• Keine Ergebnisse gefunden

Synergistic alloy design concept for new high-strength Al–Mg–Si thick plate alloys

N/A
N/A
Protected

Academic year: 2022

Aktie "Synergistic alloy design concept for new high-strength Al–Mg–Si thick plate alloys"

Copied!
13
0
0

Wird geladen.... (Jetzt Volltext ansehen)

Volltext

(1)

ContentslistsavailableatScienceDirect

Materialia

journalhomepage:www.elsevier.com/locate/mtla

Synergistic alloy design concept for new high-strength Al–Mg–Si thick plate alloys

Florian Schmid

a,

, Irmgard Weißensteiner

a

, Matheus A. Tunes

b

, Thomas Kremmer

b

, Thomas Ebner

c

, Roland Morak

c

, Peter J. Uggowitzer

b

, Stefan Pogatscher

a,

aChristian Doppler Laboratory for Advanced Aluminum Alloys, Chair of Nonferrous Metallurgy, Montanuniversitaet Leoben, Franz-Josef Strasse 18, 8700 Leoben, Austria

bChair of Nonferrous Metallurgy, Montanuniversitaet Leoben, Franz-Josef Strasse 18, 8700 Leoben, Austria

cAMAG rolling GmbH, Lamprechtshausener Strasse 61, Postfach 32, 5282 Ranshofen, Austria

a r t i c le i n f o

Keywords:

Aluminum alloys Precipitation strengthening Alloy design

Wrought alloys

a b s t r a ct

Withtheaimoffullyexploitingtheadvantageousstrength-to-weightratioevidentinAl–Mg–Sialloys,thisstudy presentsmeasuresforincreasingtheyieldstrengthofanENAW-6082typeplatealloy.Inadditiontodescribing thethermodynamicsimulation-basedadjustmentofage-hardenableelements(Si,MgandCu)andamodified artificialageingtreatment,itinvestigatestheeffectsofaddingasmallamountofZr.Thesignificantstrengthen- inginducedbyaddingZriscorrelatedwithsub-grainboundaryhardeninginarecoveredmicrostructureafter solutionannealingat570°C,comparedwiththealmostentirelyrecrystallizedmicrostructureinanunmodified ENAW-6082alloy.Incombinationwithamaximumdissolvablenumberofage-hardenableelementsandinter- ruptedquenching,whichcomprisesanimprovedheattreatmentstrategyforthickplates,itisseenthattheyield strengthcanbeincreasedbymorethan40%to411MPacomparedtoconventionalENAW-6082basematerial asverifiedbytensiletesting.Inthestudyscanningelectronmicroscopyandscanningtransmissionelectronmi- croscopywereperformedformicrostructuralcharacterizationwithafocusonparticleanddeformationanalysis.

Allindividualcontributionswhichgeneratedthesuperiorstrengtharecalculatedanddiscussedinordertoreveal themicrostructure-propertyrelationship.

1. Introduction

Weight-optimized design hasbecomea top priority in the trans- portsector becauseitfacilitates substantialCO2 reduction[1]. Low- densityAlalloys,especiallyage-hardenableAl–Mg–Si(6xxx)plateal- loys,arecurrentlyapreferredchoice.Inadditiontotheirmediumto highstrength,thickplatesmadefrom6xxx-alloyscombinehighfracture toughness,goodcorrosionresistanceandweldability.Theyalsofeature goodrecyclabilityatamoderateprice.Tofurtherexploitthepotential ofthesealloysandpromotetheirselection,anincreaseoftheirstrength- to-weightratiotomorefavorablelevelsseemstobroadentheirfieldof applications[2–4].

Workhardeningplaysakeyroleinsheetmaterials. Deformation- inducedstructuressuchasdislocationsandcellstructureswhichorig- inate from coldrolling contributesignificantlytoenhancedstrength valuesinthinsemi-finishedproducts[2].However,platealloys,which aredeformedathightemperaturesonly,donotnecessarilyexperience

Correspondingauthors.

E-mailaddresses:florian.schmid@unileoben.ac.at(F.Schmid),irmgard.weissensteiner@unileoben.ac.at(I.Weißensteiner),matheus.tunes@unileoben.ac.at(M.A.

Tunes),thomas.kremmer@unileoben.ac.at(T.Kremmer),thomas.ebner@amag.at(T.Ebner),roland.morak@amag.at(R.Morak),peter.uggowitzer@unileoben.ac.at (P.J.Uggowitzer),stefan.pogatscher@unileoben.ac.at(S.Pogatscher).

strengtheningfromthisprocessingstepbecausetheirincreaseddisloca- tiondensitywanesduetorecoveryprocesses[5].Inaddition,theso- lutionannealingtreatmentforAl–Mg–Sialloysistypicallyperformed above530°C,whichweakenshardeningviacoldworking[6].Several otherhardeningmechanismsarealsoactiveinthick plates,ofwhich strengtheningviatheformationofmetastableMgandSicontaining𝛽’’- phasesuponartificialagingisthemostimportant[2,7].AddingCuen- hancesthestrengthresponsefurther,mainlyduetotheoccurrenceofL- andQ’-phases[8,9].IncorporationofCuintohardeningphaseswhich containMgandSialsotakesplace.Allofthesealloyingelementsmay alsocontributedistinctlytosolidsolutionhardening[10].

Inthecourseofindustrialheattreatment,highstrengthAl–Mg–Si alloyswithaMgandSisolutecontentexceeding1%sufferfromaloss ofagehardeningpotentialwhichiscausedbynaturalaging(NA)dur- ingroomtemperature(RT)storagebetweensolutionannealingandar- tificial aging[11,12]. Naturalagingin plates maybe suppressedby applying interruptedquenching(IQ)toanelevated temperatureand

https://doi.org/10.1016/j.mtla.2020.100997

Received21December2020;Accepted25December2020 Availableonline2January2021

2589-1529/© 2021ActaMaterialiaInc.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense

(2)

slowfinalcoolingtoroomtemperaturedirectlyafterwards.Thisgener- atestransformableclustersandnucleiofthemainhardeningphase𝛽’’

[13].InthisrespectIQisthereforecomparabletopre-aging[14],and increasedmechanicalparameterscanbeachievedinthefinalproduct [13,15,16].

In addition to the group of age-hardening elements, dispersoid- formingelementswithlowsolubilitysuchasFe,MnandCrcanbead- vantageouslyadded.Ifprimaryconstituentsarelargelysuppresseddur- ingthecastingprocess,smallanddenselydistributedsecondaryphases willformduringhomogenization[17,18]significantlyinfluencingthe mobilityofboththegrainboundaryandthedislocationmotion[19].Es- peciallyinAl–Mg–Sialloys,effectivelystabilizingthegrainboundaries turnsout tobe quitebeneficial becauseprocessingtemperaturesare comparablyhigh,thusincreasingthekineticsofboundaryrearrange- ments[20–22].

