Open-source numerical strategy for the analysis and redesign of turbomachine blades accounting for nonlinear structural interactions

在过去的十年中，减少涡轮机器（如飞机发动机和燃气轮机）的环境足迹成为电力和航空航天工业的一个主要问题。虽然已经取得了重大进展，但有希望进一步改进的途径需要减少与接触或摩擦相关的危险非线性结构相互作用，特别是对于叶片部件。对于工程师来说，这是一个主要的障碍，因为叶片设计过程主要是由空气动力学因素驱动的。非线性结构的相互作用，充其量，解释了一个后验，从而使现代叶片设计一个低效的试错过程。当有报道的事故要求重新设计叶片时，这个漫长的设计过程意味着非常高的成本，因为发电厂必须停止运行或飞机停飞。涡轮机械叶片的各种非线性结构相互作用与许多可能发生接触或摩擦的机械界面有关。对于这些界面，最近的研究已经导致了预测特设数值方法的发展。然而，对于每种类型的接口使用不同的数值范式和算法已经划分了这些方法。在此背景下，提出的研究计划旨在开发一个开源的数值策略，用于分析和重新设计涡轮机械叶片，考虑非线性结构相互作用。首先，将提出一个统一的数值框架，对涡轮机械叶片的振动响应进行首次全面的多非线性分析，包括摩擦和结构接触（分别发生在叶片/盘界面和叶片-叶尖/机匣界面）。频率和时域方法将在双管齐下的验证方法中考虑。提出的研究的第二个目标集中在致力于非线性结构相互作用的叶尖定时算法的发展。这对于更好地理解与这些相互作用相关的潜在物理现象至关重要。这将通过开发涡轮机器数字孪生体来实现，该数字孪生体用于生成具有非线性结构相互作用的叶尖定时数据。最后，提出的研究计划的最后一个目标涉及到重新设计的数值策略。它将主要依靠原始的开放式叶片参数化来满足空气动力学家和结构工程师的需求。从前两个研究目标中获得的结果将产生创新的指导方针，这些指导方针将用于设计具有非线性结构相互作用的叶片。该研究项目将直接应用于涡轮发动机叶片设计的改进，从而在整体能源效率和维护成本方面获得显著收益。

Over the past decade, reducing the environmental footprint of turbomachines (such as aircraft engines and gas turbines) became a leading issue in both power and aerospace industries. While significant gains have already been made, promising avenues for further improvements require to mitigate hazardous nonlinear structural interactions related to contacts or friction, specifically for bladed components. This is a major roadblock for engineers due to the fact that blade design procedures are mostly driven by aerodynamic considerations. Nonlinear structural interactions are, at best, accounted for a posteriori, thus making modern blade design an inefficient trial-and-error process. When reported incidents call for a blade redesign, this lengthy design process translates into very high costs as power plants must be stopped or planes grounded. The large variety of nonlinear structural interactions involving turbomachine blades is related to the many mechanical interfaces on which contacts or friction may occur. For each of these interfaces, recent research has led to the development of predictive ad-hoc numerical methodologies. However, the use of distinct numerical paradigms and algorithms for each type of interface has compartmentalized these methodologies. In this context, the proposed research program aims at developing an open-source numerical strategy for the analysis and redesign of turbomachine blades, accounting for nonlinear structural interactions. First of all, a unified numerical framework will be proposed to carry out the first comprehensive multi-nonlinear analyses of turbomachine blades' vibration response, including both friction and structural contacts (respectively occurring at the blade/disk interface and at the blade-tip/casing interface). Frequency and time domain methodologies will be considered in a two-pronged verification approach. The second objective of the proposed research focuses on the development of blade-tip timing algorithms dedicated to nonlinear structural interactions. This will be essential to better understand the underlying physical phenomena associated to these interactions. It will be made possible through the development of a turbomachine digital twin for the generation of blade-tip timing data featuring nonlinear structural interactions. Finally, the last objective of the proposed research program relates to the redesign numerical strategy. It will notably rely on an original and open blade parameterization meeting both the needs of aerodynamicists and structural engineers. Results obtained from the two first research objectives will yield innovative guidelines, which will be used to design blades robust with respect to nonlinear structural interactions. This research program will have direct applications for the improvement of turbomachine blade design allowing for significant gains in terms of overall energy efficiency and maintenance cost.