Downloaded by PURDUE UNIVERSITY on September 17, 2016 | http://arc.aiaa.org | DOI: 10.2514/5.9781600865176.0439.0457 | Book DOI: 10.2514/4.865176 Purchased from American Institute of Aeronautics and Astronautics THE EFFECTS OF FORWARD SPEED ON A NUMBER OF TURBOJET EXHAUST SILENCERS J.R. Brooks* Rolls-Royce (1971) Limited, Bristol, England and R.J. Woodrow Hawker Siddeley Aviation Ltd., Hatfield, England Abstract Flight and static noise measurements were made on eight exhaust configurations of a turbojet engine installed in an HS125 aircraft to study flight effects on different components of turbojet exhaust noise. The configurations tested comprised plain and suppressor nozzles with and without a tailpipe acoustic lining. Jet velocities ranged from 1100 to 2100 fps. At angles less than about 40 to the jet axis exhaust noise correlated on a basis of jet velocity relative to the atmosphere; at angles greater than about 60 absolute jet velocity was a better correlating parameter. In the forward arc, noise increased, static to flight, with the datum conical nozzle. This increase disappeared with either a suppressor nozzle or an acoustically lined tailpipe. These results strongly suggest the existence of an important nozzle-based noise source influenced both by internal and external flow conditions. 1, Introduction Current noise regulations have insured that subsonic transport aircraft entering service at the present time are Presented as Paper 75-506 at the AIAA 2nd Aero-Acoustics Conference, Hampton, Va., March 24-26, 1975. The authors acknowledge with thanks the permission of their employers to publish this paper and the assistance of their colleagues in the preparation of the data. Thanks are also due to FIAT, Rolls-Royce partners in the design and development of the tailpipe and nozzle assemblies. * Deputy Head, Noise Department, Bristol Engine Division. Acoustics Engineer, Aerodynamics Department. 439 Purchased from American Institute of Aeronautics and Astronautics Downloaded by PURDUE UNIVERSITY on September 17, 2016 | http://arc.aiaa.org | DOI: 10.2514/5.9781600865176.0439.0457 | Book DOI: 10.2514/4.865176 440 J. R. BROOKS AND R. J. WOODROW significantly quieter than the first-generation jet aircraft. The next step in aircraft noise reduction could be a requirement that all aircraft in service shall comply with these regulations, which implies that considerable noise reduction will be required for most of these earlier aircraft, many of which have engines of low bypass ratio. Additionally, the first supersonic aircraft are about to enter service, bringing with them problems of noise associated with their high jet velocities. There is thus a strong interest in developing techniques for reducing the exhaust noise of all engines powering aircraft of these types. Much effort has been devoted by the engine and aircraft companies to develop effective jet silencers by testing at both model and full scale. In the past most of this work has been done statically to optimize a variety of silencer designs to give the maximum noise attenuation for a minimum performance penalty. Unfortunately, it has been found on several occasions that a silencer that performs well statically is far less acoustically effective in flight. This situation led Rolls-Royce and Hawker Siddeley Aviation to mount a collaborative test program to measure the noise and aerodynamic performance of a range of exhaust silencing systems. Eight exhaust configurations were tested statically and in flight, but consideration here will be confined mainly to a conical nozzle and a low-loss mixer nozzle, tested with and without an acoustically lined tailpipe. The static and flight results have been compared in order to obtain an improved understanding of the characteristics of exhaust noise components and the action of certain potential silencing devices upon these components. 