Node-based spacecraft radiator design optimization

Abstract

Spacecraft radiators can be designed through thermal analysis of a thermal model that is divided into nodes. Thus, the radiator is designed discretely in unit nodes; the radiator node distribution for the node division of a candidate radiator region indicates the radiator design in the form of a radiator node combination. Because the radiator size and topology in terms of shapes and locations are key design factors for radiator efficiency, the node-based design approach can cover all of these factors in unit nodes for a given node divi-sion. Radiator design optimization is mainly concerned with minimizing radiator size under the constraint of temperature limits; thus, the node-based optimization approach determines the optimum radiator design as the optimal radiator node combination with the minimum number of radiator nodes while satisfying the temperature limits. Two optimization methods are proposed that numerically search for the optimal solutions to a radiator design in the form of radiator node combinations in different ways. The first method uses an integrated numerical optimization analysis that combines an optimization algorithm with thermal analysis and executes them iteratively during the optimization process. Design variables of 0 or 1 bits are assigned to each node division to represent the physical radiator design of a node combination and serve as the interface between two sets of software; thus, the binary numbers of 0- or 1-bit design variable strings are mapped one-to-one to every possible radiator design. This optimization is concretely formulated as a multi-objective problem and further modified by user preferences of thermal design heuristics. The second method is a stepwise approach that adds one radiator node at each step according to the criterion of temperature sensitivity until the temperature limits are satisfied; this stepwise approach can also be regarded as a progressive preference-based optimization method in that it steps forward...