文档文库
手机版
投诉建议
热门搜索： 心得体会 演讲稿 思想汇报
首页 > Response of a vortex flowmeter to impulsive vibrations

Response of a vortex flowmeter to impulsive vibrations

发布时间：   来源：文档文库   
小 中 大
字号：
手机查看
FlowMeasurementandInstrumentation11(200041–49
www.elsevier.com/locate/flowmeasinst
Responseofavortexflowmetertoimpulsivevibrations
J.J.Miau
a
a,*
,C.C.Hua,J.H.Chou
b
InstituteofAeronauticsandAstronautics,NationalChengKungUniversity,Tainan70101,Taiwan,ROCb
DepartmentofEngineeringScience,NationalChengKungUniversity,Tainan70101,Taiwan,ROC
Received19May1999;receivedinrevisedform5July1999;accepted7September1999
Abstract
Experimentswereperformedtostudytheresponseofavortexflowmetertostructuralvibrationsduetoimpulsiveforcesappliedonthepipe.Vortex-sheddingsignalsobtainedbyapiezoelectricsensorembeddedinavortexshedderwereexamined.Majorfindingsaredescribedasfollows.First,byimprovingthedesignofthepiezoelectricsensor,thesensorsensitivitytostructuralvibrationscouldbereduced.Specificallyspeaking,thenoisecomponentduetoimpulsiveforcewithlevelupto13.8kNcouldberemovedeffectivelyfromtheoutput.Second,byapplyingrepetitiveimpulsiveforcesonthepipe,characterizedbyafrequencygreaterthanthevortex-sheddingfrequency,thequalityofvortex-sheddingsignalsmeasuredwasdegradedsubstantially.Thisisexplainedasbeingduetosuppressionofvortexshedding,notaprobleminconjunctionwiththecharacteristicsofthesensor.©2000ElsevierScienceLtd.Allrightsreserved.
Keywords:T-shapedvortexshedder;Piezoelectricsensor;Impulsiveforce
1.Introduction
Avortexflowmetermeasuresthefrequencyofvortexsheddinggeneratedbyflowoveravortexshedder,whichisknowntobelinearlyproportionaltotheflowvelocity.Owingtotheirwiderangeofapplicabilityinliquid,gasandsteamflowsinhostileenvironments,vortexflow-metersarefavoredonmanyoccasionsforindustrialflowmeasurements.However,performanceofavortexflowmeterusingapiezoelectricsensortodetectvortexsheddingisknowntobesensitivetopipingvibrations[1,2].Inevitably,thesensormaypickupstresscompo-nentsotherthanthoseduetovortexshedding;forinstance,stressesduetostructuralvibrations.Therefore,inpractice,onehastodealcarefullywiththeproblemofhowtoextractthevortex-sheddingfrequencycomponentfromthesignalsmeasured.
Intheliterature,OgawaandMatsubara[1]pointedoutthatthevibrationlevelsoftypicalindustrialpipingwerelessthan0.6gin99%ofcasesexamined,wheregdenotesthegravitationalacceleration.Understandably,
*Correspondingauthor.Tel.:+886-6-275-7575;fax:+886-6-238-9940.
E-mailaddress:jjmiau@mail.ncku.edu.tw(J.J.Miau
vibrationsofpipingstructuremayinducethevortexshedderintovibratingmotions,henceaffectingvortexshedding.Thisconsiderationfurthermotivatedthepresentauthorstopayattentiontoanumberofreportsintheliteratureconcerningthecontrolofvortexshed-dingbyactivemeans.Tanidaetal.[3]andStansby[4]studiedthephenomenonofvortexsheddingfromavibratingcylinder.Theyfoundthatresonanceorlock-onbetweensheddingvorticesandthevibratingcylindermayoccurwhenthecylindervibratesatfrequenciesclosetoamultipleofthevortex-sheddingfrequency.Inthelock-onrange,thevortex-sheddingfrequencyisshiftedtoeitherthevibratingfrequencyorthesub-har-monicsofthecylinder,insteadofvaryingwiththevelo-cityoftheincomingflow.Thelock-onphenomenonwasalsonoticedintheworksofBarbietal.[5]andArm-strongetal.[6],inwhichstationarycylinderssubjecttooscillatoryincomingflowwerestudied.Obviouslythelock-onphenomenonmightcauseerrorsinthemeasure-mentofoscillatoryflowwithavortexflowmeter.Mot-tramandRobati[7]andPeteretal.[8]pointedoutthaterrorsweremeasuredwhenthevortexfrequencieswereinthelock-onrangeduetooscillatoryincomingflow.Furthermore,theybothconcludedthatthelargesterrorsweremeasuredwhenlock-onoccursathalftheoscillat-oryfrequency.Ontheotherhand,Schummetal.[9]
0955-5986/00/$-seefrontmatter©2000ElsevierScienceLtd.Allrightsreserved.PII:S0955-5986(9900018-7

