Determination of current maps by svet of hot-dip galvanized steel under simultaneous straining.

DETERMINATION

OF

CURRENT

MAPS

BY

SVET

OF

HOT-DIP

GALVANIZED

STEEL

UNDER

SIMULTANEOUS

STRAINING

S.M.

Manhabosco

a

,

R.J.C.

Batista

b

,

S.

Neves

da

Silva

a

,

L.F.P.

Dick

a,1,

*

a

Lab.deProcessosEletroquímicoseCorrosão,Depto.Metalurgia,UniversidadeFederaldoRioGrandedoSul(UFRGS),Av.BentoGonçalves9500,91501.970 PortoAlegre,Brazil

b

Depto.deFísica,UniversidadeFederaldoOuroPreto(UFOP),CampusMorrodoCruzeiro35400-000OuroPreto,Brazil

ARTICLE INFO Articlehistory:

Received15January2015

Receivedinrevisedform19March2015 Accepted23March2015

Availableonline25March2015 Keywords:

Hot-dipgalvanizedsteel SVET

Strain Currentmap

ABSTRACT

Anewexperimentalprocedurewasusedtoanalysethecorrosionbehaviourofhot-dipZncoatedsteel

basedontheassociationofthescanningvibratingelectrodetechnique(SVET)anduniaxialtensilestrain

appliedupto3.1%in0.01MNaClsolutionswithoutreleaseoftheelasticstrainduringthetest.The

nucleationoflocalizedcorrosionsitesonthecoatingoccursatverylowstrainvaluesaroundtheyield

pointandthemaximumcurrentdensitiesincreasecontinuouslywiththeincreaseoftheappliedstrain.

ThenucleationoflocalizedcorrosionontheZnsurfacewasfavouredbytheruptureofthepassivefilmin

contactwiththesolutionbytheactionofslipstepsandfineintergranularcracks,ratherthanbythe

exposureofthesteelsubstrate.StrainingoftheZncoatingimmersedinthesolutionwascomparativelya

muchmoreaggressiveconditionthanstrainingthesampleinairbeforethecorrosiontests.

ã2015ElsevierLtd.Allrightsreserved.

1.Introduction

Hot-dipzinccoatings(HDZn)onsteelshavebeenwidelyused in the automotive industry, due to the excellent corrosion resistance achieved, low cost and an acceptable ductility that enablesdeepdrawingandothercoldforgingprocessesessentialto thistypeofindustry.Forthispurpose,HDZncoatingsareusually appliedonveryductilesubstrates,suchastheinterstitial-free(IF) steels.Goodformabilityofthesteelsheetcoatingisrequiredin ordertoreduce crackingand tomaintaintheadherenceof the coatingtothesteelsubstrate[1],andthereforetheinfluenceof strainingonthecoatingadhesionisaveryimportantparameter. TheinfluenceofstrainingonthecorrosionofZn-coatedsteelshas usuallybeenstudiedbythepreviousdeformationofthecoated sheet,whichisonlylaterexposedtocorrosiveenvironments[2_–4]. Vaggeet al.[3]observedthat asdeformationincreasesthereis moredelaminationofthehot-dipgalvanizedsteelsheetcoating. AccordingtoCulcasietal.[4],theharmfulin_fluenceofstrainonthe corrosion of HD Zn coatingsincreases as the grain size of the

coatingdecreases,duetothecrackingofthegrainboundariesof theFe-Zncoating.

After the introduction of the scanning vibrating electrode technique (SVET) in corrosion studies by Isaacs [5–7], several authors have used this technique to study the corrosion of artificiallyproduced,modelFe-Zngalvanicpairs[8–12]orofthose producedinthecut-edgecorrosionofhot-dipcoatings[13–16],as well asthecorrosionof Zn-electroplatedsamples[17,18].Other MicroelectrodetechniquessimilartoSVEThavealsobeenusedin these studies for the determination of ionic species on the corroding surface,suchasscanningelectrochemicalmicroscopy (SECM) [9–10] and the scanning reference electrode technique (SRET)[19].Microelectrodetechniqueswererarelyassociatedwith thestrainingofZncoatedsheets,andtheywerealwaysperformed afterapreviousdeformationofthesamples,asinthecaseofHDZn coated [19], electrogalvanized [20] or electrogalvanized and additionallycoatedwithZnrichprimer[21].

