Small-scale wind tunnel data must undergo a number of corrections to arrive at simulated full-scale, aircraft flyover noise. The key metric for aircraft flyover certification is EPNL - effective perceived noise level. The scaling and extrap-olation process involves three steps:
1) Remove test geometry, atmospheric conditions, and convection effects from the data.
2) Scale the data to full-scale frequency and magnitude.
3) Simulate the flyover scenario by extrapolating the data to the proper dis-tance and standard day conditions, converting the resulting radiation pat-tern into a time dependant flight simulation, and calculating the appropri-ate metrics.
The method that follows is based on the assumption that the noise source is lo-cated at one position. If there are strong variations in noise source location with frequency, or if measurements were made in the near field of multiple sources, then each source must be handled independently.
Removal of Test Day Effects
To begin the scaling and extrapolation process, the data are corrected for mea-surement geometry, acoustic convection, and test day atmospheric effects. In effect, these steps take the acoustic waves to where they were emitted from the noise source.
Referring to Figure 1.21, the path R is the acoustic wave path as the sound propagates with wind convection to the microphone. Spherical spreading and atmospheric attenuation only apply to the motion of the acoustic wave relative to the airstream (distance R' in Equation 1.22). To extrapolate back to the source, the sound pressure lost to spherical spreading and atmospheric atten-uation is put back into the data:
i
p (fss, 8')SS = Lp (fss, 8)SS + 20 log R' + atd(fJ') R' (l.48) where the superscripts ss indicate small-scale model data.i
p (fss, 8')SS is the sound pressure level of the acoustic wave as it is emitted from the noise source at the emission angle, 8' (Equation 1.21), and measured frequency. Lp (fss, 8 ys is the sound pressure level measured in a closed-jet wind tunnel. In the case of an open-jet wind tunnel, Lp (PS, 8)SS represents the data measured outside the jet and corrected back to the jet by removal of shear layer effects as explained in the previous section (Equation 1.40).20 log R' adds back the sound pressure level lost to spherical spreading. The absence of a denominator in the argu-ment of the logarithm implies that the data are corrected to a sound level one unit distance (e.g., one meter) from the source. atd(fJ') R' adds back the en-ergy lost because of atmospheric attenuation from source to microphone. The atmospheric attenuation coefficient, atd, is found for the test day static pres-sure, static temperature, relative humidity (Shields and Bass 1977,ANSI 1995), and Doppler-shifted frequency (Equation 1.23).Figure 1.33 shows the magnitude of atmospheric absorption in dB per me-ter of propagation distance for a typical test day. At 35 kHz, the atmospheric absorption is greater than 1 dB per meter.
Scaling from Small Scale to Full Scale
The second step in the process is to scale the data to full-scale. Both the sound levels and frequencies are adjusted based on the scale factor, SF (ratio of small-scale model to full-small-scale aircraft characteristic length). We assume that the noise sources are sufficiently incoherent that an increase in source area causes an equivalent increase in acoustic power or pressure square (area doubling causes power doubling). Thus, for the sound level adjustment:
A A [ 1 ]
Lp (Jfs, 8')fs = Lp (fss, 8')SS + 20 10glO SF (1.49) Similarly, we assume that to first order the acoustic source wavelengths vary inversely with source dimension. This is based on the observation that aircraft noise sources such as vortex wake shedding and jet turbulence have Strouhal relationships that dictate an oscillation rate inversely proportional to body di-mension. Boundary layers grow and their oscillation rate decreases with body
3.5 3
2.5 a,dE/m
2
1.5
0.5
o o
20 40 60 80 100frequency, kHz
Fig. 1.33. Variation of atmospheric absorption with frequency, static pressure 101.353 kPa, static temperature 15.6°C, and relative humidity 70% (Shields and Bass 1977)
length. And cavity oscillations vary inversely with source dimension. This as-sumption breaks down if large Reynolds number changes are associated with the scaling, in which case secondary effects come into play (see Hayes et al.
