Measuring surface temperature is part of our daily lives, yet the results of these measurements are often criticized as being unrepresentative of the temperature we're looking for. For example, let's take our hand as a temperature sensor. By applying it to a person's forehead, we can make an initial diagnosis to check for the presence of fever. This first impression is often reliable, even when quantified by a thermometer. The hand used to assess surface temperature, like an industrial sensor, can give a false impression of wall temperature. For example, touching a staircase banister heated by the sun will give a sensation of gentle warmth if it's made of wood, or of burning if it's made of steel! This assessment bias depends on the thermal characteristics (conductivity, effusivity...) of the surface, which will exchange more or less heat. Just like the hand, industrial temperature sensors can deliver false results depending on the quality of the sensor, the thermal characteristics of the substrate, the environmental conditions and the way they are used.
The aim of this article is to present good practice in the use of these sensors, in order to assess surface temperature in the best possible way. Physical principles, parasitic effects and the impact of the environment on these particular temperature measurements are reviewed.
The two main applications of surface sensors are clearly defined, to enable the reader to use the most appropriate measurement technique. The article is illustrated by technological examples.
For the correct installation of surface temperature sensors, it is important to :
identify the need (surface measurement or other);
know about disruptive effects: macroconstriction, contact resistance, fin effect, thermal effusivity... ;
use the appropriate sensors ;
control the environment ;
use appropriate methods.