Despite the importance of zinc coatings as a corrosion protection method for steels that will suffer intense straining, there are relatively few reported studies on the influence of straining onthe corrosion of Zncoated steels. These corrosion studieswerealwaysperformedafterstrainingandreleasingthe elastic strain before immersion of the sample in the solution [4,19–21].Thecorrespondingstrain(

e

)valuesbyfarexceededthe onesappliedinthisstudyaswellastheaggressivityoftheused testsolutions.Culcasietal.[4]andBerchemetal.[19]applied10% straintoHDZngalvanizedsteelbeforetheexposuretorespectively

* Correspondingauthor.Tel.:+555133087537;fax:+555133089469. E-mailaddresses:[email protected](S.M. Manhabosco),

[email protected](R.J.C. Batista),[email protected](S.N.daSilva),

[email protected](L.F.P. Dick).

1

ISEmember.

http://dx.doi.org/10.1016/j.electacta.2015.03.162

0013-4686/ã2015ElsevierLtd.Allrightsreserved.

ContentslistsavailableatScienceDirect

Electrochimica

Acta

0.05M[4]and3%NaCl[19],whileBastosetal.[20]usedvaluesof 9% and 23% before the exposure of electrogalvanized steel to 0.1molL
1NaClsolutions.Inallcases,thepreviousappliedstrain increasedthecorrosion,mostprobablyduetothefractureofthe coating at high

e

values and exposure of the substrate before immersioninchloridesolutions.

Inthepresentworkwepresentexperimentalresults introduc-ing the use of the scanning vibrating electrode technique associatedwith insitu simultaneous elasticstraining of HD Zn coatedsteel,performedimmediatelyafterapplyingplasticstrain tosamplesimmersedin0.01MNaClsolutions.

2.Experimental

The corrosion experimentswerecarried out onsamplescut froma0.7mmthickinterstitialfreesteel(IF)sheetproducedand hot-dip galvanized by USIMINAS, Brazil. The galvanizing was carriedoutbyimmersioninamoltenZnbathcontaining0.2wt% Alataround450_{Cinacontinuousprocessadjustedtoanominal}

coating thickness of 10

_m

m controlled by “N2-knifes”.

The composition of the IF steel is listed in Table 1, and is characterizedbyaC contentofonly0.0019wt%. Microstructure and chemical composition of the Zn-based coating were determined by SEM (scanning electron microscopy) and EDS (energy dispersive X-ray spectrometry) analysis of sample cross-sections in a 5800-JEOL-NORAN SEM. To increase the precision, EDS analysis was performed using metal standards and electron beam current measurement using a Faraday cup constituted of a grounded graphite sample provided with a 100-

_m

mhole.

Flatspecimensforuniaxialtensiletestswerecutfromprovided steelsheetsaccordingtoASTME8/E8M-09[22]withdimensionsas showninFig.1a.Asmallsized(Ø=12.3mm,h=5.1mm)“barnacle” electrochemicalcell with an electrolyte volumeof 600

m

L was attachedtothecoatingsurface.Goodadhesionofthebarnaclecell tothecoatingduringthestrainingtestwasachievedusingflexible cellsofpolypropyleneandlow-hardnessfastcureepoxyglue.An areaof approximately 1 mm2 was exposed to the electrolyte, delimitedwiththeuseof65

_m

mthick3MScotchtape.Thesample wasmountedonasimpledeviceforuniaxialdeformation(Fig.1b). Thebarnaclecellwasactuallygluedtothetape-coveredsurface and neitherdetachmentof celland tape or perceptiblecrevice corrosionwereobserveduptostrainvaluesof

e

4%(0.04).

Thecurrentdensitymap(j-map)acquisitionswerecarriedout ina 0.01molL
1NaClsolutionwitha measuredconductivityof 1.205mS. For this, an Applicable Electronics Inc. equipment controlled by theASET software(Science Wares Inc.) was used. Thevibratingmicroelectrodeforj-determinationswasaPtwire fromMicroProbesInc.withasphericaltipofØ=10

m

m.Toincrease thesensibilityofj-measurements,newPt

_m

-electrodeswerefirst platinized black in a (0.1gL
1 Pb(CH2COOH)2



3H2O+10gL
1

H2PtCl6



6H2O)-solutionbyapplyingacurrentof
0.2

m

Afor180

s and then
1.2

m

A for 30s, followed by 15 intermittent 1s depositions with
1.2

m

A.The integrityof the

m

-electrode was regularly controlledin anoptical microscope. Whennecessary, the Pt

_m

-electrode was recovered by repeating the second platinization step. The distance between microelectrode and sample surface was always 50

m

m, and the scanned areawas approximately 1mm1mmin thecase of j-maps,while j-line scans were acquired in a length of 1mm. Two Pt-Ir auxiliary electrodeswereadditionallyusedduringtheSVETmeasurements as a reference and for grounding, respectively. The scanning direction was always parallel to the horizontal axis of the presentedSVET-maps.

