-
CHAPTER 5
Corrosion of Duplex Stainless SteelWeldments
DUPLEX STAINLESS STEELS are two-phase alloys based on the
Fe-Cr-Ni system.These materials typically comprise approxi-mately
equal amounts of body-centered cubic(bcc) ferrite and face-centered
cubic (fcc)austenite in their microstructures and are
char-acterized by their low carbon content (
-
100 / Corrosion of Weldments
Table 1 Composition and pitting resistance equivalent numbers
(PREN) of wrought duplex stainlesssteels covered by UNS
designations
Composition(a), wt%
UNS No.Common
designation C Mn S P Si Cr Ni Mo Cu W N2 PREN
Low-alloy grades (PREN
-
Chapter 5: Corrosion of Duplex Stainless Steel Weldments /
101
Fig. 1 The development of hot-rolled duplex stainless steelsin
terms of their nitrogen versus chromium plus molyb-
denum contents. Source: Ref 1
Fig. 2 Effect of orientation plane on the microstructure of
Fe-22Cr-5.5Ni-3Mo-0.15N wrought duplex stainless steel base
materialelectrolytically etched in 40% NaOH. (a) Parallel to
rolling direction. (b) Transverse to rolling direction. (c) Plan
view. 100
show a typical microstructure representing theweld metal in the
as-welded and the solution-annealed conditions. The darker areas,
as forexample in Fig. 3(d), represent the ferrite phase,while the
lighter areas represent the austenitephase. The austenite islands
are coarser in thesolution-annealed microstructures as comparedto
the as-welded microstructures. The coarsegrain HAZ of the as-welded
condition wasremoved by solution-annealing heat-treatment,as shown
in Fig. 3(b) and Fig. 3(e). Thus, solution-annealing has the
benecial effect ofeliminating the coarse grain HAZ which isusually
detrimental due to carbide precipita-tion. The weld metal
microstructures (Fig. 3c and 3f ) revealed coarser austenite
regions in thesolution-annealed weld metal as compared to theweld
metal in the as-welded condition.
Mechanical and Physical Properties. Du-plex stainless steels
characteristically arestronger than either of their two phases
consid-ered separately. The duplex grades have yield
strengths twice those of the common austeniticgrades while
retaining good ductility (Table 2).In the annealed condition, the
duplex gradeshave outstanding toughness. With the morerecently
developed intermediate and high-alloygrades, it is possible to
retain toughness and cor-rosion resistance after welding.
The coefcient of thermal expansion and theheat-transfer
characteristics of the duplex stain-less steels fall between those
of the ferritic andthe austenitic stainless steels.
When installing a duplex stainless-steel com-ponent in an
existing austenitic stainless-steelstructure, consideration should
be given to therelative strengths and expansion coefcients ofthe
materials. The high strength of the duplexgrade and its relatively
low expansion coef-cient may impose high stresses on the
transitionwelds or the host structure.
Elevated-Temperature Properties. Thehigh alloy content and the
presence of a ferriticmatrix render duplex stainless steels
susceptibleto embrittlement and loss of mechanical proper-ties,
particularly toughness, through prolongedexposure to elevated
temperatures. This iscaused by the precipitation of
intermetallicphases, most notably alpha prime () sigma (),chi (),
and Laves () phases. For this reason,duplex stainless steels are
generally not used attemperatures above 300 C (570 F). Figure
4shows the phases that can be formed in duplexstainless steels over
the temperature range of 300to approximately 1000 C (5701830
F).
Corrosion Resistance. Duplex stainlesssteels comprise a family
of grades with a widerange of corrosion resistance. They are
typicallyhigher in chromium than the corrosion-resistant,
austenitic stainless steels and havemolybdenum contents as high as
4.0%. Thehigher chromium plus molybdenum combina-tion is a
cost-efcient way to achieve good chlo-
(a) (b) (c)
-
102 / Corrosion of Weldments
Fig. 3 Solidication morphologies of fusion welded alloy 2205.
(a) As-welded base metal. (b) As-welded composite region. (c)
As-welded weld metal. (d) Postweld heat treated solution-annealed
base metal. (e) Solution annealed composite region. (f) Solu-
tion annealed weld metal. Source: Ref 2
ride pitting and crevice corrosion resistance.Many duplex
stainless steels exceed the chlo-ride resistance of the common
austenitic stain-less-steel grades and also alloy 904L (UNSN08904)
(Table 3). SAF 2507 (UNS S32750)has chloride resistance comparable
to the 6%molybdenum austenitic stainless steels.
The constraints of achieving the desired bal-ance of phases dene
the amount of nickel in aduplex stainless steel. The resulting
nickel con-tents, however, are sufcient to provide signi-cant benet
in many chemical environments. Asshown in Table 4, alloy 2205 and
Ferralium 255(UNS S32550) compare favorably with type
317L (UNS S31703) and alloy 20 (UNSN08020) in a variety of
chemical environments.
