Sources and Remedies of High Freq Piping Vibration n Noise.pdf

7/27/2019 Sources and Remedies of High Freq Piping Vibration n Noise.pdf

1/24

SOURCES AND REMEDIES OF HIGH-FREQU ENCY PIPING VIBRATION AND NOISEby

StephenM. PriceSenior Project Engineer

andDonald R. Smith

Senior Staff EngineerEngineering Dynamics Inc.San Antonio, Texas

Stephen M. Price is a Senior ProjectEngineer for Engineering Dynamics Inc.,in San Antonio, Texas. For the past 15years, he has been actively involved insolving a variety of problems in systemsthat have experienced failures due todynamic phenomenon. He has had field andanalytical experience solving problemswith reciprocating and rotating machinery,along with structural and acousticaloroblems. Additionallv. MI: Price has beeninvolved in development of multiple channel data acquisition andmonitoring hardware and sojiware for critical environments. Hehas presented and published several papers in the areas of signalprocessing, finite element analysis, fatigue analysis, acoustics, andreciprocating pumps.MI: Price has a B.S. degree (Mechanical Engineering) fromMississippi State University and an M.S. degree (MechanicalEngineering) from Purdue University. He is a registeredProfessional Engineer in the State of Texas and a member ofASME.

Donald R. Smith is a Senior StaffEngineer at Engineering Dynamics Inc.(EDI), in San Antonio, Texas. For the past30 years, he has been active in the fieldengineering services, specializing in theanalysis of vibration, pulsation, and noiseproblems with rotating and reciprocatingequipment. He has authored and presentedseveral technical papers. Prior to joiningEDI, he worked at Southwest ResearchInstitute for 15 years as a Senior ResearchScientist, where he was also involved in troubleshooting andfailure analysis of piping and machinery.MI: Smith received his B.S. degree (Physics, 1969) rom TrinityUniversity. He is a member of ASME and the Vibration Institute.

ABSTRACTIn large diameter piping, high-frequency energy can produceexcessive noise and vibration, and failures of thermowells, instru-mentation, and attached small-bore piping. In severe cases, thepipe itself can fracture. Perhaps more precisely called "high wavenumber" problems, these problems most often manifest themselvesin centrifugal compressors, screw compressors, heat exchangers,and silencers.Two high-frequency energy generation mechanismspredominate in most industrial processes; flow induced (vortex

shedding) and pulsation at multiples of running speed (blade-passin centrifugal compressors and pocket-passing frequency in screwcompressors). Once this energy is generated, amplification mayoccur from acoustical and/or structural resonances, resulting inhigh amplitude vibration and noise.To resolve these problems successfully, an understanding of theunderlying physics of two- and three-dimensional acoustics isnecessary. With these principles in mind, modifications to thepiping system can be considered for the particular application. Thethree-dimensional wave equation is used to analyze thepropagation of the high-order (cross-wall) acoustical modes in theduct or pipe. These cross-wall modes can be diametrical (m)modes, annular (n) modes, or combined (m, n) modes. Byreformulating the resulting differential equations into polarcoordinates and applying the appropriate boundary conditions, anequation for the "cut-on" frequencies, f ( for cross-wall modescan be developed that incorporates zeroes of the first order Besselfunction, /? the speed of sound, and pipe diameter. Severalreferences provide lists of the zeroes of the Bessel function;however, most of these references only provide solutions up to m,n = 6. Field tests have identified cross-wall modes up to m = 30.Therefore, a table is provided for zeros of /?im,,, f or m = 0 t o 32 andn = O t o 8 .This paper discusses the excitation and amplificationmechanisms relevant to high-frequency energy generation inpiping systems. Mechanisms that allow efficient coupling of thisenergy with the surroundings (either structural or acoustical) arediscussed. Data from various systems are presented, as well asdesign modifications that have been shown to be effective atreducing the high-frequency energy.EXCITATION MECHANISMS

The main sources of the high-frequency pulsation are usuallyvortex shedding and/or blade-pass excitation. The high-frequencypulsation typically occurs at frequencies above approximately 500Hz , although for systems with large diameter vessels, suchproblems can exist at lower frequencies. Vortex shedding is aphenomenon that occurs due to flow over an obstruction. Inturbomachinery, the primary excitation is due to the interaction ofa blade or vane with the internal passages. Pulsation occurs at theblade-pass frequency (number of blades multiplied by theoperating speed) and multiples of blade-pass frequency. In screwcompressors, the intermeshing of the helical lobes generatepulsation at the pocket-passing frequency (the number of lobes onthe male rotor multiplied by the running speed) and multiples ofthe pocket-passing frequency.Vortex Shedding

Vortex shedding occurs due to flow over an obstruction in theflow path. At low flow velocities (low Reynolds number), a fluidparticle can flow completely around the obstruction. However at


2/24

190 PROCEEDINGS O F THE 28TH TURBOMACHINERY SYMPOSIUM

higher Reynolds numbers, the fluid boundary layer at the trailingedge of the obstruction will separate, causing shear layers that trailthe obstruction. The se shear layers tend to roll into vortices that arealternately shed from either side of the obstruction (Figure I). Thealternate shedding occurs at regular intervals and thereforeproduces an oscillating pressure field (pulsation).

Regime ofUnseparated F io r

5m156Rc


3/24

SOURCES AND REMEDIES O F HIGH-FREQUEN CY PIPING VIBRATION AND NOISE 191

where:F(,,,) = Pulsation frequency, HzP(,,,) = Zeros of the first order Bessel func tionc = Speed of sound, ft/secd = Pipe diameter, ft

The first order Bessel function is a continuous function that isroughly sinusoidal in shape. Values for the zero crossings of thisfunction are shown in Figure 2. The modes are designated byordered pair (m, n) where the integer "m" determines the numb erof radial (diametrical) nodal lines, and the integer "n" determinesthe number of the annular (circumferential) nodal circles.

Figure 2. Zeros of First Order Bessel Function.Figure 3 shows a representation of the pressure mode shapes fordiametrical modes m = 1 to 4, n = 0, and for annular modes m = 0,n = 1 to 4. The mixed modes (where both m and n are nonzero) arenot shown because of their complexity. In all cases, there is apressure maximum at the wall of the pipe and pressure minimaalong the nodal lines shown in the diagrams.Previous laboratory testing has determined that the diametricalmodes ("m" modes) are easily excited when the excitation sourceis near the pipe wall. Similarly, the circumferential modes ("n"modes) are easily excited when the excitation source is near the

center of the piping (along the axis of the pipe).The cross-wall acoustic natural frequencies are also referred toas cut-on, or cutoff frequencies. Below the first cross-wall naturalfrequency, m = 1, n = 0 (cutoff frequency), only plane wavepropagation is possible. However, at frequencies above the firstcross-wall natural frequency (cut-on frequency), the two-dimensional cross-wall modes can propagate down the piping.Furthermore, above the cut-on frequency, these modes have littleor no damping and can travel long distances unattenuated.Shell Wall Natural Frequencies

At low frequencies, pipe vibration occurs laterally along itslength (like a beam). At higher frequencies, the pipe shell wall

Plane Wavem=O,n=O

I Diametrical (m ) Modes

I Annular (n) Modes

Figure 3. Cross-1YullAmusticat 1bfode Shupes for ( n l = 1 to 4, 11 =0)and (m= 0, n = I to 4).begins to vibrate radially across its cross-section. This could beviewed as a "breathing" mo de if all points were vibrating in-phase.At progressively higher frequencies, adjacent portions of the pipeshell may vibrate out-of-phase, producing a sine-wave around thecircumference. Examples of these mode shapes are shown inFigure 4.

Cross-Secfivn ef Pipe

Circumferential Node Number - i

3-Dimensional View of Pipe

Shell Wall Vibration

Figure 4. Piping Shell Wall Vibration Mode Shapes.A number of theories can be used to calculate the shell wallnatural frequencies for shell wall vibration. For infinitely longpipe, a closed form solution has been developed, as provided inBlevins (1979). Using Equation (4) and Equation (5), it can beshown that the natural frequencies are primarily controlled by thepiping diameter and wall thickness. The frequencies are alsoslightly influenced by the internal pressure in the piping. Toaccount for internal pressure, the effective pipe wall thicknessshould increase by 5 percent to 10 percent.


4/24

192 PROCEEDINGS OF THE 28TH TURBOMACHINERY SYMPOSIUM

Experience has shown that the accuracy of these calculations issufficient for a pipe span approximately one diameter away fromsignificant discontinuities (e.g., flanges, tees, elbows, etc.).

where:Shell wall natural frequency, HzFrequency factor, dimensionlessMean radius of pipe wall, inchPoisson's ratioMass density of pipe material, lb/g-in3Modulus of elasticity of pipe wall, psiPipe wall thickness, inchMod e number, 2, 3, 4...Gravitational constant, inchlsec2

Radiation EficiencyEach of these shell w all natural frequencies (f,, i = 2, 3 , 4 ..) hasa particular mode shape defined by the number of diametrical node

lines (Figure 4). For a particular mode shape to be an efficientradiator of noise (or conversely, to be easily excited by pulsation),the bending w avelength of the shell wall vibration must be equal toor greater than the wavelength of the acoustical wave in air. Thischaracteristic is termed radiation efficiency.The point at which the two wavelengths match is called thecoincidence frequency. If the shell wall natural frequency is lowerthan the coincidence frequency, the mode will not be an efficientradiator of noise. At shell wall frequencies above the coincidencefrequency, the mode can radiate sound effectively.The coincidence frequency is defined by Equation (6).

