1、Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-4 NASA CR-1056 A STUDY OF TURBOFAN-ENGINE COMPRESSOR- NOISE -SUPPRESSION TECHNIQUES By Alan H. Marsh, I. Elias, J. C. Hoehne, and R. L. Frasca Distribution of this report is provided in the interest of
2、information exchange. Responsibility for the contents resides in the author or organization that prepared it. Prepared under Contract No. NAS 1-5256 by MCDONNELL DOUGLAS CORPORATION, AIRCRAFT DIVISION Long Beach, Calif. for Langley Research Center NATIONAL AERONAUT ICs AN D SPAC E ADM I N I ST RAT I
3、 ON For sale by the Clearinghouse for Federal Scientific and Technical Information Springfield, Virginia 22151 - CFSTI price $3.00 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-Provided by IHSNot for ResaleNo reproduction or networking permitted wi
4、thout license from IHS-,-,-. K CONTENTS e Page SUMMARY. . . . . . . . . . . . . . . . . 1 INTRODUCTION . . . . , . . . . . . . . . . . 2 Acknowledgment . . . . . . , . . . . . . . 4 “ii. SYMBOLS. . . . . . ., . . . . . . . . . 4 DESCRIPTION OF THE NOISE SOURCE . . . . . . . . . 7 NOISE-CONTROL METHO
5、DS . . . . . . . . . 9 Background . . . . . . . . . . . . . . . 9 Design Requirements. . . . . . . . .*12 Acoustical Considerations . . . . . . .* 12 Aerodynamic Considerations . . . . . . 18 %6 EXPERIMENTALWORK . . . . . 20 Ae r odynami c Wind - Tunnel Tests . . . . . . . . . 2 0 Full-scale JT3D Du
6、ct-Wall SPL Measurements . . . . . 32 Laboratory Acoustical Studies . . . . . . . . . 40 P the primary gas generator part of the engine produces slightly more than half the total thrust, the fan produces the balance. The additional stages, In either case, the fan stages are larger in diameter than t
7、he adja- Figure 1 shows a cutaway view of the P a typical duct cross dimension (radially) is about 6. 5 inches. direct the airflow; these splitters divide each duct into three approximately ,equal portions (figure 2b): tions with a splitter. The short piece visible (figure 2b) in the central por- ti
8、on on the horizontal centerline at the nozzle exit is not intended to act as a ,splitter; it is merely a structural tie across the duct. There are four full-length splitters in each duct to a central portion with no splitter and two end por- Views of the inlet to the engine are given in figures 2c a
9、nd 2d. The minimum cowl diameter (throat diameter) is about 46 inches; the bullet (centerbody) diameter is about 18 inches at the IGV station. vanes and the blades of the first rotor stage can be seen in figure 2d; there are 23 IGV and 35 first rotor stage blades on all hush-kit equipped JT3D engine
10、s. (The hush kit was a development by P only that other solutions may be more attractive. with inlet choking are related to the wide range of engine power settings or airflows that may be encountered during representative approach conditions for a typical subsonic jet transport. Figure 3 shows the r
11、eductions in inlet area necessary to choke the flow at the inlet throat at approach conditions for the JT3D turbofan-powered DC-8, considering the extremes in landing weights possible, different flap settings permissible during approach, range of rates of descent, and possibility of an inoperative e
12、ngine. Note that area reductions from 30 percent to near 70 percent of the basic inlet area must be provided. The area reduction obtained by translating a standard-type (existing) nose bullet to the throat is only 15 percent. range of inlet areas, a “lightbulb“-type bullet might be necessary with a
13、relatively complex control system to set the bullet at the appropriate positions. If area reductions of from 50 to 70 percent are to be obtained without excessive lengthening of the inlet, then diffuser angles of near 40“ would be encountered aft of the throat. These high diffuser angles would lead
14、to serious flow separa- tions and associated total pressure distortions at the engine face, and possibly result in intolerable engine surge and stall. The major problems associated To provide the required Elimination of the inlet separation and associated surge and stall problems would require lengt
15、hening the inlet to reduce the diffuser angles. To obtain a diffuser angle of 7“ would necessitate increasing the length of the inlet by a factor of 5 to 7 relative to a normal inlet for 50 percent and 70 per- cent blockage, respectively, The drag and weight of this additional nacelle length would r
16、esult in unacceptable increases in aircraft fuel consumption. Other practical considerations which tend to eliminate the movable - bullet choked-inlet solution include the facts that: 0 The loads on the bullet are such that on failure of the actuation system the bullet could go to a forward position