Thesequenceoftestswasthefollowing:Afterdelimitatingthe measuring window with tape and adapting the barnacle cell, samplesweremountedinthestrainingdeviceandexposedtothe electrolyte. j-line scans were immediately acquired on the unstrainedsamplesfollowedbyj-mapsmeasurements.Thereafter, sampleswereprogressivelyplasticallydeformedinstepsofaround

De

=0.8% alwaysimmersedinthetest solutionand j-linescans wereimmediately measured along the direction of theapplied tensilestrainfollowedbyj-mapsmeasurementswithoutremoving theappliedstress.Thismeansthatj-linescansandj-mapswere measured on samples plastically deformed immersed in the electrolyte and without relieving the elasticstrain. In order to control strain values the distance between scratches or other markswasmeasuredinthetensiledeformationdirectionunderan optical microscope. These marks were located outside the measuring window at distances within 1mm. The line scans consistedof300pointswithatotalacquisitiontimeof200s,while the j-maps, respectively, of a grid of 3030 points with an acquisitiontimeof480s.Thesettingofanewstrainvaluetook between30and60s.Therefore,theprogressivestrainstepswere appliedattimeintervalsofaround15min.Afteratotal

e

-valueof 3.1%theexperimentwasinterruptedandthesurfacecleanedfor

Table1

Chemicalcompositionoftheinterstitialfree(IF)steel,wt%.

Al C S Mn P Ti Nb Si B N

IFSteel 0.084 0.0019 0.0093 0.257 0.028 0.003 0.008 0.008 0.0022 0.0026

SEMinspectionofthecorrodedarea.Thesameprocedurewasused tomeasureOCPtransientsandcyclicvoltammograms(CV)after straining.Experimentswereperformedintriplicate.

3.ResultsandDiscussion

TheSEMmicrographsofFig.2showthetopviewoftheHDZn coating (2a) and its cross section (2b). Cross sections suffered during grinding and polishing corrosive attack due to strong galvaniccouplingassociatedwiththeremovalofpassivelayerson the Zn coating by the mechanical action. This explains the porositiesofthecoating,asseenonthecrosssectionofFig.2b. The Zncontentdetermined by EDS using metalstandards was 93.0wt%(87at%)ata distanceof1.2

_m

mfromthesteel/coating interfaceandincreasedto98.2wt%(92.5at%)forgreaterdistances fromthesteel/coatinginterface,andisthuslowerthanexpected forthecompositionoftheouter

_h

-phase(99.97wt%Zn).Thiscan be attributedto thepresence of

d

and

z

-phases (FeZn10 and

FeZn13)andclosertotheoutersurface,possiblytotheocclusion

ofdrossparticlesfromthemetalbath,increasingthemeasured [Fe]/[Zn] ratio.The coating had an apparent grainsize around 100

m

monthesurfaceplane(

h

-phase)andameanthicknessof 9.80.3

_m

m.

Current density line scans measured on the surface of unstrainedHDZnsamplesin0.01molL
1NaClshowverystable valueslowerthan3

_m

Acm2_{,whichisinthesensibilityorderofthe}

SVETmethod,asverifiedbymeasuringjonnonconductiveflat surfacesunderthesameconditions.Thestrainwasthenappliedto the immersed samples increasing in steps progressively from

e

=0 to 0.6; 1.7; 2.2 and 3.1%. The corresponding j-line scans showed very irreproducible oscillations that could only be observedimmediatelyafterstrainingthesample.Theamplitude of j oscillations increased with the applied strain as seen for examplefore=1.7%ande=3.1%inFig.3.Thejvaluesalternated from anodic to cathodic at distances ranging from 100

m

m to 10

_m

m. These valuesare in theorder oreven smallerthan the expectedlateralresolutionofSVETundertheusedexperimental conditions,whichislimitedbythedistancebetweenmicroprobe and sample surface (_50

_m

m). The time interval between two consecutive adjacent j measurement points was around one s. Thus, the observed j oscillations must be interpreted as time oscillations rather than the occurrence of stable neighbouring anodicandcathodicsites.

Possibleexplanationsfortheincreaseinthejoscillationbetween anodicandcathodicvalueswithincreasingapplied

e

are:

Fig.2.SEMmicrographsofhot-dipZncoatinga)topviewandb)crosssection.