where cai, is the speed of sound in air and kb is the bendingwavelength of the shell wall mode. For pipes, the shell wall modeshave a bending wavelength defined by:

where d is the diameter of the pipe and N is the number ofdiametrical node lines.The evaluation of whether a mode is an efficient radiator ofsound is made simpler by computing the ratio of its naturalfrequency, fi, to the coincidence frequency, f,. Modes that haveratios greater than or approximately equa l to 1.0will radiate sound.Values less than approximately 0.7 denote modes that are poorradiators of sound.The radiation efficiency concept also provides an indication ofthe degree of coupling between pulsation in the gas inside the pipe,to shell wall vibration.

where cgas s the speed of sound in the g as. Values greater than onewill denote modes that can transfer energy to the pipe wall easily,while values less than one denote modes that are less efficient attransferring their energy.Equation (4) through Equation (8) can easily be set up into aspreadsheet for computation. One such computation for a heavy

hydrocarbon flowing in a 36 inch steel pipe is shown in the table inthe section, Calculations.REDUCING HIGH-FREQUENCYPULSATION AND VIBRATION

There are several techniques available to achieve reductions ofhigh-frequency energy. The energy generation mechanisms can bereduced or eliminate d, and/or the amplification mechanisms can beremoved. Additionally, it is sometimes possible to simply controlthe vibration by using damping, or to reduce the pulsation energyby using acoustical absorption m aterial inside the pipe.Reduction of Vortex Shedding E nergy

The first step to reducing vortex shedding energy (pulsation) isto identify the particular components causing shedding. Since thefrequency of the excitation is known, Equation ( I ) can be used todetermine the approximate size of the obstruction. With thisinformation, it should be possible to determine if an obstruction (orgap) exists in the piping that conforms to the required size thatwould result in pulsation in the frequency range of interest.Sometimes, these structures can be modified to discourageshedding, or to change the shedding frequencies. Figure 5 showsseveral ways to reduce vortex shedding. Adding strakes or airfoilshapes are the most commonly used methods.

Flag

HelicalStrake

Perforated AxialShroud Slots

Plates

1:ribbonairAerodynamic

Fairing

PivotingFairing

Figure 5. Vortex Suppression Devices.If prevention of vortex shedding is not practical, the vortexshedding frequencies can be shifted away from acoustical ormechanical na tural frequencies by making the effective diam eter ofthe obstruction larger or smaller. Alternatively, the flow path couldbe altered to reduce or increase velocities across the obstruction.

Reduction of Blade-Pass PulsationSince blade-pass pulsation is caused by interactions within thecompressor, it is usually difficult to reduce blade-pass excitationwithout altering compressor performance. The compressormanufacturer should be consulted to assess possible methods to


5/24

SOURCES AND REMEDIES OF HIGH-FREQUENCY PIPING VIBRATION AND NOISE 193

reduce blade-pass energy. Increasing the internal clearancesbetween the impeller vanes and diffuser blades or cutwater ("BGap," Figure 6) has been shown to be effective in some cases.Modification of the flow-path has also been successful.Gap B

I

X-Brace in Pipe

Figure 6. Cutwater ( "B" Gap) Clearance.If it is not possible to reduce blade-pass energy amplitudes,changing the number of impeller blades can shift the frequency.Adding blades generally produces lower amplitude blade-passenergy.

Prevention of Cross-Wall Acoustical ModesChanging the internal dimensions of the pipe can shift the natural Tube Bundle in Pipe- &frequencies of cross-wall modes. However, since it is not usually Q~~~, internal pipe ~ ~ d i f i ~ ~ ~ i ~ ~o ~i~~~~~~~~ ~~~~~~i~practical to provide a significant diameter change, "flow-splitters" ~ ~ ~ ~ ~ - ~ ~ l l~d~ ~ ~ ~ ~ ~ ~can be added inside the pipe to change the internal geo metry, whichsignificantly shifts the cross-wall natural frequencies.Depending upon the frequencies and number of modes involved,an internal X-brace (Figure 7) may be sufficient. In some cases, amore complicated modification called a "tube bundle" may berequired (Figure 7) . A potential problem with such modifications isthat the cross-wall modes can re-form in the piping upstream or Idownstream of the tube bundle or X-brace. It is difficult todetermine the length of the tube bundle or X-brace required toprevent reformation of the cross-wall modes. Laboratory tests Perforated ~ ~ t a lindicated that for tube bundles that are several pipe diameters inlength, the cross-wall modes tend not to re-fir&, but for flowsplitters, the modes may re-form downstream of the splitter.Another method to attenuate cross-wall modes is withacoustically absorptive material (Figure 8) . Since all the cross-wallmodes have a maximum pressure boundary at the edge of the pipe,applying absorption material at this boundary (which enforces apressure-release condition) will discourage cross-wall modeformation. One drawback to this approach is that absorptionmaterial inside the pipe is not acceptable under manycircumstances (e.g., when liquids are present, high temperatures,etc.). Also, it can be difficult to contain the absorption material.Often, the absorption material is restrained with wire mesh orperforated metal. Figure 8. Absorption Muterial Applied to Silencer Internuls.


6/24

194 PROCEEDINGS OF THE 28TH TURBOMACH INERY SYMPOSIUMA final method to dissipate high-frequency energy is through useof reactive silencers. These silencers make use of various volumeand choke tube arrangements to attenuate the energy before it canpropagate into the piping. However, care must be taken to designthe silencer so that cross-wall modes do not form in the silenceritself. A poorly designed "silencer" can actually become anamplifier, potentially creating destructive levels of pulsation andvibration energy. Additionally, the silencer intemals can be proneto vibration induced failures.

Reduction of Shell Wall VibrationIn some cases, it is not practical to reduce the pulsation energysources or attenuate the acoustical cross-wall modes. Increasingthe pipe wall thickness can reduce the shell vibration levels.However, it is generally imp ractical to replace existing piping withthicker wall piping. In these cases, adding damping to the pipingand surrounding structures can be effective at reducing the shellvibration amplitudes. There are two basic types of dampingtreatment for structural damping: extensional (unconstrained orfree-layer damping) and shear (constrained-layer damping).Extensional damping treatment (Figure 9) consists of a singlepiece of visco-elastic damping material bonded directly to thestructure surface. As the surface deforms, the damping material issubjected to tension-compression deformation. While this type ofdamping is the easiest to design and apply, extensional damping isusually not as effective as constrained-layer dam ping.

P i p ShdlWall Damping-Extcmional

Dampinq Moterial

Pipe Shell

!amping MaterialExtends' os PlateDeforms

Domping MaterialSubjected to SheorBetween Plotes

Figure 9. Plate and Shell Damping Treatments-Extensionul andConstrained Layer:Constrained-layer damping (Figure 9) is similar to extensionaldamping, except that the damping material is constrained by anouter metal layer. As the pipe wall is deformed, the outer metal layerconstrains the damping material and forces it to deform in shear. Thethickness of the outer metal layer and the thickness and materialproperties of the damping material control the effectiveness of theconstrained-layer damping. Normally, the thickness of the outer

metal layer should be approximately 25 percent to 50 percent of thepipe wall thickness. Theoretically, the maximum damping isobtained with very thin dam ping material, since the shear is higher.For flat plates, a very thin damping layer can be used. How ever,for large diameter piping, it can be impractical to fabricate aconstraining layer with tolerances less than one-eighth inch.Thicker damping material can be used (good success has been hadusing sheets of silicone rubber or neoprene), but the effectivedamping will be reduced from the ideal case.Acoustical Lagging

For situations where vibration is not a problem, but noise levelsare excessive, it is sometimes desirable to add aco ustical "lagging"to the piping to absorb the noise, rather than to try to reduce thesource or amplification mechanisms. The choice of laggingmaterial is important. Absorption material that is appropriate forthe frequency of noise must be selected.A typical lagging ap plication is shown in Figure 10 (adaptedfrom Pelton, 1993). Note that the absorption m aterial is surroundedby a mass layer (usually lead or vinyl). Lower frequencyapplications will require thicker mass layers. In some cases, severalsandwiches of absorption material, mass layer, must be used toachieve acceptable noise attenuation.

Abso r p t i v eLay e rFigure 10.Acoustical Lagging Applied to the Exterior of a Pipe.TESTING

To diagnose these types of problems, field testing is usuallyrequired. The test program philosophy is to try to identify energygeneration mechanisms (such as blade-pass pulsation or vortexshedding) and determine transmission paths and/or amplificationmechanisms. Eliminating energy generation, short-circuiting thetransmission path, an d or changing the amplification mechanismscan reduce vibration and noise. The following sections describe atypical instrumentation and test procedure.Instrumentation

A variety of instrumentation is required, including high-frequency accelerometers, insertion pulsation probes,microphones, strain gauges, and impact hammers.

High-Frequency AccelerometersMeasurement of piping shell wall vibration requires low-mass,high-frequency accelerometers. The accelerometers should beattached to pads that are either glued or welded onto the structure.In some cases, acceleration levels in excess of 500 g's zero-peakhave been measured. Such extreme amplitudes can damageaccelerometers, or cause the electrical connections and wires tofail. Therefore, specially constructed wiring and accelerometersmay be required.


7/24

SOURCES AND REM EDIES OF HIGH-FREQUENCY PIPING VIBRATION AND NOISE

Insertion ProbesPulsation data should be acquired in the piping, especially nearlocations of high vibration and noise. Pulsation data are most oftenmeasured at "stubs" (i.e., through a valve or other connecti n).Such an installation will not provide an accurate indication of

pulsation data at intervals and frequencies given by:1ulsation in the piping due to quarter-wave resonances. Fo asimple stub, these resonances will produce false peaks in the

where:f,, = Stub frequencies, H zn = 1, 3, 5...c = Acoustic velocity, ftlsecL = Length of stub, ft

In the field, such simple stub connections are rarelyencountered. As shown in Figures 11 and 12, the stub pipingtypically consists of a number of diameter changes. These diameterchanges cause impedance discontinuities, which result in multiplepulsation responses, rather than a single frequency.Pressure Transducer\

9.2"TotalEffectiveLengthTo PressureTronsducer

Swogelok Fittings

Swagelok Adopter

Anderson GreenwoodNeedle Valve

I - " x 1/2" ReducerFigure 11. Pressure Transducer Stub Connection Through NeedleValve.

