Mid- and High-Frequency Vibroacoustics of Built-up Structures -- A Wave Approach

噪声和振动是许多机械系统的重要性能方面。高噪声和振动水平可能对结构（例如，造成损坏）和操作人员（例如，造成疲劳或伤害）有害。因此，重要的是能够理解结构如何振动和发出噪音，即，它们的振动声学行为。传统上，工程师会尝试用解析方法来描述振动声学。然而，这些仅适用于非常简单的结构。工程师在航空航天、铁路或海事领域遇到的结构通常由复合板制成，这些复合板使用复杂的结构接头连接在一起。这种复杂的组合结构的振动声学分析不能用解析法进行。多年来，研究人员已经开发了数值技术来解决这个问题。基于元素的方法（如有限元法）是发展成熟的方法，可以使用许多商业/内部代码。然而，航空航天、铁路和海事结构相对较大。例如，典型的铁路车辆可以使用高达500 Hz的有限元方法来建模。在该频率以上，有限元模型的尺寸变得太大，不切实际，并且相关的计算成本变得过高。然而，音频频率范围为20 Hz-20 kHz。在高频（10 kHz以上）下，可以使用基于能量的方法（例如，统计能量分析方法）对铁路车辆进行建模。基于能量的统计方法是有价值的，但不如基于元素的方法成熟。火车车厢的例子指出了一个频率间隙，实际上是一个中频间隙，在这个间隙中，既不能使用基于元素的方法，也不能使用基于能量的方法。我建议使用波动方法来弥合中频间隙，并进一步加强能量方法。波为振动声学提供了统一、直观的方法。波浪模型的计算成本相当小（特别是与完整的有限元模型相比），并且结构的波浪特性可以通过对任意大结构的一小部分的有限元模型进行后处理来获得。因此，本计划的目标是开发一个基于波浪的工具箱，用于模拟振动声学复杂的组合结构。来自航空航天、铁路和海事部门的工业实例将用于证明所开发方法的效率和有效性。

Noise and vibration are important performance aspects in many mechanical systems. High noise and vibration levels can be detrimental to structures (e.g. causing damage) and to the human operators (e.g. causing fatigue or injury). Thus, it is important to be able to understand how structures vibrate and emit noise, i.e., their vibroacoustic behavior. Traditionally, engineers would try to describe the vibroacoustics using analytical methods. However, these are only possible for very simple structures. Structures that engineers confront in the aerospace, railway or maritime sectors are often made of composite panels that are connected together using complicated structural joints. The analysis of the vibroacoustics of such complex built-up structures cannot be performed analytically. Over the years, researchers have developed numerical techniques to solve this problem. Element-based methods (such as the finite element method) are well-developed and well-established methods with many commercial/in-house codes that can be used. However, aerospace, railway and maritime structures are relatively large. For example, a typical railway car can be modelled using the finite element method up to 500 Hz. Above this frequency, the size of the finite element model becomes too large, impractical and the associated computational cost becomes prohibitive. However, the audio frequency range is 20 Hz-20 kHz. At high frequency (above 10 kHz), the railway car can be modelled using energy-based methods such as the statistical energy analysis method. Energy-based statistical methods are valuable, but less well-established than element-based methods. The railway car example points to a frequency gap, indeed a mid-frequency gap, where neither element-based nor energy-based methods can be used.I am proposing to use wave methods to bridge the mid-frequency gap and to further strengthen energy methods. Waves provide a unifying, intuitive approach to vibroacoustics. The computational cost of a wave model is substantially small (especially when compared to a full finite element model), and the wave properties of structures can be obtained by post processing the finite element model of a small segment of an arbitrarily large structure.Thus, the goal of this programme is to develop a wave-based toolbox for modelling the vibroacoustics complex built-up structures. Industrial examples from the aerospace, railway and maritime sectors will be used to demonstrate the efficiency and effectiveness of the developed methods.