Aerodynamic noise and its passive control of subcritical flow past a cylinder

Abstract

Noise pollution poses dual challenges of annoyance and significant health risks. The turbulent airflow generated by motor vehicles, trains, flights, and industrial machines is considered as a primary contributor to environmental noise, prompting growingly stringent noise standards on these engineering applications and their machinery. Addressing this noise challenge necessitates a concerted effort to investigate the turbulent flow and its noise generation mechanism, as a critical priority for both industry and academia.
The study of flow past a cylinder represents a pivotal area of research in aerodynamics, which has significant relevance for numerous engineering applications such as train pantographs, automotive axles, and aircraft landing gears. However, comparing to aerodynamic and flow field studies of flow past a cylinder, there has been relatively less ones focusing on the aerodynamic noise, making it a valuable area for investigation. The advancement of aerodynamic and aeroacoustic research is driven by the integration of sophisticated computational fluid dynamics (CFD) methods, notably LES, computational aeroacoustuics (CAA) methods, e.g., Ffowcs Williams-Hawkings (FW-H), and advanced experimental techniques like particle image velocimetry (PIV). These complementary tools offer increasingly precise insights into the mechanisms underlying flow-induced noise and avenues for its reduction. Through a comprehensive analysis of both experimental and numerical findings, our understanding of aerodynamic noise generation can be further enriched. Moreover, recent attention has gravitated towards bio-inspired vibrissa cylinders and splitter plates as passive noise control methods. These configurations exhibit significant potential for noise reduction through vortex suppression and wake stabilization. This project conducts a comprehensive analysis of these passive control mechanisms, delving into their aerodynamic and aeroacoustic characteristics.
The outcomes of this project encompass several key findings. LES simulations of flow past a cylinder first demonstrate good agreements with experimental results. Then, coupling the FW-H method with LES results enables accurate prediction of far-field noise. Analyzing the relationship between near-field flow and far-field noise further enhances our understanding of noise generation mechanisms. Geometric modifications to the cylinder shape, including vibrissa shape and splitter plates configurations, perform noise reductions by about 13.2 dB and 7.4 dB compared with the circular cylinder case, further revealing insights into noise reduction mechanisms through the analyses of aeroacoustic and PIV experimental results. Additionally, LES/FW-H simulations for the vibrissa cylinder and splitter plates cases validate the noise reduction mechanisms and further elucidate the flow dynamics, i.e., the vibrissa cylinder causes a three-dimensional vortex shedding pattern and the splitter plate could weaken the fundamental vortex shedding. These findings contributes to advancing our understanding of flow past a cylinder, providing valuable insights into aerodynamic and aeroacoustic research, and paving the way for innovative noise mitigation techniques in various engineering applications.

Date of Award	1 Oct 2024
Original language	English
Awarding Institution	University of Bristol
Sponsors	Chinese Scholarship Council
Supervisor	Bin Zang (Supervisor) & Mahdi Azarpeyvand (Supervisor)