PREDICTION PROGRAM VALIDATION Final Report · 2013-08-31 · NASA CR-159333 AIRCRAFT NOISE PREDICTION PROGRAM VALIDATION Final Report Belur N. Shiva-kankara Boeing Commercial Airplane

NASA CR-159333

(NASA-CE-153333) AIRCRAPT N O I S E P B E D l C T I O N N89-34219 PhOGRAlY VIILIDAT1D:I F i n a l d e p o r t ( B o z i n g Comrercial A i r p l a n e Co,, Seat t l e ) l b 3 p iiC Ad8/HF A01

AIRCRAFT PROGRAM

C S C L 20A Unclas G3/71 29323

NOISE PREDICTION VALIDATION

Final Report

blur N. Shivrdrrnkan

B a n g Commerckl Airplane Company Seattle, Washington

Prepared for Langley Re%e8rch Center under contract Nul-15526

National Aeronautics and Space Admnistration

1980

https://ntrs.nasa.gov/search.jsp?R=19800025711 2020-06-26T19:54:14+00:00Z

NASA CR-159333

AIRCRAFT NOISE PREDICTION PROGRAM VALIDATION

Final Report

Belur N. Shiva-kankara

Boeing Commercial Airplane Company Seattle, Washidigton

Prepared for langley Research Center under contract NASl-15526

National Aeronaulics and Space Administration

1980

CONTENTS

1.0 SUMMARY ....................................................... 1 2.0 INTRODUCTION .................................................. 3 3.0 FLIGHT TEST DESCRIPTION ........................................ 4

3.1 General ....................................................... 4 3.2 TestSite ...................................................... 4 3.3 Test Airplane and Its Engines ...................................... 4 3.4 Flight Test Procedure ............................................ 5

5 3.5 Instrumentation ................................................ 5 3.5.1 Acoustic Instrumentation ................................... 5 3.5.2 Airplane Position Instrumentation ............................

3.5.3 Onboard Performance Instrumentation ........................ 6 7 3.5.4 Weather Instrumentation ................................... 7 3.6 Test Conditions .................................................

4.0 ACOUSTIC DATA REDUCTION AND ANALYSIS ........................ 9 4.1 Data Reduction ................................................ 9 4.2 Ensemble Averaging Program and Data Normalization Procedure .......... 9

4.2. I Ensemble Averaging Program ............................... 9 4.2.2 Data Normalization ....................................... 10

4.3 Typical Aircraft Noise Results ..................................... 1 1 4.3.1 Spectra and Directivities .................................... 1 1 4.3.2 Repeatability ............................................ 1 1

5.0 COMPARISON OF ANOPP PREDICTIONS WITH FLYOVER DATA . . . . . . . . . . 12 5.1 General ....................................................... 12 5.2 Modules and Options ............................................ 12 5.3 Spectral Comparisons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

5.3.1 Low-Frequency Region (50 to 200 Hz) ........................ 12 5.3.2 Midfrequency Region (400 to 1000 Hz) ........................ 12 5.3.3 High-Frequency Region (1000 to 10 000 Hz) . . . . . . . . . . . . . . . . . . . 13

5.4 Directivity Comparisons .......................................... 14 5.5 PNLT and EPNL Comparisons ..................................... 14

6.0 EVALUATION OF COMPONENT PREDICTIONS ......................... 16

6.3 Fan Noise Component . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 6.3.1 FanTones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 6.3.2 Buzzsaw and Fan Broadbase Noise . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

6.4 Core Noise Component 1 9

6.5 Turbine Noise Component . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

6.1 Component Separation Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1h 6.2 Jet Noise Component 17 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

6.6 Airframe Noise Component . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

CONTENTS (CONCLUDED)

Page

7.0 CONCLUDING HLSMAKKS .......................................... 21

APPENDIX .......................................................... 23

REFERE?;! ........................................................ 24

iv

FIGURES

Page

I 1 .

3 4 5

7 8

9

10

I 1 12 13 14 15

16 17 18 19

20

21 22 23 24

25 26 27 28 29 30

31

Sclieaiotic Diagram of the Test Airplane . Boeing 747-100 .................... Specifications of the Pr.ltt & Whitney Aircraft JT9D Engine and

the 747 Nacelle Schematic ........................................... General Layout of the Test Site ........................................ Closcup View of the Groi.nd Microphone Installation ....................... Typical Markings for Boeing Airplane Position and Attitude

Targct Markings for APACS and Details of Microphone Layout

Concept of the Ensemble Averaging Technique ............................ 32 Comparison Between Individual Microphone Spectra and the

Ensenible Averaged Spectrum ........................................... 33 Typicxl Flyover Noise Spectra a t Several Angles for Takeoff. Cutback, and Approach Conditions .................................... 34

Typical Flyover Noise Directivities at Various Frequencies for Takeoff. Cutback. and Approach Conditions ............................. 35

Data Repeatability .................................................. 38 Spectral Comparisons of Data and Prediction at Takeoff Power . . . . . . . . . . . . . . . . 39 Spectral Comparisons of Data and Predictions at Cutback Power ............... 44 Spectral Comparisons of Data and Predictions at Approach Power ............. 49 Effect of INCT = 0 and 1 in Fan Noise Source Module on Data

Prediction Spectral Comparisons ...................................... 54 SPL Directivity Comparisons of Data and Prediction at Takeoff Power .......... 59 SPL Directivity Comparisons of Data and Prediction at Cutback Power .......... 61 SPL Directivity Cornpansons of Data and Prediction at Approach Power . . . . . . . . 63 Fundamental Fan Tone (F1) Directivity Comparisons of Data and

Prediction ........................................................ 65 Second Harmonic Fan Tone (Fz) Directivity Comparisons of Data

andprediction .................................................... 68 PNLT Directivity Comparisons of Data and Prediction ...................... 71 PNLT Comparisons as a Function of Corrected RPM ........................ 81 EPNL Comparisons as a Function of Corrected RPM ........................ 86 Spectral Comparisons of 747/JT9D Flyover and Model Scale

JetNoiseData .................................................... 87 Normalized Spectra for Jet Noise Component Separation .................... 88 Overall SPL Comparisons as a Function of Primary Jet Velocity . . . . . . . . . . . . . . . 89 SPL Comparisons as a Function of Primary Jet Velocity ..................... 90 SPL Directivity Cornpansons of Predicted and Derived Flight Jet Noise . . . . . . . . . 93 SPL as a Function of Fan Tip Relative Mach Number ....................... 37 SPL Directivity Comparisons of Data and Prediction for Broadband

Fan Noise and Buzzsaw Noise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I O 1 SPL as a Function of Primary Jet Velocity for Core Noise Component Separation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

25

36 27 28

Cafiicra System (APACS) ............................................ 29

at Bayview Airport ................................................. 30

V

FIGURES (CONCLUDED)

Page

32 33

A1

A2

A3

A4

AS

A6

A7

SPL as a Function of Primary Jet Velocity for Low Frequencies ............. SPL as a Function of Airplane Velocity .................................

1 15 119

Spectrii Comparisons of Data and Prediction, Engine Power Setting

Spectral Comparisons of Data and Prediction, Engine Power Setting


Spectral Comparisons of Data and Rediction, Engine Power Setting




% N 1 ~ = 9 8 . 8 .................................................... 122

% N 1 ~ = 9 3 . 7 . .................................................. 127

% N 1 ~ = 8 9 . 5 .................................................... 132

% N l ~ = 9 1 . 0 .................................................... 137

% N 1 ~ = 9 0 . 7 .................................................... 142

96N1cr84.4 .................................................... 147

% N 1 ~ = 7 8 . 7 .................................................... 152

vi


Belur N. Shivashankara Boeing Commercial Airplane Company

1.0 SUMMARY

NASA has developed a modular computer program (ANOPP) for predicting aircraft flyover and sideline noise. The program prediction methodology is to predict one-third octave hand spectra at the source for the various significant components of the aircraft noise. The sound pressure levels (SPL's) are propagated to the observer before conversion LO subjective units such as perceived noise.

Although the prediction methods in the program are considered to be the best available to NASA at this time, a number of uncertainties remain about the accuracy of the program for predicting the noise of actual aircraft. This has led to a number of NASA contracts to evaluate ANOPP against flight noise data of the current wide-body fleet.

The purpose of the validation study was to (1) assemble a highquality flyover noise data base for aircraft that are representative of the U.S. commercial fleet, (2) determine the accuracy of ANOPP with respect to the data base, and (3) analyze the data for source and propagation effects in order to suggest improvements to the prediction methodology.

Ten test conditions were selected from an existing data base obtained during a Boeing 747 airplane noise test. The aircraft was powered by Pratt & Whitney Aircraft JT9D engines fitted with hardwall nacelles (no acoustic treatment): only the primary plug and sleeve downstream of the turbine were lined. These test conditions covered a range of 98.8% to 75.4% of the raied engine rpm. The acoustic data were measured during 122m (400 ft) level flyovers using 12 ground plane microphones placed in line under the flightpath. The airplane was flown in a clean configuration with 20' flaps and landing gear up. The data were analyzed using a 0.lsec integration time and an ensemble averaging technique that established accurate spectra and well-defined directivities.

The noise predictions were made for each of the selected conditions for all the major components and the total noise. The predicted spectra, SPL and PNLT directivities, arid effective perceived noise levels (EPNLs) were compared with corresponding measured data.

On a total airplane noise PNLT basis, ANOPP was found to overpredict in the forward arc and underpredict in the aft arc at all power settings 90% corrected rpm a i d higher. The magnitude of overprediction or underprediction was about 9 dB. At lower power settings such as approach (75.4% corrected rpm), PNLT was underpredicted by about 9 dB for all angles between 80' and 160".

For all power settings 90% corrected rpm and higher, the overprediction in the forward arc was due to the buzzsaw noise component. At lower power settings (approach), the predicted buzzsaw levels are appreciably lower and total airplane PNLT is not overpredicted in thc for-

ward arc. At aft angles underpredictions are obtained at all power settings. For high power, the aft angle underprediction is due to the jet noise component. At other aft angles, for all power settings, the underpredictions may be due to several components in addition to jet noise, including fan tones, core, and turbine noise.

On an EPNL basis, the prediction was by mere coincidence within 1 EPNdB of the measured value at high power. The close agreement was due to the overprediction of PNLT in the forward arc and the underprediction in the aft arc canceling each other. At approach powr , the total airplane EPNL was underpredicted by about 5 EPNdB.

Based on preliminary estimates of the noise source components, it was found tlrut thc jet, fan, core, and turbine noise modules need significant modifications t o improve llicir prcdic- tion accuracy of flyover noise for the 747 airplane with hardwall JTW crib' 'Incs.

2

2.0 INTRODUCI‘ION

The importance of accurate flight noise prediction for both present and future aircraft is achieving widespread recognition within the Government and aerospace industry. This need has led to the development of individual noise component (fan, jet, etc.) prediction methods. Flight noise prediction is the ultimate goal, although static-to-flight estimates are also relevant due to the need to run ground static tests for developmental engines.

NASA has developed a modular computer program called the Aircraft Noise Prediction Pro- gram (ANOPP) for predicting flyover and sideline noise from aircraft. The program prediction methodology is to predict one-third octave band spectra at the source for the various significant components of aircraft noise and to propagate these spectra to the observer before conversion to subjective units such as perceived noise.

