US 2043731 A
Abstract available in
Claims available in
Description (OCR text may contain errors)
June 9, 1936. R. B. BOURNE 2,043,731
SOUND ATTENUATING DEVICE Filed Feb. 17, 1956 2 Sheets-Sheet 1 Z Z. 0 Z! Z} M I INVENTOR 304 :4/70 E. Boa/ME *i/ue ATTORNEYS June 9, 1936. RN 2,043,731
SOUND ATTENUATING DEVICE Filed Feb. 17, 1936 2 Sheets-Sheet 2 w BY @414 M ATTORNEY? Patented June 9, i936 UNHT FiE
The Ma Silencer Company,
Coma, a corporation of Connecticut Application February 17, 1936, Serial No. 64,223
The present invention relates to silencing means suitable for quieting the exhausts of internal combustion engines, blowers and similar sounds, while imposing a minimum of back pressure to the flow of exhaust or other gases therethrough.
There are, in general, two distinct methods previously available for attenuating sound waves traveling through an unobstructed pipe or conduit; attenuation by means of sound absorption in the walls of the passage, which may be lined with sound absorbing material, and attenuation by means of acoustic sidebranches suitably coupled to the conduit. Both of these means have been widely used in industrial sound attenuating. The former method is eifective in attenuating-sounds of relatively high frequency, While the latter method is readily adaptable to efficiently attenuating sounds of relatively low frequency.
In the silencing of complex sounds containing many sound frequencies over a very wide frequency range, it is generally necessary to resort to both of the above means for obtaining satisfactory attenuation. The attenuation vs. frequency characteristic, which I shall hereinafter refer to as the plot" germane to some particular device, is very useful in describing the functioning of the various embodiments of the invention to be described herein. The plot of a silencer employing reactive acoustic sidebranches only is a peaked curve and may be very irregular in shape. In other words, the attenuation for a given sound frequency may be very high, while for a. sound frequency but little difierent may be so low as to be of no practical use in silencing industrial noises. Such irregular plots are the result of the variation, with the frequency, of acoustic impedance of the sidebranches and f the main sound conducting channel to which such sidebranches are acoustically coupled; and more particularly are due to the fact that, as the frequency is varied, the impedance of the sidebranches does not change at the same rate as does the impedance of the main channel. The plot of a silencer employing a purely absorptive, lined channel, on the other hand, is a relatively smooth curve increasing with frequency. The attenuation at relatively low frequencies is so small, however, that excessively large quantities of sound absorbing material must necessarily be used, making the device unwieldy and costly. What is desired is a device which will offer a uniformly high degree of attenuation to a continuous range of sound frequencies such as are likely to be encountered in silencing industrial pipe-borne noises in the field. Such a device may be installed with positive assurance that no sound frequencies can escape therefrom without suifering the desired degree of attenuation. This ideal device would have no pass bands or regions of low attenuation excepting at or near zero frequency (since the exhaust or other gases must be passed without imposing any undue amount of back pressure and would have no high sharp attenuation peaks in its plot to yield unnecessarily high attenuation to a few very narrow ranges of frequencies. The attenuating efficiency, considering all factors, would be a maximum and would be uniform throughout the operating sound spectrum.
It is a prime purpose of my invention to provide a method-and a construction whereby the ideal results as hereinbefore discussed are practically realized. 1
It is a further purpose of the invention to provide a universal silencer of simple construction and low manufacturing cost.
Another purpose of the invention is to provide.
a silencing unit whereby several such units may be cascaded to provide any degree of attenuation desired over a wide range of sound frequencies In the following discussion and description of the principles and embodiments of the invention, use is made of graphs and an approximate theory whereby the functioning of the devices is explained, and the improvements over previously known devices are explained.
Referring now to the drawings,--'
Fig. 1 shows a schematic representation of an acoustic line;
Fig. 2 shows a typical embodiment of the invention;
Fig. 3 is a cross sectional view of the device of Fig.
Fig. 4 shows an acoustic device presented for comparison purposes;
Fig. 5 shows anembodiment of the invention employing a single section;
Fig. 6 is a side view of the device shown in Fig.
