Abstract
The increasing power requirement for electronics industry, combined with ever-shrinking size and weight allowances, is creating a greater need for optimal design of generalized heat sinks. In this study, the thermal and hydrodynamic models for confined heat sinks with various fin types have been successfully developed. Comparisons between the predicted heat transfer and fluid flow characteristics for both fully-confined and partially-confined generalized heat sinks with existing data are made with satisfactory agreements. In addition, a series of parametric studies, including effects of the size of heat source, power of heating load, the fin structures of heat sink, the conductivity of heat sink, inlet velocity, and the flow bypass conditions, for thermal design of heat sinks have been performed.To conduct the experimental investigation on the fluid flow and heat transfer characteristics for confined generalized heat sinks, an experimental system has been successfully established with an adjustable test section. From the experimental data, the generalized correlations for both the thermal performance and the pressure drop of confined heat sinks are proposed as the functions of Reynolds number, top-bypass ratio, and the side-bypass ratio. Besides, for further validation of the present theoretical results, a comparison between the present predictions with the experimental data was made with a reasonable agreement. The maximum and average deviations are 25.8% and 11.7%, respectively, for the pressure drop; and 14.3% and 6.1%, respectively, for the thermal resistance.In addition, a systematical design optimization method, which allow the thermal engineer to meet several design objectives and constraints simultaneously and effectively, has been successfully presented and applied to the optimal designs for generalized heat sinks in this study. First of all, a statistical method for the sensitivity analysis is performed to determine the key factors that are critical to the design; and a response surface methodology (RSM) is applied to establish explicit regression models for the thermal resistance and the pressure drop in terms of the design factors with an well-organized design of experiments (DOE). By employing the gradient-based numerical optimization technique, a series of constrained optimal designs can be efficiently performed. Comparisons between these predicted optimal designs and those evaluated by the theoretical calculations are also made with satisfactory agreements. With the developed design optimization method, optimal designs for generalized heat sinks, includes the parallel-plate fin (ppf), pin-fin array (pfa), and strip-fin array (sfa), were successfully explored with multiple design constraints of pressure drop, mass, and space limitations.Furthermore, an effective and user-friendly optimal computer-aided design (CAD) system for automatically predicting the optimal thermal performance for confined generalized heat sinks has been successfully developed. In the pre-processor, a user-friendly interface for problem definition has been constructed in order to efficiently collect the required data for the thermal analyzer and optimizer. And a thermal analyzer for confined heat sink has been successfully developed according to the present theoretical calculations. The performances studied include the pressure drop, local and average heat transfer coefficients, local temperature distribution, and the overall thermal resistance. Besides, corresponding to the presented design optimization method, the design optimizer has been established with the functions of the design of experiments, the construction of response surface models, and the programming of numerical optimization. After obtaining the optimal design, the real-time 3-D model, predicted color isotherms, and all the predicted performances can be displayed in the developed post-processor. Moreover, an interactive function has been applied for providing a direct communication between the user and the computer for the detailed information about the optimal design. Finally, to demonstrate the superiorities of the present developed optimal CAD system, two sample applications of thermal optimal designs for generalized heat sinks under multiple constraints have been effectively performed.