42J.J.Miauetal./FlowMeasurementandInstrumentation11(200041–49
foundthatthephenomenonofvortexsheddingcouldbesuppressedifthecylindervibratesatafrequencyaround1.8timesthenaturalsheddingfrequency.Intheirexperi-ments,thecylinderwasvibratingatasmallamplitude,i.e.,afewpercentofthecylinderdiameter.InHsiaoandShyu’sexperiment[10],acousticexcitationswereintroducedfromthesurfaceofacircularcylinder.Theynotedthat,withacousticexcitationsatfrequenciesintheneighborhoodoftheshear-layerinstabilityfrequency,thevortex-sheddingfrequencycomponentcouldbesup-pressed,asevidencedbythespectralresultsobtained.Apparently,understandingthemechanisminconjunc-tionwithsuppressionofvortexsheddingisasimportantasunderstandingthemechanismofthelock-onphenom-enon.However,relativelyfewpapersintheliteraturediscusstheformerissue.
Thepresentstudyisinterestedintheperformanceofavortexflowmetersubjectedtoexternalforcesappliedonthepipe,whichcanbeinaformofasingleimpulsiveforceormultipleimpulsiveforcesinsequence.Inapre-liminarystudy,itwasfoundthattheimpulsiveforcesappliedcausedspike-likenoisestoappearintheoutputsignal.Thesespike-likenoiseswereverydifficulttoremove,evenwithadesignoftwopressuresensorsinstalledonthevortexshedder[11].Hence,themeasure-mentaccuracyofthevortexflowmeterwasquestioned.Inthispaperweintendtopresentexperimentalresultsandphysicalunderstandingobtainedalongthisdirection.2.Apparatus
Experimentsweremadeinawatertunnelfacilitywhosetestsectionwasacircularpipe,150mmindiam-eter.Theaveragevelocityinthepipe,denotedasU,couldreach1m/s,asestimatedfromaVenturiflow-meterinstalledinthepipingcircuit.ThevortexshedderemployedhasaT-shapedcross-section,asshowninFig.1.Itsmaximumwidthnormaltotheincomingflowwas32mm,denotedasD.Thisvortexshedderspannedthetestsectionofthecircularpipe.AsseeninFig.1,apiezoelectricsensorwasfixedinthevortexshedderwithepoxy(ahardbondingmaterial,flushwithoutersurfaceofthevortexshedder.Thereforethepiezoelectricsensorbasicallydetectedtheunsteadyprocessofvortexshed-dingthroughstressvariationsofthevortexshedder.Thesensoroutputwasfedintoachargeamplifierwithabuilt-inintegrator,capableoffilteringoutspike-likenoisescausedbytheimpulsiveforcesapplied.
Duringtheexperiments,impulsiveforceswerepro-ducedbytheimpactofeitheranironmassfallingfreelyorahand-heldhammer.Theintensityofimpactwasmeasuredbytwoaccelerometersattachedontheironmassandthepipe,respectively,whosemeasuringrangewasbetween0.001gand4×104g.Inaddition,alaser-inducedfluorescencevisualizationtechniquewas
Fig.1.ExperimentalarrangementoftheT-shapedvortexshedderandaside-viewsketchofthepiezoelectricsensor,whereD=32mm.
employedtorevealflowstructuresimmediatelydown-streamofthevortexshedder.Flowmotionswererecordedbyavideocameraatarateof30frames/s.Theimageswerethenrecordedinacomputerinparallelwiththeoutputofthechargeamplifier.Hence,theflowmotioncouldbeexaminedframebyframeandcorre-latedwiththesignalmeasured.
Inthepreliminaryphaseofthisstudy,effortsweremadetoclarifyaconcernaboutwhetherthevortexsheddershouldbefixedrigidlyinthetestsectionorinsertedintothetestsectionwithoneendfree.Atestwithexternalforcesappliedbyahammeronthepipingstructurerevealedthatthesensoroutputobtainedinthelattercasecontainedsignificantenergyatlowfre-quenciesassociatedwithstructuralvibrations.Asnoted,theenergywasoverwhelminglydominantoverthatofthevortex-sheddingfrequencycomponent.Hence,theconfigurationwiththevortexshedderfixedrigidlyinthepipewaschosenforlaterexperiments.
3.Resultsanddiscussion
Forthevortexflowmeterinthisstudy,thedesignofpiezoelectricsensorembeddedinthevortexshedderwithepoxysuffersamajorproblemofbeingverysensi-tivetostructuralvibrations.Asanexample,Fig.2showstherawsignalobtainedbythepiezoelectricsensor,inwhicheachofthespikesisaresultofanimpulsiveforceapplied.Understandably,itisdesirabletoimprovethesensorcharacteristicsinordertoreducethesensitivitytostructuralvibrations,butmaintainthesensitivitytovortexshedding.

