基于LEM-CST方法的水翼空化性能优化
Optimization of Cavitation Performance of Hydrofoil Based on LEM-CST Method
吴向阳 1王巍 2李智健 1纪祥 1袁龙灿 1王晓放2
作者信息
- 1. 大连理工大学能源与动力学院,大连 116024
- 2. 大连理工大学能源与动力学院,大连 116024;大连理工大学海洋能源利用与节能教育部重点实验室,大连 116024
- 折叠
摘要
开发具备优秀水力性能的抗空化翼型对优化海洋船舶动力装置具有重要意义.该文以NACA66(MOD)水翼为基准翼型,首先基于LEM-CST方法对水翼几何进行参数化,运用最优拉丁超立方抽样方法构建训练样本库;其次,以升阻比和最小压力系数为优化目标,借助BP神经网络代理模型建立从水翼几何到优化目标的映射关系;最后,结合遗传算法对翼型进行寻优,并对优化设计后的水翼开展CFD空化流场计算分析.研究发现:①相较于普遍流行的原始CST参数化方法,LEM-CST参数化方法的拟合精度更高;②在相同工况下,相比于NACA66(MOD)水翼,优化后新水翼的升阻比提升了 6.9％,最小压力系数提升了 21.6％;③经空化流场计算发现,新水翼的时均无量纲空泡面积降低了 9.2％,抗空化能力得到提升.
Abstract
It is of great significance to develop anti-cavitation airfoil with excellent hydraulic performance for optimization of marine ship power plant.In this paper,NACA66(MOD)hydrofoil is taken as the referenced airfoil.Firstly,the airfoil geometry is parameterized based on the LEM-CST method,and the training sample library is constructed by using the optimal Latin hypercube sampling method.Secondly,the lift-drag ratio and the minimum pressure coefficient are taken as the optimization targets,and BP neural network proxy model is used to establish the mapping relationship from airfoil geometry to optimization targets.Finally,the airfoil is optimized by genetic algorithm,and the cavitation flow field of the optimized airfoil is calculated and analyzed by CFD.The results show that:① The fitting accuracy of LEM-CST parameterization method is higher than that of the popular original CST parameterization method;② Compared with the NACA66(MOD)hydrofoil after optimization,the lift-drag ratio and the minimum pressure coefficient of the optimized airfoil are increased by 6.9％and 21.6％respectively under the same working condition;③ The cavitation flow field calculation shows that the dimensionless cavitation area of the new hydrofoil is reduced by 9.2％,and the anti-cavitation ability is improved.
关键词
翼型优化设计/LEM-CST参数化方法/BP神经网络/遗传算法/空化性能Key words
Airfoil optimization design/LEM-CST parameterization method/BP neural network/Genetic algorithm/Cavitation performance引用本文复制引用
出版年
2024
NETL
NSTL
万方数据