Bayesian Parameter Estimation and Model Class Selection for Identification of Spatially Dependent Convective Heat Transfer Coefficients in a Microtube Heat Sink

The analysis of heat transfer at the microscale has been widely employed in recent decades due to technological advances in the industry of miniaturized fluid transport systems. This work presents a model class selection approach for the identification of spatially dependent convective heat transfer coefficients in a microtube heat sink. The physical system of interest is the cooling of a heat source simulating a processor through a thermal heat sink with a single microtube. The methodology consists in the application of the Transitional Markov Chain Monte Carlo Method for parameter estimation and model evidence calculation, used in the selection of adopted hypotheses. The methodology is illustrated with numerical examples representing different scenarios, demonstrating the feasibility and robustness of the approach.