1999). Thus, the recommended frequency adjustment is:
(1.50) where d is a linear source dimension and superscriptfs refers to full-scale con-ditions.
The sound pressure levels of Equation 1.49 are representative of sound one unit distance from the noise source.
Flyover Simulation
The final step in the scaling and extrapolation process is the flyover simulation of the scaled noise source. The first action is to extrapolate the acoustic data from the full-scale source to the appropriate distances for the flyover scenario.
In the following text, it is assumed that the source is moving at a nominal
sub-flight path
'\
sound emission locationsFig. 1.34. Flyover simulation geometry
sonic velocity, U, through still air along a linear path with no vehicle (source) acceleration (see Figure 1.34).
Because wind tunnel data are acquired at fixed locations, we assume that the source noise is emitted at a sequence of flight emission angles, 8'. These are the same as the wind tunnel emission angles as long as the orientation of the source relative to the airflow is the same for both the flyover simulation and wind tunnel test. If this orientation is different for the two cases, then an ad-justment should be made to the emission angles in order to obtain the correct source directivity pattern.
The data are extrapolated to the flyover distances,
Rf,
obtained from the geometry of Figure 1.34 as follows:Lp
(ffl,
8') =i
p (ffs, 8')fs - 20 log Rj - astd (HS) Rj (1.51) The second term on the right-hand-side (RHS) of Equation 1.51 corrects the sound levels obtained from Equation 1.49 for spherical spreading, and the third term on the RHS corrects the data for atmospheric absorption. The at-mospheric absorption coefficient, astd, is determined for standard day condi-tions at each full-scale Doppler shifted frequency (Equation 1.23). The left-hand-side of Equation 1.51 requires Doppler shifted frequencies because of the simulated motion between aircraft and observer. The spherical spreading is calculated relative to a distance of one meter, and the atmospheric absorp-tion is based on the entire distance from the source. The data now represent full-scale, extrapolated flyover data.Figure 1.35 illustrates a typical transformation of small-scale wind tunnel data to the corresponding full-scale flyover data. The third-octave band levels were calculated from the narrow band spectra using ANSI SI.II-1986 after the transformation/extrapolation from small-scale to full-scale was made.
140
frequency, 1/3 octave bands, Hz
100000
Fig. 1.35. Effect of scaling and extrapolating wind tunnel data to simulated flight at 497-m altitude
To calculate EPNL of the aircraft flyover, the third-octave sound levels are converted to perceived noise levels, PNL (or PNLT, if tone corrections are re-quired as explained below) versus time as witnessed by the observer. First, the perceived noisiness in units called noys, n, is found at each emission angle us-ing noy tables (FAA 1996, Smith 1989, ICAO 1998). Figure 1.36 contains curves of selected noy values versus frequency and sound pressure level taken from the tables. For example, a sound level of 100 dB in the 1000 Hz third-octave band has a noy value of 64.0. Clearly, high frequency sound is perceived to be more annoying and, therefore, to have greater noy values than low frequency sound with the same sound pressure level. The noy tables contain more detail than shown in Figure 1.36 and should be used for the computations. The FAA (1996) provides mathematical formulas for the noy tables.
The 24 noy values found for the third-octave band sound levels from 50 Hz to 10 kHz are summed as follows (FAA 1996, Smith 1989, ICAO 1998):
N{8') = 0.85 n (fJS)max + 0.15
r
24 ni{fJS) i=1where n (fff)max is the maximum noy value.
(1.52)
The perceived noise level, PNL, is calculated for each flight emission angle with the equation:
PNL (8') = 40 +
COg:~(2))
10g1o [N{8')] (1.53)140 120
~
8
100="
","
'"
80.S 01)
'"
'0
=
60"'0
.:::
01) 01)~ 40
Q.,
20
0 100 1000
100 dB
90 80
70---~
60-50
fi'equency, 1/3 octave bands, Hz
10000 Fig. 1.36. Perceived noisiness as a function of sound pressure level and frequency
The tone weighted perceived noise levels, PNLT, can be calculated from the PNL values by giving an extra weighting to noise that contain strong tones, which are perceived to be more irritating than broadband noise. For broad-band sounds, such as jet mixing noise, the tone correction is not used. In these cases, the PNLT algorithm must be used with caution (or avoided) because a rapid fall-off in the high frequency spectrum may trigger an erroneous tone correction.