J.J.Miauetal./FlowMeasurementandInstrumentation11(200041–4943
Fig.4.Asketchofthedropmechanismandthearrangementofthepiezoelectricsensorandaccelerometers;theflowdirectionisfromrighttoleft.
Fig.2.Therawsignalofthepiezoelectricsensorwiththepresenceofexternalforces,nofilteringtechniquewasappliedontheoutputsignal,Re=1.9×104.
3.1.ReducingsensorsensitivitytoimpulsiveforcesAnimproveddesignofthesensorwiththeaimofminimizingthesensitivitytostructuralvibrationsisdescribedinthissection.AsshowninFig.3,apiezoe-lectricmaterialissituateduponarubber-typematerialforisolationfromstructuralvibrations.Asiliconfilmcontactingthepiezoelectricmaterialisflushwiththeoutersurfaceofthevortexsheddertocollectpressurefluctuationsexerteduponthefilm.Inthisway,thesensorcandetectpressurevariationsassociatedwithvortexsheddingandminimizeitssensitivitytostressvariationsexperiencedbytheextendedplate.
Theperformanceofsensorswithregardtotheoriginalandimproveddesignsdescribedabovewasexaminedunderimpactforces.AsshowninFig.4,themagnitudeofimpactforcecouldbecontrolledbyliftinganironmassof1.224kgtoadesiredheightandthensetfallingfreely.Themagnitudeoftheimpactforcewasdeduced
fromtheoutputofanaccelerometerattachedontheironmass.Meanwhile,anotheraccelerometerwassituatedontheoutersurfaceofthetestsectiontomeasurevibrationlevelsduetotheimpactforce.Asanexample,Fig.5(aand(bshowstheoutputsignalsobtainedsimultaneouslybythetwoaccelerometers,respectively.Notethat,foreachoftheimpactevents,aseriesofspikesisseeninthesignaltracesinFig.5(aand(bwhichwerecausedbyre-bouncingmotionsoftheironmassaftertheinitialimpact.Forthepresentanalysisonlytheintensitiesoftheinitialimpactswereofinterest.Thesamplingratefordigitizingtheaccelerometeroutputwas50kHztoensurethatspikevaluesweresampledandrecorded.BasedonthedatashowninFig.5(aand(b,arelationshipbetweentheimpulsiveforcesappliedandthemeasuredvibrationlevelsofthetestsectioncanbeconstructedandisshowninFig.6.Theresponseappearstobealmostlinear.Fig.6alsoindicatesthattheinstan-taneousvibrationlevelofapipestructureduetoexternalforcecanreachhundredsofg,whichisremarkablyhigherthanthelevelofvibrationsnormallyseeninindustrialpipingsystems[1].
Thetwopiezoelectricsensorscorrespondingtotheoriginaldesignandtheimprovedonewereinstalledintwoidenticalvortexshedders,forcomparingtheirsensi-
Fig.3.Aside-viewsketchoftheimprovedsensordesign.