Additionally, as described in K insler, et al. (1982), the amplitudeof pulsation can be higher, even at frequencies that are away fromstub resonances, d ue to the effects of pressure reflection at a closedend. Therefore, to avoid these potential problems, it is desirable to po-sition the pulsation transducer just inside the shell wall of the piping.In many cases, it is also desirable to install and remove thepressure transducer without depressurizing the system. Onecompany has developed an "insertion probe" that allows thepressure transducer to be inserted through a valve, and positionedinside the pipe wall (Figure 13). This arrangement has beeninstalled successfully into operating system s with pressures as highas 2500 psi.

Pressure Tronsducer \L

7.2"TotalEf fect iveLengthTo PressureTransducerI 1/2' Nipple

1/4' Swagelok Fittings

1/2" Swogeiok Adopter

-- -- I /2- Bolon Boll Volve

Flow IPipe WollTest Point Locationwith Pressure Transducer

Figure 12 . Press-ure Transducer Stub Connection Through Ballvalve.Wire to Transduce~

----3/8" Stainless Tubingr uffalo ConoxFlow

Test Point Locationwith Insertion Probe andPressure Transducer

Connector

Pressure Tronsducer'Figure 13. Insertion Probe to Eliminate Stub Connection.MicrophonesSound data are used to correlate ambient noise with pipingvibration and pulsation. Several microphones positioned a few feetfrom the piping shell will produce data that allow suchcomparisons to be made. It has been found that in most cases,inexpensive self-contained sound-level meters provide adequatedata for these purposes.


8/24


Strain GaugesStrain gauges are used to determine strain levels in the pipingand at nozzle connections. Weldable gauges provide the largestrange of operating conditions, and are easy to install. However, asmall spot welder is required to attach these gauges. Using thesegauges, comparisons of measured strain frequencies andamplitudes to the vibration, pulsation, and noise can be made toprovide insight into the causes of failures.Impact HammersNatural frequency testing of the piping shell walls can beaccomplished easily with impact hammers. The hardness of thehammer tip should be selected to produce an impact having energyat the frequencies of interest. For example, if excitation at high-frequencies (>500 Hz) is desired, a steel tip should be used. Forlower frequencies, plastic or rubber tips could be used. The m ass ofthe hammer should also be selected so that the hammer reboundsquickly when the pipe is impacted. For most piping, the shell wallmodes can be excited using a 1 lb hammer with a steel tip.

Test ProcedureA typical test procedure involves natural frequency testing of thesystem and operational tests.

Structural Natural Frequency TestsStructural shell wall natural frequencies are amplificationmechanisms that can result in high vibration when excited. Thesenatural frequencies can be identified using an impact hammer andmodal analysis software. Usually, it is necessary to shut down thesystem to obtain good quality natural frequency data. In addition todetermining the natural frequency, the mode shape data are usefulfor determining the ability of a structural natural frequency tocouple with pulsation or generate noise. To measure the modeshapes of the higher modes, data are acquired at several evenlyspaced locations around the circumference of the pipe. Theminimum number of test points required is twice the mode number(the number of diametrical node lines) for the highest frequency ofinterest. The mode shape for a particular natural frequency can beidentified using these data and commercially available animatedmode-shape software.For example, consider a pipe with a 24 inch diameter and a 0.5inch thick wall. Table 1 illustrates the number of circumferentialtest points required to define the mo de sha pe. As show n, if thehighest frequency of interest is 1500 Hz, at least 14 points aroundthe circumference of the pipe would be required to define modeshapes for shell wall natural frequencies up to 1500 Hz.

Table 1. Minimum Number of Points Required to Dejne a ShellWall Natural Frequency Mode Shape.I Mode # I Natural Frequency I # Points Required I

Operational TestsAfter installing accelerometers, pressure transducers, straingauges, and microphones, the system should be tested while

operating at a variety of conditions. Note that in addition to theinstalled instrumentation, system parameters (e.g., pressures,temperatures, speed, etc.) should be logged. A tape recorder orother data acquisition system is necessary to gather all datasimultaneously.If the machine speed can be varied, a test should be performedwhere the machine's speed is varied slowly from minimum tomaximum speed. Another test could be done by varying theloading. Additional tests should be performed with abnormalconditions (recycle operation, near-surge conditions, etc.).Evaluation of DataData can be plotted in many formats. A spectrum analyzer, adigital acquisition system, and software for generating high qualityplots will simplify the process and enhance evaluation of data.Using the pulsation data, the energy generation mechanisms canbe identified. For example, if pulsation is measured only at blade-pass frequency, the machine itself is generating the pulsation.Nonsynchronous frequencies could be related to vortex sheddingor other phenomena.Pulsation data obtained during a speed sweep can be comparedwith calculations of cross-wall natural frequencies to determine ifpulsation energy is being amplified by cross-wall acoustical modesin the piping. Similarly, shell wall vibration data obtained during aspeed sweep combined with data from natural frequency tests willhelp determine if structural responses could be amplifying energy.

Strain and microphone data should be compared with thepulsation and vibration data to determine which of the measuredenergies are contributing to the fatigue failures or high noise kve ls.Once these comparisons are made, it should be possible to developsolutions to reduce the vibration, stress, and noise using theconcepts discussed in the previous section, REDUCING HIGH-FREQUENCY PULSATION AND VIBRATION.Acceptability of Shell Wall Vibration

For shell wall piping vibration, acceptability criteria must relatevibration amplitudes to stress. Such a relationship has beendeveloped by Mikasinovic (1989) in which vibration velocitymeasured on a cylindrical shell wall is related to dynamic strain:

where:V = Vibration velocity, inlsec, zero-peakc = Curve fit constant = 742,124 ids ecE = Dynamic strain, microstrain (id in x l o @ )

This relationshio assumes that the vibration measurements arezero-peak measurements and that several resonant modes areinvolved (so that the peak vibration velocity is approximately thesame around the circumference and along the axial length of thepiping between the constraints). Mikasinovic tested pipe diametersfrom 6 inch to 30 inch, wall thicknesses from 0.25 inch to 0.75inch, and lengths from 12 ft to 42 ft with satisfactory results. Goodcorrelation has been obtained between measured and calculatedstrain values for 30 inch pipe with wall thicknesses of 0.375 inchto 0.8 inch.Using this formula, it would be possible to relate the vibrationvelocity to the fatigue endurance limit. Assuming an allowableendurance limit stress for carbon steel of 13,0 00 psi zero-peak, asafety factor of 1.3, and a stress concentration factor of five (at theheat-affected zone of a weld), the allowable stress would be 2000psi zero-peak. For carbon steel, the elastic modulus is 30,000,000psi, which results in an allowable strain of approximately 67microstrain zero-peak (133 microstrain peak-to-peak). Using thisvalue for the acceptable strain in Equation (l o) , the allowablevelocity is 7.9 inlsec zero-peak.


9/24

SOURCES AND REMEDIES OF HIGH-FREQUENCY PIPING VIBRATION AND NOISE 197If the stress concentration factor is less than the maximum, theallowable vibration velocity would be higher by the ratio of theactual stress concentration factor to five. For example, a butt weldhas a stress concentration factor of approximately two; therefore,the allowable velocity would be increased to 19.8 psi zero-peak.This allowable vibration is considerably higher than the allowablelevels for lateral piping vibration and machinery vibration (e.g., ata bearing housing).The vibrational velocity of the shell wall can also be related tothe sound pressure level (C or linear weighting). Field experiencewith strain gauges installed on piping with high-frequency,

broadband, vibrations has shown that the sound pressure level(SPL) measured approximately 1 inch away from the pipe wall isproportional to the dynamic strain. Although the relationshipbetween dynamic strain and SPL amplitude is not exact, the overalllevels presented below have been used to estimate the severity ofshell wall vibrations and as a screening method to help determinewhere strain gauges should be installed on a piping system.When the SPL is measured with a sound level meter (SLM)using C weighting approximately 1 inch from the vibrating pipewall, the following criteria have been found to be applicable:130 dB is equivalent to approximately 100 microstrain peak-to-

peak136 dB is equivalent to approximately 200 microstrain peak-to-peakIn addition to the criteria outlined above, it has been shown byfield experie nce that allowable strain levels (E) for carbon steel canbe specified by:

E < 100 microstrain p-p Safe100 microstrain p-p 200 microstrain p-p Excessive

Therefore, sound pressure levels greater than 136 dB measured atapproximately 1 inch from the pipe wall would be sufficient toexpect fatigue failures.CASE HISTORIES

Over the past several years, the authors have had the op portunityto evaluate a number of situations where high-frequency vibrationwas causing high noise amplitudes and fatigue failures. Suchproblems are seemingly becoming more prevalent, as flow ratesthrough equipment and piping rise. The problems occurred withscrew compressors, centrifugal compressors, and roots-typeblowers. Excitation mechanisms included blade-passing frequency,pocket-passing frequency, and vortex shedding.In some cases, the noise problems were treated with acousticallagging. Other cases were solved by removing the energygeneration mechanisms (in the cases of vortex shedding),removing the amplification mechanisms (in cases of blade-passexcitation), installing specially designed silencers, or acombination of these methods. The following sections will brieflypresent a few of these cases.Laboratory Testing-Pipe with Zero FlowA simple laboratory test was conducted in an effort to gain agreater understanding of the phenomena involved in generationand propagation of high-order acoustical modes. The test planincluded investigations of high-order mode generation andmethods to "turn o ff' the modes once they were generated.Excitation of Cross- Wall ModesA 12 inch diameter ( 1 1.875 inch ID), 20 ft long plastic pipe wasconstructed in the laboratory. A white noise or a swept-sine signalwas applied to a speaker that was mo unted radially at the pipe wall.Several locations were selected for placement of microphones tomeasure responses in the pipe (Figure 14). Anechoic termina tionswere provided at each end of the pipe using acoustical foam inserts.