Fig.4.OCPtransientsmeasuredonprogressivelystrainedhot-dipcoatedsamples comparedtotheOCPonunstrainedonesin0.01molL
1_{NaCl.Thearrowsindicate}

themomentwhenthestrainwasincreasedandtheverticaldashedlines,the beginningofaj-mapmeasurement.

Fig.3.Currentdensitylinescansmeasuredonhot-dipZncoatingin0.01molL
1 NaClforevaluesofzero,1.7and3.1%(300points).

(a)Fractureofthewhole

h

-Zncoatingandexposureofthesteel substrate, withformation ofcathodic sitesonthesteel and anodiconesonZncoating;

(b)Ruptureofthepassivefilmonthesurfaceoftheouter

h

-Zn phaseduetotheintersectionoftheslipstepswiththesurface, andformationofanodicsitesonfreshlyexposedZnsurfaces andofcathodiconesonpassiveZn;

(c)Cracks formedonthe

h

-Znsurfaceongrain boundariesand inside grains,whichstopbeforecrossingthe

_z

and

_d

-phases forming galvanicelementsbetweentheZn-richerand more activeouter

h

-phaseandtheunderlying

z

and

d

-phases. Fig.4showsOCPtransientsmeasuredafterimmersioninthe chloridesolutionofunstrainedandprogressivelystrainedsamples. OCPoftheunstrainedsampleisstableuptoaround2500sata valuearound–700mVSHE,whichisdeterminedbytheZn/Zn2+

electrode potential (EZn2þ/Zn=
762mV SHE+29mV log [Zn2+])

and, most probably, by the potential of the oxygen reduction

reactionatthepartiallypassiveZnsurface.Afteraround6000s OCPdropsslowlyto
730mVSHEindicatingaslowactivationof thesurfaceexposedtothechloridesolution.However,whenthe sampleisstrainedwith

e

=0.6%OCPdropsrapidlyto–770mVSHE and then shifts slowly back towards the values observed for unstrainedsamples.Forprogressivelyhigher

e

valuesOCPdrops rapidlydowntovaluescloseto–790mVSHE,shiftingbacktothe values observedonunstrainedsampleswithin1,000 to1,500s. The negative shiftof OCP caused by the appliedstrain can be explainedbytheactivationofthesurfaceduetotheruptureofthe passivefilmandnot bytheexposure ofmorenobleunderlying phases,asFe-richerintermetallicsoreventheIFsteelsubstrate that would shift the corrosion potential towards the positive direction.Indeed,Culcasietal.[4]observed,forHDZncoatings10% strainedbeforeexposingto0.05MNaCl,corrosionpotential(Ecor)

shiftsinthenegativedirectionuptoapproximately60mVdueto slip steps formed by stretching. However, when the coating crackedduringrolling,Ecorwenttothepositivedirection[4,19].

Fig.5. CurrentdensitymapsofHDZncoatedsteelin0.01molL
1NaClfora)e=0.0%,and1minexposure,b)e=0.0%,and60minexposure,c)e=0.6%,d)e=1.7%,e)e=2.2%and f)e=3.1%.Thejmapsc),d),e)andf)wereconsecutivelymeasuredafterjmapsa).Theblackarrowindicatesthedirectionoftheappliedstrain.

Afterapplying

e

=3.1%,detachmentofthebarnaclecelloccurred within a certain period of time, which explains the final OCP oscillationsintheexperimentofFig.4.

ItisalsopossibletoseeinFig.4thattheOCP-transientsduring theirreturntomorepositivevalueshaveapparentlyoneortwo inflectionpoints.Thisindicatesthattherepassivationprocessof freshlyexposedsurfacesofthe

h

-Znisnotacontinuousprocess involvingonlyonemechanism.Onepossibilityistheformationof more than one Zn compound on the surface during the repassivation.Indeed,it hasbeereported[23–25]that inmoist

atmospheres enabling electrochemical reactions amorphous-Zn (OH)2 initially forms and than recrystallises to

b

-Zn(OH)2 or

dehydrates to ZnO, while in the presence of chlorides the hydroxichloride Zn5(OH)8Cl2 is formed [26]. Moreover, on Zn

electrodeposits Zn7(OH)12Cl2 and

e

-Zn(OH)2 were found after

anodicpolarizationindiluteCl
solutions[25].Theageingofthe ZnhydroxidesandhydroxychloridestoZnOtoformamorestable passive film couldexplain the differentrates of OCP variation duringthereturntomorepositivevalues.