44J.J.Miauetal./FlowMeasurementandInstrumentation11(200041–49
Fig.5.(aTheoutputoftheaccelerometerattachedontheironmass.(bTheoutputoftheaccelerometerattachedontheoutersurfaceofthetestsection.
Fig.6.Theresponsebetweentheimpulsiveforceandthevibrationlevelofthetestsection,wheregdenotesthegravitationalacceleration.
tivitiestotheimpulsiveforcesapplied.First,experi-mentswereperformedforthecasethatthewaterinthepipewasquiescent.TheresultsobtainedfromtheoutputofthechargeamplifierarepresentedinFig.7.Notethattheoutputsignalsshownwereamplifiedwithoutfilteringandthegainofthechargeamplifierwassetsosmallthatitfunctionedproperlyevensubjectedtoimpulsivevibrations.Inthefigure,opencircularsymbolsshowthe
Fig.7.Theresponsemagnitudeofthetwopressuresensorstoimpul-siveforce.᭺denotestheresponseoftheoriginalsensorinwater,¼denotestheresponseoftheoriginalsensorinair,̅denotestheresponseoftheimprovedsensorinwater,̆denotestheresponseoftheimprovedsensorinair.
sensitivityoftheoriginalsensortovariouslevelsofimpulsiveforcesandopentriangularsymbolsshowthesensitivityoftheimprovedsensor.Itisseenthatthesensitivityoftheimprovedsensortoimpulsiveforcesisreducedtoabouthalfthatoftheoriginalsensor.Hence,areductioninsensitivitytostructuralvibrationsduetotheimprovedsensorisquiteevident.
Second,experimentsweremadetostudyperformanceofthetwosensorsaswaterdrainedoutofthepipe.TheresultsarealsoincludedinFig.7bysolidsymbolsfordiscussion.Fortheimprovedsensor,thesensitivitytoimpulsiveforces,shownbythesolidtriangularsymbols,ismuchreducedcomparedwiththesituationofpipefilledwithwaterdescribedabove.Thereasonforthisisthatthissensorbasicallysensespressurefluctuationsofthefluidmedium.Whenthepipeisfilledwithwater,theintensityofpressurefluctuationsinthewaterinducedbyanimpulsiveforceisnotablyhigherthaninthesitu-ationwhenwaterisdrainedoutofthepipe.Fortheorig-inalsensor,denotedbysolidcircularsymbols,thesensi-tivitytoimpulsiveforcesisratherincreasedcomparedwiththesituationwhenthepipeisfilledwithwater.Themainreasonforthisisthatthissensorbasicallysensesstressvariationsofthevortexshedder.Aswaterdrainedoutofthepipe,mostoftheenergyassociatedwiththeimpactforceconvertstostructuralvibrations;whereas,inthesituationofthepipefilledwithwater,someoftheenergyassociatedwiththeimpactforceisabsorbedbythewater.Alsonoted,inthesituationofwaterdrainedoutofthepipe,thesensitivityoftheoriginalsensorto

