermometer as close to 90 degrees to the surface as possible. Even on a dull surface radiation radiation will be scattered especially at steep angles. This can be particularly difficult if the surface is curved or inaccessible. In addition some thermometers may be affected by polarisation of the light at extreme angles.
- Surface finish. Shiny or highly polished surfaces can be very tricky to measure, even dull smooth metal surfaces may not be equally emissive in all directions. Be especially careful where surfaces are highly curved, see viewing angle above. If the emissivity is very low, some polished surfaces have an emissivity of 0.2 or less, then an accurate reading is unlikely to be possible.
- Smoke, water vapour, dust. Any of these in the line of sight will reduce the amount of radiation received. It is extremely hard to determine how much the reading is out in these circumstances. It is probably best to use a more advanced type of thermometer, a so-called Ratio Radiation Thermometer.
- Thin surfaces, for example plastic sheeting, or surfaces covered in oil, can generate interference that causes emissivity to vary.
- Measuring through a window. Any losses associated with reflection or absorption by the window material will require you to set a lower emissivity than normal. Also bear in mind that the thickness of the window and the viewing angle may affect the setting.
Determining Emissivity
There are some things that you can do to help get accurate readings. The following selection of tips should help.
- Heat a sample of material to a known temperature (measured accurately with a sensor that you can trust). Measure the temperature with the Infrared Thermometer and either adjust the emissivity setting until you get the reading to match, or note the amount of the discrepancy and use that to compensate future readings.
- Masking tape or matt paint applied to a part of the surface can be used to give an emissivity of close to 1.0. Generally the colour of the tape or paint has no effect.
Conclusion
Infrared Thermometers are, in the right hands and when used with an understanding of the factors and limitations affecting their use, a useful and reliable way of measuring temperature. In some application their use is indispensable.
Many low cost Infrared Thermometers have a fixed emissivity (usually around 0.95), which is a reasonable value in the right circumstances and if you are careful when taking the measurements. If your application includes any of the specific issues outlined above you may be better going for a higher cost options that allows the emissivity range to be adjusted to take these external factors into account.
About the Author
Denis Laverty possesses more than 17 years experience in network management and communications, Denis has been involved with network management applications from the early DOS days; as product trainer, technical author and QA Director. In 2003 he co-founded OPENXTRA together with Jack Hughes and serves as its Managing Director.
Subscribe now with RSS