基于iSIGHT的大膨胀比5.0级向心涡轮多目标优化与分析

Multi-Object Optimization and Analysis of 5.0 Expansion Ratio Radial Turbine Based on Isight

关键词:  向心涡轮  iSIGHT  三维优化系统  实验设计方法（DOE）  多目标优化

Key words:  Radial turbine  iSIGHT  3D optimization system  Design of experiments method（DOE）  Multi-object optimization

摘要:

针对轻量化、紧凑化、高效化的向心叶轮设计需求，对5.0大膨胀比向心涡轮进行三维多目标优化。首先，基于iSIGHT，集成Numeca，CFX软件及自编程序，搭建了三维气动优化集成系统，实现向心涡轮流道、叶片的参数化、网格划分、数值计算及三维结果的自动处理与优化，且其所需的存储空间仅为原优化系统的2.4%，极大地降低了对计算机存储空间的要求；其次，针对向心叶轮造型参数众多所导致的优化规模巨大问题，采用试验设计方法（DOE）详细地开展了流道、叶片的控制参数对涡轮性能的影响研究，得到造型参数对涡轮效率的贡献度，继而给出了优化变量的选取依据，从102个造型参数中选取23个作为优化变量，减少了优化的盲目性，缩短了优化时长，极大地减轻了工作量，提高了优化效率；最后，考虑涡轮性能、排气损失以及叶轮轻量紧凑化的需求，以涡轮总对静效率η_ts，级出口气流角α₆及叶片的表面积A作为优化目标，采用NSGA-Ⅱ算法进行多目标优化并探讨相应机理。优化后，叶轮的轴向长度缩短了8.09%，叶片的表面积减小了8.46%，有效降低了叶轮的尺寸及重量；在保持涡轮进口流函数和膨胀比基本不变的情况下，设计点涡轮总对静效率提高了0.1%，级出口绝对气流角仅降低1°，且不同转速下涡轮的性能基本保持不变。以上研究表明，该三维优化系统和多目标优化方法的有效性。

Abstract:

For lightweight， compact， efficient radial turbine design requirements， 3D multi-object optimization of a 5.0 expansion ratio radial turbine is carried out. Firstly， based on optimization platform iSIGHT， a 3D integrated aerodynamic optimization system is developed with Numeca， CFX and self-program to automatically process the flow channel and the blade parameterization， mesh generation， numerical calculation， 3D calculation result and optimization， and it requires only 2.4% of the computer storage space compared with original optimization system. Secondly， in order to solve the problem of huge optimization scale caused by modeling parameters of radial impeller， the effects of flow channel and blade control parameters on turbine performance were studied using design of experiments method（DOE）. Contribution of modeling parameters to turbine efficiency is obtained， and the selection basis of optimization variable is given. 23 of 102 modeling parameters are selected as optimization variables， which reduce the blindness of optimization， shorten the optimization time， greatly reduce the workload and improve the optimization efficiency. Finally，considering the performance of turbine， exhaust loss and the requirement of impeller compactness， total to static turbine efficiency η_ts， stage outlet flow angle α₆ and blade surface area A are taken as the optimization objectives， NSGA-Ⅱalgorithm is used for multi-objective optimization and the corresponding mechanism is discussed. After optimization， the axial length of impeller is shortened by 8.09% and the surface area of blade is reduced by 8.64%， which effectively reduces the size and weight of the impeller. The total to static turbine efficiency increases by 0.1% and the stage outlet flow angle decreases by 1° under the condition of the inlet flow function and the expansion ratio remains almost unchanged. Furthermore， the performance of turbine remains unchanged at different rotor speeds after optimization. The above research shows that the 3D optimization system and multi-object optimization method are effective.