Although the prediction methods in the program are considereu to be the best available to NASA at this time, a number of uncertainties remain about the accuracy of the predictions for actual aircraft. Many of the prediction methods, such as jet noise, are based on model data. Full-scale predictions require a range of parameters, such as Strouhal number, that ex- tends beyond the available model-scale data base. The prediction methods are based largely on static test data. Flight effects are added as a modification to the static predictions, usually through an oversimplified theoretical model such as a moving point source. Installation effects such as modification of flow fields around the source, wing and fuselage shielding of the sources, and interactions between the sources are not presently included within the program. These factors require that a study be made of the program’s ability to predict noise from full- scale aircraft in flight.

The purpose of this study was to (1) assemble a highquality flyover noise data base for aircraft that are representative of the U.S. commercial fleet, (2: determine the accuracy of ANOPP with respect to the data base, and (3) analyze the data for source and propagation effects in order to suggest improvements to the prediction methodology,

3

3.0 FLIGHT TEST DESCRIPTION

3.1 GENERAL

The full-scale data base on which noise prediction methods are based or by which they are confirmed has, historically, been inadequate. This has been true especially in thc flight case, because the motion of the airplane introduces several measurement and data analysis problems. Two of the most important problems are:

1. The constantly changing position of the airplane with resped to the ground-based microphone during the integration time required to get a stable noisc sycctrum

2. Uncertainties in the position of the airplane relative to the ground-based microphone

A flight test program designed to minimize the above problems and to obtain highquality noise data was conducted by the Boeing Commercial Airplane Company in 1977. These tests were performed with a “clean airplane configuration,” that is, with gear up and minimum flaps (20°), to reduce airframe noise. The airplane was flown over a linear array of ground plane microphones at nominally constant altitude, attitude, and power setting. A wide range of power settings was tested, from flight idle to full-power takeoff. The following sections provide a detailed description of the test site, instrumentation, data acquisition, and data reduction procedures. The test program, as far as is known, is Unique and is believed to provide data of high quality to be used in assessing the flight component noise levels for high bypass ratio engines.

3.2 TESTSITE

The fi&t test was conducted at Baywew Airport in the state of Washington. This airport was chosen since it met several requirements for the research flight test. The ambient noise ‘evels at this airport are quite low because of very little traffic in and out of the airport. The airport is iocated in relatively flat terrain, enabling the airplane to get into the level flyover mode well ahead of the first target location and stay in that mode beyond the last target (the placing of targets will be discussed in sec. 3.5.2). The mcway has a paved, hard, concrete surface that is ideally suited fGr ground microphone installation.

Further, the runway surface itself is virtually horizontal over its entire length, a desirable fea- ture ft.r level flyover acoustic measurements. The final and the most important consideration was the willingness of the airport authorities to allow one of their runways (designation 3/2 1 ) to be marked with targets and instrumented with microphones over much of the runway and its vicinity. During noise recordings, the traffic at the test site was monitored to assure that no extraneous noise sources were present.

3.3 TEST AIRPLANE AND ITS ENGINES

A Boeing 747-100 airplane equipped with four Pratt & Whitney Aircraft JT9D engines was utilized. The outlines of the airplane with inajor dimensions marked are presented in figure 1. Three of the four engines were of the JT9D-3A type. The fourth engine was a JT9D-7CN

4

trimmed to the -3A takeoff p ,wer setting. The JT9D-7CN engines are JT91)-3A engines that have been uprated to JT9D-7 rstings. The JT9D-3A engines are rated at a nominal static takeoff thrust of 193.5 kN (43 5\90 lb). They are twinspool turbofan engincs with a bypass ratio of 5 and were fitted with sho,-t fan duct nacelles. The engines have 2 high-prossurc turbine stages, 4 low-pressure turbine st,iges, and 3 and 1 1 stages in the intcrmcdialc- and high- pressure compressors. The front fan has one stage with no inlet guide vanes. SOJIW sa!icn 1 dc- tails of the JT9D-3A engine are providea k tigure 2.

The nominal gross weight of the airplane was 272 230 (600 000 lb) at tlic hcginr 0 1 thc test. All engines utilized hardwall inlets and ducts, with acoustic treatment only on prinihry plug and sleeve downstream of the turbine.

3.4 FLIGHT TEST PROCEDURE

The airplane was flown over the runway at a nominally constant altitude of 122m (400 ft) with all engines set at the same nominal power setting. The landing gear was up and the flaps were set at 20°, thereby exposing only one slot out of the three-slot flap mechanism. The pitch angle was chosen to allow level flyover at the desired airspeed.

The desired power setting, airspeed, and altitude wcre maintained for about 6 to 7 sec before approaching the microphone array and for a total time of about 20 sec. This assured the airplane “on condition” for directivity angles ranging from 20” to 160’ (re: engine inlet axis). The microphones were placL 1 along the centerline of the runway, and the airplane was flown such that the projection of the flightpath on the runway would coincide with the runway centerline. The flight path relative to the microphone array is shown in figure 3.

3.5 INSTRUMENTATION

Both ground-based and airborne instruments were used. The ground-based instrumentation included microphones for noise and weather instrumentation. The onboard instrumentation monitored several engine and airplane parameters. The weather was aka monitored using an instrumented light airplane. Detailed description of the instrumentation is provided in the following sections.

3.5.1 ACOUSTIC INSTRUMENTATION

Twelve ground microphones were placed along the centerline of the test range runway, with a 15.24m (50 ft) spacing between microphones (fig. 3). These microphone: were all 13 mm (112 in.) in size and were mounted on their sides on 0.61 x 0.61m ( - ft x 2 ft) aluminum plates (fig. 4).

Time synchronization for all microphones was provided by the lnter Range Instrumentation Group (IRIG) time code, and standard calibration procedures were Lsed. The time code generator on the ground was synchronized with the airborne time code generator.

3.5.2 AIRPLANE POSITION INSTRUMENTATION

Boeing Airplane Position and Attitude Camera System (APACS) was used for determination

5

of the airplane position relative to the microphones at any instant in time. This system (ref. 1) uses a 35-mm motion picture camera mounted in the belly of the airplane looking down- ward. The camera is synchronized with the IRIG time code generator and is operated at 5 or 10 frames/sec. The camera photographs the 1.2m square (4-ft square) targets that are marked in parallel lines on the gr0uF.d directly below the flightpath. Adjacent targets are csrefully positioned on lines perpendicular to the desired airplane flightpath and are about 15.2m (50 ft) apart. Typical target markings are shown in figure 5 .

Following a flight test, the movie films are developed and read using a film reader, such as the Benson-Lehner Telereadex t 3 p 29E film reader. Tk reader converts the x and y coordinates of the targets on fi!m and iilm frame time into punched cards for computer entry.

To calculate the airplane positicn in space from the film, it is necessary to know the airplane pitch and roll angles. The airplane attitude is obtained either by a flight test gyro system or by an inertial navigation system.

The punched card information, in combination with the pitch and roll data, is used to com- pute the attitude and off-center distance of the airplane from the runway centerline as well as the visual overhead time with respect to a given microphone.

For the present flight test, two APACS cameras were installed in the electronic door of the test airplane. Both were operatec, continuously during each flyover to photograph several positioning targets painted on the runway, as shown in figure 6, The targets we12 referenced to the U.S. Coast and Geodetic Survey Benchmark 82, located on the pavement on the approach end of runway 3. It can be seen that the targets cover a distance of 165 1 m (54 17 ft) along the flightpath. The 1 ?-microphone array occupies only 183m (600 ft) in the midpor- tion of this target range (fig. 6(b)). The spread of the target locations on either side of the microphone array is necessary to obtain airplane position information for both low and high acoustic emission angles,

The IRIG time code was used for synchronization of aircraft, engine, and noise data.

3.5.3 ONBOARD PERFORMANCE INSTRUMENTATION

Several engine operating parameters were measured online by the onboard data acquisition system. Normal performance parameters for aU engines were either monitored or generated by an engine performance program. The following are some of the parameters that were available for acoustic data analysis:

1. IRIG time

2. Primary and secondary jet velocities (ideal)

3. Fan, turbine, and nozzle pressure ratios

4. Fan rotor speed and fan tip relative Mach number

5 . Combustor inlet pressure and temperature

6

6. Zombustor exit temperature

7. Primary and secondary total temperatures

8. Primary and secondary mass Pow rates

For NASA ANOPP predictions, average performance parameters werc calwh!ct! lor cach 11 y- over and used as inputs to the program. Since the variation of performanic: paranictcrs over each flyover was reasonably small, use of an averaged set of performance paramctcrs l o r cach flyover was considered adequate.

3.5.4 WEATHER INSTKUMENTATION

Surface Measurements:

Surface weather instrumentation consisted of the following measurements at the indicated heights :

1. Ambient temperature: 1.22m (4 ft), 9.14m (30 ft)

2. Ambient relative humidity: 1.22m (4 ft), 9.14m (30 ft)

3. Ambient wind speed and direction: 9.14m (30 ft)

Verticr' Measurements:

Vertical profiles of air temperatures, relative humidity, and mind speed and direction were obtained during each tcst period. The temperature and relative humidity : I easurements were performed by Meteorological Research, Inc.. who used an instrumented light airplane. Wind profiles were obtained through a ground-based BIBAL station,

The entire test was conducted within the FAA FAR 36 weather window. The ambient temperature and relative humidity on the ground were nearly constant at 12.8' C (55" F) and 55%, to 60% relative humidity during all flyovers. The vertical temperature from ground to 300m (984 ft) elevation was nearly constant (+ 1" U) and the relative humidity varied less than 10% in 300m.

3.6 TEST CONDITIONS

Test conditions were chosen to cover a wide range of engine power settings. The test cases used for this prediction validation contract are shown in table 1 . Nominal gross weight of tliL airplane was 272 223 kg (600 000 Ib) at the beginning of the test. All flyovers were flown at a nominal altitude of 122m (400 ft) with the aircraft in a relatively clean configuration (20" flaps, landing gear retracted). Pitch angle was determined by the airplane velocity and engine rpm chosen.

7

as nunkr

1 2 3 4 5 6 7 8 9

10

8

Vdocity of Airplne 9% mk (ftk)

ma 110.6 (363) 983 mm (285) 93.7 93.9 ~308) 896 m.8 (324) 91 D 8 8 (324) 91.3 92.0 (302) QQ.7 908 (298) 84.4 31.1 4299) 78.7 93.6 (307) 75.4 92.7 ~304)

4.0 ACOUSTIC DATA REDUCTION AND ANALYSIS

The acoustic data reduction and analysis procedures were carefully chosen to eliminate as many data quality problems as possible. As a matter of fad, the instrumentation and flight test procedure were dictated by the data handling procedures selected.

4.1 DATA REDUCTION

The output from all 12 ground plane microphones were recorded simultaneously on analog magnetic tapes during the tlyover test. From these tapes the microphone signals were reduced to one-third octave band spectra at the W i n g Noise Technology Laboratory.

Data reduction into one-third octave bands was done one microphone at a time, starting with the fust microphone in the array. For each microphone, spectra were obtained at consecutive 0.10029 sec using an integraticin time of 0.1 sec. The difference between the consecutive spectra time interval and this integration time-0.00029 sec-is the time required for the computer to store the spectrum and restart the one-third octave band analyzer. The beginning and end of data reduction for each microphone was selected to cover at least 15' to 165' directivity angle based on the airplane position data. The one-third octave band spectra were stored on digital magnetic tape (DMT) to be used by the ensemble averaging program discussed in the next section.

A similar procedure was used for reducing the data into narrowband spectra. The microphone signals were processed to give 0- to IO-kHz narrowband spectra with a constant band- width of 37.5 Hz and an integration time of 0.1 sec. One spectrum was obtaked every 0.12 sec. As before, the difference of 0.02 sec represents the time required by the computer io store the data and restart the FFT analyzer. The narrowband spectra were also stored on DMT for further processing. The airplane position camera pictures of the targets were reduced to provide tables of targets and microphone locations, time at which the airplane was optically overhead each target and microphone, the airplane altitude and pitch angle at that location.