After300safterstrainingandmeasuringthefastj-linescans, OCP hassufferedmostof itsvariation due totherepassivation process(Fig.4),allowingthemeasurementofmorestable_j-maps. Thetimefortheinitiationofeachj-mapisindicatedbyavertical dashed line in Fig. 4. The corresponding j-maps show that for unstrainedsample (Fig.5a),jliesinitially below8

m

Acm
2 and after 1h exposure to the solution even repassivation of localized corrosion sites are observed (Fig. 5b). However, for increasing

e

values, localized anodic and cathodic sites can be identified(Fig.5c–f).Themaximumcathodicandanodicjvalues extractedfromj-mapsforincreasingappliedstrainvalues(Fig.5f) areplottedinFig.6fortwoindependentmeasurementsforwhich nocelldetachmentcouldbeobservedupto

e

4%.Itispossibleto observethatmaximumanodicandcathodiccurrentdensitiesjmax

increaseasthestrainisincreased,differenttowhatwasobserved with unstrainedsamplesexposed for thesame time. Thus, the occurrence of anodic and cathodic sites in Fig. 5b–e can be attributed to the rupture of passive films on the Zn coating. The slopeof |jmax|versus applied

e

(Fig.6)is roughlyequal for

anodic and cathodicvalues, |

@

jmax/

@e

|70

m

Acm
1%
1 and the

extrapolation to i=zero lies at

e

0.20–0.25%. This minimum

Fig.6.Maximumanodicandcathodiccurrentdensities,jmax,ofthehot-dipZn

coatedIFsteelimmersedin0.01molL
1NaClversustheappliedstraine.

strainvalueis possiblyrelated tothe elasticstrainlimit ofthe sample,atwhichslipstepsbegintoformonthecoatingsurface. Itshouldbeobservedthatsamplesdeformedupto

e

=3.1%inair adaybefore immersionin the0.01molL
1NaClsolutionshow afterexposurejvalueslowerthan10

_m

Acm
2,asshowninFig.7 aftera)10,b)1800andc)3600sexposureinthechloridesolution. Themaximumjvaluesfor“dry”pre-strainedsamples(Fig.7)are slightlyhigherthanforunstrainedsamples,howeverwellbelow thevaluesobservedwhen

e

wasappliedtotheimmersedsample and the j-map measured without releasing the elastic strain. Additionally, the density of localized corrosion sites is much higher,when

e

isappliedtotheimmersedsample.Thisshowsthat corrosiontestswithpre-strainedsamplesarelessaggressivethan in situ straining due to the increased nucleation of localized corrosionwhenslipstepsfracturethepassivefilmincontactwith thechloridesolution.

AninspectionoftheHDZncoatingaftertheexperimentsupto

e

=3.1% (Fig. 8a, b) reveals that the steel/coating interface roughnessincreased during thecorrosion+deformation experi-ment, indicating that coating and substrate were plastically deformed,asseeninthecrosssectionsofFig.8acomparingthe interfacesbefore(Fig.8a,bottom)andafterexperiments(Fig.8a, top).Thisisinaccordancewith

e

valuesappliedabovetheyield point.Althoughthe upperpartofthecoating was damagedby preparationofthecrosssections,itcanbeseenfromFig8athat cracks do not cross the whole coating down tothe substrate. Hence,theobservedcathodicsitesarenotassociatedwithexposed steelsurfaces.

Thetopviewofthesampleshowsanapparentattackofgrain boundaries(bigblackarrowsinFig.8b),whichiscompatiblewith themediumgrain sizeof theouter

_h

-Znphase in theorderof 100

m

m(Fig.2a).Thecrackingofgrainboundariesofpreviously strainedHDZncoatingsandtheircorrosionattackafterexposure wereobservedbefore[4].However,severalsmallercracksinside the100

m

m-grainsarealsovisibleseparatedbymeandistances rangingfrom5to10

m

m(smallblackarrowsinFig.8b).Inaddition, slipstepsarevisibleonthetopofthecoatingaftertheexperiments (whitearrowsinFig.8b).Thesefinerdefectsmaybeassociated withtheinitialjoscillationsbetweenanodicandcathodicvalues observedinthenon-stationaryj-linescans(Fig.3).

To verify the influence of plastic and elastic strains onthe localizedcorrosion oftheZncoating,CVcurveswith20mVs
1 wereadditionallyacquiredin0.01MNaClfordifferentstraining

conditions (Fig. 9).Verylow current densitiesare observedfor unstrained samples in the potential range between –970 and –480mVSHE(Fig.9,labelw/odef),showingthatthe

h

-Znsurface remainspassiveinthisEinterval.

