Acoustical absorptive splitter
An acoustic absorptive splitter consisting of two absorptive faces separated by a single on-quarter wavelength cavity support is disclosed as an equal substitute for a multi-layered splitter containing a central septum. This construction takes advantage of the reflective properties of the absorptive face sheets to support standing waves in the tuning cavities and in the duct cross modes. The utilization of these design principles permits significant silencer size and weight reduction. It also permits reduced splitter manufacturing costs by elimination of two layers of materials, difficult internal bonding and improvements in quality.
 The present invention relates generally to duct silencers and, more particularly, to duct silencers wherein the sound attenuation is achieved through absorption of sound energy by multiple absorptive surfaces of a longitudinal splitter or group of splitters.DESCRIPTION OF THE PRIOR ART
 State of Art Absorptive Duct silencers, utilized for low temperature applications, such as for commercial air conditioning systems, engine silencers, etc. utilize fibrous material for lining the duct. Many fiber-like materials may be used, for example, steel wool or glass fiber. The absorptive materials may be confined behind foraminous material such as perforated sheet metal, expanded metal or screening, or simply glued in place. These silencers, however, are unsuitable for high temperature gas applications or high velocity gases, since the fibers tend to migrate, fracture or char. They also become blinded or saturated when high amounts of entrained materials such as ash, soot, dust or oily liquids are present. These materials are particularly unsuitable in aircraft applications, where the danger of fire from oil or fuel contamination is possible. In addition, it is desirable in these applications to minimize weight.
 Duct silencers in aircraft applications utilize metallic sheet absorber facing materials. These facings are positioned in the duct as splitters and are usually supported by spacing materials. These spacers are designed to create open cavity space behind the facing sheets, honeycomb or “egg-crate” structure being preferred. The facing sheets may be adhesively bonded, brazed, welded or mechanically joined to these supports. Usually, the support materials are configured to position the facings a predetermined one-quarter wavelength from a reflective hard wall. The absorber assembly is, thereby, tuned to a desired peak portion of the sound spectrum range; usually one centered on a dominant frequency emitted by a turbine, fan or other offensive noise generator. It is characteristic of sheet absorbers that they are most efficient when the maximum velocity of local standing waves is positioned at the absorptive surface. The support spacing described above assures that the facing sheet is properly positioned. It is known that the standing wave velocity at a reflective wall is zero, and that one-quarter wavelength from this wall it will be a maximum.
 Conventional splitter designs utilize two absorptive faces in a back-to-back arrangement. These splitters are fabricated with two absorptive face sheets, a central solid septum, and two one-quarter wave spacer supports. The splitter is then one-half wavelength thick. The presence of this thick splitter either forces the duct exterior size to be increased or creates excessive restriction or blocking, resulting in backpressure or pressure drop. These size increases also create excessive increases in weight or force compromises in performance. In addition, the solid septum adds weight and requires two internal surface bonds. This results in significant manufacturing complexity and associated increased cost.
 An example of an acoustic liner or splitter that utilizes facing materials and spacers are described in U.S. Pat. No. 6,209,679, which issued to Hogeboom, et al. on Apr. 3, 2001 for an “Aircraft engine acoustic liner and method of making same” and U.S. Pat. No. 5,782,082, which issued to Hogeboom, et al. on Jul. 21, 1998 for an “Aircraft engine acoustic liner,” both of which disclose a low resistance acoustic liner formed by a middle layer having partitioned cavities sandwiched between an imperforate sheet and an optional perforate sheet. A second arrangement of acoustic linings includes a splitter having a portion formed of low resistance liner. Similarly, U.S. Pat. No. 5,912,442, which issued to Nye, et al. on Jun. 15, 1999 for a “Structure having low acoustically-induced vibration response” discloses a low vibroacoustic structure comprising first and second facesheet defining a plurality of holes attached to opposed surfaces of a core defining a plurality of passages in communication with the holes to form channels through the structure.
