1

ATMOSPHERICCORROSIONDATAOFWEATHERINGSTEELS.A
REVIEW.
M.Morcillo,B.Chico,I.Daz,H.Cano,D.delaFuente
NationalCentreforMetallurgicalResearch(CENIMCSIC),Avda.GregoriodelAmo,8,28040
Madrid,Spain

ABSTRACT
Extensiveinformationontheatmosphericcorrosionofweatheringsteelhasbeenpublishedin
the scientific literature. The contribution of the present work is to provide a bibliographic
reviewofthereportedinformation,whichmostlyconcernstheweatheringsteelASTMA242.
Thisreviewaddressesissuessuchasrustlayerstabilisationtimes,steadystatesteelcorrosion
rates,andsituationswheretheuseofunpaintedweatheringsteelisfeasible.Italsoanalyses
theeffectofexposureconditions.Finallyitapproachestheimportantmatterofpredictingthe
longtermbehaviourofweatheringsteelreviewingthedifferentpredictionmodelspublished
intheliterature.

Keywords:A.Lowalloysteel;C.Atmosphericcorrosion;C.Rust

Introduction
1.
Weatheringsteels(WS),alsoknownaslowalloysteels,aresteelswithacarboncontentofless
than 0.2 wt. % to which mainly Cu, Cr, Ni, P, Si and Mn are added as alloying elements to a
total of no more than 35 wt. % [1]. The enhanced corrosion resistance of WS in relation to
mildsteelorplaincarbonsteel(CS)isduetotheformationinlowaggressiveatmospheresofa
compactandwelladheringcorrosionproductlayerknownaspatina.
This definition, however, has not remained unchanged but has evolved as new WS
compositions have been developed to achieve improved mechanical properties and/or
withstand increasingly aggressive atmospheric conditions from the corrosion point of view,
especiallyinmarineenvironments.TheAmericanSocietyforTestingandMaterials(ASTM)has
standardiseddifferentalloycompositionsforWS,fromaninitial1.5%totalweightofalloying
elements added in the first standardised WS A242 [2], to 5% in the last standardised WS A
709HPS100W[3],whichisatthelimitofthecompositionofintermediatealloysteels.Table1
setsoutthechemicalcompositionoftwocommonlyusedWS[2,4].
The patina on WS not only offers greater corrosion resistance than on mild steel, but is also
responsible for its attractive appearance and selfhealing abilities. The main applications for
WS include civil structures such as bridges and other loadbearing structures, road
installations, electricity posts, utility towers, guide rails, ornamental sculptures, faades,
roofing,etc.
The literature contains a great deal of information on WS, and there are entire chapters in
collective works dedicated to this issue, e.g. [1, 56]. However, an indepth bibliographic

Correspondingauthor:Tel.+34915538900,Fax.+34915347425.
email:morcillo@cenim.csic.es(M.Morcillo)

review on the atmospheric exposure data of WS and a rigorous analysis of the published
informationarelacking.Thereviewpresentedinthispaperseekstofillthisgap.
2.

Briefhistoricaldevelopment

AlbrechtandHall[7]publishedacompletereviewonthehistoricaldevelopmentofWS.
ThebirthofWScanbetracedbacktothedevelopmentofsteelscontainingcopper,knownas
coppersteels[8].In1910Buckobservedthatsteelsheetswith0.07%CumanufacturedbyUS
Steel and exposed in three environments of different corrositivities (rural, industrial and
marine)showeda1.52%greateratmosphericcorrosionresistancethanCS[9].Hence,in1911
US Steel started to market steel sheets with a certain copper content. Buck subsequently
reported that the improvement achieved with Cu concentrations in excess of 0.25% was
insignificant,notingthat0.15%Cuprovidedsimilarresultsto0.25%Cuinmostcases[10].
Oncethiscapacityofcoppersteelbecameknown,furtherresearchledtothedevelopmentof
WSandthustoHighStrengthLowAlloy(HSLA)steels[7].Inthe1920'sUSSteelproduceda
newfamilyofHSLAsteelsintendedprimarilyfortherailwayindustry.Finally,in1933USSteel
launched the first commercial WS under the brand name USS CorTen steel (Early CorTen
steelorCorTenB),anamewhichreflectsthetwopropertiesthatdifferentiateitfromCS,i.e.
itscorrosionresistance(Cor);andfromacoppersteel,i.e.itssuperiormechanicalproperties
(tensile strength, Ten). This product was claimed to provide a 30% improvement on the
mechanical properties of conventional CS, thus reducing the necessary thickness and
accordingly the weight of steel to be used for a given set of mechanical requirements [11].
Figure1illustratesthecorrosionofthesethreesteelsintheindustrialatmosphereofKearny
anditsevolutionwithtime[12].Attentionisdrawntothelowercorrosionexperiencedbythe
EarlyCorTensteel.
EarlyversionsofUSSCorTensteelswerebasedonFeCuCrPsystems,towhichNiwaslater
added in order to improve corrosion resistance in marine environments. USS CorTen steels
presentedtwospecifications,AandB,whosemaindifferencelayintheamountofphosphorus
present in their composition. USS CorTen A can be said to be the WS with the highest
phosphoruscontent(0.070.15%weight)andUSSCorTenBthatwiththelowestphosphorus
content(0.04%weight)[13].
Greaterknowledgeoftheroleplayedbythedifferentalloyingelements(Cu,Cr,Ni,P,etc.)in
theatmosphericbehaviourofWSwasachievedthankstotwoambitiousstudiescarriedoutin
theUnitedStates,onebeganin1941byASTMCommitteeA5[14]andanotherbeganin1942
byUSSteelCo.[12].Inthefirst,71lowalloysteelswereexposedtotheindustrialatmosphere
atBayonne,N.J.andtothemarineatmospheresatBlockIsland,R.I.andat KureBeach (250
m),N.C.Inthesecondstudy,270differentsteelswereexposedinthefollowingatmospheres:
SouthBend,Pa(semirural),Kearny,N.J.(industrial)andKureBeach(250m),N.C.(marine).
ThecurrentcompositionofUSSCorTensteelshasalteredtoacertainextent,especiallyinthe
case of specification B, with the addition of Ni (0.40 % Ni), but they all continue to be
marketed to the present day [1516]. In 1941 the first WS was standardised by ASTM
specification A242; a steel that is roughly comparable to USS CorTen A steel. Its main
characteristic is its high resistance to atmospheric corrosion, which is approximately 4 times
greater than that of CS due to the presence of copper, a high phosphorus content, and in

general the presence of nickel (0.500.65%). However, it is now somewhat obsolete as a

structural steel due to the fact that phosphorus can form iron phosphide (FeP3) during the
weldingprocess,decreasingitsweldabilityandcausingthesteeltobecomebrittle.
In 1968 ASTM standard A242 presented two specifications, one with a high phosphorus
content(<0.15%P)andtheotherwithalowerphosphoruscontent(<0.04%P).Thelatterwas
ultimatelyreplacedbyASTMstandardA588WS[4](seeTable1),whichisroughlycomparable
toUSSCorTenBsteel.Thissteelpossesseslessresistancetoatmosphericcorrosionduetoits
lowerPcontent,butforthissamereasonithasbetterweldability.
Finally,in1992theUSFederalHighwayAdministration(FHWA),theAmericanIronandSteel
Institute (AISI) and the US Navy started to develop new improved WS for bridge building,
knownasHighPerformanceSteels(HPS),andin1997thefirstbridgewithHPS70Wwasbuilt
in Nebraska [17]. Three basic targets were set to improve the overall quality and
manufacturabilityofthesteelsusedhithertoforbridgeconstructionintheUnitedStates[18]:
(a) improve weldability, achieved by lowering the carbon, phosphorus and sulphur contents;
(b)improvemechanicalproperties,suchasfracturetoughnessandyieldstrength,achievedby
raisingthemaximummanganeselimit;and(c)maintaintheformationofprotectiverustthat
characterisesWS.

3.

RequirementsfortheformationofprotectiverustlayersonWS

Ashasbeenmentioned,theenhancedcorrosionresistanceofWSisduetotheformationofa
denseandwelladheringcorrosionproductlayer.
Experiments carried out in 1969 by Schmitt and Gallagher with low alloy steel (CorTen A)
indicated that the texture of the oxide layer was dependent upon the washing action of
rainwater and the drying action of the sun [19]. Surfaces sheltered from the sun and rain
tendedtoformalooseandnoncompactoxidewhilesurfacesopenlyexposedtothesunand
rain produced strongly adherent layers. On northfacing surfaces the protective layer
developedsomewhatmoreslowlyasaresultofreceivinglesssunlight.Matsushimaetal.[20]
subsequentlystudiedtheroleofalargenumberofenvironmentalanddesignvariablesinthe
behaviourofWSinarchitecturalapplications,verifyingthedecisiveinfluenceontheformation
of the protective patina of whether or not the metallic surface was exposed to the rain, or
whether or not areas where moisture was liable to accumulate were drained. These effects
were more intense in atmospheres with higher pollution levels, in which case the protective
patinamaynotfullyform.
Extensive research work has thrown light on the requisites for the protective rust layer to
form.Itisnowwellacceptedthatwet/drycyclingisnecessarytoformadenseandadherent
rustlayer,withrainwaterwashingthesteelsurfacewell,accumulatedmoisturedrainingeasily,
and a fast drying action (absence of very long wetness times). Structures should be free of
interstices, crevices, cavities and other places where water can collect, as corrosion would
progresswithouttheformationofaprotectivepatina.ItisalsonotadvisabletousebareWSin
continuouslymoistexposureconditionsorinmarineatmosphereswheretheprotectivepatina
doesnotform[56,21].
Therefore, the ability of weathering type steels to fully develop their anticorrosive action is
dependent on the climate and exposure conditions of the metallic surface. It must also be

taken into account that a truly protective oxide film may never develop on certain areas, or
thattheirevolutionwillbeexcessivelyslow.

4.

AtmosphericcorrosiondataofWS.Worldwidesurvey.

Overtheyears,thefirstextensivefieldtestsofWScarriedoutbyCopson[14]andLarrabee
Coburn[12]havebeenfollowedbyfurtherstudiesallovertheworld.
A worldwide literature survey has been conducted to obtain data series on WS corrosion
versus exposure time [14, 2255]. Table 2 lists the countries and testing stations where
atmosphericexposurestudieshavebeenundertaken,aswellassomegeneralcharacteristics
ofthetestsperformed.
Afactorofuncertaintyisthattheatmospheresareoftenclassifiedinpurelyqualitativeterms
as rural, urban, industrial or marine, based on subjective assessments of pollution factors. It
canalsobeseenthatlargeamountsofdataareavailableforbehaviourinthefirst10yearsof
exposure,butthereisnotablylessinformationonlongtermexposureformorethan20years.
ThebulkofthedatareferstoASTMA242WS,andassuchthisisthesteelwhichisparticularly
referredtothroughoutthisreview.
The abundant literature on atmospheric corrosion of weathering steels can sometimes be
confusingandlackclearcriteriaregardingcertainconceptsthatareaddressedinthepresent
paper,namely:
(i)

Howlongdoesittaketoreachasteadystate(stabilisationoftherustlayer)inwhichthe
corrosionrateremainspracticallyconstant?

(ii)

Effectofenvironmentalexposureconditions

(iii)

WhatlawsbestfittheatmosphericcorrosionofWSandestimation(prediction)oflong
termatmosphericcorrosion?

Eachoftheseissuesisaddressedbelow.

5.

Stabilisationofrustlayersandsteadystatecorrosionrate

Bibliographicinformationonthisaspectishighlyerraticandvariable,goingfromclaimsthata
protective patina can be seen to be formed after as little as 6 weeks exposure to reports of
stabilisationtimesofoneyear,23years,8years,andsoforth[56].
Matsushima et al. [20] reported the time required for the corrosion product to stabilise
accordingtoitslocationinthestructureandwhethertheenvironmentwasindustrialorurban.
Thistimevariedfrom1to5years,exceptforlocationswherestabilisationdidnotoccurdueto
waterpoolformation.
ThegradualdevelopmentofaprotectivelayeronWStakesseveralyearsbeforesteadystate
conditionsareobtained.Thetimetakentoreachasteadystateofatmosphericcorrosionwill
obviously depend on the environmental conditions of the atmosphere where the steel is
exposed.
TodeterminethestabilisationtimesofrustlayersformedonWSinthedifferentatmospheres,
referencehasbeenmadetoinformationobtainedintheaforementionedbibliographicsurvey
(Table2).Priortoitsanalysisthedatabasehasbeenrefinedbyremovingalldataserieswhich

haveinformationforonlyasmallnumberofpoints(e.g.1or2pointsoncorrosionrateversus
timecurves),whichcorrespondtostudiesthatendedaftershortexposuretimes(<7years),or
in which it was not possible to determine the stabilisation time due to rising corrosion rates
withexposuretime,excessivelyhighcorrosionrates,etc.
The procedure established to determine rust layer stabilisation times and steadystate
corrosionrateshasbeenasfollows:
1.

Theavailablecorrosionrate,y,isplottedagainsttheexposuretime,t1,andfittedtoan
exponentialdecreaseequation:
y=A1 exp(

x
t1

)+yo

(1)

2.

Oncetheequation(1)isobtained,itisusedtodeterminethecorrosionrateatdifferent
exposure times and to calculate the decrease in the corrosion rate for each annual
incrementinexposure.

3.

A steady state (stabilisation time) is considered to have been reached when the
decreaseinthecorrosionrateforaoneyearincrementinexposuretimeis10%.(Fig.
2).

4.

The steady state corrosion rate is the rate corresponding to the year from which a
decreaseof10%takesplace(Fig.2).