Ina2006studyKniplingetal.[23]listedseveralelementsthatform trialuminides,Al3X(X=Sc,Ti,Zr,Hf,Er,Tm,YrandLu),inAlalloys, ofwhichZirconiumandScandiumhavealreadybeenintensivelyinves- tigated.DuetothehighpriceofScandthelowcoarseningresistance ofAl3Sc,Zrmaybeamoreattractivecandidateforindustrialuse.To minimizethedisadvantageoftheprolongedannealingtimenecessitated bythereduceddiffusivityofZrwithinAl,itwasshownthatminorad- ditionsofSi(upto0.18at%)acceleratetheprecipitationkinetics[24]. Generally,Al3X-particlesarenotonlyusedadvantageouslyinunalloyed Al[23,24],butalsoinage-hardenablealloysystems(2xxx[25–27]or 7xxx[28,29])and– thishasbeenquiteprominentlately– inadditively manufacturedparts[30,31].Theseparticleswereshowntohaveasig- nificantimpactondislocationandgrain boundarymotion,whichde- creasesrecrystallizationtendencyand,thus,leadstotheformationof subgrains.IncombinationwithanincreasedamountofOrowanhard- ening,anincreaseinstrengthisthusobservedforScorZrcontaining 2xxxand7xxxseriesalloys[32–34].

TheeffectoftrialuminidesonAl–Mg–Sialloyswasrecentlyinves- tigated[35–38].However,foralloyscontainingSi(inamountsgreater thantheminutenucleationagent additionshownin[24]) itwas re- portedthat Zr,Si-phasescan form [39], thus decreasing theamount ofZravailablefor theformationof finetrialuminides dispersoids.A transformationofthebeneficialmetastableL12-Al3Zrprecipitates(co- herentwithAl-matrix)intoits stableD023-structure(incoherentwith Al-matrix)is alsoobservedatelevated temperatures(roughlyaround 500°C)[40].AhighSicontentsuchasthatofAl–Mg–Sialloyswasre- portedtofavortheoccurrenceofanincoherent(Al,Si)3Zrphaseandto causeacceleratedtransformation[41–43].6xxx-seriesalloysneedcom- parablyhighhomogenizationtemperaturestoobtainthedesiredinter- metalliclandscape(dissolutionofprimary𝛽-Mg2Si,reductionofmicro andmacrosegregation,andtransformationofAl(Fe,Mn)Si-phases)[44–

46], which limitsthepositive effectsof Zron industriallyprocessed Al–Mg–Si alloys due to coarsening. However, investigations demon- stratethatZradditionscanhaveasimilar[47]andevenadditiveeffect [48]onrecrystallizationretardationcomparabletoconventionaldisper- soidscontainingFe,MnorCr.AfewAl–Mg–Sistandards(ENAW-6056, ENAW-6065,ENAW-6110andENAW-6182)allowZruptoamaxi- mumof0.20wt%[49]tomakeuseofthesebeneficialdispersoids.

Inthisstudy,followingthestandardproductionrouteforthickplate alloys(casting, homogenization,hotrolling andartificialaging), we

combinedall ofthestrengtheningmechanismspresented inamodel Al–Mg–Siplatealloy.Thenecessaryprotocolforachievingthedesired highstrengthisasfollows:(i)balancedandoptimizednumberofage- hardenableelements(suchasSi,MgandCu);(ii)applicationofinter- ruptedquenching;and(iii)additionofZr.Itisnotyetfullyunderstood howZrinteractswithMnandFeorhowitreactsat6xxx-typicallyhigh processingtemperatures,andthestrengtheningmechanismduetoZir- coniuminsuchcomplexalloysisalsostillunknown[35].Theseissues areexaminedindetailbelow.Thescientificmethodologyinthisstudy addressesthequestionofwhetheritispossibletocombineallofthese strengtheningmechanismssynergisticallyinonealloy,andwhetherthe yieldstrengthofAl–Mg–Siwroughtthickplatescanbeincreasedtoover 400MPa.

2. Experimentalmethods

Forthisstudy,fivealloysbasedonanENAW-6082wereproduced atalaboratoryscale.Afterre-meltingofprimaryindustrialmaterial(EN AW-6082)inaresistanceheatedfurnace(NaberthermK20/13/S),pure Si,Mg,CuandZirconium(intheformofAlZr10)werealloyedtoobtain thecompositionslistedinTable1.Theseweremeasuredusingoptical emissionspectroscopy(notethatthesimulation-assisteddesignconcept isdescribedinthenextsection).Afterrefiningthemelt,whichwasheld atroughly750°C,viaArgaspurgingandgrainrefineraddition,cast- ingintocustom-madepre-heatedmoldssimilartothatinindustrywas startedaccordingto[50].Acoolingrateof3–5K/swasmaintained.Af- terfilletpiececuttingandhomogenizationtreatment,theresultingslabs werehotrolledfrom40mmdownto6mminseveralrollingsteps.A residualamountof0.02wt%Crispresentinallalloys,duetotheusage ofscrapintheproductionoftheindustrialmaterial.

Solutionheattreatmentoftensiletestsampleswasperformedusinga circulatingairfurnace(NaberthermN60/85SHA)at570°Cfor20min.

Aftereitherwaterquenchingorinterruptedquenchingintoa180°Chot metalbath(Bi57Sn43)withdelayedcoolingwithin24htoRTatade- creasingcoolingrate,thesamplesunderwent14daysofnaturalaging toemulateindustrialprocessing.After7daysofnaturalaging,aplastic pre-deformationof2%tookplace.Afteranother7daysartificialaging wascarriedoutat160°Cfor14hwith10hoframpingup(tosimulate theheatingofthickplates),resultinginatotalagingtimeof24hforthe water-quenchedspecimens(T651state).Startingfromthebasealloy,ar- tificialagingtimesforreachingT651-statewerecheckedforeachalloy variationandnomajordifferencesweredetected.ConcerningIQspeci- mens,2hagingat160°Cwasapplied,i.e.intotal12h(IQstate).The wholeheattreatmentprocessisschematicallyshowninFig.1.Whereas thesolidblacklinerepresentsastandardindustrialprocessingscheme forhotrolledplatealloys,ourhighstrengthvariantsrequireanalter- ationofthehomogenizationstrategy(dashedorangeline)andtuningof thefinalheattreatment(IQ+shortenedartificialaging),asindicated bythedashedblueline(seealsoSections3.3and3.4).

To characterize the mechanical parameters uniaxial tensile tests were performed on a universal testing machine (Zwick-Roell BT1) equippedwitha50kNloadcell.Roundspecimenswerepreparedwith a30mmlongand5mmthickgaugeinaccordancewithEN-ISO6892-1.

As-quenchedspecimensweretestedwithin5minofwaterquenching.

3samplesweretestedtocalculatemeanvalueandstandarddeviation.

Table1

Measuredchemicalcompositionofallalloystested,Alinbalance(wt%).

