A Comparison of the ITS-90 Among NPL, NIM, and CEM, Above the Silver Point Using High-Temperature Fixed Points

A prototype comparison of the ITS-90, as realized by NPL, NIM, and CEM, using high-temperature fixed points (HTFPs) of Co-C (1324 °C), Pt-C (1738 °C), and Re-C (2474 °C), is reported. The local realizations of ITS-90 temperatures were assigned by NPL, NIM, and CEM to their own set of HTFPs. NIM and CEM then transported their cells to NPL, and the ITS-90 temperatures of all three sets of cells were measured using a linear pyrometer. From these measurements, a comparison reference value (CRV) was derived. At the Co-C and Pt-C points, the deviation from the CRV was <0.1 °C for all three institutes; at the Re-C point, the deviation was <0.4 °C. These deviations are significantly less than the scale realization uncertainties ascribed by the individual institutes indicating that these uncertainty estimates are conservative and could be revised to smaller values. In addition, thermodynamic temperatures were determined for these HTFPs using the current value of the thermodynamic temperature for the copper point, namely, 1357.82 K. Given the consistent performance of the HTFPs, they should be seriously considered as scale comparison artifacts of choice when comparing primary realizations of the ITS-90 and of the thermodynamic temperature.     

Comparison of Pyrometric Co-C and Re-C Eutectic-Point Cells Between VNIIM and LNE-Cnam

VNIIM and LNE-Cnam have collaborated for several years in the field of metal-carbon eutectic points. The first action was the construction of a Pt-C cell at VNIIM using the LNE-Cnam technique and cell design. The Pt-C cells constructed in each of the laboratories were studied and compared in the past. The two laboratories have followed their collaboration work by studying and comparing Co-C and Re-C cells. Different designs and filling techniques were applied. The melting and freezing temperatures observed on the Re-C cells from the two laboratories were measured at VNIIM. The Re-C and Co-C cells were compared at LNE-Cnam in the high-temperature blackbody furnace HTBB 3200pg which was thermally optimized before the measurements. The results of the comparison showed that the Co-C cells were comparable at the level of 0.03 K while the Re-C cells showed a large difference of melting temperatures of about 0.7 K. In this article, the cells used and the methodology of the comparison will be described. The temperature differences that were obtained at the highest temperature will be examined to propose an explanation for this temperature difference.     

Manufacturing of MC Eutectics and Reproducibility of Pt-C Eutectic Fixed Points using a Thermogauge Furnace

National Institute of Metrology (NIM) (China) and National Physical Laboratory (NPL) (UK) have collaborated to construct metal-carbon eutectic alloy fixed points at NPL. A modified NPL Thermogauge furnace was vertically used to construct fixed points of Pd–C, Pt–C, Ru–C, and Ir–C. Breakage of Pd–C and Ru–C cells was traced to changes in furnace temperature gradients resulting from changing from horizontal to vertical operation. Subsequently, it was found that positioning the cell being filled so that the metal melting always starts from the top and freezing from the bottom to solve this problem. The constructed Pt–C cell was then compared to a Pt–C fixed point previously constructed by NIM. The results indicate that the two cells made independently agreed to be better than 40 mK.     

Melting Temperature of High-Temperature Fixed Points for Thermocouple Calibrations

Thermocouples can be calibrated at pure metal ingot-based fixed points at temperatures up to the freezing point of copper (1084.62 °C). For Pt/Pd thermocouples, the deviation from the accepted reference function very often takes an approximately linear form up to the copper fixed point. The calibration of Pt/Pd thermocouples may therefore be more amenable to extrapolation than that of Pt/Pt-Rh thermocouples. Here, the melting temperatures of a Co?CC and a Pd?CC eutectic fixed point are determined by extrapolating the deviation functions of several Pt/Pd thermocouples, after the fashion of Edler et al. The results are compared with the melting temperatures measured using non-contact radiation thermometry. The expanded uncertainty (k = 2) of the melting temperatures determined by extrapolation of the Pt/Pd thermocouple calibrations is ±0.32 °C for the Co?CC fixed point, and ±0.49 °C for the Pd?CC fixed point. For both fixed points, these uncertainties are comparable to those of non-contact radiation thermometry measurements. While a number of assumptions are made in performing the extrapolation of the calibrations, the method does appear to offer a useful complement to non-contact radiation thermometry measurements.     

