# Incorrect IR temperature reading on stainless steel?

Using a Fluke 62 Mini Infrared Thermometer today on a stainless steel pipe with coolant I noticed that it gives incorrect temperature readings. There is a temperature sensor inside the coolant saying about 90° Celsius. I spray painted a spot on the pipe and the IR thermometer instantly gave more correct readings on the painted spot.

So why? Why is it the coolant fluid is about 90° Celsius (both the painted spot and the temperature sensor suggest that), while the unpainted clear stainless steel says about 50° Celcius? I.e. apx 55%

## 2 Answers

I would guess your meter is assuming the surface is a reasonable approximation to a black body, i.e. the emissivity is approximately one. The emissivity of polished metal can be as low as 0.1.

From the fluke manual: Emissivity Of the kinds of energy—reflected, transmitted and emitted—emanating from an object, only emitted infrared energy indicates the object’s surface temperature. Transmitted and reflected energy do not. When IR thermometers measure surface temperatures, they sense all three kinds of energy. Therefore, they have to be adjusted to read emitted energy only. The Fluke 62 Mini Infrared Thermometer has a fixed, pre-set emissivity of 0.95, which is the emissivity value for most organic materials as well as painted or oxidized surfaces. To accurately measure the surface temperature of a shiny object, cover the target surface with masking tape or flat black paint and allow enough time for the tape or paint to reach the temperature of the material underneath.

• You are suggesting that the emissivity of stainless steel is low and the conductivity is high. Sounds reasonable. A google search of "emissivity of stainless steel 316" suggest an emissivity of 0.28 @ 24 deg Celsius. Source: coleparmer.com/TechLibraryArticle/254 – wittrup Oct 17 '13 at 16:05
• @wittrup: the link I provided gives the emissivity of polished stainless steel as 0.075 and weathered stainless steel as 0.85 (they don't say exactly what weathered means). So depending on how polished your pipe is the emissivity could vary over a wide range, but could be substantially less than one. If your meter simply measures emission reducing the emissivity from one to around 0.6 would be enough to lower the apparent temperature from 363K to 323K. – John Rennie Oct 17 '13 at 16:18
• Sorry for not checking out your link before now ;) Didn't notice it. – wittrup Oct 17 '13 at 16:28

Probably your instrument is calibrated against something with an emission spectra more similar to the coolant and thus reporting a different temperature while reading the emission spectra of steel at 90° C.