Selection and use of non-contact temperature sensors

There are three main challenges when selecting and using a non-contact temperature sensor:

● Filling the field of view
● Overcoming object transparency issues
● Achieving the correct emissivity adjustment

The thermal radiation from an object in the field of view "reflects" the temperature of its surface, much like a radio beacon. The difference is that in IRT measurements, the thermal radiation comes from a spot of known size on the object's surface. You need to capture all of the radiation and not have any of it blocked along the way.

The spot diameter at a specific distance (that is, the cross-sectional area of ​​the "cone of view" at a given distance) is determined by the optical properties of the device. Generally speaking, the farther away from the sensor, the larger the diameter of the spot. Spot size is often expressed as a ratio, such as 50:1 or 10:1. This means that the smallest target spot diameter is 1/50 or 1/10 of the distance from the sensor to the object of interest (Figure 2).

Figure 2. The spot diameter at a given distance is determined by the optical properties of the sensor and is defined as a ratio;
generally, the farther the object is from the sensor, the larger the spot.

If the object has a circular cross section (Figure 3), and the only field of view is along the diameter, make sure that the size of the severely curved surface being measured is a small fraction of the object being viewed. In other words, the spot diameter should be 25% or less of the effective diameter of the curved surface.

Figure 3. When measurements have to be made on curved surfaces,
make sure the spot diameter is no larger than 25% of the effective diameter of the surface.

It is also important to ensure that the viewing cone is clear, or the detector will not receive all the radiation it needs to make a valid measurement. If the position spot is larger than the object, or the viewing cone is partially blocked, the measured temperature will be lower than the true temperature.

If you cannot reliably fill the spot diameter or keep the viewing cone clear, you may want to consider using a dual-color or ratiometric IRT. There may be a slight loss of accuracy, and possibly a slow response, and a small increase in cost, but these devices are designed to work when the line-of-sight path is partially blocked, the viewing cone changes, or objects move in and out of the field of view. They are very sensitive to changes in the ratio of emissivity, but this is beyond the scope of this article.

Transparency Issues Most (but not all) organic and building materials (such as brick, wood, metal, asphalt, rock, and minerals) are opaque. However, many plastics are semi-transparent in the IR spectrum, so special bands need to be used to make them essentially opaque to the sensor. Ircon has some very useful application notes on plastic and glass measurement, which can be viewed on the Ircon website.

Other materials that can present transparency problems include semiconductor materials (silicon, gallium arsenide), certain coatings, some optical materials (such as silver chloride, sapphire, quartz, sodium chloride, germanium), and many exotic crystal products.

Emissivity Correction Don't believe anyone who tries to tell you this is a trivial problem. While it's not trivial, it becomes fairly easy to deal with once you break it down into three scenarios you're likely to encounter:

● The object of interest is at or very close to ambient temperature.
● The object is warmer than ambient.
● The object is cooler than ambient.

The object's optical properties (emissivity being one of them) come into play here. If the target is translucent, you'll probably need help; if it's opaque, you may be able to handle the emissivity correction yourself. Let's look at specific examples for each scenario.

Scenario One If the object is about the same temperature as its environment, and its surface is not specular, its surface reflectivity will compensate for the emissivity, and no emissivity correction is necessary. Applying an emissivity correction will give a temperature reading that is higher than the true temperature.

Case 2 If the object is warmer than its surroundings, an emissivity correction is needed to get an accurate temperature reading. Emissivity can be an elusive and variable optical property, but it usually only changes when something on the surface changes, such as charring, oxidation, or melting.

If the emissivity is controlled to 1.00, a "radiance temperature" can be obtained in this case. Although this number will be lower than the true temperature, it is repeatable if the object's emissivity (although unknown) has not changed very much. So, how much is "very much"? This is a good question, and one that cannot be answered here due to space limitations, but if the object has not changed visually, there is a good chance that the spectral emissivity has not changed. There is no guarantee, just a possibility. Case

3 This is a difficult situation to deal with. When the temperature of the target is lower than the ambient temperature, the simplest solution is to change the location of the measurement. For example, if the temperature of the target is lower than the temperature of the oven or furnace it is entering, do not try to measure its temperature at the entrance or inside (while it is being heated). The temperature measurement should be moved to the exit, when the object has completed its heating cycle and is leaving the oven (at this point it may be above its ambient temperature).

This situation has a "hidden" form, when the object is exposed to sunlight, or has a very high temperature (which may still be below ambient temperature) or light level. These conditions can be tested by casting a temporary shadow on the surface and aiming the IRT at this shadow.

On processing lines where cold web products such as metal, glass, plastic, etc. enter the oven or furnace, an attempt should be made to aim the IRT at the "wedge" formed between the product and the rollers it passes over (Figure 4), especially if it changes direction (i.e. has a wrap angle of 25% or more). This can be a "real situation" because the surface of the rollers is usually reflective and creates a mirror image of the object. In reality, it is surrounded by something that is similar in temperature to itself.

Figure 4. One solution to emissivity problems caused by the target being cooler than the ambient temperature (such as with webbed products) is to aim the IRT at the “wedge” formed by the roller and product.

Advice to Readers

There are many excellent non-contact temperature sensors on the market, and manufacturers are trying hard to persuade you to buy their products. It is important to remember that you cannot assume that as long as the temperature sensors are plugged in, you can expect them to do the job you assign them. They must be provided with optimal operating conditions, or corrected for those that are not optimal.