Calibration of the Irradiance Responsivity of a Filter Radiometer for T Measurement at NIM

At the National Institute of Metrology of China (NIM), silicon photodiode-based narrow-band interference filter radiometers (FRs) have been designed for the radiometric determination of the thermodynamic temperature. The FR calibrations were performed on a new spectral comparator with a trap detector which was calibrated against the cryogenic radiometer at several discrete laser lines. The new spectral comparator is constructed from two grating monochromators assembled to give lower stray light and higher transmitted flux. Applying a transmittance measurement of the filter in the out-of-band region and careful control of the temperature, the irradiance responsivity of a 633 nm centered FR has been obtained over a dynamic range of nearly eight decades in the wavelength range from 450 nm to 1200 nm. The relative standard uncertainty of the responsivity is also analyzed and estimated to be less than 7 × 10^?4 at the 1 ?? level.     

Realization of Low-Temperature Fixed Points of the ITS-90 at NMIJ/AIST

A new model of sealed cells with three thermometer wells for calibration of capsule-type thermometers at low-temperature fixed points of the International Temperature Scale of 1990 has been developed at the National Metrology Institute of Japan (NMIJ). The melting curves of Ar and O₂ obtained using the new cells show very flat plateaux and a linear temperature dependence as a function of the inverse liquid fraction (1/F) over the range 1/F = 1 to 1/F = 20 with a narrow melting curve width of 0.1 mK. The melting curves of Ne obtained with the new cell also show very flat plateaux and approximately linear temperature dependence versus 1/F and a narrow melting curve width of 0.1 mK, though with a slight concave structure at high 1/F. The melting temperatures with the new cells agree with previous NMIJ sealed cells within 10 μK, which is similar to the reproducibility of the realization of the triple points at NMIJ. The source dependence of the triple-point temperature of Ne was investigated by filling two of the new cells from different sources of Ne. The difference in the realized triple point temperatures between the two sources is 0.031 mK, consistent with that estimated from isotope analysis. The uncertainties in the calibration of standard platinum resistance thermometers at the low-temperature fixed points are summarized. The uncertainty of the calibration at the triple point of e-H₂ has been reduced to about one-third of its value without the correction by making the isotopic correction on the basis of the technical annex for the ITS-90 in the mise en pratique for the definition of the kelvin.     

Realization of ITS-90 Above the Silver Point at the NIM

The high-temperature primary standard system was gradually improved at the National Institute of Metrology (NIM) in China after 2004. A new primary standard pyrometer (PSP) was developed, one with a size-of-source effect of 1 × 10^?4, and regional thermostats for interference filters, photoelectric detectors, and I/V converters. The relative spectral responsivity of the entire PSP was calibrated by means of a new facility. A new LED-based measurement facility and novel systematic error correction model were utilized to characterize the PSP nonlinearity and extend the photocurrent to PSP temperature readings of about 2680 °C. As an improved scheme, the fixed-point blackbody pyrometer assembly was utilized to realize and disseminate the International Temperature Scale of 1990 above the silver point. This scheme can avoid the influences of instability and inhomogeneity of tungsten strip lamps and corrects pyrometer drifts, thereby improving the realization uncertainty and simplifying the transfer chain. The expanded uncertainties of the scale realization ranged from 0.04 °C at the silver point to 0.48 °C at 2474 °C.     

Optimizing the Implementation of High-Temperature Fixed Points through the Use of Thermal Modeling

