Understanding and improving energy dissipation and vibration damping in structures subject to self-excited irregular vibrations – linking data driven approaches with modelling

该项目的重点是发展一种方法来分析和描述阻尼设计元素和耗散机制在复杂系统状态和多种负载情况下自激多组件系统中的影响和作用。而对于规则动力学，即平稳周期响应或线性瞬态，阻尼元件和机制可以用标准技术来描述，对于不规则动力学，识别能量的来源和汇甚至结构中的能量流是一个很大程度上未解决的问题。此外，大多数技术系统包含大量的局部非线性和阻尼元件，例如系统的关节和接触界面，并且受到多种不同载荷情况的影响。这给阻尼装置的评估带来了挑战。在不规则振动状态下，多种能量汇处于不断的相互作用中。最近，由于多种载荷情况造成的额外复杂性使得应用简单的自下而上的基于物理的阻尼或耗散描述来评估阻尼装置是不可能的。因此，需要新的技术来表征阻尼设计元素。在此背景下，我们提出了深入分析振动响应以及结构、阻尼和载荷参数的发展方法。第一部分将重点发展分析和计算方法，用于研究受不规则振动的摩擦激励系统中的阻尼装置，并且仅限于计算机建模和从数值模型中获得的数据。本文采用了非线性时间序列分析和多元统计相结合的方法。中心元素是数据矩阵M，该矩阵由多个负载情况下不同类别的特征量填充。最后的结果是评估和优化阻尼装置和耗散机制的系统受到不规则振动的程序。第二部分将努力将方法扩展到包括从物理测试中获得的数据的系统，并将因此作为验证元素。所考虑的系统是销盘系统和汽车摩擦制动器。几乎总是在验证中，对结果的理想期望是，早期开发的方法可以转移到真实的更大的系统中，而不会进一步复杂化。然而，我们期望我们在这里提出的验证工作也将导致新的见解，即如何调整或修改方法本身，以达到稳健和令人信服的结果。该提案努力克服基于物理的系统建模、数据分析和时间序列分析以及设计领域之间的现有界限。因此，它具有高度的远见和雄心勃勃的基调。

The focus of the project is on the development of a method to analyse and characterise the influence and role of damping design elements and dissipation mechanisms in multi-component systems with self-excitation subject to complex system states and manifold load cases.While for regular dynamics, i.e. stationary periodic response or linear transients, damping elements and mechanisms can be described with standard techniques, for irregular dynamics it is a largely unsolved problem to identify sources and sinks of energy or even the flow of energy in structures. In addition, most technical systems contain a plethora of local nonlinearities and damping elements, e.g. the joints and contact interfaces of the system, and are subject to a multitude of different load cases. This causes a challenging scenario for the assessment of damping devices. In the irregular vibrational state the manifold energy sinks are in constant interaction. At the latest, the additional complexity caused by the manifold load cases makes the application of simple bottom-up physics-based damping or dissipation description for the assessment of damping devices impossible. Hence, new techniques to characterize damping design elements are required. In this context we propose developing methods for the thorough analysis of the vibration response as well as the structural, damping and load parameters. The first part will focus on developing the analysis and computational methods for studying damping devices in friction-excited systems subject to irregular vibrations and is restricted to computer modelling and data derived from numerical models. In this context methods from nonlinear time series analysis and multivariate statistics are employed and combined. The central element is the data matrix M which is filled with characteristic quantities from different classes for multiple load cases. The final result is a procedure for the assessment and optimization of damping devices and dissipation mechanisms in systems subject to irregular vibrations.The second part will strive to extend the approach to systems where data obtained from physical testing is included, and will thus serve as a validation element. The considered systems are a pin-on-disk system and an automotive friction brake. As almost always in validation, the ideal expectation for the outcome is that the approach developed earlier can be transferred to real and larger systems without further complications. However, we expect that the validation work that we propose to conduct here will also lead to new insights into how the approach itself has to be tailored or perhaps modified to reach robust and convincing results. The proposal strives to overcome existing boundaries between the domains of physics-based system modelling, data analysis and time-series analysis, as well as design. Thereby it has a highly visionary and ambitious tone.