Measurement of Specific Heat Using Bayesian Inference to Account for Uncertainties in the Calorimeter Performance
A. F. Emery, K. Abernethy · 2006
One way of estimating thermal conductivity is to measure the temperature history for transient conduction. Unfortunately, the property estimated is usually not the conductivity but the thermal diffusivity, α = k/ρcp. While α can often be estimated with good precision, and ρ is usually well defined, the specific heat is often poorly known. As a consequence the uncertainty in the conductivity can be large. This paper reports the use of a calorimeter designed to measure the specific heat with high accuracy. The accuracy depends upon a precise characterization of the thermal performance of the calorimeter. Even when much care is taken, the calorimeter's behavior introduces uncertainty in the measured specific heat. The effects of these uncertainties are accounted for by using Bayesian inference to estimate the confidence intervals of the specific heat.