Si Mg Cu Mn Fe Zr Si + Mg Si/Mg Notes

Base alloy – 6082 1.02 0.67 0.08 0.42 0.33 0.00 1.7 1.5 Remolten industrial primary material 6082-Si 1.19 0.70 0.08 0.44 0.39 0.00 1.9 1.7 Addition of Si

6082-Cu 0.98 0.66 0.29 0.42 0.33 0.00 1.6 1.5 Addition of Cu 6082-Zr 1.03 0.66 0.08 0.42 0.37 0.20 1.7 1.6 Addition of Zr

Optimal alloy 1.07 0.81 0.30 0.41 0.36 0.21 1.9 1.3 Co-Add. of Si/Mg, Cu and Zr

(3)

Fig.1.Schematicillustrationoftheappliedheattreatmentconsistingofhomogenization,hotrolling,solutionannealing,pre-deformationundartificialaging.

Fig.2. Thermodynamicequilibriumcalculationsof6082,6082-Siand6082-Cu(seeTable1forchemicalcompositions).(a)InfluenceofSiandCucontentonthe onsetofliquidphaseformation.At570°Cannealingtemperature1.25wt%Siisthethresholdvaluefortheoccurrenceofaliquidphaseinthe6082basealloyand 1.16wt%in6082-Cu.(b)PhasefractionofMg2SiandQ-phaseinrelationtothetemperature.Volumetricpercentagesarecalculatedwithadensityof2.91g/cm3 forQ-phase[60]and1.90g/cm3forMg2Si[61].

ThermodynamiccalculationswereperformedwithFactSageTM 7.3 and FactSageTM 8.0 software using the FACT FTLite and FACT PS databases(2020)[51,52].

Tocharacterizethemicrostructureascanningelectronmicroscope (SEM)(JEOL7200FFEG-SEM),equippedwithanEBSD-measurement system (NordlysNano detector, Oxford Instruments), was deployed.

Samples wereprepared viaa standard procedure of grinding, oxide (OPS)andelectrolyticpolishing.Forgrainboundaryanalysisanarea of0.5mm2wasanalyzedoneachsample,withastepsizeof0.6μm.

Therecrystallizedareafractionwasdeterminedfromasurveyedareaof atleast2.35mm2(stepsizesof0.6and3𝜇m)takenatt/2representing thecenteroftheplate.Theminimumgrainboundarymisorientationan- glewassetto5°[53]fortherecrystallizationanalysis.Furtherdetails concerningbasicEBSDdataprocessingbymeansofthemtextoolbox canbefoundelsewhere[54–56].Grainsexhibitingagrainaverageker- nelaveragemisorientation(KAM)below0.5° alongwithagrainaverage bandcontrastof>70%ofthemaximumgrainaveragebandcontrast wereidentifiedasfullyrecrystallized.Toanalyzesub-grains,grainswith aminimumboundarymisorientationof0.8° wererecalculated.

BSEmicrographsweretakenatanaccelerationvoltageof5kVanda workingdistanceofroughly4mmtoevaluateintermetallicdispersoid phases.Eachparticle’sroundnessisderivedbydividingtheactualarea bythecalculatedcircularareausingthemaximumdiameter(thevalueis inverselyproportionaltotheaspectratio).Valuescloseto1correspond toanalmostperfectcircularshape.

Scanning transmissionelectronmicroscope(STEM)measurements werecarriedoutonaThermoScientificTMTalosF200XG2.200kVwas takenastheaccelerationvoltage.Sampleswerecutfromundeformed partsoftensiletestsamples.Aftergrindingdownto100𝜇m,3mmdisc specimens wereprepared viaastandardrouteconsistingof grinding andelectrolyticpolishingwithHNO3inmethanol(1:3)at−20°Cand 15V.ConcerningthedeterminationofSicontainingZr-particlesthefol- lowingmethodologywasapplied.Deployingapictureanalysissoftware (ImageJ1.51f)andusingacolorthresholdfortheSiandZrmapping,bi- naryblackandwhitepictureswithparticlescontainingSiandZrwere made. OverlappingthesetwobyusingtheBooleanoperation“AND” broughtbackathirdpicturewhereonlyparticlesweredepictedthat containedbothZrandSiatthesamespot.

(4)

Table2

Overviewofmechanicalparametersofalltestedalloysafterstandardartificialaging(T651)andinter- ruptedquenching+shortenedartificialaging.

As-quenched Standard AA (T651) IQ + shortened AA

R p0.2 R p0.2 R m A R p0.2 R m A

MPa MPa MPa % MPa MPa %

Base alloy – 6082 58 ± 2 289 ± 3 309 ± 2 19 ± 2 349 ± 8 352 ± 9 11 ± 1 6082-Si 66 ± 2 319 ± 8 339 ± 7 15 ± 1 364 ± 4 373 ± 5 12 ± 1 6082-Cu 64 ± 2 313 ± 3 337 ± 3 17 ± 1 355 ± 6 362 ± 6 12 ± 1 6082-Zr 85 ± 0 334 ± 5 362 ± 4 15 ± 1 384 ± 7 398 ± 8 12 ± 1 Optimal alloy 90 ± 3 362 ± 9 392 ± 9 15 ± 1 411 ± 4 425 ± 4 14 ± 0

Fig.3. Isoplethof6082withadditionsofZr.TheoccurrenceofAl3Zrishighly dependentontemperature.Atthesolutionannealingtemperatureof570°CZr hasasolubilityofabout0.1wt%inthesolid𝛼-Al.

3. Designconcept

Inthefollowingwepresentourdesignconcept,whichisbasedonthe exploitationofallscenariosforimprovingstrengthinamodelAl–Mg–Si alloywhichcontainsFeandMn.Theconceptinvolvestuningboththe elementcompositionandtheprocessingparameters.

3.1. Additionandoptimizationofprecipitation-hardeningelements

Asthefirststepinouralloydesign,weadjustedthenumberofage- hardenableelements:SiwasadjustedtoMg,andCuwasaddedasa strength-increasingelement.TheCuamountwaslimitedbecauseofthe decreasingcorrosionresistance(especiallyresistanceagainstintergran- ularcorrosionin6xxx-alloys[57])associatedwithincreasingamounts ofthiselement[2].AfterconsideringtheoptionofincreasingtheCu contenttoacorrosion-irrelevantamountof0.3wt%,andthespecified processparameters(i.e.solutionannealingtemperatureof570°C;see Section2andFig.1forfurtherinformationonthestandardindustrial processingroutefortheproductionofthickplates),wecarriedoutther- modynamiccalculations.TheresultsofwhicharepresentedinFig.2. Fig.2ashowstheliquidphasefractioncoursewithalteredSicontentin the6082basealloyat570°C.Noliquidphaseisallowedtoformatthis temperature,becauseotherwisethematerialwouldbeunusableafter- wards.AthresholdSicontentof1.25wt%canbederivedfromthegraph (notethataddingCuto0.3wt%decreasestheonsetofliquidphasefor- mationtoanSicontentof1.16wt%).TodeterminetheinfluenceofSi onlythiselementwasincreasedfrom1.02to1.19wt%inthe6082-Si alloy,butitwaskeptatthereferencelevelin6082-Cu.MgandSiwere

bothaddedtotheoptimalalloytoapproachthepreferentialSi/Mgratio of1.2inthemainhardeningphase𝛽’’(Mg5Si6[58,59])(seeTable1for thechemicalcompositionsofallalloys).