High-temperature fixed points (HTFP) have the potential to make a step-change improvement in high-temperature metrology, significantly reducing the uncertainty of scale realization of the current ITS-90 and improving dissemination of high-temperature scales to industry. However, in a practical implementation, the performance of HTFP could be limited, by, for example, injudicious use of insulation in the vicinity of the fixed point, furnace gradients, or incomplete filling. This article investigates some of these aspects for a selection of HTFP. Steady-state modeling of the influence of insulation on the radiance temperature was performed for Co–C (1,324°C), Pd–C (1,492°C), Pt–C (1,738°C), Ru–C (1,953°C), and Re–C (2,474°C) fixed points. This included studying mitigation scenarios through the insertion of different types and designs of insulation. The optimum design was identified to minimize the temperature drop in a particular furnace. It was found that, for the furnace and fixed-point combination modeled, the actual effect of the insulation was almost insignificant. Transient modeling was performed for a Re–C fixed point, to track the evolution of the radiance temperature through the melting transition. The starting point of the model was the beginning of the melt. The evolution of radiance temperature with time in “perfectly” filled cells was modeled with a range of linear temperature gradients across the eutectic cell. The gradient had a significant effect on the duration of the transition and on the structure of the melt itself. Despite the model’s simplicity, it qualitatively demonstrated that the melt transition temperature, as identified by the point of inflection, could be significantly affected by the presence of furnace gradients.     

Influence of the Opening of a Blackbody Cavity Measured at the Ag and Cu ITS-90 Fixed Points

The International Temperature Scale of 1990 blackbody fixed points are commonly composed of a graphite crucible containing a pure metal enclosing a radiating blackbody cavity. The shape of the cavity is determined to behave as much as possible as a perfect blackbody; however, the opening from which the radiance is measured induces radiative losses. The measured temperature is therefore underestimated by a few tens of millikelvins at \(1000\,^\circ \) C, compared to that of a perfect blackbody. The difference is due, on the one hand, to the drop of emissivity caused by the opening, and on the other hand, to the temperature drop between the solid/liquid interface and the inner wall of the cavity, observed by the radiation thermometer. The temperature drop is generally estimated by modeling the emissivity and the temperature difference across the cavity wall. This approach is relevant as long as the temperature distribution along the cavity and the graphite properties are known, but in many cases, the lack of data does not allow precise determination of the corrections. The corrections for the temperature drop and emissivity drop, which both depend on the cavity opening, can be determined experimentally with a low uncertainty by measuring the temperature of a fixed point for different cavity openings. To be significant, the measurement requires a source stable within a few millikelvins. In this study, this constraint has been solved by changing the cavity opening during the phase transition of the fixed point, with a rotating wheel supporting apertures of different dimensions. Measurements have been performed at the Ag and Cu fixed points during the freezing plateaus. Experimental results are presented and compared to those obtained by modeling.     

Calculation of the Temperature Drop for High-Temperature Fixed Points for Different Furnace Conditions

High-temperature fixed points (HTFPs) based on eutectic and peritectic reactions of metals and carbon are likely to become, in the near term, reference standards at high temperatures. Typically for radiation thermometry applications, these HTFPs are generally formed of a graphite crucible, with a reentrant well, an included 120?? cone, and a nominal aperture of 3?mm. It is important to quantify the temperature drop at the back wall of the cavity, and to understand the influence of the crucible configuration and furnace conditions on this drop. In order to study these influences, three different situations have been modeled by means of the finite volume method for numerical analysis. The first investigates the influence of the furnace temperature profile on the temperature drop by simulating four different furnace conditions. The other two study variations in the crucible configuration, namely, the thickness of the graphite back wall and the length of the blackbody tube.     

Development of High-Temperature Fixed Points of Unknown Temperature Suitable for Key Comparisons

During the last key comparison of local realizations of the International Temperature Scale of 1990 above the silver point, which used high stability tungsten strip lamps, it became clear that these artifacts can no longer be used to evaluate the real calibration and measuring capabilities (CMCs) of the participant laboratories. The intrinsic uncertainty of the lamps is actually larger than the claimed CMCs of most national laboratories. Ideally a set of driftless robust artifacts, preferably of unknown temperature, should be used for this purpose, as this would allow CMCs to be probed at the highest level. Currently such artifacts do not exist. High-temperature fixed points (HTFPs) have been the subject of intense study for more than 10 years. The research has come to an advanced state so much that the temperatures of some of them are well known to be within 1 K. This has rendered their use as comparison artifacts questionable as any comparison would not be blind. To address this issue, doped HTFPs have been developed which have had their transition temperature altered from that of the eutectic composition. Two Ni–C–Cu cells and two Ni–C–Sn were constructed by Inmetro with different quantities of Cu and Sn, respectively. These were compared to a reference Ni–C cell (nominal transition temperature of 1329  \(^{\circ }\) C) and the temperature differences from the pure state determined. In this paper the design, construction, and results of long-term stability are described. These promising results indicate that it is possible to make HTFPs with altered temperatures which are stable enough to serve as comparison artifacts.     

