ML17256A708

App C to Structural Reanalysis Program, Utility Pole Tornado Missile Trajectory Analysis.

ML17256A708
Person / Time
Site:	Ginna
Issue date:	03/08/1983
From:	DUNN W L, TWISDALE L A APPLIED RESEARCH ASSOCIATES, INC.
To:
Shared Package
ML17256A706	List: ML17256A705 ML17256A707 ML17256A708 ML17256A709 ML17309A300
References
44T-2488-1, C556, NUDOCS 8305310157
Download: ML17256A708 (33)

Text

APPENDIXCtotheSTRUCTURAL REANALYSIS PROGRAMForTheR.EGINNANUCLEARPOWERPLANT8305310157 830519PDRADOCK05000244PPDR March8,1983ARAReportC556RTIReport44T-2488-1 UTILITYPOLETORNADOMISSILETRAJECTORY ANALYSISL.A.TwisdaleW.L.DunnAppliedResearchAssociates, Inc.Southeast Division4917Professional CourtRaleigh,NorthCarolina27609PreparedforResearchTriangleInstitute P.O.Box12194ResearchTrianglePark,NorthCarolina27709andGilbertAssociates, Inc.P.O.Box1498Reading,Pennsylvania 19603UnderPurchaseOrder201494 TABLEOFCONTENTSl.INTRODUCTION

.2.APPROACH.a.TornadoWindfield

...................b.Trajectory Modelc.MissileAerodynamics d.Injection ModelPage53.TRAJECTORY SIMULATION RESULTS...............

84.COMPARISONS TOOTHERWORKANDFIELDOBSERVATIONS......

13a.6-DTrajectory ModelPredictions

...........

14b.3-DBallistic Model..................

14c.UtilityPoleTransport:

Xenia,OhioTornado.....15d.StoredUtilityPoles:Brandenburg, Kentucky.....165~CONCLUSIONS~~~~~~'~~~~~~~~~~'~~~~~~18REFERENCES 19

-UTILITYPOLETORNADOMISSILETRAJECTORY ANALYSIS1.INTRODUCTION Theobjective ofthisstudywastoperformtrajectory calculations ofutilitypolemissilesintornadowindfields.

TheNuclearRegulatory Coomission (NRC)-defined utilitypole[1]wasspecified asthepostulated missilefortheseanalyses.

Tornadowindfields withpeakvelocities of132,150,and188mphwerespecified.

Theutilitypoleswereinjectedat20ftabovegrade,whichcorresponds approximately tothecenterofmassofastanding35ftpole.Trajectory calculations weremadeusingtherandom-orientation six-degree-of-freedom (R06-D)trajectory model[2,3,4g,whichaccountsfordrag,lift,andsideaerodynamic forcesinatimehistoryintegration oftheequations ofmotion.Themaximumheight,range,andspeedattainedbythemissileswereextracted fromthetime-history flightdata.Inadditiontothesenumerical calculations, severalcomparisons offieldobservations andtrajectory predictions ofutilitypolemissileshavebeenmade.Thisreportdocuments themethodsandresultsofthisstudy.2.APPROACHTheapproachusedforthetrajectory analysesisbasedprimarily onthemodelsanddatareportedinRefs.2-10.Abriefsummaryofthetornadowindfield model,traj'ectory model,missileaerodynamics, andinjection modelarepresented inthefollowing paragraphs.

a.TornadoWindfield

,Thetornadowindfield modelusedhereinisdocumented indetailinRef.4.Thissynthesized windfield modelwasdeveloped explicitly formissiletransport analysisandincludes5basicparameters thatdefinethe3-dimensional flowcharacteristics giventhepeakspeedUmaxandpathwidth

