Effect of texture on the residual stress response from laser peening of an aluminium-lithium alloy Zabeen, S, Langer, K & Fitzpatrick, ME Author post-print (accepted) deposited by Coventry University’s Repository Original citation & hyperlink:

Zabeen, S, Langer, K & Fitzpatrick, ME 2018, 'Effect of texture on the residual stress response from laser peening of an aluminium-lithium alloy' Journal of Materials Processing Technology, vol 251, pp. 317-329 https://dx.doi.org/10.1016/j.jmatprotec.2017.07.032

DOI 10.1016/j.jmatprotec.2017.07.032 ISSN 0924-0136 Publisher: Elsevier NOTICE: this is the author’s version of a work that was accepted for publication in Journal of Materials Processing Technology. Changes resulting from the publishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in this document. Changes may have been made to this work since it was submitted for publication. A definitive version was subsequently published in Journal of Materials Processing Technology, [251, (2017)] DOI: 10.1016/j.jmatprotec.2017.07.032 © 2017, Elsevier. Licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International http://creativecommons.org/licenses/by-nc-nd/4.0/ Copyright © and Moral Rights are retained by the author(s) and/ or other copyright owners. A copy can be downloaded for personal non-commercial research or study, without prior permission or charge. This item cannot be reproduced or quoted extensively from without first obtaining permission in writing from the copyright holder(s). The content must not be changed in any way or sold commercially in any format or medium without the formal permission of the copyright holders. This document is the author’s post-print version, incorporating any revisions agreed during the peer-review process. Some differences between the published version and this version may remain and you are advised to consult the published version if you wish to cite from it.

Effectoftextureontheresidualstressresponsefromlaserpeeningofanaluminium-lithiumalloy

S.Zabeen1,3,K.Langer2,andM.E.Fitzpatrick1

1CoventryUniversity,FacultyofEngineeringandComputing,GulsonRoad,CoventryCV12JH,UK2AirForceResearchLaboratory,Wright-PattersonAirForceBase,OH45433,USA

3Formerlyat:TheOpenUniversity,WaltonHall,MiltonKeynesMK76AA,UK

Email:Suraiya.Zabeen@coventry.ac.uk

Abstract

Lasershockpeeningcanimprovethedamagetoleranceofmetallicmaterialsbyintroducingdeep compressive residual stress that inhibits crack initiation and growth. This studyinvestigatesthelaserpeeningofaluminiumalloyinaproductform–anextrudedT-section– thathasdifferentcrystallographic textures indifferent locations.Thealloystudied isAl2099,analuminium-lithiumalloythatshowsanisotropyinthemechanicalpropertieswhentextureispresent.Specimensextractedfromdifferentregionsoftheextrusionwerelasershockpeenedwithapowerdensityof3GW/cm2 in single shocksaswell as inapatternarrangement. The effect of number of laser shocks on the residual stresses werecharacterized primarily using incremental hole drilling. The results show 20% higherresidual stresses in theweb area of the extrusion compared to the flange after peeningwithasinglelasershock,withthisdifferencedecreasingasthenumberofshocksincreases.Thiseffectcanbeexplainedbythedifferenceinyieldstrengthbetweenthoselocations.Nosignificantdifferenceswereobserved intheresidualstresses frompeeningontodifferentplanesofatexturedsampleatagivenlocation.

1. IntroductionAl 2099 alloy is a third-generation aluminium-copper-lithium alloy used for aerospacestructuralapplications,particularly inairplane internalstructureandlowerwingstringers.The lithium content in an Al-Li alloy increases the specific properties over traditionalaluminium alloys. For example, compared to the commercially-used non-lithium Al 2024alloy, Al 2099 shows a 5% reduction in densitywith 20% increase in longitudinal tensileyieldstrength(Giummarra,Thomas&Rioja2007).Incommonwithmanyaluminiumalloys,Al 2099 shows significant in-plane and through-thickness anisotropy, particularly in therolledproducts, andaxisymmetric flowanisotropy in theextrudedproducts (Rioja1998).Generally,thisanisotropyresultsfromthestrongcrystallographictexturethatformsduringmultistagethermomechanicalprocessing.Al2099extrudedproductsusuallyexhibit<111>+ <100> fibre texture that gives a high longitudinal yield strength, with only a slightdifferencebetweenthetensileyieldstrengththeultimatetensilestrength,andlowductility

inthetransversedirection(Denzeretal.1992).Bois-Brochuetal.(Bois-Brochuetal.2014)studiedtheeffectoftextureontheanisotropyofthemechanicalpropertiesofanAl2099extruded integrally-stiffened panel. They found a variation in mechanical propertiesbetweendifferentlocationsofthepanel.Inalllocations,thehigheststrengthwasfoundinthe extrusion direction and the lowest strength was observed at 45° to the extrusiondirection.Reductionsintheyieldstrengthof17%and25%werereportedforthetransverseand 45° directions, respectively. This anisotropic behaviour drives the design minimumstrength to the45°direction, therebyneglecting thestrengthbenefits inotherdirectionsandtakingbenefitonlyforweightsavings.

