Newsletter  2021.11  Index

Theme : "Mechanical Engineering Congress, 2021 Japan (MECJ-21)” Preface Masaaki MOTOZAWA, Hideo MORI Turbulence of viscoelastic fluid-from practical examples to turbulent coherent structure Yasuo KAWAGUCHI (Tokyo University of Science) Integration method of measurement and simulation in flow analysis and its applications Toshiyuki HAYASE (Tohoku University) Workshop on Experimental Fluid Dynamics (EFD) Chair: Shouichiro IIO (Shinshu University), Masaki FUCHIWAKI(Kyushu Institute of Technology), Ayumu INASAWA(Tokyo Metropolitan University), and Satoshi KIKUCHI (Gifu University) The Aerodynamic Noise not Made Clear With Acoustic Wind-tunnel Yoshiyuki MARUTA (Chuo University) Analysis of the flow-induced noise with the Extended Proper Orthogonal Decomposition Osamu TERASHIMA, Reon NISHIKAWA (Toyama Prefectural University), and Miyu OKUNO (Kanazawa University) Fan Noise Characteristics and Reduction Hidechito HAYASHI (Nagasaki University)

 

Analysis of the flow-induced noise with the Extended Proper Orthogonal Decomposition

Osamu TERASHIMA Toyama Prefectural University, Reon NISHIKAWA Toyama Prefectural University, Miyu OKUNO Kanazawa University

Abstract

In this presentation, we showed our newly developed method for the analysis of the relation between the flow information and emitted flow-induced noise directly using Proper Orthogonal Decomposition.

Even though the measurement of wind noise in low-noise wind tunnels and its numerical predictions are extensively and commonly performed nowadays, identification of the noise source and understanding the mechanism of its generation is potentially time-consuming because it is difficult to correlate the velocity field and far-field acoustic sound information precisely. Therefore, simple techniques have been developed to correlate these variables easily based on proper orthogonal decomposition (POD).

In this study, modal analyses were performed first using POD to extract the information of the flow field. Subsequently, extended POD was performed on the obtained data, and the dominant mode of the far-field sound pressure was determined (See Fig. 1). Using these methods, the speed, efficiency of identification, and the understanding of the generation mechanism, are expected to improve considerably.

We presented three examples of analysis results in the presentation. One was the flow-induced noise emitted from a circular cylinder, the second was that from a forward-facing step, and the other was that from a back-ward facing step. Simultaneous measurements of the velocity fluctuations in the wake of the circular cylinder or wall pressure fluctuations on the surface of the forward- and backward-facing step, and emitted far-field sound pressures were performed in a low-noise wind tunnel, using flows of 20-50 m/s. Then, the data obtained were analyzed using the proposed method.

Analysis results showed that the use of these methods allowed the determination of the dominant modes of the velocity fluctuations and surface pressure fluctuations from noise generated from the cylinder or forward- and backward-facing steps.

Key words

Flow-induced noise, Proper Orthogonal Decomposition, Extended Proper Orthogonal Decomposition, Noise source detection

Figures

Fig. 1 Procedure of the analysis of the flow-induced noise with extended POD analysis.