17、. This would lead to an excessive loss in thrust at high power and possibly result in an unsafe condition, 13 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-Y Id 0 L- 5 Id E a Rate of descent during approach (fthin) Figure 3.- Area reduction require
18、d to choke inlet on approach. IO0 14 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-0 There is a large increase in noise level when the airflow through the blades is turbulent. Thus, choking the inlet may increase the noise radiated out the fan and
19、aggravate * I, rather than reduce the neighborhood1 noise problem. Therefore, it was considered that development of a choked inlet device would be outside the scope of this program and no experiments to demonstrate the effect of choking on inlet noise were carried out. It is felt that choking is not
20、 an acoustical design problem since it is known to be effective; it is an engineer- ing development problem, though a very complex one, with some difficult sible effects on the safety and reliability of the aircraft. decisions to be made about the magnitude of the penalties involved and the pos - 2.
21、 Resonators. - A resonator is defined as a device of the Helmholtz type which usually has strong selective absorption (i. e. , it is absorptive only in a narrow band of frequencies) and which, in combination with other acoustical elements, is often used in reactive mufflers. A resonator con- sists o
22、f a trapped volume of air which is connected to the external medium by some kind of channel. Rayleigh, reference 29, gives a derivation for the frequency of resonance. With the usage of reference 22, the frequency is where f is the frequency in Hertz (or cycles/second), c is the speed of sound in ft
23、/sec, G is the conductivity of the opening to the resonator in feet, and V is the trapped volume in ft3. The conductivity is related to the acoustic mass of air contained in the channel connecting the volume to the medium. proportional to the acoustic compliance. Resonance occurs between the kinetic
24、 energy of the acoustic mass of air oscillating in the channel under the influence of the imposed sound field and the potential energy stored in the compliance of the volume where the air acts like a spring. The volume of the cavity is A resonator of this type absorbs energy principally by two mecha
25、nisms: The resonator is assumed to have little or no acoustic friction or viscous losses in or near the connecting channel, and energy storage at resonance. damping and thus can only respond to the frequency to which it is tuned. ever, when it does respond is very large and the absorption is very gr
26、eat. the use of this concept as a duct lining. the acoustic energy to enter the resonator since there are many duct modes which carry energy in axial or circumferential modes compared to the radial modes which would have a perpendicular or normal incidence and thus a better coupling to the resonator
27、., quirement that a practical suppression device be absorptive over a broad fre- quency band in order to be effective not only against the fundamental frequency but also against the second and higher harmonics of the fundamental over a range of engine operating conditions How- at the resonance frequ
28、ency, the energy storage The first is that it is difficult to get There are two problems with The second problem results from the re- 15 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-In order to broaden the bandwidth of the absorption spectrum, som
29、e type of acoustic resistance must be added. For maximum effectivity the resistive material should be added where the acoustic particle velocities are highest, i.e., near the opening to the channels. However, addition of resistive material decreases the amount of absorption attained. Thus, the real
30、difficulty and the major design problem lies in the proper and practical choice of the resistance. Perforated panels spaced in front of rigid back walls can be analyzed by an extension of the theory of single resonators, with certain restrictions (reference 22), and can be designed to have rather hi
31、gh absorptivity (greater than 50 percent) over a rather broad frequency band (two octaves). in order to have low aerodynamic friction drag, the perforations must be very small. The panel should be thin to reduce the weight penalty. More- . over, the choice of the size of the perforations is dependen
32、t upon knowledge of the behavior of the resistance of the perforations as the particle velocity of the air molecules increases. Increasing the particle velocity of the air- flow through an orifice raises the effective resistance of the orifice above that measured under low flow conditions due to tur
33、bulence, acoustic streaming, and other nonlinear phenomena. As the particle velocity approaches the speed of sound, flow in the perforations approaches a choked condition and the acoustic resistance becomes infinite. Bies and Wilson, reference 30, showed that the resistance of an orifice starts to i
