Extrapolation of radiation thermometry scales for determining the transition temperature of metal–carbon points. Experiments with the Co–C

Abstract

Four independent radiation temperature scales approximating the ITS-90 at 900 nm, 950 nm and 1.6 µm have been realized from the indium point (429.7485 K) to the copper point (1357.77 K) which were used to derive by extrapolation the transition temperature T 90 (Co-C) of the cobalt-carbon eutectic fixed point. An INRIM cell was investigated and an average value T 90 (Co-C) = 1597.20 K was found with the four values lying within 0.25 K. Alternatively, thermodynamic approximated scales were realized by assigning to the fixed points the best presently available thermodynamic values and deriving T (Co-C). An average value of 1597.27 K was found (four values lying within 0.25 K). The standard uncertainties associated with T 90 (Co-C) and T (Co-C) were 0.16 K and 0.17 K, respectively. INRIM determinations are compatible with recent thermodynamic determinations on three different cells (values lying between 1597.11 K and 1597.25 K) and with the result of a comparison on the same cell by an absolute radiation thermometer and an irradiance measurement with filter radiometers which give values of 1597.11 K and 1597.43 K, respectively (Anhalt et al 2006. The INRIM approach allows the determination of both ITS-90 and thermodynamic temperature of a fixed point in a simple way and can provide valuable support to absolute radiometric methods in defining the transition temperature of new high-temperature fixed points.

Indirect Determination of the Thermodynamic Temperature of a Gold Fixed-Point Cell

“…On the other hand, the temperature scales realized with the Si standard thermometer proved to be more repeatable and reproducible than those with the InGaAs thermometer which was used to derive T Cu [12]. The 12 freezing temperatures obtained during the four measurement cycles are summarized in Table 2.…”

“…It is generally accepted that such an approach may be safely used within the extreme end points and that extrapolation beyond these points is not recommended. However, it has been proved that in some circumstances, namely, when a low uncertainty can be associated to the fixed-point realizations and when the temperature difference between the highest calibration point and the extrapolated point is relatively small, the uncertainty in the extrapolated temperature can be kept reasonably low [11,12].…”

“…The measurement setup was part of that already described in [12] and used for the determination of the thermodynamic temperature of the copper point. Essentially it is based on a standard radiation thermometer, a set of fixed-point cells, and the furnaces for realizing the fixed points.…”

“…The two furnaces used for this exercise were those already described in [12]. The compact fixed-point furnace, for which constructional details may be found in Ref.…”

“…Error bars associated to the mean measured values represent uncertainties with k = 1 as reported in [2] thermometers for many years. Recently, the technique has been improved and already used for the indirect determination of the thermodynamic temperature of the copper point [11] and of the transition temperature of the Co-C eutectic fixed point [12]. In the present exercise the technique has been implemented for deriving T Au both by interpolation and extrapolation of a thermodynamic scale.…”