Document 7RJ8zLd3M3nqmqwLJapzNwbwR

192 CHAPTER 14 1965 Guide And Data Boole Table 1 .... Octave Pass Bands frequency. Cycles per Second A. Previously Used* Bands: Frequency/Lower 37.5* 75 150 300 600 1200 2400 4800 T.imU* \Upper 75 150 300 600 1200 2400 4800 9600 Approximate Geometric MidFrequency 53 106 212 425 850 1700 3400 6900 B. New Preferred* Bands: Mid-Frequency 63 125 250 500 1000 2000 4000 8000 Approximate Frequency/Lower 45 90 180 355 710 1400 2800 5600 linuta (Upper 90 180 355 710 1400 2800 5600 11200 * meat date an baaed on the frequency Omits bated ondar A, these bare been used in this chapter. Alt crash* in this chsptcr carry raid- iit>qocsey scale* eo thal data can be interpotated. * Many octave band analysers now in tree and pubBsbed data have a Inver limit <4 to epa far the first band. Tins does not significantly affect the design procedure for airborne eound set forth in this chapter, or say the design data. * See Reference 16. tsAere W = the sound power in watts. In this chapter, the reference value for sound pressure is always 0.0002 microbar, and for sound power is always 10"" watt In the text, these reference values usually are not re peated, in order to amplify-the presentation. When data in this chapter is compared with other publications, reference values must be considered, because other values (notably 10"u watt for sound power) have often been used. Earlier edi tions of the Guide And Data Book used 10~u watt for sound power. Sound Frequency The quality of a sound is determined primarily by, its fre quency spectrum. Unlike a tone, which has a single frequency, the sounds generated by air-conditioning system components have a multitude of frequencies. A plot of eound pressure level or sound power level against frequency is called a frequency spectrum. For engineering calculations, it would be too com plicated to deal with the levels at all individual frequencies. Instead, the audible frequency range of approximately 2010,000 cycles per second is divided into bands, usually one octave wide. The term octave means that the frequency of the upper limit of the pass band is twice that of the lower limit. An octave band sound pressure spectrum is obtained with- an octave band analyzer containing electronic, filters which pa> only those components of the measured-sound which have frequencies within the limits of the filter. The second set of octave band limits shown in Table 1 is presently specified by an ASA standard.11 New instruments and most new published data can be expected to comply with this set of limits. However, the first set given has been in gen eral use for some time, and the bulk of literature now available will be' found with this division by frequency. Data in this chapter are presented so that values, are apparent for either set of band limits. Experimental data obtained in one sys tem can usually be interpolated to the other-system--with sufficient accuracy for all practical design problems. _'j Attenuation In the design of sound control for air handling systems' it is usually necessary to calculate the extent to which sound is attenuated (reduced) as it travels through a duct system.A certain percentage of the .entering sound is lost in'each'duct' element (elbow, bare or lined duct section, sound trap, etc.). ' The ratio of the sound power entering the element to the sound power continuing beyond the element, expressed in decibels, is called the attenuation of that duct element: Attenuation, in db -- 10 log)* (W./W.) (4) tnhere W, " sound power entering a duct element, watts. . W, = sound power continuing beyond a duct element,- watts. Fig. 1 may be used to convert the ratio of sound powers in Equation 4 to decibels and vice versa. If the sound power con tinuing beyond a bare duct'section is 25 percent of the enter-" tog sound power, the ratio is 100/25 = 4, and the attenuation is 6 db. If a sound trap absorbs 09.9 percent of the entering sound, the ratio is 100/0.1 -- 1000, and the attenuation b 30 db. Since the attenuation ratio is dimensionless, no refer ence value is needed. When the sound power entering a duct element and the power continuing beyond the element are expressed as power levels in decibels, Equation 4 reduces to a simple arithmetic expression: - Attenuation, in db -- Lw. -- Lw, (5) Lw, " sound power level entering a 'duct element, decibels. Lw, " sound power level continuing, decibels. .. The attenuation of a group of elements through which the sound passes in succession is the arithmetic