4.2 ENSEMBLE AVERAGING PROGRAM AND DATA NORMALIZATION PROCEDURE

4.2.1 ENSEMBLE AVEPYiGITVC PROGRAM

The production of a spectrum from a microphone signal requires a finite time called the sam- p h g time or integration time. In the flyover noise measurements, the airplane position is constantly changing relative to the stationary microphone on the ground during the integration time. This leads to a loss in spatial resolution or the ability to define a spectrum of a precise angle. For level flyovers the greatest loss of spatial resolution occurs at the overhead position. In the prcsent flyover-at a level 122m (400 ft) altitude and 9 1 m/s (300 ft/s) speed -this lass of spatial resolution was calculated as ?10.6' for a standard 0.5 sec integration time. This uncertainty can be reduced to t?. I ' by choosing a smaller integration time of 0. I sec, as was done in the present analysis.

9

Tht reduction of integration time from 0.5 to 0.1 stc, however, increases the statistical uncertainty in the spectd level estimates. To overcome this problem, the ensenible averaging technique is used. This technique is discusstd in detail in section 3.3.3 of reference 1.

The ensemble averaging concept is illustrated in figure 7. A straight-line array of microphones is &. The separation distance between adjacent microphones is chosen such that it is larger than the distance traveled by the airplane, in kvd flyover, duzing the chosen integration time. For example, in the present case, at 91 m/s (300 ft/s) the airplane traveled 9.lm (30 ft) during the chosen integration time of 0.1 sec. The spacing between microphones was 15.24m (50 ft). This assures that the signals measured at the microphones are statistically independent. During a level flyover a 0.1- integration spectrum can be obtained at each microphone for any desind emission angle by choosing appropriate delayed signals. Sin= all these spectra correspond to the same range and amissiOn angle, one could determine an average spectrum using the independent s p e c - from each microphone. The net result is vastly improvad s ta t i s t idy confidence. For the present case using a 12inicrophone array, the effective integration time for the average spectrum is 0.1 x 12 = 1.2 sec. Thus, by the use of lewel flyover and the 12-microphone array, it was poss'ble to increase the integration time From the standard 0.5 sec to 1.2 sec while reducing the spatial uncertainty from k10.6" to k2.lo. In f- 8,O.l-c spectra from several individual microphones are plotted along with the ensemble average of all 12 microphones. It can be seen that the data scatter has been greatly reduced by the ensemble averaging technique.

The ensemble averaging of the data is accomplished using a computer program. The inputs to the program are the one-third octave band (or narrowband) spectra with a 0.1%~ htegration time on digital magnetic tape as described in section 4.1. The optical overhead times for the targets and microphones are provided in a tabular form along with the corresponding airplane altitude, pitch angle, and physical location of the targets and microphones.

4.2.2 DATANORMALIZATION

The ensemble averaging program also norma!ized the data to the following : 0 122m (400 ft) altitude

0 Zero engine angle to the horizontal

0 Atmospheric propagation effects corrected to standard day, 2 5 O C (77" F), 70% relative humidity using ANSI procedure (ref. 2) for tones

No correction to ground microphone data to convert it to free field

The above data normalization is done prior to ensemble averaging, that is, on the raw data. This is a necessary step in the ensemble averaging technique since it removes small deviations of airplane altitude and attitude during a given flyover from the measurements made at each microphone and brings the SPL's at all the microphones in the array to a compatible basis.

10

4.3 TYPICAL AIRCRAFT NOISE RESULTS

4.3.1 SPECTRA AND DIRECITWTIES

The ensemble averaged data fer operation at nominal takeoff, cutback, and approach engine power conditions are presented in fuures 9 and 10. These data are nonnalized as described in section 4.2.2, Le., to 122m (400 ft) level flyover, zero pitch angle, standard day atmosphere, and as measured by ground plane microphones.

The spectra (fa. 9) show relatively smooth behavior in the lower frequency region for aft angles. This is achieved by the use of ground plane microphones which eliminate the ground interference modulations from the low-frequency region. Further, the use of the ensemble averaging techniq;... reduces random error, thus contributing to even smoother spectrum shapes.

Directivities at selected frequencies are presented in figure 10 for the same three power conditions. Once again, the data show smooth behavior.

The discussions of data trends and separation of total noise into various components is included in section 6.0.

4.3.2 REPEATABILITY

The data repeatability was good. Spectra at .I. -e angl;s are compared in figure 11 for two consecutive runs. The data can be seen to repeat within It1 dB at all frequencies up to 5 0 0 0 Hz. For the 160" case, there is a wide variation at frequencies abovc 8000 Hz. This is believed to be due to large atmospheric attenuation corrections made to high-frequency data to correct from test day to standard day conditions. At shallow angles to horizontal, the propagation path is large and when corrections are made to change the pitch angle to zero, the propagation distance corrections become appreciable. The change in propagation distance requires two corrections to be made to the SPL. The fm. ~herical divergence correction-is only about 1 dB for a pitch angle correction of 3" at 16b' directivity. However, the second correction-to standard day atmospheric conditions-is substantial at frequencies greater than SO00 Hz. Due to inherent limitations of the atmospheric coirtxtion procedure, very erroneous corrected SPL's are xcasionally obtained for shallow radiation angles and high frequencies.

1 1

5.0 COMPARISONS OF ANOPP PREDICTIONS WITH FLYOVER DATA

5.1 GENERAL

The NASA ANOPP was run to obtain predictions for total noise and several noise components. Comparisons of predictions with 747 flyover data wen made every 10" between 20" and 160° for 10 cases. The predictions and data provided in the comparisons are for ground plane microphones on an acoustically hard surface, 122m (400 ft) level overhead flyover, engines, zero pitch angle, and standard day (25' C, 7096 relative humidity) weather conditions. Comparisons are provided on a spectral basis followed by directivity comparisons. Compari- sons are also provided for PNLT and EPNL.

5.2 MODULES AND OPTIONS

The NASA ANOPP has some alternative source modules. Further, within each module there a s several user-selected options. The modules used and the options selected within each module are provided in table 2. The selections for the options were based on their appropri- ateness to the particular flight test situation.

5.3 SPECTRAL COMPARISONS

5.3.1 UIW-FREQUENCY REGION (50 to 200 Hz)

The total airplane noise in the low-frequency fegion is underpredicted at all power settings and angles (fig. 12, 13, and 14). The magnitude of underprediction varies from 5 to 15 dB. At high powers, the far aft angle (e&, 150") underprediction is due to the jet noise component. It will be shown in section 6.2 that at high power, jet noise may be underpredicted at all directivity angles. At low powers, the far aft angle underprediction is again mostly due to the jet noise component. However, at other angles, it may be due to underpredidion of core noise and airframe noise components.

5.3.2 MlllFREQUENCY REGION (250 to IO00 Hz)

At high power (fig. 12 and 13), the total airplane noise h the midfrequency region of the spectrum is overpredicted at all angles. The maximum overprediction-of the order of 10 dB -is fou.nrl in the forward arc. At approach power (fig. 14). SPL's in this frequency region are underpredicted by about 2 to 6 dB.

Exarninatio 1 4 the component predictions shows that fan noise is resmnsible for the OM:- predictim of the midfrequency region at high powers. The fan noise subcomponent expected at ?h .e frequencies is the buzzsaw noise. In the ANOPP FAN modde, INCT = 1 includes the i .A combination tones (i-e., buzzsaw) and INCT = 0 excludes it. The effect of selecting F' :CT = 1 and 0 upon the fan noise prediction is examined in figure IS. If INCT = 1, large werprediction is obtained. If INCT = 0, large underprediction of fan noise in the midfre-

queiwies is seen. Thus, it is clear that buzzsaw noiso, prediction is responsible for the over- predictisn of total noise in the midfrequencies at high engine power settings, At approach power, the buztsaw is generally not expected to be dominant and the prediction also appears

TABLE 2: MODULES AND OPTIONS

~ ~ ~-

NOISE COMPONENT

JET FAN a) Inlet FanNoise

b) Aft Fan Noise

CORE

TURBINE AIR FRAME

MODULE

JRSJET

FAN

FAN

GECORE

TUR AFM

OPTIONS SELECTED

IOPT = 6

IDBB = O IDRS = O IGV = O INBB = 1 INCT = l INDIS = O INRS = 1 IOBB = 1 lDRS = 1 IGV = O INBB = O INCT = O INOlS = O INRS = O

None needed in this

ITOPT = 2

N = 1 ITEWN = 1 ITEHTN = 1 ITEVTN = 1 ITEFN = 1 ILESN = O IMGN = O INGN = O ITYPW = 1

module

13

to lower the level of this component appreciably.

The underpredictions at low power in the midfrequency region could be due to core noise prediction being too low, at least in the aft quadrant. Other components, such as airframe noise and jet noise, may also be underpredicted

5.3.3 HIGH-FREQUENCY REGION (loo0 to 10 OOO Hz)

At high powers the fonward arc total noise is again too high due to erroneous buzzsaw noise component predictions.

At low power (approach), the forward arc is generally underpredicted except for the fan tones. From 90" aft, the entire! high-frequency region is underpredicted by up to about 10 dB.

5.4 DIRECIIVITY COMPARISONS

Measured SPL directivities are compared with ANOPP predictions for total and component noise in figures 16, 17, and 18 at the takeoff, cutback, and approach power settings. Four representative frequencies are included in these comparisons. Directivity comparisons of fan tone f i t and second harmonics (F 1 and F2) are presented in figures 19 and 20. These tones do not always occur in the Same onathird octave band since Doppler frequency shift is ex- perienced due to the motion of the airplane relative to the fixed microphones. In figures 19 and 20, the one-third octave band center frequencies at which F1 and F2 occur are also indicated.

The directivit.1 compvisons show that at 63 and 160 Hz (fig. 16(a), 17(a), 18(a)), the total noise is underpredicted at all three power settings. At 400 and 1000 Hz, forward arc noise is overpredicted 3r takeoff and cutback powers. At approach power, the total noise is generally underpredicted at thew frequencies for all directivity angles.

The directivity comparisons for fan first harmonic (fig. 19) show that at high powers the f b t harmonic is overpredicted in the forward arc and underpredicted in the aft arc. At approach power, the l int harmonic is underpredicted for all angles greater than 50". Similar results are obtained for the fa-i tone second harmonic, as shown in figure 20.

5.5 PNLT AND EPNL COMPARISONS

Predicted and measured PNLT's are compared in figures 21 and 22. The predictions include total noise and various components. Figures 21(b), (9, and (j) correspond to takeoff, cutback, and approach power settings, respectively.

At takeclff and cutback powers, the total noise PNLT is overpredicted from 30" to 70' and underpredicted for all directivity angles greater than 80". The large overprediction in the forward arc is due entircly to the erroneous buzzsaw noise prediction. At approach power setting, the total PNLT is underpredicted at most angles.

The total noise PNLT's for both data and prediction are plotted as a fmction of corrected

14

rpm at five directivity angles in feure 22. In generai, the agreement is poor, in particular at angles 90" and higher where PNLT's are underpredicted. In the forward arc, overpredictions are seen at all power settings 90% corrected rpm and higher.