Figure 14. Laboratory Test-Pipe with Zero Flow (Showi ngInsertion of Tube Bundle).Cross-wall mode frequencies were calculated for this pipe form

= 1 to 6 and n = 0 to 1. Responses were measured from white noiseexcitation applied radially at the pipe wall. Agreement betweencalculated and measured values was good, as shown in Table 2.Note, howev er, that for this excitation arrangement, only the purelydiametrical modes (m = 0 to 6, n = 0) were excited much strongerthan the circumferential modes.Table 2. PredictedMea sured Cross-Wall Modes for Air in 12 InchPipe-Radial Excitation.

I I Frequency (Hz) I

The test was repeated with excitation applied at the pipecenterline (axially). In this configuration, it was much moredifficult to excite cross-wall modes. The response data were not asclear (more broadband). However, as shown in Table 3, responsesat 1 400 Hz, 1950 Hz, and 2450 Hz were measured thatcorresponded to circumferential modes for the (0, I), (1, I), and (2,1) modes, respectively.These tests indicated that the diametrical modes seem to beeasier to excite than the circumferential modes. Furthermore, theexcitation must couple well with the mode shape to "turn on"cross-wall modes.

0

Elimination of Cross-Wall ModesExperiments were conducted to determine if the cross-wallmode propagation in the piping could be reduced or eliminated.Several different devices including a perforated disk, an "X-divider," and a bundle of 2 inch tubes (Figure 7) were evaluated.The devices were inserted into the pipe, downstream of the second

Calc0

Meas Calc1387

Meas1388


10/24

198 PROCEEDINGS OF TH E 28TH TURBOMACHINERY SYMPOSIUM

Table 3. Predicte dM easured Cross-Wall Modes for Air in 12 InchPipe-Axial Excitation.Frequency (Hz)

microphone. Excitation was provided at the pipe wall (radialdirection) near the front end of the pipe and the noise data weremeasured at several locations in the piping.Response data with the perforated plate showed that this elementprovided no attenuation of cross-wall modes.Although the X-divider turned off the cross-wall modes in thesection with the X-divider, the modes tended to re-form afterleaving the X-divider section. This test indicated that adding X-dividers to the piping could eliminate the cross-wall modes;however, the plates would have to be installed over the full lengthof the piping, which would probably not be practical.The tube bundles that were evaluated were similar to flowstraighteners used to improve the accuracy of flow meters. Thetube bundles consisted of 18, 2 inch diameter pipes. One bundlewas 2 ft long (two pipe diameters) and a second bundle was 4 ftlong (four pipe diameters). The tests with the tube bundlesindicated that both bundles were effective in preventing theformation of the cross-wall modes in the section of piping with thetube bundles. However, the 4 ft length bundle was more effectivein preventing reformation and propagation of cross-wall modesdownstream of the bundle. This suggested that the commerciallyavailable flow straighteners may not be as effective in preventingthe propagation of cross-wall modes, du e to the short length of thetubes (typically less than one pipe diameter). Note that cross-wallmodes could also occur in the 2 inch diameter tubes, but thefrequencies would be much higher, and would not be as likely tocause shell wall vibration of the 12 inch piping.It should be remembered that du e to the design of the test facility,these tests were performed with zero mean flow. If there had be enflow, some attenuation wo uld have likely been observed through theperforated plates due to the inherent pressure drop. Additionally, thegreater attenuation due to pressure drop may have prohibitedreformation of cross-wall modes downstream of the X-divider.Limited information was obtained from these tests. Therefore, itis recommended that additional research should be conducted withmore rigorous testing to evaluate the potential for "turning on"cross-wall modes, and to evaluate modifications that could reducecross-wall mode formation.

n = 0I

Compressor Station-Suction Side

n = lCalc I Meas I Calc I Meas

A series of new pipeline compressor stations was experiencinghigh amplitude, high-frequency noise and vibration of the suctionpiping associated with the compressors. Failures of instruments onthe suction piping near the compressors resulted. Additionally, thenoise levels in the buildings were excessive for personnel (> 125dB), requiring both ear plugs and muffs to be worn.

Each compressor station utilized a centrifugal compressordriven by a gas turbine. The comp ressors had a single, overhungimpeller . Because of f low requirements, som e stations utilizeda 17-blade impeller while others used a 14-blade impeller. Thecompressor operating speed range was 3120 rpm to 5040 rp mwith a rated speed of 4800 rpm. Design operating pressureswere 1050 psi suction and 1460 psi discharge. Suction anddischarge piping for the compressors were 36 inch diameter(0.75 wt).It was hypothesized that the noise and vibration were the resultof amplification of energy by high-order acoustical modes in thepiping an do r piping shell wall modes. The excitation source wasnot known; however, blade-pass energy or vortex shedding wereconsidered to be probable mechanisms.

CalculationsThe calculations discussed in previous sections, Acoustic Cross-Wall Natural Frequencies, and Shell Wall Natural Frequencies,were performed to see if there were potential coincidences ofacoustical natural frequencies and shell wall natural frequencieswith excitation mechanisms.Acoustical cross-wall mode natu ral frequencies-Using theequations in the section Acoustic Cross- Wall Natura l Frequencies,the high-order acoustical mod es for the 36 inch (34.5 inch ID) pipefor m = I to 32 and n = 0 to 7 were calculated. The results forseveral of the modes are shown in the table in the section, Analysisof Measured and Calculated Data.Piping shell wall mode natural frequencies-Using theequations in the section, Shell Wall Natural Frequencies, theshell wall mod es of the piping were calculated and are shown inTable 4. Also shown are the radiation efficiency parameters,calculated using the equations in the section, RadiationEficiency .

Tuble 4. Shell Wall Modes and Radiation Eficiency.


11/24

SOURCES AND REMEDIES OF HIGH-FREQ UENCY PIPING VIBRATION AND NOISE 199

It is interesting to note that for this piping geometry, most ofthe potential shell wall modes will radiate sound efficiently.Furthermore, some of the shell wall modes have nodal patternsthat are similar to the predicted acoustical mod e at that frequency.For example, the seven-diameter shell wall mode was calculatedto be at 1118 H z. The seven-diameter acoustical naturalfrequency was predicted to be 1173 Hz (refer to Table 7). Thesecalculations indicate that the acoustical and structural modes ofthe piping could act together to produce even largeramplifications of energy.0 Vortex shedding-Flow velocities calculated from measuredplant operational ;onditions indicated that obstructions of 0.1 inchto 0.5 inch in diam eter could produce vortex shedding excitation inthe range of the expected acoustical natural frequencies.Obstructions in this size range were thought to exist at the strainer(stiffener rings), and at various flange gaps an d stub connections inthe vicinity of the strainer (Figure 15).0 Conclusions from prelim inary calculations-The calculationsshowed that for the piping geometry and flows, either blade-passenergy or vortex shedding energy could excite acoustical naturalfrequencies and shell wall natural frequencies of the piping. Thesecalculations therefore tended to verify the hypotheses for the highamplitude noise and vibration.

Figure 15 . Origin al Straine r Installatio n Upstream of Pipelin eCompressor:Field Data-Station 7Small, high-frequency accelerometers were used to acquire shellwall vibration data. Each accelerometer was attached to a pad,which was glued to the pipe using an adhesive.Noise data were measured using a soun d-level meter (SLM). Forthe frequency range of interest, these devices are preferred overmicrophones becau se of their durability and ease of use. The SLM swere mounted on a tripod and aimed at the pipe. Prior to use, eachSLM was calibrated using a 94 dB, 100 0 Hz sound source.Pulsation data were acquired using piezoelectric transducers,installed in insertion probes (refer to section Insertion Probes).

Shaft vibration data were recorded from the permanentlyinstalled proximity probes. These instruments record radial or axialshaft vibration. A once-per-revolution (key-phase) pulse was alsoused to monitor compressor speed.The installed station transducers and process computer providedadditional data (suction pressure, discharge pressure, temperatures,gas composition, etc.).0 Operatin g data-The comp ressor was operated over theoperating speed range achievable on the day of the test. A11 datawere tape recorded during the test for later analysis.

A trending program was set up to monitor process data duringthe tests. The compressor speed, station flow, and suction anddischarge pressures during the test are shown in Figure 16. Asshown, the speed ramp of 312 0 rpm to 4400 rpm took place duringan interval of approximately 20 minutes. Suction pressures rangedfrom 1140 psi to 1045 psi, while discharge pressures went from1298 psi to 1392 psi. Station flow was varied from 1760 MMscfdto 2500 MM scfd.

Figure 16 . Station 7 Operating Conditions During TestingUsing station flow, temperature, gas properties, and pressuredata, the average flow velocities in the suction pipe during the testwere computed. As shown in Figure 17, the average flow velocityin the suction pipe ranged from approximately 32 ftlsec to nearly

50 ftlsec. For the period of the testing, the flow velocity variedessentially linearly with compressor speed.

