随着科技的发展，大型飞机、无人机群及高速列车等产生的气动噪声已成为一个迫切需要深入研究和解决的问题。研究气动噪声的主要方法有两种：试验和数值模拟。虽然前者在研究中不可或缺，但只能在设计后期进行，且费时费力。数值模拟在产品设计初期便可对噪声进行预估，提供降噪解决方案。然而，目前还没有一个有效的数值方法可以有效地用于大规模气动噪声分析。本项目将探索数值分析气动噪声的新方法。首先，通过计算流体力学计算，对近场气动噪声源进行分类建模；其次，研究新的快速气动声学边界元方法，基于声源模型对远场气动噪声进行有效数值评估并采用快速多极和自适应十字近似算法、区域分解、时域及频域插值、和高性能并行计算等方法对边界元加速；最后，对几个典型问题（大型飞机、无人机群和高速列车）建模计算，并进行试验验证。本项目将结合计算流体力学近场计算和边界元法远场计算的优势，为大规模气动噪声分析提供一个既准确又高效的数值模拟工具。

The noise produced by airplanes, unmanned aerial vehicles (UAVs), high-speed trains, wind-turbine farms and other man-made machines are increasingly causing troubles to people’s everyday life. This aeroacoustic problem needs to be studied and methods need to be developed to reduce such noise. Experimental methods with measurements and tests are crucial and important in studying the noise. But experimental methods are limited to the late stage of a product design, and often time-consuming and costly. Computational methods can be applied in such studies in the early stage of a product design. However, current computational methods are not effective and efficient for large-scale aeroacoustic modeling and simulation, such as in the analysis of the environmental noise in the scale of several kilometers around an airport or a high-speed train...In this project, we will study a new computational approach in modeling and simulation of large-scale aeroacoustic problems. First, we will explore the computational fluid dynamics (CFD) tools in near-field simulations in order to establish models of various noise sources in a few typical applications (for example, airplanes, trains, and UAVs at low and high speeds). Second, we will develop a new fast aeroacoustic boundary element method (BEM) for the far-field simulations based on the models of noise sources. The BEM will be accelerated with the fast multiple, adaptive cross approximation, new domain decomposition methods, time-domain and frequency-domain interpolations, and high-performance parallel computing algorithms. Finally, several model problems (an airplane, a group of UAVs, and an entire high-speed train) will be studied using the developed approach and the fast BEM solver, and the computational results will be validated using testing data. ..The proposed research is innovative and unique in that it combines the advantages of the CFD methods for near-field calculation and those of the BEM for far-field calculation. If successful, the outcome of the research project will be an effective and efficient computational tool for large-scale modeling and simulation of aeroacoustic problems, that can be applied by engineers in designing products with better sound quality to reduce the noises in a modern world...We plan to organize one workshop on computational aeroacoustics in Shenzhen in 2023, and co-organize an international conference on the boundary element method (IABEM Symposium) in Hong Kong in 2020. The research results will also be disseminated in journal publications, at national and international conferences, and introduced in undergraduate and graduate courses taught by the principal investigator.