Reduction of Background Noise in the NASA Ames 40- by 80- Foot Wind Tunnel Stephen M. Jaeger* Christopher S. Allen* Sterling Federal Systems NASA Ames Division Moffett Field, California and Paul T. Sodermant NASA Ames Research Center Moffett Field, California September 12, 1994 Abstract Background noise in both open-jet and closed wind tunnels adversely affects the signN-to-noise ratio of acoustic measurements. To measure the noise of increasingly quieter aircraft models, the background noise will have to be reduced by physical means or through signal processing. In a closed wind tunnel, such as the NASA Ames 40- by 80- Foot Wind Tunnel, the principle background noise sources can be classified as: 1) fan drive noise 2) microphone self-noise 3) aerodynamically induced noise from test-dependent hardware such as model struts and junctions, and 4) noise from the test section walls and vane set. This paper describes the steps taken to minimize the influence of each of these background noise sources in the 40 x 80.
Aeroacoustics Engineer, Member AIAA.
t Group Leader, Aeroacoustics and Test Techniques, Associate Fellow AIAA.
At low frequencies andat testsectionvelocitieslower than 120knots,the wind tunnel
fans arethe dominant sourceof tonal andbroadbandnoise.In the 40 x 80 most of this
noise is observedat frequenciesbelow 500 Hz.I Soundintensity mapping of the test
section,diffuser, and contractioncone,was done with the drive fans at flat pitch. The
resultsindicatedthat mostof thefan noisetendsto propagate upstream from the diffuser
exceptfor frequencies below500Hz wheresoundpropagates equallyfrom the contraction
coneandthe diffuser.The diffuseractslike a hornto allow efficientradiationof noiseinto
the testsection.The contractionconereflects noisebecause of the relatively abruptarea
change. Acoustictreatment on thevanesetsupstream of the contraction conealsoattenuate
someof the fan noise.
Partof the solution for reducingthedrive noiseof thewind tunnelis to lower the fan
speedbelow 180 RPM and control the test sectionvelocity with the fan blade angle. 2
Recent testsindicatethattheeffectof varyingthefan angleis mostadvantageous atlow test
sectionvelocities but has little effect at high velocities where the background noise is
dominated by the microphone self-noise.
Microphone self-noise appears to be the main source of background noise for measurements above 500 Hz. 2 Microphone self noise comes from several sources including: 1) boundary layer noise on the microphone, 2) noise created by onset turbulence striking the microphone 3 and 3) high frequency nose cone tones. One solution was the development of the FITE (Now-Induced Tone Eliminator) nose cone which eliminated high frequency tones generated at the nose cone screen, a Coherence measurements were made between two closely spaced in-flow microphones in the test section. The results show high coherence for frequencies below 500 Hz which is attributed to the wind tunnel fan tones. The maximum level of the coherence drops with increasing wind tunnel speed indicating the presence of increased incoherent self-noise. 5 The phase between the two microphones is also ambiguous above 500 Hz indicating the presence of uncorrelated self-noise at each microphone.
Noise created by test hardware in the test section have a wide range of characteristics depending on the particular source. Noise from microphone struts, model struts etc. have haystack-shaped frequency spectra, but sharp tones may also result from steady vortex shedding from cylinders and other objects. 6 During a test of various microphone strut designs, it was found that elimination of strut braces and junctions reduced the background noise significantly. This lead to the development of struts with maximum thickness airfoils
such as the McMasters-Henderson airfoil 7 which is designedto delay transition and
eliminatelaminarflow separation at high Reynoldsnumbers. Maximum thicknessairfoils
arestrongandcanoftenbeusedwithoutbraces.
Various signal processing methodshavebeenemployedfor discriminating against
background noiseto increase the signal-to-noise ratio.The paperwill includesomeresults
from cross-spectrummeasurements, multiple-element arrays, and time-delayed dual-
microphonemeasurements. With carefulattentiondevotedto elimination of background
noisesources andutilizing advanced signalprocessing methods, the signal-to-noise ratio in
the 40- by 80- Foot Wind Tunnel testsectionwill be maximizedfor the future needsof
aeroacoustic wind tunneltesting.
Reference_ .
Hudgins, L. H., "Advanced Sensor Study for NASA's 40' x 80' NFAC Wind Tunnel," Contract No. NAS2-13699, Final Report, March 1993.
, Soderman, P. T., "Sources and Levels of Background Noise in the NASA Ames 40- x 80- Foot Wind Tunnel - A Status Report," NASA TM 100077, May 1988.
.
Young, K. J., "Noise Characteristics of a 1/4-Inch Brtiel and Kjrer Microphone with UA 0385 Nose Cone in Airflow," Presented at the NASA Ames Workshop on Aeroacoustics Tunnel Testing Techniques, March 1979, also: NASA TM 84219, 1982.
.
Allen, C. S., and Soderman, P. T., "Aeroacoustic Probe Design for Microphone to Reduce Flow-Induced Self-Noise," AIAA 15th Aeroacoustics Conference, October 1993, AIAA Paper 93-4343.
Bendat, J. S., and Piersol, A. G., Engineering Applications of Correlation and Spectral Analysis, pp. 167 - 176, John Wiley & Sons, 1980.
o Brooks, T. F., Pope, D. S., and Marcolini, M. A., "Airfoil Self-Noise and Prediction," NASA Ref. Pub. 1218, July 1989.
o McMasters, J. H., Nordvik, R. H., Henderson, M. L., and Sandvig, J. H., "Two Airfoil Sections Designed for Low Reynolds Number," Technical Soaring, Vol. VI, No. 4, XVIIth OSTIV Congress, Paderborn, Germany, May 1981.