Figure 17 . Average Flow Velocity Versus Compressor OperatingSpeed.Waterfall plots were made for the recorded data. As shown inFigure 18, the suction line pulsation data measured do wnstream ofthe strainer show discrete frequency peaks that line up vertically.These peaks were especially apparent in the data acquireddownstream of the strainer.Excitation at blade-pass frequency (17x running speed) and 2xblade-pass frequency (34x running speed) can be seen as peaks


12/24


Frequency (Hz)

Figure 18. Suction Line Pulsation Measured Downstream ofStrainer-Station 7.that lie along the 17x and 34x order lines. However, these peakswere typically much lower amplitude than the vertically alignedones. This behavior suggests the presence of acoustical resonances(cross-wall modes) at the frequencies of the vertically alignedpeaks.Another behavior that can be observed from the data is that ascompresso r speed (and as shown by Figure 17, flow velocity)increases, peaks at lower frequencies die out while new peaks athigher frequencies appea r. The se data show that acoust icalresonances (vertically aligned peaks) we re being excited by vortexshedding, since the vortex shedding phenomenon generates bandedenergy at higher frequencies for higher flow velocities.Discharge pulsation data (Figure 19) show behavior similar tothe suction side, but at much lower amplitudes. Broadbandturbulent energy and blade-pass excitation predominate these data.

Frequency (Hz)

Figure 19 . Discharge Line Pulsat ion Measured n ear Com pressorStation 7.The noise data and the measured shell wall vibration of thesuction and discharge piping indicated that most of the vibrationand noise were caused by pulsation in the suction pipe. Vibrationampli tudes were measured to be as high as 100 g's zero-peak at

several different frequencies. While these levels were not thoughtto be sufficient to cause failures of the piping itself, attachments tothe piping (thermowells, pressure connections, instruments, etc.)could be expected to fail.Shaft vibration data showed that the vibrations were lowamplitude. Furthermore, the majority of the vibration was at the firstfew multiples of running speed. However, a few low amplitudepeaks were apparent in the axial vibration data that corresponde d topulsation in the suction piping. These data showed that the pulsationand piping vibration were not adversely affecting shaft vibration.Impact data-With the compressor shut down, impact responsedata of the suction and discharge piping were gathered. Both theimpact and response were in a radial direction.The impact tests produced a large number of response peaks(Figure 20). However, the peaks were in distinct groups. Althougha full modal analysis was not done, experience has shown that theindividual peaks in these groups of natural frequencies willprobably have similar mode-shapes (e.g., a five-diameter mode),but are at slightly different frequencies due to local discontinuitiessuch as branch connections or weldolets.

90 DEGREES FROM TOP I

45 DEGREES FROM TOP

2

TOPOF PIPE A 1

Frequency Hz)

Figure 20. Inzpact Response Data of Suction Piping Shell WallNear Strainer:

The measured shell wall natural frequencies (from impact tests)did not correlate well with the measured shell wall vibration peaks(when operating), although some frequencies were in close prox-imity. These data indicate that the piping shell wall vibration wasnot being significantly amplified by structural natural frequencies.Field Tests-Station 9Similar testing was performed at Station 9. This station hadhigher flow rates than the Station 7. The compressor had an

impeller with 14 vanes.As shown in Figure 21 , pulsation data show ed that Station 9 wa salso experiencing high pulsation due to excitation of cross-wallacoustical modes. For purposes of this paper, the remainder of thedata is not presented in detail, but will be included in summ ary forcomparison purposes.Analysis of Measured and Calculated DataAs shown in Table 5 , there was good agreement between themeasured pulsation frequencies and the calculated cross-wallacoust ical natural frequencies. These data confirm that themeasured pulsation was the result of excitation of cross-wallacoustical modes in the piping.


13/24

SOURCES AND REMEDIES OF HIGH-FREQUENCY PIPING VIBRATION AND NOISE 20 1

Frequency (Hz)

Figure 21. Suction Line Pulsation Measured Downstream ofStrainer-Station 9.Table 5. Comparison of Measured and Calculated Cross- Wall Modes.I Mode I Calculated I Measured Frequencies I

Frequency Frequency Frequency INo. ValuesI (m. n) I

In situations where variable frequency energy excites systemnatural frequencies, "interference diagrams" can be used to helpunderstand the coincidence of excitation and resonance. Lines areplotted on the diagram representing system resonances. Additionallines are plotted that show variable frequency excitation. Points ofintersection between excitation lines and system resonancesindicate areas where the energy can excite a resonance. Such aninterference diagram was generated and is shown in Figure 22.The vertical axis is assigned to be frequency in Hz. Naturalfrequencies (n = 0 cross-wall modes) were plotted with horizontaldashed lines intersecting the vertical axis. These lines correspondto the "Calculated Values" column of Table 5.For this case, the variable frequency energy is provided byvortex shedding, which is a function of the obstruction diameterand geometry (which are constant), and flow velocity (which isvariable). Therefore, by assigning the horizontal axis to representflow velocity, diagonal lines were plotted to represent the variablefrequency excitation for a particular obstruction size.

Station 7

. ... . .- alcvlaledVodex Sheddmg Frequency For Enecllve ObriructionSireCalculatsdAwunic Cmrr-Wall Mode Fiequsncy- tll8on 7- tation9 VeloaV Range O v e r W n h Ciorr-Wall Mode war Measured

Station 9

23 w,ameter . . - .12 0 , a m t e r . . . - - - - -11 Dlametei - ~

4 Diameter

0 10 20 30 40 M 60 70 80 00 100Flow Vclosity FVSec)

Figure 22 . Interference Diagram for Stations 7 and 9-Cross- WallAcoustic N atural Frequ encies and Vortex Shedding Frequencies.For this diagram, the variable frequency lines are calculatedvortex shedding frequencies for a Strouhal number of 0. 5 and flowvelocity of zero to 100 ft/sec. Three diagonal lines are shownrepresenting a potential for energy generation for obstruction sizesof 0.1 inch, 0.2 inch, and 0.4 inch.For example, a 0.2 inch obstruction in a 100 fthe c flow streamwill generate energy near 3000 Hz. The same obstruction in a 30ft/sec flow stream will generate energy near 800Hz. The diagonalline drawn through these points shows excitation over a range offlow velocities.Measured pulsation responses were plotted along this diagram toestimate obstruction size. The measured cross-wall naturalfrequencies (from the "Measured Frequencies") columns of Table5 were plotted using horizontal lines with square ends. The lengthof these lines show the velocity where the particular cross-wallmode turned on or off. Dotted vertical lines are shown to remesentthe range of velocities over which the compressor operated.With all these data plotted, an imaginary excitation line wasdrawn through the middle of the measured cross-wall mode ranges.This line represents an estimated ob struction size. Therefore, using

this interference diagram, an obstruction size of 0.15 inch to 0.2inch was predicted. Note however that this was based upon averageflow velocities around a cylinder, oriented perpendicular to theflow. Since all these factors are approximations, the actualobstruction size may have been somewhat different. However,these data provided an indication of the range of obstruction sizesthat could have been causing the vortex shedding.ConclusionsBased upon the measured data and calculations, it wasconcluded that the strainer was generating vortex shedding energythat was being amplified by cross-wall acoustical modes in thepiping. Stiffener rings on the strainer had approximately thecorrect dimensions to p roduce the vortex energy. However, it wasnot completely proven that these rings generated the vortex

energy.It was recommended that a modification be considered thatwould both reduce the potential for vortex energy generation, andalso prevent acoustical cross-wall formation in the frequency rangeof interest.ModificationThe proposed strainer modification is shown in Figure 23. Thestiffener rings were removed and longitudinal plates were added toprevent the formation of acoustical cross-wall modes, and providestructural support for the perforated metal of the strainer. Theperforated metal was rolled at the edges to reduce stressconcentration at the attachment point to the longitudinal plates.


14/24

PROCEEDINGS OF THE 28TH TURBOMACHINERY SYMPOSIUM

Figure 23. Proposed Strainer Modijkation.This modified strainer was installed at Station 9. Upon startup,it was immediately obvious that the strainer modification hadsolved the noise problem.Pulsation, noise, and vibration data were acquired with thecompressor operating and the modified strainer installed. As shownin Figure 24, the high-frequency vertically aligned peaks inpulsation were still present, but they were greatly reduced inamplitude (compare Figure 21 with Figure 24). Note that theblade-pass energy pulsation (14x running speed) was

approximately the same before and after modification.pK5Y/ / / / / / / / iULSATI N 5 PSI P-PIDIV

I0 250 500 750 1000 1250 1500 1750 2044 2250 2500

Frequency (Hz)

Figure 24. Suction Line Pulsation after Installation of ModsedStrainer-Station 9.The suction pulsation levels were reduced by a factor of six (12psi peak-to-peak to 2 psi peak-to-peak). Piping vibration

amplitudes do wnstream of the silencer were reduced by a factor of25 (125 g's zero-peak to 5 g's zero-peak). Noise levels were alsoreduced to acceptable levels.Based upon these data, it was concluded that the strainermodification prevented formation of the cross-wall modes in thepiping section where the strainer was installed. These modesapparently re-formed either upstream or downstream of thestrainer, indicating that som e vortex shedding was still occurring inthe system. However, the vibration and noise levels were muchlower with the modifications.This case history illustrates potential problems with strainersand other obstructions installed in the piping. In situations such asthese where an obstruction is intended to be permanently installed,the possibility for vibration and noise problems should be consid-ered. Calculations of coincide nces between potential vortex sheddingfrequencies, shell wall modes, and acoustical cross-wall modesshould be performed in the design stage. Note, however that addition-al factors not included in such calculations (such as flow angles,Iocation of the obstructicrrr, damping, artached s ma l l -b re piping,etc.) can, in practice, significantIy influence whether such coinci-dences will actually combine to cause noise and vibration problems.Compressor Station-Discharge Side

A pipeline compressor station was experiencing excessiuet ~ h r a t i o nnd nmse Ievels on the discharge piping. Even &mgh thecompressor was installed in an enclosed, acoustically insulatedbuilding, nezrby residents were comphining of excessive noise.The noise levels measured n ear the house were amroximatelv 7 0dBA. The Federal Energy Regulatory commissidn' (FERC) ioiserequirements are 55 dBA LDN, which for a continuous source isequivalent to 48 dBA.The single-stage overhun g centrifugal compressor was driven byan electric motor through a speed increaser. The motor wasequipped with a variable frequency drive (VFD) that allowed thecompressor speed to be varied over a speed range of approximately4600 rpm to 6700 rpm.Field tests indicated that the noise levels were predominantly atthe compressor vane-passing frequency (17x running speed) andtwice the vane-passing frequency ( 3 4 ~ unning speed). Themaximum noise levels occurred at approximately 1900 Hz (17x )when the compressor was running near 6600 rpm. The overallsound levels in the building were approximately 114 dBC.Measured data and calculadons revealed that the excessive noiselevels were due to the coincidence of acoustical cross-wall naturalfrequencies and mechanical shell wall natural frequencies of thepiping. Pulsation at the compressor blade-passing frequencies wasamplified by the cross-wall natural frequencies. The pipingvibration due to the pulsation at the blade-passing frequencies wasfurther amplified by the shell wall natural frequencies of thepiping, resulting in high amplitude shell wall vibration andexcessive noise levels.