 Another example of an acoustic liner having a double wall type structure is described in U.S. Pat. No. 4,294,329, which issued to Rose, et al. on Oct. 13, 1981 for a “Double layer attenuation panel with two layers of linear type material.” Rose, et al. discloses a double degree acoustic attenuation sandwich panel having a plurality of stacked components adhered together to form a unitary sandwich structure comprising an impervious facing of thin sheet material, a first honeycomb core with end wise directed cells, a first perforate facing of thin sheet material, a first thin layer of porous fibrous material, a second perforated facing of thin sheet material, a second honeycomb core with end wise directed cells, a third perforate facing of thin sheet material, the second perforated facing sheet having substantially larger perforations than the first and third sheets, and a second layer of porous fibrous material. A method for manufacturing such a panel is disclosed in U.S. Pat. No. 4,421,811, which also issued to Rose, et al. on Dec. 20, 1983 for a “Method of manufacturing double layer attenuation panel with two layers of linear type material.” This patent discloses a method of manufacturing a double degree acoustic attenuation sandwich panel having a plurality of stacked components adhered together to form a unitary sandwich structure comprising an impervious facing of thin sheet material, a first honeycomb core with end wise directed cells, a first perforate facing of thin sheet material, a first thin layer of porous fibrous material, a second perforated facing of thin sheet material, a second honeycomb core with end wise directed cells, a third perforate facing of thin sheet material, the second perforated facing sheet having substantially larger perforations than the first and third sheets, and a second layer of porous fibrous material.
 Sound absorbing panels are described in both U.S. Pat. No. 4,667,768, which issued to Wirt on May 26, 1987 for a “Sound absorbing panel” and U.S. Pat. No. 4,084,366, which issued to Saylor, et al. on Apr. 18, 1978 for a “Sound absorbing panel.” The former discloses a sound absorbing panel comprising an array of walls configured to provide two or more contiguous hollow cells having adjacent open ends and adjacent closed impermeable ends, the open ends defining a sound receiving end of the array, at least one impermeable three dimensional closed surface disposed in at least one of the cells, and a flow resistive permeable facing sheet covering the sound receiving end. The Saylor et al. patent, on the other hand, discloses a wall panel having a rigid rectangular frame with a core structure disposed within the region bounded by the frame, which core structure comprises at least one honeycomb layer and sheetlike skins fixedly secured to opposite sides of the frame and extending across the region bounded by the frame for confining the honeycomb layer therebetween.
 The particular applicability of such acoustic liners for jet engines is shown in U.S. Pat. No. 5,594,216, which issued to Yasukawa, et al. on Jan. 14, 1997 for a “Jet engine sound-insulation structure,” and which discloses a jet engine sound insulator having a structure that includes a framework and an acoustical insulation material disposed within the framework composed of a rigid matrix of randomly oriented, fused silica fibers having fiber diameters predominantly in the range 2-8 .mu.m, a three-dimensionally continuous network of open, intercommunicating voids, a flow resistivity between about 10-200K rayls/m, and a density of between about 2 and 8 lb/ft3.
 Such devices, however, fail to offer the unique advantages of the acoustical absorptive splitter as contemplated by the present invention, namely, to provide an effective absorptive splitter that is thinner in cross section, easier to manufacture and available at reduced costs.
 The following publications are all directed to the scientific and engineering elements of acoustics, and are incorporated herein by reference: (1) “Notes on Sound Absorption Technology”, K. U. Ingard. Noise Control Foundation, Poughkeepsie, N.Y., 1994 ISBN 0-931784-28-X; (2) “Elements of Acoustical Engineering” Harry F. Olson, D. Van Nostrand Company, New York, N.Y. 1947; (3) “Acoustical Engineering” Harry F. Olson, D. Van Nostrand Company, Inc., New York, 1957, Library of Congress Card No. 57-8143; and (4) “Handbook of Noise Control” Cyril M. Harris, McGraw-Hill Book Company, New York, N.Y. 1957 Library of Congress Catalog Card No. 56-12268SUMMARY OF THE INVENTION
 The present invention provides an absorptive splitter that consists of two absorptive face sheets spaced apart one-quarter wavelength by a single honeycomb or egg-crated supporting structure. The invention eliminates the solid septum and its two bonded joints. It also eliminates one of the quarter-wave thick supports required in state of art splitters. The new splitter is then substantially (almost 50%) thinner and provides the advantage of reduced weight and simplified manufacturing. It also provides reduced overall duct size, pressure drop and back pressure.
 Noise attenuation testing in laboratory ducts has shown that the attenuation performance of this design is equal to or better than the state of art design. This result contradicts present theory in that the one-quarter wavelength cavity between the two face sheets is active and continues to control tuning; this despite the absence of a central hard wall reflector. This unexpected result indicates that the two absorptive face sheets are sufficiently reflective to maintain the internal standing wave.
