REPAIRGUIDEFOR DENSOCOMMONRAILINJECTORREPAIR

DISCLAIMER THEFOLLOWINGDOCUMENTISBASEDONTHE RESULTSANDEXPERIENCEGAINEDFROMSTRIPPING, REBUILDINGANDTESTINGNEWGENUINEDENSO INJECTORSUSINGHARTRIDGEALLMAKESTEST EQUIPMENT. THEINFORMATIONCONTAINEDWITHINISNOT INTENDEDFORUSEASASUBSTITUTETOTHEGENUINE DENSOREPAIRPROGRAM.ITSCONTENTISFOR INFORMATIONPURPOSESONLYANDISTOBEUSEDAS AGUIDEANDINTRODUCTIONTODENSOCOMMON RAILINJECTORREPAIR. DIESELDISTRIBUTORSAUSTRALIAPTYLIMITEDAND DIESELDISTRIBUTORSLIMITEDINNEWZEALAND ACCEPTSNORESPONSIBILTYFORANYDAMAGEOR INJURYTHATMAYOCCURFROMTHEUNAUTHORISED REPAIR/ADJUSTMENTOFDENSOCRINJECTORS.
2


Introduction
TechnicalandrepairinformationsurroundingDensoCommonRailinjectorsoutsideofthe Densonetworkislimitedandscarce,yettheiruseandpresencewithintheAustralasian dieselmarketiswidespread. Thefollowingdocumentaimstoprovideaguidetousingthetoolingandshimkitstogether withgenuineandnongenuineparts,tostrip,rebuildandtestDensoG2typecommonrail injectors.Theinformationhereinhasbeengainedfromthedismantling,rebuildand consequentflowandresponsetestingofnewDensoCRinjectorsusingtheHartridgeCriPC andIFT70allmakesprogramme.Alterationstotheshimadjustmentofindividualinjectors weremadeandtheresultsrecorded.Theseresultshavebeenusedtodeterminewhat effectsanychangestotheshimadjustmenthaveontheoverallperformanceoftheinjector. Correctioncodes(QRcodes)cannotbewrittenandproducedforanyDensoinjectorusing thisinformation.TheproductionandcreationofQRcodesrequirescomplexsoftwareand intricateflowtestingusinggenuineDenso/Hartridgeauthorisedequipment.DensoCR injectorrepairinvolvesstringentcleanroompracticesandverytightqualitycontrolonbuild tolerances.PleasetakethisintoconsiderationwhenattemptinganydismantlingofDenso CRinjectors. Hopefullythisguidewillprovideaninsightintowhatcanbeachievedwithtoolingavailable outsideoftheDensoofficialrepairprogramme.Considerationshouldbegiventothe importanceofcorrectioncodeswhenundertakinganyadjustmentorunofficialrepairof ANYcommonrailinjectorthatusesacorrectioncode.

ComponentOverviewandIdentification:
ThephotobelowshowsanexplodedstrippedDensoG2injectorandlabelledcomponents:
Solenoid CalibrationShim SolenoidSpringShim SolenoidSpring SolenoidValve HalfBallValve ValveSeatRetainingNut ValveSeat NOPShim NOPSpring PressureSpindle CommandPiston Nozzle

Dismantling
Thefollowingdismantling,rebuildandtestinstructionsarebasedontheuseoftheCR injectorkit(partnumber850901)andShimKits(850276,850296,850301)availableto purchasefromDieselDistributors,togetherwithgenuineandnongenuinerepairparts. STEP1SolenoidRemoval

Figure1 Mounttheinjectoruprightintheviceusingthecorrectholdingjaws,dependinguponthe bodytypeandusingthe27mmcrowsfootgentlyloosenthesolenoid(fig1).Onceloose continuetoloosenbyhandandCAREFULLYandGENTLYremovethesolenoid,takingcare nottodropthesolenoidvalvespringortensioningshim(Fig2). Removethefuelcalibrationshimthatwilleitherstillbesatontheinjectororsatupinside thesolenoidbody(Fig2). Thesolenoidvalvecanthenberemovedusinglongnosepliers.TAKECARENOTTODROP THEHALFBALLVALVE(Fig3).Thisissituatedintheendofthesolenoidvalve.Itissmall andveryeasilylost!!(ReplacementvalvesandhalfballvalvesareavailablethroughDiesel Distributors). Figure2 Figure3

