项目名称: 叶轮机械颤振稳定性的不确定性预测方法研究
项目编号: No.51475025
项目类型: 面上项目
立项/批准年度: 2015
项目学科: 机械、仪表工业
项目作者: 唐文忠
作者单位: 北京航空航天大学
项目金额: 84万元
中文摘要: 导致叶轮机械颤振发生的很多影响因素,在设计、制造和使用过程中难以精确控制,使得叶轮机械颤振发生具有明显的不确定性。因此只有用统计特征参数描述颤振特性,进行颤振稳定性的概率预测,才能更准确、全面地把握颤振发生与发展的客观规律。本项目考虑颤振影响因素的随机性和颤振判据边界的模糊性,发展高精度且高效率的概率统计分析方法,与颤振预测的确定性方法相结合,建立叶轮机械颤振稳定性的不确定预测方法,主要包括基于能量法、频域法、时域法的叶轮机械颤振稳定性概率预测和模糊概率预测方法,以及随机失谐(包括结构参数失谐和气动参数失谐)叶轮机械颤振稳定性的不确定性预测方法。用本项目所提出的基于多学科分解协同策略建立近似代理模型,进行叶轮机械颤振稳定性概率预测,可以显著地提高计算精度和效率。本项目研究成果将对拓展气体弹性动力学的研究领域具有重要的学术意义,也在叶轮机械气弹稳定性设计中具有现实的应用前景。
中文关键词: 叶轮机械;颤振;概率预测;模糊预测
英文摘要: Many factors leading to flutter in turbomachinery is difficult to be accurately controlled in design, manufacture and operation, so that the flutter appears obvious uncertainty. Therefore, only describing the flutter characteristics with statistical parameters and predicting the flutter with probabilistic methods, the essential characteristics of flutter occurrence and development can be revealed. In this project, the nondeterministic methodologies to predict flutter in turbomachinery are presented, in which the random influential factors and the fuzzy criterion of the flutter are considered, and the high-precision and high-efficiency probabilistic statistical methods which are integrated into the deterministic flutter prediction methods are developed, and which include the probabilistic and fuzzy probabilistic prediction methods based on the energy method, the frequency domain method, and the time domain method, respectively, and the nondeterministic methods of prediction flutter in random mistuned turbomachinery. By means of the approximate surrogate models constructed in the light of multi-disciplinary decomposition coordination strategy, the computational precision and efficiency of the probabilistic prediction of flutter in turbomachinery can be markedly improved. The study of this project possesses important academic significance to develop aeroelastic dynamics, and realistic application prospect in aeroelastic dynamic design of turbomachinery.
英文关键词: turbomachinery;flutter;probabilistic prediction;fuzzy prediction