Fig. 7 is a view similar to Figs. 2 and 4 showing a different form of structure;
Fig. 8 shows a ventilating duct employing the principles of the invention;
Fig. 9 is a cross sectional view of the duct shown in Fig. 8;
Figs. 10, 11 and 12 show aphs germane to various devices shown; and
Figs. 13, 14 and 15 show composite structures embodying the principles of the invention.
All of the devices shown herein have straight sound and gas conducting passages, although the application of the principles of the invention is by no means limited to such channels or conduits.
In all the devices shown herein, acoustic resistance plays an important part. The acoustic resistance of the main channel may be taken as the radiation resistance of the medium. The resistance in the sidebranches is predominantly that due to viscosity in the interstices of the sound absorbing material used. The specific material used will depend to some extent upon the use to which the device is to be put, and particularly upon the conditions of heat, moisture, and deposit formation of the gas passing through the device. For conditions involving no substantial heat, moisture, or deposit formation hair felt may be used. For other purposes resort may be had to other standard materials such as metal wool, vermiculite, etc. This viscous resistance varies greatly with different materials and with the manner of using them. It is, for instance, increased for a given material by packing such material into greater compression. Absorption of moisture, etc., changes the viscous resistance of the material. This is one reason why it is somewhat difficult to present an exact acoustic theory of my invention which will accurately predict quantitatively the actual acoustic performance of a device built in accordance with the principles disclosed herein. An approximate theory has been worked out which is useful in explaining the unique performance obtained.
Fig. 1 represents an acoustic line having series impedances Z1 and shunt impedances Z2. It can readily be shown that the propagation constant per section, P, can be expressed in terms of Z1 and Z2. P is in general a complex quantity, having real and imaginary components. We may write This may be also written explicitly for the propagation constant as quantity P. In the devices of my invention, phase change is of little importance compared to attenuation and we may study attenuation only by simply considering the absolute value of the quantity where H is the quantity whose logarithm to the base e is P in equation (2). It will be shown that the devices of my invention yield measured plots of attenuation vs. frequency which are substantially flat over a wide frequency range. This means that the expression (3), above, must be independent of frequency over the range involved. In my invention, I consider the main acoustic line to comprise an inertance in series with a resistance. The sidebranches I consider to comprise a resistance, inertance and capacitance, all in series. For the frequencies involved over the 'flat portion of the characteristic, the capacity reactance is relatively negligible. If now the impedances Z1 and Z2 be considered to have the same phase angle for any one frequency, we may say that Z1 is equal to a real constant In times Z2. This would be true of all he quencies in the range. Making this substitution in equation (3), we have I: W H= +1+ /k+ (4) Since it is impossible to predict precisely what the value of I0 is for a given device, we may assume a number of values and; by noting the actual attenuation, see readily what the probable value of la is for a given device. Since k is a factor which applies to both the resistance and inertance, it is entirely feasible to predict that to increase the attenuation for a given device It must be made as large as possible without departing from the desired independence of the attenuation per section from variations with frequency. In connection with some of the embodiments of the invention herein shown, it will be apparent that the devices do function in accordance with the simple theory outlined above. Certain graphs will be analyzed in accordance with the theory.
Turning now to specific embodiments of the invention, Fig. 2 shows a cross sectional longitudinal view of a typical silencer. It comprises the generally cylindrical casing l, of diameter D, fitted with perforate end closures 2, 3 at either end thereof, a centrally disposed tubular member 4 of perforated metal having a diameter d which forms the straight main sound and gas conducting channel 5 extending from one end of the device to the other. The laminated transverse headers 6 extend from the outer casing I to the tube 4 to form the sidebranch compartments ll each of longitudinal length L. Contiguous to the outside of the perforate tube 3 is a relatively thin, uniform layer'of sound absorbing material 8, held in place by the outer perforate tubes 9 of diameter d which extend from one transverse header or partition to another, as shown. The volume of each cavity 1 is made sufiiciently large so that the capacity reactance thereof is small for the frequencies for which the theory applies. The layer of sound absorbing material 8, together with the two perforate members 4, 9 form the acoustic resistance and the acoustic inertance of the sidebranch, and.
in the major portion of the frequency range of the device can be considered as forming the major portion of the impedance -of the sidebranch, since as pointed out above the impedance of the cavity beyond the sound absorbing material is very much less than the impedance of the material itself.