Total and component noise EPNL's are compared with data as a function cf corrected rpm in figure 23. In this plot, the predicted and measured EPNL's are in close agreement at aU power settings except for two lower power settings where predicted values are low. If taken at face value this could lead to a rather deceptive conclusion regarding thc accuracy of ANOPP for predicting EPNL. The close agreement at high power is due not to the accuracy of the prediction but to iarge overprcdiction of PNLT in the forward arc and underprediction in the aft arc. When combined. the underpredicted and overpredicted PNLT's compensate and a relatively accurate EPNL prediction results.

At approach power, the total airplane EPNL is underpredicted by about 5 EPNdB.

The comparisons between measured data and ANOPP predictions indicate significant errors in the prediction of total noise of the 747 airplane. Since the predicted total noise is the s u m of predictions for individual components, it is necessary to separate the measured total noise in- to various components before determining which of the component predictions need im- provement. It is beyond the scope of this project to achieve a total separation of components from the measured total noise. Component separation for a limited range of frequencies and angles is attempted in order to provide some insight regarding the accuracy of ANOPP in predicting component noise levels.

6.1 COMPONENT SEPARATION TECHNIQUE

The separation of components from the flyover data is a very complex exercise. Components that dominate at certain angles and frequencies are relatively easy to identify. However, where several components contribute to comparable noise levels, it becomes difficult to ac- curatcly isolate them. In the present flyover data, for example, the fundamental fan tone F1 can be identified at almost all andes and power settings without much difficulty. Also, jet noise can be detected at takeoff to approach power at angles greater than 130" for the low- frequency end (50 to 80 Hz) of the spectrum. However, at frequencies greater than 80 Hz and at angles less than 130°, the levels of jet noise are not clearly identifiable and additional analysis becomes necessary.

In the aft arc, between 100" and 130" there is an indication of a reasonably strong nonjet noise source even at takeoff power. This could be corc noise or possibly aft-radiated fan noise. Again, more analysis is required to resolve this issue.

Inlet fan noise is another component that shows up as a rather dominant component in the flyover data. This is because hardwall nacelles were selected for tMs test. The buzzsaw noise or the inlet combination tones can be Seen at several angles in the 250- to 1000-Hz range although not as clearly as the fan tone F1.

There are no clearly defined paths to achieve source isolation and component separation. There are, however, a number of techniques that will permit a reasonable estimation of component noise levels to be made. They include the following:

1. Examine the flight spectra (several angles) and directivities (several frequencies) over a range of airplane/engine operating conditions, both on a one-third octave band and narrowband basis.

2. Plot flight data into various trend curves such as: 0 SPL (f,e) vs Log Vp for jet noise (where VP = primary jet velocity)

SPL-OASPL vs Log (Strouhal number) for jet noise SPL Cf,O) vs Log Va for airframe noise (where Va = airplane velocity)

0 SPL (f,O) vs Log (rpm) for fan noise o SPL vs relative tip Mach number for fan and buzzsaw noise

16

3. Compare data with cwliponent predictions. In the ideal case, rile y ~ r u ~ c r t u ~ r J U I I C L .

ivity or spectrum shapes will be correct such that only level adjustments as a function of power setting will be needed to get a good match between the data and prediction.

4. Compare model scale test results with the full-scale data. Boeing has model scale data for the JT9D jet noise, both static and with relative velocity.

5 . Examine static engine spectra and directivities. Boeing has conducted static tests with the JT9D engine, acquiring data on several sidelines for a wide range of power conditions.

6. Study measured static source locations for the engine. For example, jet noise sources will be located downstream of nozzle exit stations. Source locatioris may be determined by the Boeingdeveloped multiple sideline technique (ref. 3) or microphone correlation techniques.

7. Determine static-to-flight effects. Comparison of flight data with extrapolated static data will provide clues regarding the nature of the source.

8. Use correlation techniques including internal-to-far-field cross-correlation and coherence function analysis (ref. 4, 5, and 6) and far-field correlations as suggested in reference 7.

The component separation effort for this study was limited to three power conditions representative of takeoff, cutback, and apprcach. Techniques 1 through 7 were used to varying degrees, with emphasis on the fiwt three techniques.

In the following sections, the component separation and comparisons with ANOPP are discussed on a component-by-comgonent basis.

6.2 JET NOISE COMPONENT

Typical comparisons between JT9D model jet noise spectra and the 747 flyover data are presented in figure 24. The model data were measured in a wind tunnel and have been scaled and extrapolated to the full-scale conditions. Good agreement is seen at 140' and 150" in the low- to intermediate-frequency region of thc spectrum where jet noise is expected to dominate. At 120°, the model jet noise levels are lower than the total airplane levels indicating that sources other than jet noise either contribute to or dominate the airplane noise. It should be noted that the close agreement between wind tunnel model and full-scale flight data at 140' and 150' may have been somewhat fortuitous. The comparisons do, however, show that the spectral and directivity shapes obtained for model scale jet noise can be applied to full-scale flyover jet noise prediction. Carefully conducted model scale tests also provide static-to-flight effects and scaling laws for jet noise.

In figure 25, the difference between the SPL and OASPL (50 to 1000 Hz) is plottcd as a function of a Strouhal number based on relative velocity. The relative velocity. i n this figure, is the difference between the primary jet velocity and the airplane velocity. Under ideal conditions, all jet-noise-dominated portions of the spectra would collapse onto a single curve

17

since jet noise scales with Strmhal number. In figure 25, at 120’ and 150°, the 98.2% N 1c (takeoff) and 91.3% (cutback) spectra appear to be reasonably close to each other at the low- frequency end. The 75.4% N lC(approach) case, by this analysis, appears to be dominated by sources other than jet noise, even at the low-frequency end. This type of analysis was used by Blankenship, et al. (ref. 8) to separate the jet noise components from flyover data. De- tailed analysis of this type appears to be useful in separating the jet noise component.

Another method is to examine the trend of data with selected airplanelengine parameters. In figure 26, OASPL (50 to 1000 Hz) is plotted as a function of primary jet velocity. Most probable jet-noisedominated regions are indicated on the figure. SPL’s at two frequencies are plotted as a function of log (VP) in figure 27 for several directivity angles. Again, regions of probable jet noise domination are noted on the figure.

Using analysis techniques such as those illustrated in the previous paragraphs, levels of jet noise were deduced for the takeoff power condition. In figure 28, derived jet noise component SPL directivities are plotted along with the ANOPP predictions. The difference between the derived and predicted jet noise levels are also indicated in the figure.

The frequencies were limited to those identified as being clearly dominated by jet noise at aft arc angles. It is noted that judgments regarding ANOPP prediction accuracy will bt more valid in the aft arc region (120’ and above). At lower angles the flight jet noise levels are estimates and may change somewhat as more detailed analysis is accomplished.

In summary, the ANOPP prediction for jet noise is relatively inaccurate. The level is underpredicted at all angles and power settings (2 to 15 dB). The predicted SPL directivity shape is roughly similar to that of the derived jet.

6.3 FAN NOISE COMPONENT

6.3.1 FAN TONES

The SPL’s at the blade passage frequency (F1) and its second harmonic (F2) are compared at three power conditions in figures 19 and 20. These figures indicate that neither the levels nor the directivity shapes are predicted correctly by ANOPP for th,; fan tone first harmonic. At certain power conditions and directivity angles, the turbine fundamental tone (T1) also re- sides in the same one-third octave band as F2. Based on the present analysis, separation of T I and F2 does not appear to be possible. Source location studies using static full-scale test data may provide a way to separate these two components.

6.3.2 BUZZSAW AND FAN BROADBAND NOISE

The flight data shows regions where buzzsaw noise is dominant at higher power settings such as takeoff and cutback. To recognize this component, various frequencies were plotted agziiist fan tip relative Mach number for several directivity angles, as shown in figure 29. A rather steep increase in noise level as the relative tip Mach number approaches 1 .O from the subsonic region is an indication of the “cut-on”,of buzzsaw noise. In figurz 29, this can clearly be seeii at 630 and 1000 Hz.

18

In figures 30(a) and (b), data at takeoff power is compared with fan noise prediction at several frequencies typical of buzzsaw and broadband fan noise. An overprediction of 10 to 15 dB is seen in the forward arc. In the aft arc, the predictions agree reasonably with data at most frequencies. At cutback power similar results are obtained (figs. 30(c) and (d)). The overprediction in the forward arc has been shown in section 5.3 to be due to the buzzsaw noise component prediction. For broadband noise at approach power (tb. 30(e) and (f)), underpredictions of up to 10 to 20 dB are evident. However, at this power, tlic total roisc at these frequencies may inc.13e significant noise 'from components other tllittl I'an. I n sum- nldry, fan noise prediction appears to be generally inaccurate for toncs, huzxsuw, and hroad- band fan noise subcomponents.

6.4 CORE NOISE COMPONENTS

The spectral comparisons at takeoff power (fig. 12) show a potentially signilicitnt contribu- tion of nonjet noise components in the vicinity of 250 to 400 Hz at all angles between 60" and 120'. The directivity plots (fig. 16) show maxima at 60" and 100" for 400 Hz. Near- field sideline data for the complementqry static test showed a very flat directivity at these frequencies, indicating widely separated sources of comparable strength. The preliminary assessment is that the spectrum in the vicinity of 250 to 400 Hz is dominated by inlet-radiated fan noise in the forward arc, which perhaps peaks around 60'. The aft arc is dominated by core noise at these frequencies, with a peak most probably between 1 10" to 130".

At cutback power, the spectra of figures 13 and 17 show a similar result. At this power setting the levels of inlet fan noise in the forward arc are more pronounced than those for the takeoff case. Again, indications are that core noise dominates in the aft arc, peaking around 120' and in the frequency region near 250 to 400 Hz.

At approach power, spectra in figure 14 indicate a dominant component centered around 250 Hz. In figure 18, it is seen that 250 Hz peaks at 110". Also, figure 18 shows that 400-Hz directivity has two maxima, one at 40" and the other at 120". Again, this leads to the ex- pectation of the prominence of fan noise in the forward arc and core noise in the aft arc.

The flyover data and core noise prediction at several frequencies are plotted against log Vp in figure 31. At 70", 90", and 120", the data for the 250- and 31 5-Hz cases is close to that predicted for the core noise. This along with the observations made so far indicate possible core noise dominance in the 250- to 3 15-Hz and 90" to 120" regions at all power settings including takeoff power. Having recognized this, the effect of moving the entire predicted core noise spectruni to peak at a frequency one-third octave band lower and increasing its level by about 3 dB was examined using data shown in figure 12. This modification to the core noise prediction provided better data-to-prediction comparisons in the core-noisedominated region.

6.5 TURBINE NOISE COMPONENT

The spectral comparisons at takeoff power shown in figure 12 indicate that at 150" through 120" the turbine tone can be easily recognized in the data. The turbine tone levels at takeoff power are underpredicted by 15 to 20 dB at 150' and by 10 to 12 dB at 120". At cutback power at 150' (fig. 13), it is not straightforward to draw definite conclusions regarding the turbine tone since it happans t -I occur at the same one-third octave band (4000 €Iz) as the fan

19

tone F2. However, the measured level at 4000 Hz is about 5 dB greater than that for the fan fundamental blade passage frequency. The 4000 Hz may, therefore, be domina'2d by the turbine tone. If it is, then the prediction for turbine tone at 150' is low by about 20 dB. At other angles the separation of the turbine tone needs further analysis. At approach power, from figure 14, similar conclusions are reached at 150'. At 120' the turbine tone noise prediction may be about 5 to 7 dB lower than 747/JT9D flyover values.

The broadband turbine noise levels appear to be far below the total measured noise and, as such, no !cveis can be derived from data for this component. The predicted turbine tone level at aft arc angles is substantially lower than measured levels. Additive corrections are needed to improve the match between 747/JT9D flyover data and the ANOPP prediction.