Mechanical surface treatments, suchas laser shockpeening (LSP), applied in appropriateorientationswithrespecttotheextrusiondirectionmayprovebeneficialtoimprovingthedamagetoleranceofanisotropicalloys.Lasershockpeening introducesdeepcompressiveresidualstressesthatcanenhancethefatiguelifeofamaterial.Themagnitudeanddepthoftheinducedresidualstressesdepend,amongotherfactors,onthemechanicalpropertiesandtextureofthevirginmaterial.Thelevelofinducedresidualstressisdirectlyrelatedtotheyieldstressofthematerialalongwiththehardeningresponse.

Peening on an alloy having a significant anisotropy inmechanical properties, not only indifferentlocationsbutalsoindifferentorientations,requiresathoroughunderstandingofthe influence of texture on the residual stress response due to peening. Although asubstantialamountofresearchdataisavailableontheeffectsoftextureonthemechanicalproperties ofAl 2099, very little attentionhasbeen focusedon (1) the effects of LSPonresidualstressgeneration in thisalloysystem,and (2) theeffectof initial textureonLSP-induced residual stresses inanymaterial.Therefore, this study investigates theeffectsoftextureonresidualstressprofilesinlaserpeenedAl2099.

2. MaterialandExperimentalMethod

Al2099alloywasusedintheT8condition(solutionheattreated,stretched,andartificiallyaged) tostudy theeffectsof textureonresidual stressgenerationowing toLSP.Materialwas received fromAlcoa Inc. in the formof an extrudedT-barwith gross cross-sectionaldimensionsof139×60mm2andalengthof1000mm.Figure1showstheT-bargeometryandthelocationoftheregionsthatwerefertohereasthewebandflange.Testcouponswith different geometries were extracted from different locations to evaluate texture,tensilestrength,andresidualstressafterlasershockpeening.

FivetypesofspecimenswereextractedfromtheextrudedTbarfortexturemeasurements,tensiletesting,singlelocationpeening,andmultiplepatternpeening:

• Ninespecimenswithdimensionsof10×10×5mm3wereextractedfromtheweb

and flange sections of the extruded T-bar to study the texture variation at these

locations.

• Eight longitudinal tensile specimens with geometry as shown in Figure 2 were

extracted, four from theweb and four from the flange,with the longitudinal axis

alignedwiththeZ-directionoftheextrusion.

• 18 samples for residual stress characterization following single peening on XY, YZ

andZXplanesinthreepeeningconditions,wereextracted,ninefromtheweband

ninefromtheflange.ThespecimengeometriesaredescribedinTable1.

• Four specimenswere used to study the effect of pattern peening on the residual

stressgenerationXYandZXplanes,withthespecimengeometrieslistedinTable2.

• OnespecimenwasusedtomeasurethevariationinhardnessacrosstheXY-plane.

Thisspecimenwasa10-mm-thickT-shapedsliceoftheextrusion.

Figure1:SpecimengeometryoftheextrudedAl2099T8T-barshowingthewebandflangelocations.Alldimensionsareinmm.

Figure2:Specimengeometryusedfortensiletesting.Alldimensionsareinmm.M12threadswereusedforfixingthesamplesintothetestmachine.Thedimensionofthegaugelengthis15mm.

2.1 Texturemeasurementusingneutrondiffraction

BulkcrystallograhictexturewasmeasuredusingneutrondiffractionattheGeneralMaterialsDiffractometer(GEM)instrumentattheUK’sISISneutronsource(Kockelmann,Chapon&Radaelli2006).Abeamsizeof20×20mm2wasused.Therecordeddatawerenormalizedtotheincidentneutronfluxdistribution,correctedfordetectorefficiencies,andconvertedtoadiffractionpattern(intensityasafunctionofd-spacing).ThediffractionpatternswerethenRietveldfitted(Wenk,Lutterotti&Vogel2010)inMAUDsoftware(Lutterottietal.1997).The(111),(200)and(220)polefigureswereplotted.AnexampleofthefittedspectraisshowninFigure3.

Figure3:NeutrondiffractionspectrafittinginMAUDsoftwareforthe2099material

2.2 HardnessandTensileStrengthTests

Ahardnessmapwasgeneratedfromthefullcross-section(XYplane)oftheTbarasshowninFigure1.ThesurfacewasmetallographicallypolishedtoOPSfinishforVickershardnessmeasurements.Indentationwascarriedoutwithaloadof5kgF.

Tensiletestswerecarriedoutona50kNInstron3369universaltestingsystem,followingASTMstandardE8/E8M-13a

2.3 Lasershockpeening

EighteencouponswithdimensionsasshowninTable1wereextractedfromthewebandflangesectionsoftheextrudedbarandpeenedatMetalImprovementCompany(Earby,UK)usingthepeeningparametersandgeometrieslistedinTable1.A5×5mm2laserspotwithanintensityof3GW/cm2wasused,withpeeningatthegeometriccentreofeachtestcouponasshowninFigure4a.Formultiplelasershocks,eachpeenspotwasapplieddirectlyoverthepreviousspot.

FourcouponswithdimensionsasshowninTable2werepattern-peenedtoinvestigatetheeffectsofshotoverlapinamoreconventionalpeenpatternoveralargerarea.Twospecimenswereextractedfromeachofthewebandflangelocations,withonespecimenfromeachlocationpeenedontheXYplaneandtheotherontheZXplane.Allspotswere5×5mm2withanintensityof3GW/cm2andapulsewidthof18nanoseconds.Eachlayerofspotswasformedwiththeindividualspotsappliednominallyedge-to-edge(a5%overlapwasusedbetweenthespotswithineachlayer).Threelayerswereusedwitha50%offsetbetweeneachlayer(equivalentto300%coverage).