The inertance of the sound absorbing material, together with the capacitance of the sidebranch, shows, for some dispositions, a desirable tendency toward resonance at a frequency lower than the range of frequencies for which the above elementary theory holds. The net result is that the attenuation is held up to a higher value, as the frequency is lowered, than would obtain were the phenomenon of resonance in the sidebranch entirely absent. The attenuation at the resonant frequency is of the same order of magnitude as that obtained at higher frequencies. The attenuation falls ofi sharply on the low frequency side of resonance, as might be expected. The
resonance frequency can be computed with sufiicient accuracy on the basis of efiective inertances determined experimentally, and the known capacitances. The resistance of the layer of sound absorbing material plays a less important direct part at such low frequencies, but is suificient to produce a very broad resonance peak instead of the usual sharp peak, ordinarily associated with resonators having a minimum of absorption. The use of sound absorbing material as described for coupling the main sound conducting channel to the chambers not only produces a generally flat attenuation characteristic throughout the medium and high frequency range, but also serves to lower the frequency response of chambers of given physical dimensions below that which would be obtained by conventional methods of coupling. These two effects combine to produce a wholly novel uniformity of attenuation throughout the frequency range being operated upon, and to permit the use of shorter sections than would otherwise be possible. As is elsewhere pointed out, the use of relatively short sections is desirable in producing the greatest possible uniform attenuation for a given length of apparatus. It is a feature of the invention that the same characteristics are obtainable from a single section as are obtained from a device having several sections. Thus the device does not depend upon interactions between adjacent sections for its functioning. Adding sections results in additive attenuation.
Curve A, Fig. 10, is the measured plot of a device of four sections constructed in accordance with Fig. 2 for the condition that L=D=3d. d'=1.5d. In this as in other plots herein shown the fre quency is represented by arbitrary abscissae, since the actual frequency is a function of the physical dimensions of the device. The slight hump at point 35 is due to a tendency to resonance in the sidebranches. The slight dip at point 36 is due to series resonance in the sidebranch and occurs at a frequency for which L equals a half wave length. The insertion loss in decibels is plotted as ordinates, the device being inserted in a long acoustic line. Curve A is for the condition that the transverse headers 6 are of plain, non-laminated metal.
Curve B shows the plot for the device when laminated headers are used. The increased attenua tion is due to the fact that the headers B transmit less sound by diaphragm action into adjacent sections, tl'nis preventing partial short-circuiting.
Curve C, Fig. was obtained from the device of Fig. 4, which is of the same size and proportions as that of Fig. 2 and comprises a casing in, perforate end headers M, 92, a, centrally disposed perforate tube is which forms the main conducting channel it; and transverse headers i5 which form the compartments l B. These compartments 0r sidebranches are completely filled with sound absorbing material, there being approximately six times the amount used as in the case of the device of Fig. 2. The device of Fig. 4 represents a common form of muiiier in use commercially. The improvement in attenuation due to my invention is marked, especially at the lower frequencies, even though but one-sixth of the sound absorbing material is used. It is clear from this comparison that the operation of devices constructed in accordance with my invention is wholly different from that of devices constructed in accordance with Fig. 4 in which it is attempted to secure attenuation solely by the operation of the sound absorbing material itself. This will be even more evidentfrom the form of the invention next to be described.
Curve D, Fig. 10 was taken with a device made in accordance with Fig. 2 for the condition that D=3d; d'=l.5d, L=.5D. Four sections were used. The attenuation is much greater, even though the overall length of the device is half that used to obtain curve A. The resonance frequency in this case does not show itself as a hump in the curve,
but is in fact higher due to the fact that, while the volume of the sidebranch has been cut in half, the conductivity of the layer of sound absorbing material is only about .7 its former value. This is due to the fact that the conductivity of the cylindrical area of the layer is proportional to the diameter of a circle of equal area. of perhaps more importance is the remarkable flatness of the attenuation frequency curve. Curve E was obtained from a four section device similar in all respects excepting that L=.25D. The attenuation is even more uniform; in fact it remains at 20 decibels over the entire operating portion of the frequency range. The attenuation for the device of curve E is more than half of that for the device of curve D, showing that there is some advantage in the shorter sections. Had the device of curve E been made as long as the device of curve D, therefore including twice the number of sections in the same overall length, the total attenuation would have been substantially greater.