Withtheinformationselectedinthisway,tables3and4havebeenpreparedfornonmarine
(rural,urbanandindustrial)andmarineatmospheresrespectively,listingthetestingstationsin
orderofcorrosivitycategoriesaccordingtoISO9223[57],obtainedfromcorrosiondataforCS
afteroneyearofexposure.
The tables contain the following information: ISO corrosivity category (ISO 9223), site and
country, firstyear CS corrosion (m), stabilisation time (years), steadystate corrosion rate
(m/y),max.exposuretime(years)andtypeofWSaccordingtoASTMspecifications.
Boxwhisker plots obtained from statistical processing of the stabilisation times and steady
statecorrosionratesforASTMA242WSshownintables3and4arepresentedinfigures3to
5. The boxwhisker type graph represents the statistical data as separate boxes. The box is
determinedbythe25thand75thpercentiles,andthebandinsidetheboxisthemedian(50th
percentile).Theendsofthewhiskersaredeterminedbytheminimumandmaximumofallof
thedata.Thesmallsquaresymbolinsidetheboxisthemeanvalueofthedataandxsymbols,
the1stand99thpercentiles.Thesefiguresrevealthefollowingtrends:
(i)

The stabilisation time of the rust layer decreases as the corrosivity category of the
atmospherewheretheWShasbeenexposedrises(Fig.3),goingfrom68yearsinless
aggressiveatmospheresto46yearsinmorecorrosiveatmospheres.
Thestabilisationtimeoftherustlayersdepends,amongotherfactors,ontheexposure
time,theexistenceofwet/drycycles,thecorrosivityoftheatmosphere,andinshorton
the volume of corrosion products formed. However, a shorter stabilisation time does
notimplyagreaterprotectivecapabilityoftherust.Inthisrespect,thestabilisationof
therustlayeroccursfasterinmarineatmospheresduetotheirgreatercorrosivity,but

the protective value of this rust is lower than that of rusts formed in nonmarine
atmospheres(rural,urbanandlightindustrial)wherethestabilisationtimesarelonger.
(ii)

The steadystate corrosion rate rises in line with the corrosivity of the atmosphere in
bothrural,urbanandindustrialatmospheres(Fig.4)andmarineatmospheres(Fig.5).
AmatterofthegreatestpracticalrelevanceistoknowwhenanunpaintedWScanbe
used. The cost of painting a CS bridge usually represents around 12% of its initial
construction cost, with repainting taking place every 6 to 15 years, depending on the
aggressivityoftheenvironmenttowhichitisexposed.WSstructuresnormallybecome
more economical than painted CS structures [58] after 15 years in service in
environmentsofmoderateaggressivity.
The problem is how to define the environmental conditions of an atmosphere of
moderate aggressivity, which is ultimately what determines the applicability of an
unpainted WS. From a practical point of view, the criterion followed has been to limit
the steadystate corrosion rate of WS in the atmosphere to an acceptable value for
material safety, where maintenance operations are not required [59]. Thus, in 1960,
aftera15yearatmosphericcorrosiontest,LarrabeeandCoburn[12]suggestedthat5
m/year was an acceptable corrosion rate for the use of unpainted WS. In Japan
unpaintedWScanbeusedforbridgeconstructioninenvironmentswherethecorrosion
loss is 6 m/year for the first 50 years of exposure [60], although this criterion has
subsequently become more restrictive, and their use is now only advised in
environmentswherethecorrosionlossis5m/yearforthefirst100yearsofexposure
[61]. In the USA, Cook [62] reported that the use of unpainted WS was acceptable in
placeswithanaveragecorrosionofupto120minthefirst20yearsofexposure,i.e.a
corrosionrateof6m/year.
The fairly widespread criterion of 56 m/year for longterm atmospheric corrosion
suggests only a marginal benefit of using conventional WS in preference to CS in
aggressiveatmospheres,especiallyinmarineatmospheres[63].Aviewoffigures4and
5, which plot a steadystate corrosion rate of 6 m/year, indicates that unpainted
conventionalWSshouldnotbeusedinatmospheresofcorrosivitycategoryC3,marine
orotherwise.

(iii)

With regard to the corrosion resistance of the two traditionally most widely used WS,
ASTMA242andASTMA588,viewoftables3and4revealsthatASTMA588exhibits
poorer corrosion resistance than ASTM A242 (its steadystate corrosion rates are 50
100%higherthanthoseofASTMA242)andashorterexposuretimeintheatmosphere
beforestabilisationoftherustlayer.

6.

Effectofexposureconditions

ThecorrosioncharacteristicsofWSdependinacomplexwayonclimateparameters,pollution
levels,exposureconditions(openair,shelter,etc.)andthecompositionofthesteel.
WithregardtotheeffectofexposureconditionsontheatmosphericcorrosionofWS,onlya
limited amount of information is available. In this respect it is interesting to cite the work
carried out by K. Park in 2004 [64], which makes an exhaustive bibliographic review

consideringdesignaspectssuchas:angleofexposuredependingonthelatitudeofeachsite,
skywardorgroundwardorientation,shelteredconditions,continuousmoistureconditions,and
heavyconcentrationsofpollutantsfromrainfallorsaltspray,deicingsalts,dirt,anddebrisin
industrialandmarineenvironments,etc.
In order for patinas with optimum protective properties to develop, the environmental
conditions must act in the appropriate way on the metal surface. In this sense, as has been
noted above, it is important that rainwater washes the steel surface well, that accumulated
moistureiswelldrained,andthatdryingactionisfast.Akeyconditionfortheformationofa
protectivepatinaisacyclicalvariationbetweenwetanddryperiods.Periodicflushingfollowed
bydryingisdesirable.Inrainshelteredareas,thetypicaldarkpatinaformedonopenexposure
isnotobtainedandinsteadthesurfacebecomescoatedwithalayeroflightercolouredrust.
Nevertheless, the rust layers are more compact and adherent than on CS and the corrosion
rateisusuallysolowthattheabsenceofdarkpatinaseemstobeofonlyaestheticsignificance
[32]. An exception is atmospheres with high chloride contents, where the accumulation of
hygroscopicchloridescangiverisetoveryhighcorrosionrates[65].
6.1

Geometryofexposure

The good performance of WS, low corrosion rate, good surface appearance, and early
formationof protectiverustfilms,dependonthegeometryof exposurewhichdeterminesa
goodwashingactionofrainwaterandeasydrainageofmoisture.Thus,Matsushimaetal.[20]
bymeansofperiodicobservationsoftheappearanceofrustandcorrosionratemeasurements
showedthatgoodwashingofthesurfacebyrainwaterwasthemostimportanteffectforthe
earlydevelopmentofaprotectiverustfilm.
Larrabee [66] and Schikorr [67] noted the importance of the orientation and geometrical
configurationofspecimensinfieldexposuretests,i.e.theanglewithrespecttothehorizontal,
and differences between the upper and lower sides of the specimens, due primarily to the
timeofwetnessandamountofdeposition.
The orientation and angle of exposure significantly affect the corrosion rate. To obtain
comparable results, ASTM practice G50 [68] recommends exposing specimens at an
orientationandananglethatyieldtheleastamountofcorrosion.Thisconditionismetinthe
northernhemispherewhenthesunisabovetheequatorandthespecimensfacesouthatan
angleequaltothelatitudeofthetestsite.Conversely,inthesouthernhemisphere,specimens
should be exposed facing north at an angle equal to the latitude of the test site. Most
atmosphericcorrosiontestsintheUnitedStateshavebeenperformedatanangleof30facing
south. However, atmospheric corrosion studies in Europe are usually done with specimens
facingsouthat45fromthehorizontal,owingtothehigherlatitudes.
6.1.1. Anglewithrespecttothehorizontal
Several researchers have addressed this issue. In 1951 Laque [69] noted that specimens
exposed vertically do not develop protective rusts as those exposed at an angle of 30,
especially in the very severe environment close to the ocean. Laque concluded that vertical
exposurehasthedisadvantagethatafewdegreesvariationfromthespecified90willhavea
much greater effect on the extent to which the surfaces are shaded than a similar variation
fromaspecified30position.In1966Larrabee[70]alsoreportedthattheanglerelativetothe

horizontalatwhichspecimensareexposedtotheenvironmentaffectedthecorrosionrateof
WS.
Coburnetal.[41]presentedresultsobtainedafter16yearsofexposureofASTMA588Gr.A
(originallyCorTenB)WSspecimensindifferentatmospheresattwodifferentangles,30and
90(vertical),concludingthatexposureat30ledtolowercorrosionrates.Theseresultswere
consistentwiththosepreviouslyreportedbyLaque[69]andLarrabee[70].Meanwhile,Coburn
etal.notedthatWSspecimensexposedfacingsouthshowedlowercorrosionratesthanthose
facingnorthduetothefactthatdirectsolarradiationfromthesouthresultedinashortertime
ofwetness[41].
Studies carried out by Cosaboom et al. [71] confirm the observations of Coburn et al. [41].
WhileCosaboometal.found that0 (horizontal)specimenscorrodedlessthan90(vertical)
specimensbecausetheywashedanddriedwithinashorttimeafterrainfall,Zoccolaetal.[72],
in a study carried out in Detroit, noted the reverse to be true, with horizontal specimens
corrodingmorethanverticalspecimens.Zoccolaetal.notedthatthetestsitewassubjectto
the accumulation of road dirt and salts on the steel surfaces, and the horizontal specimens
were easily contaminated, promoting prolonged wetness and accelerating steel corrosion
rates.Thus,specialattentionmustbepaidtohorizontalsurfaceswherehygroscopicdeposits,
chlorides and sulphates, may accumulate and prolong the time during which the surface is
wetted.
The effect of panel orientation was studied also by Knotkova et al [32]. Based on the
appearanceofthesurfaces,therustfoundonthespecimensfacingsouth,west,andeasthad
an attractive fine texture and dark colour and was comparable in all three orientations. The
horizontal specimens were different in appearance: the WS had a lightcoloured, coarse
texturedbut adherentrustlayerwhile theruston theCSwasflakingofflocally.Gravimetric
measurementsrevealedthateastwardexposurewasthemostaggressive.Moistureprobably
dried more slowly on the eastfacing specimens and this increased the time during which
activecorrosionoccurred.ThelargestdifferencebetweenWSandCSwasseenonthesouth
facingspecimens.
6.1.2. Orientation
Larrabee [66] and Zoccola [73] showed that the skyward or groundward orientation of the
specimen can affect the atmospheric corrosion rate of WS. The upper surface of steel
specimens is generally expected to be wet for a shorter time than the lower surface, which
receivesnosunshine.Inthisrespect,OeschandHeimgartner[74]notedthatrustonthelower
sideofaWSexposedtoanatmospherepollutedwithSO2presentedahigherconcentrationof
sulphatenests,whiletherustformedontheuppersidewasmorecompactwithfewerflake
offs.
Inastudyofgeometricalfactors,MoroishiandSatake[75]assessedtheupperandlowersides
ofCSandWSspecimensusingspecimenswithonesidepaintedthatwereexposedatvarious
angles and directions. The lower side tended to corrode as much as 6070% more than the
upperside,andtheextentofcorrosionincreasedastheorientationapproachedthevertical.
Theseobservationscorrelatedwiththesulphateconcentrationincorrosionproductlayerson
specimensthatwereonlyexposedontheirlowerside.Ontheotherhand,specimenswhose

uppersideswerealsoexposedbehaveddifferentlybecauseofthewashoutcausedbyrainfall.
Thesecorrosionphenomenaweremoreprominentforcarbonsteelthanforweatheringsteel.
Knotkovaetal.[32]alsomeasuredthecorrosionrateseparatelyontheupperandlowersides
ofCSspecimensduringexposureinasoutherlyorientationatanexposureangleof45from
thehorizontal.The7yearresultsshowedarelativeretardationofthecorrosionprocessonthe
lower sides of the specimens, but this conclusion cannot necessarily be applied to all
exposures.Incaseswherelargestructuresareexposedandventilationconditionsaredifferent
ondownwardfacingsurfaces,corrosionrateswillbesubstantiallydifferenttothosenotedfor
standardpanelsinatmosphericexposuresites.
6.1.3. Otherdesignparameters
Important studies of design parameters and the formation of protective corrosion product
layers were conducted by Satake et al. [76] and Matsushima et al. [20] in 1970 and 1974,
respectively.TheseresearchersconstructedmodelswiththeanticipatedconfigurationsofWS
structures: Hshaped pillars and beams, ceilings, window frames, siding, louvers and stairs.
TheypointoutthatduringthedesignandmaintenanceofWS,specialattentionmustbepaid
tocreviceareasthatcan retainmoisturemuchlongertimethan opensurfacesandinwhich
moreaggressiveelectrolytesmayform.
6.2.