2. Engine and Silencer Configurations The exhaust noise of the hot stream of turbine engines conventionally is said to comprise both jet mixing noise generated externally to the engine installation and "tailpipe" noise generated in the region of the nozzle exit. Statically, at takeoff power (jet velocity typically 2000 fps) jet mixing noise is usually the dominant component; at approach power (jet velocity typically 1200 fps) tailpipe noise is usually dominant, whereas at intermediate powers both components are significant. In flight, the tailpipe noise component is more significant relative to the jet mixing noise. The research program described here was designed to investigate, statically and in flight, the behavior of these components in a turbojet engine fitted with a standard convergent nozzle and alternatively with several silencer configura- Purchased from American Institute of Aeronautics and Astronautics 441 Downloaded by PURDUE UNIVERSITY on September 17, 2016 | http://arc.aiaa.org | DOI: 10.2514/5.9781600865176.0439.0457 | Book DOI: 10.2514/4.865176 EFFECTS OF FORWARD SPEED tions aimed at attenuating each noise component individually and in combination. Various types of mixer nozzles were used for jet mixing noise attenuation whereas "tailpipe" noise attenuation was anticipated from a tailpipe acoustic lining. The major part of this program was performed on a RollsRoyce Viper 601 engine, the flight phase being carried out with the engine installed in a Hawker Siddeley Aviation HS125 executive jet aircraft. The Viper 601 is a single-spool turbojet having a maximum sea-level, standard atmosphere, static thrust of 3750 Ib with a corresponding jet velocity of 2050 fps at a nozzle pressure ratio of 2.3:1. Five different test nozzles were used; all were tested in conjunction with a standard unlined engine tailpipe and some also were tested with an acoustically absorbent tailpipe. The test configurations investigated are given in Table 1. The following sections outline the considerations used in the design of the various silencers. 2.1. Nozzle Design This research program was performed in association with a silencer development program designed to develop a hushkit that would reduce the noise of the HS125-600 to permit it to comply with the requirements of FAR Part 36 with a minimum performance penalty. The design of the eight-lobed nozzles was based^ upon work done at Rolls-Royce at the time of the introduction of the first commercial jet aircraft. The first nozzle to be designed and tested had six lobes, and, although at this time Table 1 Test configurations Flight noise test Nozzle Tailpipe Datum conical Standard Absorbent Yes Yes Yes Yes 8-lobed (small core) Standard Absorbent Yes Yes Yes No 8-lobed (large core) Standard Absorbent Yes Yes Yes Yes 6-chuted Standard Absorbent Yes Yes Yes Yes Convergentdivergent Standard Absorbent Yes No Yes No Static noise test Purchased from American Institute of Aeronautics and Astronautics Downloaded by PURDUE UNIVERSITY on September 17, 2016 | http://arc.aiaa.org | DOI: 10.2514/5.9781600865176.0439.0457 | Book DOI: 10.2514/4.865176 442 J. R. BROOKS AND R. J. WOODROW Fig. 1 Eight-lobed large-core nozzle. Fig. 2 Six-chuted nozzle. the Viper 601 had not yet flown, the attenuation achieved was considered insufficient against the predicted aircraft noise levels. Consequently, a revised version of the nozzle was designed with eight lobes, now known as the "small-core" version. Extensive noise and performance tests were made on models of both the six-lobed and eight-lobed nozzles, and as a result it was concluded that the performance losses (both internal and external) of the eight-lobed nozzle could be reduced significantly by a redesign to give increased flow through the central core. At the same time, the function of the vanes in the lobes could be changed from being both structural members and flow straighteners to that of structural members only as a result of improvements to the flow profile of the lobes. Figure 1 shows the "large-core" eight-lobed nozzle installed on the aircraft. Minimum installed thrust loss is a basic requirement for any practical silencing system. Experience with "chuted" nozzles as fitted to some Rolls-Royce Spey engines has shown that such nozzles have low losses, and so a scaled-down Spey nozzle was included for comparison. It was recognized that the chuted nozzle probably would give less jet noise attenuation than the lobed nozzles. Figure 2 shows one of these nozzles during a test-bed performance calibration, 