However, if the sample is strained up to

e

=3.1% under dryconditions andCVacquired24hlater,jstarts increasingat E
650mV SHE and, at the beginning of the reverse scan, continuesgrowing(Fig.9,labeldry(pl)def),asexpectedforthe onset of pitting corrosion. Under these conditions, pit repassivation is observedat E
725mVSHE (Fig. 9, labeldry (pl)def).Iftheelasticstrainisnotreleasedafterapplyingtheplastic strain,asseeninFig.8inthecurvelabelled“dry(pl+el)def)”,the pittingpotentialshiftsslightlyinthenegativedirection,without significantvariation oftherepassivationpotential.Eventhough, theanodicchargeflowedduringpitting approximatelydoubles. Thus, elastic

e

promotes the further growth of the active dissolution front without interfering much on nucleation and repassivationphenomena.

Acompletedifferentbehaviourisobservedwhenthesampleis strained while immersed in the chloride solution and CV is measured a few seconds after straining the sample without releasingtheelasticstrain(Fig.9,labelwet(pl)def).Inthiscase,the

Fig.8. SEMmicrographsofhot-dipZncoatingafter3.1%strainingin0.01molL
1NaCl:a)Topviewandb)crosssection.Smallblackarrowsindicateslipsteps,largeblack arrowindicatesgrainboundarycrack,andsmallwhitearrowsindicatetransgranularcracks.

Fig.9.Cyclicvoltammetriccurves(0.01molL
1NaCl,20mVs
1)ofhot-dipZn coatingunderdifferentconditions.Undeformed:w/odef.Strainedpreviouslyinair: dry(pl)def).Strainedinairwithoutreleasingtheelasticstrain:dry(pl+el)def. StrainedinthesolutionafewsecondsbeforeCVacquisition:dry(pl+el)def.

current rise is observed at a potential close to the standard equilibrium potential of Zn2+/Zn, at E
775mV SHE without showingapassivecurrentplateau,andjapproacheszeroasthis potential is achieved again during the reverse scan. The jhysteresis,typicaloflocalizedcorrosion,isnolongerobserved, becauseofthechangeofthecorrosionmechanismfrompittingof thepassive

h

-Znsurfacetoitsuniformcorrosion.Itisnoteworthy thatcathodiccurrentsalsoslightlyincreasewiththeappliedstrain, indicatingthatthepassivelayerformedonthe

_h

-Znphaseunder dryconditionswas disruptedandnotcompletelyrestoredfora certaintime. The observedcathodic j-increaseis in accordance withtheoccurrenceoflocalizedcathodicsites,asobservedonthe jmapsofSVET(Fig.5)whenstrainisapplied.

Itcanbeconcludedthataslipdissolutionphenomenonof

_h

-Zn ofthecoatingoccursatlowstrainvalues,promotingthenucleation of pitsand lowering thepitting potential. The “slipdissolution model”hasbeenproposedinthe60’s[e.g.[27]]andintensively studiedin the80'sasa mechanismofstresscorrosioncracking (SCC)ofstainlesssteels[28,29]andAl-alloys[30–32].Thestrain in_fluenceonSCCofthose alloyswasexperimentallyveri_fiedby observingthenucleationofcracksonslipsteps,whichformatthe interceptofpersistentslipbands(PSB)withthemetalsurface[27]. Additionalevidencewasfoundbyobservingmicrotunnelsatslip steps,united by crack propagation[33]. According tothis SCC mechanism,anincrementofcracklengthwillleadtotheformation ofPSBatthenewpositionofthecracktipandtonewmicrotunnels inacontinuingprocessofcrackgrowthofthestressedmaterial.In thiswork,theinfluenceofslipstepdissolutiononthecorrosionof the Zn coating was directly verified by measuring the current densityby CVand SVETexperimentsat verylow strainvalues. Morerecently,Zhuetal.[34]observedpitnucleationandSCCof austeniticstainlesssteelsubmittedtolowelasticstrainsimilarto thevaluesusedinthiswork.

IftheCVismeasuredjustafterstrainingtheimmersedsample, thesurfacerepassivationisnotcomplete,enablingahigherdegree ofactivedissolutionoftheZnsurfacebyuniformcorrosionbesides theincreaseofcathodiccurrents.AsverifiedintheOCPtransients ofFig.4,around400safterstrainingandexposureofthesurfaceto the chloride solution, the passive film is reconstructed and localizedcorrosionwillagainbethemajorcorrosionmechanism, however maintaining the influence of previous applied plastic strain.Theelasticstrainhelpwillhelpthedisruptionofthepassive filmenhancinguniformandlocalizedcorrosion.