Environmentalconditions

Itisofinteresttoknowtheeffectofclimateandpollutionvariablesonthecorrosionresistance
ofWS.
6.2.1. Outdoorexposure
Theliteraturecontainsabundantreferencesandrepresentationsofcorrosionversustimefor
WSexposedinatmospheresofdifferenttypes:rural,urban,industrialandmarine.Figure6is
an example of this. Such information, however, presents a factor of uncertainty in that the
atmospheresareclassifiedinapurelyqualitativewayaccordingtothelocationofthetesting
station and its surroundings, based on a subjective assessment of climate and pollution
factors.
AccordingtoKnotkovaetal.[32],inruralareasthecorrosionrateisusuallylow,andthetime
needed to develop a protective and nicelooking patina may be quite long. For CS the
corrosionrateisalsoratherlowintheseconditions(Fig.7).
InurbanenvironmentswithSO2levelsnotexceedingabout90mgSO2/m2.d,WSusuallyshow
stabilised corrosion rates in the range of 26 m/year, i.e. only slightly higher than in rural
atmospheres.CSshowsmarkedlyhighercorrosionratesintheseconditions(Fig.7).
In more polluted industrial atmospheres with SO2 levels exceeding 90 mg SO2/m2.d,
significantlyhighercorrosionratesarealsofoundforWS(Fig.7),indicatingthattherustlayer
formed at very high SO2 pollution is not altogether protective [32]. Although the surface
coatingmayhaveadark,pleasantappearance,itcannotbeclassifiedasatruepatina.Inthese
conditionslooserustparticlesarealsoformed.However,thecorrosionratehasalwaysbeen
foundtobelessthanthatofCSintheseenvironments.
Inmarineenvironmentspollutedwithchloridesaprotectivepatinadoesnotdevelopandthe
corrosion rate may be high, especially close to the shore. This applies especially to rain

10

sheltered surfaces, where the corrosion rate may be very high due to the accumulation of
chlorideswhichareneverwashedaway.
6.2.2. Indoorexposure
In indoor exposure no systematic differences have been observed between WS and CS
corrosionrates.Thelowcorrosionratesobservedinallcaseswereduetothelowcorrosivity
of the environment and not the steel composition. Thus, in agreement with Knotkova [32],
thereisnojustificationforusingWSinindoorconditions.
6.2.3. Shelterexposure
ThemajorityoflongtermexposuretrialsshowingtheadvantagesofWShavebeencarriedout
inopenexposureconditionswherethetestspecimensarefreelyexposedtotheactionofrain,
wind,andsunshine.However,alargepartofallsteelworkisshelteredfromanydirectaction
ofrainandsunshineanditisquestionablewhethertheresultsfromopenexposuretestingcan
be applied to sheltered steelwork. Unfortunately, little information is available on corrosion
ratesinshelteredconditions.
Larrabee[70],Zoccola[73],Cosaboometal.[71]andMckenzie[77]conductedtestswithboth
sheltered and open exposure to clarify the effect of sheltering on WS corrosion. Sheltered
racks were used for the experiment by Larrabee and Zoccola, whereas Cosaboom et al.
conductedexperimentsbyattachingthespecimensonaninteriorgirdertoachieveasheltered
effect, and on a roof of a building to simulate open exposure. Mckenzie also examined
specimensonaninteriorgirder,andexposedasetofspecimensnotonaroofbutintheopen,
facingprevailingwinds.
Larrabee [70] was the first researcher to carry out tests involving partly and completely
sheltered test specimens facing north, south, east and west. He found that for a rural site
there were only relatively small variations in weight loss that could be attributed to the
amount of sheltering and direction of exposure, although sheltered specimens generally had
considerablyhighercorrosionratesthanfreelyexposedspecimens.Atamarinesitehefound
that fully sheltered specimens corroded more than partly sheltered specimens. This was
attributed to the better washing action of rain on the chloride deposits on partly sheltered
specimens. However, in later tests at the same site the opposite was found, with fully
shelteredspecimenscorrodinglessthanpartlyshelteredspecimens.
Althoughtherelevanceofsuchinformationtotheparticularshelteringofacompositehighway
bridge is clearly limited, a paper analysing the applicability of results of longterm corrosion
tests with WS to the corrosion of highway bridges concluded that steel specimens sheltered
fromdirectrainorsunlightwouldcorrodemorethansteelfullyexposed[78].
In the research carried out by Zoccola [73], vertical specimens were attached on an interior
girderofabridgelocatedonanindustrialsiteandontheroofofabuildingnearthebridge.WS
was two times more corroded on sheltered interior girders than in open exposure. It was
arguedthatWSexposedtotrafficfumes,roadsaltanddirtcausedthespecimenstodevelop
heavy,flakyrustandasignificantaccumulationofdeposits.
Cosaboom et al. [71] mounted specimens vertically and horizontally on interior and exterior
girdersofabridgeclosedtotrafficandontheroofofanearbybuildingonanindustrialsite.

11

Both the vertical and horizontal bridgesheltered specimens lost up to 65% more thickness
thanthoseopenlyexposedonthebuildingroof.Moreover,verticalspecimenscorrodedmore
on interior girders than on exterior girders while horizontal specimens corroded almost the
sameoninteriorandexteriorgirders.
After5yearsoftestingataruralsiteMcKenzie[77]foundlittleindicationofareductioninthe
corrosion rate with time in sheltered conditions. In the absence of high chloride levels,
corrosionwaslowerinbridgeshelteringthaninopenexposure.However,inshelteredmarine
environmentsthecorrosionratesofWSwerehigherandseverepittingdeveloped.
6.2.4. Continuouslymoistexposure
Inconditionswithverylongwetnesstimesorpermanentwetness,suchasexposureinwater
orsoil,thecorrosionrateforWSisusuallyaboutthesameasforCS.
TheresultsofexperimentscarriedoutbyLarrabee[70]showedthattheamountofcorrosion
lossonWSandCSwashighinthewettunnelwherethesteelsremainedcontinuouslymoist;
especially with water of a low pH, the WS did not perform well in open exposure. The
atmosphere in the wet tunnel was so severely corrosive that after one year of exposure the
thicknesslossofWSwasmorethan11timesthatofthesamesteelexposedontheroof.
6.2.5. Airbornepollutants
Oneofthefactorsaffectingsteelcorrosionisindustrialpollutionproducedbytheburningof
fossil fuels. Atmospheric corrosion increases strongly if the air is polluted by smoke gases,
particularly sulphur dioxide, or aggressive salts, as in the vicinity of chimneys and marine
environments.Atmosphericcorrosionisthereforeparticularlystronginindustrialandcoastal
areas.Corrosionisfurthermoremuchhigherifthemetalsurfaceiscoveredbysolidparticles,
suchasdust,dirt,andsoot,becausemoistureandsaltsarethenretainedforalongtime[79].
According to Singh et al. [80], both SO2 and chlorides change the structure and protective
propertiesoftherustlayer.
6.2.5.1

Sulphurdioxide(SO2)

ParticularmentionshouldbemadeoftheeffectofatmosphericSO2pollutiononWScorrosion,
where the existing literature is rather confusing. While some authors argue that WS are less
sensitivetoSO2thanCS,especiallyforlongexposuretime,othersbelievethatWSneedaccess
to SO2 or sulphatecontaining aerosols to improve their corrosion resistance or report that a
low but finite concentration of SO2 in the atmosphere can actually assist the formation of a
protectivelayeronWS[6].
ISO 9223 [57] notes that "in atmospheres with SO2 pollution a more protective rust layer is
formed".LeygrafandGraedel[6]considersthatacertainamountofdepositedSO2orsulphate
aerosols is beneficial for the formation of protective patinas on the surface of WS, but large
amounts result in intense acidification of the aqueous layer, triggering dissolution and
hinderingprecipitation.
(a)

Openexposures
Studies into the effect of SO2 pollution in the atmosphere on WS corrosion are very
scarce.Perhapsoneofthefirststudiesinwhichthismatterhasbeendirectlyaddressed

12

isthatcarriedoutbySatakeandMoroishi[81],whoexposedspecimensofalowalloy
steel in a network of 12 stations throughout Japan. These authors found a close
relationshipbetweenfirstyearcorrosionlossesandtheSO2concentrationintheair,but
thisrelationshipwasnotclearlyseenafterfiveyears.
ApartfromthisworthyattemptbySatakeandMoroishi,thefactisthatfewstudieshave
been carried out (supported by an extensive experimentation) to determine the
relationshipbetweentheatmosphericSO2contentandWScorrosion.
Knotkovaetal.[32]carriedoutafieldstudyindifferentatmospheres,citedabove,to
determine the effect of SO2 on the atmospheric corrosion of an ASTM A242 type WS
with the denomination "Atmofix" (see Fig. 7), and the following considerations were
made:
(i)

Inruralandurbanatmosphereswithlowpollutionlevels,i.e.SO2levelsofabout40
mg/m2.day or less, the conditions are suitable for the formation of a highly
protectiverustlayeronlowalloysteelsleadingeventuallytoarelativelystablelow
rate of corrosion. The appearance of the rust layer on WS is characterised by an
attractivedarkbrowntovioletcolourandacompactstructure.

(ii) In more polluted urban and industrial atmospheres, where the annual average
pollution level reaches up to 90 mg/m2.day, the rust layer formed on WS is again
dark brown to violet, but its structure is coarser than that observed in rural
atmospheres. It is likely that at certain times of the year the higher SO2 content,
coupled with higher humidity, causes local cracking and spalling of the rust layer,
requiringthelayertoreformintheseareas.
(iii) ForheavilypollutedindustrialatmosphereswheretheSO2levelishigh,theresults
indicate a higher corrosion rate than is generally seen on WS with a stable rust
layer. This indicates that the process of spalling and regrowth of the rust layer is
occurringregularlyandintensively.
AccordingtoKnotkovaetal.[32]thecorrosionofWSmustbeaconstantfunctionofthe
SO2 content, increasing as the SO2 content rises. For this reason it was desirable to
specify a maximum SO2 level in the atmosphere where WS may be used without
protective coatings. Detailed analysis of the data set of longterm test results,
supplementedbytheresultsof3yeartestsintheNorthBohemiannetworkofstations
with more finely graduated SO2 levels, was very helpful to more precisely define the
critical SO2 level. The results of this analysis indicate a maximum annual average SO2
contentof90mg/m2.day.AsummaryplotshowingthisanalysisisgiveninFig.8.
It seems that 90 mg SO2/m2.day, the limit established by Knotkova et al. [32] may
perhapsbeexcessive.Inaccordancewiththesteadystatecorrosionratecriterionof6
m/year(seechapter5(ii)),whichiswidelyacceptedtoconsiderthatunpaintedWSmay
be used, Fig. 9 has been prepared from data obtained in a 8year study performed at
numerous locations within the framework of the UN/ECE International Cooperative
Programme on effects on materials, including historic and cultural monuments [26].
Figure 9 shows that the steadystate corrosion rate of 6 m/year is obtained for
atmospheric SO2 contents of less than 20 mg SO2/m2.day. Above this SO2 level WS

13

corrosion is accelerated. These results seem to confirm the opinion of Leygraf and
Graedel that large amounts of deposited SO2 result in intense acidification of the
aqueous layer existing on WS during the corrosion process triggering dissolution and
hinderingprecipitation[6].
(b)

Shelteredexposures
TheresultsobtainedbyKnotkovaetal.[33]forshedexposures(Stevensonscreen)after
5yeartestsonWSspecimensyieldedthefollowingconclusions:
(i)

The rust that forms on WS in shed exposures does not develop the protective
qualityfoundinnormalexposuretotheweather,and

(ii) The most favourable conditions for WS application came from urban shed
exposures.WScorrosionlossesatthesamelocalitieswerelowerthanthoseofCS,
and locations with good ventilation proved to be especially favourable; e.g. the
lower decks of bridges and open galleries. Even though the rust layers in these
cases never had the typical appearance of exterior exposures, they were more
compact and more adherent on WS than on CS. The difference in steadystate
corrosionratesisrelativelylow,sothefinalcorrosionratescannotbeamotivation
forusingWSinsteadofCS.
Table5hasbeenpreparedusingdataobtainedinamorerecentstudycarriedoutinthe
frameworkoftheUN/ECEexposureprogramme[26],whereASTMA242WSspecimens
were exposed to the open air and inside ventilated boxes (shelters). The steadystate
corrosion rate has been calculated, according to the procedure mentioned above in
chapter 5, for different testing stations in the two exposure conditions: open air and
shelter.
Inatmospheresoflowcorrosivity,wherefirstyearCScorrosionremainsbelow40m,
the steadystate corrosion rate in sheltered conditions is similar to or less than that
correspondingtoopenairexposure.However,whenthefirstyearCScorrosionishigher
than40m,theWScorrosionrateinshelteredexposureexceedsthatfoundinopenair
exposure.IntheseatmospherestheSO2contentintheatmosphereishigher,promoting
greater steel corrosion in sheltered conditions due to moisture retention and the
absenceofwashoffbyrainwater.
6.2.5.2

Seachlorides

IfbibliographicinformationontheeffectofSO2ontheatmosphericcorrosionofWSisscarce,
dataanalysingtheeffectofatmosphericsalinityinmarineatmospheresisevenscarcer.
TheonlyrigorousstudyfoundintheliteraturewascarriedoutinJapanon41bridgesthrough
longtermexposuretests(9years)between1981and1993bythreeorganisations:thePublic
Works Research Institute of the Construction Ministry, the Japan Association of Steel Bridge
Construction,andtheKozaiClub[60].Figure10showstherelationshipbetweenatmospheric
salinity and WS corrosion, differentiating between two zones according to the adherent or
nonadherentnatureoftherustformed.