Table1:SinglelocationPeeningconditions

SpecimenNumber

Extractionlocation

Peenedplane

PeeningconditionSingleSpot(NotationisGW/cm2–ns–numberoflaserhits)

Specimendimensions/mm3

1a

Flange

YZ 3-18-1

50×50×25

2a XY 3-18-13a ZX 3-18-11b YZ 3-18-32b XY 3-18-33b ZX 3-18-31c YZ 3-18-52c XY 3-18-53c ZX 3-18-5

4a

Web

YZ 3-18-1 60×25×155a XY 3-18-16a ZX 3-18-14b YZ 3-18-35b XY 3-18-36b ZX 3-18-34c YZ 3-18-5

5c XY 3-18-56c ZX 3-18-5

Table2:MultiplePatternPeeningconditions

2.3.1.1.1.1 SpecimenNo

2.3.1.1.1.2 Extractionlocation

2.3.1.1.1.3 PeenedPlane

2.3.1.1.1.4 Specimendimensions/mm3

7C2.3.1.1.1.5 W

eb 2.3.1.1.1.6 XY

2.3.1.1.1.7 26 cimendim

2.3.1.1.1.8 8C 2.3.1.1.1.9 Z

X2.3.1.1.1.10 2

6 cimendim

2.3.1.1.1.11 7R

2.3.1.1.1.12 Flange

2.3.1.1.1.13 XY

2.3.1.1.1.14 50 ngeendim

2.3.1.1.1.15 8R 2.3.1.1.1.16 Z

X2.3.1.1.1.17 5

0 ngeendim

Figure4:(a)singlelocationpeenedspecimens,(b)patternedpeenedspecimens

2.4 ResidualStressMeasurementbyIncrementalHoleDrilling

Incremental hole drilling is a relatively fast, straightforward, and inexpensivemethod forresidual stressmeasurement in the laboratory. In this techniquea smallhole isdrilledormilledintothespecimen,whichcauseselasticstrainrelaxationasmaterialisremoved.Theelasticstrainrelaxationcausesachangeindisplacementinthesurroundingmaterialthatismeasured by a strain gauge attached to the specimen surface. The pre-existing residualstressesarethencalculatedfromthemeasureddisplacements.

Holedrillingmeasurementswerecarriedoutusingaset-updevelopedbyStresscraft,UK.Foraccuratemeasurements,theUKNPLGoodPracticeGuideNo.53andASTM837standardwerefollowed(ASTME837−13a2013,Grant,Lord&Whitehead2006).

A2-mm-diameterholewasmilledinanorbitalmotionwithfourdepthincrementsof32µm,fourdepthincrementsof64µm,andeightdepthincrementsof128µmforatotalof16incrementsandatotaldepthof1.4mm.StraingaugerosettessuppliedbyVishaywithpartnumbersCEA-13-062UL-120andEA-13-062RE-120wereused.

2.5 ResidualStressMeasurementbylaboratoryX-rayDiffraction

Residualstressesarealsocharacterizedbysin2ψmethod(Noyan,Cohen1987).Thismethodmeasuresthein-planenear-surfacestresses.ThemeasurementparametersarelistedinTable3.

Table3:LaboratoryX-raydiffractionparametersusedforresidualstressmeasurements

X-rayTube CrK-α

Wavelength/nm

0.229107

2-Theta/˚

139.3

{hkl}

311

TiltAngles/˚

±45

TiltOscillation/˚

5

MeasuringMode

Modifiedχ

CollimatorDistance/mm

11.9

CollimatorDiameter/mm 3

2.6 FiniteElementAnalysis

Finite element modelling was conducted to predict the stress and deformation statesresultingfromlaserpeening.ThecommercialcodeAbaqus/EXPLICIT6.14-1wasusedforallsimulations.Eachimpactpulsewasmodelledusingtwodistinctsolutionstepssothatthecomputationalparameterscouldbeadjustedforimprovedefficiencyandconvergence.The

firstsolutionstepstartswiththeinitialapplicationofthepressurepulseandendswhennofurtherplasticdeformationoccurs.ThesecondsolutionstepintroducesRayleighdampinginto the model, to return the system to a state of near-equilibrium, and lasts until thekineticenergyissufficientlydampedout.Onepairofanalyseswasperformedforeachlaserpulse.

Becausestrain-rateeffectsonthematerialhardeningwerenotavailableforthisalloy,asimpleisotropichardeningmodelwasused.Itshouldbenotedthatthisassumptioncouldleadtooverpredictionofcompressivestressesandunderpredictionofthecompensatorytension(Langeretal.2015).

