The key question: which temperature matters
A furnace has several temperatures: that of the air or gases, the walls, the load and the interior of the load. The control sensor usually measures a point close to the heating element or the air, and the part may be at a different temperature. Defining which one matters to the process determines where and how it is measured.
Choosing the sensor
- Thermocouples: the usual choice because of their high range and robustness. The type is chosen according to temperature, oxidizing, reducing or inert atmosphere, and the required stability.
- RTDs: used in lower ranges, where their stability and accuracy are an advantage.
- Radiation pyrometers: these measure without contact with the surface, which is useful when the sensor cannot touch the part, but they depend on emissivity and on anything that comes into the line of sight.
Installation
The sheath protects the sensor from the environment but adds thermal inertia. Insertion depth must be sufficient so that conduction along the sheath does not cool the tip. The extension cable and connectors must match the thermocouple type and be kept out of zones that exceed their limit.
What the engineer sees
It is normal for the controller to show the setpoint while the part is still colder, especially during fast ramps. Gradients inside the furnace show up as differences between sensors located at different points. A drifting thermocouple makes the furnace operate above or below the true value without the controller noticing.
How it is verified
A uniformity survey is carried out with sensors distributed across the working zone, and the control sensor is compared against a standard or against test sensors. This reveals the differences between the control reading and the temperature at the load. These checks are repeated at a frequency based on process risk and observed drift.
Work on furnaces involves thermal and electrical hazards and must be carried out by qualified personnel following the applicable safety procedures.