Fig.2bshowsthephasefractionsoftheequilibriumphasescontain- ingSi,MgandCuoverthetemperature,whichcanbeusedasarough estimationofthemainhardeningprecursorphasesformingduringartifi- cialaging.AddingSiincreasestheamountofMg2Siconsiderably,with- outaffectingtheamountofQ-phase(Al4Cu2Mg8Si6[60]).Inthe6082- Cualloy,adrasticriseoftheQ-phasecontainingCuisseen,alongwith areductionofMg2Si.At160°Casignificantlyincreasedtotalhardening phasevolumeispresentinbothalloyvariations(6082-Siand6082-Cu), whichshouldultimatelyleadtoanenhancedagehardeningresponse.

3.2. AdditionofZirconium

Whereas Zr also precipitates from the super-saturated matrix as Al3Zr,itsnucleationandgrowthtakeplaceatmuchhighertemperatures thanthoseassociatedwiththeelementsdiscussedabove.Fig.3shows anisoplethofthe6082basealloywithZrupto0.3wt%.Amaximum amountofroughly0.1wt%isdissolvablein𝛼-aluminumatthetarget solutionannealingtemperature(570°C).Becausebothhomogenization andhotrollingarecarriedoutat6xxx-typicaltemperatureswellabove 500°C,thereisatendencytodissolutionoflargesharesofpreviously formedAl3Zr.Therefore,anincreasedamountof0.2wt%Zirconiumad- dition(comparedtoamaximumamountof0.15wt%inENAW-7050 forexample[49])waschosen.Becausenoprimarynon-hardeningAl3Zr precipitates,whichimpairelongation,shouldbeformed,thetempera- tureduringthecastingproductionprocesswaskeptabove730°C(see redlineinFig.3).

3.3. Adaptionofhomogenization

DuringhomogenizationtreatmentofconventionalAl–Mg–Sialloys, thedissolutionofpreviouslyformedcoarseMg2Sirequiresaratherhigh temperatureofabout550°C(seeFig.2b).Slowheatingtothistemper- ature alsogeneratesthedesiredformation ofdispersoids.Inorder to formdispersoidsinsmallsizesandhighnumberdensities,wesuggest anadaptationofthehomogenizationtreatmentstep.Ontheonehand, itisknownthatAl(Mn,Fe)Si-dispersoidscontainingFeandMnexhibit thehighestnucleationrateat370 °Cuponheatingat0.01K/s [62]. Conversely,anintermediateannealingstepat360°Chasbeenshownto promotetheformationofAl3Zr-nucleiandthusreducethetimeneeded forfullprecipitationat550°C,whichisinfluencedbytheslowdiffusion ofZr[23,63].Ithasalsobeenreportedthatthesetwophase-familiesin- fluenceeachotherbypromotingheterogeneousnucleationsites[64], whichnecessitatesaholdingannealingsteptopreventcoarseningand non-uniformdistributionofdispersoids.Withthisinmind,optimized homogenization(indicatedbythedashedyellowlineinFig.1)wasap- pliedtothealloyscontainingZr.Thishomogenizationconsistedofan intermediateholdingannealingstepat350°Cfor16handasecondstep at550°Cfor2h.

(5)

Fig.4.Stress–straincurvesof(a)thebasealloy6082,6082-Siand6082-Cuand(b)6082-ZrinT651state

3.4. Interruptedquenching

BecauseAl–Mg–Sialloyslosetheiragehardeningpotentialdueto naturalaging[11,12],quenchingtoanelevatedtemperaturewasap- pliedasanadditionalmeasuretoalltheexperimentalalloystestedhere [13].Accordingly,allsampleswerequenchedto180°Crightaftersolu- tionannealingandslowlycooledtoroomtemperature(thistakes24hin total,atadecreasingcoolingrate)tosimulatethecoolingofthickplates (indicatedbythedashedbluelineinFig.1).Thisshouldleadtothefor- mationoffavorableclustersandnucleiofthemainhardeningphase𝛽’’, andiscomparabletopre-aging[13,14].Becausethistreatmentalready precipitatesmanyoftheage-hardenableelements,ashortenedfinalar- tificialagingstepwasapplied.

4. Results

ThissectionbuildsonthemeasurespresentedinSection3,whichad- dressedalloyandprocessdesign,andoptimalalloyandmicrostructure characterization.Asanoverview,Table2liststhemostimportantstress–

straincurvevaluesfortheagedsamplespresentedbelow.Allalloyswere alsotestedrightaftersolutionannealingintheas-quenchedstate.These valuesarealsolistedinTable2andarediscussedinSection5. 4.1. Alloydesign

4.1.1. Additionandoptimizationofprecipitation-hardeningelements Fig.4ashowsthestressstraincurvesofalloysoptimizedbyadding SiorCuincomparisonwiththebasealloy6082inthefullystrength- enedT651state.Regardingmechanicalproperties,bothmeasurements (Si/Cuaddition)generateaconsiderableincreaseinstrength.Whereas 6082-Cushowsatotalyieldstrengthincreaseof24MPa,addingSire- sultsinanincrementof28MPa.Elongationtofractureissignificantly reducedin6082-Si.6082-Cushowshardlyanychangesinthisrespect.

4.1.2. AdditionofZirconium

Fig.4bshowsthestress–straincurvesofthealloywithaZraddition comparedtothe6082basealloyinT651state(6082-Zrwashomog- enizedby2-steptreatment). Theachievable propertychanges gener- atedbyaddingZrarequiteremarkable.Asignificantincreaseinyield

strengthofroughly45MPaisobserved.Theelongationtofracturestays highataround15%.Notethatasimilareffectonmechanicalparame- tersisachievedwhenusingpureAlinsteadofindustriallyproducedEN AW-6082asbasematerial(seeSupplementaryMaterial:TableS1and Fig.S1).

4.2. Processdesign

4.2.1. Adaptionofhomogenization– formationofdispersoids

Fig.5showsSEM-backscatteredelectron(BSE)micrographs,where all dispersoids containing Fe, Mn, Cr and Zr appear as bright dots (Fig.5a–d)afterone-step andtwo-step homogenizationtreatmentin theundeformedstatebeforehotrollingfor6082and6082-Zr.Thecor- respondingparticledistributionstatistics areshown inFig.5eandf.