Thepressureprofilefromthelaserpulsewasestimatedusingtheone-dimensionalmodelproposedbyFabbroetal.(Fabbroetal.1990):

𝑃𝑚𝑎𝑥 𝐺𝑃𝑎 = 0.01 𝛼2𝛼+3 𝐼0

𝐺𝑊𝑐𝑚2 𝑍 𝑔

𝑐𝑚2𝑠 (2)

where𝛼isacorrectivefactorcorrespondingtothefractionofinternalenergycomprisedofthermalenergy(typicallyabout0.25)(Bertheetal.1997);𝐼0isthepeakpowerdensityofthelaser;and𝑍istheeffectiveacousticimpedanceoftheinterfacebetweentheablativecoatingandthetransparentoverlay,definedby:

2𝑍 =

1𝑍𝑐𝑜𝑎𝑡𝑖𝑛𝑔

+ 1𝑍𝑜𝑣𝑒𝑟𝑙𝑎𝑦

(3)

Here𝑍𝑐𝑜𝑎𝑡𝑖𝑛𝑔and𝑍𝑜𝑣𝑒𝑟𝑙𝑎𝑦aretheacousticimpedancesoftheablativecoatingandthetransparentoverlay,respectively.Withwaterastheoverlay(𝑍 ≈ 0.15 × 10!g cm!!)andaluminumtapeasthecoating(𝑍 ≈ 1.7 × 10!g cm!!),𝑍hasavalueofabout 0.3 × 10!g cm!!.TheFWHMofthepressurepulsewastakentobethreetimestheFWHMofthelaserpulse,followingtheobservationsin(Peyre,Fabbro1995).

3. ResultsandDiscussions

3.1 Mechanicalproperties

Figure5bshowstheVickersmicrohardnessmapforacross-sectionoftheextrudedT-bar.Theresultclearlyshowsahardnessvariationwithinandbetweenthewebandtheflangesections.Intheflange,ahardenedsurfacelayerwithathicknessof2mmatthetopsurfaceand1mmatthebottomsurfaceisobserved.Theseouterregionsareapproximately18%harderthanthecentralsectionoftheflange.Thehardnessofthewebsectionisabout140HV5(VickersHardness),whichissimilartothehardnessoftheskinlayerintheflange.

Figure 5: (a) Specimen Geometry of the extruded bar showing the axis system used; (b) correspondingVickersmicrohardnessmapinXYorientation.

Figure 6 shows tensile test results for specimens extracted from theweb and flangesections,with gauge lengthparallel to theZ-axis as shown in figure5a.Results fromfourspecimenswereaveraged.Specimensextractedfromtheweblocationshow38%higher yield stress than the specimens extracted from the flange section (Table 4).Higher UTS is also found in the web compared to the flange. A comparison ofmechanicalpropertiesbetweenthewebandflangespecimensisshowninTable4.Alsoplotted in Figure 6 the stress-strain curve for a conventional Al-Cu alloy forcomparison.

Table4:MechanicalpropertiesofAl2099alloy

Material Elastic Modulus/GPa

Yield Strenght σy / MPa

E/σy

Ultimate Tensile

Strength/MPaElongation/%

Al-2099 Web 70 575 122 617 8.6Al-2099 Flange 70 414 169 536 8.6

Al-2624 T39 70 460 152 487 15

Figure 6: Engineering stress-strain curves for the web and flange material, plus data from Al 2624 forcomparison(sourceofdata?)

3.2 MacrotextureandMicrostructuralevolution

Preferredcrystallographicorientationortextureresultsinanisotropicmechanicalpropertiesinpolycrystallinematerial.Al-Liextrusionproductsshowin-planeandthrough-thicknesstexturevariation,andalsoatexturedifferencebetweenthewebandflangesectionsofextrudedT-bar(Hales,Hafley1998).

Figure7presentsthereconstructed(111),(200)and(220)polefiguresforthespecimensextractedfromdifferentlocationsintheextrusionasshowninFigure5a.Specimens1,2and3(extractedfromtheweb)showtypical<111>fibretexture,consistentwithwhathasbeenreportedintheliterature(Denzeretal.1992,Jata,Singh2014).Theintersectionoftheflangeandthebaseoftheweb(specimen5)hasarolling-typetexture.Aweakrollingtypetexturecanalsobeobservedinspecimens6and7.

Figure7:(111),(200),and(220)polefiguresforspecimens1to7asshowninFigure5.

3.3 SurfaceProfileoftheSinglePeenedSpecimens

Figure8aandb show the surfaceprofilesof single-spotpeenedweb (specimen5b) andflange (specimen 2b) specimens, respectively (with conditions 3GW cm–2 for 18 ns, andthree sequential shocks, 3-18-3, onto the XY plane). An uneven depression from thepeening is evident in both specimens. This may be a result of a non-uniform energydistribution across the laser spot. This phenomenon is not uncommon in laser peening(Dormanetal.2012,Hfaiedhetal.2015,Toparli2012).Forthisreason,intypicalindustrial

applications,multiplelayersofpeeningwithoverlapoftheshotsisusedtoachieveamoreuniformsurfaceprofileaswellasresidualstress.

A horizontal line profile through the centre of the peen spot was extracted from thecontourdataandcomparedwithsimilar lineplots frompeeningontheYZandZXplanes(Figures8candd).Amaximumdepthofabout30-40µmisobservedforthewebspecimen,with a pile up at the peen boundary of about 15-25 µm, as shown inFigure 8c. Similarresults are found for the flange specimens,Figure 8d. The difference in the depressiondepthbetweenXY,YZ,andZXorientationsisverysmallforbothwebandflangelocations.