 These and other aspects of the invention will become apparent from the following detailed description and the accompanying drawings, which illustrate by way of example the features of the invention and its application.BRIEF DESCRIPTION OF THE DRAWINGS
 FIG. 1 is a plot of the acoustic impedance of Fibermetal measured with an acoustic impedance tube instrument;
 FIG. 2 is a drawing showing the structure of a typical state of art acoustical splitter as used in a duct silencer;
 FIG. 3 is a drawing showing a duct containing an acoustic splitter as described in FIG. 2;
 FIG. 4 is a drawing showing the simplified structure that replaces the splitter shown in FIG. 2 and is the subject of this patent;
 FIG. 5 is a drawing showing a duct containing an acoustic splitter of the type shown in FIG. 4;
 FIG. 6 is an end view of an acoustical silencer duct containing a state of art splitter of the type shown in FIG. 2;
 FIG. 7 is an end view of an acoustical silencer duct containing a simplified splitter as shown in FIG. 4;
 FIG. 8 is an end view of an acoustical silencer duct containing multiple splitters of the type shown in FIG. 2;
 FIG. 9 is an end view of an acoustical silencer duct containing multiple splitters of the type shown in FIG. 4;
 FIG. 10 is an end view of an acoustical silencer duct containing a splitter of the type shown in FIG. 2 plus treated side walls;
 FIG. 11 is an end view of an acoustical silencer duct containing a splitter of the type shown in FIG. 4 plus treated side walls;
 FIG. 12 is an end view of a circular acoustical silencer duct containing a splitter of the type shown in FIG. 4; and
 FIG. 13 is an end view of a hexagonal acoustical silencer duct containing a splitter of the type shown in FIG. 4.DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
 Duct silencers for use in aircraft have been designed and manufactured generally as shown and configured in FIG. 6, FIG. 8 and FIG. 10. These may contain acoustical absorptive splitters in proximity to the flowing gases as shown in these figures. The splitters are tuned by sizing the spacing established by the thickness of the cavities 40 in supporting structure behind the face sheets 50. This spacing is referred to in the industry as the “cavity depth” 42. The cavity depth 42 is designed to approximate one-quarter wavelength at a desired peak frequency. This spacing establishes a peak resonant standing wave frequency in the cavity 40 and, thereby, a peak absorptive frequency for the assembly. Usually this frequency is chosen to be the first harmonic of the dominant noise being treated. For example, it could be the blade passing frequency of a fan or turbine, or it could be a center frequency of a noise spectrum of a hiss from an air blast.
 The tuning mechanism involved in the cavity depth design is based on the establishment of a standing wave in the cavity 40 behind the face sheets 50. Some of the energy from the sound wave that impinges on the face sheets 50 penetrates through the face sheets 50 and excites and perpetuates a standing wave in the cavity 40. The state of art termination of the cavity 40, the backing plate or septum 30, reflects energy and its face becomes a zero velocity, maximum pressure point. A corresponding point, one-quarter wavelength away, will then be at the maximum velocity of the standing wave. The media face sheet 50 is located at this point. Further, as a consequence, the face sheets 50 is then at a maximum velocity point and the gas is forced to cyclically permeate through the porous face sheet. The tortuous resistive path through the face sheets 50 converts the sound energy to heat and thereby absorbs the sound.
 The face sheet 50 is designed to offer an acoustical impedance match R to that of the gas, &rgr;c, in the duct 20. When that match is correct, as measured in an impedance tube or by air permeability, theory says that the face sheets 50 will not reflect sound, but will optimally absorb it. Ingard states, pages 1-3, that if the absorption coefficient, &agr;, of a material is measured, it's value will be &agr;=1−(R)2, where R is the reflection coefficient. If &agr;=1, then R must be zero or very small.
 FIG. 1 is a plot of the normalized resistance and reactance of a typical absorber face sheet 50 versus frequency as measured in an acoustical impedance tube. The resistance curve is equal to 1.0, the optimum match, i.e., the acoustical resistance of the face sheet 50 matches the acoustical impedance of air. This match is displayed for a substantial useable range of the test frequencies. The impedance plot for this material then indicates that reflectance must be very low. The reactance curve indicates the tuned response of the sample and the resonant frequency of the supporting cavity 40. This occurs at the frequency (2750 Hertz) where the reactance curve crosses zero. This face sheet 50 material should then be non-reflective and, if used to terminate the tuning cavity 40, should not support a standing wave or control tuning.