To simulate a flyover, the wind tunnel emission angles illustrated in Figure 1.37 must be converted to the corresponding flight observer time. The time for the source to move to the emission location and the time for sound
propaga-Fig. 1.37. Geometry for transformation from flight emission angle to observer time
tion to the observer are required. The calculated observer time, ti , corre-sponding to a given emission angle is given by Equation 1.54.
Xi
Rfi
i is an emission angle counter that starts from i = 1 at the shallowest aircraft approach angle and is incremented through the emission angles. The perpen-dicular distance from the observer location to the flyover line is represented by y. Once the observer time has been calculated for each emission angle, thecal-culated PNL (or PNLT) versus observer time is used to calculate the effective perceived noise level, EPNL, as follows.
Figure 1.38 shows jet aircraft noise plotted as PNL versus observer time from scaled and extrapolated wind tunnel data. The curve was faired with a cubic spline curve fit.
If at all possible, the range of angles comprising the wind tunnel data should result in PNL (or PNLT) curves that cross the 10 dB-down line during ap-proach and departure. If data is lacking, the curve must be carefully extrapo-lated to the 10 dB-down line graphically as shown in Figure 1.38 or by follow-ing trends of similar, more complete data sets if available. A poor extrapolation can cause an error in the computed EPNL.
The EPNL computation requires a summation of PNL (or PNLT) values every 0.5 second that fall within 10 dB of the peak as follows (FAA 1996, Smith 1989, ICAO 1998):
[ 2d PNL(kl ]
EPNL = 10 10glO
?:
010-10- - 13 (1.57)where d is the number of half-second steps in the flyover period for which PNL is above the 10 dB-down line. If the value of PNL at the 10 dB-down point is 90 PNdB or less, the value of d may be taken as the time interval between the ini-tial and the final times for which PNL equals 90 PNdB. The EPNL units are decibels, often noted as EPNdB.
Not mentioned in the above discussion is any correction for Reynolds num-ber effect because that is more a simulation quality issue than a measurement
118
Fig. 1.38. PNL variation with observer time and the 10 dB-down line
issue. In general, we know that low Reynolds number flows can create jet plumes that lack adequate turbulence scale and broadband sound (Soderman and Allen 1992)_ Likewise, Hayes et aL (1999) showed that variations in Reynolds number can cause variations in airframe noise leveL
References
Accredited Standards Committee Sl (1995) Acoustics: Methods for Calculation of the Ab-sorption of Sound by the Atmosphere. ANSI S1.26-1995 (ASA 113-1995)
Ahuja, KK, Tester, BJ, and Tanna, HK (1978) Calculation of Far-Field Jet Noise Spectra from Near-Field Measurement Using True Source Location. AIAA 78-1153, AIAA 11th Fluid and Plasma Dynamics Conf., Seattle, WA
Ahuja, KK, Tanna. HK, and Tester, BJ (1981) An Experimental Study of Transmission, Re-flection and Scattering of Sound in a Free Jet Flight Simulation Facility and Comparison with Theory,J. Sound and Vib, 75(l),pp 51-85
Ahuja, KK, Tester, BJ, and Tanna, HK (1978) The Free Jet as a Simulator of Forward Velocity Effects on Jet Noise. NASA Contractor Report 3056
