Large Eddy Simulation of Acoustical Sources in a Low Pressure Axial-Flow Fan Encountering Highly Turbulent Inflow

[+] Author and Article Information
Hauke Reese

Institute of Fluid- and Thermodynamics,  University of Siegen, D-57068 Siegen, Germanyhauke.reese@uni-siegen.de

Chisachi Kato

Institute of Industrial Science,  The University of Tokyo, 4-6-1 Kamaba, Meguro-ku, Tokyo, 153-8505, Japanckato@iis.u-tokyo.ac.jp

Thomas H. Carolus

Institute of Fluid- and Thermodynamics,  University of Siegen, D-57068 Siegen, Germanythomas.carolus@uni-siegen.de

J. Fluids Eng 129(3), 263-272 (Oct 05, 2006) (10 pages) doi:10.1115/1.2427077 History: Received October 25, 2005; Revised October 05, 2006

A large eddy simulation (LES) was applied to predict the unsteady flow in a low-speed axial-flow fan assembly subjected to a highly “turbulent” inflow that is generated by a turbulence grid placed upstream of the impeller. The dynamic Smagorinsky model (DSM) was used as the subgrid scale (SGS) model. A streamwise-upwind finite element method (FEM) with second-order accuracy in both time and space was applied as the discretization method together with a multi-frame of reference dynamic overset grid in order to take into account the effects of the blade-wake interactions. Based on a simple algebraic acoustical model for axial flow fans, the radiated sound power was also predicted by using the computed fluctuations in the blade force. The predicted turbulence intensity and its length scale downstream of the turbulence grid quantitatively agree with the experimental data measured by a hot-wire anemometry. The response of the blade to the inflow turbulence is also well predicted by the present LES in terms of the surface pressure fluctuations near the leading edge of the blade and the resulting sound power level. However, as soon as the effects of the turbulent boundary layer on the blades become important, the prediction tends to become inaccurate.

Copyright © 2007 by American Society of Mechanical Engineers
Your Session has timed out. Please sign back in to continue.



Grahic Jump Location
Figure 1

Fan assembly with main flow from right to left. Clean inflow (CI) condition is achieved by removing the turbulence generator.

Grahic Jump Location
Figure 2

Numerical grid of the complete flow domain (top) and details in the vicinity of the fan blades (bottom)

Grahic Jump Location
Figure 3

Overview of the numerical grid for the case without fan blades

Grahic Jump Location
Figure 4

Monitoring points in the reference plane 0.56 D downstream of the turbulence generator

Grahic Jump Location
Figure 5

Local turbulence intensity, integral length scale, and mean streamwise velocity behind the turbulence generator at three monitoring points: comparison of predictions (G1=1.5 million, G2=3 million grid elements) with measurements

Grahic Jump Location
Figure 6

Comparison of the predicted and measured distribution of the mean streamwise velocity normalized by the inflow velocity in the reference plane behind the turbulence generator. The dashed lines indicate the position of the struts of the turbulence generator.

Grahic Jump Location
Figure 7

Comparison of the predicted and measured power spectral density (PSDw) of the nondimensional axial velocity fluctuations at three monitoring points in the reference plane

Grahic Jump Location
Figure 8

Comparison of predicted total to static pressure rise and efficiency with measurements (16); G3=2.9 million, G4=3.3 million grid elements

Grahic Jump Location
Figure 9

Predicted fluctuations of impeller forces in a fully developed state

Grahic Jump Location
Figure 10

Root mean square values of static pressure coefficient cp; left: clean inflow (CI), right: highly turbulent inflow (HT)

Grahic Jump Location
Figure 11

Highly turbulent inflow (HT): instantaneous distributions of the static pressure coefficient (black lines) and surface streamlines (white lines) on a blade surface and relative velocity magnitude (black lines) on the mid-span plane r=0.37D (normalized by the impeller’s tip velocity) at four consecutive time instances

Grahic Jump Location
Figure 12

Comparisons of the predicted and measured (22) power spectral density levels of the wall pressure fluctuations on the blade suction-side at approximately mid span radius for clean (CI) and highly turbulent inflow (HT)

Grahic Jump Location
Figure 13

Surface pressure coefficient at three blade sections predicted by LES with numerical grid G4

Grahic Jump Location
Figure 14

Predicted power spectral density of the unsteady forces for clean inflow (CI) and highly turbulent inflow (HT) cases (upper: representative force on one blade, lower: force on the complete impeller)

Grahic Jump Location
Figure 15

Predicted and measured duct sound power spectral density (upper: predicted from the single blade force, middle: predicted from the overall impeller force, lower: measurements (25))




Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In