InFig.5a-d.Thenumberdensitiesofallparticlesafterhomogenization showthefollowingtrends(seeFig.5f):(i)thenumberdensityincreases whenZrisadded;(ii)forbothalloysthenumberdensityincreaseswhen switchingfromone-totwo-stephomogenization;(iii)two-stephomoge- nizationappliedtothebasealloy6082generatesahighernumberden- sitythanstandardone-stephomogenizationofthe6082-Zralloy.Com- paredtothematerialcontainingZr,anincreasednumberofelongated particles(0–0.33roundness;seeSection2Experimentalmethods,foran explanationofthisterm)formsinthebasealloy6082,especiallyafter one-step homogenization.Withtwo-stephomogenizationthenumber ofroundparticles(withanevensmallerdiameter;seeFig.5e)ismore thandoubled.Generally,two-stephomogenizationleadstoanoverall decreaseinthediameterofallparticlesin6082(from116nmto85nm).

In6082-Zrthediameteroftheparticlesstaysonalevelcomparableto anyhomogenization(between70and80nm),especiallyinthemost importantfractionofroundish,smallparticles(0.66–1.00roundness).

However,thenumberdensityincreasessignificantlywithtwo-stepho- mogenization,asalsoreportedinRef.[63].Fig.5gandhshowthedis- tributionofparticlesafterhotrollingandsolutionannealingat570°C inT4state.Thenumberdensitiesarereducedinbothalloys.Inparticu- lar,thecompletedisappearanceofstronglyelongatedparticles(0–0.33 roundness)andastrongreductioninthesecondshareofparticles(0.33–

0.66roundness)isalsoapparentinboth.Whereasthemeandiameterof allparticlesincreasesfrom85to101nmin6082,noalterationinthe particlesizeisseenfor6082-Zr(Fig.5i).

(6)

Fig.5.SEM-BSEmicrographsofthemicrostructureindifferentstates(afterhomogenization– a–e;inT4-state– g–i)afterone-stephomogenization(a,c,g)and two-stephomogenization(b,d,h)ofthebasealloy6082(a,b,g)and6082-Zr(c,d,h).Meandiameter(e,i)andnumberdensities(f,i)ofalldetecteddispersoids arecategorizedinthreegroupsaccordingtotheirshape.

Fig.6. Stress–straincurvesfor6082and6082-ZrinT651stateandafterin- terruptedquenching.Inbothcasesanincreaseinyieldstrengthofabout40to 50MPaisobserved.

4.2.2. Interruptedquenching

Fig.6showsstressstraincurvesfor6082(onestephomogenization) and6082-Zr(two-stephomogenization)afterstandardAA(T651state) andinterruptedquenching+shortenedAA(160°C/2h).Thistreatment generatedanincreaseinyieldstrengthofabout40to50MPaforboth alloys.Obviouslythisprocessvariationnotonlystrengthensthebase alloy,butalso– almostequally– 6082-Zr,whichdemonstratesthatIQ andtheeffectofZradditionworksynergistically.

4.3. Optimalalloy

Afterdemonstratingthegaininstrengthcausedbyapplyingthein- dividual measurespresented,we now investigate whetherthese can be combinedtocreate anoptimalalloy. Themeasures consistof (i) amaximumdissolvableamountof SiadjustedtoMg,combinedwith theadditionof0.3wt%Cu;(ii)theadditionof0.2wt%Zr;and(iii) theapplicationofIQtosuppressboththeformationofnon-hardening phasesand,subsequently,thelossofhardeningpotential.Fig.7shows theresultsofthislastalloydesignstep.Comparedtothereferenceal- loy,thesumofmeasures(i)and(ii)generatesayieldstrengthincrease of73MPa.Interruptedquenchingfurtherincreasesstrengthby49MPa (lightgreendashedcurveinFig.7),resultinginasignificantlyincreased yieldstrengthof411MPa,ascomparedtocommonlyprocessedENAW- 6082referenceplatematerialwith289MPayieldstrength.Notethat

(7)

Fig.7. Demonstrationofanoptimalstrategyforhighstrengthplatematerial viastress–straincurves.Theminuteoptimizationofchemicalcompositionand modifiedindustrialprocessingcanboosttheyieldstrengthoftoday’sstate-of- the-artENAW-6082platematerialbymorethan40%.

theseeffectscanalsobeachievedusingpureAlasthebasematerial(see supplementarymaterial:TableS1andFig.S1).

4.4. Microstructureanalysis

4.4.1. 6082-Zr

Fig.8showstheresultsofSEM-EBSDmeasurementsofthebasealloy 6082and6082-Zr.Adepictionofthe3rdorderKAM(stepsize0.6μm, threshold5°)inFig.8aandbhighlightsthenoticeabledifferencebe- tweenthesetwoalloysaftersolutionannealingat570°C.Whereasonly littledeformationisstoredin6082aftersolutionannealing,misorien- tationisstillhighin6082-Zr.ThiscanalsobeseenfromFig.8cand d,whichshowsthebandcontrastforeachspecimen.Afurtherdepic- tionisfoundinthesupplementaryFig.S2aandb,whererecrystallized areasareshadedin blue.Thedistributionofallboundaries forthese twoalloysinFig.8eandffurtherconfirmsthesignificantdifferencein themicrostructurespresented.Nosharptransitionofsmallanglegrain boundaries(SAB)andhighanglegrainboundariesisevident.However, McQueendefinesSABuptoathresholdvalueof8°[5],Gottsteinuntil 15°[65].Exemplarilytakingthelattervalueof15°,thereisaroughly uniformdistributionin6082,whereasin6082-Zrthevastmajorityof grainboundaries(88%)aremeasuredatunder15°.

4.4.2. Optimalalloy

Fig.9showsthemicrostructureoftheoptimalalloyafterstandard AAcontainingseveraldifferenttypesofdispersoids.Twodistinctphase familiesarepresent:(i)phasescontaining Fe,MnandCr(Fig.9b–e) and(ii)phasescontainingZr(Fig.9g–i).Fig.9b–eshowsthatFe,Mn andCrareallmixedanddonotprecipitateontheirown,whichcon- firmstheseparticlesasAl(Fe,Mn,Cr)Si-phases.Incontrast,Zr(seeFig.9g andh)eitherformspureAl3Zr-precipitatesorasubstitutionofacertain amountofAlbySitakesplacetoform(Al,Si)3Zr(afewexamplesare indicatedbyredarrows).OfalldetectedphasescontainingZr,62%had adetectableamount ofSi(seeFig.9g–i).Fig.9ialsoshows thatZr- particlesareoftenattachedtoAl(Fe,Mn,Cr)Si-particles,whichindicates heterogeneousnucleation(forbettervisibilityonlyMnwasselectedin multi-elementEDXmapping).Notethatusingvarioustiltinganglesin theSTEMconfirmedthatthesetwophasesarelocatedoneachother (seesupplementarymaterialFig.S3).ThesizedistributioninFig.9fin- dicatesthattheZrparticlesaremuchsmaller(74±32nm)thanthe othertypeofdispersoid(134±72nm).Althoughthecumulativesum oftheFe,MnandCrcontent(0.38at%)ismorethansixtimeshigher

thanZr(0.06at%),thearealdensityof(Al,Si)3Zrismorethantwiceas high(7.52×1012m2comparedto2.97×1012m2).

