HIGH-FREQUENCY NOISE GENERATION IN SMALL AXIAL-FLOW FANS

Citation
Da. Quinlan et Ph. Bent, HIGH-FREQUENCY NOISE GENERATION IN SMALL AXIAL-FLOW FANS, Journal of sound and vibration, 218(2), 1998, pp. 177-204
Citations number
15
Categorie Soggetti
Acoustics
ISSN journal
0022460X
Volume
218
Issue
2
Year of publication
1998
Pages
177 - 204
Database
ISI
SICI code
0022-460X(1998)218:2<177:HNGISA>2.0.ZU;2-3
Abstract
This paper presents results from an investigation of the broadband sou rces of acoustic noise in small axial flow fans. Observations drawn fr om flow visualization experiments and fluid dynamic measurements indic ate that secondary flows are primary contributors to the broadband noi se generated by small axial flow fans. More specifically, how unsteadi ness associated with tip gap flows is identified as a primary source o f high frequency noise. As air is forced through the tip gap (i.e. the space between the rotating blade tip and the stationary housing), the flow rolls up forming vortices at the blade tip. These vortices conve ct into the blade passage and become the dominant source of unsteadine ss in the blade passage and at the fan exit plane. The data presented indicates that this turbulence is the dominant source of noise above 1 .5 kHz for the fan tested. The likely radiation mechanisms are trailin g edge scattering, and radiation from free turbulence and/or boundary layers. Three types of experiments were performed as part of the study . First, flow visualization tests were run in an attempt to obtain a s ubjective evaluation of the flowfield. Then, stationary and rotating h ot-wires were used to provide mean velocity and turbulence intensity d ata for non-radial flow components. Using the results from the flow vi sualization and hot-wire tests, possible noise generation mechanisms w ere postulated. Fan modifications were then made to test the viability of the proposed noise contributors. The addition of flanges to the bl ade tips and of fabric near the blade trailing edges provided up to a 9 dB decrease in sound power above 1 kHz. These results will need to b e coupled to an analogous study of low frequency noise generation (i.e . below 1 kHz) for a significant reduction in perceived noise level to be achieved. (C) 1998 Academic Press.