Document LZyXKwkdrGQKajwbn4Rxe4rd

204 CHAPTER 14 1965 Guide And Data Book Table 11 .... Power Level Division at Brandi Take-Offs Arse of Cawffawtoo Dttd m Percent of the Total Area of aO Owcfi after Brandi Tafcw-Off 5 10 !5 20 SO 40 50 80 Decibels to be subtracted from power level before take off in order to get power level in continuing duct 13 10 8 7 5 4 3 1 Table 12 .... Approximate Allotment of Fan Sound Power Level to Each Air Outlet Cfn of OeHaf in Percent of Total Fan Cf X Vt 1 2 S 10 Decibels to be subtracted from fan power level in order to get fan sound power level per outlet 27 23 -20 17 13 10 20 7 50 3 W -- horsepower input to the fan. p, static pressure across fan, inches water gage. The term 10 log (IFp.) can be read from fig. 12. The base sound power levels which can be estimated from Table 10, depend on how well the fan has been sized for the particular operating condition and also indicate the relative quietness of a particular type of fan. Sound output of centrifugal fans should be estimated from Table 10 only when actual test data cannot be obtained, be cause this table is based on rather limited published data.**-** As more test data become available, Table 10 may need revision. Note that the values derived in this manner represent $ the total sound power output from the fan, or approximately:the amount expected from either inlet or discharge. .. Example S: For the optimum size fan of Fig. 11, estimate the approximate octave band power levels of the noise delivered to the supply duct of an air-conditioning system. Solution; From Fig. 11, the fan power input at 6000 dm is approximately 2 hp. The base sound power levels can be ex pected to be close to the maximum values for an efficiency of over 65 percent in Table 10,- say. 1 db below the maximum. These values of base sound power level, together with the value of the term of 10 k>gt (FPp), found from fig. 12 (for IF = 2 hp and p,' " 1.5 in.) are in lines 1 and 2 below. On Line 3, a 5 db allowance for the blade passage peak is added in the 150- 300 cps band. The approximate sound power levels of the fan, obtained by adding Ones 1, 2, and 3 are entered in line 4. Octave Band Center, cps 106 212 425 850 1700 3400 6900 84 79 74 69 64 59 54 2. 10 log,, (Wp.)................. 5 5 5 5 5 5 5 3. Blade'Passage Peak....... -- 5 -- -- -- -- -- 4. bwf................................... 89 89 79 74 69 64 59 accurate, it is also quite time consuming when there are many branches. Asuming that all branch ducts are sized for approximately equal air flow velocities and that ducts and fittings are de signed and sized so that air flow noise is insignificant, the fan power level for each outlet can be estimated in a tingle step, from Table 12. The error introduced by this simplified method is generally not greater than the tolerances of the present data on tiie power level of fan*. Attenuation of Untreated, Ductwork . The total natural attenuation, in decibels, of duct work not acoustically treated is the of: (1) the transfer of acoustical energy to the duct walls, (2) the reflection of sound waves at the open end of the duct, and (3) the sound reflection by elbows. . Table 13.... Approximate Natural Attenuation in Bare Rectangular Sheet Metal Ducts1** Duct Stza, in. Odova Bond Center, cp* 53 106 ' 212 Above 63 125 250 250 ' Attenuation, db/ft , Small Large 6X6 24X24 72X72 0.2 0.2 0.1 0.2 0.2 0.1 0.15 0.1 0.1 0.05 * if duet a covered withthermal tnanlatlng material, attenuation will fa* Published data on axial flow fans at this time are insuffn dent to set up a general table covering not only the broad band part of the spectrum, but also the narrow band peaks: Axial flow fan* are frequently used in self-contained air-con ditioning ' units, which often provide considerable sound attenuation and contain sound sources other than the fans: Round ratings of self-contained units, with either axial or cen trifugal fans, should be based on tests of the complete unit. Sound Distribution At branch-off points, the sound power flow is divided in approximately the gune ratio as the duct areas in the duct sys-' tern (see Table 11). Hus power division is practically indepen dent of frequency and therefore applies to all octave,bands.' While this area ratio method of calculation is reasonably (e) Above ewW* baaed on directivity factor, 0*2, o roe* Birroc* mww net rrora gi*r w* ^ (fa) tf opening' it. at-the junction of two room cwrface*, Q -- 4, eod refection pttunucrfwa wtf be redoced to that of a dod with twin <f one. (1.4 too* the radar* or tqwwo root of orooj .v . (c) far other rehe* of O.'mMpfy ochnJ area by 0/2 far eqeindent ' area fa be ated fas this flgvn. . Fig.1 3 . . Sound Reflection at Air Outlet . Sound;ControK r hi ^cVv 205 Table 14..... Approximate'Attenuation'of'Round Elbows Jwenfinnt .