Genetic Algorithm Optimization for Primary Surfaces Recuperator of Microturbine

Abstract

Recent years, Genetic Algorithm (GA) technique has been gotten much more attention in solving real-world problems, more successful genetic algorithms applications to engineering optimization have shown the technique has strong ability of global searching and optimizing based on various objectives for their optimal parameters. The technique may be applied to more complicated heat exchangers and is particularly useful for new types. It is important to optimize heat exchanger, for minimum volume/weight to save fabrication cost or for improved effectiveness to save energy consumption, under requirement of allowable pressure drop; simultaneously it is mandatory to optimize geometry parameters of heat plate from the technical and economic standpoints. In this paper, GA is used to optimize the Cross Wavy Primary Surface (CWPS) and Cross Corrugated Primary Surface (CCPS) geometry characteristic of recuperator in 100kW microturbine, in order to get more compactness and minimum volume and weight. Two kinds of fitness assignment methods were considered. Furthermore, the GA parameters were set optimally to yield smoother and faster fitness convergence. The comparison shows the superiority of GA and confirms its potential to solve the objective problem. When the rectangular recuperator core size and corrugated geometries are evaluated, in the CWPS the weight of recuperator decreases 12% or more, the coefficient of compactness increases by up to 19%, with an increase of total pressure drop by a percentage of 0.84 compared to the original design data, the total pressure drop as a percentage of the operating pressure is controlled to be less than 3%. In the CCPS area compactness is increased to 70% the initial design result by decreasing pitch and relatively high height of the passage, the weight decreases by 17% to 36%, depending on different inclination angle (θ). Comparatively the CCPS shows superior performance for use in compact recuperators of the future. The GA technique chooses from a variety of geometry characters, optimizes them and picks out the one which provides the closest fit to the recupertor for microturbine.