J.J.Miauetal./FlowMeasurementandInstrumentation11(200041–4945
structuralvibrationsisabout20timeslargerthanthatoftheimprovedsensor.
Formeasuringtheflowrate,thegainofthechargeamplifierwassethigherthanthatforthesignaltracesshowninFig.7,enablingtheoutputsignalstoshowappreciablevariationsinthevortex-sheddingfrequencycomponent.Moreover,inordertoremovenoisecausedbyimpulsiveforcesfromtheoutputsignals,abuilt-inintegratorinthechargeamplifierwasconfiguredasalow-passfilter.Fig.8(ato(cshowtheoutputsignalsoftheoriginalsensorsubjecttothreelevelsofimpulsiveforceonthetestsection,upto11.3kN.Theintensitiesofimpulsiveforcesareindicatedinthefigure.Foreachofthecasesshown,theimpulsiveforceappliedwassoseverethatthechargeamplifierwasapparentlyoutofrangetemporarily.Incontrast,Fig.9showstheperform-anceoftheimprovedsensorunderidenticalimpulsiveforces.ItcanbeseenthatthenoiseduetoimpulsiveforcesappearstobereducedremarkablycomparedwithFig.8.ItshouldbenotedthattheverticalscaleshowninFig.8isoneorderofmagnitudelargerthanthatshowninFig.9.
Toeliminatethenoiseinducedbystructuralvibrationswiththeoriginalsensoremployed,itwasfoundthatinstallingalow-passfilterbetweenthesensorandthechargeamplifiercouldbeeffective.ThiscanbeseenfromacomparisonofthreecasesshowninFig.10.Fig.10(apresentstheoutputoftheoriginalsensorwithoutimpulsiveforceapplied.Fig.10(bpresentstheoutput
Fig.8.Thefilteredsignaltracesoftheoriginalsensorsubjecttodif-ferentvaluesofimpulsiveforce,Re=1.9×104:(aimpulsiveforce=9.7kN,(bimpulsiveforce=10.7kN,(cimpulsiveforce=11.3kN.
Fig.9.Thefilteredsignaltracesoftheimprovedsensorsubjecttodifferentvaluesofimpulsiveforce,Re=1.9×104:(aimpulsiveforce=9.7kN,(bimpulsiveforce=10.7kN,(cimpulsiveforce=11.3kN.
Fig.10.Re=1.9×104:(atheintegratedsignaltracesoftheoriginalsensorwithoutimpulsiveforce,(btheintegratedsignaltracesoftheoriginalsensorwithanimpulsiveforce=13.8kN,(cthesignalobtainedbyapplyinganRClow-passfilterpriortotheamplifierwithanimpulsiveforce=13.8kN.