* , *' ,i`*41-80 Oefarw Band Center, cp* 125. 212 425 250 . 500 650' 1700 3400 6900* 1000 2000 .4000 8000 Attenuation, db 0 .0 0` 0 0- -2 0 . '2 3 12 23 3 3 ' 3 - 3` 3 .3, 3 3. 33 3 - '3' The walls of rectangular sheet metal ducts are readily set' into vibration by sound waves, especially at low frequencies." As a reult, sound is radiated to the space surrounding the ducts at the expense of the sound within the duct. The natural attenuation effect of bare ducts listed in Table 13 may swm small; but in long ducts, it may provide a very significant amount of low frequency attenuation. If bare ductwork is located within the air-conditioned room, most of the sound will still remain in the room. Untreated ducts should be run only through non-critical areas. , i; When ducts are covered with thermal insulating material, wall vibrations are damped. This will not only lessen the sound rsrliatinn from the duct walls,but will also increase the attenu ation effect0 to about twice the values shown in Table 13. The .walls of round ducts used in high velocity systems are not set into vibration as readily as those of rectangular ducts. For diameters of 4 to 12 in., the natural attenuation with or without external thermal insulation is about 0.03 db per foot below 1000. cps, rising irregularly to 0.1 db per foot at high frequencies.* ' Only a percentage of the low frequency sound energy in a duct is radiated from the air outlet, because the sudden enlargement of the duct to the wide open room acts' as an invisible sound reflecting barrier. For a 10 X 10 in. duct out let, for example, 90 percent of the sound power in the 75-150 cps band is reflected back into the duct system, and only 10 percent is radiated into the room. In decible notation, this amounts to'a;10-db reduction unsound power level'in'that band. Fig. 13 shows the end-reflection attenuation for various duct rises as a function ed frequency! The duet size to be used for calculating end-reflection attenuation is that which pre vails for several wavelengths before the opening. No consider ation,need be given to the actual size of the opening itself or to tire cross section of a very short length (up to several inches) preceding it. Sound reflection is also responsible for the attenuation effect of elbows. In round elbows, much of tire sound is re flected around tire elbow rather than back to the fan. The attenuation of such elbows,' as shown in Table' 14,* is relatively smalL In square elbows,* however, tire attenuation (Bow A of Table 15) is considerable, especially at frequencies where the wavelength of the sound is in tire same order of magnitude as the elbow width D (see Fig. 14). Noise From Duct Fittings and Mixing Units 1 Air flow generates sound energy in elbows, vanes, sound traps,- or-any other element of the duct system which creates turbulence. Sound power levels for 90-deg elbows with'turn ing vanes0 are shown in Table 16 for a duct velocity of 2000 fpm. This table is also valid for 30, 45, and' 60 deg elbows. Vane shape and size have little effect on tire noise generated, Fig. 14 .... Diagram of Square Bbow without Turning Vanes Table 15 .. . .'Attenuation of SquareBbows without Turning Vanes,* in Decibels Octave Sand Center, cps * 53 106* 212 425 850 1700 3400 6900 .'63' .125 .250 500 1000 2000 4000 8000 (A) No lining . 5' Duct width (D) 10* Duct width "AT Lhict width . 40' Duct width. () Lining* Ahead of Elbow 5' Duct width - 10* Duct width 20* Duct width 40* Duct width (L) lining* After. Elbow 5' Duct width . 10* Duct width ' 20* Duct width 40* Duct width (L>) lining* Ahead of and After Bend ,^ 5* Duct width"' ` 10* Duct width 20' Duct width - 40' Duct width 4 4 '8 "4 ' .8 8 '55 ,3. 8' 5 53 33 3 `3 " S' 3 3 3; 1 V 8 "6 ' 7 ' -"i 8 6 -7 11 4 8- 6 7 11 11 4 8 6 i7 11 ir 11 1 4 11 "'IO-' `10 4 11 10' 10 10 ~4 11 10 10 10 10 4 11 10 10 10 10 10 1 '-4 11r ' 13 15 4 11 13 15 18 4 11- 13 15 18 18 4 11 13 15 - 18 18 18 1 "1 15 57 1 15 5 .8 1 16 6 11 1 ' 5 57 7 5 53 1 '.5' 58 .8 6 6 '8 6 6 11 11 10 10 10 7 .5 . .3 5 .. 3, 3 3 . 3. : 3 3 8 6 :$ 6 8 11 8 11 ,11 1 11 11 ir 11* 10., 10 10 10 10 10 10 10 10 10 10 1 6 12 14 16 1 6 12 14 16 18 1 6 12 14 16 18 18 6 12 14 16 18 18 18" 4 on lining extending for * dirt*dob et it least two dact widths "D" and lining un*a length uost he looser. * tor aquare elbows with ahart taimng vine*, use avenge between Tahie* 14 and IB. erf 10% <rf duet width *'D." (See Fig. 14). For thinner Enin*. .................... - >