Figure 8: Surface profile for single-location peened specimens for (a) and (c) the web, and (b) and (d) the flangelocationsafterpeeningat3-18-3;(c)and(d)comparethesurfacelineprofilesforpeeningontheXY,YZ,andZXplanes.

3.3.1 Comparisonwiththenumericalprediction

Figure9presentsacomparisonbetweenpredictedandexperimentalsurfaceprofilesforthesingle-locationpeenedspecimenswith3-18-3peeningconditions.Thepredictedsurfaceprofilefortheflangecompareswellwiththeexperimentalsurfaceprofile,withamaximumdepressiondepthofabout40µm.However,fortheweb,themodelpredictsalowerdepth(20µm)comparedtothemeasuredsurfaceprofile.Thehigheryieldstressusedtomodelthewebresponsegeneratedalowerdeformationdepththanisobservedexperimentally.Itisalsotobenotedthattheprofilesofthepeenedspotvarysignificantlybetweenthepredictionandtheexperimentalresult.Duringlaserpeeningthematerialundergoeshigh-strain-ratedeformation,andthetypeofhardeningexperiencedbythealloycannotbepredicted(Langeretal.2016).Whilstthemodeldidnottakeaccountofthe

hardeningandassociatedhigh-strain-rateeffects,thedepressiondepthwaspredictedrelativelyaccuratelyfortheflange.Itcanbeconcludedthatstrainhardeningplaysanimportantroleinthesoftermaterialwithlowerstrength.

Figure9:Comparisonofsurfaceprofilesofsinglelocationpeenedspecimen(experimentallymeasuredandpredicted).

3.4 ResidualStress

Figure10aand10bcomparetheresidualstressdistributionsasafunctionoforientationforthe 3-18-1 peening condition for specimens extracted from the web and flange,respectively.Two in-planestresscomponentsweredetermined.Afterasingle lasershockexposurethenear-surfaceresidualstressesinthewebspecimensareintherangeof–75to–200MPa(Figure10a);underthesamepeeningconditions,themeasuredstresses intheflange specimens are in the range of –100 to –150MPa (Figure 10b). A similar level ofsurface stress is reported in (Hfaiedh et al. 2015) for single location peening of a similaraluminium-lithiumalloy,Al2050-T8,withanenergyof3.5GW/cm2for10nsusinga1.5mmdiameter circular spot; however, the depth at which balancing tensile stress appears isreportedas675µm,whereasforoursamplesatthisdepth,asshowninFigure10a,–200MPastressisstillpresent.Thismaybeduetotheshorterpulselength–andhencelowerenergy–thattheyused.Theredoesnotseemtobeasignificantvariation intheresidualstressesbetweentheXY,YZ,andZXorientationsineithertheflangeorwebspecimens.

Figure10:Comparisonofresidualstressprofilesbetween(a),(c),(e)theweband(b),(e),(d) the flange after laser peening with one, three and five successive shock exposuresrespectively at apeeningenergyof3GW/cm2. Thedifference in residual stressbetweenpeeningontheXY,YZandZXplanesofthespecimensarealsoshown.

Figure10cand10d showsimilar residual stressmeasurements for specimensexposed tothree layers of peening (the 3-18-3 condition). No noticeable difference seems to beevidentintheresidualstressprofilesbetweentheorientationswithintheboundsofscatterin the data, in both the web as well as the flange. The maximum compressive residualstressesforthewebandtheflangeare–330MPaand–300MPa.intherangeof–240to–290MPa.However,theσxandtheσycomponentsofstressforpeeningontotheXYplaneareslightlyhigherinmagnitudethanthecorrespondingstresscomponentsintheothertwoorientations.WhilepeeningontheXYplane,the laser-inducedshockwaveactsalongthe

extrusiondirection,whichhasthegreateryieldstrengththanthetransverse(X)direction.Thismaybethereasonwhyahighervalueofresidualstressesisobservedforpeeningonthis plane. In the flange section little differences in the residual stress were observedbetweenXY,YZandZXplaneduetolowertexture.

ResidualstressdistributionsafterfivesuccessivelasershocksareshowninFigures10eand10f for theweb and flange specimens, respectively. Themaximum compressive residualstressinthewebisabout–350MPa,withslightlylesscompressionobservedintheflange(about –300 MPa). Following 5 hits the residual stresses remained largely unchangedshowingstrainhardeningsaturation.someeffectsoforientationareobserved inthewebspecimens, with no noticeable effects observed in the flange. The residual stressesobtainedfromincrementalholedrillingareverifiedwiththeresultsfromlaboratoryX-raydiffractionasshowninTable5.Agoodcorrelationexistsbetweentheresultsfromthetwotechniques.