STEP2ValveSeatRemoval

Figure4 InserttheThreeProngedValveSeatSocketintotheguide(Fig4),ensuringthatthe3prongs arelocatedinthe3holesinthevalveseatretainingscrewandthenscrewtheguideover thesocket.Tightentheguideallthewaydownbyhanduntilitisfelttobottomout,then unscrewbyapproximately1/8thofaturn(Fig5).Weneedtoensurethattheinner3 prongedsocketwillturn,whilstmaintainingmaximumsupportfromtheguide,minimising theriskofthesockettilting. WARNINGThisnutisusuallyverytight!.Useasuitablebreakerbaranda14mmsocket placedoverthehexonthe3prongedsockettoundothevalveseatretainingnut(Fig6).If unsuccessful,useanelectricheatguntoapplyGENTLE,MODERATEheattothenut,then usingthebrassdrift,strikethenutandretry.DONOTUSEABLOWTORCH!!Excessiveheat willdamagetheinjector! Figure5 Figure6

Figure7 Withthevalveseatremoved(Fig7),carefullyremovethe2locatingdowels(Fig8).Remove theinjectorfromtheholdingjigandturnupwards,whilstholdingyourhandovertheopen endofthebody(Fig9).Thecommandpistonshouldslideout,oratleastbegintoprotrude fromthebore.Withdrawthepistonfromthebody.Ifthereisnomovementofthe commandpistonatthispoint,itmaysignifythatitisseizedortight.Ifthisisthecase,then itcanbeleftinsitutoberemovedafterthenozzleendhasbeendismantled. Figure8 Figure9

STEP3CapnutRemoval Remounttheinjectorintheholdingjigwiththenozzleassemblypointingupwards.

Figure10 Aswiththevalveseatretainingnut,thenozzleretainingnutcanalsobenotoriouslytight. Onemethodthatcanbeusedtoremovethenutmoreeasilyandminimisetheriskof damagetothebodythreadandlocatingdowelsinvolvestheuseofapipecuttertoscore thecapandfreethethreadpriortoremoval.Itiscommonpracticetorenewthecapnut onCRinjectorsanyway,sothecapnutwillhavetobediscardedandanewone(available fromDieselDistributors)willbefittedonrebuild. Toremovethecapnut,scorealinenodeeperthan0.5mmapproximately15mmfromthe baseofthenut(oppositetonozzleside)aroundthecapnut(Fig10).Then,usingagood 15mmsinglehexsocket,undothenut.Thereisa15mmcrowsfootsuppliedinthekit,but thisisbestreservedfortightening,ifusinganinjectoraligningjig. Oncetheinjectorcapnuthasbeenremoved,thenozzlecanberemovedalongwiththe locatingdowels(Fig11).Finally,removetheinjectorbodyfromthejigandcarefullyturn over,whilstholdingtheneedlespringandremovethespringandspringtensioningshim (Fig12). Ifthecommandpistonwasnotremovedearlierduetobeingtightorseized,thenitcannow begentlytappedoutfromoneend.USUALLY,IFTHECOMMANDPISTONREQUIRESFORCE TOBEREMOVED,THEINJECTORBODYANDPISTONCANBECONSIDEREDUNSERVICABLE ANDNOTWORTHREPAIRING. Figure11 Figure12

Inspection
Solenoidassembly

Figure13 Resistancechecksshouldbemadetothesolenoid.Thespecificresistancemeasuredwill dependuponthetypeofinjectorandshouldbecomparedtoanewunit.Theresistance recordedontheparticularG2injectorsthatwetestedwasapproximately0.50.7Ohms. Obvioussignsofanydamagetotheoutercasingorsplits/cracksontheelectricalconnector shouldbeobserved.Checkforerosion/corrosionoftheelectricalterminalsandensurethat theinnersurfaceofthesolenoidiscleanandfreefromdentsetc(Fig13). Ensurethatthesolenoidspringisingoodcondition.Itmaybeagoodideatocheckand recordthelengthagainstanewone.