46J.J.Miauetal./FlowMeasurementandInstrumentation11(200041–49
signalsubjectedtoanimpulsiveforce,13.8kN,withoutinstallationofalow-passfilter.Evidently,thechargeamplifiersaturatestemporarilyattheinstantofforceimpact.Ontheotherhand,Fig.10(cshowsthatwithinstallationofanRClow-passfilterthenoisesduetoimpulsiveforcesarereducedgreatly.Forthiscase,thecut-offfrequencyoftheRClow-passfilter,124Hz,ismuchhigherthanthevortex-sheddingfrequency;there-foretheintensityofthevortex-sheddingfrequencycomponentispreservedthroughfiltering,whichcanbeconfirmedbycomparingFig.10(cand(a.
Insummary,theresultspresentedaboveindicatethatsensitivitytoimpulsivevibrationscouldbereducedwiththeimprovedsensor.Furthermore,installinganRClow-passfilterbetweenthesensorandthechargeamplifiercaneffectivelyremovespike-likenoisesfromtheout-putsignal.
Onthebasisoftheresultsshownabove,itisfurthersuggestedthattheoriginalsensorwithinstallationofanRClow-passfilterinthesignalprocessorisagoodcom-binationforliquidflowmeasurementbecauseoflong-termdurabilityofthesensorwithepoxypackaging.Forairflowmeasurement,theimprovedsensorisrec-ommendedbecauseitismuchmoreinsensitivetostruc-turalvibrations.
3.2.Suppressingvortexsheddingbyrepetitiveimpulsiveforces
Duringtheexperiments,itwasnotedthatwhenasequenceofimpulsiveforceswasappliedtothepipe,thevortex-sheddingfrequencycomponentcouldbesup-pressed,inparticularwhenthefrequencyofrepetitiveforceswashigherthanthevortex-sheddingfrequency.Itwasalsoobservedthattheappearancesofthesignalsobtainedbythetwosensorsemployedwerequitesimi-lar.Hence,onespeculatesthatthisisnotaproblemasso-ciatedwiththesensorcharacteristics,whichcanbeseeninthefollowingfigures.
Figs.11and12presentoutputsignalsoftheoriginalsensorandtheimprovedsensor,respectively.Figs.11(aand12(aprovidetheoutputsignalsobtainedbytheaccelerometersforreference,inwhichthespikessignifytheeventsofrepetitiveimpulsiveforcesapplied.Thespikesappearnotperfectlyperiodicintime,becausetheimpulsiveforceswereproducedbyahand-heldhammer.TherepetitionfrequenciesoftheimpulsiveforcesshowninFigs.11(aand12(aarecomparable,about4Hz,whichisapproximatelytwicethevortex-sheddingfre-quency.ItshouldalsobementionedthattheimpulsiveforcesappliedinthepresentsituationaremuchsmallerinmagnitudethanthoseseeninFigs.8and9,hencethechargeamplifierworkedwellinthepresenceofimpul-siveforces.Outputsignalsofthechargeamplifiercorre-spondingtothetwosensorsemployedareshowninFigs.11(band12(b.Theappearancesareratherirregularin
Fig.11.Asequenceofexternalforceswasappliedonthepipe:(athesignalsobtainedfromanaccelerometersituatedonthepipewall,(bthefilteredsignalsmeasuredbytheimprovedsensor,atRe=1.9×104.
Fig.12.Asequenceofexternalforceswasappliedonthepipe:(athesignalsobtainedfromanaccelerometersituatedonthepipewall,(bthefilteredsignalsmeasuredbytheoriginalsensor,atRe=1.9×104.
time.Variationsinthevortex-sheddingfrequencycomponentarebarelyidentifiableatsomeinstants.Theseobservationsevidencethatsignalqualityisdegradedsoseverelythatitisdifficulttoextractinformationofthevortex-sheddingfrequencyfromthesignalsmeasured.Thisphenomenonwasinvestigatedfurtherbycompar-ingthesignalsandtheflowvisualizationimagesobtainedsimultaneously.Notablefindingsaredescribedbelow.Fig.13providesthesignaltraceobtainedbytheimprovedsensor(solidline,whichistheoutputoftheintegrator,foratimeintervalof10s,whiletherepetitiveforceswereappliedbetweent=1.5sand7.4s.Forcom-parison,theflowvisualizationimagestakenatthetimeinstantsatohmarkedinFig.13arepresentedinFig.14.Twosegmentsofdashedlinesarealsoincludedin

J.J.Miauetal./FlowMeasurementandInstrumentation11(200041–4947
Fig.13.Externalforceswereappliedwithinthemeasuringtimeperiod,atRe=1.28×104;solid-linetracedenotesthefilteredsignal,dashed-linetracedenotestherawsignalmeasuredbeforebeingintegrated.
Fig.14.Pictures(a–(hcorrespondingtotheinstantsindexedasatohinFig.13.
Fig.13,representingthesignalsbeforebeingfedintothetimeintegrator.Comparingthedashedlinesandthesolidlinerevealsthat,whenperiodicvortexsheddingprevailsintheflow,theintegratormerelyproducesaphaseshiftof90°totheinputsignal.
InFig.13,thefirstspikeisseenatt=1.55smarkingthefirsteventofforceimpact.Priortothistimeinstant,vortexsheddingisquiteperiodicintime.Seealsotheflowvisualizationimagestakenatt=0.78s,1.11sand1.25sinFig.14(a,(band(c,respectively.TheimageofFig.14(ashowsthattheseparatedshearlayerappearstobesituatedattheoutermostpositioninthetransversedirection,signifyinganinstantthatavortexisabouttobeshed.Atthismoment,thevortex-sheddingsignalmeasuredbythesensor,indicatedbythedashed-linetraceinFig.13,appearstoreachamaximumvalue.Fig.