2.2. Tailpipe Acoustic Lining Design A silencer, or mixer, nozzle is most effective at high jet velocity when jet mixing noise is dominant. At reduced powers other sources, usually considered to be engine based, become relatively more significant. A tailpipe acoustic Purchased from American Institute of Aeronautics and Astronautics Downloaded by PURDUE UNIVERSITY on September 17, 2016 | http://arc.aiaa.org | DOI: 10.2514/5.9781600865176.0439.0457 | Book DOI: 10.2514/4.865176 EFFECTS OF FORWARD SPEED 443 lining was considered to be the most promising method of attenuating the noise from these other sources. So far as is known, acoustic liners have not been flown previously in the exhaust of a turbojet engine, and thus little previous experience was available as design background. The acoustic design procedure adopted was briefly as follows. An estimate of the jet mixing noise level and spectrum shape was made at the flight engine conditions of interest. This estimate then was subtracted from the total measured noise spectrum to leave a so-called "tailpipe" noise spectrum, which peaked at about 500-600 Hz with a very broad Noy spectrum. Thus to obtain KINKED CONICAL NOZZLE LINE Fig. 3 Tailpipe acoustic lining. Purchased from American Institute of Aeronautics and Astronautics Downloaded by PURDUE UNIVERSITY on September 17, 2016 | http://arc.aiaa.org | DOI: 10.2514/5.9781600865176.0439.0457 | Book DOI: 10.2514/4.865176 444 J. R. BROOKS AND R. J. WOODROW maximum attenuation a very thick lining was required which posed severe mechanical design problems because of the high thermal gradients present in the exhaust systems, especially during transient conditions such as startup. The design finally adopted is shown in Fig.3. The outer assembly consists of six panels supported in a partially "floating" manner to permit differential thermal expansion to be accommodated. The final item tested was a conventional convergentdivergent nozzle designed to give ideal expansion of the jet at takeoff pressure ratio, 2.3:1. The intention was to obtain a small amount of attenuation by reduction of the noise associated with the shock structure of the supersonic jet. It was found that, although a small attenuation was obtained, this attenuation was inadequate against the requirement, and so the convergent-divergent nozzle was discarded after one flight test. 3. Test Program and Techniques 3.1. Static Noise Tests All of the exhaust configurations were tested on the same Rolls-Royce static rig. All acoustical field shapes, whether static or in flight, are referred to the direction of the nozzle at exit, since aerodynamic performance investigations have established that the jet leaving the nozzle is coaxial with the nozzle at exit. Noise measurements were made at 100 ft linear from the engine using Bruel and Kjaer ^-in.-diam microphones at angles to the jet axis from 30 to 150 , at 15 intervals. The entire measurement area was paved with level concrete. In order to minimize the ground reflection problem as far as possible, two microphones were used at each position: one at engine centerline height (10 ft) and one at 2-in. height. Combination of the spectra produced from the two microphones at each position permitted a close approximation to a free-field spectrum to be deduced. Measurements were made at seven engine conditions in low wind conditions only. The engine conditions were chosen to cover the jet velocity range of interest in flight and were set up on the basis of constant corrected engine speed for each test configuration to insure that closely similar jet velocities were obtained each time. Two independent tests were carried out on the datum conical nozzle to check the repeatability of results. Analysis of the noise recordings was Purchased from American Institute of Aeronautics and Astronautics EFFECTS OF FORWARD SPEED 445 Downloaded by PURDUE UNIVERSITY on September 17, 2016 | http://arc.aiaa.org | DOI: 10.2514/5.9781600865176.0439.0457 | Book DOI: 10.2514/4.865176 performed on a General Radio type 1921 real-time 1/3-octave analyzer using an averaging time of 16 sec. All results were fully corrected for system frequency response. 