6.6 AIRFRAME NOISE COMPONENT

For all test conditions reflected in this report, the airplane velocity (V,) variod h-om 86.9 to 110.6 m/s (285 to 363 ft/s). Although this may appear to be a masoliable tanEe of Va variation to study airframe noise, the situation is not straightforward. This u s e , as seen in table 1, the engine power settings were also varying simultaneously sc observed SPL changes may be due to changes in either or both engine noise and airfram,

'

Four low-frequency SPL's are plotted as a function of primary jet velocity in figure 32. In this case primary jet velocity is used only as a parameter indicative of enginc power setting. At 70' and 90°, the low-power data vary little with jet velocity (or engine power setting). Therefore, airframe noise may be important at low frequencies in the forward quadrant at or near approach power settings.

The SPL's are plotted as a function of airplane velocity in figure 33. In this figure, the prediction of ANOPP for airframe noise is shown. If the SPL is mainly due to airframe noise, then good collapse of data should be obtained when plotted as a function of airplane velocity. No such collapse is obvious in figure 33 and, therefore, no inferences rdn be drawn regarding the validity of the NASA ANOPP airframe noise predictions.

20

7.0 CONCLUDING REMARKS

NASA aircraft noise prediction program, ANOPP, was run to obtain predictions for 747 airplane total noise and the jet, core, turbine, fan, and airframe voise components. Progain inputs corresponded to flyover noise test conditions of a Boeing 747 airplane fitted with four Pratt & Whitney Aircraft JT9D turbofan engines. The predictions were compared with measured noise data in several ways, including:

0 One-third octave band spectra at 30", 60°, 90°, 120°, and 150"

0 Directivities at several frequencies

0 Directivities for fan tone first and second harmonics

Tone-corrected perceived noise level (PNLT) as a function of directivity angle

0 Effective perceived noise level (EPNL) as a function of engine power setting

On a total airplane noise PNLT basis, ANOPP was found to overpredict in the forward arc and underpredict in the aft arc at all power settings 90% corrected rpul and higher. The magnitude of overprediction or underprediction was about 9 :'?. At lower power settings such as approach (75.4% corrected rpm), PNLT was underpredicted oy about 9 dB for all angles between 80' and 160'.

On an EPNL basis, the prediction was by mere coincidence within 1 EPNdB of the measured value at high power. The close EPNL agreement is the result of overpredictions of PNLT in the forward arc and underpredictions in the aft arc canceling each other. At approach power, the total airplane EPNL was underpredicted by about 5 EPNdB.

To enable identification of the source modules that need correction and the magnitude of the correction method, an attempt was made to separate the measured noise into various components predicted by ANOPP. The scope of the contract precluded a detailed investigation in this regard. In the beginning of the contract work it was hoped that the directivity s h ~ p e and spectrum shape predicted by ANOPP would be generally correct. Under that circumstance, the flight component levels could be more easily identified. Improbed predictiolis could be achieved by applying appropriate corrections to the current predictions such as Ad9 versus angle or AdB versus frequency.

For jet and fan noise, at least, neither the predicted spectrum shape nor the directivity shape was correct and, therefore, component separation proved to I.,* very complicated. Based cln preliminary estimate of the components, the following conclusions are drawii on the validit;' of the NASA ANOPP program :

1. The jet noise is anderpredicted by a 3 to 7 dB in the aft arc (e.g., 150') at tlie low- frequency end of the spectrum where it can be clearly identified in the data. The prediction of je noise in the forward arc also appears too low ;Itliough accumtc x p - aration of jet noise in this region is difficult.

2. The fan tones arc overpredicted in the forward arc and underpredicted in the aft arc at takeoff and cutback power and underpredicted at most angles at approach power. The predicted directivity has two pesks, one in the aft arc and one in the forward arc whereas the data shows a prominent peak around 90'. At 90' underpredictions are of the order of 5 to 10 dB. The buusaw and broadband fan noise is overpredicted by 10 to 1 dB in the forward arc.

3. The core noise prediction will fit flyover data better for angles between 90' and 120' if about 3 dB is added to its power kvel and the entire spectruni is shifted to one- third octave band lower frequency at a11 power settings.

4. The turbine tone appears to be underpredicted in the far aft arc where it can be distinguished. Under several conditions, the turbine tone lies within the Same one-third octave band as the fan tone second harmonic, making it difficult to separate the two components. The broadband turbine noise could not be distinguished in the data.

5. No comments are made on the airframe noise prediction since it was not possible to - d y distinguish this component.

To better assess the modifications required to improve the ANOPP prediction, accurate separation of components is essential. A detailed examination would be cquired of all available static and wind tunnel model data as well as static engine and flight test data. The study would include d e f ~ t i o n of model and engine source locations and static-to-flight effects. This should aid in the identification of jet and noqjet noise components at certain angles and frequencies. Correlation and coherence techniques may also prove useful. Combining various techniques will make it possible to separate components accurately. Once components are separated and evaluated from various data bases, prediction procedures can be improved,

22

APPENDIX

Spectral comparisons of data and ANOPP predictions at seven test conditions art: provided in the appendix. These comparisons ah: in addition to those shown in fgures 12, 13, and 14 corresponding to takeoff, cutback, and approach power settings. The engine power setting for each condition is identified on the fiures. The modules and options used for ANOPP predictions are provided in table 2. For each condition, comparisons are shown at directivity angles of 30°, 60°, 90°, 120°, and 150". The SPL's in these figures correspond to:

0 122m (400 ft) altitude

0 Zero engine angle to the horizontal

Standard atmosphere (25" C, 70% relative humidity)

0 Ground plane microphones

Fourengines

REFERENCE3

1 .. Chun, K. S., Berman, C. B., and Cowan, S. J., “Effccts of Motion on Jet Exhaust Noise From Aimaft.” NASA CR-2701,1976.

2. American National Standard Method for the Calculation of the Absorption of Sound by the Atn~16phen. ANSI, S1.26,1978.

3. Jaeck, C. L., ”Static and W W T d Neu Fieid/Far Field Jet Noise Measurements From Model Scak SingleFbw Baseline and Suppressor Nozzles-Summary Report,” NASA CR-2841 , 1977.

4. Shivashankam, B. N., “Cas Turbine Engine Con N o k Source Isolation by Intmal-to- Far Field Correlations,” Joumal of Ahmuff. vol. 15, No. 9, September 1978, pp. 597- 600.

5. Karchmer, A., and Reshotko, M., “Core Noise Source DiagnosticS on a Turbofan Engine Using Correlation and Coherence Techniques,” NASA TMX-73535,1976.

6. Wag, Hun En, “The Application of Coherence Function Techniques for Noise Source Identification,” Ph.D. thesis, Purduc University, 1978.

7. ParthaSarathy, S. P.* Cuffel, R. F., and Massier, P. F., ‘Separation of Core N&se a d Jet Noise,” AlAA paper 79-0589, March 1979.

8. Blankenship, G. L.. Low, J. K. C., Watkins, J. A., and Merriman, J. E., “Effect of For- ward Motion on Engine Noise,” NASA CR-l34954,1977.

24

TY PL

INTAKE AND FRONT FAN

C(WPRESSOR ( 1NTERMEDt ATE)

CWRESSOR (HIGH-PRESSURE)

CWBUSTOR

TURilINE

RkT ING

Turbofan, 2-shaft, byp. ss -at.io 5.0.

)(o i n l e t gutde vanes, 1-sta e fan,pressure r a t i o 1.51:1, and a i r f l a r 678 kg 91495 lb)/sec, a t 3330 rpm.

3 rows o f blades driven by l o w pressure turbine a t 3330 rpm.

Variable i n l e t vanes, 3 rows o f variable stator blades, 8 FOW o f f i x e d blades driven by high pressure turbine a t 7415 rpin. Overall pressure ra t to 22:l. A i r mass f low 113 kg (250 lb)/sec a t 7415 rpa.

Annular. 20 fuel nozzles.

6 stages, 2 high pc s u r e (7415 rpn) and 4 low pressure (3330 rnn) . 193.5 kN

Ref: Af rcraf t Engines o f the World, Paul H. Wilkinson, Published by Paul H. Wilkinson, Washington, D . C . 20014

(43,500 lbf)/7415 h.p. turbine rpn/static, dry.

Figure Z.-Specifications of the Pratt & Whitney Aircraft JTSD Engine and the 747 Nacelle Schematic

26

0 &EVATION VIEW

DATA ON CONDITION

.?CAN V I M

DATA OFF CONDITION

-

Figure 3.-General Layout of the Test Site

168 m

27

AIRPLANE FLIGHT 111- --m--- PATH I ! , V

RUNWAY Q ---

29

Target Number

01 01 01 02 0103 0201 0202 0203 0301 0 302 0303 0401 0402 0403 0501 0502 0503 0504 0505 0601 0602 0603 0604 0605 0701 0702 0703 0801 0802 0803 0901 0902 0903

Stat ion

-500.00 -500.00 -500.00 400.00 400.00 400.00 1100. do 1100.03 1100.00 1800.00 1800.00 1800.00 2225.00 2225.00 2225.00 2225.00 2225.00 2600.00 2600.00 2600.00 2600.00 2600.00 3300.00 3300.00 3300.00 4000.00 4000.00 4000.00 4917.78

4917.78 4917.78

0:'fse t + so. -40.00

c, . 00 40.00 -40.00 0.00 40.00 -40.00 0.00 40.00 -40.00 0.00 40.00 -40.00 0.00 40.00 80.00 120.00 -40.00 0.00 40.00 80.00 120.00 -40.00 0.00 40.00 -40.00 0.00 40.00

-kO.OO 0.00 40.00

ULI > l A I I P

El c vt4 t I 011

Bo. 59 19. C'? 77.94 81.13 80.79 30.39 79.40 79 00 78.60 80.30 79-80

81.40 80.93 80.55 80.08

82.41 81.93 81.60 81.05 80.65 85.29 84.81 84.41 89.62 89.22 88.82

79 50

79-70

93.84 92.al 92.68

Dimensions for Station, O f b t and El.vation are in 0.3048 m

i

0101 OIOZ 010)

- I I I I I 4 -* 44 a

(8) TARGET LAYOUT OfflC1

Figure 6.-Target Markings for APACS and Details of Microphone Layout at Bayview Airport

30

OFFSET

2,-3.' I

2400 - 1

K

J t H

2300 -

2200 - 9 6

t

D 2100 - i zm- 0 c

'P 100

I

Nicrophone Elevation on Station I.D. Station Runway Centerline

A B C D E P G H J K L M x i P Q

1900.00 1950.00 2000.00 2050.00 2100.00 2150.00 2200.00 2250.00 2300.00 2350.00 2Loo. 00 2450.00 25G0.00 2510.00 2520.00

80.00 80.10

80.40 80.30

80.55 80.75 80.85 80.99 81.12 81.23 81.38 81.53 81 . 67 81.71 81.73

Dimtionr for Station and Elevation am in 0.3048 m

t ' 1900 -

(b) MICROPHONE POSITIONS

Figure 6. -(Concluded)

31

32

FREOUENCV, Hr

- SPECTRA FROM 6 INDIVIOUAL MICROPHONES

L_ ENSEMBLE AVERAGE OF ALL 12 MICROPHONES

Figure 8. -Comparison Betwen Individual Microphone Spectra and the Ensemble Averaged Spectrum

33

900

%"lc 0 98.2 (TAKEOFF)

0 91.3 (CUTBACK)

0 75.4 (APPROACH)

1200

150°

FREQUENCY Hz

Figure 9. -Typicat Ftyover Noise Spectra at Several Angtzs for Takeoff, Cutback, aa:d Approach Conditions

34

0 88.2 (TAKEOFF) 0 91.3 (CUTBACK) 0 75.4 IAPPROACH)

I

!ii n a f cr W

- i5

DIRECTIVITY AhiGLE dogmas DIRECTIVITY ANGLE degrms

(11 FREQUENCY 63 TO 260 Hr

Fuure 10. -Typical Flyover Noise Directivities at Various Frequencies for Takeoff, Cutback, and Approach Conditions

35

A 10 -*

t 6300 92

DlRECTlVlTY ANGLE dqpes 40 160

DIRECTIVITY ANGLE deipes

(b) FREQUENCY 400 TO 1600 H t

Figure 10. --(Continued)

36

4 0 P t?Z

w >

0 P a I k w 2 0

8 -

DIRECTIVITY ANGLE

~ I ~ ~ ~ I ~

4 0 6 0 8 0 !