Allen, RM and Reed, DH (1992) Development of the Boeing Low Speed Aeroacoustic Facil-ity (LSAF). DGLRI AIAA 92-02-030, 14th DGLRI AIAA Aeroacoustics Conf., Aachen, Ger-many
Allen, CS and Soderman, PT (1993) Aeroacoustic Probe Design for Microphones to Reduce Flow-Induced Self-Noise. AIAA 93-4343, 15th AIAA Aeroacoustics Conf., Long Beach, CA Allen, CS and Soderman, PT (1998) Effect of Freestream Turbulence on the Flow-Induced Background Noise of In-Flow Microphones. AIAA/CEAS-98-2297, 4th AIAA/CEAS Aeroacoustics Conf., Toulouse, France
Allen, CS and Soderman, PT (1997) Scaling and Extrapolating Small-Scale In-Flow Wind Tunnel Jet Noise to Full-Scale Flyover Jet Noise. AIAA/CEAS 97-1602, 3'd AIAA/CEAS Aeroacoustics Conf., Atlanta, GA
Allen, CS, Vandra, K, and Soderman, PT (1995) Microphone Corrections for Accurate In-Flow Acoustic Measurements at High Frequency. CEAS/AIAA-95-150, 1st Joint CEASI AIAA Aeroacoustics Conf., Munich, Germany
American National Standard: Specifications for Octave-Band and Fractional-Octave-Band Analog and Digital Filters. ANSI S1.11-1986 (ASA 65-1986), revision of ANSI S1.11-1966 (R1976).
Amiet, RK (1975) Correction of Open Jet Wind Tunnel Measurements for Shear Layer Re-fraction.AIAA Paper 75-532
Anon.: International Standards and Recommended Practices Environmental Protection -Annex 16 to the Convention on International Civil Aviation, Vol. 1, Aircraft Noise. Second Edition - 1998. International Civil Aviation Organization (lCAO), Montreal.
Arnold, F (1999) A Cross Correlation Method for Sound Power Determination in Flow Ducts and Frequency Corrections for the Standardized In-Duct Method. 6th IntI. Con-gress on Sound and Vibration, Copenhagen, Denmark
Bauer, AB (1994) Airborne Sensors for Listening to Acoustic Signals. United States Patent 5,339,287
Bendat, JS and Piersol, AG (1980) Engineering Applications of Correlation and Spectral Analysis. John Wiley & Sons, New York
Brock, M (1986) Wind and Turbulence Noise of Turbulence Screen, Nose Cone and Sound Intensity Probe with Wind Screen, Bruel and Kjaer Technical Review, No.4
Bruel & Kjaer (1982) Condenser Microphones and Microphone Preamplifiers for Acoustic Measurements
Bruel & Kjaer (1971) Instructions and Applications for Quarter Inch Condenser Micro-phones Type 4l35/4l36
Bruel & Kjaer Instruments Inc (1989) Master Catalog Electronic Instruments. BF0l36-11, pg. 218 (measurement by Neise)
Candel, SM, Guedel, A, and Julienne, A (1976) Radiation, Refraction and Scattering of Acoustic Waves in a Free Shear Flow,AIAA Paper 76-544
Dassen, T, Holthusen, H, and Beukema, M (1996) Design and Testing of a Low Self-Noise Aerodynamic Microphone Forebody. AIAA 96-1711, 2nd AIAA/CEAS Aeroacoustics Conference
Dickinson, SC (1990) Appendage Hull Junction Flows and Their Application to Sailboat Keels. Proceedings of the 2nd Inter. Symposium on Performance Enhancements for Marine Applications, Newport, RI
Federal Aviation Agency (1996) Federal Aviation Regulation - Part 36. Noise Standards: Air-craft Type and Airworthiness Certification
Fields, RS Jr, Tso, J, and Soderman,. PT (1997) An Experimental Investigation of Flow Induced Tones of the Bruel & Kjaer In-Flow Microphones. AIAA-97-1674-CP. 3rd AIAA/CEAS Aeroacoustics Conf., Atlanta, GA
