Sutherland, R.; Seward, A.M. 2024 Calibration and measurement of thermal properties using portable electronic divided-bar apparatus. Lower Hutt, N.Z.: GNS Science. GNS Science report 2024/24. 20 p.; doi: 10.21420/BRSG-8P81
Abstract
The divided bar apparatus provides a convenient, quick and accurate way to measure thermal resistance and heat capacity of rock samples that are sandwiched between internal standards. A temperature step is applied across the apparatus and temperatures at each interface are monitored. This report summarises protocols of measurement and recommends improvements in methodology. In particular, a new method of inference of thermal properties is presented: (1) a heat loss correction is made by extrapolation of temperature step values across the sample and internal standards at equilibrium to a median sample temperature at room temperature; (2) thermal resistances of internal (polycarbonate) standards and the average contact resistance (includes brass plates) is determined using measurements of external standards with known resistances; (3) a metric to infer thermal resistance of an unknown sample is constructed using Fourier’s law, corrected temperature steps, and the thermal resistances of internal standards and contacts. This semi-empirical approach is constructed to minimise the effects of heat loss, but uncertainty in the heat loss model is suggested to be the most important remaining source of error, with contact resistances being the second most important source of error. The heat loss model depends on sample geometry and thermal properties, and equipment setup, e.g. configuration of insulation, and equilibration temperature. The heat capacity of a sample is determined by integration of heat accumulation in the sample during a period of temperature increase and re-equilibration. An empirical correction is made to account for inaccuracies related to determination of thermal resistances within the apparatus or any other systematic (repeatable) aspects of assumptions or experimental apparatus or procedure. Our approach is repeatable and accurate. Remeasurement of some or all samples in a set would provide an independent way to estimate repeatability, i.e. uncertainty in thermal resistance, that is specific to the sample and insulation geometry used, but the fit to known external standards provides an adequate estimate of uncertainties.