Cross- Wall Natural FrequenciesThe discharge pulsation amplitudes generated by the compressorwere approximately 0.2 psi to 0.5 psi peak at the blade-passing

frequency (17x), and 0.1 psi to 0.2 psi peak-to-peak at twice theblade-passing frequency (34x). These pulsation amplitudes wereconsidered to be typical for compressors of this type. The pulsationsgenerated by the compressor at the blade-passing frequencies wereamplified when the vane passing frequencies were coincident withthe cross-wall natural frequencies of the piping.A comparison of the calculated acoustic cross-wall naturalfrequencies with the measured pulsation response frequencies isshown in Table 6. As shown, the calculated diametrical modesfavorably agreed with the acoustic natural frequencies excited bythe pulsation at multiples of the compressor blade-passingfrequency. Acoustic natural frequencies up to the sixteenth modenear 3700 Hz were measured.


15/24

SOURCES AND REMEDIES OF HIGH-FREQUENCY PIPING VIBRATION AN D NOISE 203Table 6. Comparison of Calculated and Measured Cross-WallNatural Frequencies in 26 Inch Piping (wt = 0.5 in).

Calculated Measured Pulsation I

,000 0 500 1000 $5043 2000 2500 3000 3500 40W 4500 5000Frequency (Hz)

M o d e(m, n)

Figure 25. Pulsation Due to Quarter-Wave Stub Resonances.

Excited by Energy at 17xFrequency

(Hz)

Excited by Energy at 34x

Quarter Wave Stub Natural Frequencies

( 9 , o ) I 2202

The pulsation data summarized in Table 6 were obtained with aninsertion probe. As shown, the pulsation amplitudes were in therange of 0.5 psi peak-to-peak to 6 psi peak-to-peak. However,pulsation data obtained by installing pressure transducers at theend of stubs indicated pulsation levels of approximately 100 psipeak-to-peak at the quarter wave stub natural frequencies. Thesepulsation data illustrate the importance of using insertion probeswhen acquiring high-frequency pulsation data.An example of pulsation due to the quarter wave stubresonances is shown in Figure 25. The pulsat ions weremeasured at a thermowell test location in the flange thatresulted in a stub with an effective length of ap proximately 9.4inch between the pressure transducer and the inside of the pipe.The stub resonances were excited by pulsation at 17 x runningspeed. As shown, the stub created major acoustic responseswith pulsation levels of approximately 25 psi peak-to-peak at6 00 Hz , 125 psi peak-to-peak at 1400 Hz, and 75 psi peak-to-peak at 1800 Hz .

These quar ter wave s tub responses were ver if ied bycomputing the acoustic natural frequencies of the stub. Thecomputer model included the tubing f i t t ings , the s tub(thermowell), and a short section of the discharge piping (Figure26) . The computed response f requencies and pulsat ionamplitudes favorably agreed with the measured pulsation data,which also verified that the high amplitude responses measuredat the end of the stubs were acoustic natural frequencies of thestubs, and were not cross-wall natural frequencies of thedischarge piping. The pulsation amplitudes at the stubresonances were maximum at the closed end of the stub at thepressure transducer and were almost zero at the open end of thestub at the main piping (Figure 27).

Frequency(Hz)

Figure 26 . Computer Model of Thermowell Pressure Tap.

Pulsationpsi, p-p

-

Frequency (Hz)

-

Figure 27. Computed Passive Acoustic Frequency Response forThermowell .


16/24

204 PROCEEDINGS O F THE 28TH TURBOMACHINERY SYMPOSIUMAlthough pulsations at the stub frequencies are considered tobe measurement errors when measuring pulsation in the mainpiping, the high pulsation in the stubs is also a potential source ofhigh-frequency energy that can excite the piping shell wallmechanical natural frequencies. The available shaking force canbe estimated by multiplying the c ross-sectional flow area (in2) bythe pulsation amplitude (Ib/in2) (Force = area H pulsation). Forexample, the shakin g forces at the closed end of a 1 inch diameterthermowell with 100 psi peak-to-peak pulsation would beapproximately 80 Ib peak-to-peak. This is a significant high-frequency shaking force, which can excite the shell wall natural

frequencies. Therefore, the number and lengths of stubconnections should be minimized.Shell W all Natural FrequenciesImpact tests were performed to measure the shell wallmechanical natural frequencies of the suction piping, dischargepiping, and suction scrubber. The piping was impacted with asteel tipped instrumented hammer to excite the high-frequencyshell wall natural frequencies. The responses at the shell wallnatural frequencies were measured with accelerometers, straingauges, and sound level meters. The accelerometers weremounted on pads that were glued to the piping. The straingauges were installed near the accelerometers and were orientedto measure the strain in the circumferential direction (hoopstress). Th e sound level meters were mounted on tripods near the

piping. The shel l wall responses measured with theaccelerometers, strain gauges, and sound level meters favorablyagreed (Figure 28).

Figure 28. Measured Shell Wall Natural Frequencies Due toImpacts.The frequencies of several of the discharge piping shell wall

natural frequencies increased when the piping was pressurized to825 psi. The increase in natural frequencies was due to stiffeningeffects from the internal pressure. The computed and measuredshell wall natural frequencies for the 26 inch diameter dischargepiping are compared in Table 7. As shown, the increase in naturalfrequencies due to the pressure could be approximated byincreasing the wall thickness (wt) by 10 percent to 0.55 inch. Thegood agreement between the measured and calculated frequenciesconfirmed that equations in the section, Shell Wall NaturalFrequencies, can be used to approximate the shell wall naturalfrequencies.The m aximum piping vibration levels measured on the dischargepiping were approximately 1 00 g's peak-to-peak near 19 00 Hz

Table 7. Comparison of Calculated and Measured Shell WallNatural Frequencies in 26 Inch Piping (wt = 0.5 in) .Computed MeasuredNumber of w t =

( 1 7 ~t 6700 rpm) (Figure 29). These piping vibration levels werenot considered to be excessive with regard to fatigue failures of themain piping; however, the resulting noise levels were excessive.Furthermore, experience has been that such vibration levels cancause failures of small diameter lines and instrumentation attachedto the piping.9000 d / 17x I I I I I I I ?- .- - -u Acceleration, 50 g ' s p - p / d i v l o ~21$a000 I I I I I I l/l,

0 5 W 10 W 1500 2000 25 W 3000 35 W 4000 4553 5000Frequency(Hz)

Figure 29 . Discharge Piping Shell Wall Vibration During Startup.Although the pulsation levels were higher at the cross-wallnatural frequency near 1300 Hz, the maximum vibration, strain,and noise levels occurred at the resonance near 1900Hz. The testdata and the calculations indicated that the eight-diameter shellwall mechanical natural frequency and the eighth diametrical (8,O)cross-wall acoustic natural frequency were both coincident withthe vane passing frequency when the compressor was operating atapproximately 6700 rpm. This suggests that the shell vibrationlevels were easier to excite when the pulsation mode shapematched the mechanical mode shape.


17/24

SOURCES AND REMEDIES OF HIGH-FREQUENCY PIPING VIBRATION AND NOISE 205Pipe Clamp Mechanical Natural FrequenciesThe maximum noise levels were measured near a large thermalstop pipe clamp at the compressor discharge nozzle (Figure 30).The vibration levels were approximately three times higher on thepipe clamp compared with the levels on the pipe wall. Impact testson the pipe clamp revealed several mechanical natural frequenciesnear 190 0 Hz. The sh ell wall vibration of the discharge piping at1900 Hz excited the mechanical natural frequencies of the pipeclamp. The vibration of the pipe clamp caused the noise levels tobe further increased. As shown in the photograph, this therma l stoppipe clamp was not originally treated with sound insulation.

Figure 30. Pipe Clam p on Discharge Piping. (N ote: The clamp waslater treated with sound insulation.)Noise TreatmentThe preferred modification to reduce the excessive noise levelswas to reduce the discharge pulsation energy at the blade-passingfrequencies, which was the source of the piping vibration andnoise. One possibility was to install a high-frequency pulsationfilter in the discharge piping; however, it would require majorpiping m odifications to install the filter. Another possibility was to

prevent the amplification of the pulsation by eliminating the cross-wall natural frequencies at the blade-passing frequencies byinstalling modifications inside the discharge piping, such as thetube bundle discussed in the section, Prevention of Cross- WallAcoustical Modes. However, due to time limitations, the tubebundles could not be installed.Since the measured strain levels were considered to be low withregard to potential fatigue failures of the main piping, it was decidedthat noise treatment would be installed to reduce the noise levels,rather than reducing the pulsation and vibration levels. Additionalnoise treatments w ere installed on all the inside discharge p iping, thepipe clamps, the scrubber, and the outside block valves and controlvalves. The following modifications were installed.

