迷宫密封结构设计优化

Shape Optimization of a Labyrinth Seal Configuration

耦合Kriging代理模型、迷宫密封参数化方法和自适应差分进化算法,在数值验证的基础上,完成了某迷宫式密封多目标鲁棒性设计优化。优化时,选取压比和密封间隙作为不确定性变量,同时选取齿厚等10个几何参数作为优化变量,以迷宫密封泄漏量的均值及其方差最小为目标,进行多目标优化。优化后,在给定压比和间隙条件下,最优设计泄漏量降低了29.85%,在压比及间隙同时变化时,密封性能提升同样明显,从而验证了多目标设计优化方法的正确性。

A multi-objective robust optimization method is proposed and implemented for the design of a labyrinth seal upon nu- merical validation. This method combines the Kriging surrogate model with parameterization method of a labyrinth seal, and self-adaptive multi-ohjective differential evolution algorithm （SMODE） as well. The multi-objective optimization is conducted for mini- mizing averaged leakage flow with minimum variance. In the optimization process, pressure ratio and seal clearance are set as un- certainty variables, and other 10 geometrical parameters are set as optimization variables. After optimization, the leakage flow rate of the optimal solution is reduced by 29.85% when pressure ratio and seal clearance are set as constant. And such smaller leakage flow maintained when pressure ratio and seal clearance are varied. Therefore, the correctness and effectiveness of the proposed method is demonstrated.