基于Kriging模型及近似降阶技术的失谐叶盘振动响应预测与优化研究

The major accidents of aero-engine are related to the high-cycle fatigue of blades vibration, and the coupled vibration of blades disk is an important reason for producing the high-cycle fatigue of blades. In order to reduce the risk of high-cycle fatigue accident of blades, it is necessary to accurately predict and control the vibration response of mistuned blades disk in aero-engine. Taking a scale model of mistuned blades disk as research object, the prediction and optimization of vibration response of mistuned blades disk are carried out based on Kriging model and approximate reduced-order technique. Research contents: Identification of mistuned parameters of blades by the static frequency test. Prediction of vibration response of mistuned blades disk based on prestressed fixed interface free interface component mode synthesis reduced-order technique. Optimization of vibration response of mistuned blades disk based on Kriging model and intelligent optimization algorithm. Verification of the prediction and optimization accuracy of vibration response by the vibration test of NSMS. Key issues: Accurate prediction of vibration response of mistuned blades disk. Vibration optimization of mistuned blades by the scale finite element model of blades disk. Innovation point: The prediction method of vibration response of mistuned blades disk is proposed based on prestressed fixed interface free interface component mode synthesis reduced-order technique. The vibration optimization method of mistuned blades of blades disk is presented based on the combination of Kriging model and intelligent optimization algorithm.

航空发动机重大事故几乎都与叶片振动的高周疲劳有关，失谐叶盘的耦合振动已成为诱发叶片产生高周疲劳的重要原因。为了降低叶片发生高周疲劳事故的风险，必须对发动机失谐叶盘系统的振动响应进行准确预测与抑制。本项目以某失谐叶盘系统缩比模型为研究对象，开展基于Kriging模型/近似降阶技术的失谐叶盘系统振动响应预测与优化研究。研究内容：基于静频实验的叶片失谐参数识别；基于固定界面预应力-自由界面CMS降阶技术的失谐叶盘系统振动响应预测；基于Kriging模型/智能优化算法的失谐叶盘系统振动优化；基于NSMS振动测试的振动响应预测与优化精度验证。关键问题：失谐叶盘系统振动响应的准确预测；基于缩比叶盘系统有限元模型的失谐叶片振动优化。创新点：提出基于固定界面预应力-自由界面CMS降阶技术的失谐叶盘系统振动响应预测方法；提出基于Kriging模型与智能优化算法相结合的叶盘系统失谐叶片振动优化方法。

航空发动机重大事故几乎都与叶片振动的高周疲劳有关，失谐叶盘的耦合振动已成为诱发叶片产生高周疲劳的重要原因。为了降低叶片发生高周疲劳事故的风险，必须对发动机失谐叶盘系统的振动响应进行准确预测与抑制。本项目开展航空发动机失谐叶盘结构振动响应预测与振动优化研究不仅对解决发动机运行安全可靠性的关键问题，提高我国航空发动机研制、设计水平，增强我国航空工业的国际竞争力具有重要而紧迫的实际意义，而且对促进基础学科和工程技术应用学科的相互交叉融合具有重要的理论意义。.本项目以某失谐叶盘系统缩比模型为研究对象，开展了基于Kriging模型/近似降阶技术的失谐叶盘系统振动响应预测与优化研究。主要研究内容：基于静频实验的叶片失谐参数识别；基于移动界面预应力-自由界面CMS降阶技术的失谐叶盘系统振动响应预测；基于Kriging模型/智能优化算法的失谐叶盘系统振动优化；基于NSMS振动测试的振动响应预测与优化精度验证。取得的重要成果：搭建了叶片静频测试实验台和基于叶尖定时的旋转失谐叶片振动非接触测量(NSMS)实验台。采用叶片静频测试与有限元相结合的方法，实现了叶片失谐参数的识别。基于固定界面预应力-自由界面CMS降阶法预测了失谐叶盘系统振动特性，在固定界面预应力-自由界面CMS降阶法的基础上，提出了基于移动界面预应力-自由界面CMS降阶法的失谐叶盘系统振动响应预测方法；通过直接法及NSMS振动测试验证了该方法预测精度。该方法具有更高的预测精度，更适于预测失谐叶盘系统的振动特性。讨论了不同叶片-轮盘的耦合程度对失谐叶盘系统振动局部化程度的影响。基于Kriging模型及假设弹性体法获得了热-流-结构多场耦合作用下叶盘系统的力学特性，讨论了不同载荷对叶盘系统力学特性的影响。利用智能优化算法，对失谐叶片进行了减振排布优化，有效降低了叶片的振动幅度和叶盘系统的振动局部化程度，显著改善了叶盘系统的振动能量分布。