%0 Journal Article
%A Schmit, Timothy S.
%A Dhingra, Anoop K.
%A Landis, Fred
%A Kojasoy, Gunol
%D 1996
%I Begell House
%N 4
%P 281-290
%R 10.1615/JEnhHeatTransf.v3.i4.40
%T A Genetic Algorithm Optimization Technique for Compact High Intensity Cooler Design
%U http://dl.begellhouse.com/journals/4c8f5faa331b09ea,0a32880b5a8e6054,569def237ef46aa0.html
%V 3
%X This paper initially reviews the operation and design criteria for a compact high intensity cooler (CHIC) unit as used in avionic equipment. Here high heat loads are dissipated via multiple impinging jets fed sequentially through a series of fins connected with a bus bar to the heat source. The analytical basis for the heat transfer design, most of which has been published previously, is shown to predict the performance of CHIC units to a high degree of accuracy. This then permits optimizing the design. Most optimization techniques depend on continuous variables, while in the design of CHIC unit many of the critical geometrical variables must assume discrete values. A genetic algorithm, generally not well known in engineering circles, that looks for an optimum by simulating an evolutionary process was found to be satisfactory for this problem with its mixture of discrete and continuous variables. It is also shown that in an actual optimization problem, where the fluid pressure drop across the unit has to be balanced against a low overall thermal resistance, an optimum geometrical design can be determined. This design is an improvement over the empirical "best" design previously reported in the literature.
%8 1996-12-01