Thermodynamic Temperature of High-Temperature Fixed Points Traceable to Blackbody Radiation and Synchrotron Radiation

Absolute spectral radiometry is currently the only established primary thermometric method for the temperature range above 1300 K. Up to now, the ongoing improvements of high-temperature fixed points and their formal implementation into an improved temperature scale with the mise en pratique for the definition of the kelvin, rely solely on single-wavelength absolute radiometry traceable to the cryogenic radiometer. Two alternative primary thermometric methods, yielding comparable or possibly even smaller uncertainties, have been proposed in the literature. They use ratios of irradiances to determine the thermodynamic temperature traceable to blackbody radiation and synchrotron radiation. At PTB, a project has been established in cooperation with VNIIOFI to use, for the first time, all three methods simultaneously for the determination of the phase transition temperatures of high-temperature fixed points. For this, a dedicated four-wavelengths ratio filter radiometer was developed. With all three thermometric methods performed independently and in parallel, we aim to compare the potential and practical limitations of all three methods, disclose possibly undetected systematic effects of each method and thereby confirm or improve the previous measurements traceable to the cryogenic radiometer. This will give further and independent confidence in the thermodynamic temperature determination of the high-temperature fixed point’s phase transitions.     

Thermodynamic Temperatures of High-Temperature Fixed Points: Uncertainties Due to Temperature Drop and Emissivity

This study forms part of the European Metrology Research Programme project implementing the New Kelvin to assign thermodynamic temperatures to a selected set of high-temperature fixed points (HTFPs), Cu, Co–C, Pt–C, and Re–C. A realistic thermal model of these HTFPs, developed in finite volume software ANSYS FLUENT, was constructed to quantify the uncertainty associated with the temperature drop across the back wall of the cell. In addition, the widely applied software package, STEEP3 was used to investigate the influence of cell emissivity. The temperature drop, \(\Delta T\) , relates to the temperature difference due to the net loss of heat from the aperture of the cavity between the back wall of the cavity, viewed by the thermometer, defining the radiance temperature, and the solid–liquid interface of the alloy, defining the transition temperature of the HTFP. The actual value of \(\Delta T\) can be used either as a correction (with associated uncertainty) to thermodynamic temperature evaluations of HTFPs, or as an uncertainty contribution to the overall estimated uncertainty. In addition, the effect of a range of furnace temperature profiles on the temperature drop was calculated and found to be negligible for Cu, Co–C, and Pt–C and small only for Re–C. The effective isothermal emissivity \((\varepsilon _{\mathrm{eff}})\) is calculated over the wavelength range from 450 nm to 850 nm for different assumed values of surface emissivity. Even when furnace temperature profiles are taken into account, the estimated emissivities change only slightly from the effective isothermal emissivity of the bare cell. These emissivity calculations are used to estimate the uncertainty in the temperature assignment due to the uncertainty in the emissivity of the blackbody.     