34、ncrease, above its linear laminar flow value, at an equivalent sound pressure level of 138 to 140 dB (i. e. , at a sound pressure corresponding to the steady flow particle velocity). However, Bies and Wilson conducted their experiments in an unconvected, stationary medium. typical SPLs are on the or
35、der of 150 to 160 dB. Therefore, the effective resistance of a perforated resonator is undoubtedly much greater than that calculated from classical small-amplitude theory and the determination of the proper design becomes a matter for empirical determination requiring simulation of the actual full-
36、scale environment. In a turbofan engine, the medium is convected and In summary, utilization of the perforated panel type of resonators as a noise suppression technique did not seem worthwhile because: Resonators have too narrow an absorption bandwidth A high value of absorption over a wide range of
37、 frequencies is needed (at least two octaves) The direction of propagation and modal structure of the sound field in the engine ducts makes it difficult to effectively couple the sound energy into resonators installed on the duct walls A wide absorption baandwidth requires the addition of acoustical
38、ly resistive material The resistance of the perforations or the channel into the resonator increases nonlinearly with increasing particle velocity starting at sound pressure levels of about 140 dB The required acoustic damping is hard to predict and requires considerable full-scale empirical effort
39、to select an optimum value 16 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-e Absorbers. - Absorbers are defined here as a class of acoustical The acoustic energy is converted into heat which is Friction Drag of the design might be high. % mqterial
40、s which absorb energy through viscous resistance to the motion of the air molecules. dissipated in the air and in the material. the low-density rockwool or fiberglass batt commonly used for insulation in air conditioning ducts. 80 percent of the incident acoustic energy in a band of frequencies thre
41、e octaves wide, for frequencies greater than 1000 Hz. However, these mate- rials have to be porous to admit the acoustic pressure oscillations and to The low-density (0.6 lb/ft3) fiberglass of the type com- monly used in aircraft insulation is over 99-percent porous; i. e., less than 1 percent of a
42、given volume of a sample is glass fibers, the rest is air. An example of this material is These materials can be made to absorb well over * absorb the energy. Use of these materials as duct linings alone was not considered to be feasible because: e The material would erode away rapidly due to the hi
43、gh- velocity stream of air in the duct 0 The porous materials would wick the various fluids that might be present in a duct (fuel, oil, water, etc. ) 0 Water retained in the material would freeze and possibly cause damage. the duct It might also tend to corrode the structure of a Fuel or other flamm
44、able fluid retained in the material would present a fire hazard. The use of acoustical absorbers can be considered for duct linings if a suitable facing material can be found. The facing material would prevent erosion of the material. Warm bleed air, if needed and if available, could be circulated t
45、o prevent freezing damage. However, considerable detailed engineering design work would be needed before these materials could be utilized in an actual ins tallation. Drains could be provided to drain away liquids. Broadband resonator. - The most promising noise suppression technique consists of lin
46、ing the walls of the fan inlet duct and the fan-discharge ducts with a broadband resonator. Such a resonator consists of sheets of porous metal installed over a fixed cavity. There are several requirements which can be specified to guide the selection of an acoustical material suitable for lining a
47、duct wall for the application intended. approach a pure resistance with almost no reactive components. The resistance of the resonator should be almost independent of frequency, main- taining a nearly constant value as frequency is increased. The lining should be comparatively smooth, so that fricti
48、on losses may be held to a minimum. The acoustic impedance of the porous surface should 17 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-The lining would replace part of the existing sheet-aluminum duct lining and would be backed up by a cavity. ca
49、vity resonating with the distributed reactance of the air in the pores of the liner that produces the broadband resonator.) Wirt, references 31 and 7 32, showed the feasibility of this approach for small gas-turbine exhaust muffler applications. (It is the compliance of the volume of trie A lining material which meets the requirements stated above