3.2. Performance Tests Each exhaust configuration was calibrated on an engine test-bed in order to permit the calculation of accurate values of the relevant exhaust parameters (primarily jet velocity). Performance measurements were made during the noise tests but only in sufficient detail to enable the operating points on the engine characteristics to be determined. 3.3. Flight Noise Tests The flight tests were conducted on the basis of making comparative noise measurements on each configuration in turn from nominally straight and level flyovers conducted under closely controlled conditions. Each of the engine exhaust assemblies to be tested was installed on the port engine of the test aircraft, a standard HS125-600, the starboard engine being left unmodified as a control. Four engine speed settings were used: 100% (maximum), 95%, 90%, and 82%, corresponding to jet velocities of about 2100, 1900, 1600?and 1200 fps. For each flyover, the aircraft was powered by the "test" engine only. In all cases, the "nontest" engine was operated at flight idle engine speed and did not contribute to the total noise. Test conditions were constant engine speed, constant forward speed of 160 knots indicated airspeed (270 fps) at a nominally constant altitude of 2000 ft, except for the lowest power case, which was made at an altitude of 400 ft. At least three flyovers were made at each condition to insure an adequate data sample. For reference purposes, single flyovers were made on the starboard engine at 90% engine speed at intervals throughout the test period. None of the starboard engine results have been used in this study. The total number of flyovers was 227. Aircraft and engine performance parameters were recorded by an onboard autoobserver, which was synchronized with the ground noise recorders. The noise recordings were made under the flight path using the standard procedures prescribed in FAR Part 36; the flight path was aligned with the side of the runway at the test airfield. The aircraft height and any lateral offset from the nominal flight path line were measured by photographic techniques, a camera adjacent to each microphone being operated as the aircraft passed overhead. Operation of the camera shutter also caused a pulse to be recorded on the noise Purchased from American Institute of Aeronautics and Astronautics Downloaded by PURDUE UNIVERSITY on September 17, 2016 | http://arc.aiaa.org | DOI: 10.2514/5.9781600865176.0439.0457 | Book DOI: 10.2514/4.865176 446 J. R. BROOKS AND R. J. WOODROW tape for synchronization purposes. All analysis of the noise recordings was also in accordance with the requirements of FAR Part 36, using the same systems as used for analysis of the static noise data to avoid the possibility of hardware or software discrepancies affecting the flight-to-static comparison. 4. Test Results 4.1. Static Results It was mentioned earlier that the datum conical nozzle configuration was tested twice. Comparison of the "asmeasured" overall sound pressure levels (OASPLfs) for the two runs showed that the measurement repeatability was generally better than 1 dB. Also mentioned earlier was the problem of ground reflection, which severely obscures the true spectral shapes when measurements are made close to a reflecting surface, such as the ground. For research purposes, free-field data are almost essential, and efforts have been made to remove ground reflection effects as far as possible. The high- and lowmicrophone technique was used for the static tests to deal with this problem. Although the resultant "free-field" spectra undoubtedly contain imperfections, they are very much more satisfactory than those measured directly. The deduced free-field spectra from the static measurements have been extrapolated to 2000 ft linear (using the atmospheric absorption data of the latest revision of SAE ARP 866) to permit comparison with the flight data. Figure 4 compares extrapolated spectra for three nozzles of interest at maximum power (a jet velocity Vj of 2020 fps). Figure 5 shows the effect of the acoustic lining for the conical and eight-lobed large-core nozzles, respectively, at low power, Vj = 1170 fps. Free-field OASPL's have been obtained by integration of the 1/3-octave SPLfs for each spectrum over the frequency range 50 Hz -10 kHz. Typical results for the datum conical nozzle and for the three mixer nozzles are shown in Fig. 6. Figure 7 compares OASPLfs for the conical nozzle and eight-lobed large-core nozzle with and without the tailpipe acoustic lining. 