-I-Mt.Theseparameters are:translational speed(UT);theratioofradialtotangential flowcomponents (y);theradiustomaximumwindspeed (pm);coreslope(S);andreference boundarylargerthickness (6o).Asensitivity analysiswas,performed I4,6]usingaone-at-a-time experimental designandthreelevelinputpattern.Thebasicconclusions ofthisanalysisfortornadoes withUax=300mphandtwotypesofmissileswere:(1)Forgiventornadointensity, thenumberofmissilesgenerated andtheirtransport characteristics aremostsensitive tothetranslational speed(UT)ofthestorm.LowvaluesofUTresultinmoremissileinjections andhighermissilevelocities forspecified Umax.(2)Forgiventornadointensity, anincreaseintheradialinflowcomponent relativetothetangential component increases thenumberofmissilesinjectedandleadstohigheraveragevaluesofmaximumvelocities, ranges,andaltitudes.

(3)Missilesinjectedandtransported bylarge-core tornadoes generally attainhighermaximumvelocities butlowerpeakaltitudes thanthosepredicted withsmallerpm.Theabsolutenumbersofmissilesproducedareproportional totheradiusofthecore.(4)Theslopeofthecoredoesnothaveanappreciable effectonmissiletransport, evenformissilesinjectedathighelevations.

(5)Relatively smallvariations inairdensitycanproduceproportional changesinmissilerange,buttheeffectofairdensity(duetoentrained dust,etc.)onmaximumvelocities isheavilydependent onthemissileinjection height.Hence,fromthisanalysiswehaveabetterunderstanding ofhowtocharacterize thetornadowindfield (giventhepeakwindspeed) tomaximizemissiletransport parameters.

Asecondsensitivity analysiswasmadetoassesstheimportance ofcertainuniqueflowcharacteristics thatexistinseveralprominent tornado models.Toevaluateexplicitly theeffectsofbasicdifferences among'indfield definitions, apairwisecomparison studywasperformed L4,6]withthesynthesized modelinaseriesofmatchedcomparisons withothermodels.Fromtheresultsofthefirstphaseofthesensitivity

analysis, themoreimportant variables inthesynthesized windfield wereidentified asUT,y,andpm.Inthisphase,thepairwisemodelcomparisons weremadewiththeUT,y,andpmvaluesinthesynthesized modelmatchedtotherespective valuesusedinthewindfield modelselectedforcomparison.

Threemodelswereselectedonthebasisoftheirdistingui shingfeaturesrelativetothesynthesized modelandrecentness ofdevelopment:

theFujitaDBT-77tornadomodel[11],theFujitasuctionvortexDBT-78flip,andtheTRWPhaseIIImodelL12].Theresultsofthecomparative missiletransport analysisindicateinsignificant differences formostofthevelocityandrangestatistics.

FortheFujitaDBT-77comparisons, thesynthesized modelinjectsmoremissileswithhighermeanvaluesofmaximumvelocities, whereastheFujitamodelpredictsslightlyhighervariances andextremevalues.Thecomparison dataexhibitdifferences thataremuchlessthanthoseobtainedfromvariations inUTandyforthesynthesized modelalone.Forthesuctionvortexmodel,anumberofsimulations weremadewithsingleandmultiplesuctionvorticeswithboththepipeandautomissiles.

Theresultsindicatethatthemissilegeneration andimpactpositions areinfluenced byembeddedvortices.

However,forthesamereference windspeed intensity, atornadowithnosuctionvorticesyieldshighermissiletransport characteristics whencomparedtoasysteminwhichthesamemaximumwindsoccurinthefast-moving embeddedvortices.

Thus,forconservative l

-predictions ofmissiletransport, thereisnoneedtomodelsuctionvorticesformissi1etrajectory analysis.FortheTRWmodel,moreinjections resultforthepipemissile,butthesynthesized modelpredictshighervelocityandrangestatistics.

Ingeneral,thedatasuggestthatthetransport differences inthemodels,withthesameUma,p,andy,arelimitedprimarily tolowinjection heights.TheTRWmodelgenerally dominates atz=10ft,andthesynthesized modelat33ftwithsimilartransport statistics overthecombinedelevations.