Table5:ComparisonofresidualstressesmeasuredbyholedrillingandlaboratoryX-raydiffractionforspecimenspeenedwith3-18-5

SpecimenIdandpeenplane

LocationResidualstress:IncrementalHoleDrilling/MPa

Residualstress:X-raydiffraction/MPa

σx σY σx σY

1CYZ Web –223 –212 –195 –222 2CXY –272 –249 –236 –229 3CZX –198 –248 –218 –268 4CYZ Flange –275 –208 –264 –214 5CXY –151 –236 –163 –232 6CZX –232 –206 –267 –250

Figure 11 presents the FEA simulated residual stress results for the 3-18-3 peeningcondition. To model the residual stresses in the web and flange, the correspondingexperimentally-determined yield strengths of 575 and 414 MPa were used. ComparingFigures11a and 11b, it is evident that a relatively uniform layer of surface compressiveresidualstress(–375MPa)isgeneratedinthewebascomparedtotheflange.Abalancingtensile stress of 100 MPa appeared for both specimens. The extent of the residually-stressed region in theweb is slightly smaller than in the flange inbothwidthanddepth.Thisobservation is furtherclear inFigure11c,whichshows thatalthough in theweb thepredictedvalueofmaximumcompressive residual stress isonly30MPahigher thanwaspredicted in the flange, the depth at which themaximum compressive residual stressesappeararesignificantlydifferent:0.2mmfor theweband1mmfor the flange.Asimilardifference can also be seen in the location of balancing tensile stresses. This is simplybecause the higher yield strength of the web material causes a resistance to plastic

deformationduring the shockwavepropagation. Thedepthsof the compressive residualstressesforthewebandflangeare2.5and3.2mm,respectively.Notethatonlyasimpleisotropichardeningmodelwasusedforprediction.

Figure11:FEApredictionof residual stress forpeening conditionof3-18-3: (a)weband(b)flange,(c)thelineprofilesextractedfromthe2Dplotsandcomparedwiththeresidualstressesforpeeningat3-18-1.

3.5 Comparisonofmeasuredandpredictedresidualstress

Figure12aand12bshowthecomparisonbetweenpredictedandexperimentalresidualstressprofilesinthewebandflangespecimens.Thefiguresalsocompareresidualstressesgeneratedafter1and3shocks.Inagreementwiththeexperimentalresults,lowerlevelsofresidualstressarepredictedfor1shock.Byincreasingthenumberofshocksfrom1to3,approximately100MPahighercompressiveresidualstresseswereintroducedforbothlocations.

Figure12:Comparisonofpredicted(bluelines)andmeasured(greenlines)residualstressesafter1and3shocks,for(a)weband(b)flangelocation.FromFigure12,itisclearthatthemodelpredictionofadifferenceofaround70MPainthesurfacestressfor1and3shocksisdifferentfromthatseenexperimentally,wherethenear-surfacestressesarewithin20MPafor1and3hits.Betteragreementisseenforthemodellingofthesingleshock.Thereissomereverseyieldingnearthesurface,thepredictionofwhichrequiresdetailedknowledgeofthecyclichardeningwhenthreeshocksareused.Thisisthesubjectoffurtherworkontherefinementofthemodelperformance.

3.6 Effectofnumberofpeenlayers

Figure 13 shows the effect of the number of peen layers (the number of shocks at eachlocation) on the residual stress distribution for peening onto the XY orientation for (a) aweb specimen and (b) a flange specimen. For the flange specimen, the surface residualstress wasmeasured to be about –100 to –125MPa, regardless of the direction of thestress component or the number of layers. However, considering the subsurfacedistribution of residual stress, the magnitude of the maximum compression increasessignificantlyas thenumberofpeen layers increases.Themaximumvalueof compressionafter3lasershotsreachedabout–275MPaatadepthof300μm.

The effects were more pronounced in the web specimens, with significantly highercompressionafterthreeshocksthanafterone.Thesurfaceresidualstressesafteroneandthree layers of peening were about –140 and –255 MPa, respectively; the maximumcompressiveresidualstressafteroneandthree layersofpeeningwereabout–250and–375MPa,respectively.Thevaluesofthecompressiveresidualstressesaregreatlyincreasedwith increasing number of peen layers; however, as is shown, the overall shape of theresidualstressesprofilesaresimilar.

Figure14 shows a linear increase in thedepthof the surfacedepression (peak-to-valleydistance) after peening as the number of layers is increased.No difference between thepeeningoftheflangeandwebregionscanbedetected.It is interestingtonoteherethat

even though there is a strength variation between web and flange, no variation in thedepressiondepth isobserved.Also,the linearrelationshipbetweennumberof layersandthedepressiondepthdoesnotcorrelatewiththemeasuredchangesintheresidualstress.

Figure13:Effectofnumberofpeenlayersontheresidualstressprofiles in(a)flangeand(b)web.

Figure14:Effectofnumberoflayersonthedepressiondepthforsinglelocationpeeningat3GW/cm2

3.7 Residualstressesfrompatternpeening

SurfaceresidualstressesweremeasuredbylaboratoryX-raydiffractionforaspecimen,asshowninFigure1,thatwaspeenedovertheentireXYsurface.TheresultsarepresentedinFigure15aand15b for theσxandσycomponents, respectively.Theσxcomponentof theresidualstressintheflangeis lesscompressivethanthestressmeasuredintheweb.Thiscanbeattributedsimplytothefactthatthewebhasahigheryieldstressthantheflange:i.e.,itisnotnecessarytoinvokethedifferencesseeninthetextureinordertoexplaintheresult. Also note that for both the web and flange sections, the σx component of the

residualstressisgenerallylesscompressivethantheσycomponent.Lineplots(Figure15cand15d)areextractedfromthe2DresidualstressmapsfrompeeningontheXYplaneandcomparedtothesurfaceresidualstressesmeasuredfrompeeningontheZXplane.ThereisnosignificantdifferenceinresidualstressbetweenpeeningontheXYandZXplanes,withtheresidualstressesbeingfairlyscatteredoverthecentral18mmofthewebsection.Themagnitudeofthemaximumcompressiveresidualstress(about–300MPa)generatedinthewebsectionfortheσycomponentisofthesameorderofthecompressiongeneratedintheflange section,but theσx component is lower in the flange sectionataround–200MPa.Balancingtensileresidualstresseswerefoundattheedgesofbothwebandflange.