' loo 120 140 lb DIRECTIVITY ANGLE

(d FREQUENCY 2500 TO l0,OOO Hz


37

FREQUENCY, Ht

Figure 1 l.-Data Repeatability

OCAS€ NUMBER 2 @ENGINE POWER SERING, X N1C - gaZ0

DATA 0 TOTAL PREDICTION CI TOTAL 0 JET

I AIRFRAME

f

i 10

FREQUENCY, Hz

(a) DIRECTIVITY ANGLE = 300

Figure 12. - Spectral Comparisons of Data and Prediction at Takeoff Power

39

.CASE NUMBER 2 OENGINE POWER SElTING, X NlC - 86.20

DATA 0 TOTAL PREDICTION D TOTAL Q E T 0 FAN A CORE D TURBINE I AIRFRAME

(b) DlRECTlVlfY ANGLE = so0

Figure 12. - (Continued)

40

f . m 10

zi t 8

z n E I

2 0

@CASE NUMBER 2 @ENGINE POWER SETTING, % N1C = B6.20

DATA 0 TOTAL PREDICTICIN D TOTAL e JET

I AIRFRAME

-

FREQUENCY, Hz

(c) DIRECTIVITY ANGLE = 900


41

OCAS€ NUMBER 2 @ENGINE POWER SETTING. K N1C -

DATA a TOTAL PREDICTION fl TOTAL

JET

I AIRFRAME

(d) DIRECTIVITY ANGLE = 1200


42

O W E NUMBER 2 @ENGINE POWER SETTING, W N1C = 8@,2o

DATA 0 TOTAL PREDICTION [II TOTAL e m 0 FAN A CORE D TURBINE V AIRFRAME

(e) DIRECTIVITY ANGLE = 1500

Figure 12. - (Concluded)

43

.CASE NUMBER 6

.ENGINE POWER SETTING, X N1C = 91.30 DATA Q TOTAL PREDICTION I9 TOTAL

6 FAN A CORE D TURBINE I AIRFRAME

e JET

(8) DIRECTIVITY ANGLE = 300

Figure 13. - Spectral Comparisons of Data and Predictions at Cutback Power

44

*CASE NUMBER 6 @ENGINE POWER SETTING, W N1C - 91.30

DATA (3 TOTAL PREDICTION

TOTAL e JET 0 FAN A CORE (3 TURBINE F AIRFRAME

f 10

i

00 FREQUENCY, Hz

(b) DIRECTIVITY ANGLE = 600


45

OCAS€ NUMBER 6 OENGINE POWER SETTING, % N1C 91.90

PATA 0 TOTAL

A CORE , D TURBINE

Q AIRFRAME

II

P m

n rr S t



46

OCAS€ NUMBER 6

A CORE . D TURBINE I AIRFRAME

OENGlNE POWER SETTING, % N1C - 91.30 DATA

I

I I

50 125 315 800 2000 FREOUENCY, Hz


Figure 13. - (Contittimi)

47

OCASE NUMBER 6 OENGINE POWEH SETTING, % NlC = 91.30

DATA (3 TOTAL PREDICTION 0 TOTAL 0 JET 0 FAN

P AIRFRAME

8

a a m IO

2

n f z

; i I- o 0 n pc I

z 0 z

50 125 315 800 2000 5000 12500 FREQUENCY, HL

(e) DIRECTIVITY ANGLE = 150°

Figure 13. -- (Concluded)

48

OCASE NUMBER lo OENGINE POWER SETTING, % N1C = 76.40

DATA 0 TOTAL PREDICTION

TOTAL JET

0 FAN A CORE D TURBINE I AIRFRAME

L

50 1 I

125 315 800 20 00 FREQUENCY, Hr

(a) DIRECTIVITY ANGLE = 300

%O@

Figure 14. - Spectral Comparisons of Data and Predictions at Approac,$ Power

49

OCAS€ NUMBER 1 OENOlNE POWER SElTlNG. X N1C - 7S.40

DATA 0 TOTAL PREOlCTlON 0 TOTAL @ JET

I AIRFRAME

8

i k2T 2 10

g i 5 Q

I a - 5

I'REQUENCY, Hz

(b) DIRECTIVITY ANGLE =


50

OCASE NUMBER 10 OENGINE POWER SETnNG, X N1C = 75.40

DATA (3 TOTAL PREDICTION fl TOTAL 0 JET

I AIRFRAME

8

i

$ L i T m 10

c 8 Q

I

W z 0

a F -

0 FREQUENCY, H t


Figure 14. - (Contintied)

5 1

@CASE NUMBER I OENGINE POWER SETTING, 96 N1C - 7 S M

DATA 0 TOTAL PREDICTION 0 TOTAL 0 -

II AIRFRAME

8

s g - t k

g T m 10

0 5

& t 0

E

FREQUENCY. Hz

(d) DIRECTIVITY ANGLE - lW


52

0- NUMBER Ul OENOlNE #JWER SETTING. X N1C - 78.70

DATA Q TOTAL PREDICTION CI TOTAL 0 JET 0 FAN A CORE D TURBINE I AIRFRAME

FREQUENCY, Hr



53

CASE NUMOER 2 ENOlNE MWER SETTING, 9b N1C = 962 DATA pREDlCNON Q TOTAL

PI FAN WTM INCT- 1 in INLET FAN NOSE WILMCTIOIY e FAN WITH INCT- 0 in INLET FAN NOIS PREMCTllMl

.

;-,...-&,it r FREQUENCY, Ht

DlREMlVtTY ANGLE - 3@ Figure 15. - Effect of INCT = 0 and 1 in Fan Noise Source Module on Data

Prediction W t r a l Comparisons

54

CASE NUMBER 2 ENGINE W E R SETTING. % N1C - 96.2


FAN WITH INCT - 1 h tNLET FAN NOISE PREDICTION 0 FAN WITH INCT - 0 in. INLET FAN NCISE PREDICTION

B

ii Sf m 10

a E I L

DIRECTIVITY ANGLE = 6@


5 5

CASE NUMBER 2 ENGINE #)MR SETTINO, % NlC - W.2

DATA Q TOTAL

fl FAN WITH INCT - 1 b INLET FAN NOISE PREOICTlON e FAN WITH INCT- 0 In INLET FAN NOISE PREOlCflON

~ I c m

DIRECTIVITY ANGLE =


56

CASE NUMBER 2 ENGINE POWER SETTING, % N1C - 88.2

DATA 0 TOTAL PREDICTION 13 FAN WITH INCT = 1 in INLET FAN NOISE PREDICTION 0 FAN WITH INCT = 0 in INLET FAN NOISE PREDICTION

FREQUENCY, H t

DlRECSlVlTV ANGLE - 1200


5 7

CASE NUMBER 2 ENGINE POWER SETTING. X N1C = Sa2

OATA 0 TOTAL PREDICTION PJ FAN WITH INCT - 1 in INLET FAN NOISE PREDICTION 0 FAN WITH INCT - 0 in INLET FAN NOISE PREDICTION

DIRECTIVITY ANGLE - IMP


58

OEENGINE POWER SETTING, X N1C = MEASUREMENT 0 TOTAL

PREDICTION c) TOTAL

JET 0 FAN A CORE Ip TURBINE P AIRFRAME

6 3 k

Gc18 D l R E C T i V l N ANGLE, DEGREES

160 Hr

DIRECTIVITY ANGLE, DEGREES

a) 63HtAND180Hz

Figure 16. - SPL Directivity Comparisons of Data and Prediction at Takeoff Power

59

OENGINE POWER SETTING, 96 N l C - a 2 MEASUREMFNT 0 TOTAL

PREDICTION (II TOTAL 0 JET 0 FAN A CORE D TURBINE V AIRFRAME

DIRECTIVITY ANGLE, DEGREES DIRECTIVITY ANGLE, DEGREES - b) 100 H t AND loo0 Hz


60

@ENGINE POWER SETTING, % N1C = 91.3 MEASUREMENT 0 TOTAL

PREDICTION D TOTAL Q JET 0 FAN A CORE D TURBINE I AIRFRAME

r

m U

rn 0

DIRECTIVITY ANGLE, DEGREES DI RECTlVlTY ANGLE, DEGREES

a) 63 Hz AND 160 Hz

Figure 17. - SPL Directivity Comparisons of Data and Prediction at Cutback Power

61

eENGlNE POWER SETTING, % N1C = 81.3 MEASUREMENT 0 TOTAL

PREDICTION 13 TOTAL 0 JET 0 FAN A CORE I3 TURBINE I AIRFRAME

DIRECTIVITY ANGLE, DEGREES OIRECTlVfTY ANGLE, DEGREES

b) 400 H t AND loo0 H t


62

OENGINE POWER SETTING, % N1C = 76.4 MEASUREMENT 0 TOTAL

PREDICTION Cl TOTAL 0 JET Q FAN A CORE D TURBINE P AIRFRAME

%

- . 63Hz

W

DlRECTiVlTY ANGLE, DEGREES DIRECTIVITY ANGLE, DEGREES 11 83 Hz AND 160 H t

Figure 18. - SPL Directivity Comparisons of Data and Prediction at Approach Power

c.3

OEffilNE POWER SETTING, X N1C - 76.4 MEASUREMENT 0 TOTAL

PREDICTION CI TOTAL Q JET 0 FAN A CORE [P TURBINE I AIRFRAME

5

DlRECTiVlN ANGLE, DEGREES

% A-


Fiwre 18. - (Concluded)

64

0 0

DIRECTIVITY ANGLE, dagraer

a) Takeoff Power

Figure 19. - Fundamental Fan Tone ( F and Prediction

Directivity Comparisons of Data

65

DATA (TOTAL) 0 O o o

DlRECTlVlW. ANGLE,

b) CutbrdrPowmr


66

Engirn Powsr Stting % N, = 75.4

DIRECTIVITY ANGLE, ab,*-

c) AppraachPower


T i

.

. o 0

0

L

DIRECTIVITY ANGLE, drpnrr

Fiwm 20. - Second Harmonic Fan Tone (Fd Directivity Comparisons of Data and Prediction

68

0 0

0

DIRECTIVITY ANGLE, &grees

b) CutbrckPower


69

i Q

0 e 0 0

8

Engin Powu Settiq Y N,, = 75.4

0

DATA(T0TAL) 0 O 0 0

0

@ a 0

FAN PREDICTION m .