The length of a section must not be made too long with respect to its diameter A device was made in accordance with Fig. 7 wherein L=4D, the other proportions remaining the same. The parts being the same as in Fig. 2, except for the elimination of the partitions 6, they have been indicated by primed numerals. In all devices for which curves are given in this specification, d is the same. Curve F, Fig. 11 is the plot for the device of Fig. 7. The dips 31 are due to series resonance longitudinally in the side branch. The average attenuation is relatively low. Such proportions do not represent an economic use of material, and as will be apparent the device of Fig. 7 lacks the desirable fiat-top characteristic of curves D and E. The length of the section in Fig. 7 is a very appreciable part of a wave length and the tendency is greatly increased for sound to enter the side branch at one end, travel longitudinally therethrough, and re-enter the main channel at the other end thereof. For best approximation to the conditions implied by the 'heory, the direction of the sound wave entering the side-branch should be normal to the surface of the sound absorbing material. Indeed experience with devices of this type shows that maximum possible attenuation, for a given total available length is aiforded by the use of very short sections, other things being equal. A compromise must generally be effected, however, since too many transverse headers raise the cost of manufacture. To obtain low frequency response, it is preferable to use short sections of large diameter, rather than long sections of small diameter, for the reasons above given. The lengths of the sections are preferably made less than their diameters. For economic reasons it is generally desirable to form the sound absorbing layer as a continuous body, so that each cavity is coupled to the main sound conducting channel through sub stantially its entire available coupling area. This is also desirable for acoustical reasons, since as is elsewhere pointed out the attenuation is improved by keeping Z2 low provided this is done without departing from the condition that the ratio between Z1 and Z2 is maintained independent of frequency. This is best accomplished by using a large coupling area. A further advantage in this manner of construction is that the entire length of main channel contained"within the device is acted on, there being no spaces intermediate constructed in accordance with Fig. 5, wherein D=5d, L=.3d and d'=2.5d. Curve G is for a unit amount of sound absorbing material I! uni-,
formly distributed. This curve is seen to be relatively flat and of average attenuation comparable to one section of the device of Fig. 2 which yielded curve A, Fig. 10. Curves H and I are for the same device, but with one-half and one-quarter unit amount of sound absorbing material, respectively, the same being uniformly distributed in each case. Our elementary theory shows that in order to increase the attenuation, It must be larger. One way in which to increase k is to decrease Z2. In the case of these curves G, H, I, it is apparent that the attenuation does increase with a decrease of Z2, since, by decreasing the amount of sound absorbing material, the resistance is decreased and the inertancelikewise decreased. Since the curves are each of the same type, it is evident that both the resistance and inertance have been changed without greatly changing the phase angle between them. Curve I is beginning to depart from the conditions imposed by the elementary theory. Decreasing Z2 to a limit would result in the ordinary resonance curve.
Disposing the sound absorbing material as shown in the several devices described in this specification produces the results described. The same amount of sound absorbing material disposed contiguous with the outer shell yields a wholly different plot, having high attenuation for a relatively narrow range of frequencies and relatively low attenuation over the remainder of the frequency spectrum, especially at low frequencies. The reason for the improvement in the shape of the plot may be further said to be that in my invention, the sound absorbing material is disposed at a point of high acoustic velocity whereas having the material disposed contiguous with the outer shell instead of at the entrance to the sidebranch, places the material at a point of low acoustic velocity.