Because the laboratory X-ray measurements only provide near-surface residual stress,incremental hole drilling was used to characterise the sub-surface residual stress. AcomparisonofresidualstressfrompeeningontheXYandZXplanesispresentedinFigure16. For theweb specimens (Figure 16a) amaximum compression of about –300MPa isgeneratedatadepthof290μm.FortheXYorientationtheσxstresscomponenthasslightlylowerstressthantheσycomponent. Intheflangespecimens, thestressesresultingfrompeening on the XY plane are about 15% less compressive than those generated frompeeningontheZXplane.ThisresultisconsistentwiththeX-raymeasurements.

Alalloysgenerallyshowstrainhardeningeffects.Whenthespecimenswerepeenedwithasingleshockexposure,acleardifferenceinresidualstresswasobservedbetweenthewebandflange.Intheweb30%higherresidualstresses(calculatedbyaveragingresidualstressvalues inalldirectionsat288µmdepth)aregenerated than the flange,whichcorrelateswellwiththefactthatthewebhas38%initialhigheryieldstrength.Itisinterestingtonotethat the differencedisappears for 3 or 5 shock exposures at the samedepth. Themostlikely reason for this is thatmultipleshockshavecausedstrainhardeningsaturation.Theresults from the pattern-peened specimens reconfirm this fact, as minimal difference isobserved between theweb and flange, with the only exception being the stresses frompeeningontheXYplane.

Figure 15: Surface residual stresses for pattern peened specimens (a) σx componentfollowingpeeningontheontheXYplane, (b)σycomponent followingpeeningontheXYplane; line profile of stress with position along the X-axis in the (c) web and (d) flangeshowing the comparison residual stresses between XY and ZX plane. The uncertainty intheseresidualstressresultsliewithin±25MPa.

Figure 16: Depth-resolved residual stress profile measured by incremental hole drillingshowing the comparison between two orientations in the (a) web and (b) flange. Theuncertaintyintheseresidualstressresultsliewithin±15MPa.

Conclusions

Laserpeeningwasappliedtodifferentlocationsofanal2099extrudedT-section,tostudytheeffectsoftextureontheresidualstressgeneration.

1. The extruded Al 2099 T‒section shows a variation in strength between different

locations, with the web having 38% higher yield strength than the flange. This

correlateswiththedifferentcrystallographictexturesseenacrossthesection,witha

<111>+<100>fibretextureinthewebsections.

2. After laser peening with single shots onto the same location, the residual stress

induced in theweb isabout25%morecompressive than in the flangespecimens.

However, the difference in induced residual stresses is greatly minimized after

multiple laser shocks owing to strain hardening and strain hardening saturation.

Thedifferencesinresidualstressesforidenticalpeeningindifferentlocationscanbe

linkedtothedifferenceinbaselinemechanicalpropertiesthatarisesinthedifferent

regionsoftheextrusion.

3. Nosignificantvariationintheresidualstresscouldbedetectedbetweenpeening

ontosamplesfromdifferentorientationswhenastrongtextureispresent.

4. Agoodagreementwasfoundbetweentheexperimentalandnumericallypredicted

residualstressprofiles;however,themodelscouldnotaccuratelypredictthe

surfacedeformation.Thisislikelybecausethenumericalmodelidealizesthe

distributionofenergyacrossthelaseranddoesnotaccountfornonuniformity.

5. Patternpeeninginthewebandflangesectionsinducedsimilarlevelsand

distributionsofcompressiveresidual,suggestingthatthebeneficialeffectsof

layeringcanovercomevariationsinthemechanicalpropertiesofagivenalloyasa

resultofthermomechanicalprocessing.

Itcanbeconcludedthatlaserpeeningofformedaluminiumproductsmusttakeintoaccountoflocaldifferencesinproperties.Forgingsandextrusionswillhavedifferenttexturesandyieldstressindifferentlocations,andwillthereforehavedifferentlevelofinducedresidualstress.Thedifferenceinresidualstressescanberationalisedsolelyintermsoftheelastic-plasticproperties:noeffectsoftexturepersewereobserved.

Acknowledgements

This researcheffortwas sponsoredby theAirForceOfficeof ScientificResearch,AirForceMaterialCommand,USAF,undergrantnumberFA8655-12-1-2084,and theAirForce Research Laboratory’s Aerospace VehiclesDirectorate. TheU.S. Government isauthorized to reproduce and distribute reprints for Government purposenotwithstandinganycopyrightnotationthereon.Theviewsandconclusionscontainedherein are those of the authors and should not be interpreted as necessarilyrepresenting theofficialpoliciesorendorsements,eitherexpressedor implied,of theAirForceOfficeofScientificResearchortheU.S.Government.TheauthorswouldliketothankDrMarkusHeinimannatAlcoa Inc. for theprovisionof thematerial studied inthe project. Thanks are also due to Dr Andy Fitch at ESRF, Grenoble, France; DrWinfried Kockelmann, ISIS, UK; Mr Pete Ledgard and Mr Stan Hiller at The OpenUniversity, and Dr PhilipWhitehead at Stresscraft, UK for technical support. MEF isgrateful for funding from the Lloyd’s Register Foundation, a charitable foundationhelping to protect life and property by supporting engineering-related education,publicengagementandtheapplicationofresearch.

