遺伝的アルゴリズムを用いたCAMUI型燃料グレインの最適設計

概要

論文の詳細を見る
The authors have been developing Cascaded Multistage Impinging-jet (CAMUI) type hybrid rockets. A CAMUI type hybrid rocket uses a fuel grain consisting of several cylindrical fuel blocks with two ports. To minimize both of c* loss due to O/F shifting and residual fuel weight after burning, an appropriate design of initial fuel grain shape is necessary. However, obtaining an optimum design of initial grain shape is not easy because there are many design variables influencing one another. To solve this problem, the authors employed Genetic Algorithm (GA) combined with a numerical model forecasting performance history of a CAMUI type motor. GA can acquire an approximate optimum solution for problems with a vast search space in practical time. The numerical model gives c* loss and residual fuel weight to evaluate the degree of performance of each initial fuel grain shape. A fuel grain design proposed by this method showed residual fuel weight as small as 6.36% of the initial weight and c* loss less than 1%.