Reduction of the Compressor Inlet Infrasound Emission

The infrasounds attenuation is inefective mainly because of the lon~ sonic wave diffraction, so that the reduction of the infrasounds source is the only acceptable method. Unfortunately low frequency pressure pulsation has the great influence on the compressor operation. -Each change in suction line results in change of compressor volumetric efficiency. The paper presents examples of numerical investigation and results of some experimental verification for several inlets with mufflers of different types. The infrasound emission reduction, its efficiency that influence the compressor operation and correlation with mufflers dimensions are in th9 centre of interest. INTRODUC'IION Thoup;h the percerJtion of sound by the human ear is limited to frequency higher than 16-20 Hz, as the man responce to infrasound ex:oo R-,_lre, the mechanical, physiological ana psychological effects are observed. The low speed air compressors suction installations very often are one of a strong sources of infrasounds. The one of such installation became a subject of presented investigation.Vibration was the prob1em. The measurements carried out near the house located about 150 m from the compressor plant gave the results shown in Figure 1 and J!'igure 2. Figure 1 presents acoustic spectrum measured near the house and Figure 2 is the spectrum of acceleration of the window glass. Some further measurements excluded other sources of vibration and the coclusion was that the mechanical vibration is due to the acoustic waves. Noise rating curves additionaly drawn in "F'igure 1 show that rather high sound pressure levels are not annoying from the acoustic roint of view. On the other hand, mechanical vibration in the low frequency range caused oy the infrasound components produce annoyances which are individually very different. 426 So the -rroblem of reduction of vibration was ac~om~anied by the problem of the ini'rasound l'eduction. OBJECT 0 ,~' HTVlDSTHi-A'J'ION' The considered col!lnressor plant was equiped with three two-stage dc•'Jle-acting reciprocating compressors VC 750. Each comyressor had its 0\>111 sepaprate suction 1 i .e / Figure 3 /. In Figure 4 the results of computation and the results of measurements conducted at one of the com~ressors inlet are presented. The small schematic drawing in Figure 4 explains used nomenclature. I1ETHOTIS OF AT'rENU AT TON Attenuation of infrasouncls in far field is rather problematical because of very long waves diffraction. So the only practical method of sound pressure levels decreasing is the reduction of sound source. The several different methods of infrasound emission have been cosidered. The schemes of considered types of mufflers ~,2,3,4, 5] are shown in Figure 5. The simplest method is the change of the suction pipe length /b/. It is possible to use a multiple expansion chamber /c/, or Helmholtz resonator type of muffler /d/. Also resistive types of mufflers /e,f/ have been taken int0 account. REQUIREJV[ENTS Each method of attenuation has its own advantages and disadvantages [1,2,3] • So it is necessary to specify technical requirements for mufflers. There are many different requirements [5,6] but three of them seem to be the most important. They are: sufficiently high attenuation, little influence on compressor operation and small dimensions. And in this case cosideration was limited only to the re~uiremen­ ts mentioned above. MATHEMATICAL IJIODln The simple mathematical model of compressor and. suction installation has been used to optimise the muffler configuration. The overal mathematical model consists of compressor cylinders model, model of flow through the valves, suction line model with mufflers and acoustic model. Compressor cylinders model For modeling the double-action cylinder a modified yolytropic compression equation was used[ 4 J • Assuming that the density during suction and discharge period is constant one can get dPc _ n Pc ( dVc . dt T 0vs0vddt)' for the outer and inner side of cylinder where If. = A1·n r ( 1-cos(.J t) + 1/.:m. m m and For reference purposes crank angle '(J~CJt is considered zero at bottom dead center. Ivlodel of flow through the valve Because of limited solution only to the low frequencies, sim:')lified model of flow [7, 8] was used Oy= (X F V2 6.P/J) Qv= 0; for both suction and discha.r.o;e valve. Suction line model with mufflers The lumped parameter rr.odel of nonstationary flow seemed to be sufficiently accurate [4,8,9,10] for low frequencies oscillations. So the set of differenciaJ. eouations dQ dt dp dt = _J_ !.':::. p Ma _1_ t0:. Q Ca Ra Q. Ma I M. = S'LlX. a A ) C _Al!x. aSa2' ?\Uo Ra=~; was build for suction •1ire. Also the mufflers models were done -oasing on the lum1'ed ?arameter model. And only the resistive type of muffler needs some explanation. The flow through the oriffice was modeled in the way similar to that used for valves. It was assumed that kinetic energy of flow in the oriffice is completely dissipated inside the chamber. Acoustic model Far field acoustic pressure level was found by exnansion of the flow rate at the inlet of suction pipe into a Fourier series and by the use of model presented in [8) • !J. = J'o CJn On On I.TTR NUHJi~RICAJ, SOlUTION The set of differential equations has been solved by the Runge-I:utta ITJ.ethod. For TJrOgraming, the symulating language HH1IC and the standard Fortran have been used. The use of Fortran significantly reduces the time of computation in comparison with MIHIC.