In-situ traceability to the ITS-90 using integrated self-validating thermocouples – Trials of the INSEVA thermocouple.

Thermocouples are one of the most commonly used temperature sensors in industry, but over time they suffer from calibration drift due to exposure to a variety of factors such as high temperatures, thermal cycling, chemical attack, and transmutation. As a result, thermocouples can suffer from unknown temperature measurement errors, causing higher energy usage, increased product wastage due to incorrect control temperatures, and possible safety issues [1]. The integrated self-validating thermocouple (INSEVA) device was originally developed as part of the European Metrology Programme for Innovation and Research (EMPIR) project "Enhancing process efficiency through improved temperature measurement", EMPRESS [2]. The device is a standard alumina sheathed noble metal thermocouple with a miniaturized metal fixed-point cell in close proximity to the measurement junction. When the thermocouple is heated above the melting temperature of the fixed-point ingot, the melting curve is detected and used to validate (and therefore calibrate in situ) the thermocouple's temperature-electromotive force (emf) relationship at the fixed-point temperature, in the same way that full sized metal fixed-point cells are used to calibrate thermocouples in calibration laboratories. The aim of this investigation was to assess the performance of the INSEVA thermocouples in a real industrial heat treatment process, building on previous work [3]. We report on the industrial tests performed using five Type S INSEVA thermocouples in a vacuum furnace, whilst heat treatment processes were being carried out. Of the five INSEVA thermocouple fixed-point cells tested, two contained silver (961.78 °C), and the other three contained gold (1064.18 °C), copper (1084.62 °C), and an iron-carbon eutectic alloy (∼1153 °C) respectively. Regular Type S thermocouples were also used alongside the INSEVA thermocouples. In this paper, we report the results of tests when thermally cycling the thermocouples regularly up to 1000 °C, and occasionally up to 1200 °C. Additionally, the thermocouple robustness was assessed when the furnace was routinely quenched with nitrogen at an ambient pressure of up to 10 bar. We also discuss the mechanism of extracting the melting curve values from the data, and the steps required to fully automate the process of thermocouple self-validation for end users. [ABSTRACT FROM AUTHOR]