SolenoidValveandHalfBallValveassembly

Inspectball,seatandshank

Figure14 Checktheconditionofthesolenoidvalveassembly,payingparticularattentiontothehalf ballvalvefaceandtheseatinsidethevalvestem.Theflatfaceofthehalfballvalveis responsibleforsealingthecontrolchamber,soitmustbeingoodconditionwithnoerosion orwear.Thesecomponentsarebestinspectedunderamicroscope.Checktheconicalseat thattheballseatsoninsidethevalve,aswellasthevalvestemforanysignsofwear, scratchesanderosion(Fig14). Renewbothcomponentsifthereareanyabnormalitiesordoubtontheircondition(both availablefromDieselDistributors).

 ValveSeatandRetainingNut
Inspectallsurfacesforerosion&wear

Figure16 Inspectthevalveseatforerosionandwear(Fig16).Aswiththeballandvalveassembly,this isprobablybestviewedunderamicroscopeorothermagnificationequipment.Ifthereare anymarksacrosstheseatingareathatwouldsuggestasealingproblemrenewthevalve seat(availablethroughDieselDistributors). Inspectthesealingfaceoftheretainingnutforsignsoferosionandwear(Fig16). DuetotheveryfinetolerancesinvolvedinCRinjectors,lappingofthesealingsurfacesof thesecomponentsisnotrecommended.Anyattempttolapthesurfacesshouldbe undertakenwiththeutmostcareandwiththeuseofVERYFINElappingpastes.

10

 CommandPiston&InjectorBody

Figure17 Checkthepistonforscoringandwearmarksalongwiththeinnerboreoftheinjector.If thereisconsideredtobeexcessivescoringorwearoneithercomponent,thentheunitwill morethanlikelybescrapped(Fig17).Thepistonshouldbeclean,freefromscoringand movefreelyinsidetheinjectorborewithnosnaggingortightspots. NozzleAssembly,PressureSpindleandSpring. Guidelinesforinspectingthenozzleassemblycanbeconsideredthesameasforany conventionalnozzle.Checkforexcessivewearorscoringontheneedleshankandseatand renewthenozzleifnecessary.(GenuineandnongenuinenozzlesareavailablefromDiesel Distributors,dependinguponapplicationandavailability). Checkthespringforconditionandthespindleforexcessivewearonthespringseat.

11

ReAssembly
Whenreassemblinganycommonrailcomponents,cleanpracticesshouldalwaysbe observedandtheutmostprecautionshouldbetakentoavoidanycontaminationofthe componentsbeforeandduringassembly. WARNINGThetorquevaluesmentionedinthefollowingprocedureareaSUGGESTED VALUEbaseduponourexperienceandNOTDensospecification. STEP1.2ValveSeatAssembly

Figure18 Mounttheinjectorintheholdingjigandplacethe2locatingdowelsintotheirboresonthe topfaceoftheinjector(thesedowelsarelargerinlengthanddiametertothenozzle locatingdowels).Ensurethatthevalveseatisthecorrectwayroundbeforeinsertingitinto theinjector.Theboreontheundersideofthevalveneedstobealignedwiththeborehole ontheinjectorface(Fig18&19). Usingthe3prongedsocket,insertandhandtightentheretainingnutintotheinjectorwith thesealingfacetowardsthevalve.Screwtheguideoverthesocket,ensuringthatthesocket canstillmove,butwithmaximumsupportfromtheguideandtightento7075NM(Fig20). Figure19 Figure20

12

STEP2.2Solenoid&SolenoidValveAssembly

Figure21 Fitthehalfballvalveintothesolenoidvalvestemandensurethattheflatfaceofthehalf ballvalveisvisibleandlevelwiththestem.DONOTINSERTTHEVALVESTEMINTOTHE VALVESEATIFYOUCANSEETHESPHERICALPARTOFTHEHALFBALLVALVE(Fig22).IT MUSTBETHEFLATSIDE(Fig21). Figure22

Figure23 Thehalfballvalveshouldstayinthevalvestemwhenturnedoverandinsertedonthevalve seat.Oncethevalvestemisinpositionensureitisfreebyrotatingit(Fig23). Placethefuelcalibrationshimontopoftheinjector(Fig24). Next,insertthesolenoidspringshimandsolenoidspringintothesolenoidandcarefullyfit thesolenoidontotheinjector,ensuringthatthespringseatsinthecentreofthevalveand handtighten.Oncehandtight,tensionthesolenoidassemblyto1012NM(Fig25). OVERTIGHTENINGOFTHESOLENOIDASSEMBLYCANHAVEANADVERSEAFFECTONTHE OPERATIONOFTHEINJECTORANDMAYRESULTINDAMAGETOTHESOLENOID. Figure24 Figure25