4.Conclusions

ThecorrosionofHDZncoatedsteelstrainedincontactwith 0.01MNaClsolutionwithoutreleaseoftheelasticstrainismuch more intense compared to unstrainedsamples exposed during thesametime,ortosamplespreviouslystrainedinairandthan exposedtothesolutionduringthesametime.

Inthe caseof strainingthe immersedsurface,thecorrosion mechanismchangesfromlocalizedtouniformcorrosion,specially ifthepolarizationisappliedwithinafezsecondsafterstraining. After repassivation of the surface, the influence of previous straining is still verified by a lower pitting potential and higher anodic current densities. This is explained by slip dissolutionthathelpsthenucleationofpitsonthepassivesurface or to the disruption of the passive layer promoting uniform corrosion.

Maximum anodicand cathodic current densitiesincrease at approximatelythesamerateastheappliedstrainiscontinuously increasedwithavalueofroughly70

_m

Acm
2perpercentagestrain increment.

Thenucleationoflocalizedcorrosionsitesoccursatultralow strainvaluesaroundtheyieldpointorevenlower.Thenucleation isfavouredbytheruptureofthepassivefilmontheZnsurfaceby theactionofslipstepsandfineintergranularcracksincontactwith thesolution,ratherthan duetothefractureofthecoating and exposureofthesteelsubstrate.

Acknowledgments

TheauthorsacknowledgethefinancialsupportofUSIMINASand CAPES,andtheCentrodeMicroscopiaEletrônicaofUFRGSfortheuse offacilities.

References

[1]A.K.Gupta,D.R.Kumar,Formabilityofgalvanizedinterstitial-freesteelsheets, JournalMaterialsProcessingTechnology172(2006)225.

[2]E.A.Sacco,N.B.Alvarez,J.D.Culcasi,C.I.Elsner,A.R.DiSarli,Effectoftheplastic deformationontheelectrochemicalbehaviorofmetalcoatedsteelsheets, Surface&CoatingsTechnology168(2003)115.

[3]S.T. Vagge, V.S. Raja, R.G. Narayanan, Effect of deformation on the electrochemical behavior of hot-dip galvanized steel sheets, Applied SurfaceScience253(2007)8415.

[4]J.D.Culcasi,C.I.Elsner,A.R.DiSarli,Effectofzinccrystalssizeongalvanized steeldeformation and electrochemicalbehaviour, Materials Research 12 (2009)273.

[5]H.S.Isaacs,Y.Ishikawa,Applicationsofthevibratingprobetolocalizedcurrent measurements, Proceedings NACE Corrosion/85, March 25–29,Paper 55 (1985).

[6]H.S. Isaacs, Initiation of stress corrosion cracking of sensitized type 304 stainless steelindilute thiosulfatesolution,Journal Electrochemical Society135(1988)2180.

[7]H.S.Isaacs,Themeasurementofthegalvaniccorrosionofsolderedcopperusing thescanningvibratingelectrodetechnique,CorrosionScience28(1988)547. [8]R.M. Souto, Y. Gonzalez-Garcıa, A.C. Bastos, A.M. Simões, Investigating

corrosionprocessesinthemicrometricrange:ASVETstudyofthegalvanic corrosionofzinccoupledwithiron,CorrosionScience49(2007)4568. [9]A.M.Simões,A.C.Bastos,M.G.Ferreira,Y.Gonzalez-Garcıa,S.Gonzalez,R.M.

Souto,UseofSVETandSECMtostudythegalvaniccorrosionofaniron–zinc cell,CorrosionScience49(2007)726.

[10]J.Izquierdo,L.Nagy,S.González,J.J.Santana,G.Nagy,R.M.Souto,Resolutionof theapparentexperimentaldiscrepanciesobservedbetweenSVETandSECM for thecharacterization ofgalvanic corrosion reactions, Electrochemistry Communications27(2013)50.

[11]S.Kallip,A.C.Bastos,M.L.Zheludkevich,M.G.S.Ferreira,Amulti-electrodecell forhigh-throughputSVETscreeningofcorrosioninhibitors,CorrosionScience 52(2010)3146.

[12]S. Kallip, A.C. Bastos, K.A. Yasakau, M.L. Zheludkevich, M.G.S. Ferreira, Synergisticcorrosioninhibitionongalvanicallycoupledmetallicmaterials, ElectrochemistryCommunications20(2012)101.

[13]M.Challis,D.A.Worsley,Cutedgecorrosionmechanismsinorganicallycoated zinc–aluminiumalloygalvanisedsteels,BritishCorrosionJournal36(2001) 297.

[14]I.M.Zin,S.B.Lyon,A.Hussain,Under-filmcorrosionofepoxy-coatedgalvanised steelAnEISandSVETstudyoftheeffectofinhibitionatdefects,Progressin OrganicCoatings52(2005)126.