14

Thereseemstobeacriticalairbornesalinityconcentrationofaround3mg Cl/m2.day(0.05
mgNaCl/dm2.day)belowwhichthesteadystatecorrosionrateofconventionalWSislessthan
6m/year,thecriterionalreadymentionedforallowingtheuseofunpaintedWS.
6.2.5.3

Deicingsalts

Extensive use of deicing salts for snow removal, such as sodium chloride (NaCl) and calcium
chloride(CaCl2),beganintheearly1960s.Thecommonuseofsalthasbeenassociatedwitha
significantamountofdamagetotheenvironmentandhighwaystructures(Murray1977)[82].
AlbrechtandNaeemi[83]listthedifferentwaysinwhichsaltcontaminatesthesteelstructure.
Intheliteratureitispossibletofindresultswhichindicatethatspecimensatinteriorandfascia
girdersweremorecorrodedthanroofspecimensbecauseofcontaminationbysaltandwater
fromthebridgedeckduringthewinterseason[71],andresultsobtainedbyZoccola[73]which
showthatroadspray,dirtandsaltswerecarriedbytheairblastcreatedbytheheavytrafficon
the expressway and quickly contaminated horizontal specimens. The prolonged wet period
caused by deposits, chlorides and sulphates in close contact with the steel tended to
acceleratepoulticecorrosion.
Other references related with this point include Park [64] and Hein and Sczyslo [84]. The
resultsofthelatterstudyshowedthatmorecorrosionoccurredinfreezingweather,especially
onthehighway.IntheturbidandfreezingweatherofMerklingen,ASTMA588WScorrosion
on the highway showed no difference from CS corrosion. In contrast, the corrosion rates for
WSandCSattachedtotheguardrailalongthehighwayinDuisburg,whereitwassunnyand
warm,werecomparablylowerthaninMerklingen.
6.2.6. Otherspecificmicroclimates
ResultsobtainedbyKnotkovaetal.[33]inmicroclimatesofchemicalplants,agriculturalareas,
metallurgical and textile production plants, as well as in automotive and streetcar
environments, showed that WS are generally not suited for such applications and their
applicabilitymustbeestablishedbyexperimentalresultsineachspecificcase.

7.

LongtermatmosphericcorrosionofWS

AstheuseofWSincivilengineeringbecamemorecommon,itbecamenecessarytoestimate
inservicecorrosionpenetration.
ThefollowingapproachescanbeusedtopredictthecorrosionofWS:
a)

Directreferencetoexperimentaldataintheliteratureonthebehaviourofthesesteels
in rural, urban, industrial and marine environments after different time periods,
consideringthemostsimilaratmospherestothesiteoftheplannedstructure.

b)

Taking estimations made for CS and adapting them to WS, applying a reduction
coefficient(R)thatmustlogicallyvarywithexposuretimeandconditions.

c)

BasedonempiricalformulaesimilartothosedevelopedforCS.

Forapproacha)itwillbenecessarytousethebroadestpossiblecompilationofcorrosiondata
forWSinadiversityofcircumstances.Table2presentsacompendiumofinformationfoundin
the literature which may be of assistance in this respect. As has been pointed out before, a
factor of uncertainty is that atmospheres are generally classified in purely qualitative terms,

15

based on a subjective assessment of corrosion factors. Moreover, despite the profusion of

dataonbehaviourinthefirst1020yearsofexposure,thereisanotablelackofinformation
forexposurebeyond20years.
InthisapproachcouldbealsoofinteresttopredictatmosphericcorrosionresistanceofaWS
accordingtoitscomposition,i.e.thepresenceandproportionofeachalloyingelementinthe
alloy. This issue is addressed by ASTM standard G101 [85]. For this purpose the standard
establishes two corrosion resistance indices based on two sources of historic atmospheric
corrosiondataforWS,onepublishedbyLarrabeeandCoburn[12]andtheotherbyTownsend
[86]. Basically it consists of fitting the historic corrosion mass loss data to the typical
bilogarithmicequation(seeSection7.1.1),andafterstatisticalanalysistoobtainanindex(for
each historic data source) in the form of an equation in which alloying elements are
independentvariables.
Bothindicesaredimensionless,onascalethatgoesfromzero(noresistancetoatmospheric
corrosion)to10(highresistancetoatmosphericcorrosion).
ASTM standard G101 is currently the only available guide to quantify the atmospheric
corrosionresistanceofWSasafunctionoftheircomposition.However,itspredictionsneedto
betakenwithadegreeofcautionduetoaseriesoflimitationsthataffectthestandard[87].
Thus, for instance, the atmospheric corrosion resistance index obtained from the mass loss
dataofLarrabeeandCoburn[12]cannotbeusedasamethodtopredictcorrosionforother
WSwithadifferentalloyingcomposition.Andontheotherhand,theatmosphericcorrosion
resistance index obtained from the historic mass loss data reported by Townsend [86], who
used a wider compositional range and a larger number of alloying elements, presents the
disadvantage that atmospheric exposure was carried out only in industrial environments,
althoughthisdoesnotnecessarilypreventthebehaviourpredictionfrombeingtransferableto
othertypesofatmospheres.
Withregardtoapproachb),theinformationobtainedintheworldwidesurvey(seechapter4)
hasbeenusedtoprepareTable6,whichshowstherelationship(R)betweencorrosionrates
forWSandthecorrespondingtoCSwhentherustlayerstabilisationtimehasbeenreachedon
these two materials. R is not seen to vary greatly according to the corrosivity of the
atmosphere, and remains in the 0.30.5 range, indicating a notable reduction in the
atmosphericcorrosionofsteelwhenusingWS.
Finally, with regard to approach c), it is first necessary to examine what laws fit better the
behaviourofweatheringsteels.
7.1

ModelsgoverningtheevolutionofatmosphericcorrosionofWSwithexposuretime
(timedependentmodels)

7.1.1 Powermodel
AsBohnenkampetal.notedin1974[35],thecorrosioncurvesoflowalloysteels,likethoseof
mildandcarbonsteels,arereminiscentofthetypicalplotofparaboliclawsorpowerfunctions
(Fig.6).
Thus,mostoftheexperimentalatmosphericcorrosiondatahasbeenfoundtoadheretothe
followingkineticrelationship:

16

C=Atn

(2)

whereCisthecorrosionaftertimet,andAandnareconstants.
Thus, corrosion penetration data is usually fitted to a power model involving logarithmic
transformationoftheexposuretimeandcorrosionpenetration.
logC=logA+nlogt

(3)

This power function (also called the bilogarithmic law) is widely used to predict the
atmosphericcorrosionbehaviourofmetallicmaterialsevenafterlongexposuretimes,andits
accuracyandreliabilityhavebeendemonstratedbyagreatnumberofauthors:Bohnenkamp
et al. [88], Legault and Preban [50], Pourbaix [49], Feliu and Morcillo [89], and Benarie and
Lipfert[90],amongothers.
If the parabolic law is fulfilled, corrosion behaviour will clearly be characterised by only two
parameters:corrosionAafterthefirstyearofexposureandthetimeexponentn.WhenAand
nareknownforagivensteelandexposuresite,thepredictionsmaybeextendedtoanylength
oftime.Thevalueofndoesnotaffectfirstyearcorrosionbutcorrosioninsuccessiveyears,
anditscontributionbecomesincreasinglyimportant;evenwhenthefirstyearcorrosion(A)is
greaterinoneplacethaninanother,itisperfectlypossiblefortheorderofimportanceofthe
corrosionprocesstobereversedafteranumberofyearsduetodifferencesinthevalueofn.
Pourbaix[49]alsostatedthatthebilogarithmiclawisvalidfordifferenttypesofatmospheres
and for a number of materials and is helpful in extrapolating corrosion results up to 2030
yearsfromfouryeartestresults.
The value of exponent n can serve as a diagnostic tool to indicate the nature of the
relationship:linear(n=1),parabolic(n=0.5),cubic(n=0.33),etc.Thisisusuallyexplainedin
termsofdeviationfromparabolicbehaviourasaresultofchangingdiffusionconditionsasthe
filmgrows[50].
According to Benarie and Lipfert [90], equation (2) is a massbalance equation showing that
thediffusionprocessisratedetermining,andthisratedependsonthediffusivepropertiesof
thelayerseparatingthereactants.Theexponentiallaw,equation(2),withncloseto0.5,can
resultfromanidealdiffusioncontrolledmechanismwhenallthecorrosionproductsremainon
themetalsurface.Thissituationseemstooccurinslightlypollutedinlandatmospheres.Onthe
otherhand,nvaluesofmorethan0.5ariseduetoaccelerationofthediffusionprocess(e.g.as
a result of rust detachment by erosion, dissolution, flaking, cracking, etc.). This situation is
typicalofmarineatmospheres,eventhosewithlowchloridecontents.Conversely,nvaluesof
less than 0.5 result from a decrease in the diffusion coefficient with time through
recrystallisation,agglomeration,compaction,etc.oftherustlayer.
Asarule,n<1.Inthespecialcasewhenn=1,themeancorrosionrateforoneyearexposure
isequaltoA,theintersectionofthelineonthebilogarithmicplotwiththeabscissat=1year.
There is no physical sense in n > 1, as n = 1 is the limit for unimpeded diffusion (high
permeable corrosion products or no layer at all). Values of n > 1 occur practically as
exceptions,duetooutliersintheweightlossdeterminations.
The lower n is, the more protective the corrosion product layer on the metal surface is.
Therefore,ncouldbeusedasanindicatorforthephysicochemicalbehaviourofthecorrosion

17

layerandhenceforitsinteractionswiththeatmosphericenvironment.Thevalueofnwould
thusdependbothonthemetalconcerned,thelocalatmosphere,andtheexposureconditions.
Ontheotherhand,theparameterAprovidesacriterionforgaugingshorttermatmospheric
corrosionsusceptibility.Itprovidesameasureoftheinherentreactivityofametalsurfaceas
reflectedinthetendencyforthatsurfacetoproduceacorrosionproductlayerinshortterm
atmosphericexposure[91].
On the basis of Eq. (2), Townsend and Zoccola [42] use as an indicator of the atmospheric
corrosionofdifferentWSthetimetoexhibitalossof250mcalculatedbysolvingEq.(2)for
time:

t * =(

C*
A

1
n

(4)

where t*isthetimeinyearstoachievea250mcorrosionloss; C*istheselectedcorrosion

loss(250m),andAandnareconstantsofEq(2).
Usingtheinformationobtainedintheliteraturesurvey(chapter4)wehaverepresentedusing
bilogarithmic coordinates the graphs of the evolution of WS corrosion with exposure time,
subsequently calculating the exponent n and the correlation coefficient (). Tables 7 and 8
show the results obtained in nonmarine and marine atmospheres, respectively. In the vast
majority of the stations the data is well adapted to straight lines with high correlation
coefficients.
Bywayofexample,Fig.11displaysonabilogarithmicscaletheevolutionoftheatmospheric
corrosionofASTMA242WSindifferenttypesofatmospheres.Ascanbeseen,thepointsare
almostperfectlyalignedandthecorrelationcoefficientsarefairlyclosetounity.
The data series do not adapt well to this rule in only a small percentage of cases with
regressionlinesthatpresentlowercorrelationcoefficients(Fig.12).
Statisticalprocessingofthe nslopesasafunctionofthetypeofatmosphere(Fig.13)shows
the clear tendency of marine atmospheres to present higher average n values (close to 0.5)
thanatmosphereswithoutamarinecomponent,wherenpresentsanaveragevalueofaround
0.3,irrespectiveoftheatmosphericcorrosivitycategoryoftheexposuresite.
Table9liststheaveragevaluesofnforCSandWSindifferenttypesofatmospheres.Thisdata
maybeusefultohelptopredictlongtermatmosphericcorrosionofCSandASTMA242WS
[92].Nodifferencesareseenwithintheruralurbanindustrialgroupinthevalueof nforeach
of these types of atmospheres, nor in the atmospheres of a marine nature, whether or not
theywereclosetotheshoreline.Thelatteragainconfirmsthehighsignificanceofthemarine
characteroftheatmosphere(whichdoesnotdistinguishbetweenthevariouschloridelevels)
inexponent nseeninapriorstatisticalstudy[93].Attentionisalsodrawntothe6570%drop
inthevalueofnforWScomparedtoCSinanyatmosphere.
AstudyofWSbyCoburnetal.[41]whichconsideredtwodifferentanglesofexposure(30and
90) and orientations (facing south and facing north) did not show any great effect of either
variable on the exponent n value of the Eq. (2). The value of n only acquired slightly higher
valuesonthespecimensexposedata90slopeandorientedfacingnorth.

18

Withregardtotheeffectofshelteringonexponent ninEq.(2)forWS,onthebasisof8year
data observed in the UN/ECE exposure programme [26] in a set of 36 testing stations
correspondingto14countries,Fig.14pointstoanimportanteffectwherebyexponentntakes
on significantly higher values in sheltered exposure than in unsheltered exposure. In a
publicationbyTidbladetal.[94]basedondataobtainedinthissameUN/ECEprogramme,the
authorsnotethat nchangesdrasticallyfromabout0.5forshorterexposuretimes(12years)
to less than 0.2 after 8 years of exposure for unsheltered samples, which practically means
that the corrosion products are almost completely protective. In sheltered position the
exponentnchangesfromabout1.0to0.5duringthesameperiod.
7.1.2. Numericalpowermodel
There are examples in the literature [40, 95] of data that does not fully fit Eq. (2). After a
certainexposuretimein which thisfunctionisindeedfollowed,theresults divergefromthe
predictedbehaviourandarefittedbyanotherstraightlinewhoseslopeisshallowerthanthe
first(Fig.15).Thebilineargraphobtainedonloglogcoordinatesinsteadobeyanequationof
thefollowingtype:

C =C 1 t p n1 n2 t n2 (t tp)