Fig.10ashowsadarkfieldSTEMmicrographwithbrightparticles alongdislocationsandentangleddislocationsthattendtobestuckto theseparticles.Themulti-elementmappinginFig.10bshowsMncon- tainingaswellasZrcontainingdispersoidsandMgandSirichharden- ingphases.Heterogeneousprecipitationofthetwodispersoid-phasesis alsoobservedinthismicrograph(asinFig.9h),aswellasatendency towards theformationofthemainhardeningparticlesondispersoid phases(indicatedwithredarrowsinFig.10b).Besidesthenucleation onthephaseboundariesofsuchparticles,coarsenedphasescontaining MgandSiareobservedalongdislocations(indicatedbydashedyellow linesinFig.10b).ThisisdepictedinmoredetailinFig.10c–f.

Fig.11showsseveralSTEMmicrographsofasub-grainwithanori- entationclosetothe<001>-axisoftheoptimalalloy(indicatedbythe redcircleinFig.11a).InFig.11aandbmanybrightconstituents,identi- fiedaseitherZr-orMn-rich(Fig.11b),arelocatedexactlyattheborder ofthesub-graininthemiddleofthepicture(indicatedbytheyellowand redarrowsalongthegrainboundary).Thisindicatestheinterferenceof theseparticleswithsub-grainboundaries.Thedark-fieldmicrographof theoptimalalloyinFig.11crevealsthatmanydislocationsarepinned insidesub-grainsbydispersoids(seeredcircle).

5. Discussion

Astheabove datashows,alteringthechemicalcomposition(age- hardenableelementsMg,SiandCu;Zrasdispersoidformer)andadapt- ing the heat treatment (optimized homogenization and interrupted quenching) generated an exceptionally high yield strength of over 400MPainanoptimal6082-typeplatealloy.Inthefollowingwedis- cusstheunderlyingmechanismswhichcausedthisgaininstrength.We alsoprovideadetailedanalysisoftheroleplayedbyminorZraddition inmicrostructureevolution.

5.1. InfluenceofSiandCuaddition

AddingSiandCuisknowntocauseanincreaseinstrengthviasolid solutionstrengthening andprecipitationstrengthening[66]. Table 3 showsthemechanicalparametersofthealloystestedhereindifferent stateswhileundergoingstandardartificialaging.Rightafterquenching (“As-quenched”)onlyaminordifferenceinyieldstrengthismeasured.

Intheabsenceofanyhardeningphasesthisdifferencecanbeattributed solelytosolidsolutionhardeningcausedbydifferentSiandCucontent [10].However,accordingtothemodelofsolidsolutionhardeningby Leysonetal.[67]wecalculatearelativestrengthdifferenceoflessthan 1MPa tobecausedbythealterationsin6082-Siand6082-Cu[10]. Thereforethisincreaseintheas-quenchedstrengthisattributedtoclus- ters,whichformeitherduringquenchingorduringtheveryshortperiod oflessthan5minbetweenquenchingandtensiletesting[10,68].Con- cerningartificialaging,bothalloyvariationsshowanenhancedstrength responseofroughly20MPa.FromFig.2bitisqualitativelyseenthat addingSiandCugeneratesanincreasedvolumefractionofstrengthen- ingphasesat160°C(1.56vol%inthereference,1.63vol%in6082-Si and1.87vol%in6082-Cu),fromwhichanincreasedprecipitationpres- sureandthusanincreasedhardeningeffectcanbederived.

5.2. InfluenceofZraddition

AddingZrtothebasealloy6082results inastrengthincreaseof 28MPaintheas-quenchedstateafterhotrolling(6082-Zr).Artificial agingcausesafurtherincreaseinyieldstrength(17MPa),whichgen- eratesatotalincreaseof45MPainthefullyhardenedT651statecom- paredto6082.FurtheralloyingwithSi,MgandCuintheoptimalalloy showsanadditiveeffect(increasedyieldstrengthintheT651stateof 73 MPacomparedto6082).Toattributethese findingstothecorre- sponding mechanisms,adetailedinvestigation ofthemicrostructure,

(8)

Fig.8.ComparisonofthemicrostructureinT4temperaftersolutionannealing(570°Cfor20min)of6082(a,c,e)and6082-Zr(b,d,f).a,b)3rdorderkernel averagemisorientationuptoaKAMthresholdof5°.Boundariesabove5° areshowninblue,aswellaslargernon-indexedareas.(c,d)Depictionofthebandcontrast.

(e,f)Misorientationdistributionofallmeasuredboundaries;infosonthethresholdusedcanbefoundinthetext.

Table3

OverviewofstrengthvaluesofallalloystestedindifferentstatesforstandardAA.Datawereacquiredrightafter solutionannealingandquenching(as-quenched)andafterfinalartificialagingtreatment(allvaluesinMPa).

Solidsolutionhardeningiscalculatedaccordingto[67].Thecontributionduetoartificialagingiscalculatedas thedifferencebetweenas-quenchedstrengthandstrengthintheT651state.

As-quenched Artificial aging T651 (conv. AA)

R p0.2 Gain to the base alloy Strength gain Gain to the base alloy R p 0.2

Base alloy – 6082 58 ± 2 + 232 ± 3 289 ± 3

6082-Si 66 ± 2 8 ± 3 + 253 ± 8 + 21 ± 9 319 ± 8

6082-Cu 64 ± 2 7 ± 3 + 249 ± 3 + 17 ± 5 313 ± 3

6082-Zr 85 ± 0 28 ± 2 + 249 ± 5 + 17 ± 6 334 ± 5

Optimal alloy 90 ± 3 32 ± 3 + 272 ± 10 + 40 ± 11 362 ± 9

(9)

Fig.9. STEM-EDXmicrographsofthedispersoidlandscapeoftheoptimalalloyinconditionT651.(a)HAADF-overviewofthemicrostructure.(b-e)EDXmapping ofFeMnCr(b),Fe(c),Mn(d)andCr(e).(f)ParticledistributionofFeMnCr-phasesand(Al,Si)3Zr-phases.Thedash-dottedlinerepresentsthemeandiameter.(g–i) EDXmappingofZr(g),Si(h)andMnZr(i).TheredarrowsindicateparticlesthatcontainbothSiandZr.

whichconsideredthedispersoidsinparticular,wascarriedout.Bothof thealloyscontainingZrarediscussedbelow.

