Evaluation of methods for characterizing the melting curves of a high temperature cobalt–carbon fixed point to define and determine its melting temperature

Abstract

The future mise en pratique for the realization of the kelvin will be founded on the melting temperatures of particular metal–carbon eutectic alloys as thermodynamic temperature references. However, at the moment there is no consensus on what should be taken as the melting temperature. An ideal melting or freezing curve should be a completely flat plateau at a specific temperature. Any departure from the ideal is due to shortcomings in the realization and should be accommodated within the uncertainty budget. However, for the proposed alloy-based fixed points, melting takes place over typically some hundreds of millikelvins. Including the entire melting range within the uncertainties would lead to an unnecessarily pessimistic view of the utility of these as reference standards. Therefore, detailed analysis of the shape of the melting curve is needed to give a value associated with some identifiable aspect of the phase transition. A range of approaches are or could be used; some purely practical, determining the point of inflection (POI) of the melting curve, some attempting to extrapolate to the liquidus temperature just at the end of melting, and a method that claims to give the liquidus temperature and an impurity correction based on the analytical Scheil model of solidification that has not previously been applied to eutectic melting. The different methods have been applied to cobalt–carbon melting curves that were obtained under conditions for which the Scheil model might be valid. In the light of the findings of this study it is recommended that the POI continue to be used as a pragmatic measure of temperature but where required a specified limits approach should be used to define and determine the melting temperature.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 34 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.