4.2. Flight Results All of the flight test results have been corrected to 2000-ft altitude, standard day (25°C, 70% relative humidity) conditions. As mentioned earlier, all of the flying was performed at an airspeed of 270 fps, allowing the aircraft to Purchased from American Institute of Aeronautics and Astronautics 447 EFFECTS OF FORWARD SPEED Downloaded by PURDUE UNIVERSITY on September 17, 2016 | http://arc.aiaa.org | DOI: 10.2514/5.9781600865176.0439.0457 | Book DOI: 10.2514/4.865176 30° TO JET AXIS 60 VELOCITY = 1170 FPS JET VELOCITY =2020 FPS CONICAL NOZZLE 6-CHUTED -.— 6-LOBED l.C. —•- i Q 50 % | 60 I 50 ^ 53 125 250 500 10K 2 OK I OK 8-QK 1 /3rd OCTAVE CENTER FREQUENCY ~ Hz Fig. 4 Viper 601 static noise (comparison of spectra: unlined tailpipe; free-field). 63 725 250 500 1 -OK 2 OK kOK 6 OK 1 /3rd OCTAVE CENTER FREQUENCY-Hz Fig. 5 Viper 601 static noise (eight-lobed and conical nozzles: effect of tailpipe lining; free-field). climb or descend as necessary to maintain the airspeed constant. During the data processing phase, all of the results have been corrected to level flyover conditions, including corrections for aircraft incidence, so that all of the flight results can be compared with the static results on a common basis of angle to the jet axis. The low power flights at 400-ft altitude also have been converted to 2000-ft altitude to permit comparison with the high-power results. The basic analysis of the noise recordings was performed at J-sec intervals as defined by FAR Part 36 but the analyzer output has been interpolated to provide spectra at 15 intervals to the jet axis. In the case of the flight noise measurements, the microphones were positioned over grass at one height only, 4 ft, so that corrections to free-field equivalent spectra are less well defined compared with the static case. The method adopted was to draw in by eye the free-field curves on the basis of previous Rolls-Royce experience with various microphone and ground surface configurations. OASPLfs were computed from the deduced freefield spectra, as for the static spectra. Purchased from American Institute of Aeronautics and Astronautics Downloaded by PURDUE UNIVERSITY on September 17, 2016 | http://arc.aiaa.org | DOI: 10.2514/5.9781600865176.0439.0457 | Book DOI: 10.2514/4.865176 448 J. R. BROOKS AND R. J. WOODROW Flight engine performance was computed from the data recorded by the aircraft auto-observer in combination with results obtained from performance calibration running carried out on the engine test-bed, as mentioned earlier. In all cases, the jet velocity at each test condition has been derived from the tailpipe temperature and pressure, assuming isentropic expansion to ambient pressure, so that all jet velocity values have been calculated on the same basis. Particular care was taken when setting up each exhaust configuration for performance calibration to insure that the nozzle area was trimmed suitably at takeoff engine rating. Thus, not only could the jet conditions be compared accurately, but also the exhaust system losses could be deduced by comparison with the datum nozzle assembly. Figure 8 shows typical free-field OASPL!s plotted as a function of flight jet velocity for two nozzles (datum conical and eight-lobed large-core, with and without the tailpipe sor 80 - CONICAL NOZZLE UNLINED + LINER 8-LOBED L.C. UNLINED + LINER DATUM CONICAL NOZZLE 9-LOBED SMALL CORE 8-LOBED LARGE CORE 6- CHUTED —o—— •—•—• ——"—— —*— 70 - 70 - 60 60 L- "2 "1 0 -1 2 -3 LOG*)(JET VELOCITY / SPEED OF SOUND) Fig. 6 Viper 601 static noise (comparison of nozzles: unlined tailpipe; free-field). LOGn(JET VELOCITY / SPEED OF SOUND) Fig. 7 Viper 601 static noise (effect of tailpipe acoustic lining: free-field). Purchased from American Institute of Aeronautics and Astronautics Downloaded by PURDUE UNIVERSITY on September 17, 2016 | http://arc.aiaa.org | DOI: 10.2514/5.9781600865176.0439.0457 | Book DOI: 