Onthebasisofthesesensitivity studiesandtheresulting updatingoftheUT,y,andpparameters inRef.4,thesynthesized windmodelprovidesatestedwindfield modelfor'utilitypoletransport calculations.

b.Trajectory ModelrTrajectory modelsthathavebeenusedintornadomissiletrajectory analysesinclude:(1)theballistic 3-Dmodel,whichassumesaconstantdragforceandneglectsliftandsideforces;(2)therandomorientation, 6-D(R06-D)model,inwhichaerodynamic drag,lift,andsideforcesaredependent onmissileorientation, whichisperiodically updated;and(3)theconventional 6-0model,whichtracksmissiletranslation androtationusingasystemof6coupleddifferential equations.

Detaileddiscussion andcomparisons ofthesemodelsarepresented inRef.4.Trajectory comparisons ofthesemodelshavebeenmadeusingutilitypolemissilesL2],.12-inpipeandautomobile missiles[43.Onthebasisofthesecomparisons, theballistic 3-Dmodelhasbeenshowntounderpredict

velocity, lift,andrangecharacteristics.

TheR06-Dmodelprovidespredictions thattendtoboundthoseofthe6-Dmodelanditisconsiderably morecomputationally efficient.

c.MissileAerodynamics Amodeloftheaerodynamic coefficients forageneralclassofmissileswasdeveloped inRef.2andlaterupdated[4)toreflectnewaerodynamic databasedonfullandsubscaletests(12,133.Amodifiedcrossflowtechnique hasbeendeveloped topredictdrag,lift,andsideforcecoefficients asafunctionofangleofattackandrollangle,giventhedragforcecoefficients fortheobjectinflownormaltothemajorbodyaxes.Table1summarizes themodelforcylindral missilessuchastheutilitypole.Aplotofthemodelpredicted vs.utilitypolewindtunneldata[13]for3different RenumbersisshowninFigure1.Theseresultsindicatecloseagreement betweentheaerodynamic modelandmeasuredcoefficients.

d.Injection ModelAsthetornadowindfield passesoveranobject,thedynamicpressureinducesaerodynamic forcesthataredependent onthemissileshape,orientation, surfaceroughness, andproximity tootherobjectsandsurfaces.

Iftheseaerodynamic forcesaregreaterthantherestraining forces,suchasgravity,slidingfriction, andfoundation embedment, theobjectwillbedisplaced bythewindfield.

Ingeneral,theseaerodynamic forceswillnotactthroughthecenterofmassofthebodyandthemissiletumblesandinteracts withthegroundandotherobjectsduringthisinjection phase.Hence,detailedmodelingofinjection requiresinformation onrestraining forcetime'histories andinteraction modelstosimulatemissilecollisions.

Inviewofthecomplexities ofmissileinjection, tornadomissiletrajectory analysesgenerally treatinjection parametrically throughthespecification oftheinitialconditions ofthemissileattheinstantitisreleasedtothetornado.Oncereleased, themissileisassumedtobeactedon ll

'TABLEl.AERODYNAMIC COEFFICIENTS FORCYLINDRICAL MISSILESGeometrical shapeMissiletypeMissilesetnumbers[2]Axialdragcoefficient, CDaSkinfrictioncorrection, fCross-flow coefficient, CDcRightcircularcylinderRods,pipes,poles1,2,3,41.16Solid(rods,poles)0.812Hollow(pipes)1,L/dc10.724+0.276e2(L/d-1)1<L/d<4'0.681+0.0108L/d,L/del.25,(subcritica1)1.80.85[1.9--a],(supercritical)

Aspect-ratio correction, k1-8(d/L),d/LC.02058q042e-(0.51

+5.6(d/L-0.02))d/L>0.02Dragcoefficient, CDmd4La--.CDf)cosa(+CDcksinaLiftcoefficient, CLSidecoefficient, CSRef.area,A1fd---CDafcosa[cose)sinu+4LCDckcosasjn2aLd 0lI Table1oooo44oo<<IW~4I.<<4I.NoRef.(13)p~<<r<<<<<<<<<<<<<<w(a)DragCoefficient Table144oo4oo4ooo<<~l04E.S04lgloLwooRef.(13)oo~<<<<<<<<<<<<<<<<~y<<4(b)LiftCoefficient Figure1.Aerodynamic Coefficients forUtilityPoleHissile