5.2.1. Formationofdispersoids

AddingZrleadstotheformationofAl3Zr.Herealmosttwo-thirds ofallprecipitatesshowthatasubstitutionofAlbySiforms(Al,Si)3Zr (seeFig.9i)intheoptimalalloy.ForboththeZr-freebasealloy6082 and6082-Zr,theoptimized2-stephomogenizationhasprovedeffective increatinga dispersoidlandscapewithanincreasednumberdensity anddecreasedaveragediameterofallparticles(seeFig.5).STEMmea- surements(seeFig.9h)haveconfirmedpreviousobservationsofhet- erogeneousnucleationofMn-andZr-richparticlesoneachotherinthe optimalalloy[64,69].

5.2.2. EffectofZronthemicrostructure

SEM-EBSDmeasurementsinFig.8revealcompletelydifferentmi- crostructuresinthebasealloyand6082-Zr.Thecauseliesinquitediffer- entdispersoidlandscapesandtheincreasednumberdensityofsmaller particlesin6082-Zr(seeFig.5)duetoboththeZradditionandanop- timizedhomogenizationstrategy.Itisknownthatsuchsecondphase particlesinterferewithmovinggrainboundariesandblocktheirmove- mentduringrecrystallization,andotherprocessesthatincludeabound- aryrearrangement[70].Dispersoids,suchasAl3Zrthathaveacoherent interfacewiththematrixarepostulatedtohavearetardingpressureon boundarieswhichistwiceashighasincoherentones.Neverthelessit isquiteremarkablethattheincoherentMn-dispersoidsdonotinhibit recrystallizationmoreeffectively(Fig.8c),eventhoughthereference alloyactuallycontainssufficientlevelsofMn (0.42wt%) toactwell againstgrainboundaryanddislocationmovements[20,71].Thestan- dardhomogenizationone-steptreatment,however,causestheforma- tionofcomparativelycoarse particleswithlownumberdensity.This

resultsinalesseffectiveretardingpressureagainstrecrystallizationin thebasealloy,thusrenderingtheMnadditionfairlyineffective.

5.2.3. Influenceonmechanicalparameters

Intheas-quenchedstate,6082-Zr,whichismodifiedbyZr-addition only,showsa strengthincrease of28 MPacompared tothebase al- loy.DispersoidstypicallycontributetoastrengthincreaseviaOrowan hardening,duetotypicalsizes,whichforeclosesthemtobecutbydislo- cations[10,35].StrengthcontributionsofZr-particlesaccordingtothe Orowanmechanismcalculatedaccordingto[9]aregiveninTable4, andareonlyofminorimportance.In6082-Zronly10MPaandinthe optimalalloyonly14MPaarecalculatedinadditiontothebasealloy.

AssumingthatparticlescontainingZrthatprecipitateheterogeneously atpresentMn-richdispersoids(countedinthecalculationfortheopti- malalloyseparately)donotcontributemuchtoincreasedstrength,this valueisreducedevenfurther.Consequently,theOrowanmechanism alonecannotexplaintheincreaseinstrengthcausedbyaddingZr.

Followingthestandardproductionrouteofthickplates,hotrollingat 540°Cwasappliedforallalloystested.Duetothehighstackingfaulten- ergyofAl,sub-grainscaneasilyformduringthisproductionstep.More severeconditions(increaseofdeformationspeeḋεanddecreaseoftem- peratureT)generateadecreaseinthesub-grainsize[5,73].Afterhot rolling,similarelongatedmicrostructures,indicatingnon-recrystallized areas,arepresentinboththereferencealloyandin6082-Zr(Fig.S4).

Asalreadydescribedabove,recrystallizationin6082-Zrissuppressed duringsolutionannealingduetothebeneficialdispersoidlandscapes, resultinginalargelyrecoveredmicrostructure.Wecanthereforeexpect sub-grain boundaryhardeningasanadditionalstrengtheningmecha- nisminthealloyscontainingZr.

Anevaluationofthemicrostructureintheoptimalalloyshowsan averagesub-grainsize(𝛿)ofroughly5𝜇minrollingdirection(equiv-

(10)

Fig.10. STEMmicrographsoftheoptimalalloyintheT651-state.(a)Darkfieldmicrographshowingdispersoidsalongwithentangleddislocations.(b)EDX-mapping ofMg,Si,MnandZrwhichshowstheheterogeneousnucleationofseveralphases(hardeningphasecontainingMg,Siondispersoidsindicatedbyredarrows)anda tendencytowardsprecipitationofthehardeningphaseondislocations(seedashedyellowline).Thiscanalsobeseeninmoredetailinc–f.

Fig.11. STEMmicrographsoftheoptimalalloyintheT651-state.Thegraininthemiddle(lightgreyishgrain,indicatedbyaredcircle)in(a)isnearthe<001>- axis.(a)HAADFmicrographshowingsmallsub-grainswithasizeofroughly3–5𝜇manddispersoids.(b)Multi-elementEDXmappingofthesub-grain(grainfrom (a),circledinred)inthe<001>-axis.Manyparticles(indicatedbyarrows)arelocatedrightontheboundary.(c)Dark-fieldmicrographshowingthedislocation distribution.Anentanglementofdislocationscanbeobservedinthevicinityofbrightparticles(redcircle).Sub-grainboundariesaremadeupofahighnumberof stackeddislocations.

alenttothetestingdirection).PuttingthisintoEqs.(1)and(2),which weredevelopedbyGinterandFarghalli[74],withaburgersvector(b)of 2.86×1010m,ashearmodulus(G)of25.4GPaandaTaylorfactor(M) of3givesastrengthincrease(Δ𝜎)of44MPaduetosub-grainboundary hardening[74].Takingintoaccountthat– dependingonthethreshold value(see4.4.16082-Zr)– 74%or88%ofthetotalboundariesaremade

upofSAB(Fig.8f),thisincrementisreducedtoavaluebetween33and 39MPa.Heterogeneousnucleationof50%ofallZr-containingparticles (Fig.9i)leadstoaconsiderablereductionoftheirhardeningcontribu- tion.ReducingOrowanhardeningtoabouthalftheamountwouldleave 33MPafortheoptimalalloy,whichisonly3MPahigherthaninthe basealloy.Thereforethiscontributionjudgestobenotsignificantand

(11)

Table4

SimulatedstrengthcontributionsduetoOrowanhardeningofthereference,6082-Zrandoptimalalloy(allvaluesinMPa).Calculatedaccordingto [10].

T651 (conv. AA) Total dispersoids strengthening (calculated) Input parameters

R p0.2 Al(FeMnCr)Si (Al,Si) 3Zr Diameter [nm] Volume fraction [m 3/m 3] a

Base alloy – 6082 289 ± 3 30 SEM ( Fig. 5 g)

101 ± 59

0.0090

6082-Zr 334 ± 5 40 SEM ( Fig. 5 h)

75 ± 47 0.0115 b

Optimal alloy 362 ± 9 23 21 STEM ( Fig. 9 b and d)

134 ± 72 (Mn) 74 ± 32 (Zr)

0.0093 (Mn) 0.0024 (Zr)

aThevolumefractioniscalculatedbyusingtheat%ofMnandZrandassumingthatallofitprecipitatesasAl12Mn2FeSi[72]orasAl3Zr[35]. Thereforethisvaluerepresentsthemaximumamount,becausenotallMnorZrwillprecipitateentirelyduringprocessing.

b Thevolumefractionofthedispersoidsin6082-ZristhesumoftheparticlescontainingMnandZr.Thereforethediameterusedheredoesnot differentiatebetweenthesetwodifferentphases.

