Effect of Annealing on Drift in Type S Thermocouples Below $$900\, ^{\circ }\hbox {C}$$ 900 ∘ C

Abstract

Type R and Type S platinum/platinum–rhodium thermocouples are amongst the most widely used high-temperature thermocouples, both for process measurement and as reference thermocouples. To achieve the lowest practical uncertainties, below $$1\, ^{\circ }\hbox {C}$$ , the thermocouples must be in a well-defined thermoelectric state. There are two annealing procedures in common use that leave the thermocouples in different states, leading to a potential ambiguity and uncertainty. This paper reports on experiments with Type S thermocouples clearly exposing the different drift characteristics for the two different annealed states. Thermocouples used above $$800\, ^{\circ }\hbox {C}$$ show the least drift when annealed at $$1100\, ^{\circ }\hbox {C}$$ and then passively quenched to room temperature. If used at lower temperatures, they exhibit drift, at temperatures as low as $$170\, ^{\circ }\hbox {C}$$ , with the drift peaking at $$0.3\, ^{\circ }\hbox {C}$$ to $$0.4\, ^{\circ }\hbox {C}$$ around $$300\, ^{\circ }\hbox {C}$$ to $$600 \, ^{\circ }\hbox {C}$$ . Thermocouples used below $$800\, ^{\circ }\hbox {C}$$ are best annealed at $$1100\, ^{\circ }\hbox {C}$$ , and then again at $$450\, ^{\circ }\hbox {C}$$ . In this state, they exhibit practically zero drift for temperatures up to about $$600 \, ^{\circ }\hbox {C}$$ . Advice on calibration procedures to minimise the effects of drift is given.