10.2514/4.865176 EFFECTS OF FORWARD SPEED CONICAL NOZZLE —o + LINER —* 449 B-LOBED LC. NOZZLE—• •- + LINER—I •2 -3 0 -1 LOCmUET VELOCITY I Fig. 8 Viper 601 flight noise (results with datum and silenced exhausts: free-field). acoustic lining). It is evident that in all cases the results at the lowest power condition, measured at 400-ft altitude and extrapolated to 2000 ft,are consistent with the higher-power results, measured at 2000 ft. 5. Discussion of Results Of the various suppressor nozzles tested, the eight-lobed large-core nozzle was found to be the most effective in terms of attenuation for minimum thrust loss, the latter being about 1|% net thrust at takeoff conditions.1 Consequently, this nozzle has been chosen as the most appropriate for comparison with the datum conical nozzle. 5.1. Effect of Flight: Rearward Arc It is conventional to assume that the intensities of jet mixing noise and tailpipe noise are a function of an exhaust velocity to some power. Fully expanded jet velocity is that most commonly used, and all of the comparisons made here are on this basis. However, for the case of a jet from a nozzle with forward speed, jet mixing noise theory suggests that the velocity of the jet relative to the stationary atmosphere should be the controlling parameter. One important objective of the present study was to examine how much of the noise field was dominated by jet Purchased from American Institute of Aeronautics and Astronautics Downloaded by PURDUE UNIVERSITY on September 17, 2016 | http://arc.aiaa.org | DOI: 10.2514/5.9781600865176.0439.0457 | Book DOI: 10.2514/4.865176 450 J. R. BROOKS AND R. J. WOODROW 0 -1 2 -3 LOGmUET VELOCITY/Gael 0 -1 -2 -3 LOdolREL. JET VELOCITY/Coc) Fig. 9 Viper 601 with datum conical nozzle (comparison of flight and static results: free-field). mixing 'noise. Consequently, where the relative strengths of the noise sources are unknown as in the present study, it is not obvious whether absolute or relative jet velocity is the more relevant parameter at any given angle with respect to the jet axis, and so both parameters have been examined. Figure 9 compares static and flight OASPLTs at 30°, 60°, and 90 to the jet axis for the datum conical nozzle as a function of both absolute and relative jet velocities divided by the speed of sound (C^. It is clear that only at 30° do the static and flight results collapse on a relative velocity basis. Figure 10 shows similar data for the conical nozzle with the tailpipe lining installed, whereas Fig. 11 shows results for the eight-lobed nozzle, also with the lined tailpipe; both of these again show that, at 30 , relative jet velocity is the better correlating parameter. At other angles, all three diagrams show that Vj is the better parameter, in respect to both level and slope. The consistency of all of the results at angles of 60 and greater indicates that the sources are similar statically and in flight and, unlike jet mixing noise, are not affected by forward motion. Although the emergence of a noise source, other than from conventional jet mixing, is not unexpected at low jet velocities, it is perhaps surprising that there is no apparent tendency, as jet Purchased from American Institute of Aeronautics and Astronautics EFFECTS OF FORWARD SPEED 451 Downloaded by PURDUE UNIVERSITY on September 17, 2016 | http://arc.aiaa.org | DOI: 10.2514/5.9781600865176.0439.0457 | Book DOI: 10.2514/4.865176 velocity is increased, for the rear arc to become progressively more dominated by jet mixing noise. An alternative method of examining the results is to compare the flight and static noise levels at the same source (nozzle) thermodynamic conditions. Since in this test program all of the jet velocities have been calculated in the same way from tailpipe conditions, this is effectively equivalent to comparing results at the same value of jet velocity. Figures 12 and 13 show flight and static field shapes from the datum and the eight-lobed nozzles at three jet velocities: supersonic, transonic, and low subsonic. Examination of the static-to-flight changes indicates clearly that any reduction in noise level due to relative motion between the jet and the surrounding air is confined to angles well in