摘要: |
为满足某涡轴发动机平原、高原、高寒以及海洋环境下的全疆域设计需求,综合采用尺寸优化和拓扑优化对核心机转子盘体形状进行了减重优化。首先提出了一种基于本征正交分解的径向基神经网络代理模型构造方法,通过在关键区域补充采样,在样本总数相当的前提下,提高了寻优效率,缩短了优化所需的时间。利用该方法对核心机转子盘体尺寸进行优化,优化后结构重量减轻,应力分布和强度储备更趋合理。然后采用变密度法对涡轮盘螺栓孔附近区域进行拓扑优化,得到一种带凸耳形状的螺栓孔创新构型,解决了由于优化后盘体减薄而导致的螺栓孔应力上升过多的问题。结果表明:在转子强度、寿命满足规范要求的前提下,优化后的核心机转子减重15%,满足了全疆域设计需求对转子重量的要求。 |
关键词: 全疆域设计 核心机转子 尺寸优化 拓扑优化 本征正交分解 径向基神经网络 |
DOI:10.13675/j.cnki.tjjs.210328 |
分类号:V232.4 |
基金项目: |
|
Weight Reduction Optimization of Turboshaft Core-Engine Rotor for Whole Territory Design Requirements |
BAO You-lin, ZHANG Li-zhang, ZHAO Yan-yun, LI Li-xin
|
AECC Hunan Aviation Powerplant Research Institute,Zhuzhou 412002,China
|
Abstract: |
In order to meet the whole territory design requirements of a turboshaft engine in plain, plateau, high cold and marine environment, the size optimization and topology optimization were used to reduce the weight of the core-engine disc. Firstly, a Radial Basis Function (RBF) network surrogate model construction method based on Proper Orthogonal Decomposition (POD) is proposed. Through supplementary sampling in the key areas, the optimization efficiency is improved and the optimization time is shortened on the premise that the total number of samples is equal. This method is used to optimize the disc size of core-engine rotor, which reduces the weight and makes the stress distribution and strength reserve more reasonable. Then, the topology optimization method of Solid Isotropic Material with Penalization (SIMP) is used to optimize the topology of the region near the bolt hole of the turbine disc, and an innovative configuration with lug shape is obtained, which solves the problem of bolt hole stress rising too large caused by the disc thinning after optimization. The results show that the weight of optimized core-engine rotor is reduced by 15% which meets the requirement of rotor weight for the whole territory design on the premise that the strength and life of core-engine rotor meet the standard requirements. |
Key words: Whole territory design Core-engine rotor Size optimization Topology optimization Proper orthogonal decomposition Radial basis function network |