Frederiksen, E. and Christensen, JI (1998) Pressure Reciprocity Calibration - Instrumenta-tion, Results and Uncertainty. Bruel & Kjaer Technical Review No.1
Guedel, A (1985) Scattering of an Acoustic Field by a Free Jet Shear Layer. J. of Sound and Vibration, Vol. 100, No.2, pp. 285-304
Hayes, JA, Horne, WC, Jaeger, SM, and Soderman, PT (1999) Measurement of Reynolds Number Effect on Airframe Noise in the 12-Foot Pressure Wind TunneI.AIAA-99-1959, 5th AIAA/CEAS Aeroacoustics Conf., Seattle, WA
Herkes, WH and Stoker, RW (1998) Wind Tunnel Measurements of the Airframe Noise of a High-speed Civil Transport. AIAA 98-0472, 36th Aerospace Sciences Meeting & Exhibit, Reno,NV
Hersh, AS, Soderman, PT, and Hayden, RE (1974) Investigation of Acoustic Effects of Lead-ing-Edge Serrations on Airfoils. J. of Aircraft, Vol. 11, No.4
Horne, WC and James, KD (1999) Concepts for Reducing the Self-Noise of In-Flow Acous-tic Sensors and Arrays. AIAA 99-1815, 5th AIAA/CEAS AeroacousAcous-tics Conf., Bellevue, WA
International Electrotechnical Commission: Draft IEC 1094-3 (1995) Measurement Micro-phones, Part 3: Primary Method for Free-Field Calibration of Laboratory Standard Mi-crophones by the Reciprocity Technique. ref. 291294/DIS, project 2911094-3/Ed. l.
ISO DIS 5136 (1999) Acoustics - Determination of Sound Power Radiated into a Duct by Fans - Induct Method, (Revision ofISO 5136: 1990)
Jacob, MC, Robert, G, Fremion, N, and Guerrand, S (2000) In-flow Acoustic Measurements on High Speed Trains. AIAA-2000-2011, 6th AIAA/CEAS Aeroacoustics Conf., Lahaina, HI
Jaeger, SM, Allen, CS, and Soderman, PT (1995) Reduction of Background Noise in the NASA Ames 40- By 80- Foot Wind Tunnel. CEAS/AIAA-95-152, 1'1 CEAS/AIAA Aeroa-coustics Conf., Munich, Germany
Jaeger, SM, Horne, WC, and Allen, CS (2000) Effect of Surface Treatment on Array Micro-phone Self-Noise. AIAA-2000-1937, 6th AIAA/CEAS Aeroacoustics Conf., Lahaina, HI Juhl, P (1994) A Numerical Investigation of Standard Condenser Microphones. JS&V
177( 4), pp. 433-446
Krothapalli, A, Soderman, PT, Allen, CS, Hayes, JA, and Jaeger, SM (1996, 1997) Effects of Forward Flight on the Far-Field Noise of a Heated Supersonic Jet. AIAA 96-1720, 2nd AIAA/CEAS Aeroacoustics Conf., State College, PA,Also - AIAA Journal, Vol. 35, no. 6, pp 952-957
McMasters, JH, Nordvik, RH, Henderson, RH, and Sandvig, JH (1981) Two Airfoil Sections Designed for Low Reynolds Number. Technical Soaring, vol. VI, no. 4, XVIIth OSTIV Con-gress, Paderbon, Germany
Miles, J.W (1957) On the Reflection of Sound at an Interface of Relative Motion, JASA Vol.
29No.2
Miller, WR, Meecham, WC, and Ahtye, WF (1982) Large-Scale Model Measurements of Air-frame Noise Using Cross-Correlation Techniques. JASA 71 (3), pp. 591-599
Morse, PM and Ingard, KU (1968) Theoretical Acoustics. Princeton Univ. Press, Princeton, Nakamura, M, Komine, T, Tsuchiya, M, and Hald, J (1993) Measurement of Aerodynamic N.J
Noise using STSF. Bruel & Kjaer Application Note BO 0392-11
Neise, JW and Stahl, B (1979) The Flow Noise Level at Microphones in Flow Ducts. JS&V, 63(4)
Norum, TD, Garber, DP, Golub, RA, Santa-Maria, OL, and Orme, JS Supersonic Jet Noise from F-15 Active Aircraft Flyovers. NASA TP in progress.