13

 STEP3.2NozzleandCapNutAssembly

Figure26 Repositiontheinjectorassemblyinthejigwiththesolenoidtowardsthebottom. Insertthecommandpistonintotheinjectorbodywiththemainshankmachinedwith groovestowardsthesolenoidend(Fig26). Nextinsertthenozzleopeningpressureshimandthespringintothecentralinjectorbore andtheninsertthepressurespindleontopofthespringwiththesmallerdiameterfacing thenozzleneedle(Fig27.) Insertthelocatingdowelsintothesealingfaceandplacetheinjectoroverthem.Placethe (new)capnutoverthenozzleandscrewdownuntilhandtight(Fig28). Ideallysomekindofinjectorclampingrigshouldbeusedtounsurealignmentofthenozzle, ifyoucanfindsomethingormanufactureanythingsuitabletoholdthebody.The arrangementpicturedbelow(Fig29)featurestheStanadyneInjectorAligningTool(Part# 29617).Aninjectoraligningtoolwasnotusedinthisexerciseandnooperationalproblems wereexperienced,howeveritwouldbebettertouseoneifavailabletominimisetheriskof misalignmentorbreakingthedowelpins. Figure27 Figure28

Figure29
14

Testing
AlltestingoftheDensoCRinjectorsusedtoobtaintheinformationgivenherewas completedusingtheHartidgeCRIPCandtheIFT70.(bothavailablefromDiesel Distributors) WhentestingDensoCommonRailinjectorsthefollowingelementsshouldbeobservedto evaluateanddeterminethefunctionalityandconditionoftheinjector: 1Nozzleopeningpressure 2InjectorResponseTime 3FuelDelivery 4Backleakage 5Backleakagetemperature

Eachoftheabovewillbediscussedinmoredetailinthefollowingguidewhichhopesto provideanunderstandingofwhateffectalteringtheshimadjustmenthasontheinjectors performance.

1. NozzleNeedleopeningpressure Thiscanbesetbytheshimthatsitsbehindthenozzlespring.Thenozzleopening pressuresettingonaG2injectorwillhaveaneffectonfueldeliveryandResponse Time(seefurtherdowninsection2fordefinition)particularlyatlowerengine speedsandrailpressures.Iftheopeningpressureissettoolow,theidledeliveryand preinjectionwillincrease.Iftheopeningpressureissettoohightheidleandpre injectionwilldecreaseandresponsetimewillincrease.Fuellingandresponsetimeat mainfuellingsettingswillremainlargelyunaffected. WecansettheopeningpressurewiththeuseofaHartridgeIFT70injectorfunction testerthatcanactuatethesolenoidwhilsttheusercanobservethespraypattern andcontroltherailorfeedpressure.Thisexercisecouldpossiblybeperformedwith aconventionalinjectortesterandamultifunctionCRinjectortriggerboxsuchasthe IFTC(availablefromDieselDistributors). THISTESTSHOULDBECARRIEDOUTANDSET(IFNECESSARY)BEFOREANYFLOW TESTINGOFTHEINJECTOR.
15

Mounttheinjectorintothemachineandconnectthehighpressurepipeand appropriateelectricalconnector.

Figure30 Usingalargepulsewidthofaround2.5m/sandlowinjectionspeedofsay 100injectionsperminute(IPM)(Fig30).Setthefeed(railpressure)tozero andgraduallyincreaseuntilthefuelcanbeseentoinject.Atthispointtake noteofthepressureonthegaugeatwhichinjectionstartedtotakeplace. (Fig31). TheopeningpressureontheparticularDensoinjectortestedwas approximately130140bar.Thisfiguremaydifferdependinguponinjector partnumber.Itwillbeuptotheusertogainsomespecificationsfromnew injectors. Ifthepressureneedstobeadjusted,thentheopeningpressureshimwill needtobechangedtoincreaseordecreasespringtensionaccordingly.Since thenozzlecapnutwillnothavebeenputintoservicethelooseningandre tensioningshouldnotcauseanyissuesandshouldbeabletobereusedafter removal. Acheckofthespraypatternandgeneraloperationoftheinjectorcanthen beperformedwhilststillmountedontheIFT70aftertheopeningpressure hasbeenchecked.Asageneralrulethereshouldbenomorethan3040% backleakagedeliverycomparedwithfueldeliveryonfullloadsettings(e.g. 500bar1200PPM1.2m/spulsewidthwhenusingtheIFT70). Iftheinjectorisfunctioningproperlywithbasicdelivery,backleakageand spraypatternallsatisfactory,theinjectorcanbeflowedmoreaccuratelywith theHartridgeCRIPCorAVM2withCRinjectorallmakescapableofchecking fueldelivery,backleakageandresponsetimeathighpressures(1600bar+). Figure31