In Figs. 8 and 9 is shown a ventilating duct treated in accordance with my invention, and illustrating the adaptability of the invention to conduits of other than circular shape. In this case the duct forming the main sound conducting channel 20 is formed by side members 2| and perforated members 22. The side members 2! extend beyond these perforated members and are joined by members 23 so as to form opposed closed spaces on either side of the main channel. Perforated members 24 are located parallel to the members 22, the space between them being filled with sound absorbing material 25. Partitions 26 extend between the members 23 and 25 to divide,
the longitudinally extending chambers into short chambers 21. The partitions serve as described above to prevent any substantial longitudinal travel of sound waves through the chambers, preventing by-passing of the sound waves in the manner referred to in connection with Fig. '1. Depending upon the amount of attenuation required, and forming a practical balance between economy of construction and theoretical improve-.
the sidebranches occurs at comparatively low fre-- .tween the cylinder 43 and the. inner perforated quencies. Where the lengths of the sidebranches are shorter as in curve B (Fig. 10) the effect of such resonance is shifted to still higher frequencies, and where the length is still shorter, as in curves D and E (Fig. 10) resonance in the sidebranches may be shifted to so high a frequency as to be of little importance for the ordinary sound spectrum. The shorter the section, other things being equal, the flatter t e plot over the usual frequency range. It is 1 preferred to keep the lengths of the sections shorter than one-half wave length for the highest important frequency which the device is to be called upon to attenuate. The use of excessively short sections is, however, undesirable on account 1 of the added expense and on account of the raising of the point at which low frequency resonance occurs. It is possible by combining sections of different characteristics to utilize both the flattopproperty of short sections and the low frequency response of longer sections, or to simulate these properties by other means.
Fig. 13 shows a composite structure having a number of sections of different lengths, other factors being equal for each section. This structure utilizes perforated tubes 40 and 4|, between which is a layer of sound absorbing material 42. An outer cylinder 43 is provided as in the case of Fig. 2, fitted with end headers 44 and 45. Be-
tubes extend partitions 46, in this instance so spaced as to provide a plurality of chambers 41 of differing sizes.
Another method of obtaining a composite structure is shown in Fig. 14. Two sections only are shown, the device comprising the casing 50, the perforate tube 5| forming the main channel 52, the two sidebranches 53, 54, having the layers of sound absorbing material 55,56,1'espectively. The material 55 is packed more densely than the material 56, thus yielding a greater resistance and inertance and giving to the two said sidebranches different frequency response characteristics.
The composite structure of Fig. 15 comprisesthe casing 60, the main channel 6|, the sidebranches 62, 63 and the layers of sound absorbing material 64, respectively. In this case, the layer 64 is made thicker than the layer 65. This construction is useful where the sound absorbing material is available in one density only and is not susceptible of different degrees of packing into its confining space. It is obvious that a composite structure could also be formed by sections of different diameter. This type of construction is not always feasible due to the constructional dimoulties, as well as the resulting appearance of the device. It is further to be stated that the sound absorbing layers shown in the variousembodiments of the invention may be replaced with suitable constructions not involving the use of fine, porous material. The essential thing is to have an acoustic resistance and inertance at the entrance to the sidebranch.
1. An acoustic sidebranch comprising a body of sound absorbing material in series with a closed cavity.
2. A sound attenuating device comprising a main sound conducting channel and one or more closed cavities each coupled to the main sound conducting channel through a body of sound absorbing material.
3. A sound attenuating device comprising a main sound conducting channeland one or more closed cavities each coupled to the main sound? the effect of series I conducting channel through a body of sound abpassing of sound waves through the chambers sorbing material, the acoustic impedance of the cavity beyond the sound absorbing material being substantially less than the acoustic imped ance of the sound absorbing material.
4. A sound attenuating device comprising a main sound conducting channel and one or more closed cavities each coupled to the main sound conducting channel through a body of sound absorbing material, the distance the sound absorbing material extends along the main sound conducting channel in each coupling zone being substantially shorter than one-half wave length for the highest frequency which the device is designed to attenuate.
5. A sound attenuating device comprising a main sound conducting channel and one or more closed cavities each coupled to the main sound conducting channel through a body of sound absorbing material, the acoustic impedance of the cavity beyond the sound absorbing material being substantially less than the acoustic impedance of the sound absorbing material, and the distance the sound absorbing material extends along the main sound conducting channel in each coupling zone being substantially shorter than one-half wave length for the highest frequency which the device is designed to attenuate.
