Why it is difficult
A polished metal emits very little radiation of its own and reflects most of what it receives. The camera therefore sees mainly the reflected surroundings. The typical result is an apparent temperature close to ambient, even when the part is considerably hotter.
Strategies, from most to least reliable
- Change the surface: apply a tape or matte paint of high, known emissivity, and wait for its temperature to equalize with the metal. Measure on the coating.
- Measure by contact: a well-coupled contact sensor gives the actual temperature, and can also be used to calibrate the camera’s emissivity.
- Take advantage of different areas: oxidized, dirty, rough or marked areas of the same metal have higher emissivity.
- Use the shape: a cavity, hole or tight corner behaves like a blackbody because it traps radiation, and its apparent emissivity is high.
- Qualitative interpretation: if none of the above is possible, treat the image only as a comparison, without stating absolute values.
Complementary measures
- Determine the reflected apparent temperature carefully, because in this case it dominates the result.
- Eliminate or block reflection sources, such as lamps, hot equipment and the operator.
- Avoid very oblique angles, since emissivity falls with inclination.
- Change the angle slightly: if the spot moves, it is a reflection.
What an engineer sees in practice
New copper or aluminum bars or terminals that look colder than the paint around them, despite carrying the same current. A rusty bolt next to a shiny lug gives a difference that is not one of temperature.
Placing tape or sensors on energized parts must be done only by qualified personnel following the procedures and applicable regulations.