Paterson, RW, Vogt, PG, Fink, MR, and Munch, CL (1972) Vortex Shedding Noise of Isolat-ed Airfoils. AIAA Paper 72-656, AIAA 5th Fluid and Plasma Dynamics Conf., Boston, MA
Plumblee, HE (Editor) (1976) Effects of Forward Velocity on Turbulent Jet Mixing Noise, NASA CR-2702
Ribner, HS (1957) Reflection, Transmission, and Amplification of Sound by a Moving Medium, JASA Vol. 29 No.4
Roshko, A (1968) Boundary-Layer Theory by H. Schlichiting, translated by J. Kestin, 6th edi-tion, McGraw-Hill Book Co., New York
Ross, R (1981) Spectral Broadening Effects in Open Wind Tunnels in Relation to Noise Assessment. AIAA Journal, Vol. 19, No.5, pp. 567-572
Savell, CT (1977) Precision and Accuracy 00 et Noise Measurements. AIAA 77-1302, AIAA 4th Aeroacoustics Conf., Atlanta
Schultz, KJ and Splettstoesser, WR (1992) Helicopter Main Rotor I Tail Rotor Noise Radia-tion Characteristics from Scaled Model Rotor Experiments in the DNW. DGLR/AIAA 92-02-023, DGLRI AIAA 14th Aeroacoustics Conf., Aachen, Germany
Sheplak, M, Seiner, JM, and Breuer, KS (1999) A MEMS Microphone for Aeroacoustic Mea-surements. AIAA 99-0606, 3]lh AIAA Aerospace Sciences Meeting & Exhibit, Reno, NV Shields, FD and Bass, HE (1977) Atmospheric Absorption of High Frequency Noise and
Ap-plications to Fractional-Octave Bands. NASA CR 2760
Smith, MT (1989) Aircraft Noise. Cambridge Univ. Press, Cambridge
Soderman, PT (1999) Aeroacoustic Research Techniques - Jets to Autos. FEDSM99-7240, 3rd ASMEIJSME Joint Fluids Engineering Conf., San Francisco, CA
Soderman, PT (1988) Sources and Levels of Background Noise in the NASA Ames 40-by 80-Foot Wind Tunnel - A Status Report. NASA TM 100077
Soderman, PT and Allen, CS (1992) On the Scaling of Small-Scale Jet Noise to Large Scale.
DGLRI AIAA 92-02-109, DGLRI AIAA 14th Aeroacoustics Conf., Aachen, Germany Soderman, PT and Hoglund, LE (1979) Wind-Tunnel Fan Noise Reduction Including Effects
of Turning Vanes on Noise Propagation. AIAA 79-0642, AIAA 5th Aeroacoustics Conf., Seattle, WA
Soderman, PT, Jaeger, SM, Hayes, JA, and Allen, CS (2000) Acoustic Performance of the 40-by 80-Foot Wind Tunnel Test Section Deep Acoustic Lining. AIAA 2000-1939, 6th AIAA/CEAS Aeroacoustics Conf., Lahaina, HI
Soderman, PT, Julienne, A, and Atencio, A (1991) J -85 Jet Engine Noise Measured in the ON-ERA SI Wind Tunnel and Extrapolated to Far Field. NASA TP 3053
Soderman, PT and Mort, KW (1983) Aeroacoustic Characteristics of a Large, Variable-Pitch, Variable-Speed Fan System. Inter-noise 83, Proceedings, Vol. 1, Edinburgh, Scotland, pp.
Soderman, PT and Mort, KW (1983) Aeroacoustic Characteristics of a Large, Variable-Pitch, Variable-Speed Fan System. Inter-noise 83, Proceedings, Vol. 1, Edinburgh, Scotland, pp.