One of the primary reasons for using duplexstainless steels is
their excellent resistance to ClSCC. Compared with conventional
austenitics,they are clearly superior (Fig. 5). The morehighly
alloyed superduplex grades are moreresistant to Cl SCC than those
with lower alloy-ing contents. The SCC resistance in the an-nealed
condition of the superduplex grades iscomparable to that observed
with highly alloyedaustenitic grades like 20Cb-3 (UNS N08020)and
the 6% molybdenum superaustenitics likeAL-6XN (UNS N08367).
-
Chapter 5: Corrosion of Duplex Stainless Steel Weldments /
103
Table 2 Comparison of mechanical properties of commonly used
stainless steels in the annealedcondition
Tensile strength
UNS No. Common designation MPa ksi MPa ksi Elongation, %
Austenitic grades
S30400 Type 304 515 75 205 30 40 88N08020 20Cb-3 585 85 275 40
30 95Duplex grades
S31803 2205 620 90 450 65 25 30.5 HRC(b)S32750 2507 800 116 550
80 15 32 HRC(b)Ferritic grades
S40900 Type 409 415 60 205 30 22(c) 80S44625 E-Brite 26-1 450 65
275 40 22(c) 90(a) At 0.2% offset. (b) Typical values. (c) 20%
elongation for thicknesses of 1.3 mm (0.050 in.) or less. Source :
Ref 1
Yield strength(a)
Hardness (max), HRB
General Welding Considerations
The performance of duplex stainless steelscan be signicantly
affected by welding. Due tothe importance of maintaining a
balancedmicrostructure and avoiding the formation ofundesirable
metallurgical phases, the weldingparameters and ller metals
employed must beaccurately specied and closely monitored.
Thebalanced microstructure of the base material(that is, equal
proportions of austenite and fer-rite) will be affected by the
welding thermalcycle. If the balance is signicantly altered andthe
two phases are no longer in similar propor-tions, the loss of
material properties can beacute. Because the steels derive
properties fromboth austenitic and ferritic portions of the
struc-ture, many of the single-phase base materialcharacteristics
are also evident in duplex mate-rials. Austenitic stainless steels
have excellentweldability and low-temperature toughness,whereas
their Cl SCC resistance and strengthare comparatively poor.
Ferritic stainless steelshave high resistance to Cl SCC but have
poortoughness, especially in the welded condition.A duplex
microstructure with high ferrite con-tent can therefore have poor
low-temperaturenotch toughness, whereas a structure with
highaustenite content can possess low strength andreduced
resistance to Cl SCC (Ref 3). The highalloy content of duplex
stainless steels also ren-ders them susceptible to formation of
inter-metallic phases from extended exposure to hightemperatures.
Extensive intermetallic precipita-tion may lead to a loss of
corrosion resistanceand sometimes to a loss of toughness (Ref
4).
Duplex stainless steels weldability is gener-ally good, although
they are not as forgiving asaustenitic stainless steels or as prone
to degrada-tion of properties as fully ferritic stainless
steels.The current commercial grades are low in carbon(less than
0.03 wt%), thereby essentially elimi-nating the risk of
sensitization and intergranularcorrosion from carbide
precipitation. The basematerial and ller metals also have low
sulfur andphosphorus levels (less than 0.03 wt%), which
incombination with the ferritic solidicationreduce the likelihood
of solidication cracking(hot cracking). Hydrogen cracking (cold
crack-ing) resistance is also good due to the high hydro-gen
solubility in the austenite and the high per-centage of austenite
in the matrix. Nevertheless,hydrogen cracking can occur in duplex
alloys,and is discussed later in the section CorrosionBehavior of
Weldments.
Fusion Welding. Nearly all of the arc weld-ing processes that
are employed for other stain-less steels can be used with duplex
alloys,except where the process characteristic is toweld
autogenously, such as electron-beamwelding and laser-beam welding.
In such cir-cumstances a PWHT is nearly always necessaryto restore
the correct phase balance to the weldmetal and remove any
undesirable precipitates.There are few reported differences in
corrosionresistance between welding processes, but thenonmetallic
inclusion distribution would beanticipated to have an effect. In
most instancesof pipe welding where access is from one sideonly,
gas-tungsten arc welding (GTAW) isalmost exclusively employed for
the root pass (Ref 5). This provides a controllable, high-
-
104 / Corrosion of Weldments
Fig. 4 Time-temperature-transformation (TTT) curve for alloy2205
(UNS 31803) showing the effect of alloying ele-
ments on precipitation reactions. These phases can
negativelyaffect the corrosion resistance and the ductility of the
materialand are the most serious threats to the successful
applications ofduplex grades. Source: Ref 4
Table 3 Critical crevice corrosion temperatures
Critical crevice temperature in 10%FeCl3 6H2O; pH = 1;24 h
exposure
UNS No. Common name C F
Duplex grades
S32900 Type 329 5 41S31200 44LN 5 41S31260 DP-3 10 50S32950 7-Mo
PLUS 15 60S31803 2205 17.5 63.5S32250 Ferralium 255 22.5
72.5Austenitic grades
S30400 Type 304
-
Chapter 5: Corrosion of Duplex Stainless Steel Weldments /
105
Fig. 5 Stress-corrosion cracking (SCC) resistance of
selectedduplex stainless steels (S31803, S32304, and S32750)
relative to austenitic stainless steels (S30400, S30403,
S31600,and S31603) as a function of temperature and chloride
concen-tration in neutral O2-bearing solutions (approximately 8
ppm).Test duration was 1000 h. Applied stress was equal to
yieldstrength. Source: Ref 1
the as-welded phase balance and increase aus-tenite content. The
ferrite content of a weld madewith a nickel-enriched consumable
would de-crease signicantly if it underwent a PWHT. Itmay suffer
from slightly reduced weld metalstrength and could also be more
susceptible to phase formation during heat treatment (Ref 6).The
nickel level in the enriched weld metal willbe approximately 2.5 to
3.5% greater than in thebase material (for example, for the
Fe-22Cr-5.5Ni-3Mo-0.15N duplex stainless steel basematerial
containing 5.5% Ni, the ller metal willcomprise 8.0 to 9.0% Ni,
depending on consum-able manufacturer and form).