(5)

where Cisthecorrosionaftertyears, C1isthefirstyearcorrosion, tp istheduration(inyears)

ofthefirstexposureperiodwhoseslopeis n1,and n2istheslopeofthesecondperiod.One
possiblereasonforthissingularbehaviourmaylieintheformationwithtimeofmorecompact
rustlayerswhichimpedethediffusionofthereactivespeciesthatparticipateinthecorrosion
reactions.
McCuenetal.[48]proposedtoimprovethepowermodelbynumericallyfittingcoefficients A
and nwiththenonlinearleastsquaresmethoddirectlytotheactualvaluesofvariables Cand
t, not the logarithms of the variables, since the logarithmic transformation lends too much
weighttothepenetrationdataforshorterexposure.Thiseliminatestheoverallbias,andmore
accuratelypredictspenetrationforlongerexposuretimes.Theyrefertothisnewmodelasthe
numericalpowermodel,whichhasthesamefunctionalformasthebilogarithmicmodel.
In order to know what law best predicts the atmospheric corrosion of WS, it is necessary to
refer to a greater volume of information that makes reference to exposure times of > 1020
years.Inthisrespect,McCuenetal.[48]considerthatpenetrationdataforexposuretimesof
atleasttenyearsisneededtoreliablyestimatethe penetrationattheendofa50100year
servicelife,incontrasttotheopinionofPourbaix[49],whostatedthat1to4yearsexposure
datawassufficienttomakelongtermforecasts(2030years).
7.1.3. Powerlinearmodel
McCuen et al. [48] saw that WS corrosion penetration data revealed behaviour differences
thatcouldnotallbeexplainedbytheparabolicmodelandthuspreferredacompositemodel
(powerlinear model) consisting of a power function for short exposure times, up to 3 to 5
years, followed by a linear function for longer exposure times. This model is similar to that
usedtodevelopstandardISO9224[96],whichenvisagestwoexposureperiodswithdifferent
corrosionkinetics.Inthefirstperiod,coveringthefirsttenyearsofexposure,thegrowthlawis

19

parabolic(averagecorrosionrate,rav),whileinsecondperiod,fortimesofmorethan10years,
thebehaviourislinear(steadystatecorrosionrate,rlin).
ISO 9224 offers information on guiding corrosion values for CS and WS in each time period
according to the atmospheric corrosivity as defined in ISO 9223 [57]. The guiding corrosion
values are based on experience obtained with a large number of exposure sites and service
performances.
ThequestionofwhetherthislawprovidesabetterpredictionofWScorrosioncannotbefully
answereduntilagreater volumeofdataisavailableforanalysis,withreference toexposure
timesofatleast20years.McCuenandAlbrechtcomparedbothmodels(thepowermodeland
the power linear model) using atmospheric corrosion data reported for WS in the United
Statesandconcludedthattheexperimentaldatafittedthepowerlinearmodelbetterthanthe
power model and thus provided more accurate predictions of longterm atmospheric
corrosion [48]. The improvement in accuracy was greatest for plain carbon steel and copper
steeldata,andlessforASTMA588andA242steeldata.
7.1.4. Bilinearmodel
Some refinement of the powerlinear model has been reported for Albrecht and Hall [7],
where the authors proposed a new bilinear model based on ISO 9224, called modified ISO
9224 (Fig. 16a), as well as an adjustment of this new bilinear model that accounts for a
modifiedcorrosionrateduringthefirstyearofexposureandasteadystateduringsubsequent
years. An application of this adjustment for the upper and lower curves in the medium
corrosivitycategoryC3isreportedinFigure16b.
7.2.

Damage(doseresponse)functions

Differentmodelshaverecentlybeendevelopedwiththeaimofgeneralisingthecorrosionloss
overtimefordifferentenvironments,reportingclimateandpollutantvariablesasindependent
factors.Twoofthesemodelsarepresentedbelow.
Several doseresponse functions have been developed within the International Cooperative
Programme(ICP)onEffectsonMaterials,includingHistoricandCulturalMonumentsinthe
framework of the UN/ECE convention on long range transboundary air pollution [97]. These
functionshavebeenformulatedforWS(Table10)andarebasedonbothlongtermexposures
and trend analysis based on repeated oneyear measurements, also taking into account
unshelteredorshelteredexposure.ThedegradationofWSovertimeisexpressedbymeansof
massloss(ML)asafunctionofclimaticparameters(Rh,T)andSO2concentrationasreported
inTable10.Furtherdetailscanbefoundinreference[97].
Finally, Klinesmith et al. [98] mention that all variation related to environmental conditions
appearsaserrorvariationinthetimedependentmodelsformodelsthatpredictcorrosionloss
as a function of time only. Furthermore, timedependent models will yield inaccurate
predictionswhenusedtoestimatecorrosionlossinenvironmentsthataredifferentfromthe
environmentwherethemodelwascalibrated.Toovercomethisproblem,amodelisproposed
that incorporates multiple environmental factors such as time of wetness (TOW), sulphur
dioxide(SO2),chloridecontent(Cl)andtemperature(T):

20

SO
Cl
( TOW
)
(
1+
)
(
1 + )e
D
F
H

C =At

I J(T+To)

(6)

whereA,B,D,E,F,G,H,I,JandT0areempiricalcoefficients.
Themodelwasformulatedfordifferentmetalsandtheresultsindicatethatitwasreliablefor
useinabroadrangeofconditionsorlocations.
According to Landolfo et al. [99], as far as Klinesmith's corrosion model is concerned, the
thicknesslossvaluespredictedforoneyearexposureareveryclosetoEN12500[100],butin
thelongtermthemodelprovidesanaveragethicknesslossofabout1.7mmforadesignlife
of 50 years, which is approximately 10times more than the forecasts given by the other
models.
Inordertoevaluatethedifferencesbetweenseveralcorrosionmodelsandcorrosionratesfor
WS, Landolfo et al. have undertaken a comparison [99]. Good agreement between the
selected corrosion models and corrosion rates can be observed in the case of WS for both
shortandlongtermexposures.
Despite mentioned inconvenients, recent studies on the longterm atmospheric corrosion of
WScontinuetousethepowerfunction(C=Atn)becauseofitssimplicity[86,101],although
ignoringthelinearpart(paraboliclinearmodel)willintroduceconsiderableerrorinthickness
estimatesforlongexposuretimes.

8.

Conclusions

Thefollowingconclusionscanbedrawnfromtheindepthbibliographicreviewofinformation
ontheatmosphericcorrosiondataofconventionalWS:
a)

The rust layer stabilisation (consolidation) time decreases as the corrosivity of the
exposureatmosphererises,droppingfrom68yearsinlesscorrosiveatmospheres(C2
C3)to46yearsinmorecorrosiveatmospheres(C3C5).

b)

ThesteadystatecorrosionrateofWSincreaseswiththecorrosivityoftheatmosphere.
The use of unpainted WS (corrosion rate 6 m/year) is restricted to nonmarine
atmosphereswithlowcorrosivity,C2C3(low).

c)

The exposure geometry (angle of exposure, orientation) influences WS corrosion

although its impact is only moderate. In contrast, the presence of atmospheric
pollutantsSO2inurbanandindustrialatmospheresandClinmarineatmosphereshave
a rather considerable effect on WS corrosion. There seem to be a critical SO2 level of
close to 20 mg SO2/m2.d and a critical Cl level of around 3 mg/m2.d above which the
applicationofunpaintedconventionalWSisnotfeasible.

d)

Withregardtotheeffectofshelteredexposure,inatmospheresoflowcorrosivity(C2
C3) practically no effect is seen, but when the corrosivity category approaches C4,
shelteredexposureacceleratestheWScorrosionrate.

e)

Continuouslymoistexposuredoesnotallowtheformationofprotectiverustlayersdue
to the lack of the necessary wet/dry cycles. Furthermore, as in indoor exposure, WS
corrosion is not so different to that experienced by CS, and thus the use of WS is not
justified.

21

f)

It is very common to use a power function (C = Atn), or its logarithmic transformation

(logC=logA+nlogt),topredictthelongtermatmosphericcorrosionofWS. nvalues
forWS,whichareinaverage33%lowerthanforCS,arehigherinmarineatmospheres
(n=0.5)thaninothertypesofatmospheres(ruralurbanindustrial)(n=0.3).

22

9.

References

[1]

T.Murata,Weatheringsteel,in:R.W.Revie(Ed.),Uhlig'sCorrosionHandbook,J.Wiley
&Sons,NewYork,2000.

[2]

ASTMA242/A242M04,Standardspecificationforhighstrengthlowalloystructural
steel,AmericanSocietyforTestingandMaterials,Philadelphia,2007.

[3]

ASTMA709/A709M,Standardspecificationforstructuralsteelforbridges,American
SocietyforTestingandMaterials,Philadelphia,2009.

[4]

A588/A588M,Standardspecificationforhighstrengthlowalloystructuralsteelwith
50 ksi [345 MPa] minimum yield point to 4in. [100 mm] thick, American Society for
TestingandMaterials,Philadelphia,2005.

[5]

V. Kucera, E. Mattsson, Atmospheric corrosion, in: F. Mansfeld (Ed.), Corrosion

Mechanisms,MarcelDekker,NewYork,1987,pp.211284.

[6]

C.Leygraf,T.Graedel,AtmosphericCorrosion,ElectrochemicalSocietySeries,J.Wiley
&Sons,NewYork,USA,2000.

[7]

P. Albrecht, T.T. Hall, Atmospheric corrosion resistance of structural steels, J. Mater.

CivilEng.15(2003)224.

[8]

F.B. Fletcher, Corrosion of weathering steels, in: Vol 13B: Corrosion: Materials, ASM
Handbook,2005,pp.2834.

[9]

D.M. Buck, Copper in steel The influence on corrosion, J. Ind. Eng. Chem. 5 (1913)
447452.

[10]

D.M.Buck,Recentprogressincorrosionresistance,IronAge(1915)12311239.

[11]

G.Smith,Steelsfitforthecountryside,NewScientist(1971)211213.

[12]

C.P. Larrabee, S.K. Coburn, The atmospheric corrosion of steels as influenced by

changes in chemical composition, First International Congress on Metallic Corrosion,
London,1961,pp.279285.

[13]

W.K. Boyd, Corrosion of metals in the atmosphere, in: Metals and Ceramics
InformationCenter,Columbus,1974,pp.118.

[14]

H.R. Copson, Longtime atmospheric corrosion tests on lowalloy steels, Proceedings

ASTM60,1960,pp.116.

[15]

Material Specification 531, Weathering fine grain structural steel, CORTEN A,

ThyssenKruppSteelEurope,March2007.

[16]

Material Specification 532, Weathering fine grain structural steel, CORTEN B,

ThyssenKruppSteelEurope,August2005.

[17]

A. Azizinamini, High Performance Steel: New Horizon in Steel Bridge Construction,

TransportationResearchBoard,Washington,D.C.,1998.

[18]

A.D. Wilson, Properties of recent production of A709 HPS 70W bridge steels, in:
International Symposium on Steel for Fabricated Structures, ASM International,
Cincinnati,1999,pp.4149.

[19]

R.J. Schmitt, W.P Gallagher, Unpainted high strength low alloy steel, Materials
Protection8(1969)7077.

23

[20]

I.Matsushima,Y.Ishizu,T.Ueno,M.Kanasashi,K.Horikawa,Effectofstructuraland
environmental factors on the practical use of lowalloy weathering steel, Corrosion
Engineering23(1974)177182.

[21]

S.Feliu,M.Morcillo,CorrosinyProteccindelosMetalesenlaAtmsfera,Bellaterra,
Barcelona,1982.

[22]

A.Bragard,H.Bonnarens,Atmosphericconditionsanddurabilityofweatheringsteels,
C.R.M.ReportN.57,December,1980.

[23]

A.A.Bragard,H.E.Bonnarens,Predictionatlongtermsoftheatmosphericcorrosionof
structuralsteelsfromshorttermexperimentaldata,in:S.W.DeanJr.,E.C.Rhea(Eds.),
Atmospheric Corrosion of Metals, ASTM STP 767, American Society for Testing and
Materials,1982,pp.339358.

[24]

R. de O. Vianna, M.F. Cunha, A.C. Dutra, Comparative behavior of carbon steel and
weatheringstelsexposedforfiveyearstodifferentclimatesinBrazil,Proceedings9th
International Congress on Metalic Corrosion, Toronto, Vol. 3, National Research
Council,Ed,1984,pp.211215.

[25]

A.C.Dutra,R.deO.Vianna,AtmosphericcorrosiontestinginBrazil,in:W.H.Ailor(Ed.),
AtmosphericCorrosion,TheElectrochemicalSociety,JohnWileyandSons,NewYork,
1982,pp.755774.

[26]

UN/ECE International Cooperative Programme on Effects on Materials including

HistoricandCulturalMonuments,ReportN.22:Corrosionattackonweatheringsteel,
zincandaluminium.Evaluationafter8yearsofexposure,SVUOM,Prague,May1998.

[27]

W. Hou, Atmospheric corrosion of steels, Proceedings 11th International Congress on

MetallicCorrosion,Firenze,1990,pp.2.552.61.

[28]

W. Hou, C. Liang, Eightyear atmospheric corrosion exposure of steels in China,

Corrosion55(1999)6573.

[29]

W.T. Hou, C.F. Liang, Effects of alloying on atmospheric corrosion of steels, in: H.E.
Townsend (Ed.), Outdoor Atmospheric Corrosion, ASTM STP 1421, American Society
forTestingandMaterials,WestConshohocken,2002,pp.368377.

[30]

W.Hou,C.Liang,Atmosphericcorrosionpredictionofsteels,Corrosion60(2004)313
322.

[31]

D. Knotkova, L. Rozlivka, J. Vlckova, Corrosion properties of weathering steels,

Proceedingsof8thInternationalCongressonMetallicCorrosion,Frankfurt/Main,Vol.
II,1981,pp.17921799.

[32]

D. Knotkova, J. Vlckova, J. Honzak, Atmospheric corrosion of weathering steels, in:

S.W. Dean Jr., E.C. Rhea (Eds.), Atmospheric Corrosion of Metals, ASTM STP 767,
AmericanSocietyforTestingandMaterials,1982,pp.744.

[33]

D.Knotkova,K.Barton,M.Cerny,AtmosphericcorrosiontestinginCzechoslovakia,in:
W.H.Ailor(Ed.),AtmosphericCorrosion,TheElectrochemicalSociety,JohnWileyand
Sons,NewYork,1982,pp.9911014.

[34]

G. Burgmann, D. Grimme, Investigations into the atmospheric corrosion of plain

carbon and lowalloy steels at various environmnetal conditions, Stahl Eisen 100
(1980)641650.