 However, contrary to this theory, the splitter 10 of the present invention, as illustrated in FIG. 4, provides a tuning curve similar to tests of the “state of art” splitter as shown in FIG. 2. The splitter 10 of the present invention eliminates the backing plate 30 and instead utilizes a pair of acoustically matched absorptive face sheets 52 separated by supporting cavities 40. The acoustically matched material in the face sheets 50 of the present invention is sufficiently reflective to support the tuning standing wave.
 In the preferred embodiment, the absorptive face sheets 52 should be thin foraminous sheet materials, and may be fabricated out of a variety of materials, although test have shown that either perforated metal, fibermetal, metal screening or a combination of these absorptive metals is preferred. For high temperature applications or in aircraft, the absorptive face sheets 52 may be constituted of fibrous metals consisting of randomly oriented or felted fibers, closely woven metal screens, perforated metals or layers of any or all of these. These materials may be metallurgically sinter or diffusion bonded or bonded with resins. In some applications, mechanical bonds or simple layering may be adequate. Low temperature applications may use resin bonded open weave fiberglass or carbon fiber. In general, however, any thin, porous, sheet material or sandwich that is impedance matched to ½ to 2 &rgr;c of the ambient air and is sufficiently rigid to provide internal cavity reflection will suffice. It should be appreciated that while a non-metallic material may be utilized for the absorptive face sheets 52, for example in those situations where weight considerations play a pivotal role, such non-metallic material must display the same absorptive and reflective qualities of the metallic absorptive face sheets 52 of the preferred embodiment. Soft cloth, felt or silk-like material is not recommended.
 The cavities 40 are supported by a cavity support structure 44 arranged in a variety of configurations. In the preferred embodiment, the cavity support structures 44 are configured as a honeycomb core material, or an “egg-crate”-type structure. Generally, any cell-like structure that will adequately support the absorptive face sheets 52 and provide wall structures to guide the acoustical standing waves into a normal approach to the absorptive face sheets 52 will be satisfactory. The “cell size” in the preferred embodiment is on the order of one-half the cavity depth 42. The walls may be metal or non-metallic.
 The absorptive face sheets 52 may be joined by a variety of methods, including welding, brazing, resin bonding or mechanical attachment by crimping or through the use of bolts, rivets or screws. Regardless of the method, however, care should be taken to minimize the binding of the absorptive face sheets 52 from excessive wicking of braze alloys, adhesive resins or mechanical attachment means.
 The fact that the acoustically matched material in the absorptive face sheets 52 of the present invention is sufficiently reflective to support the tuning standing wave, makes the acoustical absorptive splitter 10 of the present invention ideal for use as a silencer. First, the assembly without the septum 30, as shown in FIG. 4, is much simpler and easier to manufacture, since the solid central septum 30, two braze or adhesive bond joints 32 and one of the cavity supports (not shown) have been eliminated. The thinner splitter 10 presents less restriction to flow and, therefore, reduces pressure drop. In addition to the ease of manufacture, the elimination of the septum 30, joints 32 and cavity supports significantly reduces the weight of the splitter 10 of the present invention, making it ideal for aircraft applications. Manufacturing cost is likewise reduced through both elimination of unneeded materials and reduction in complexity. The design of the splitter 10 of the present invention also facilitates quality control, since the splitter 10 does not contain the two hidden central bond joints 32. Finally, the splitter 10 assembly can be of all welded construction utilizing state of art processing, which is not possible with the old design.
 By illustration of the simplification permitted through the use of the splitter 10 of this invention, FIG. 6 and FIG. 7 show comparative end views of single splitter duct silencers, FIG. 6 illustrating the current “state of the art” splitter using a central backing plate 30 and FIG. 7 illustrating the splitter 10 of the present invention. In both figures, the duct cross sections 20 have been sized as one-quarter wavelength, thereby placing maximum pressure at the exterior walls 70 and maximum velocity at the faces 50, 52 of the splitter. FIG. 7 shows the significant size reduction achieved by the splitter 10 of the present invention.
 FIG. 8 and FIG. 9 illustrate the use of multiple splitters in the same duct 20, FIG. 8 illustrating the “state of the art” splitter and FIG. 9 illustrating the splitter 10 of the present invention. Referring to FIG. 8, the sizes of the ducts 20, at the sidewalls, are one-quarter wavelength and the central duct 60, is one-half wavelength. The central duct 60 is made wider in the “state of the art” splitter based on the assumption that the faces 50 are non-reflective. This spacing establishes a full wavelength between the two solid septa 30 and places maximum velocity points at both absorptive faces 50.
