Application of parameter estimation techniques to thermal conductivity probe data reduction
Abstract
A data reduction method is described that makes use of experimental temperature data at both early and late times, does not require full development of the classical linear temperature increase versus log-time asymptotic behavior, is not as sensitive to data noise as alternative techniques, and permits evaluation of the experiment design and data assumptions through statistical analysis of final residuals. The approach consists of obtaining a nonlinear least squares fit between an analytical solution containing the parameters of interest (conductivity, contact resistance and in some cases density-specific heat product) and experimental data. The method estimates the parameters of the analytical solution from experimental data. Examination of the sensitivity coefficients (derivatives of the parameters with respect to the measured variable) demonstrates the adequacy of thermal conductivity as a fit parameter. Application of the method to analysis of laboratory and in situ field data are presented for illustrative purposes.
- Publication:
-
Presented at the Intern. Joint Conf. on Thermophys. Properties
- Pub Date:
- 1981
- Bibcode:
- 1981thpr.confQ....K
- Keywords:
-
- Data Reduction;
- Least Squares Method;
- Parameter Identification;
- Thermal Conductivity;
- Contact Resistance;
- Physical Chemistry;
- Fluid Mechanics and Heat Transfer