Application and calibration of low-temperature radiation thermometers
Large numbers of low-temperature radiation thermometers are finding application in the food processing, storage and transport industries. These instruments suffer severely from reflection and emissivity problems, and an additional error due to the radiation emitted by the thermometer itself. Unfortunately, the thermometer manufacturers have designed the instruments to deliberately obscure some of these errors from the user.
Low-temperature radiation thermometers are often low-cost and use uncooled thermopile detectors operating over the wavelength range from 8 mm to 14 mm. Because the detectors are uncooled, the amount of radiation emitted by the detector itself is often a significant fraction of the total detected radiation, and therefore must be taken account of. For most applications, reflection errors from ambient surroundings are also an issue (see Reflection Errors in Radiation Thermometry). In order to simplify the use of these low-temperature radiation thermometers, manufacturers use a built-in algorithm that automatically corrects for the detector temperature, reflection errors, and emissivity, so long as two specific conditions hold. While this algorithm usually leads to tolerably low errors in many situatiuons, in cases where the assumed conditions don’t hold, the errors can be large , partially obscured from the user, and can cause confusion.
At MSL, we have investigated the consequences of the internal processing carried out inside these instruments. It turns out that the readings have zero error only when the temperature of the surroundings exactly matches the temperature of the detector inside the thermometer and when the instrumental emissivity is set to the actual emissivity of the target. The first condition allows the thermometer to correct for reflection errors by making an internal measurement of its own temperature. However, this is clearly only valid if the surroundings are all at ambient temperature. Measurements inside cool stores, or in environments where objects are being heated, violate this condition. The second condition complies with the ‘normal usage’ philosophy (see Reflection Errors in Radiation Thermometry). Because of this philosophy and because many materials such as paper, plastics, wood, and paint have emissivities close to 0.95 in the 8–14 mm region, many of the less expensive low-temperature radiation thermometers have their instrumental emissivity fixed at 0.95. These instruments are not suitable for measurements on other types of surface, such as metals.
In order to extend the utility of these instruments to situations where the two conditions above do not hold, we have investigated two complementary measurement strategies, and quantified the resulting residuals errors. The first strategy, which we have called the ‘detector compensation method’, requires the instrumental emissivity to be set to the emissivity of the target (the assumption of the manufacturers). When using this strategy, the error in the reading increases as the difference between the temperature of the surroundings and the temperature of the detector increases. There is also an additional error if the instrumental emissivity is set incorrectly. In the second strategy, the instrumental emissivity is set to 1.0, which makes the thermometer reading independent of the temperature of the detector. The reading has zero error when the temperature of the surroundings is the same as the target temperature, and the error increases as the difference between these two temperatures increases. We have called this method the ‘blackbody approximation method’ because the error is zero when the target combined with its surroundings appear to be a blackbody.
The blackbody approximation method is more flexible than the detector compensation method. It can be taken advantage of by measuring inside gaps or cavities between objects of similar temperature, and can be used, for example, inside cool stores and walk-in freezers, where the detector temperature is far higher than the temperature of the surroundings. The error for both methods increases as the target emissivity decreases, but does so much less rapidly for the blackbody approximation method. Fixed-emissivity instruments are, by design, restricted to the detector compensation method.
Issues relating to detector temperature, emissivity, and reflected radiation must also be considered when calibrating low-temperature radiation thermometers (see MSL Technical Guide 22). When the emissivity setting on the thermometer doesn’t match that of the calibration source, which is generally a blackbody cavity or a flat-plate calibrator, then corrections (of up to 15 °C) need to be applied to the calibration temperature before comparison with the thermometer under calibration. An example of this is a calibration at the ice point, using an ice-point blackbody cavity (see Technical Guide 2). The graph below shows the expected reading for an 8–14 mm thermometer viewing an ice-point blackbody, as a function of the instrumental emissivity setting. For a 0.95 fixed-emissivity instrument, the readings should be about 1.2 °C below the ice point. There is also a considerable detector-temperature dependence of the correction. Understanding of these corrections is crucial to the design of calibration procedure for low-temperature radiation thermometers.
Expected reading for an 8–14 mm thermometer as a function of instrumental emissivity setting and detector temperature, Td, when viewing an ice-point blackbody cavity with effective emissivity of 0.999 at an ambient temperature of 20 °C.
P Saunders, “Reflection errors for low-temperature radiation thermometers”, in Proceedings of TEMPMEKO 2001, 8th International Symposium on Temperature and Thermal Measurements in Industry and Science, edited by B Fellmuth, J Seidel, G Scholz, VDE Verlag GmbH, Berlin, 149–154, 2002.
P Saunders, “Calibration and use of low-temperature direct-reading radiation thermometers”, Measurement Science and Technology, 20, 025104, 2009.
P Saunders, MSL Technical Guide 22: “Calibration of Low-Temperature Infrared Thermometers”, June 2009.