[15]D.J.Penney,J.H.Sullivan,D.A.Worsley,Investigationintotheeffectsofmetallic coatingthickness onthe corrosionproperties ofZn–Al alloygalvanising coatings,CorrosionScience49(2007)1321.

[16]F.Thébault,B.Vuillemin,R.Oltra,K.Ogle,C.Allely,Investigationofself-healing mechanismongalvanizedsteelscutedgesbycouplingSVETandnumerical modeling,ElectrochimicaActa53(2008)5226.

[17]A.C.Bastos,A.M.Simões,M.G.Ferreira,Corrosionofelectrogalvanizedsteelin 0.1MNaClstudiedbySVET,PortugaliaeElectrochimicaActa21(2003)371. [18]A.M.Simões,J.C.S.Fernandes,Studyingphosphatecorrosioninhibitionatthe

cutedgeofcoilcoatedgalvanizedsteelusingtheSVETandEIS,Progressin OrganicCoatings69(2010)219.

[19]K.Berchem,M.G.Hocking,Theinfluenceofpre-strainingonthecorrosion fatigueperformanceoftwohot-dipgalvanisedsteels,Corrosionscience48 (2006)4094.

[20]A.C.Bastos,M.G.S.Ferreira,A.M.P.Simões,Effectsofmechanicalformingon thecorrosionofelectrogalvanisedsteel,CorrosionScience69(2013)87. [21]A.G.Marques,A.M.Simões,EISandSVETassessmentofcorrosionresistanceof

thin Zn-55% Al-rich primers: Effect of immersion and of controlled deformation,ElectrochimicaActa148(2014)153.

[22]EASTME8/E8M-13a,StandardTestMethodsforTensionTestingofMetallic Materials,ASTM,International,WestConshohocken,PA,2013.

[23]M.Ueda,M.Matsuda,T.Ohtsuka,AtmosphericCorrosionofZincandZinc AlloysUsingIn-SituRamanSpectroscopy,TheElectrochemicalSociety14 (2004)166–172.

[24]T.E. Graedel,CorrosionMechanismsforZinc ExposedtotheAtmosphere, JournalofTheElectrochemicalSociety136(1989)193–203.

[25]M.C. Bernard,A. Hugot-LeGoff, N.Phillips, InSitu RamanStudy ofthe CorrosionofZinc-CoatedSteelinthePresenceofChloride,JournalofThe ElectrochemicalSociety142(1995)2167–2170.

[26]W.Miao,I.S.Cole,A.K.Neufeld,S.Furman,PittingCorrosionofZnandZn-Al CoatedSteelsinpH2to12NaClSolutions,JournalofTheElectrochemical Society154(2007)C7–C15.

[27]P.R.Swann,H.W.Pickering,ImplicationsoftheStressAgingYieldPhenomenon withRegardtoStressCorrosionCracking,CorrosionNACE19(1963) 369t–372t.

[28]T.Nakayama,M.Takano,ApplicationofaSlipDissolution-RepassivationModel forStressCorrosionCrackingofAISI304StainlessSteelinaBoiling42%MgCl2

Solution,Corrosion42(1986)10–15.

[29]P.L. Andresen, Environmentally Assisted Growth Rate Response of NonsensitizedAISI316GradeStainlessSteelsinHighTemperatureWater, Corrosion44(1988)450–460.

[30]F.P. Ford, Quantitative examination of slip-dissolution and hydrogen-embrittlement theories of cracking in Al alloys, Materiasls Science and Technology12(1978)326–334.

[31]M.P.Mueller,A.W.Thompson,I.M.Bernstein,StressCorrosionBehaviorof 7075Aluminumin1NAluminumChlorideSolutions,Corrosion41(1985) 127–136.

[32]K.R.Cooper, R.G.Kelly,Measurementand Modellingof CrackConditions duringtheenvironment-assistedcrackingofAl-Zn-Mg-CuAlloy,TMSAnnual Meeting,Proc.ofSymp.onChemistryandElectrochemistryofCorrosionand StressCorrosion,Warrendale,PA,2001.

[33]P.R.Swann,J.D.Embury,in:V.F.Zackay(Ed.),HighStrengthMaterials,John Wiley, 1965,pp.327–362.

[34]L.K.Zhu,Y.Yan,L.J.Qiao,A.A.Volinsky,StainlessSteelPittingandEarly-stage StressCorrosionCrackingUnderUltra-lowElasticLoad,CorrosionScience77 (2013)360–368.