0 3150Hz 8 0 W H r e0 5 0 0 0 H t

0

a 0

0 a

0

i[ 1

I I I I 1 I 1

20 120 140 180 40 L

60 80 100 I I I L

DIRECTIVITY ANGLE, .-

c) AppoachPower


70

OCASE NUMBER 1

DATA 0 TOTAL PREDICTION CI TOTAL

D TURBINE P AIRFRAME

m U

f 5 t

0 DIRECTIVITY ANGLE, DEGREES

(a) ENGINE POWER SETTING, % NlC = W.8

Figure 21. - PNL T Directivitv Comparisons of Data and Prediction

71

OCASE NUMBER 2

DATA 0 TOTAL PREDlCTlW 13 TOTAL ‘3 JET Q FAN A CORE D TURBINE 17 AIRFRAME

b

l * a . l , l . n . l . a . ,

40 60 80 100 120 140 160 180 DIRECTIVITY ANGLE, DEGREES

(b) ENGINE POWER SETTING, % N l C = m2


72

OCASE NUMBER 3

DATA 0 TOTAL PREDICTION L3 TOTAL 0 JET c3 FAN A CORE 13 TURBINE F AIRFRAME

I . l . 1 1 I a 1 a I - 1 - J


(C) ENGINE POWER SETTING, % N1C = 93.7


73

@CASE NUMBER 4

DATA 0 TOTAL

0 TOTAL @ JET Q FAN

0, TURBINE F AIRFRAME

A CORE

1 - . . * . I . . . l . . . ,

40 60 80 100 120 140 160 180

DI R ECTlVl TY ANGLE, DEGREES

(d) ENGINE POWER SETTING, 96 N1C = 89.5


74

OCAS€ NUMBER 5

DATA 0 TOTAL PREDICT104 [I] TOTAL 0 JET 0 FAN A CORE P TURBINE c7 AIRFRAME

a , m . n . a . i . ~ . a . ~ 40 60 80 100 120 140 160 180

DIqECTIVITY ANGLE, DEGREES

(0) ENGINE POWER SETTING, % N1C = 91.0


75

@CASE NUMBER 6

DATA 0 TOTAL PREDICTION 13 TOTAL 0 JET Q FAN 4 CORE U TURBINE C AIRFRAME

2b 40 60 80 100 120 140 160 1hO - * - . - . - . - . . 1 . . .

DI R ECTlVlTY ANGLE, DEGREES

(f) ENGINE POWER SETTING, % N l C = 91.3

Figure 2 1. - (Continued)

76

OCASE NUMBER 7


(0) ENGINE POWER SETTING, % N1C = 90.7


77

OCAS€ NUMER 8

) . ' - n - n * * . ' * ' - * - J 20 60 80 1CO 120 140 160 i80


{h) ENGINE POWER SETTING, % N l C = 84.4

Figur 3 2 1. - (Continuedl

78

.CASE NUMBER 9


TOTAL 0 JET 0 FAN

D TURBINE V AIRFRAME

A CORE

T 1 . 1 . 1 . 1 . 1 . l . 1

40 60 80 100 120 140 160 18, DlRECTlVlTY ANGLE, DEGREES

( i ) ENGINE POWER SETTING, % N l C = 78.7

Figure 21. - (Contintiedl

OCAS€ NUMBER 10

DATA 0 TOTAL PRED l CTl ON 13 TOTAL 0 JET 0 FAN A CORE C TURBINE F AIRFRAME

A I . n . s . l . l m ~ - . - l


Cj) ENGINE POWER SETTING, % N1C = 75.4


f 10

TOTAL AIRPLANE NOISE 0 DATA 0 PREDICTION

b a 41

CORRECTED RPM, X N ~ c

Figure 22. -- PNL T Compartkons as a Function of Corrected RPM


00 e 0 Q

CORRECTED RPM, 96 N1C

b) DimctiuityAngl0800


Q

82

mo f 5

f i

70


0

0

c1 1'1

0

80 90

CORRECTED RPM, % N ~ c

100

figure 22. - (Continued)

83

0 (3

Q

m p1

T 10


00

CORRECTED RPM, % N ~ c

d) Directivity Angb 12@

Figure 22. - iConti1tuedl

TOTAL AIRPLANE NOISE 0 DATA 0 PREOfCTlON

CORRECTED RPM, % N1C

e) Directivity Ai* 1500


as

% E

E W

Ai

W

2 P

13 e @ I [3 D

- A

0

0 A

v I C

V VV B - I

0

& I I 1

T i

MEASUREMENT 0 T.OTAL

PREDICTION

0

a

Q

CI TOTAL 0 JET Q FAN A CORE D TURBINE

AIRFRAME

Figure 23. - EPNL Comparisons as a Function of Corrected RPM

86

TAKE OFF POWER 0 747/JT9D DATA (TOTAL) 0 MODEL DATA EXTRAPOLATED TO FULL SCALE

6300

Figure 24. -Spectral Comparisons of 747/JT9D Flyover and hlodel Scale Jet Noise Data

87

0

-6

-10

-16

-20

-25

"

- 30

-361 -8

120"

1500

D

Figure 25. -Normalized Spectra for Jet Noise CoTponent Separation

88

TOTAL AIRPLANE NOISE

0 DATA 0 PREDICTION

0 00 0

0 0

0 0 0

0 0

Figure 26. - Overall SPL Comparisons as a Function of Primary Jet Velocity

0 DATA (TOTAL) 0 JET (PREDICTION)

0

0

I- 1-1

10 !a910 v,

(a) 63 Hz

Figure 27.--SPL Comparisons as a Function of Primary Jet Velocity

90

ODATA (TOTAL) 0 JET (PREDICTION)

m U

4 1 0

m 0

m

> w a

(br 125 Hz

Figure 2). -/Continued)

91

00 ’ TA (TOTAL) n J E 1 (PREDICTION)

m 0 m

0

(01 250 Hz


92 A

Engine Power !Setting. % N ~ c = 93.2 Airplane Veloccty 86.9 m/t (285 ft/d

!

Figure 28. -SPL Directivity Comparisons of Predicted and Derived Flight Jet Noise

93

Engine Power Setting, % NIc = 98.2 Airplane Velocity 86.9 m/t (285 fth)

DIRECTIVITY ANGLE, degrees

Figure 28. -(Contimed)

94

Engine Power Setting, '% N ~ c = 98.2 Airplane Velocity 86.9 m/s (285 f th)

U

DIRECTI\'ITY ANGLE, degrees

(d 125 Hz

Figure 28. -IContinuedi

9s

En(pin Powst Setting, X N ~ c = 98.2 Airpkn Velocity 88.9 ds (285 fds)

m


96

OATA (TOTALI

H t

- . . _- . .-

lo00 HI

0.76 0 . 1 0.95 1.05 1.15 1.25 1.35 FAN TIP RELATIVE MACH NUMBER

b) DIRECTIVITY ANGLE 30'

Fbure 29.-SPL as a Function of Fan T b Relative Mach Number

97

FAN TIP RELATIVE MACH NUMBER

L

(b) DIRECTIVITY ANGLE

Figure 29, -(Continued)

98

830 H t

r

- .__ - -

I_ _-

- -- .- .-/A ooo HZ ______ __.__ __ . -

I

.TA

(e) DIRECTIVITY ANGLE 90'

Figure 29. -(Continued)

99

9 b

250 Ha

.- -..

. -

- i i 1.25 1.35

FAN TIP RELATIVE MACH NUM3ER

(d) DIRECTIVITY ANGLE 120'


1 00

0 DATA (TOTAL) 0 FAN PREDICTION

1600 Hz

m P

& n z U m

> 0 0

LL

a

w

2 n i

. k w 2 0

t

lo00 Hr

0


I ) TAKEOFF POWER 1OOO-1800 H t

Figure 30. - SPL Directivity Comparisons of Data and Prediction for Broadband Fan Noise and Buzzsaw Noise

101

0 OATA (TOTAL) 0 FAN PREDICTION

800 HZ


b) TAKEOFF POWER MK)-8o(H(z


102


1600 Hz

1260 Hr

Hz


c) CUTBACK POWER looQle00 Hz


103


8QO Ht

t 630 Hx

t 500 Hz


d) CUTBACK POWER 500.800 Hz


104

0 DATA (TOTAL) 0 FAN&DlCTION

rn 9

e 2

f i

z 8

!i

a m w

c-

0 a I

z 0


e) APPROACH POWER 800-1250 Hz


105

0 DATAO’OTAC) 0 FANPREDICTION

lo0


l i 0

fl APPROACH POWER 400630 Hz


106

0 DATA (TOTAL) 0 ANOPP PREDICTION (CORE)

(d DIRECTIVITY ANGLE 700; 250 TO 500 HZ

Figure 31.-PL as a Function Jf Primary Jet Velocity for Core Noise Component Separation

107


630 Hz

a k 0

u z 0 'F lo00 Hz

0

#. t .*.

1250 Hz

(b) DIRECTIVITY ANGLE 700; 630 TO 1250 Hz

Figure 3 1. -(Continued)

108

0 OATA (TOTAL) 0 ANOPP PREDICTION (CORE1

8 i 0

10

1) @ 6 0

QOO

2.

315 Hz

1 i 10

aP 9

(3

. . a 1

EE!? i f 10

(d DlRElZTlVlTY ANGLE mo; 2W TO 5on Hr

Figurc 71. -(Continued)

109

B i 0 2 s 3 ti w

0 0 a - tu 2 0


630 Hz

os",' ,

C. 0

1350 Hz

(dl DIRECTIVITY ANGLE SOo; 630 TO 12#) HZ


110


400 Hz

(e) DIRECTIVITY ANGLE lM; 250 TO 500 Hr

Figure 31.-(Continued)

1 1 1

0 DATA (TOTAL) 0 ANOPP PREDICTION JCORE)

630 Hr 1

i 10

0

. I J d

a s t

lo00 Hz ?

j 10

? i 10

1250 Hr i f 10

(f) P'. tCTlVlTY ANGLE 1200; 630 TO 1250 Hz


112

O DATA (TOTAL) 0 ANOPP PREDICTION (CORE)

400 Hz

315 Hr

(0) DIRECTIVITY ANGLE 140'; 250 TO 500 Hr

Figure 31.4Continuedl

113

CJ DATA (TOTAL) 0 ANOPP PREDICTION (CORE)

w > 4

(h) DIRECTIVITY ANGLE 140'; 630 TO 1260 Hz

Figure 3 1. - - (Concluded)

114

0 DATA (TOTAL)

m U

63 Hz - '? .f 10

0 0

> a

(11 DIRECTIVITY ANGLE 70'

FiQure 32.-SPL as a Function of Primary Jet Velocity :or Low Frequencies

115

0 DATA (TOTAL)

63 Hz - 80 H t -

0 0

00

10 Log10 vp

(b) DIRECTIVITY ANGLE Boo


116

0 DATA (TOTAL)

63 iit

0 0 Q

0 Q

0

-1 .f 10

0

!25 Hz

8 8 O

0

(c) DIRECTIVITY ANGLE 120'


117

0 DATA (TOTAL)

63 Hz

8 00 Q

0 0

00 Q

80 Hr -

a 00 0

0 0

0

10 Log10 v p

(d) DIRECTIVITY ANGLE 160'

Figure 32. -IConcluded)

118

0 DATA(T0TAL) - ANOQP AIRFRAME NOISE PREDICTION

so . - 4 100 . 116 AIRPLANE VELOCITY mls

8

(01 DIRECTIVITY ANGLE 30'

so 100 1 i o .1

AlRPLANE VELOCITY m/s

Figure 33. -SPL as 6 Function of Airplane Velocity

a

0 DATA(T0TAL) - ANOPP AIRFRAME NOISE PREDICTION

W k

F 10

4

a o k

90 100 110 AIRPLANE VELOCITV m/s

8

(b) DIRECTIVITY ANGLE

100 110 * 90 AIRPLANE VELOCITY Ws

Figun? 33. -(Continuedl

120

0 OATA(T0TAL) - MK)n AIRFRAME WISE PREMCllON

a- loo 110 AIRPLANE VELOCITY nJt

a _.