[35]

K.Bohnenkamp,G.Burgmann,W.Schwenk,Corrosionatmospheriquedel'acierdoux.
Expositiondel'acierauxintemperies,GalvanoOrgano445(1974)587589.

24

[36]

T. Moroishi, J. Satake, Effect of alloying elements on atmospheric corrosion of high

tensilestrengthsteels,TransactionsoftheIronandSteelInstituteofJapan13(1973)
165165.

[37]

T. Fukushima, N. Satu, Y. Hisamatsu, J. Matsushima, Y. Aoyama, Atmospheric

corrosion testing in Japan, in: W.H. Ailor (Ed.), Atmospheric Corrosion, The
ElectrochemicalSociety,JohnWileyandSons,NewYork,1982,pp.841872.

[38]

C.R. Southwell, J.D. Bultman, A.L. Alexander, Corrosion of metals in tropical

environmentsFinalreportof16yearexposures,Mater.Performance15(1976)926.

[39]

F.I.Wei,AtmosphericcorrosionofcarbonsteelsandweatheringsteelsinTaiwan,Br.
Corros.J.26(1991)209214.

[40]

J.H.Wang,F.I.Wei,Y.S.Chang,H.C.Shih,Thecorrosionmechanismsofcarbonsteel
andweatheringsteelinSO2pollutedatmospheres,Mater.Chem.Phys.47(1997)18.

[41]

S.K.Coburn,M.E.Komp,S.C.Lore,Atmosphericcorrosionratesofweatheringsteelsat
test sites in the eastern United States. Effect of environment and test panel
orientation,in:W.W.Kirk,H.H.Lawson(Eds.),AtmosphericCorrosion,ASTMSTP1239,
AmericanSocietyforTestingandMaterials,Philadelphia,1995,pp.101113.

[42]

H.E. Townsend, J. Zoccola, Eightyear atmospheric corrosion performance of

weatheringsteelinindustrial,ruralandmarineenvironments,in:S.W.DeanJr.,E.C.
Rhea (Eds.), Atmospheric Corrosion of Metals, ASTM STP 767, American Society for
TestingandMaterials,1982,pp.4559.

[43]

C.R. Shastry, J.J. Friel, H.E. Townsend, Sixteenyear atmospheric corrosion

performance of weathering steels in marine, rural and industrial environments, in:
S.W.Dean,T.S.Lee(Eds.),DegradationofMetalsintheAtmosphere,ASTMSTP965,
AmericanSocietyofTestingandMaterials,Philadelphia,1988,pp.515.

[44]

F.H. Haynie, J.B. Upham, Effects of atmospheric pollutants on corrosion behavior of

steels,MaterialsProtectionandPerformance10(1971)1821.

[45]

G.B. Mannweiler, Corrosion test results of fifteen ferrous metals after sevenyears
atmospheric exposure, in: Metal Corrosion in the Atmosphere, ASTM STP 435,
AmericanSocietyforTestingandMaterials,1968,pp.211222.

[46]

H.R. Copson, Atmospheric corrosion of low alloy steels, Proceedings ASTM 52 (1952)
10051026.

[47]

J.B.Horton,Thecomposition,structureandgrowthoftheatmosphericrustonvarious
steels,Ph.Thesis,LehighUniversity,Bethlehem,1964.

[48]

R.H.McCuen,P.Albrecht,J.G.Cheng,Anewapproachtopowermodelregressionof
corrosion penetration data, in: V. Chaker (Ed.), Corrosion Forms and Control for
Infrastructure, ASTM STP 1137, American Society for Testing and Materials,
Philadelphia,1992,pp.4676.

[49]

M.Pourbaix,Thelinearbilogaritmiclawforatmosphericcorrosion,in:W.H.Ailor(Ed.),
AtmosphericCorrosion,TheElectrochemicalSociety,JohnWileyandSons,NewYork,
1982,pp.107121.

[50]

R.A. Legault, A.G. Preban, Kinetics of atmospheric corrosion of lowalloy steels in an

industrialenvironment,Corrosion(NACE)31(1975)117122.

[51]

Ch.Cao,SomeresultsoflongtermatmosphericexposuretestsofsteelsinChina,16th
International Corrosion Congress, Chinese Society for Corrosion and Protection,
Beijing,2005.

25

[52]

C. Liang, W. Hou, Sixteen year atmospheric corrosion exposure study of steels, 16th
International Corrosion Congress, Chinese Society for Corrosion and Protection,
Beijing,2005.

[53]

H.E.Townsend,Atmosphericcorrosionperformanceofquenchedandtempered,high
strengthweatheringsteel,Corrosion(NACE)56(2000)883886.

[54]

L. Damian, R. Fako, Weathering structural steels corrosion in atmospheres ofvarious

degreesofpollutioninRomania,Mater.Corros.51(2000)574578.

[55]

A.U. LeuenbergerMinger, B. Buchmann, M. Faller, P. Richner, M. Zobeli, Dose

response functions for weathering steel, copper and zinc obtained from a fouryear
exposureprogrammeinSwitzerland,Corros.Sci.44(2002)675687.

[56]

I.Daz,H.Cano,B.Chico,D.DelaFuente,M.Morcillo,Someclarificationsregarding
literature on atmospheric corrosion of weathering steels, International Journal of
Corrosion,ID812192(2012)19.

[57]

ISO9223,Corrosionofmetalsandalloys.Clasifficationofcorrosivityofatmospheres,
InternationalStandardOrganization,Geneve,1991.

[58]

D.C. Cook, An active coating and new protection technology for weathering steel
structures in chloride containing environments, NACE International, Nashville,
Tennessee(USA),2007.

[59]

H. Kihira, M. Kimura, Advancements of weathering steel technologies in Japan,

Corrosion67(2011)113.

[60]

Technical Report: Guideline for designing and construction of bridges by weathering

steel,KozaiClub(Ed.),Tokyo,1993.

[61]

Specificationforhighwaybridges,JapanRoadAssociation,Tokio,2002.

[62]

D.C.Cook,Thecorrosionofhighperformancesteelinadverseenvironments,ISIAME,
Madrid,2004,pp.6372.

[63]

D.C. Cook, Spectroscopic identification of protective and nonprotective corrosion

coatingsonsteelstructuresinmarineenvironments,Corros.Sci.47(2005)25502570.

[64]

K.Park,Corrosionresistanceofweatheringsteels,Ph.Thesis,DepartmentofCiviland
EnvironmentalEngineering,UniversityofMaryland,USA,2004.

[65]

J.Gullmanetal.,Corrosionresistanceofweatheringsteelstypicalcausesofdamage
in building context and their prevention, Bull. 94, Swedish Corrosion Institute,
Stockholm,1985.

[66]

C.P. Larrabee, The effect of specimen position on atmospheric corrosion testing of

steel,Trans.Electrochem.Soc.85(1944)297306.

[67]

G. Schikorr, The atmospheric oxidisation of Iron II, Zeitschrift fur Elektrochemie und
AngewandtePhysikalischeChemie43(1937)697704.

[68]

ASTM G50, Conducting atmospheric corrosion tests on metals, American Society for
TestingandMaterials,Philadelphia,1991.

[69]

F.L.LaQue,Corrosiontesting,ProceedingsASTM51(1951)495582.

[70]

C.P. Larrabee, Corrosion resistance of highstrength lowalloy steels as influenced by

compositionandenvironment,Corrosion(NACE)9(1953)259271.

26

[71]

B. Cosaboom, G. Mehalchick, J. Zoccola, Bridge construction with unpainted high

trength lowalloy steel: eightyear progress report, New Yersey Department of
Transportation,NewYersey,1979.

[72]

J. Zoccola, A.J. Permoda, L.T. Oehler, J.B. Horton, Performance of Mayari R

weatheringsteel(ASTMA242)inbridges,eightmileroadoverJohnLodgeexpressway,
Detroit,Michigan,BethlehemSteelCorporation,1970.

[73]

J. Zoccola, Eight year corrosion test report Eight mile road interchange, Bethlehem
Steel,Bethlehem,1976.

[74]

S. Oesch, P. Heimgartner, Environmental effects on metallic materials Results of an

outdoor exposure programme running in Switzerland, Mater. Corros. 47 (1996) 425
438.

[75]

T. Moroishi, J. Satake, The influence of the inclination and direction of specimen

surfaceonatmosphericcorrosionofsteels,TetsutoHagane59(1)(1973)125130.

[76]

J.Satake,T.Moroishi,Y.Nishida,S.Tanaka,SumitomoMetals22(4)(1970)516520.
(inJapanese)

[77]

M. McKenzie, The corrosion performance of weathering steel in highway bridges, in:

W.H.Ailor(Ed.),AtmosphericCorrosion,TheElectrochemicalSociety,JohnWileyand
Sons,NewYork,1982,pp.717736.

[78]

NationalAcademyofSciences,Ananalysisofatmosphericcorrosiontestsinlowalloy
steelsApplicabilityoftestresultstohighwaybridges,HighwayResearchRecord,No.
204,Nat.Acad.Sci.,WashingtonDC.,1967.

[79]

J.M.Costa,M.Morcillo,S.Feliu,Effectofenvironmentalparametersonatmospheric
corrosion of metals, in: P.N. Cheremisinoff (Ed.), Encyclopedia of Environmental
Control Technology: Air Pollution Control, Gulf Publishing Company, Houston, 1989,
pp.197238.

[80]

D.D.N. Singh, S. Yadav, J.K. Saha, Role of climatic conditions on corrosion

characteristicsofstructuralsteels,Corros.Sci.50(2008)93110.

[81]

J. Satake, T. Moroishi, Various factors affecting atmospheric corrosion of steels, 5th

InternationalCongressonMetallicCorrosion,Tokyo,Japan,1972,pp.744749.

[82]

D.M.Murray,Economicanalysisoftheenvironmentalimpactofhighwaydeicingsalts,
Transportation Research Record 647, Transportation Research Board, Washington,
D.C,1977.

[83]

A. Naeemi, P. Albrecht, Atmospheric corrosion of weathering steels, Proceedings of

9th International Congress of Metallic Corrosion, Toronto, Vol. 1, National Research
Council,Ed,1984,pp.418427.

[84]

W. Hein, S. Sczyslo, Corrosion behaviour of weathering steels under deicing salt

exposure, Interim Report n 4, Bundesanstalt Fuer Wasserbaw, Federal Republic of
Germany,1981.

[85]

ASTMG10104,Standardguideforestimatingtheatmosphericcorrosionresistanceof
lowalloysteels,AmericanSocietyforTestingandMaterials,Philadelphia,2004.

[86]

H.E.Townsend,Theeffectsofalloyingelementsonthecorrosionofsteelinindustrial
atmospheres, Proceedings 14th International Corrosion Congress, Cape Town,
CorrosionInstituteofSouthernAfrica,1999.

27

[87]

H.E.Townsend,Estimatingtheatmosphericcorrosionresistanceofweatheringsteels,
in: H.E. Townsend (Ed), Outdoor Atmospheric Corrosion, ASTM STP 1421, American
SocietyforTestingandMaterials,PA,USA,2002,pp.292300.

[88]

K.Bohnenkamp,G.Burgmann,W.Schwenk,Investigationsofatmosphericcorrosion
of plain carbon and lowalloy steels in industry, country and sea air, Stahl Eisen 93
(1973)10541060.

[89]

S. Feliu, M. Morcillo, Atmospheric corrosion testing in Spain, in: W.H. Ailor (Ed.),
AtmosphericCorrosion,J.Wiley&Sons,NewYork,1982,pp.913922.

[90]

M.Benarie,F.L.Lipfert,Ageneralcorrosionfunctionintermsofatmosphericpollutant
concentrationsandrainpH,Atmos.Environ.20(1986)19471958.

[91]

R.A.Legault, V.P.Pearson,Atmosphericcorrosioninmarineenvironments,Corrosion
(NACE)34(1978)433437.

[92]

M. Morcillo, J. Simancas, S. Feliu, Longterm atmospheric corrosion in Spain: results

after13to16yearsofexposureandcomparisonwithworldwidedata,in:W.W.Kirk,
H.H. Lawson (Ed.), Atmospheric Corrosion, ASTM STP 1239, American Society for
TestingandMaterials,Philadelphia,1995,pp.195214.

[93]

S. Feliu, M. Morcillo, S. Feliu Jr, The prediction of atmospheric corrosion from

meteorologicalandpollutionparameters.2.Longtermforecasts,Corros.Sci.34(1993)
415422.

[94]

J.Tidblad,J.K.Kreislova,D.Knotkova,V.Kucera,Exposureofweatheringsteelinthe
UN/ECEICPMaterialsProgramme,InternationalWorkshoponAtmosphericCorrosion
andWeatheringSteels,CartagenadeIndias,2004.

[95]

M. Morcillo, S. Feliu, J. Simancas, Deviation from bilogarithmic law for atmospheric

corrosionofsteel,Br.Corros.J.28(1993)5052.

[96]

ISO9224,CorrosionofmetalsandalloysCorrosivityofatmospheresGuidingvalues
forthecorrosivitycategories,InternationalStandardsOrganization,Geneve,1992.

[97]

V. Kucera, Mapping Effects on Materials in Manual Mapping Critical Load, ICP

Materials, Coordination Centre: Stockholm, 2004. Available online:
http://icpmapping.org.
D.E. Klinesmith, R.H. McCuen, P. Albrecht, Effect of environmental conditions on
corrosionrates,J.Mater.CivilEng.19(2007)121129.

[98]
[99]

R. Landolfo, L. Cascini, F. Portioli, Modeling of metal structure corrosion damage: A

stateoftheartreport,Sustainability2(2010)21632175.