!onlybygravityandaerodynamic forces.Amissileinjection methodology wasdeveloped inRef.[2]andsubsequently refined[4]toconservatively account"forthecomplexities anduncertainties inaparametric injection model.Theapproachinvolvesatwo-stepprocedure:

(1)theverticalandhorizontal aerodynamic forcetimehistories onthemissilearecalculated asfunctionoftornadoposition, and(2)thepositionofthetornadocorresponding topeakaerodynamic forcesarethendetermined.

Thispositiondefinesthetimeofreleaseofthemissilewithrespecttothemovingwindfield.

Injection studieshaveshownthatthismethodprovidesforoptimummissiletransport andtendstoresultinmissiletrajectories thatboundthosedocumented infieldobservations.

Thisoptimumreleasecriterion isusedhereinintheutilitypoletrajectory analysis.

!-:3.TRAJECTORY SIMULATION RESULTSUsingthemodelspreviously described, trajectory calculations havebeenmadefortheutilitypolemissle[1].Thepostulated missileis35ftlongwithadiameterof13.5inchesandweightof1,122lbs.Thecenterofmassofthemissileispositioned at20ftabovegrade.PeakUmaxwindspeeds of132,150,and188mphareconsidered.

Giventhesepeakwindspeeds, theremaining tornadoparameters havebeendefinedfromtheinformation inRef.4.Amediancasewindfield, corresponding tothemeansofthedistributions onUT,y,pm,S,and6fortherespective intensity levelhasbeenspecified, asnotedinTable2.Forexample,fora150mphtornado,thedistribution ontranlational speedisassumedinRef.4tobetruncated normal,rangingfrom5to55mphwithameanof35mphandastandarddeviation of11mph.Hence,forthemediancase,UTisassignedavalueof35mph.Inaddition, amoreseveresetofparameters hasbeendefinedusingtheresultsofthepreviously reported

TASLE2.TORNADOWINDFIELD PARAMETERS

(-.Parameter CaseParameter ValuesForEachWindfield Um=132mphUm=150mphUm=188mphTranslational MedianSpeed,UT(mph)2o30535134520RadialInflow,yMedian260.71.1,0.71.10.71.1Rax(<<)Median20375200375200500300Median200.1500.1500.150I-s(ft)Median20450500450500450500 IIsensitivity analysis[4,6].Thissetisdenotedasthe2acaseinTable2sinceeachparameter hasbeensetatitsp+2a(ory-2adepending onthesignthatmaximizes missiletransport).

Thus,sincelowUT,highY,arelowpmaximizemissiletransport (givenU),theseparameters aresetmax'espectively atp-2a,p+2o,andp-2o,respectively.

This2vcasewasincludedtostudytheinfluence ofvariations inthethreedimensional windfield onthetrajectory oftheutilitypolemissile.Theycorrespond toaboutthe95percentile ofeachrespective distribution.

Thetornadoispositioned(see Fig.2)relativetothemissileatthat'offsetpositionthatcorresponds tothepeakwindswithinthetornado.AsnotedinRefs.L3,4],thisoffsetpositionisgivenbypcos(tany).Themissileisreleasedtothemovingtornadoatpeakaerodynamic force.and.theequations ofmotionarenumerically integrated totrackthemotiontimehistoryofthemissile.Drag'ndliftforces(theradiallysymmetric utilitypolehasnosideforce)arecalculated usingthecross-flow equations inTable1.Themissileistrackeduntilthecenterofmassofthepolereachesgroundelevation (z=0).Thehorizontal distancetraveleduntilthecenterofmassofthepolefallsfromz=20fttoz=0isdefinedastherangeofthetrajectory.