 While the spacing illustrated in FIG. 8 can certainly be utilized with the new design splitter 10 of the present invention, if the pressure drop constraints will allow it, the arrangement shown in FIG. 9 can also be used. In this configuration, the absorptive faces 52 are sufficiently reflective to support quarter wave standing waves across the duct 20, as was found with the cavities 40. The permits significant size and weight reduction.
 FIG. 10 and FIG. 11 illustrate yet another configuration, FIG. 10 illustrating the “state of the art” splitter and FIG. 11 illustrating the splitter 10 of the present invention. Referring to FIG. 10, the central duct 60 sizes are half wavelength, and quarter wave treatments are applied to the sidewalls. Again, while this same spacing can also be utilized by the new design splitter 10 of the present invention, FIG. 11 shows an alternate configuration, where the reflective properties of the absorptive face sheets 52 have been utilized.
 The acoustical absorptive splitter 10 of the resent invention is preferably used in applications where the noise frequencies are above 450 Hertz. This includes applications such as gas turbine inlet and exhaust, fan noise (particularly when blade tip velocities exceed Mach one), hiss from the blow-down of high pressure gases, and high velocity or high temperature duct applications. The splitters 10 perform particularly well in non-rectangular ducts, i.e., round, oval, hexagonal, etc., as is illustrated in FIGS. 12 and 13. The diffused nature of the reflected waves from the curved or angled walls of the duct provide excellent excitation of the standing waves in the tuned splitter cavities 40. Tests have shown that the splitter 10 also works well with non-symmetrical walls. It is therefore not necessary to have normally oriented standing waves reflected from flat wall surfaces to achieve high insertion loss. The main criteria is the ratio of treated surface area to duct cross-sectional area.
 Furthermore, crossed splitters 10 that are tuned to complementary wavelengths are particularly useful. The cavity depth 42 of a splitter 10 will provide maximum insertion loss at the frequency approximated by the quarter wave tuning. The splitter 10 will provide very low insertion loss at the frequency corresponding to the corresponding half wave tuning, i.e., absorption peaks will occur at the first, third, fifth, seventh, etc., frequencies with deep gaps between. These gaps can be filled by the use of a second splitter that is one-half as thick as the first. It will tune to these gaps and provide insertion loss at the second, fourth, sixth, etc. frequencies. The result is a broadband silencer that is simple and inexpensive.
 Having thus described the invention with particular reference to the preferred forms thereof, it will be obvious that various changes and modifications can be made therein without departing from the spirit and scope of the present invention as defined by the appended claims.
1. An acoustic splitter for use in attenuating sound through the absorption of sound energy having a fixed wavelength, said splitter comprising a plurality of one-quarter wavelength cavities defined by a cavity support structure, said cavities being bounded on either end by an absorptive face sheet.
2. The acoustic splitter of claim 1, wherein said absorptive face sheets are composed of perforated metal.
3. The acoustic splitter of claim 1, wherein said absorptive face sheets are composed of fibermetal.
4. The acoustic splitter of claim 1, wherein said absorptive face sheets are composed of metal screening.
5. The acoustic absorptive splitter of claim 1, wherein said absorptive face sheets are composed of a combination of absorptive materials.
6. The acoustic splitter of claim 1, wherein said absorptive face sheets are composed of a porous non-metallic material.
7. The acoustic splitter of claim 1, wherein said cavity support is arranged in a honeycomb configuration.
8. The acoustic splitter of claim 1, wherein said cavity support is arranged in an egg crate-type configuration.
9. An acoustical silencer for use in attenuating sound through the absorption of sound energy having a fixed wavelength, said acoustical silencer comprising at least two acoustic splitters comprising a plurality of one-quarter wavelength cavities defined by a cavity support structure, said cavities being bounded on either end by an absorptive face sheet, wherein said splitters are spaced at quarter wavelengths.
10. An acoustical silencer for use in attenuating sound through the absorption of sound energy having a fixed wavelength, said acoustical silencer comprising first and second acoustic splitters comprising a plurality of one-quarter wavelength cavities defined by a cavity support structure, said cavities being bounded on either end by an absorptive face sheet, wherein said splitters are spaced at quarter wavelengths, said second acoustic splitter having a thickness one half the thickness of said first acoustic splitter.
International Classification: E04B001/82;