90 100 110 AIRPLANE VELOCITY mh

121

(4 DIRECTIVITY ANGLE = 300 Figurn A 1. - Spectral Comparisons of Data and Prediction, Engine Power setting

% N l ~ ' 9 8 . 8

122

*CASENUMBER 1 bEff i lNE HMUER SETTINO, K NlC - Qcy)

DATA 0 TOTAL PREDICTION El TOTAL 6 JET

P AIRFRAME

8

i

U g i

9

%f m 10

b 0

t

FREQUENCY, Hz

(bl DIRECTIVITY ANGLE = 600

rigwe A I . - (Continued)

123

O W E NUMBER 1 .ENGINE POWER SElTlNG, X N1C - ocbo

DATA 0 TOTAL PREDICTION cf TOTAL 0 JET 0 FAN r

P AIRFRAME

8

f T Po 10

I' 0 E I k cir B

FREQUENCY, Hr

(c) DIRECTIVITY ANGLE - 900

Figum A 1. - (Continued)

124

OCAS€ NUMBER 1 eENGlNE POWER SETTING, 9b N1C = Quo

DATA 0 TOTAL PREDICTION Cl TOTAL 0 JET

I AIRFRAME

9

f. rn 10 w z + 5

!2 L

0

I

z 0

FREQUENCY, Hr

(d) DlRECTlVlTV ANGLE = lZOO

Figure A 1. - (Continued)

I25

*CASE NUMBER 1 *ENGINE POWER SETlING, K N1C -

DATA 0 TOTAL PREDICTION p1 TOTAL *JET

E k I N E P AIRFRAME

(e) DlRECTlVlTY ANGLE = 1500

Figure A 1. - (Concluded)

I26

w > a k

OCASENUMBER 3 OENGINE POWER SETTING. % NlC = a 7 0

DATA Q TOTAL PREDICTION c1 TOTAL 0 JET

I AIRFRAME

f 10

i

FREQUENCY, Hr

(a) DIRECTIVITY ANGLE = 300 Figure A2. - Spectral Comparisons of Data and Prediction, Engine Power Setting

% N I C = 93.7

O W E NUMBER 3 OENGINE POWER SETTING, % N1C -

DATA 0 TOTAL PREDICTION CI TOTAL 0 JET 0 FAN A CORE D TURBINE

AIRFRAME


Figure A2. - (Continued)

128

OCASE NUMBER 3 OENGINE POWER SETTING, % N1C - @3.70

DATA 0 TOTAL PREDICTION II] TOTAL 0 JET 0 FAN A CORE 0 TURBINE I AIRFRAME



129

OCASE NUMBER 3 OENGINE POWER SETTING, X NlC * 83.70

DATA (9 TOTAL PREDICTION [II TOTAL 0 JET 0 FAN A CORE D TURBINE I AIRFRAME



*CASE NUMBER 3 .ENGINE POWER SEITING, 95 N1C = B3.70

DATA 0 TOTAL

c1 TOTAL 0 JET

1 &NE I AIRFRAME

(e) DlRECTlVlTY ANGLE = 1500

Figure A2. - (Concluded)

131

O W E NUMBER 4 OENGINE P9WER SETTING, X NlC = 89.60

DATA 0 TOTAL PREDICTION (I] TOTAL 0 JET

N E I N E

t

i n f a m 10


% N I C = 89.5

O W E NUMBER 4 OEWINE POWER SETTING, K N1C = OaW


V AIRFRAME

(b) DIRECTIVITY ANGLE = 60°


.ENGINE POWER SETTING, 96 N1C = 1460 DATA O r o T A L PREDICTION tl TOTAL +JET

I AIRFRAME

f

i 10

FREQUENCY. Hz



134

.CASE NLlMBER 4 OENGINE POWER SEmNG, X N1C - l)bo

DATA 0 TOTAL PREDICTION CI TOTAL QJET 6 FAN A CORE 13 TURBINE I AIRFRAME

(d) DIRECTIVITY ANGLE = 1 s

F &e A3. - (Continued)

135

.CASE NUMBER 4

.ENGINE IWER smm, w Nic = mo DATA 0 TOTAL PREDICTION

* C I TOTAL 0 JET

A CORE Ip TURBINE F AIRFRAME

e FAN



136

8

*CASE NUMBER 6 *ENGINE POWER SETTING, X N1C = 81.00


I AIRFRAME

f

1, 10

FREQUENCY, Hz


% NIC = 91.0

137

OCAS NUMBER 6 OENGINE POWER SElTlNo, % N1C - 0I.W

DATA 0 TOTAL

TOTAL e JET 0 FAN A CORE D TURBINE P AIRFRAME

50 800- 2000 FREOUENCY, Hz

(b) DIRECTIVITY ANGLE - 600 Figure A4. - (Continued)

138

O W E NUMBER IC OLNGlNL #mER SETTING, X NlC = 9140

PATA 0 TOTAL PREQICTION Cl TOTAL OJET 0 FAN A CORE b TURBINE I AIRFRAME

L I

50 125 315 800 12500 FREQUENCY. HZ


Figure A4. -- (Continued)

I39

(d) DlRECTlVlTV ANGLE - l2tP


140

.CASE NUMBER 5 aENGlNE WWER SElTlNG, X N1C - 91.00

DATA 0 TOTAL PREDICTH)IY c) TOTAL 0 JET

I AIRFRAME

m U

ii g T

gi ii

2 10

0

I w 2 0

E ?

125 315 800’ 2000 --3m 12500 FREQUENCY, Hz



141

O W E NUMBER 7 *ENGINE WWER SETTING, 96 NlC - a 7 0

DATA 0 TOTAL PREDICTION Cl TOTAL

JET

I AIRFRAME

t


% N t ~ x 9 0 . 7

142

O W E NUMBER 7 OENGINE POWER SEITING. K NlC = 9470


TOTAL CI JET

I AIRFRAME

f

i 10

00 FREQUENCY, H2

(b) DIRECTIVITY ANGLE = 60°


143

OCASE NUMBER 7 OENGINE POWER SETTING, W NlC - 80.10

DATA 0 TOTAL PREDICTION (II TOTAL 0 JET

FREQUENCY, H t



144

A CORE . D TURBINE I AIRFRAME

-

(d) DIRECTIVITY ANGLE - :200


145

O W E NUMBER . OENGINE POWtR SETTING, X N1C - a70

DATA 0 TOTAL PREDICTION cl TOTAL 0 JET

Q AIRFRAME


Figure A5. - (ConcJuded)

146

.CASE N-ER 8 @ENGINE #)MR SElTlNG, K N1C = U

DATA Q TOTAL PREoICTtoN PI TOTAL 0 JET

P AIRFRAME

a P

f n f $ i a 10

G n a

z

0

- I

2 0

FREOUENCY, H t

{a) DIRECTIVITY ANGLE = 300 Figure A6. - Spectral Cornparisors of Data and Prediction, Engive Powr Setting

% N I c = 84.4

I47

@CASE NUMBER 8 ~ E ~ I N E IWER sEmffi. x NIC = lylo

DATA 0 TOTAL PREDICIIOSY CI TOTAL 0 JET c) FAN A CORE

TURBINE V AIRFRAME



1 48

@CASE NOMBER 8 @ENGINE P O m R SETTING. X NlC = lMA0

DATA 0 TOTAL PREDICTION cl TOTAL 0 J€T Q FAN 4 CORE 0 TURBINE V AIRFRAME

ic) DlRECTlVlTV ANGLE - 900 Figure A6. - (Continued)

I49

(d) DlRFCfIVlTY ANGLE = 1200

Figure A6. - (Cont!nued)

150

O W E NUMBER 8 OENGlNE PW&R SETTING. K NlC =

DATA 0 TOTAL PREDICllmJ 0 TOTAL 0 JET 0 FAN A CORE D TURBINE I AIRFRAME

-

FREQUENCY, Hr

(e) DlRECTlVlN ANGLE = 1500


151

@CASE NUMBER 8 OENGlNE POWER SETTING, W N1C - 78.70

DATA 0 TOTAL PREDICTION D TOTAL 0 JET 0 FAN

CORE TURBINE

I AIRFRAME

Q 0

s p' m 10

i

FREQUENCY. Ht


% N ~ c = 78.7

I52

O W E NUMBER 8 OEffilNE POWER SETTING. % N1C - m70

DATA 0 TOTAL PREDICTION D TOTAL 0 JET 0 FAN A CORE 17 TURBINE I AIRFRAME

FREQUENCY, Hz

Ib) DIRECTIVITY ANGLE = 600


153

P m

A CORE , D TURBINE

1 AIRFRAME

-

n a

5 8

2

P,

U c



154

@CASE NUMBER 9 .ENGINE POWER SElTlNG, X N1C - lam

OATA 0 TOTAL PREDICTION b TOTAL 0 JET 0 FAN A CORE 0 TURBINE v AIRFRAME



155

OCASE NUMBER 8 @ENGINE POWER SETTING, K N1C = 78.70


TOTAL @ JET 0 FAN

I AIRFRAME

8

L L

n f zi 3 10

$

L

I-

O

I

z 0

s

0 FREQUENCY, Hz



156

3. Recipient's Catalog No. I I

4. Title and Subtitle


7. Author(s1

Belur N. Shivashankara - - . 9. hrfaming Orgnization Name and Address

2. Govrnmont Accuion No. I 5. Repon. Date

6. Performing Organlidtion Coda

October 1980

8 Performing Orpniiation Report No

D6-48 7 2 7 10. Work Unit No

- 12. Sponsorit-; Agmcy Name md Address

Langlev Research Center National Aeronautics and Space Administration Washington, D.C. 20546

13. Type of Repon and Period Covered

Contract report 14. Sponsoring Agency Code

NASA Contract Monitor: W. E. Zorumski

19 Security Classif. (of this report)

Unclassified

Y

11. Contract or Grant No

NAS 1 - I 5 5 26

20 Security Classif (of this pagel 21. No. of Pages 22 Rice'

Unclassified 162

~ ~ -~ ~~ ~ ~~

16. Abstract

NASA has developed a modular program called Aircraft Noise Prediction Program (ANOPP) for predicting aircraft flyover and sideline noise, To evaluate its accuracy, the ANOPP predictions were compared with the flyover noise measurements of a Boeing 747 airplane fitted with four Pratt & Whitney JT9D engines with hardwall nacelles. In general, the predictions did not agree well with the data. At high powers, airplane perceived noise levels (PNLT s) were overpredicted in the forward arc and underpredicted in the aft arc. At low powers the PNLT s were underpredicted at all angles greater than 80". Preliminary estimates of the various components indicate that significant modifications to the jet, fan, core, and turbine noise source modules may be required to improve the prediction-todata match. Since the airplane was flown with flaps 20" and landing gear up, the airframe noise was not dominant and, therefore, evaluation of the accuracy of the airframe noise prediction was not possible.

17. Key Words (Suggested by Authoris) J

Aircraft noise Fan noise Jet noise Engine noise Core noise Airframe noise Turbine noise Noise prediction

18. Distribution Statement

Unclassified-unlimi ted

N-305 For sale by the Natlonal Technical Inlormation Servtce Sprinefteld Vi ig in ta 22161

Documents

PREDICTION PROGRAM VALIDATION Final Report · 2013-08-31 · NASA CR-159333 AIRCRAFT NOISE PREDICTION PROGRAM VALIDATION Final Report Belur N. Shiva-kankara Boeing Commercial Airplane