A Phase-Field Solidification Model of Almost Pure ITS-90 Fixed Points

A two-dimensional axisymmetric phase-field model of thermo-solutal solidification in freezing-point cells used for calibrating standard platinum resistance thermometers for realization and dissemination of the International Temperature Scale of 1990 is presented. The cell is essentially a graphite crucible containing an ingot of very pure metal (of order 99.9999 %). A graphite tube is inserted along the axis of the ingot to enable immersion of the thermometer in the metal. In this study, the metal is tin (freezing temperature of \(231.928\,^{\circ }\hbox {C}\) ). During the freezing of these cells, a steady, reproducible temperature is realized, with a defined temperature that can be used to calibrate thermometers with uncertainties \({<}1\)  mK. The model is applied to understand the effect of experimental parameters, such as initiation technique and furnace homogeneity, on the measured freezing curve. Results show that freezing curves whose behavior is consistent with the Scheil theory of solidification can be obtained with a specific furnace temperature profile, and provided that the freeze is of a long duration, the results are consistent with previous one-dimensional models and experiments. Morphological instability is observed with the inner interface initiation technique, causing the interface to adopt a cellular structure. This elevates the measured temperature, in accordance with the Gibbs–Thomson effect. In addition, the influence of initiation techniques on the solidification behavior is examined. The model indicates that an initially smooth inner mantle can ‘de-wet’ from the thermometer well-forming agglomerated solid droplets, following recalescence, under certain conditions. This manifests as a measured temperature depression due to the Gibbs–Thomson effect, with a magnitude of \(100\, {\upmu }\hbox {K}\) to \(200\,{\upmu }\hbox {K}\) in simulations. The temperature rises to that of the stable outer mantle as freezing progresses and the droplets re-melt. It is demonstrated that the effect occurs below a critical mantle thickness. A physical explanation for the origin of the effect is offered showing that it is consistent with solid-state de-wetting phenomena. Consideration is also given to the limitations of the current model configuration.     

Methods for the Realization of its-90 Fixed Points: Necessity of Improvement

Measurement Techniques - We consider the problems of consistency of the results of international comparisons, equivalence of the national standards, and improvement of the methods aimed at the...     

High-Purity Fixed Points of the ITS-90 with Traceable Analysis of Impurity Contents

The International Temperature Scale of 1990 (ITS-90) is based on thermodynamic equilibrium states of ideally pure substances. The largest contribution to the uncertainty budgets of most metallic fixed points is the influence of impurities on the fixed-point temperature. Therefore, a traceable chemical analysis of the remaining impurities with small uncertainty is the basis of further progress. Further requirements are better knowledge of the phase diagrams at very low impurity contents, impurity segregation, and the quantification and correction of thermal effects during a fixed-point realization. In this article, current and future activities at PTB and BAM in order to develop improved metallic fixed-point cells of the ITS-90 are reviewed.     

Temperature at the Reference Points of Scale MTSh-90

It is shown that the lack of a clear definition of one of the main concepts underlying temperature scales – the reference-point temperature – is causing uncertainty in the realization of the temperatures corresponding to the reference points on scale MTSh-90. An analysis is made of the advantages and disadvantages of different approaches to determining the temperature at reference points. It is proposed that this temperature be the temperature at which a given substance of a specified degree of purity begins to crystallize. That temperature corresponds to thermodynamic equilibrium of the liquid phase having the specified initial concentration of impurities.     

高温温标复现与传递技术的改进

为了改善高温温标复现,研制了新的基准光电高温计和光谱响应度测量装置.采用基于参考点-高温计的亮度比测量的温标量值确定方式,将显著减小温标复现的不确定度.利用单固定点分度方法和Sakuma-Hattori方程的多固定点内插方法校准标准光电高温计,可大幅度减少温标传递环节,提高校准不确定度水平.     

The Comparison of MTDATA with the Melting/Freezing Point Curves of ITS-90 Metal Fixed Points

The International Temperature Scale (ITS-90) is defined in part by a series of metal freezing points between 156 °C and 1,084 °C. These freezing-point cells provide reference temperatures with an uncertainty of realization claimed to be in the range of several tenths of a millikelvin. The impurities in the nominally 99.9999 % pure metals make a major contribution to the uncertainty of realization of the fixed-point temperature. Recently, a new method to correct for the influence of the impurities by summing the individual contributions of each impurity has been suggested. This method is referred to as the “sum of individual estimates” (SIE). NPL is a partner in a Euromet project to improve the realization of ITS-90 metal fixed points. As part of this project, NPL is particularly interested in the fixed-point cells of tin (231.928 °C) and aluminum (660.323 °C). This article describes the use of a thermodynamic model, embodied in NPL’s MTDATA software, to estimate the initial drop in the freezing temperature and the temperature decrease during freezing using both equilibrium and “Scheil” approaches. Calculations of this type establish the effect of single elements, and with the chemical analysis of the metal, enable an estimation of the whole freezing curve. This has been done for a sample of tin from this laboratory, and thereafter, the theoretical curves are compared with previously published experimental data on impurity-doped aluminum, with good agreement, e.g., better than 1 mK over most of the curve for 76 μg · g⁻¹ Ag in Al.