4. ReferencesASTME837−13a2013,StandardTestMethodforDeterminingResidualStressesbytheHole-DrillingStrainGageMethod1,ASTMInternational,100BarrHarborDrive,POBoxC700,WestConshohocken,PA19428-2959.UnitedStates.

Berthe,L.,Fabbro,R.,Peyre,P.,Tollier,L.&Bartnicki,E.1997,"Shockwavesfromawater-confinedlaser-generatedplasma",JournalofAppliedPhysics,vol.82,no.6,pp.2826-2832.

Bois-Brochu,A.,Blais,C.,Goma,F.A.T.,Larouche,D.,Boselli,J.&Brochu,M.2014,"CharacterizationofAl–Li2099extrusionsandtheinfluenceoffibertextureontheanisotropyofstaticmechanicalproperties",MaterialsScienceandEngineering:A,vol.597,no.0,pp.62-69.

Denzer,D.K.,Hollinshead,P.A.,Liu,J.,Armanie,K.P.&Rioja,R.J.1992,"TextureandPropertiesof2090,8090and7050ExtrudedProducts,Ibid.(II)(1992),pp.903.",VIInternationalAluminium-LithiumConference,vol.ii,pp.903-909.

Dorman,M.,Toparli,M.B.,Smyth,N.,Cini,A.,Fitzpatrick,M.E.&Irving,P.E.2012,"Effectoflasershockpeeningonresidualstressandfatiguelifeofclad2024aluminiumsheetcontainingscribedefects",MaterialsScienceandEngineering:A,vol.548,no.0,pp.142-151.

Fabbro,R.,Fournier,J.,Ballard,P.,Devaux,D.&Virmont,J.1990,"Physicalstudyoflaser-producedplasmainconfinedgeometry",JournalofAppliedPhysics,vol.68,no.2,pp.775-784.

Giummarra,C.,Thomas,B.&Rioja,R.J.2007,"Newaluminumlithiumalloysforaerospaceapplications.InProceedingsofthelightmetalstechnologyconference.",lightmetalstechnology,.

Grant,P.V.,Lord,J.D.&Whitehead,P.2006,TheMeasurementofResidualStressesbytheIncrementalHoleDrillingTechnique-Issue2,TheNationalPhysicalLaboratory(NPL).

Hales,S.J.&Hafley,R.A.1998,"TextureandanisotropyinAl-Lialloy2195plateandnear-net-shapeextrusions",MaterialsScienceandEngineering:A,vol.257,no.1,pp.153-164.

Hfaiedh,N.,Peyre,P.,Song,H.,Popa,I.,Ji,V.&Vignal,V.2015,"Finiteelementanalysisoflasershockpeeningof2050-T8aluminumalloy",InternationalJournalofFatigue,vol.70,no.0,pp.480-489.

Jata,K.V.&Singh,A.K.2014,"Chapter5-TextureandItsEffectsonPropertiesinAluminum–LithiumAlloys"inAluminum-lithiumAlloys,eds.N.E.Prasad,A.A.Gokhale&R.J.H.Wanhill,Butterworth-Heinemann,Boston,pp.139-163.

Kockelmann,W.,Chapon,L.C.&Radaelli,P.G.2006,"NeutrontextureanalysisonGEMatISIS",PhysicaB:CondensedMatter,vol.385–386,Part1,no.0,pp.639-643.

Langer,K.,Olson,S.E.,Brockman,R.A.,Braisted,W.R.,Spradlin,T.&Fitzpatrick,M.E.2015,HighStrain-RateMaterialModelValidationforLaserPeeningSimulation,iet.

Lutterotti,L.,Matthies,S.,Wenk,H.-.,Schultz,A.S.&Richardson,J.W.1997,"Combinedtextureandstructureanalysisofdeformedlimestonefromtime-of-flightneutrondiffractionspectra",JournalofAppliedPhysics,vol.81,no.2,pp.594-600.

Noyan,I.C.&Cohen,J.B.1987,ResidualStress-MeasurementbyDiffractionandInterpretation.Springer-Verlag,Berlin.

Peyre,P.&Fabbro,R.1995,"Lasershockprocessing:areviewofthephysicsandapplications",OpticalandQuantumElectronics,vol.27,no.12,pp.1213-1229.

Rioja,R.J.1998,"FabricationmethodstomanufactureisotropicAl-Lialloysandproductsforspaceandaerospaceapplications",MaterialsScienceandEngineering:A,vol.257,no.1,pp.100-107.

Toparli,M.B.2012,AnalysisofResidualStressFieldsinAerospaceMaterialsAfterLaserPeening,TheOpenUniversity.

Wenk,H.-.,Lutterotti,L.&Vogel,S.C.2010,"RietveldtextureanalysisfromTOFneutrondiffractiondata",PowderDiffraction,vol.25,no.03,pp.283-296.