the rear arc. It is interesting to notice that, at 30 to the jet axis, log Vj/CQf 0.20 (i.e.,Vj = 1750 fps approximately), the reduction of noise with forward speed is effectively the same for the conical and silencer nozzles, whereas with the assumption of independent sources a reduced attenuation with forward speed would have been expected with the silencer nozzle because of the expected lesser significance of jet mixing noise in relation to other sources. Figure 14 compares the field shapes at two jet velocities of the conical, six-chuted9and eight-lobed 0 1 2 3 LOGwUET VELOCITY/CccJ 0 1 2 -3 LOGwlREL. JET VELOCITY/Cai Fig. 10 Viper 601 with conical nozzle and lined tailpipe (comparison of flight and static results: free-field). Purchased from American Institute of Aeronautics and Astronautics Downloaded by PURDUE UNIVERSITY on September 17, 2016 | http://arc.aiaa.org | DOI: 10.2514/5.9781600865176.0439.0457 | Book DOI: 10.2514/4.865176 452 J. R. BROOKS AND R. J. WOODROW 0 -1 -2 3 LOGmUET VELOCITY/Coc) 0 -1 2 -3 LOG10(REL. JET VELOCITY/Coc) Fig. 11 Viper 601 with eight-lobed large-core nozzle and lined tailpipe (comparison of flight and static results). nozzles with the unlined tailpipe and shows that, as expected, the six-chuted nozzle is intermediate between the other two. It is clear that, even at angles where jet mixing noise is dominant, the behavior of suppressor nozzles is not as simple as often supposed, nor does there appear to be any support for the hypothesis that noise sources can be separated meaningfully into mixing noise and tailpipe noise. 5.2. Effect of Flight; Forward Arc Figure 12 shows that, with the datum conical nozzle and no tailpipe liner fitted, the flight noise levels at high angles to the jet and in the forward arc are higher than the static levels, suggesting either a new noise source or amplification of an existing one. This increase with the datum nozzle occurs at all jet velocities covered by the tests (1100 - 2100 fps) and so cannot be related to shock noise or any other source that is peculiar to supercritical pressure ratios. The additional noise clearly is not airframe noise or engine compressor noise since,as Fig. 14 shows, use of the eight-lobed nozzle reduces the forward arc noise level at both high and low jet velocities. It is concluded that, for all of the test conditions.the source of the forward arc Purchased from American Institute of Aeronautics and Astronautics EFFECTS OF FORWARD SPEED 453 Downloaded by PURDUE UNIVERSITY on September 17, 2016 | http://arc.aiaa.org | DOI: 10.2514/5.9781600865176.0439.0457 | Book DOI: 10.2514/4.865176 noise is associated with the engine exhaust system and that this source may be amplified with forward speed. 5.3. Effect of Tailpipe Acoustic Lining Both the static and flight results show some interesting features. The static spectra for the conical nozzle (Fig.5) show that introducing the liner appears to have increased the noise at 30 an angle where the liner would be expected to have no effect because of the dominance of jet mixing noise. Figure 7 shows that this was not an isolated (and so possibly unrepresentative) result whereas Fig. 8 shows much the same result in flight. At the other angles, small attenuations are produced statically, the liner attenuation with the eight-lobed nozzle being generally greater than with the conical nozzle. This is compatible with the theory that the mixer nozzle has suppressed some of the jet mixing noise so that any reduction of "tailpipe" noise by the acoustic lining is more conspicuous. However, it will be noticed that the principal attenuation occurs at the high-frequency peak around 800-2000 Hz. This peak, common in multielement suppressor nozzle spectra, often is assumed to be mixing noise from the individual elemental jets, and consequently its attenuation by the acoustic lining is unexpected. It seems likely, however, that the liner is effective in modifying some fluid-dynamic mechanism that is important in generation or propagation processes in the region of the nozzle exit rather than inside the tailpipe. STATIC UNLINED ——o—— FLIGHT UNLINED —-*• — FLIGHT LINED ——i-—— EFFECT OF FLIGHT^/ LINED ~ UNLINED 6-CHUTED NOZZLE J^" LINED s 50L 80 r- 6-LOBED LC. NOZZLE UNLINED ^ LINED X LogJJET VELOCITY/SPEED OF SOUND } Fig. 15 Viper 601 noise (comparison of flight and static results in the forward arc (120° to the jet axis: free-field). Purchased from American Institute of Aeronautics and Astronautics 456 J. R. BROOKS AND R. J. WOODROW 70 Downloaded by PURDUE UNIVERSITY on September 17, 2016 | http://arc.aiaa.org | DOI: 10.2514/5.9781600865176.0439.0457 | Book DOI: 10.2514/4.865176 50 / 70 i50 \ § ^ 60° STATIC ^ UNLINED -----LINED — —— •—— 00 50 \ STATIC ft 50 _ 60 \ i •* 80 \ 120° s-^^ \\ V\ ^— — _.