[100] EN12500,Corrosionlikelihoodinatmosphericenvironment,EuropeanCommitteefor
Standardization(CEN),Brussels,2000.
[101] S.W. Dean, D. Knotkova, J.K. Kreislova, ISOCORRAG International Atmospheric
Exposure Program: Summary of Results, ASTM Series 71, ASTM International, West
Conshohocken,2010.

28

FIGURECAPTIONS
Figure1.

AtmosphericcorrosionofCorTenBsteelanditsevolutionwithexposuretimein
the industrial atmosphere of Kearny [12]. Comparison with Cubearing steel and
plaincarbonsteel(CS).

Figure2.

Determinationofrustlayerstabilisationtimeandsteadystatecorrosionratefrom
a plot of exponential decrease function obtained with corrosion rate data at
differentexposuretimes.

Figure3.

Boxwhiskerplotsofrustlayerstabilisationtimeofweatheringsteelasafunction
ofatmosphericcorrosivitycategory(ISO9223[57]).

Figure4.

Boxwhiskerplotsofweatheringsteelsteadystatecorrosionrateasafunctionof
atmospheric corrosivity category (ISO 9223 [57]). Nonmarine (ruralurban
industrial)atmospheres.

Figure5.

Boxwhiskerplotsofweatheringsteelsteadystatecorrosionrateasafunctionof
atmosphericcorrosivitycategory(ISO9223[57]).Marineatmospheres.

Figure6.

Typicalplotsofcorrosionversusexposuretimeforweatheringsteels(ASTMA242
(a),duetoSchmittandGallagher[19],andASTMA588(b),duetoTownsendand
Zoccola [42] and Shastry, Friel and Townsend [43]) in different types of
atmospheres.

Figure7.

Corrosion versus exposure time for weathering steel (WS) and reference plain
carbon steel (CS) in atmospheres with different SO2 content. Plots have been
elaboratedfromgraphsshowedinreference[32].

Figure8.

Corrosion of weathering steel in outdoor atmospheres as a function of SO2

pollutionlevel[32].

Figure9.

Variation of atmospheric corrosion rate of conventional weathering steel with

atmospheric SO2 content. Constructed from data corresponding to 8year
exposure,providedbyUNECE/PICMaterials[26].

Figure10. Influenceofairbornesalts(atmosphericsalinity)oncorrosionrateofconventional
weatheringsteel.Thefigurehasbeenobtainedfromreferences[60]and[62].
Figure11. Typical loglog plots of weathering steel corrosion versus exposure time at
differenttestsites.
Figure12. Loglogplotsofweatheringsteelcorrosionversusexposuretimeindataseriesnot
welladaptedtothepowerfunction(C=Atn),withlowcorrelationcoefficients().
Figure13. Boxwhisker plots of n values in power function (C = Atn) for different types of
atmospheres.
Figure14. Boxwhisker plots of n values in power function (C = Atn) for sheltered and
unshelteredexposures.
Figure15. Bilinearloglogplotsofweatheringsteelcorrosionversusexposuretimefordata
seriesnotwelladaptedtothepowerfunction(C=Atn)[95].

29

Figure16. Bilinear plots of weathering steel corrosion versus exposure time [7]: (a)
ModifiedISO9224fordifferentISO9223[57]corrosivitycategories;(b)Albrecht
etal.adjustmentandmodifiedISO9224forC3corrosivitycategory.

Cor-Ten B
Cu-bearing steel
Plain carbon steel (CS)

300

Corrosion, m

250

200

150

100

50

0
0

10

15

Time, years

Figure1.

20

Corrosion rate, m/y

y=A1 exp(

x
)+y o
t1

Steady-state
10%

Stabilization time

Time, years

Figure2

All types of atmospheres

140

80
60
40

C2

20

C3
C4
25
50
80
Corrosivity category (ISO 9223)

100

st

1 year CS corrosion, m

120

C5

WS: ASTM A-242

4-6 years

6-7 years

7-8 years

Stabilization time, years

Figure3

Steady-state corrosion rate, m/year

30

Rural, urban and industrial atmospheres

WS: ASTM A-242

25

20

15

10

6 m/y

C2

C3

C4

ISO corrosivity category (ISO 9223)

Figure4

C5

Marine atmospheres
WS: ASTM A-242

Steady-state corrosion rate, m/year

24
22

y = 6.73 + 0.11x
R = 0.783

20
18
16
14
12
10
8
6
4

6 m/y

ISO 9223 CORROSIVITY CATEGORY

2
C2

0
0

10

C3

20

30

40

C4

50

60

70

C5

80

st

90 100 110 120 130 140

1 year CS corrosion, m

Figure5

(a)

Corrosion, m

50

40

30

ASTM A-242
(CORTEN A)

20

Rural (South Bend)

Industrial (Kearny)
Marine (Kure Beach, 250)

10

0
0

10

15

20

Time, years

180
160
140

Corrosion, m

(b)

Rural (Saylorsburg)
Heavily Travelled Expressway (Newark)
Industrial (Bethlehem)
Marine (Kure Beach, 250)

120
100
80
60
40

ASTM A-588
(CORTEN B)

20
0
0

10

Time, years

Figure6

12

14

16

Heavily polluted atmospheres (> 90 mg SO2/m d)

180

CS

Urban and weakly polluted industrial atmospheres

2
(40-90 mg SO2/m d)

160

Rural atmosphere (< 40 mg SO2/m d)

Corrosion, m

140

WS

120
100

CS

80
60

CS

40

WS

20

WS

0
0

Time, years

Figure7

Corrosion,g/m 2

1000

500

50

100
2

SO2,mg/m .d

Figure8

150

100

SO2, mg/m .d

10

0.1
0

10

12

14

16

Corrosion rate, m/y

Figure9

18

20

22

24

Steadystatecorrosionrate,m/y

200

.
20
6m/y
2
0.001

0.01

0.05

0.1

Salinity,mgNaCl/dm2.day
Adherent rust layer (protective)
Not adherent rust layer (nonprotective)

Figure10

100

Quingdao, China (marine atmosphere)

= 0.999

80

Detroit, USA (urban atmosphere)

= 0.999

Corrosion, m

60

40

Saylorsburg, USA (rural atmosphere)

= 0.999

20

0.1

10

Time, years

Figure11

Corrosion, g/m

400

300

Kopisty (Czech Republic): = 0.38

Steubenville (USA): = 0.79
200
1

10

Time, years

Figure12

1.0

Exponent n in Eq. (2)

0.8

Rural, urban and

industrial atmospheres

Marine atmospheres

0.6

0.4

0.2

0.0

Type of atmosphere

Figure13

1.0

Exponent n in Eq. (2)

0.8
0.6
0.4

SHELTERED

0.2
0.0

UNSHELTERED

-0.2
-0.4

Type of exposure

Figure14

C = C 1 t p n1 n2 t n2

log(corrosion)

Slopen2
logC
Slopen1

logC1

logtp
log(time,years)

Figure15

logt

ModifiedISO9224
(a)

350

C5

Corrosion (m)

300
C4
250
200
C3

150
100

C2
C1

50
0
0

10

15

20

Time(years)

ModifiedISO9224andAlbrechtetal.adjustmentforcorrosivity
categoryC3
200
(b)

Corrosion (m)

Modified ISO9224
Albrechtetal.adjustment

C3

100

0
0

10
Time(years)

Figure16

15

20

Table1.

Chemicalcompositions(weight%)ofcommonlyusedWS
Weatheringsteel
ASTMA242
(CORTENA)[2]
Typical
concentrations
ASTMA588Gr.A
(CORTENB)[4]
Typical
concentrations

Si

Mn

Cu

Cr

Ni

0.15

1.00

0.15

<0.05

0.20

0.15

0.250.40

0.500.80

0.500.65

0.19

0.300.65

0.801.25

0.04

<0.05

0.250.40

0.400.65

0.40

0.020.10

0.04

0.300.40

0.601.00

0.020.30

Table2.

List of atmospheric corrosion exposure tests involving weathering steels carried out in
differentpartsoftheworld.Generalcharacteristicsofthetests
R=Rural,I=Industrial,M=Marine

Country

Reference

Belgium

[2223,49]

Brazil

[2425]

Canada

[26]

China

[2730,5152]

CzechRep.

[26,3133]

Estonia

[26]

Finland

[26]

France

[2223,49]

Germany

[2223,26,34
35,49]

Italy

[2223,26,49]

Japan

[3637]

Norway

[26]

Typeof
Weathering Max.exposure
atmosphere steel(ASTM) time,years
Eupen
R*
4
Lieja
I*
10
A242
OstendeI
M*
10
OstendeII
M*
10
Aracaju13
M
2
Betim
I
5
Canoas
I
5
Cubatao
I
A588
5
Fortaleza
M
2
MadredeDeus
M
2
S.Mateus
R
5
Dorset
R*
A242
8
Beijing
U*
8
Guangzhou
R*
8
Jiangjin
I*
8
Qonghai
R*
Other
8
Quingdao
MI*
8
Wanning
M*
8
Wuham
U*
8
Hurbanovo
RU*
10
KarsperskeHory
R*
8
A242
Kopisty
I*
8
Prague
UI*
10
Lahemaa
R*
A242
8
Ahtari
R*
8
A242
Espoo
U*
8

Helsinki
UI*
8
Biarritz
M*
4
A242
StGermain
RU*
4
Aschaffenburg
U*
8
Bottrop
I*
8
Cuxhaven
M
8
Duisburg
I
16
Dsseldorf
I*
4
Essen
R*
8
A242
Gelsenkirchen
I
4
Langenfeld
R*
8
Mlheim
I
8
Olpe
R
8
GarmischPartenkirchen
R*
8
WaldhofLaugenbrugge
R*
8
Bari
R*
4
Lasaccia
R*
8
A242
Milan
UI*
8

Rome
U*
8
Venice
U*
8
Amagasaki
I*
5,7
Other
Kitakyushu
I*
5
Birkenes
R*
8
Borregaard
I*
A242
8
Oslo
U*
8
Testsite

Table2(continued)

Netherlands

[2223,26,49]

Panama

[38]

Portugal

[26]

Romania

[54]

Russia

[26]

Spain

[26]

Sweden

[26]

Switzerland

[55]

Taiwan

[3940]

United
Kingdom

[2223,26,49]

USA

[12,19,26,41
50,53]

*SO2concentrationdataareavailable

Delft
DenHelder
Eibergen
Vlaardingen
Vredepeel
Wijnandsrade
LimonBay
MirafloresLock
Lisbon
UrbanIndustrial
Rural
Moscow
Bilbao
Madrid
Toledo
Aspureten
Stockholm(Centre)
Stockholm(South)
Bern
Cadenazzo
Davos
Dubendorf
Harkingen
Lagern
Payerne
Sion
ChinaSteel
Nat.Tsing.HuaUniv
Clateringshowsloch
LincolnCathedral
Rye
StokeOrchard
Stratford
WellsCathedral
Bayonne
Bethlehem
BlackIsland
Cincinnati
Columbus
Detroit
EastChicago
Kearny
KureBeach,25m
KureBeach,250m
LosAngeles
Newark
Philadelphia
PointReyes
PotterCounty
Rankin
ResearchTrianglePark
Saylorsburg
SouthBend
StateCollege
Steubenville
Washington
Whiting

I*
M*
R*
UI*
R*
R*
M
R
U*
UI
R
UI*
U*
U*
R*
R*
U*
U*
U
R
R
U
U
R
R
R
I*
RU*
R
U*
M*
R*
I*
U*
I
I
M
U
U
U
I
I
M
M
U
UI
U
M
R
U
R*
R
R
R
I*
U
I

A242

A242
A242
A242and
A588
242
A242
A242

A242

Other

A242

A242
A242/A588
A242
A242
A242
A242
A242
A242/A588
A242
A242/A588
A242
A242/A588
A242
A242
A242/A588
A242
A242
A242/A588
A242
A242
A242
A242
A242

4
4
8
8
8
8
16
16
8
20
20
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
6
5
8
8
4
8
4
8
18.1
16
9.1
5.3
10
5.3
20
20
7
16
5.3
8
5.3
7
16
17
8
16
22
7
8
5.3
7

Table3. Rust layer stabilisation times and steadystate corrosion rates of weathering steels exposed in
atmosphereswithdifferentcorrosivitycategories.Nonmarineatmospheres(rural,urbanandindustrial)
1styearCS
ISOcorrosivity
category[57] corrosion,m

C2

C3

C4

C5

16.8
18.7
19.0
22.4
22.7
23.6
24.0
24.0
24.7
27.1
27.3
28.3
29.0
29.0
29.2
29.6
29.9
29.9
30.0
30.0
30.3
31.2
33.0
33.5
33.6
34.5
34.8
36.1
37.2
37.3
39.1
40.1
42.0
43.0
43.6
43.8
46.6
47.5
50.0
51.0
54.8
64.0
67.0
69.0
71.0
72.0
76.0
76.0
80.0
86.0
88.0
114.0
116.0
119.0
119.0

Testsite

Country

Stabilization
time,years

Ahtari
Aspureten
Dorset
R.Triangle
Rome
Lahemaa
PotterC.
PotterC.
Birkenes
Aschaffenburg
Steubenville
Madrid
Kearny1
Kearny1
Oslo
Eibergen
Casaccia
Clatteringshaws
Saylorsburg
Saylorsburg
StockholmS
Venice
Wijnandsrade
StockholmC
Waldhof
Espoo
Helsinki
Vredepeel
SouthBend
Langenfeld
StokeOrchard
LincolnCath.
Hurbanovo
Olpe
Essen
Vlaardingen
Milan
Bottrop
Newark1
Columbus
Borregaard
Lieja
Withing
Bethlehem2
Bayonne
Rankin
Bethlehem1
Bethlehem1
Newark2
Prague
Mlheim
Duisburg
Kearny3
Kearny2
Kearny2