48J.J.Miauetal./FlowMeasurementandInstrumentation11(200041–49
14(brepresentsaninstantthattheshearlayerissituatedattheinnermostpositioninthetransversedirection.Cor-respondingly,thedashed-linetraceshowninFig.13indicatesthat,atthisinstant,thepressuresignalmeas-uredreachesaminimumvalue.Physicallyspeaking,atthisinstanttheshearlayerstartstorollupnearthetrai-lingedgeofthevortexshedder.Subsequently,avortexdevelopsthroughaccumulationofvorticityinthesepar-atedshearlayerasshowninFig.14(c,wherethevortexisseentogrowinsizewithtime.TheimagesofFig.14(ato(cshowaseriesofmotionsthattheseparatedshearlayerisflappinginanundulationamplitudeasvor-texsheddingisprevailinginthewake,consequentlythepressurefluctuationsappearedinFig.13.
AsseeninFig.13,theappearanceofperiodicvari-ationsinthesignalgetsdiminishedaftertheimpulsiveforcesareappliedrepetitively.ThisphenomenoncanbeexaminedfurtherwiththeimagesshowninFig.14(dto(g.Aremarkablefeaturenotedfromtheseimagesisthatsmall-scalevorticesaredistinctlypresentintheseparatedshearlayerwhicharebarelyseenintheimagesinFig.14(ato(c.Infact,visuallyonecouldfurtherconceivethateachofthediscretevorticesintheshearlayerwasoriginallydevelopedrightbehindthesharpedgeofthevortexshedderattheinstantofimpulsiveforcebeingapplied.Owingtoasequenceofimpulsiveforcesbeingappliedthediscretevorticeswerepresentintheseparatedshearlayer,whichbehavedlikeindividualeddiestoentrainsurroundingfluidandgrewintosizablescales.Thediscretevorticesdidnotnecessarilyamalga-mateintoacoherentvortex.Theinstantsofthediscretevorticesproducedwerenotedtobeverycrucialtothedevelopmentofasheddingvortex.
TheimagesofFig.14(dto(fshowthattheseparatedshearlayeroriginatedfromthevortexshedderisflappingatasmalleramplitudethanthatseeninFig.14(ato(c.AsnotedinFig.13,theinstantsd,eandfarecompara-bletotheinstantsa,bandcasfarasthephaseofvortexsheddingisconcerned,butsignalvariationsassociatedwiththevortex-sheddingfrequencycomponentappeartobesmalleratthesetimeinstants.
Aftertheexternalforceshaveceased,theprocessofvortexsheddingtakessometimetorelaxtothestatepriortowhentheexternaldisturbanceswereapplied.ThisisrealizedfrombothFigs.13and14.ReferringtoFig.13,aftert=7.4sthedashed-linetracereachesaminimumvalueatthetimeinstanth.Thisinstantisequi-valenttotheinstantb,asfarasthephaseofvortexshed-dingisconcerned.However,adiscrepancybetweentheimagesofFig.14(band(hrevealsthattheseparatedshearlayerseeninFig.14(bappearstobesituatedclosertotheshedder.Hence,alowerpressurevalueatthisinstantisshowninFig.13.
4.Conclusions
Inthisstudy,theperformanceofavortexflowmetersubjectedtoimpulsiveforcesonthepipewasexamined.Itwasfoundthatthesensitivitiesofsensorstoimpulsivevibrationscanbereducedbyimprovingthedesign.Itisalsofoundthathavingalow-passfilterinstalledbetweenthesensorandthechargeamplifierhelpstokeepthechargeamplifierinworkingrangeunderimpulsiveforces,enablingthesignalstobeprocessedaccurately.Thisfindingsuggeststhepossibilityofintegratingalow-passfilterintothesignalprocessor.Finally,suppressionofthevortex-sheddingphenomenonwasnoticedwhilerepetitiveimpulsiveforceswereappliedtothepipe.Flowvisualizationsrevealedthatthesuppressionisduetotheseparatedshearlayeroriginatingfromthevortexshedderflappingirregularlywithsmalleramplitude.Impulsivevibrationsofthevortexshedderpromotetheformationofdiscretevorticesintheseparatedshear-layerthatentrainsurroundingfluidandgrowintosizablescales.Thesediscretevorticesdonotnecessarilyamal-gamateintoacoherentvortex.Hence,theundulationamplitudeassociatedwithflappingoftheseparatedshearlayerisreduced,andsoistheintensityofvortex-shed-dingsignals.Acknowledgements
TheauthorswouldliketoacknowledgethesupportoftheNationalScienceCouncilofthisworkundercontractnumberNSC86-2622-E-006-018.References
[1]Y.Ogawa,N.Matsubara,Newtypeofvortexsheddingflow-meter,in:FLUCOME’85,ProceedingsoftheInternationalSym-posium,Tokyo,Japan,1985,pp.1009–1014.
[2]S.Takahashi,I.Itoh,Intelligentvortexflowmeter,in:FLOME-KO’93,Proceedingsofthe6thInternationalConferenceonFlowMeasurement,Seoul,Korea,1993,pp.107–113.
[3]Y.Tanida,A.Okajima,Y.Watanabe,Stabilityofacircularcylin-deroscillatinginuniformfloworinawake,J.FluidMech.61(4(1973769–784.
[4]P.K.Stansby,Thelock-onofvortexsheddingduetothecross-streamvibrationofcircularcylinderinuniformandshearflow,J.FluidMech.74(4(1976641–665.
[5]C.Barbi,D.P.Favier,C.A.Marsca,D.P.Telionis,Vortexshed-dingandlock-onofacircularcylinderinoscillatoryflow,J.FluidMech.170(1986527–544.
[6]B.J.Armstrong,F.H.Brane,I.Grant,Theeffectofaperturbation
ontheflowoverabluffcylinder,Phys.Fluidsvol.29(7(19862095–2102.
[7]R.C.Mottram,B.Robati,Comparisonofpulsationeffectsonvor-texandorificeflowmeter,in:InM.Harada(Ed.,FluidControlandMeasurement,vol.2,PergamonPress,Oxford,1986,pp.957–963.
[8]M.C.A.M.Peter,E.vanBokhorst,C.H.L.Limpens,Impactof
pulsationsonvortexflowmeter,in:FLOMEKO’98,Proceedings