Table3summarizes theresultsofthesetransport simulations forboththe.medianand2awindfields.

Forthe132mphtornado,theutilitypoledoesnotliftforanyoftheorientations considered.

Thepeakaerodynamic forceatinjection isabout1,800,lbsandisdirectedhorizontally fortheverticalandhorizontal poleorientations.

Whenthepoleispitchedintowind(orientations 3and4),thepeakverticalinjection forceisabout700lbs.Hence,thepoledropsassoonasitisreleased.

Forthe132mphtornado,weestimateapeakrangeofabout34ftandapeakvelocityof37mph.10 0

PlantTargetArea~mo(X;,Y;,Z;)MissiIeInitiaIPositionfi,gure2.TrackLengthandOffsetCoordinate SystemforMissileInjection Model11 MI'.'lTABLE3.TRANSPORT CHARACTERISTICS OFUTILITYPOLEMISSILESimulation Parameters Transport Characteristics ByPeakNndspeed(mph)CaseDescription HissileOrientation AAA(x,y,z)IHaximumHeight(ft)2HaximumRange(ft)3HaximumVelocity(mph)Um~132Um~150Um~188Um~132Um~150U'm~188Q~132UmI150Um~188Hedianllindfield:(UT~YPm,60~5atmidvalue)Vertical(0,0,1)Horizontal (0,1,0)45X(-0.71,0,0.71) 45XY(-0.5,-0.5,0.71) 2020202020202020202020.1202726232735342941495411188243330304040323650536457Hedian+2oWndfield:

(UT,Y,P,ao,satu+2o)Vertical(0,0,1)Horizontal (0,1,0)45'(-0.71,0,0.71) 45'Y(-0.5,-0.5,0.71) 202020202020202020202221273127343641345854571549435-4335383735364555577658Heightabovegrade;missileinjectedat20ft.lRangeinX-Yplaneuntilgroundimpact.3Usuallyoccursatgroundimpact.

!--Theresultsforthe150mphtornadoaresimilartothoseforthe132mphtornadointhattheaerodynamic forcesarenotsufficient toliftthepole.Thepeakaerodynamic forceatinjection isabout2,200lbs,ofwhichabout900lbsactvertically forthefavorable orientations.

Sincethepoleweighs1,122lbs,itaccelerates downwarduponrelease.Maximumpredicted rangeandspeedare41"ftand45mph,respectively.

Forthe188mphtornado,thewindspeeds produceverticalaerodynamic forcesthatexceedtheweightofthemissile(fororientations 3and4),whichproducesliftatinjection.

Themaximumliftisabout2ft(from20to22ftabovegrade),amodestamountthatisconsistent withthepeakverticalaerodynamic forceofabout1,400lbs.Thisliftresultsinamuchlongerrange(upto154ft),andimpactvelocity(76mph)sincetheobjectissustained inthewindsabouttwiceaslongasbefore(2secvs1sec).Afewadditional simulations weremadewiththepolepositioned atzerooffsetforthecasesgiveninTable3.Theresultsshowreducedtransport characteristics whencomparedtothevaluesinTable3.Itisnotedthatotheroffsetsmightresultintransport thatcouldapproachorslightlyexceedthoseinTable3.However,previousstudieswithratherdenseinjection gridshaveshownthattheoffsetpmcos(tan-1.y)generally providesaccurateestimates ofpeaktransport parameters.

4.COMPARISONS, TOOTHERWORKANDFIELDOBSERVATIONS Thepreviousresultsindicateverylittleliftandtransport lessthan200ftforutilitypoletypemissilesinjectedintornadoes withpeakwindspeeds upto188mph.Thesepredictions canbecomparedtoothercalculations andfieldobservations.

Theavailable comparisons generally 13 correspond tohigherwindspeeds, butwillnevertheless providesomebasisforjudgingtheseresults.a.6-DTrajectory ModelPredictions Redmannetal.513)simulated thetrajectories ofutilitypolemissilesin255mphtornadoes.

