Temperature measurement with single-, dual- and multi-wavelength technologies

An infrared thermometer calculates temperature by using a model, or algorithm, that correlates measured infrared energy to temperature. The appropriate model for any given application depends on several factors, including the measured material and its size and emissivity properties, and the presence of any stray background infrared energy. The three available types of infrared sensor - single, dual, and multiwavelength - are all based on assumptions about these factors. Any deviation from these assumptions will result in a measurement error; therefore, the selection of the most appropriate temperature measurement model is a critical sensor selection criterion.

Temperature measurement is complicated by the fact that the amount of infrared energy emitted by a surface is proportional to the emissivity as well as the temperature of the measured target. Therefore, any temperature measurement model must effectively address the issues associated with the emissivity of the measured target if an accurate temperature value is to be obtained.

Emissivity, or the tendency of a material to emit infrared energy, is a physical property of the measured material. Emissivity is defined as the per cent of infrared energy emitted by an object as compared to the amount of infrared energy that would be emitted by a theoretically perfect emitter at that same temperature. For opaque objects, emissivity is the opposite of reflectivity (at the measured wavelength). A highly reflective material will have a low emissivity and a non-reflective material will have a high emissivity. For some materials and applications, the emissivity is a constant value, while for others it can vary widely.

Single-wavelength infrared thermometers

A single-wavelength infrared thermometer assumes that the emissivity is constant and known. In addition, the single-wavelength model assumes that the optical path is unobstructed, and that the target fills the measured area. This is a valid assumption for many applications, but any change in emissivity will result in an error. The size of the error varies with wavelength and temperature, and with the per cent change in emissivity at the measured wavelength. The lower the temperature, and the shorter the wavelength, the smaller the error for any given per cent change in emissivity.

Single-wavelength sensors are available in a wide range of wavelengths. Short wavelength single-wavelength sensors are used to minimise errors resulting from a change in emissivity or optical obstruction. These sensors are especially effective when measuring low-temperature, low-emissivity materials. Long wavelength single-wavelength sensors provide broader temperature spans and are generally lower in cost, but are significantly more sensitive to emissivity variation. Short wavelength sensors can view through common window materials, whereas longer wavelength sensors must view through more exotic window materials. In general, short wavelength sensors provide a more robust temperature reading when changes in emissivity or optical obstructions are encountered; however, longer wavelength sensors are less influenced by interference from hotter background sources.

Dual-wavelength infrared thermometers

A dual-wavelength (ratio) infrared thermometer assumes that the ratio of emissivity values at the two measured wavelengths is known and constant. For most materials, this is a valid assumption, and the ratio is unity. The dual-wavelength model tolerates emissivity variation, optical obstruction and misalignment, so long as any variation affects both measured wavelengths equally. The dual-wavelength model provides a temperature value that is heavily weighted towards the hottest temperature viewed.

Dual-wavelength sensors are available in a variety of wavelength pairs. When viewing through translucent or transparent interference sources (steam or water spray, for example) care should be taken to select a wavelength pair that is equally affected at both wavelengths. Dual-wavelength sensors are popular for a wide range of applications, including the measurement of materials where the emissivity is likely to vary (metals, for example), of small objects where alignment may prove difficult and within hostile environments where optical obstructions may occur (such as dirty lenses, scale, water spray, steam, dust and mechanical obstructions).

Multi-wavelength infrared thermometers

Single-wavelength and dual-wavelength models utilise general-purpose algorithms that are appropriate for a wide range of applications. There exist, however, a number of industrial applications for which these general-purpose algorithms are not appropriate. Unlike the general-purpose algorithms, there is no one multi-wavelength algorithm. Each algorithm is designed to account for the unique emissivity character associated with a particular application.

Multi-wavelength sensors are required for the measurement of materials for which the single- and dual-wavelength sensors are not appropriate. Some of the more popular applications include the measurement of aluminium, brass, copper, stainless steel, chrome, tin and zinc. For each of these materials, emissivity varies over time or from part to part, making single-wavelength sensors inappropriate, and emissivity varies differently at different wavelengths, making dual-wavelength sensors inappropriate.

Figure 1: Accuracy of single-, dual- and multi-wavelength sensors when measuring steel strip of varying emissivity.

The graph in Figure 1 shows the accuracy of single-, dual- and multi-wavelength sensors when measuring steel strip of varying emissivity. The single-wavelength sensor is set for an emissivity of 0.300. Note that the dual-wavelength sensor is highly accurate for emissivity values higher than 0.600. As the strip emissivity varies below about 0.500, however, only the multi-wavelength sensor provides a high degree of accuracy.

*Tom Larrick is Technical Manager for Williamson Corp., USA

Heastern Industries
www.heastern.com.au