Development of high temperature fixed points for application in radiometry and thermometry

Abstract

Metrology is the science of measurement. It involves the formulation of the theoretical principles related to the definition of the International System of Units (Système International d Unités, SI) and the development of the technology for realising these units in practise. Reliable measurement is important all areas of human activity. Decisions are based on measurement results. Incorrect measurement results can lead to incorrect decisions, which may have significant health, economic or environmental consequences. Applications of sources of optical radiation are wide-ranging, including: medical treatment, diagnosis and sterilisation; quality control in production processes; stimulation of plant growth in agriculture; lighting and signalling for transport; interior and exterior lighting for buildings; solar power generation; and earth observation. The spectral power distribution of a source must often be known in order to ensure that the selected radiation source (and detector) is suitable for the intended purpose and to optimise process efficiency. Sometimes spectral measurements are required to demonstrate compliance with regulatory requirements (such as Occupational Health and Safety, Aviation Authority regulations, Emergency Lighting and Road Lighting regulations). Spectral measurements in the ultraviolet region are especially important due to its associated biological hazard. The accuracy of a measurement result can only be known if it is traceable to a national measurement standard. The National Metrology Institute of South Africa (NMISA) realises and maintains the national measurement standards for South Africa. Sources of optical radiation can be calibrated in terms of spectral irradiance over the wavelength range 250 nm to 1 300 nm. The NMISA currently imports traceability for spectral irradiance by sending primary standard lamps to another National Metrology Institute (NMI) for calibration. The uncertainties of calibration are 4 % to 8 % (k = 2) over the wavelength range of 400 nm to 1 500 nm. A primary spectral irradiance facility is being developed, which will allow the NMISA to realise the scale independently, eliminating the need to import traceability. Through developing this facility, the NMISA aims to achieve an uncertainty of < 1 % over the visible wavelengths. Many NMIs realise the spectral irradiance scale by obtaining traceability from a primary standard cryogenic radiometer through calibrated filter radiometers. The filter radiometers are used to determine the temperature of a high temperature black body functioning as a reference source, which spectral radiance can be determined from Planck s equation. The uncertainty of the temperature measurement makes the most significant contribution to the uncertainty of realising the spectral irradiance scale. High temperature fixed points, above the copper point, can be used to improve these uncertainties. After more than ten years of research, results obtained on metal-carbon eutectic fixed points by several NMIs, showed that these novel high temperature fixed points could lead to significant improvements in high temperature metrology and could be considered as potential fixed points in a future International Temperature Scale. This dissertation describes the development and characterisation of high temperature metalcarbon fixed points at NMISA. It is demonstrated that these fixed points can be utilised as reproducible, stable reference standards for temperatures above the copper point. The melt temperature of Re-C cells was repeatable within 60 mK, which is equal to a relative spectral radiance value of 0,02 % at 650 nm. The melt temperature of the d(MoC)-C cell was repeatable within 100 mK, which is equal to a relative spectral radiance value of 0,03 % at 650 nm. Without implementing eutectics, NMIs typically achieve a best measurement capability of 3 K (k = 2) at 2 800 K, which contributes approximately 1 % to the overall spectral radiance measurement uncertainty. By using eutectics the reproducibility of spectroradiometric scales can be improved by a factor of 10. The NMISA result for Re-C (2 747,51 K ± 2,43 K, (k = 2)) is consistent with international values and agrees with the preliminary consensus value (2 747,35 K ± 1 K, (k = 2)) within the stated uncertainties. Only one published result for d(MoC)-C eutectics could be found, which was by NMIJ (Japanese NMI). The NMISA result for d(MoC)-C (2 856,76 K ± 2,89 K, (k = 2)) corresponds with that published by NMIJ (2 856 K ± 4 K, (k = 2)) and compares well with the result of measurements performed at VNIIOFI (Russian NMI) (2 856,40 K ± 2,0 K, (k = 2)) on this particular cell. As far as can be ascertained, the most comprehensive study of d(MoC)-C was done by the NMISA. Once internationally agreed to melt temperatures for a selected set of high temperature eutectics are approved and incorporated into an updated mise en pratique for the definition of the kelvin, the NMISA could assign the international consensus value to its Re-C cells and re-realise its high temperature scale from 961,78 ºC (the silver freeze point) up to 2 474 ºC (or higher) and immediately, in retrospect, realise the substantially reduced uncertainty of 1 % (k = 2).