On the outside yard piping, an extra layer of sound treatmentwas installed o ver the existing sound insulation, which was sim ilarto the noise insulation shown in Figure 10. The final configurationconsisted of the following: 4 inch of glass fiber adjacent to the pipewall, a sound barrier jacketing, a vinyl layer, 2 inch of glass fiber,and another sound barrier jacketing.The discharge piping and clamps inside the compressorbuilding were lagged with noise insulation similar to Figure 10.Sound blankets were installed on all vents, block valves,

control valves, and check valves.The suction scrubber and the attached inlet and outlet pipingwere lagged with no ise insulation similar to Figure 10.The seal between the building and the discharge piping wa sr rmrwxi to prrvrrrt the piping vibration lev& from k.iagmechanically transferred to the building walls where the noiselevels could be radiated from the building walls.Seals were installed under the building doors to prevent the

noise from passing though the cracks under the doors.After these modifications were installed, the noise levels at thenearby house were reduced to acceptable levels. No noisetreatment was installed on the suction piping inside the compressorbuilding, because the test data indicated that the pulsation andvibration levels were very low in the suction piping. It is felt that

the pulsation levels were significantly lower on the suction side ofthe compressor due to the geometry of the compressor case thatprevented the pulsation from being directly transferred into thesuction piping.Refine?-Compressr,r Wh eel Failure

The compressor wheel of an overhung, motor driven centrifugalcompressor had experienced cracking at four locations on thewheel. The compressor wheel had 17 blades, and was operated ata constant speed of 4558 rpm (76 Hz).The location of the failures corresponded with node lines of athree-diameter structural natural frequency of the compressorwheel (Figure 3 1). Therefore, it was suspec ted that excitation of acompressor wheel natural frequency might have caused thefailures. An exp erimental mod al analysis of the compressor whee lwas conducted to determine if one of its natural frequenciescoincided with running speed or blade-pass excitation.

Crack

C r a c kFigure 31. Location of Compressor W heel Fatigue Failures.


18/24

206 PROCEEDINGS OF THE 28TH TURBOMACHINERY SYMPOSIUMDuring the impact testing, the compressor wheel was supportedby a mandrel and hun g from an overhead crane. The numb er of testpoint locations was selected to provide information for structuralmodes having up to four diametrical (m) and two annular (n) nodelines. A 1 Ib instrumented hammer was used to excite the wheel.Response data were obtained at each test point location from a 0.8oz wax-mounted accelerometer.The mod al analysis identified several wheel natural frequencies,three of which are shown in Table 8. Since blade-pass excitationwas 1292 Hz , it appeared initially that the (3,O) wheel mod e wouldnot be excited, and therefore could n ot be the cause of the failures.

However, pulsation data were acquired to determine if other energysources were present.Table 8. Compressor Wheel Mode Natural Frequencies

Wheel Mode Natural Frequency(Hz)

Pulsation data acquired at the inlet to the compressor showed astrong response peak at 1140 Hz (Figure 32). Note that thepulsation data were not acquired with an insertion probe.Therefore, several stub frequency peaks can be seen in the data.While the pulsation at 1140 Hz would provide the excitationnecessary to excite the (3, 0) compressor wheel mode (and hencecause the failures), it was unknown what was generating thepulsation energy.

Table 9. Cross- Wall Acou stical Natural Frequencies.High-Order Acoustical Natural Frequency (Hz)0 1 2 I 3

acoustical mode, which in turn excited the (3,O) structural mode ofthe compressor wheel, ultimately causing the experienced failures.It was therefore suggested that the butterfly valve be relocatedfurther upstream of the compressor, or that longitudinal vanes beinstalled downstream of the valve to "turn off' the high-ordercross-wall mode.Since this valve was not mandatory to the process, plantpersonnel decided to remove the valve completely during aplanned outage. After the valve was removed and the processrestarted, pulsation data were acquired. The data showed that thepulsation energy peak at 11 40 Hz was no longer present. The unithas been operating without failures for several years since the valvewas removed.LNG Plant-Turning Vane Failure

A refrigeration compressor experienced a catastrophic failure afterturning vanes in the first interstage crossover piping failed and wereingested by the second stage impeller. The compressor operates on100 percent nitrogen in a closed loo p process (Figure 33).

""""""'1Flap ValveFigure 32 . Pulsation Data at Com presso r Inlet.Since a butterfly valve was located just upstream of thecompressor inlet, it was hypothesized that vortex shedding fromthis valve might be exciting high-order acoustical modes in theinlet piping, thereby generating the measured pulsation energy.Using the equations in the section, Acoustic Cross-Wall NaturalFrequencies, the high-order acoustical modes for this pipe werecomputed and are presented in Table 9. As shown, the calculationspredicted a high-order ( I , 2) mode at 1 137 Hz .The measured pulsation data and the results of the calculationsseemed to confirm the hypothesis that vortex shedding energy gen-erated by the butterfly valve was exciting a high-order cross-wall

Figure 33. Refrigeration Compressor (First and Second Stuges).The compressor was driven by an induction electric motor atapproximately 3000 rpm (50 Hz). The first and second stageimpellers are mounted on the low speed pinion shaft, whichoperates at approximately 20,540 rpm (342.3 Hz). Similarly, thethird and fourth stage impellers are mounted on the high-speedpinion shaft that operates at approximately 28,400 rpm (4 73.3 Hz).


19/24

SOURCES AN D REMEDIES OF HIGH-FREQUENCY PIPING VIBRATION AN D NOISE 207After operating for approximately 250 hours, the compressorexperienced a m ajor failure of the second stage impeller and pinionshaft when a piece of one of the second stage turning vanes brokeoff and entered the second stage impeller. The second stageimpeller was damaged, which caused the vibration levels on thesecond stage pinion to increase suddenly due to the increasedunbalance. When the compressor was shut down, the vibrationamplitudes increased to excessive levels as the pinion shaft coasteddown through its first lateral natural frequency. When thecompressor was disassembled, it was determined that majordamage had occurred on the second stage impeller, pinion shaft,

and seals.Inspection of the turning vanes revealed that each of the fiveturning vanes was cracked where the vanes were welded to thepipe. Metallurgical examination of one of the vanes confirmed thatthe cracks were fatigue cracks that appeared to be initiated atsurface pits on the concave surface near the edge of the vaneadjacent to a weld. A sketch of the typical vane cracks is shown inFigure 34. The propagation pattern of the cracks suggested that thecracks were probably due to the excitation of a particularmechanical natural frequency, because the location and shapes ofall the cracks were very similar. In addition, the cracks did notpropagate along the fixed edges of the vanes that would probablybe the location of maximum stress for a simple bending of thevane.

WELDED TO P I P E --/Figure 34 . Turning Va ne Failure Locations.

Due to the problems with the second stage turning vanes, thefourth stage turning vanes were also examined for cracks; however,no cracks were found in these vanes. This indicated that the failuresof the second stage turning vanes could be associated with thereported high vibration levels on the interstage piping between thefirst and second stages. The vibration levels were reported to beparticularly high near the miter joint at the entrance to the secondstage impeller (as shown in Figure 33, the turning vanes are insidethe m iter joint).Mechanical Natural Frequencies of the Turning VanesIn an effort to solve the failures of the turning va nes, a new miterjoint was fabricated with thicker vanes. The thickness of the newvanes was increased from 4 mm to 6 mm.Prior to installing the new miter joint, the mechanical naturalfrequencies and vibration mode shapes of the turning vanes weremeasured using impact testing. Each of the vanes was divided intoa grid with 25 test points. Each vane was impacted near the middleof the vane with an instrumented hammer. A small (I gram)accelerometer was used to obtain the vibration data at each of the25 test points. The accelerometer was attached using wax.Each of the turning vanes had several mechanical naturalfrequencies between 1000 Hz and 10,000 Hz. The impact tests

were made prior to running the compressor; therefore, it wasunknown which of the many mechanical natural frequencies couldbe excited. Initially, it was felt that the turning vane failures couldpossibly be due to excitation of one or more of the mechanicalnatural frequencies by pulsation at the impeller vane-passingfrequency at approximately 5477 Hz (20,540 rpm x 16 vanes =328,640 cpm = 5477 Hz). Therefore, the modes near the vane-passing frequency were analyzed in greater detail. However, dataobtained during operation indicated that the pulsation and strainlevels at the vane-passing frequency were low, which indicated thatthe failures were not due to excitation at the vane-passingfrequency.Strain and Pulsation Data During O perationStrain gauges were installed on each of the van es before the newmiter joint was installed (Figure 35). Since the strain gauges wereinstalled inside the piping, the wires to the strain gauges had to berouted through special fittings to provide a pressure seal.