sum of all indi vidual attenuation values, in decibels. The maximum net attenuation that can be obtained is limited by noise regenera tion due to air flow. Since attenuation cannot.be measured directly, duct ele ments are generally rated in terms of differences between measured sound pressure levels. Depending on the way in which these imMuanwnm'tii are mad* (see Reference 3), the ratings will differ in magnitude, usefulness and name,'although the unit is always the decibel. Insertion loss (IL) ratings of duct silencers most'accurately describe the reduction of fan noise level as measured in an airconditioned room. Insertion loss data, however, are relatively costly to establish because full sets of readings have to be tftkpn both before and after the insertion of the silencer into a typical duct system. Insertion loss measurements can be made with air flowing through the duct and silencer. Noise reduction (NR) ratings of duct silencers are merely the difference in the sound pressure levels measured inside the duct at points just ahead of and just after the silencer. Due to peculiarities of wave propagation,1 the effect which' will be achieved by installing the rileniw in the actual system is usually less than the measured noise reduction. Since induct measurements generally have to be made without cur flow, the usefulness of NR measurements is limited. Transmission loss (TL) ratings are a measure of sound transmission from one room to another, and are generally used to'describe the sound isolating value of walls. When this technique is used to rate duct silencers, it is necessary to sub tract the TL of the test set-up. Table 2 .... Combining Two Sound Levels Diffmon hehxan Two UtWi To 6* 0 1 2 4 69 Number of Decibels To Be Added to Higher Level to Obtain Com bined Level 3 2H 2 IK 1 H Sou^y-Gcntrol = 193 pgcibel Calculations ' Eauatian 5 shows that calculations involving attenuation mb Amplified by the .decibel, notation system, because the "7^3 db values are simply added or subtracted. The same for the conversion of sound power levels to sound. Uresui* levels, and vice versa. To further gimplify sound con- J^lcalculatibns, all data in this chapter are presented directly - commonly used noise reduction methods in' Table 3, and some of the case histories in Reference 19. It cannot be overempha sized that every effort should be made to install the major noise sources as far away from critical areas as possible. This phase of noise control must be jointly considered in the ad- - vanced planning stage?* by the architect, the mechanical engi neer and, if be is available, the acoustical consultant.. * m However, when the decibel system' is used, it should be DESIGN PROCEDURE acted that the levels of two'or more sounds cannot be added In order to deal with each of the various noise sources and arithmetically- Table 2 may be used to combine two sound, ... paths in the simplest manner, the following seven steps should levels. If more than two levels are to be combined, the two- - , be taken: highest levels should be combined, and the1resulting value 1. Determine the design goal for the air-conditioning system combined witb the next highest level. Levels which'are more " sound for each room, according to its use and construction!;.1 - ^,n jo db below the others in a combination can be neglected. 2. Select equipment radiating directly into the room that will be quiet enough to meet the design goal. Example 1: Find the combined level of two sounds having 3. Calculate the amount of duct attenuation required to reduce levels of40 and 34 db respectively. .. ,, . central system eound to design goals. Satation:' The 'difference between the ` two : tevels'-is 1 40 ; 4. Design or select sound attenuating duct elements' .to meet _ 34 - 6 db. From Table 2, for 6 db difference,in;ievetiy' itis'- ' the requirements established in Step 3. , found that l!db must be added to the higher level to obtain the'-' 5. Make sure noise radiated by outdoor equipment' will ;nbt f^mhined level of 40 + 1 --41 db. disturb neighbors. - 6. Prevent transmission of vibration through the structure. analysis of the sound control problem 7. Examine the completed installation and make necessary adjustments. ' Every major piece of air-conditioning equipment is,. to/ some extent, a source of sound. The quantity and quality/oh 1 Steps 2 to 4 are illustrated in the text of this chapter by a sound that a.source may generate without being considered - typical problem,.covered in segments.as the material is dis-. too noisy depends on two factors: cussed (Examples 2 to 6). The same problem solution" is 1. The extent'to,which the sound is dispersed and attenuated , shown on a sound control work sheet at the end of this chap*lm>g the path from the sound source to the 'receiver (the person , ; ter. These work sheets ore available from ASHRAE. hearing the sound). 