Withtheutilitypoleat20ftelevation pitchedintothewindata45degreeangle,themissileliftedtoamaximumheightof40ftduringa339ftflightandimpactedthegroundat113mph.Foraninitialangularvelocityof10rpm,whichisamorerealistic injection condition, thepoleliftedonly4fttoanelevation of24ftandlandedat91mphwitha140ftrange.Theseresultstendtosupportthetrendestablished inTable3.At188mph,wenotedthattheverticalaerodynamic forceswerebeginning toexceedmissileweightandsomeliftwasnoted.At255mph,onewouldexpectthemissiletoliftsubstantially highersincetheaerodynamic forceswouldbeabout(255/188)2

='1.8timesgreater.AsafurthertestoftheR06-Dmodelusedinthedevelopment ofTable3,10simulations weremadeusinga255mphtornadowithinitialorientation 3,similartothatreportedinRef.13.TheR06-Dmodelpredictsanaveragemaximumheightof43feetwithrangesuptoabout500ftandimpactspeedsupto133mph.Theseresultstendtoboundthe6-Dmodelpredictions, andaresimilartopreviouscomparisons ofthe6-DandR06-DmodelsI2,4].b.3-DBallistic ModelSimiuandCordesf14]usedthesimplified ballistic 3-Dmodelforcalculation ofmaximumhorizontal missilespeeds.Forthe35ftutilitypole,theypredictpeakhorizontal speedof60mphina240mphtornadowhentheutilitypoleisinjectedat131ftelevation.

Theseresultsareclearly

unconservative whencomparedtothe6-DandR06-Dmodelpredictions (forzo=20ft)presented previously andraisequestions regarding theadequacyofthe3-Dballistic modelforslenderbodyshapes.~~I.:c.UtilityPoleTransport:

Xenia,OhioTornadoVMcDonald$15]andMehtaetal.I16]reportthetransport ofautilitypoleintheXenia,Ohio,tornadoofApril3,1974.Thepolefailed2ftabovethegroundandwastransported atotaldistanceof160ftbyF5tornadowindsestimated atabout'250 mph.Thepolewas10in.indiameterand25.5ftinlength.Mehtaetal.[16]notethattheotherutilitypolesthathadfailedatthislocationinthetornadopathwerefoundwithin10to15ftoftheiroriginalpositions.

TheR06-Dtrajectory modelandpeakaerodynamic forceinjection modelhasbeentestedagainstthesefieldobservations

[4g.TosimilatetheF5tornadowindspeeds, a250mphtornadowithpm=500ft,y=0.7,andUT=40mphwasusedastheinputtothetransport model.The250-mphintensity at33ftisbaseduponTwisdale's

[2,8]estimated midrangeofF'5storms.Theutilitypolewaspositioned atzo=15fttocorrespond totheinitialheightofthecenterofmassabovethegroundplane.Simulations withaninitialverticalorientation ofthepoleresultinmaximumpredicted transport rangeslessthan53ft.Usingfavorable orientations toaccountforinitialrepositioning afterthepolefails,maximumtransport rangesof283,651,and161ftarepredicted formissileoffsetsontherightsideofthetornadocenter.Amaximumimpactvelocityofabout140mphispredicted.

Fora200-mphtornadowithpmo=300ft,thepredicted rangesare62,121,and91ftforoffsetsof100,150,and200ft,andthemaximumimpactvelocityis63mph.Theresultssuggestthat,withafavorable initialmissileorientation, windspeeds inthe15

I

200to250mphintervalcouldhaveproducedtheobserved160-fttransport range.Thefactthatmanyofthefailedpoleswerenotsignificantly displaced

confirms, thepredictions oftheR06-Dtrajectory analysiswithunfavorable initialorientations andtendstosupporttheuseoftheR06-Dtransport model.d.StoredUtilityPoles:Brandenbur, KentuckTheBrandenburg tornadoofApril3,1974,witharatedintensity ofF5passeddirectlythroughthestorageyardoftheRuralElectricCooperative.