Finland
Sweden
Canada
USA
Italy
Estonia
USA
USA
Norway
Germany
USA
Spain
USA
USA
Norway
Netherlands
Italy
UK
USA
USA
Sweden
Italy
Netherlands
Sweden
Germany
Finland
Finland
Netherlands
USA
Germany
UK
UK
CzechR.
Germany
Germany
Netherlands
Italy
Germany
USA
USA
Norway
Blgica
USA
USA
USA
USA
USA
USA
USA
CzechR.
Germany
Germany
USA
USA
USA

78
67
78
56
78
67
78
78
56
78
78
67
67
78
67
67
67
56
78
67
67
67
67
78
67
67
78
67
1011
67
67
45
78
67
67
67
67
78
56
67
56
45
67
67
45
56
78
78
56
67
56
56
89
67
45

Steadystate
corrosionrate,
m/y
4.4
5.7
3.8
6.9
4.1
5.6
4.2
8.7
7.6
5.9
5.2
4.3
3.9
5.1
6.1
7.1
6.6
9.7
5.4
9.7
7.5
7.6
7.4
7.1
9.3
8.1
8.9
8.8
3.4
8.9
9.6
19.8
5.9
9.7
9.1
9.7
6.8
8.2
10.3
10.7
13.7
19.2
11.4
10.0
12.4
12.1
5.8
9.7
12.1
10.2
20.0
29.5
4.9
5.4
7.7

Max.
exposure
time,years
8
8
8
8
8
8
16
16
8
8
8
8
16
16
8
8
8
8
16
16
8
8
8
8
8
8
8
8
22
8
8
8
10
8
8
8
8
8
8
10
8
10
7
10
18
10
16
16
7
10
8
16
20
20
7.5

WS(ASTM)
A242
A242
A242
A242
A242
A242
A242
A588
A242
A242
A242
A242
A242
A588
A242
A242
A242
A242
A242
A588
A242
A242
A242
A242
A242
A242
A242
A242
A242
A242
A242
A242
A242
A242
A242
A242
A242
A242
A588
A242
A242
A242
A242
A242
A242
A242
A242
A588
A242
A242
A242
A242
A242
A588
A242

Table4.

Rustlayerstabilisationtimesandsteadystatecorrosionratesofweatheringsteelsexposed
inatmosphereswithdifferentcorrosivitycategories.Marineatmospheres

1styearCS
ISOcorrosivity
category[57] corrosion,m

28.5

C3

C4
C5

32.2
35.0
35.0
39.0
39.0
40.0
40.0
41.0
47.9
52.5
64.0
64.0
65.0
95.0
140.0

Testsite

Country

Stabilization
time,years

Lisbon
WellsCathedral
MirafloresLock
MirafloresLock
KureBeach2
KureBeach2
KureBeach1
KureBeach1
Bilbao
KureBeach5
KureBeach4
LimonBay
LimonBay
Cuxhaven
PointReyes
KureBeach3

Portugal
UK
USA
USA
USA
USA
USA
USA
Spain
USA
USA
USA
USA
Germany
USA
USA

67
56
67
67
56
56
56
45
67
67
45
67
78
56
67
56

Steadystate
corrosionrate,
m/y
6.9
9.8
11.3
12.2
11.0
14.4
10.4
17.4
7.3
8.8
15.1
19.5
15.8
15.3
16.1
20.0

Max.
exposure
time,years
8
8
16
16
16
16
16
16
8
8
16
16
16
8
7
7

WS(ASTM)
A242
A242
A242
A242
A242
A588
A242
A588
A242
A242
A242
A242
A242
A242
A242
A242

Table5.

Rustlayerstabilisationtimesandsteadystatecorrosionratesofweatheringsteels
exposed in atmospheres with different corrosivity categories [26]. Unshelter
versusshelterexposure

ISO
1styearCS
corrosivity
corrosion,m
category[57]
19.0
22.4
22.7
23.6
27.1
28.3
29.6
29.9
30.3
31.2
33.0
33.5
33.6
34.5
34.8
36.1
37.3
39.1
40.1
43.6
43.8
47.5
54.8

C2

C3

C4

Testsite

Country

SO2,
g/m3

Dorset
R.Triangle
Rome
Lahemaa
Aschaffenburg
Madrid
Eibergen
Casaccia
StockholmS
Venice
Wijnandsrade
StockholmC
Waldhof
Espoo
Helsinki
Vredepeel
Langenfeld
StokeOrchard
LincolnCath.
Essen
Vlaardingen
Bottrop
Borregaard

Canada
USA
Italy
Estonia
Germany
Spain
Netherlands
Italy
Sweden
Italy
Netherlands
Sweden
Germany
Finland
Finland
Netherlands
Germany
UK
UK
Germany
Netherlands
Germany
Norway

2.8
9.8
24.5
0.6
14.8
11.7
7.7
6.0
8.4
15.7
10.2
8.5
9.6
8.8
11.8
8.9
19.4
14.6
17.6
24.0
28.2
44.8
34.2

Unshelter
Shelter
Steadystate
Steadystate
Stabilization
Stabilization
corrosion
corrosion
time,years
time,years
rate,m/y
rate,m/y
78
3.8
78
3.7
56
6.9
56
6.9
78
4.1
67
3.3
67
5.6
78
4.4
78
5.9
67
4.6
67
4.3
56
4.1
67
7.1
45
6.2
67
6.6
56
5.4
67
7.5
78
4.2
67
7.6
34
9.0
67
7.4
45
6.5
78
7.1
78
4.2
67
9.3
45
9.1
67
8.1
56
6.3
78
8.9
45
10.1
67
8.8
45
8.3
67
8.9
67
6.6
67
9.6
67
8.7
45
19.8
23
16.2
67
9.1
34
11.4
67
9.7
23
14.8
78
8.2
34
10.7
56
13.7
34
26.8

Table6.

RURAL,URBANANDINDUSTRIALATMOSPHERES

Relation (R) between weathering steel corrosion rate and plain carbon steel
corrosionrateattherustlayerstabilisationtime

ISOcorrosivity
category[57]

1styearCS
corrosion,m

Country

Testsite

R(Cws/Ccs)atthe
stabilizationtime

C2

24.0

USA

PotterCounty

0.33

C3

29.0
30.0
37.0
42.0
43.0

USA
USA
USA
CzechRep.
Germany

Kearny1
Saylorsburg
SouthBend
Hurbanovo
Olpe

0.42
0.32
0.28
0.43
0.57

C4

C5

MARINEATMOSPHERES

35.0
35.0

C4

Panama
Panama

USA
USA
Germany

53.0
64.0
65.0

C5

USA
CzechRep.
Germany
Germany

80.0
86.0
88.0
114.0

C3

USA
Belgium
USA
USA
USA
USA
USA

51.0
64.0
67.0
69.0
71.0
72.0
76.0

140.0

Cws=WS(ASTMA242)corrosionrate
Ccs=Plaincarbonsteelcorrosionrate

USA

Ave.
Columbus
Lieja
Withing
Bethlehem2
Bayonne
Rankin
Bethlehem1
Ave.
Newark2
Prague
Mlheim
Duisburg

0.40
0.49
0.39
0.58
0.38
0.38
0.34
0.35
0.42
0.51
0.41
0.37
0.47

Ave.

0.44

MirafloresLock1
MirafloresLock2

0.52
0.56

Ave.

0.54

KureBeach4
LimonBay
Cuxhaven
Ave.
KureBeach3

0.41
0.53
0.47
0.47
0.29

Table7.

Values of exponent n and correlation coefficient in bilogarithmic plots of the

powerfunction(C=Atn).Nonmarineatmospheres

ISOcorrosivity
category[57]

C2

C3

1styearCS
corrosion,m

Testsite

Country

Exponentn

Correlation
coefficient

5.7
14.1
16.8
16.9
18.7
19.0
21.4
22.4
22.7
23.1
23.6
24.0
24.0
24.7
27.1
27.3
28.2
28.3
28.5
29.0
29.2
29.6
29.9
29.9
30.0
30.2
30.3
31.2
33.0
33.5
33.6
34.5
34.8
36.1
37.3
38.0
39.1
40.1
42.0
43.0
43.6
43.8
45.0
46.6
47.5
50.0

Toledo
LosAngeles
Ahtari
Garmisch
Aspureten
Dorset
Cincinati
R.Triangle
Rome
Moscow
Lahemaa
PotterCounty
Washington
Birkenes
Aschaffenburg
Steubenville
Philadelphia
Madrid
KasperskeHory
Kearny1
Oslo
Eibergen
Casaccia
Clatteringshaws
Saylorsburg
Detroit
StockholmS
Venice
Wijnandsrade
StockholmC
Waldhof
Espoo
Helsinki
Vredepeel
Langenfeld
SouthBend
StokeOrchard
LincolnCath.
Hurbanovo
Olpen
Essen
Vlaardingen
StateCollege
Milan
Bottrop
Newark1

Spain
USA
Finland
Germany
Sweden
Canada
USA
USA
Italy
Russia
Estonia
USA
USA
Norway
Germany
USA
USA
Spain
CzechR.
USA
Norway
Netherlands
Italy
UK
USA
USA
Sweden
Italy
Netherlands
Sweden
Germany
Finland
Finland
Netherlands
Germany
USA
UK
UK
CzechR.
Germany
Germany
Netherlands
USA
Italy
Germany
USA

0.47
0.35
0.47
0.23
0.37
0.26
0.27
0.35
0.19
0.25
0.26
0.27
0.33
0.40
0.33
0.14
0.29
0.11
0.33
0.22
0.24
0.30
0.30
0.43
0.35
0.45
0.24
0.34
0.31
0.24
0.39
0.38
0.35
0.35
0.32
0.16
0.35
0.58
0.14
0.51
0.32
0.30
0.22
0.09
0.28
0.26

0.955
0.996
0.969
0.886
0.994
0.916
0.998
0.995
0.938
0.874
0.979
0.977
0.998
0.996
0.946
0.793
0.996
0.854
0.866
0.948
0.970
0.959
0.993
0.993
0.999
0.999
0.992
0.987
0.974
0.940
0.991
0.992
0.957
0.983
0.998
0.980
0.992
0.999
0.867
0.987
0.983
0.981
0.919
0.888
0.947
0.999

Table7(continued)

C4

C5

51.0
54.8
55.8
64.0
67.0
69.0
70.9
71.0
72.0
76.0
80.0
86.0
88.0
100.0
114.0
116.0
119.0
210.0

Columbus
Borregaard
Prague
Lieja
Whiting
Bethlehem2
Kopisty
Bayonne
Rankin
Bethlehem1
Newark2
Prague
Mlheim
Kopisty
Duisburg
Kearny3
Kearny2
Gelsenkirchen

USA
Norway
CzechR.
Blgica
USA
USA
CzechR.
USA
USA
USA
USA
CzechR.
Germany
CzechRep.
Germany
USA
USA
Germany

0.36
0.35
0.22
0.64
0.17
0.17
0.07
0.31
0.32
0.26
0.15
0.18
0.50
0.51
0.61
0.30
0.15
0.50

0.965
0.998
0.867
0.998
0.947
0.950
0.380
0.996
0.978
0.999
0.974
0.933
0.998
0.999
0.996
1.000
0.973
0.998

Table8.

Values of exponent n and correlation coefficient in bilogarithmic plots of the

powerfunction(C=Atn).Marineatmospheres

ISOcorrosivity
category[57]

C3

C4

C5

1styearCS
corrosion,m
28.5
32.2
35.0
35.0
39.0
40.0
41.0
47.9
52.5
64.0
64.0
65.0
95.0
100.0
140.0
181.0

Testsite

Country

Exponentn

Correlation
coefficient

Lisbon
WellsCathedral
MirafloresLock1
MirafloresLock2
KureBeach2
KureBeach1
Bilbao
KureBeach5
KureBeach4
LimonBay1
LimonBay2
Cuxhaven
PointReyes
Ostende2
KureBeach3
Ostende1

Portugal
UK

0.31
0.38
0.53
0.52
0.58
0.55
0.26
0.45
0.49
0.51
0.48
0.56
0.46
0.72
0.41
1.04

0.988
0.989
1.000
0.998
0.996
0.992
0.973
0.986
0.998
0.998
1.000
0.996
0.989
0.999
0.996
1.000

USA
USA
USA
Spain
USA
USA
USA
Germany
USA
Belgium
USA
Belgium

Table9.

Average values of exponent n in bilogarithmic plots of the power function (C =

Atn) for plain carbon steel (CS) and weathering steel (WS) in nonmarine (rural
urbanindustrial)andmarineatmospheres

Steel
CS(from
ref.[92])
WS(from
tables78)

Nonmarine(RuralUrbanIndustrial)
atmospheres
Numberof
n
Rangeofn
data
Ave.
inEq.(2)

Marineatmospheres
Numberof
data

n
Ave.

Rangeofn
inEq.(2)

25

0.49

0.260.76

13

0.73

0.370.98

58

0.33

0.110.64

16

0.48

0.260.72

Table10.

Doseresponsefunctionsforweatheringsteelinunshelteredandsheltered
exposures[86]

Material
Weatheringsteel(unsheltered)
Weatheringsteel(sheltered)

Massloss,g/m2
ML=34[SO2]0.33exp{0.020Rh+f(T)}t0.13
ML=8.2[SO2]0.24exp{0.025Rh+f(T)}t0.66

where: t = exposure time (years); Rh = relative humidity (%); T = average annual temperature
(C);f(T)=a(T10)whenT<10C,otherwiseb(T10),witha,bbeingconstantvaluesdepending
onthespecificmetal;SO2=sulfurdioxideconcentration(g/m3)