Asmentionedearlier,time-varyingcomponentofthewind-inducedloadingismodeledbytheWISPERloadsignal.Whilescalingthisloadsignal(whosevaluesrangebetween1and64withthelevelof25correspondingtoazeroload),theWISPERmean-valuewasassumedtocorrespondtothepreviouslycomputedwindmeanspeedof7.6m/s.
WhentheprocedureoutlinedinSection2.3wasapplied,thefatiguelifeofthebaselineHAWT-bladewasestimatedas32.8years.InFig.12(a),theelementswhichcontrolthefatiguelifeofthebladeareidenti ed.Ascouldhavebeenexpected,theseelementsarelocatedintheairfoil/roottransitionregionwhichissubjectedtothehighestin-serviceloads.
3.2HAWT-BladeDesignModi cations
Thegeometrical-/meshed-modelgeneratorprogramandthestructuralandfatigue-lifeassessmentanalysesdevelopedinthisworkareatalevelthattheycanbereadilyincorporatedintoadesign-optimizationalgorithm.Thiswillbedoneinourfuturecommunication,andthesamedesign-optimizationmethodol-ogyaspresentedinourrecentwork(Ref15)willbeused.Inthissection,however,acoupleofexampleswillbeshowninordertodemonstratehowfewminorchangesintheHAWT-bladedesignandcomposite/laminatelay-upcanhavesignif-icantchangestothebladeresponse/functionality.
InFig.10(b),aplotisshownofthespatialdistributionofdisplacementmagnitudeovertheHAWT-bladesurface.IncomparisontothebaselineHAWT-bladedesign(Fig.10a),thedesignassociatedwiththeresultsdisplayedinFig.10(b)correspondstorepositioningoftherightshearwebfromx/c=0.5tox/c=0.45.Also,thecomposite-laminatelay-upusedinthemodi eddesigncasewaschangedbyincreasingthebalsacorethicknessby15%relativetothebaselinecase.Thenewcomposite-laminatelay-upisdisplayedasaninsetinFig.10(b).
AcomparisonoftheresultsdisplayedinFig.10(a)and(b)showsthatedge-wisestiffnessoftheHAWT-blade,asmea-suredbyitstip-displacement,isafairlysensitivefunctionofthelateralpositionoftherightshearweb.
Achangeinthemodi ed-designHAWT-bladethicknessasafunctionofthenormalizeddistancefromthebladerootisdisplayedinFig.11(thecurvelabeledModi edDesign).Acomparisonoftheresultsdisplayedinthis gureforthetwoHAWT-bladedesignsanalyzedindicatesthatrepositioningoftherightshear-webhasmeasurablycompromised ap-wisebendingstiffnessoftheblade.
InFig.12(b),aspatial-distributionplotofthevonMisesstressovertheinteriorsparisdisplayedforthesameHAWT-bladedesignasthatusedtogeneratetheresultsdisplayedinFig.10(b).AcomparisonbetweentheresultsdisplayedinFig.12(a)and(b)showsthatthestressesaresomewhathigherinthemodi edbladedesign.Consequently,thesafetyfactorobtainedusingthesameprocedureasinthebaselinecasewasfoundtobereducedfrom1.57to1.52.Combiningthis ndingwiththatmadeinconjunctionwithFig.10(a)and(b)and11suggeststhatthereisaneedfortheuseofdesign-optimizationtechniquetoidentifythebladedesignwithanoptimalcombinationofitsfunctional-performancemeasures.
Theeffectofachangeintheshear-web/spar-capcomposite-laminatelay-uprelativetothatusedinthebaselinecaseontheextentofbending-to-torsioncouplingoftheHAWT-bladeisshowninFig.13(thecurvelabeledtheModi edDesigncase).Inthis gure,avariationofthegust-inducedtwistanglealong
JournalofMaterialsEngineeringandPerformancethelengthofthebladeisdisplayed.AcomparisonofthetwosetsofresultsdisplayedinFig.13showsthatsigni cantchangesintheextentofbending-to-torsioncouplingarefeasiblethroughmodi cationsinthecomposite-laminatelayup.Thefatigue-lifeassessmentprocedurebasedontheuseoftherain owcycle-countingalgorithm,theGoodmandiagram,andtheMinerÕsruleyieldedafatiguelifeof27.4yearsfortheHAWT-bladedesignusedtogeneratetheresultsdisplayedinFig.10(b)and12(b).Again,itisclearthatboththebladeperformanceandlongevityaresensitivetothebladedesignandthattheuseofdesign-optimizationmethodscouldbequitebene cial.ThisaspectoftheHAWT-bladedesignwillbeaddressedinourfuturecommunication.
3.3HAWT-BladeMaterialSelection
InSection2.4,itwasdiscussedthatthethreemostimportantmaterialpropertieswhichcontrolsuitabilityofagivenmaterialforuseintheHAWT-bladeapplicationsarethedensity,theYoungÕsmodulus,andthefatiguestrength/life.Threerelatedmaterial-selectionindiceswerealsoderived,anditwasarguedthatoneofthem,i.e.,indexMindexwhich2,Eq3,isessentiallyamaterial-propertylimitisusedtoscreenoutthematerialswhichdonotpossessthesuf cientlevelofstiffness,asquanti edbytheirYoungÕsmodulus.Consequently,theconditionM2 20GPawasappliedattheonsetofthepresentmaterial-selectionprocesstoeliminateunacceptablematerials.Then,indicesMcarryouttheHAWT-bladematerialselection1andMin3areusedtoaccordancewiththeprocedureoutlinedinSection2.4.Inconstructingthecorrespondingmaterial-selectionchartsvariousopen-literaturematerial-datasourceswereconsulted.TheYoungÕsmodulusdatausedpertaintothemeanin-planevalueofthisquantity.Theresultsofthematerial-selectionprocedurecarriedoutinthisworkaresummarizedinFig.14(a)to(c).Itshouldbenotedthat,forclarity,materialsappearinginFig.14(a)to(c)arelabeledusingnumbersandalegendisprovidedinFig.14(d)forthenumberlabelsused.
InFig.14(a),theYoungÕsmodulusversusdensitydataarecompiledforanumberofthermosetting-polymermatrixcomposites.Alog-logplotwasusedinFig.14(a)andthreelineswithaconstantslopeweredrawninaccordancewiththede nitionofthe rstmaterial-selectionindex,M1(Eq2).Thethreeguidelines(withaslopeof2.0)areassociatedwiththeMthe1-levelsof10,500,22,500,and40,000GPa0.5/(kg/m3)withlargerM1valuecausingtheguidelinetobeshiftedupward.Also,a20GPadashedguidelineisdisplayedinFig.14(a)inordertodenotethepreviouslyestablishedlower-boundfortheYoungÕsmodulusofthecandidatematerialsforuseintheHAWT-bladeapplications.WithrespecttotheM1material-selectionindexalone,theoptimalmaterialsarethoselocatedabovethetopmostguideline.
InFig.14(b),alinear-linearplotisshownofthedatapertainingtotheendurancelimit(x-axis)andthetoughness(y-axis)forthesamesetofmaterialsasthatusedinFig.14(a).ThreesolidguidelinesarealsoshowninFig.14(b),andtheycorrespondtotheMtothe3-levelsof1,7,and13MPa2m.Again,withrespectMare3material-selectionindexalone,theoptimalmaterialsthoselocatedabovethetopmostguideline.
Sincedifferentmaterialareidenti edasoptimalifmaterialselectionisbasedontheuseofasingleindex(M1orM3,Fig.14aandb),aprocedurewasdevelopedherewhichtakes
Volume19(6)August2010—799
热门文档