16

 2. ResponseTime SimplyexplainedthisisthetimetakenfromwhentheECUsendsasignal(current)to theinjectortowhentheactualinjectiontakesplace. Itisanimportantfactorindeterminingtheconditionoftheinjector.Ingeneral increasedresponsetimepointstowardsapoorlyfunctioninginjector.Onvehicleeffects ofpoorresponsetimecanshowaswhitesmoke,roughrunning,idlingproblems, detonation(knocking)andpoorperformance. InjectorresponsetimeisanimportantfactorfortheECUtobeabletocompensate drivepulseforrecalculatingfuelquantitiesforonvehiclelearningofQRcodes. Responsetimecanbeaffectedbyopeningpressure,commandpistoncondition,nozzle condition,solenoidvalveconditionandsolenoidefficiency.Anythingthatcaneffectively slowdowntheoperationoftheinjectorwillresultinincreasedresponsetime. Ifresponsetimecanbereadandseenthenaclearerpicturecanbegainedfromthe injectortestonthegeneralconditionoftheinjector,asopposedtoonlyrelyingupon measuringdeliveryandbackleakage.Itcanbeconsideredimportantinourdiagnosisof injectorcondition.Itcanbeaffectedbychangingfuellingshimsandopeningpressure. TypicalresponsetimesrecordedfortheDensoinjectorstestedwereintheregionof 430490sforhigherpulsewidthandpressuresettings(fullandmidloadscenario). Responsetimewillgenerallyincreaseatidleandlowerpulsewidthsandpressures(idle, partloadandpreinjection).Typicalidleresponsetimesmaybearound490560s. Again,theabovefiguresarebaseduponthetestingofcertainDensoinjectorsandto gainamorethoroughunderstandingofinjectorspecificresponsetimefigures,new injectorswouldneedtobetestedandresultsrecorded.

17

 3. FuelDelivery Aftertheopeningpressurehasbeensetandtheinjectorhasbeentestedforbasic functionitcanthenbeproperlyflowedforamoreindepthtestontheHartridgeAVM2 PCwithinjectorallmakeskitortheHartridgeCRIPC. Theinjectorneedstobegivenatestplanofsettingsfromwhichtheresultscanbe recorded.Thesevaryingsettingsshouldputtheinjectorthroughdifferingtestsituations thatwouldbestsimulatevariousengineconditions.Anexampleofthesetestconditions hasbeenincludedbelowforreference:(Fig32).

Injector# Pressure(bar) Idle400 MidRange1000 FullLoad1600 StaticLeak1600 PreInjection600 MidRange800 Start950

Figure31TestPlanExample Onceasetoftestsorconditionshavebeendecidedupon,theycanthenbeusedasa generictestplanforalmostallinjectors.Thetestplanisthenusedtorecordtheresults fromnewinjectorsandusedforcomparisonagainstusedorrepairedunits.Adatabase ofresultscanbebuiltupandtolerancescanevenbeformedbytestingmultiplenew unitsandtakingaveragereadings,fromwhichtolerancebandscanbedecided. Iftheflowresultsfromausedorrepairedinjectoraredeemedtobeoutofspecwhen comparedtoatestresultfromanewinjectorthenthefueldeliverycanbealteredby changingtheCalibrationShimwhichsitsinbetweenthesolenoidandtheinjectorbody. CHANGINGTHISSHIMWILLHAVEANEFFECTONTHEOVERALLFUELDELIVERYOFTHE INJECTOR. ToINCREASEthefuellingaTHICKERshimmustbefitted ToDECREASEthefuellingaSMALLERshimmustbefitted.

