ExperimentalevaluationofabilityofRelap5,Drako®,Flowmaster2™andprogramusingunsteadywallfrictionmodeltocalculatewaterhammerloadingsonpipelinesOriginalResearchArticle

Mechanicalloadingsonpipesystemscausedbywaterhammer(hydraulictransients)belongtothemostimportantandmostdifficulttocalculatedesignloadingsinnuclearpowerplants.ThemostcommonprocedureinSwedenistocalculatethewaterhammerloadingsonpipesegments,accordingtotheclassicalone-dimensional(1D)theoryofliquidtransientflowinapipeline,andthentransfertheresultstostrengthanalysesofpipelinestructure.Thisprocedureassumesthatthereisquasi-steadyrespondofthepipelinestructuretopressuresurges—nodynamicinteractionbetweenthefluidandthepipelineconstruction.Thehydraulicloadingsarecalculatedwith1Dso-called“network”programs.CommonlyusedinSwedenareRelap5,DrakoandFlowmaster2—allusingquasi-steadywallfrictionmodel.AsathirdpartyaccreditedinspectionbodyInspectaNuclearABreviewscalculationsofwaterhammerloadings.ThepresentedworkshallbeseenasanattempttoillustrateabilityofRelap5,Flowmaster2andDrakoprogramstocalculatethewaterhammerloadings.AspecialattentionwaspaidtousingofRelap5forcalculationofwaterhammerpressuresurgesandforces(includingsomeaspectsofinfluenceofCourantnumberonthecalculationresults)andalsotheimportanceofconsideringthedynamic(orunsteady)frictionmodels.Thecalculationsarecomparedwithexperimentalresults.TheexperimentshavebeenconductedatatestrigdesignedandconstructedattheSzewalskiInstituteofFluidFlowMachineryofthePolishAcademyofSciences(IMPPAN)inGdansk,Poland.TheanalysesshowquitesmalldifferencesbetweenpressuresandforcescalculatedwithRelap5,Flowmaster2andDrako(thedifferencesregardmainlydampingofpressurewaves).Thecomparisonofcalculatedandmeasuredpressuresandalsoaforceactingonapre-definedpipesegmentshowssignificantdifferences.Itisshownthatthedifferencescanbereducedbyusingunsteadyfrictionmodelsincalculations.Recently,suchmodelshavebeensubjectsofworksofseveralresearchesintheworld.