2. The m&ximam sound level which, the receiver .will consider acceptable. For. noise radiated to the inride of a'building, this is STEP 1--DETERMINATION OF DESIGN GOAL' the design goaL which is established to suit the'requirements of The degree of quietness required in each type of room in the occupants (see Design Slep.l in later text). For noise radiated to the outside, local ordinances may dictate the design goal.**-17-11 the building should be specified by the architect. On large projects and special buildings, such as music balls, it is usually In.order to meet the established desigji' goal, the system'-' -" necessary to obtain the advice of an acoustical consultant. designer may either(.reduce the source strength, or increase the path attenuation, whichever is more practicable. Due to How to Express Design Goals ' the logarithmic nature of the decibel,' this can again be ex In order to be an adequate criterion of the degree of quiet- pressed as a simple arithmetic relationship: , ness required, the design goal must .be expressed in such,a r Source Lstrength, dbj Sound Disperrion"| : -' ["m At^WH S Along Path, db J . db] '` . lTOuinpt as to define sound quality as well as magnitude.; <0f the many methods available for doing this, the following are . most widely used: (1) various criterion curves or tables defin- .' ing mftrimiiTn permissible sound pressure levels for each fre- Table 3 lists the major sound sources in an air-conditioning . quency .band (usually octave or J octave band), (2) loudness, system, together with the sound path from pa?h source to in sones, or loudness level, in phons, and (3) the A-scale sound the receiver, and gives suggested methods for sound reduc level, in decibels. tion. It should be noted that in most cases there are several Each of these methods will be discussed in some detail. parallel paths; all of which have to be considered for effective < Design Specifications Based on Frequency Band Levels. Single sound control. number systems of evaluation are useful in describing the rela-'. Sound paths are classified in two categories, (1) airborne tive acceptability of sound quality and magnitude in an occu Mid (2) solid borne. Each category must be approached dif pied space. Accordingly, .they may be useful for purposes.of. ferently because different mechanisms of sound transmission sound control specification and equipment rating in the case ere involved.*' * of equipment radiating directly into the occupied spade' (such" 1. Along'the e, ntirely by.a--ir iths 1, 3, 7, 8, and 9 (Table 3), --om the source to the receiver. sound is carried Airborne sound "Mttnufflion, whether through ducts or.through open maces, can De calculated readily. This chapter contains the necessary-data 9 ;for "ohing Equation 6 for these paths. i.1"ia hd 10, in Table 3,-contain solid'p&rtitious wtwtural elements through which the sound- must pass. In **?> l^_ "V be-possible, to calculate solidiborne,sound but the prqcedur^-^ria quitp complicated. Design iaolstirm t^se P^hfl this chapter are, therefore, limited totni awation of machine vibrations. - sure that sound reaching the listener through "tlctiire;will not .exceed the design goal,-the,engineer.has empirical methods, relying mostly on bis judgment and Psst experience. It may be helpful to, reviewthe check list of as found originating,at grilles anddiffuters). Single number systems' are not adequate, however, to express the nature of remotely located sound sources.'- Fre quency has such an important effect oh all three 'terms,of Equation 6 that this.equation must be solved separately, for, a number of frequency hands hot exceeding one' octave' in width (see Table 1). This procedure is not as time.consuming as it may seem because, with some experience, it is usually posrible to pick two or three octave bands which could be the; critical ones in a given situation. ' . / '!', ' While a breakdown into octave band levels is sufficient, for design specifications and-calculations, it-is necessary'-to-uae narrower band widths when establishing ratings'fo'r 'sound sources producing pure tones.1 Pure tones are' often'foiindrm