Guide Voner Stage ' I 2nd Stoge

Figure 35. Instrumentation Test Locations.Pulsations were measured at several locations in the pipingusing dynamic pressure transducers that were installed using shortpipe fittings to minimize the quarter wave stub effects (Figure 35).The flow rates through the compressor were controllkd usingvariable inlet guide vanes (IGVs), which were installed upstreamof the first stage impeller. The bypass valve discharged into the firststage inlet piping just upstream of the IGVs. The IGVs weredesigned to throttle the inlet flow and to direct the flow into thefirst stage impeller.Strain and pulsation spectra were plotted versus time as the inletflow was varied. When the bypass valve was fully opened, thestrain levels on the top turning vane increased to approximately400 microstrain peak-to-peak (Figure 36). These strain levels wereconsidered to be excessive an d could result in fatigue failures. Thestrain levels were primarily at a response near 2100 H z, which wasone of the vane mechanical natural frequencies. This mechanicalnatural frequency was excited by the pure tone pulsation near 2100Hz (Figure 37).The turning vane mechanical natural frequencies appeared to beexcited by flow induced pulsation that was coincident with thevane mechanical natural frequencies. It was thought that thepulsations were due to low amplitude energy vortices formed byflow across an obstruction (vortex shedding).As shown in Figure 37, the pulsation frequencies shifted as thebypass valve position was changed. When the bypass valve wasI00 percent open, the pulsation near 2100 Hz was coincident withthe mechanical n atural frequency of the top turning vane. When thepulsation frequency was lowered from 2100 Hz to 1900 Hz, thestrain levels on the top turning were significantly reduced becausethe pulsations were no longer coincident with the mechanicalnatural frequency near 21 00 Hz.


20/24


FreguencY (Hz)

Figure 36. Strain on Up per Turning Vane as Bypass Valve PositionVaried.4W

NOTES PULSATION NEAR TURNING VANE NATURAL FREQUENCIESI N

I PULSATION FREQUENCIESVARIED WTH FLOW VELOCITY I

M0 I TURNING VANE MECHANICAL NATURAL FREQUENCIES

PULSATION 20 PSI P-PIDIV

Frequency (Hz)

Figure 37. Pulsation Upstream of Turning Vane as Bypass V alvePosition Varied.Pulsation at a second acoustic natural frequency tracked fromapproximately 1600 Hz to 1400 Hz as the bypass valve was closedfrom 100 percent open to 50 percent open (Figure 37). When thepulsation frequency was reduced to approximately 1400 Hz, thepulsation energy was coincident with the m echanical natural frequencyof the secon d turning vane and the strain levels were increased on thesecond turning vane (Figure 38 ). These data indicated that the turningvanes could be excited over a large range of operating conditions.The acceleration frequencies measured on the second stage miterjoint correlated with the strain frequencies measured on the turningvane. As shown in Figure 39, there were several major responsesbetween 1000 Hz and 7000 Hz. The acceleration levels weregenerally reduced as the bypass valve was closed to 5 0 percent open.The good correlation indicated that the acceleration data measuredon the miter joint c ould be used to d etermine if the vibration levelsand resulting strain levels on the turning vanes would be excessive.Acoustic Cross- Wall Natural FrequenciesThe cross-wall acoustic natural frequencies for the first tosecond interstage piping were computed and compared with themeasured pulsation responses in Table 10.

,I"

STRAIN 200 MICR0STRAINX)IV

Frequency (Hz)

Figure 38. Strain on Sec ond Turning Vane us Bypass Valve PositionVaried.

Frequency (Hz)

Figure 39 . Acceleration on Miter Joint as Bypass Valve PositionVaried.Table 10. Comparison o f Calculated and Measured Cross- WallAcoustical Natural Frequencies.

I Fi r s t - Second I n te r s t ap ( P ipe D iamete r = 9.8 inch)Cross-Wall Acoustic Natu ral Frequen cies (Hz)

Calculated Measured


21/24

SOURCES AND REM EDIES OF HIGH-FREQUENCY PIPING VIBRATION AND NOISE 209The cross-wall natu ral frequencies are a function of the speed ofsound, which for this case is primarily controlled by thetemperature an d the pressure of the nitrogen. The calculations w erebased upon a speed of sound of 1260 fthe c.As shown, there was excellent correlation between thecalculated acoustic natural frequencies and the measured responsefrequencies. The pulsation data indicated that several of thesecross-wall modes occurred simultaneously and were thought to bethe primary cau se for the excessive vibration and noise levels in theinterstage piping. Several of these pulsation frequencies were alsocoincident with the mechanical natural frequencies of the turning

vanes.

One obvious modification to prevent the failures of the turningvanes was to remove the turning vanes. Therefore, it wasrecommended that the turning vanes should be removed andreplaced with a flow splitter to direct the flow into the second stageimpeller. In addition to directing the flow, the flow splitter wouldalso break up the cross-wall modes. The miter joint was replacedwith a long radius elbow and a flow splitter was installed as shownin Figure 40.

"1 st-"*or-H----Z;H

Figure 40. Mod$cations Installed to Prevent Formation of Cross-Wall Modes.

Although it was thought that these modifications would improvethe flow into the second stage impeller, the pulsation in theinterstage piping upstream of the flow splitter could still exist andcould possibly cause fatigue failures of the flow splitter. It was feltthat the required solution was to improve the flow into the IGVs.There were several possible modifications; ho wever, it was felt thatthe most feasible modifications would be to install a second flowsplitter just upstream of the IGVs. As shown in Figure 40, a flowsplitter was designed which consisted of a 4 inch pipe mountedinside an 8 inch pipe. The flow from the bypass valve entered theflow splitter through perforations in the outer 8 inch pipe. Theperforations were designed to also dampen the pulsations andreduce the turbulence from the bypass valve.The new flow splitters were installed directly upstream of thefirst and second stage impellers. Since these flow splitters couldpossibly experience fatigues, it was decided to install strain gaugeson both flow splitters. The strain data indicated that the vibrationand strain levels on the splitters were low.The inlet flow splitter reduced the pulsation in the first stageinlet piping upstream of the IGVs. Reducing the pulsation alsoreduced the strain levels on the IGVs. The pulsation levels at themidpoint of the first to second stage interstage piping weresignificantly reduced after the modifications were installed. Thepulsation levels at the entrance to the second stage impeller werealso reduced. The acceleration levels on the elbow at the secondstage suction were also significantly reduced. After the

modifications were installed, the vibrations on the interstagepiping were low amplitude.The compressor continued to operate satisfactorily after themodifications were installed. These modifications were laterinstalled in another compressor at the site and both compressorscontinued to operate with no additional failures.Steel Plant-Screw Compressors

The discharge silencers installed on screw compressorsexperienced fatigue failures after only a few hours of operation.Noise levels near the compressors were extremely high and thenoise levels were still excessive outside the plant boundaries. Fieldtests and calculations showed that the excessive noise levels andthe fatigue failures were due to coincidences of aco ustic cross-wallnatural frequencies and mechanical shell wall natural frequencies.

Instrumentation ProblemsDue to the High-Frequency VibrationIn an effort to determine the cause(s) for the fatigue failures ofthe silencers and the excessive noise levels, the plant personnelattempted to measure the vibration levels on the silencers. Initialvibration spectra indicated vibration levels of approximately 1000mils peak-to-peak (1 inch peak-to-peak) at low frequencies lessthan 10 Hz. Although the plant personnel felt that the data werecorrect, it was unlikely that the compressor was actually vibrating

with displacements of 1 inch peak-to-peak at any frequency. Inaddition, this high amplitude vibration should be easily visible.It was later determined that these vibration data were indeedincorrect. The actual vibrations were not low-frequency vibration,but were high-frequency vibration above 1000 Hz with amplitudesgreater than 500 g's zero-peak. Subjectively, the vibration wascharacterized by a "buzzing feel." Initially, a high sensitivityaccelerometer (100 mvlg) had been used in an attempt to measurethe high amplitude vibrations. The high-frequency accelerationscaused the accelerometer amplifier to be overloaded, whichresulted in a clipped signal (similar to a square wave). This clippedsignal was then electronically double-integrated, which resulted inthe high displacement amplitude at the low frequency. Thisillustrates the types of problems that can occur when measuringhigh amplitude, high-frequency accelerations.The actual acceleration levels were later measured using a lowsensitivity accelerometer (10 mvlg). Even with the low sensitivityaccelerometers, it was difficult to measure the high amplitude,high-frequency accelerations. Initially, the accelerometers weremounted on pads that were epoxied to the pipe; however, the epoxyglue joint would fail within a few seconds after the compresso rstarted. To eliminate the problems with the epoxy, theaccelerometers were attached to steel pads that were welded to thepipe. The accelerometers remained attached to the piping, but thecoaxial connectors on the accelerometers also failed after a fewseconds. This connector failure problem was improved bysoldering the wires to special adapters that were installed on theaccelerometers. Also, it was determined that the quality of thecoaxial connectors varied between manufacturers and that certainbrands were able to withstand the high amplitude vibration without

experiencing a failure.The high vibration levels also caused similar problems with thepressure transducers. To avoid stub resonances, the pulsation datawere obtained using insertion probes (refer to the section , InsertionProbes).The high-frequency vibration also caused fatigue failures ofthe coaxial connectors on the pressure transducers, and frettingproblems on the stainless steel tubing of the insertion probes. Inaddition, one of the pressure taps experienced a fatigue failure thatcaused loss of both the tap and insertion probe. The failure did notresult in a fire since the com pressor was operating on nitrogen at thattime. Although no on e was injured when the tap broke away from themain pipe, this failure illustrates the potential dangers of conductingtests on piping with excessive high-frequency vibration levels.


22/24

210 PROCEEDINGS OF THE 28TH TURBOMACHINERY SYM POSIUM

Pulsation Generation in Screw CompressorsThe dry screw compressor consists of a female rotor with sixhelical lobes and a male rotor w ith four helical lobes that intermeshwith each other (Figure 41). The intermeshing of the helical lobesgenerate pulsation at the pocket-passing frequency that is equal tothe number of lobes on the male rotor (four) multiplied by thecompressor running speed. In addition to the pocket-passingfrequency, pulsations are also generated at multiples of the pocket-passing frequency.

Figure 4

Sources and Remedies of High Freq Piping Vibration n Noise.pdf

Documents

Transcript of Sources and Remedies of High Freq Piping Vibration n Noise.pdf