Estimating temperature-dependent thermal conductivity of copper oxide using an inverse method

Temperature-dependent thermal conductivity of copper oxide is of great significance for the research on the thermal hazards caused by poor electrical contact. In addition, copper oxide is also a promising material in energy storage. In the aforementioned fields, the heat transfer and temperature distribution are determined by the thermophysical properties of copper oxide. However, thermal conductivity of copper oxide is seldom mentioned in the available literature. Moreover, it is impractical to test the copper oxide's thermal conductivity by the existing instruments directly due to the difficulty in sample preparation and the limitations of the equipment. Therefore, we investigate an approach to determine the temperature-dependent thermal conductivity of copper oxide using an inverse method. Temperature-drop experiments are conducted to record the heat transfer process over a broad temperature range. Three optimization algorithms, including SNOPT (Software for Large-Scale Nonlinear Programming), particle swarm optimization, and simulated annealing, except for the optimization methods, the effects of the baseline temperature and measurement errors are also tested. Results demonstrate that the particle swarm optimization is the most applicable method to solve the thermal conductivity problems with minimum errors. The average, lower and upper 95%\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\%$$\end{document} confidence intervals of the parameter estimation results are provided, which can be used for further heat transfer modeling.