HcDonald[15]andMehtaetal.[16)presentphotographs thatdocumenttheeffectsofthestormonvariousobjectsinitspath.Ofparticular interestisanumberof8-in.-diameter by20-ft-long utilitypolesthatwerestoredhorizontally onarackabout5ftabovethegroundelevation.

Thepolesweredisplaced fromtherack,butnoneweretransported significantly.

TheeffectsofanF'5tornadoontheseobjectshasbeensimulated

[4gwiththeuseofboththeinitialhorizontal storageconditions ofthemissilesandafavorable initialorientation.

A250-mphtornadowasassumedinbothcases.Forthehorizontal injection mode,themodelpredictsthatthepolesmayexperience totalforcesthatapproach1,400lb,butonlyasmallfractionofthisisdirectedvertically.

Hence,thepredicted trajectories areparabolic, andthepolesdroptothegroundwithin30ftoftherack,asindicated inFigure3.Thistransport comparescloselytothepost-tornado observations ofthemajorityofthepoles.Favorable initialmissileorientations werealsosimulated thatcouldhaveresultedforseveralmissilesinthestackastheyinteracted duringtheirinitialresponsetothetornadicwinds.Transport rangesbetween50and273ftresultedforthesesimulations, depending upontheexactorientation 16

Uma=250mph5ft//Xw/Predicted Range~30ft(a)Horizontal InitialOrientation Umax250mph32ftPredicted Range<273ft(b)Favorable RandomInitialOrientation Figure3.Predicted RangesforBrandenburg UtilityPoles17 CJandtheoffsetfromthetornadocenter.Themaximumheightandrangepredicted are32feetand273ft,respectively, asnotedinFigure2.Thepeakvelocityattainedbythepolewas95mph.Itisnotedthatsimulations withUmax=200mphpredictrangesof13ftand100ft,respectively, forthehorizontal andrandominitialorientations.

TheBrandenburg polesaresignificant inthesensethattheobjectsresponded tothewihdfield, butfew,ifany,alignedfavorably tobeliftedbythewindsandhencetobetransported thedistancethatwaspredicted forfavorable initialorientation.

5.CONCLUSIONS Onthebasisofthisbriefstudy,thefollowing conclusions aremade:'2.Tornadoes withpeakwindspeeds of132and150mphgenerateaerodynamic liftforcesthatarelessthantheweightofthepostulated utilitypolemissile.Estimates ofthepeakrange,foraninjection height(heightabovegradeofpolecenterofmass)of20ftare27ftand41ft,respectively, forthe132and150mphwindfields.

Rangeisdefinedasthehorizontal distancetraveledasthepolecenterofmassfallsfromz=20fttoz=0.Maximumvelocities areestimates atabout37and45mph,respectively, forthe132and150mphtornadoes.

Tornadoes withpeakwindspeeds of188mphcanproduceabouta2ftlift(from20ftinitialelevation),

arangeupto154ft,andapeakvelocityofabout76mphfortheutilitypolemissile.These.conditions occurwhenthepoleispitchedintothewindatabouta45degreeangleandreleasedatpeakaerodynamic force.Theseareidealized andveryconservative releaseconditions thatwouldbedifficult toduplicate inanactualtornadostrike.3.Maximumheightattainedduringatumblingwind-borne transport forUmax<188mphbyanypartofa35ftutilitypolewouldprobablynotexceed35to40ft.Otherinjection modes,suchasaramp-type injection couldproduceupwardricochetofahorizontally-translating pole.However,thisinjection modewouldrequire"ideal"missileoriginposition, terrain,andtargetconfiguration inordertoposearealistic threattoelevatedtargets.Theseresultsareconservative inthemissileinjection releasecriterion, definition ofrangeandpeakvelocity, positioning ofthemissilerelativetothewindfield, andthewindfield flowcharacteristics.

Forexample,thetipofthepolewillstrikethegroundasthepolebeginstodropandthis18 interaction willreducethehorizontal momentum.