6. A sound attenuating device comprising a cylindrical tube forming a main sound conduct-' ing channel, an annular chamber surrounding the tube, and a body of sound absorbing material interposed between the channel and the chamber and furnishing the sole acoustic coupling between them, .the tube being perforated to permit passage of sound waves from the main sound conducting channel into the chamber through the sound absorbing material.
7. A sound attenuating device comprising a main sound conducting channel, a chamber surrounding, the channel and divided by transverse partitions into a plurality of acoustically separate compartments spaced longitudinally along the channel, and a body of sound absorbing material interposed between the channel and each compartment and furnishing the sole acoustic coupling between them.
8. A sound attenuating device comprising a cylindrical tube forming a main sound conducting channel, an annular chamber surrounding the tube and divided by transverse partitions into aplurality of acoustically separate compartments spaced longitudinally along the tube and each having a length parallel to the axis of the tube substantially less than the diameter of the chamber, a body of sound absorbing material interposed between the tube and each compartment and furnishing the sole acoustic coupling between them, the tube being perforated to permit passage of sound waves from the main sound conducting channel into the compartments through the sound absorbing material.
9. A sound attenuating device comprising a main sound conducting channel, a chamber surrounding the channel and divided by transverse partitions into a plurality of acoustically separate compartments spaced longitudinally along the channel, and a body of sound absorbing material interposed between the channed and each compartment and furnishing the sole acoustic coupling between them, the length of the coupling zones of the sound absorbing material in a direcalong the main sound conducting channel.
10. A sound attenuating device comprising a main sound conducting channel and one or more terial interposed between the channel and the v chamber and furnishing the sole acoustic coupling between them, and means within the chamber for dividing it into a plurality of cavities each independently coupled to the main sound conducting channel through a body of sound absorbing material, the length of each cavity being suificiently short in proportion to its diameter so that the attenuation will be substantially uniform throughout the sound spectrum to be attenuated.
12. A sound attenuating device comprising a main sound conducting channel and one or more closed cavities each coupled to the main sound conducting channel through a body of sound absorbing material, the length of each body of sound absorbing material being substantially the same as the'length of the cavity with which it is associated, the lengths of at least some of the cavities being made difierent from the lengths of other cavities so that a composite structure is obtained having substantially uniform attenuation throughout the sound spectrum to be attenuated.
13. A sound attenuating device comprising a main sound conducting channel and one or more closed cavities each coupled to the main sound conducting channel through a body of sound absorbing material, the length of each body of sound absorbing material being substantially the same as the length of the cavity with which it is associated.
14. A sound attenuating device comprising a main sound conducting channel and one or more closed cavities each coupled to the main sound conducting channel through a body of sound absorbing material, the amount of sound'absorbing material per unit area in some of the bodies being diflerent from the amount thereof in other of the bodies.
15. A sound attenuating device comprising a main sound conducting channel and one 'or more closed cavities each coupled to the main sound conducting channel through a body of sound absorbing material, the density of the sound absorbing material in some of the bodies being different from the density thereof in other of the bodies.
16. A sound attenuating device comprising a main sound conducting channel and one or more closed cavities each coupled to the main sound conducting channel through a body of sound absorbing material, the thickness of the mass of the sound absorbing material in some of said bodies being different from the thickness thereoi' in other or the bodies.
ROLAND B. BOURKE.
CERTIFICATE OF CORRECTION.
Patent No. 2,045,751. June 9 1956.
ROLAND B. BOURNE.
It is hereby certified that error appears in the printed specification of the above numbered patent requiring correction as follows; Page 1, second column, line 5, after the word "pressure" insert a parenthesis; page 4, first column, line 2, for ".5d" read .5D; and that the said Letters Patent should be read with these corrections therein that the same may conform tothe record of the case in the Patent Office.
Signed and sealed this 11th day of August, A. D. 1956.
Henry Van Arsdale (Seal) Acting Commissioner Patents.