J.J.Miauetal./FlowMeasurementandInstrumentation11(200041–4949
ofthe9thInternationalConferenceonFlowMeasurement,Lund,Sweden,1998,pp.185–190.
[9]M.Schumm,E.Berger,P.A.Monkewitz,Self-excitedoscil-lationsinthewakeoftwo-dimensionalbluffbodiesandtheircontrol,J.FluidMech.271(199417–53.
[10]F.B.Hsiao,J.Y.Shyu,Influenceofinternalacousticexcitationuponflowpassingacircularcylinder,J.FluidsStruct.5(1991427–442.
[11]J.J.Miau,Y.P.Li,J.H.Chou,Y.C.Huang,C.C.Yang,Integration
ofvortexshedderandsensorforavortexflowmeter,in:FLOME-KO’96,Proceedingsofthe8thInternationalConferenceonFlowMeasurementStandardPressofChina,Beijing,1996,pp.89–94.

本文来源：https://www.2haoxitong.net/k/doc/aba37b42a76e58fafab003b2.html

《Response of a vortex flowmeter to impulsive vibrations.doc》
将本文的Word文档下载到电脑，方便收藏和打印
推荐度：
点击下载文档

文档为doc格式
相
关
案
例

相关推荐

推荐内容

C语言程序设计--记忆力测试试题及注释 16秋《英语语法》作业2 压力容器钢板厚度负偏差 空间位置关系的判断与证明.板块三.平行关系的判断与证明.学生版 中国水貂制品行业市场前景分析预测年度报告(目录) 2014年福建省莆田市教师资格每日一练(5.12) 小学三年级日记范文300字【6篇】 肛肠病术后并发尿潴留原因分析和预防 观看“三个以案”警示教育片《罪途》心得体会 小学校园安全管理实施方案
网站内容来自网络，如有侵权请联系我们，立即删除！ Copyright © 范文写作网京ICP备16605803号