Hence,theestimates ofpeakrangeandhorizontal velocityareveryconservative.

Inaddition, theweightofthepoleislessthanthe1,490lbsusedinsometornadomissilecalculations=

[e.g.,17].Usingthe1,490lbweight,themaximum154ftrangeinthep+2cr188mphtornado(seeTable3)reducestolessthan100ft.Ingeneral,fieldobservations donotconfirmsignificant utilitypoletransport forthewindspeeds considered herein.OurbestestimateoftypicalutilitypoleresponseforUmax<188mphwouldbeatrajectory rangefrom0to50ftwithhorizontal missilespeedsapproaching 50-60mph.REFERENCES l.U.S.NuclearRegulatory Commission, StandardReviewPlan,"Missiles Generated byNaturalPhenomena,"

Section...,Washington, D.C.,November1975.2.Twisdale, L.A.,etal.,"TornadoMissileRiskAnalysis,"

ElectricPower,ResearchInstitute, PaloAlto,California, NP-768andNP-769,May1978.!3.Twisdale, L.A.,Dunn.W.L.,andDavis,T.L.,"TornadoMissileTransport Analysis,"

NuclearEnineerinandDesin,52,1979,pp.296-308.4.Twisdale, L.A.,andDunn,W.L.,"TornadoMissileSimulation onDesignMethodology,"

EPRINP-2005,ElectricPowerResearchInstitute, PaloAlto,California, August1981.5.Dunn,W.L.,andTwisdale, L.A.,"ASynthesized Windfield ModelforTornadoMissileTransport,"

NuclearEnineerinandDesin,52,1979,pp134-144.6.Twisdale, L.A.,"AnAssessment ofTornadoWindfield Characteristics forMissileLoadingPrediction,"

Proceedin s,FourthU.S.NationalConference onWindEngineering esearch,University ofWashington, Seattle,Washington, July1981.7,.Twisdale, L.A.,"ARisk-Based DesignAgainstTornadoMissiles,"

Proceedin softheThirdASCESpecialty Conference onStructural DesignoucearPlantFacilities, Boston,Massachusetts; April1979.8.Twisdale, L.A.,"TornadoDataCharacterization andWindspeed Risk,"JournaloftheStructural

Division, Proceedings ASCE,Vol.104,No.ctober8.9.Twisdale, L.A.,"Regional TornadoDataHaseandErrorAnalysis,"

Prerints,AMS12thConference onSevereLocalStorms,SanAntonio,exas,anuary1982.19 10.Twisdale, L.A.,andDunn,W.L.,Probabilistic AnalysisofTornadoWindRisks,"JournalofStructural Enineerin,Vol.109,No.2,February1983.12.13.14.Fujita,T.T.,"Workbook ofTornadoes andHighWinds,"SMRP165,University ofChicago,Chicago,Illinois, September 1978.Redmann,G.H.,etal.,"WindFieldandTrajectory ModelsforTornado-Propelled Objects,"

ElectricPowerResearchInstitute, PaloAlto,California, Draft,1980.Redmann,G.H.,etal.,"WindFieldandTrajectory ModelsforTornado-Propelled Objects,"

EPRINP-748,ElectricPowerResearchInstitute, PaloAlto,California, May1978.Simiu,E.,andCordes,M.,"TornadoBorneMissileSpeeds,"NBSIR-76-10-50, NationalBureauofStandards, Washington, D.C.,April1976.15.McDonald, J.R.,"Tornado-Generated MissilesandTheirEffects,"

Proceedings oftheSymposium onTornadoes, Texas,TechUniversity, Lubbock,Texas,June1976.16.Mehta,K.C.,etal.,"Engineering AspectsoftheTornadoes ofApril3-4,1974,"Committee onNationalDisasters, NationalAcademyofsciences, 1976."SafetyRelatedSiteParameters forNuclearPowerPlants,"WASH-1361, U.S.AtomicEnergyComission, Directorate ofLicensing, OfficeofSafety,Washington, D.C.,January1975.20 4tI