^ 8! 40 00 •o 70 120° '• Uj 60 70 // j.'' \\\ N 'X' % 60 - UNLINED ' LINED • / ""XN FLIGHT \ 50 \ FLIGHT \ \x \ \ \ § 50 1 20 8 tj 20 60* •£! ATTENUATION SPECTRA 10 ^FLIGHT JC X ^ o x Hr _^» * ~ ~* " "x"x" x 120° . ATTENUATION SPECTRA 10 < ~ "*"*' 0 -10 -10 I63 /25 250 500 10K 2 OK l-OK 8-OK 1/3rd OCTAVE CENTER FREQUENCY ^ Hz Fig. 16 Viper 601 with conical nozzle (effect of tailpipe acoustic lining in the rear arc (60 to the jet axis: log (Vj/Ca) =0.2; free-field). 63 125 250 500 10K 2-OK 4-OK 8 OK 1/3rd OCTAVE CENTER FREQUENCY ^ Hz Fig. 17 Viper 601 with conical nozzle (effect of tailpipe acoustic lining in the forward arc (120° to the jet axis: log (Vj/Ca) = 0.2; free-field). principal source is associated with the nozzle lip, since it is hard to see how internal engine conditions upstream of the nozzle could change sufficiently from static to flight to account for the observed datum nozzle results. However, since the addition of the acoustic liner had a significant effect, it is clear that internal flow conditions upstream of the nozzle have a strong influence on the noise source concerned, even if the effective source is in the region of the nozzle exit. 6. Conclusions This study has shown that, for both conical and suppressor nozzles, only at angles near the jet does turbojet exhaust noise correlate with relative jet velocity, suggesting that only in this region does conventional jet mixing noise dominate. At other angles and all jet velocities in the range 1100-2100 fps, absolute jet velocity appears to be a Purchased from American Institute of Aeronautics and Astronautics Downloaded by PURDUE UNIVERSITY on September 17, 2016 | http://arc.aiaa.org | DOI: 10.2514/5.9781600865176.0439.0457 | Book DOI: 10.2514/4.865176 EFFECTS OF FORWARD SPEED 457 more relevant parameter. Increase of noise in the forward arc has been observed, static to flight, with the datum conical nozzle. This increase vanished with the addition of a tailpipe acoustic liner and did not occur with an eight-lobed suppressor nozzle, suggesting that the relevant noise source was associated with the nozzle lip, was affected by flow conditions upstream of the nozzle, and was amplified by forward speed. Statically, the use of the acoustically lined tailpipe with the suppressor nozzle produced attenuation of high frequency noise, previously supposed to be mixing noise of the "elemental jets," whereas in flight the acoustic lining gave more attenuation at all jet velocities with the conical nozzle than with the suppressor nozzle. Thus the conventional breakdown of turbojet exhaust noise into "mixing" and "tailpipe" components has not been substantiated by these flight experiments. It follows that attempts to predict the combined effects of a jet mixing noise suppressor and a "tailpipe" acoustic lining on the basis of treating mixing noise and tailpipe noise as relatively independent components will be unsuccessful. Further work must be done to study the characteristics of so-called "tailpipe noise" before a more complete understanding of engine exhaust noise can be achieved. References Brooks, J.R. and Woodrow, R.J., "Silencing the HawkerSiddeley HS125 Aircraft," Inter-Noise T74 Conference, Sept. 1974, Washington, B.C. 2 Greatrex, F.B. and Brown, D.M., "Progress in Jet Engine Noise Reduction," 1st International Council of the Aeronautical Sciences Congress, Sept. 1958, Madrid, Spain. Downloaded by PURDUE UNIVERSITY on September 17, 2016 | http://arc.aiaa.org | DOI: 10.2514/5.9781600865176.0439.0457 | Book DOI: 10.2514/4.865176 Purchased from American Institute of Aeronautics and Astronautics This page intentionally left blank