18

 0.1mmdifferenceinshimsizewillequatetoanapproximate7%changeinfuelling acrossthewholerangeoftheinjector.However,theresponsetimewillalsochangeas wehaveeffectivelychangedthetotaldistancethatthesolenoidvalvehastotravel.In anefforttocounteracttheresultingresponsetimechange,thesolenoidspringshimcan bechangedtotryandlimittheresultingdifferenceinresponsetime.Althougha differenceinresponsetimeisseenwhenthesolenoidspringshimischanged,itis minimalanddoesnotreturntotheoriginalresultsafteramaincalibrationshim alteration. Itwouldthereforebeprudenttomakeonlysmallalterationstothecalibrationsetting whenconsideringanychangestothefuellingoftheinjector. Changingthesolenoidspringshimwillalsoaffectfueldeliveryaswellasresponsetime, aswearereducingorincreasingspringtensiononthesolenoidvalve,makingiteasieror harderforthesolenoidtopullupthevalve.Again,considertheeffectstothefuelling throughouttherangewhenattemptinganyshimadjustment. 4Backleakage&BackleakTemperature TheamountofbackleakagefuelthataCRinjectorproducesisanindicatorofthe internalsealingandconditionoftheinjector.Anyproblemswiththesealingofthe solenoidvalveorthenozzleneedlewillshowupashighbackleakage.Usually,ifhigh backleakageisexperienceditwillbeseenineverytestandcanindicatevalveseat problems,solenoidvalveproblemsandnozzlesealingorwearissues. Backleakagetemperatureisalsoworthmonitoring,asanyexcessiveleakageofhigh pressurefuelwithintheinjectorwillcausethetemperatureofthebackleakagetorise. Thisisbecausethehighpressurefuelcreatedintherailisalothighertemperaturethan thelowerpressureofthebackleakage,thereforeanyexcessiveriseinbackleakedfuel willshowupasahighbackleaktemperature.

19

CorrectionCoding&Batching
CRinjectorcorrectioncodingisimportantandallowstheCRsystemasawholeto compensateformechanicalandhydraulictoleranceswithintheinjector.Notwo injectorswillnaturallyflowthesameandthevehicleECUcancompensateforthisby alteringthepulsewidthtoindividualinjectorswhenthecorrectioncodegiventothe injectorduringmanufactureisprogrammedintotheECU. Wecannotproducethiscodewithconventionalallmakesequipment.Itrequiresa considerableinvestmentandOEMbackingtodoso. Onemethodoftryingtolimittheproblemofourinabilitytoproduceanindividual correctioncodeistobatchinjectors.Thismethodinvolvesflowtestingandadjusting thefueldeliveryoninjectorssothatthedifferencesbetweenthemareminimal.Sayfor instancewehadaround30injectorsthatwererepairedandflowtested.Thebatching conceptassumesthatoutofthose30injectorstherewouldbesetsthatwoulddisplay similarflowandresponsecharacteristics.Theseinjectorscouldthenbefurtherfine tunedwithminoradjustmenttomakethemassimilartoeachotheraspossible. Setsof4or6injectorsarethensoldorfittedtogethertominimiseanymajordifferences inflow. Anotherconceptistobuildlengthydatabaseontheflowratesofdifferingnewinjectors fromsamepartnumberfamilybyrecordingtheirflowandresponsecharacteristicsand thecodeonthatparticularinjector.Overtimeadatabasecanbebuiltandanyfuture injectorsthatdisplayedsimilarresultstoapreviouslyrecordedunitwouldbegiventhe codefromthatsimilarflowingnewinjector.Thiswouldbealongandcomplicated process,butnotimpossible.

20

 SUMMARY Thereareparts,testequipmentandtoolingavailablethroughDieselDistributorsto serviceandtestDensoCRinjectors.Thereisnoconcretesetofinstructionsortestdata toproduceaperfectlyrepairedproduct.Thereishowever,anopeningtobeableto offersomethingmorethanjustanewunitinbox. Therepairconceptofferedhereistoenabletheusertheabilitytolearnasyougoand baserepairsondataobtainedfromnewunits. Hopefully,theinformationinthisdocumentisenoughtogetyoustarted. Forfurtherinformationpleasecontact: DieselDistributorsAustraliaPeterHunterpeterh@ddal.com.au DieselDistributorsNewZealandKevinMorriskevin@dieseld.co.nz

21