Simulation-based optimization of thermal systems

Abstract

This paper considers the design and optimization of thermal systems on the basis of the mathematical and numerical modeling of the system. Many complexities are often encountered in practical thermal processes and systems, making the modeling challenging and involved. These include property variations, complicated regions, combined transport mechanisms, chemical reactions, and intricate boundary conditions. The paper briefly presents approaches that may be used to accurately simulate these systems. Validation of the numerical model is a particularly critical aspect and is discussed. It is important to couple the modeling with the system performance, design, control and optimization. This aspect, which has often been ignored in the literature, is considered in this paper. Design of thermal systems based on concurrent simulation and experimentation is also discussed in terms of dynamic data-driven optimization methods. Optimization of the system and of the operating conditions is needed to minimize costs and improve product quality and system performance. Different optimization strategies that are currently used for thermal systems are outlined, focusing on new and emerging strategies. Of particular interest is multi-objective optimization, since most thermal systems involve several important objective functions, such as heat transfer rate and pressure in electronic cooling systems. A few practical thermal systems are considered in greater detail to illustrate these approaches and to present typical simulation, design and optimization results.