Table5

Overviewofmechanicalparametersofallalloystestedforinterruptedquenching(allvaluesinMPa).

T651 (conv. AA) Interrupted quenching + shortened AA (2 h/160 °C)

R p0.2 R p0.2 Strength gain

Base alloy – 6082 289 ± 3 349 ± 8 60 ± 9

6082-Si 319 ± 8 364 ± 4 45 ± 9

6082-Cu 313 ± 3 355 ± 6 42 ± 7

6082-Zr 334 ± 5 384 ± 7 50 ± 8

Optimal alloy 362 ± 9 411 ± 4 49 ± 10

themeasuredincreaseinyieldstrengthlistedinTable3,of28MPain 6082-Zrand32MPaintheoptimalalloy,ismostlyattributedtosub- grainboundaryhardening(as-quenchedstrengthgainwithoutcontri- butionofartificialaging).AdditionalalloyingwithSi/MgandCudoes notsignificantlyinfluencetheas-quenchedstrength;seeSection5.1.

𝛿𝑏=10⋅(𝜏 𝐺

)−1

(1)

Δ𝜎=𝑀𝜏 (2)

Nowweconsiderthestrengthgaincausedbyartificialagingofthe alloyscontainingZr,6082-Zrandtheoptimalalloy.Table3showsen- hancedvaluesduetoartificialaging(+17MPa)for6082-Zrcompared tothebase alloy.Becausethenumberofage-hardenable elementsis notalteredin6082-Zr,anadditionalmechanismisproposedhere.We notethatduringthe14daysofnaturalagingaplasticpre-deformation of2%wasapplied,whichgeneratedastrain(dislocation)hardeningof approximately39MPa.TheresearchofGruberetal.[75]showedthat duringagingat185°Casignificantreductionofthedislocationdensity takesplaceinleanAl–Mg–Sialloys.Suchsofteningcanalsobeassumed tooccurduringartificialageing(160°C/14h)of thebase alloy.Be- causethemodifieddispersoidlandscapepresentin6082-Zriscapable ofpreventingdislocationmotion(seeFig.11),thereforepreventingan- nihilationofdislocations,astrengthdecreaseduetoareductioninthe dislocationdensityisatleastpartlysuppressed.Thesameappliestothe optimalalloy.

Summingup theindividualstrength gainsachievedbyadding Si (6082-Si,21MPa),Cu(6082-Cu,17MPa)andZr(6082-Zr,17MPa), amaximumincrement of roughly55MPa shouldbe reachedduring artificialagingintheoptimalalloy,providedthesemechanismsactad- ditively.However,itisseenthatbothdispersoids(Fig.10b)anddisloca- tions(Fig.10d–f)actasheterogeneousprecipitationsitesintheoptimal alloy.Ontheonehand,hardeningparticlesprecipitatedondispersoids donotcontributemuchtotheoverallstrength,butdecreasethehard- eningpotential.Ontheotherhand,itcanbeassumedthathardening particlesnucleatingondislocationstendtocoarsenfaster,probablydue topipediffusionalongdislocations.Itisseenthatthesizeofparticles differsdistinctlyaccordingtowhethertheylieonornexttodisloca-

tionsinFig.10d–f.Consequently,thehardeningpotentialisweakened asparticlesgrowbigger.Someof theCucontentisalsoincorporated into(Al,Si)3Zr[26,76],whichlowerstheCucontentavailableforpre- cipitationformationaswell.Thisexplainswhythemeasuredgaininthe optimalalloycausedbyartificialagingcomparedtothebasealloysis actuallyabitlowerthan55MPa,showinganincreaseinyieldstrength of40MPa.

5.3. Interruptedquenching

ThethirdmeasureforincreasingthestrengthofAl–Mg–Sialloys,in- terruptedquenchingduringsolutionannealing,alsoprovedsuccessful (formechanicalparametersseeTable5).Whenquenchingthealloysto anelevatedtemperature(180°C)andslowlycoolingthemtoRT,which simulatesthecoolingofthickplates(roughly100mmthickness)inam- bientair,aminimumstrengthincreaseof42MPacouldbeseeninall alloyscomparedtotheconventionalT651state.Previousinvestigations showedthatunfavorableclustersformduringRTstorage,causedbythe highlevelofsuper-saturatedsolutesandquenched-invacanciesatthe endofsolutionannealing.Thisdecreasestheartificialagingresponse [77,78].Byapplying interruptedquenchingweexploited theadvan- tageousstateofthealloyrightattheendofsolutionannealingtopro- motetheformationofdenselydistributednucleifor𝛽’’ordirectlytrans- formableprecursors[13].Thisprocessingmethodworksforanyalloy variationandpromptedafurtherincreaseinstrength,finallyreaching ayieldstrengthof411MPaintheoptimalalloy.

6. Conclusions

This study investigates measures for significantly increasing the strengthofanENAW-6082alloy,whichisprocessedaccordingtoan industrialschemefortheproductionofthickplates.Themostimportant findingsaresummarizedasfollows:

• Athermodynamic simulation-basedadjustmentof age-hardenable elementsgeneratesastrengthincreaseof~ 40MPa.

• Interrupted quenching (IQ)can increase thestrength by at least 42MPa.

Referenzen

ÄHNLICHE DOKUMENTE

To form inlet and outlet channels as well as a space for the cooling water in the cylinder heads (so-called water chamber), sand cores are suspended in the casting dies.

Many factors like the temperature, grain size, CRSS value and texture can affect the yield strength at which dislocation slip or twinning occurs. The uniaxial tensile tests and

- analysis of the structure of Al-Mg-Si and Al-Mg-Ge casting alloys in as-cast state together with its change during solution treatment and aging;.. - determination of

1) During SHT, 6xxx series alloys undergo recrystallisation and precipitation of Si-rich precipitates within a short SHT time. Afterwards, dissolution of previously

The clustering kinetics in quenched pure Al, Al-Mg and Al-Si, as well as in pure or Cu/Ge- containing Al-Mg-Si alloys during natural ageing have been studied by applying

Nowadays the Al-Mg-Si alloys are the most important group of age hardenable aluminum alloys commercially [1,19]. They are widely used in cast, wrought and extruded form for

In this paper, transmission electron microscopy (TEM) and atom probe tomography (APT) have been employed to char- acterise precipitates microstructures formed during ageing a

The grain size of the material should not increase significantly during homogenization as this would decrease the extrusion properties. In figure 4.13 the microstructure