Estimation of Thermal Networks Using Singular Value Decomposition Method

Abstract

A model is developed using a thermal network analysis approach. An experimental setup is designed and built to validate the model. The model is based on an inverse thermal network analysis using a singular value decomposition algorithm. The data from the experiments are used to determine the thermal conductances using inverse model's unsteady- and steady-state forms. The model development and experimental results are presented together with possible sources of error and how to avoid or decrease their influence. Good agreement between the thermal network model predictions and experiments was found.

Keywords:

• thermal network
• thermal conductance
• singular value decomposition
• ill-conditioned
• inverse method