The higher alloy ller metals are sometimesused for welding a
less alloyed base material (forexample, a duplex stainless steel
ller metal with25% Cr could be used for the root run in a
Fe-22Cr-5.5Ni-3Mo-0.15N base metal). This is usu-ally done to
improve root weld metal corrosionresistance and thereby pass the
qualication testrequirements. In most cases, this does not lead
toloss of mechanical properties; indeed, the morehighly alloyed
ller metal in the case previouslyis likely to have greater
strength.
To avoid all the requirements for weld metalphase balance and
microstructural control nec-essary with duplex ller metals,
nickel-baseconsumables (for example, AWS A 5.14 ERNi-CrMo-3) have
been used. The yield strength,however, is slightly below that of
the morehighly alloyed grades, and the lack of nitrogen
and the presence of niobium in the ller metalmay contribute to
unfavorable metallurgicalreactions and the formation of
intermetallic pre-cipitates and areas of high ferrite content in
theHAZ (Ref 7, 8).
Preheat is generally not recommended forduplex stainless steels,
but may sometimes bespecied in low-nitrogen grades, because
thicksections and low heat input welding processesmay, in
combination, develop highly ferriticHAZs (Ref 9). For the more
highly alloyedduplex stainless steels, a preheat can be
highlydetrimental and reduce corrosion resistance andmechanical
properties.
Postweld heat treatment is not commonlyused except in autogenous
welds or welds witha ller metal composition that exactly matchesthe
base steel. Although not always necessary,particularly if a
nickel-enriched ller metal isused, it is common to PWHT duplex
stainlesssteel welded pipe after longitudinal seam weld-ing. The
PWHT will largely be for the purposeof restoring the correct phase
balance and redis-solving unwanted precipitates. Postweld
heattreatment temperatures of approximately 1050to 1100 C (1920 to
2010 F) are used, depend-ing on grade, followed by the same heat
treat-ment applied to the base material during solu-tion
annealingusually water quenching. Theheat treatments commonly used
for structuralsteels (for example, 550 to 600 C, or 1020 to1110 F)
are totally inappropriate for duplexalloys and should never be
considered.
Corrosion Behavior of Weldments
Corrosion characteristics of duplex stainlesssteel weldments are
complex. The HAZ suffersmore corrosion attack than either the base
metalor the weld metal because of the unbalance inaustenite/ferrite
fractions in the HAZ (Ref 2).Pitting corrosion resistance of
wrought duplexstainless steels is superior to the cast version.Less
austenite is typically present in the caststructure. Thus, the
duplex stainless steel weldconsumables are enriched with nickel
toachieve higher levels of austenite in their as-welded
microstructures, since the weld metal isessentially a cast
material.
The welding heat input affects the pitting cor-rosion resistance
of the duplex stainless steelweldments. As shown in Fig. 6, the
best pittingcorrosion resistance is achieved when the weld-
-
106 / Corrosion of Weldments
Fig. 7 Effect of ferrite-austenite balance on pitting
resistanceof Fe-22Cr-5.5Ni-3.0Mo-0.12N gas tungsten arc stain-
less steel welds. Source: Ref 3Fig. 6 Inuence of heat input on
corrosion of welded S31803
steel in ferric chloride. Source: Ref 3
ing practice involves higher heat inputs (Ref 3).In addition,
cooling rates also affect the pittingcorrosion resistance. The
slower the coolingrate, the better is the pitting corrosion
resistance(Ref 2). Best corrosion resistance and mechani-cal
properties are achieved when approximatelyequal amounts of
austenite and ferrite are pres-ent in both the weld metal and the
HAZ (Ref 2).A balanced austenite/ferrite content can beachieved by
slowing the cooling rate, throughhigh heat input, preheating in
multipass weldingoperations, and controlled interpass
tempera-tures. The interpass temperature is usually keptbetween 150
and 200 C (300 and 400 F).
Alloying elements in duplex stainless steelsplay a key role in
determining the mechanicaland corrosion properties. Due to the
highchromium content, duplex stainless steels haveexcellent
high-temperature oxidation resis-tance. However, they are prone to
carbide pre-cipitation and to phase and chromium nitride(e.g., CrN,
Cr2N) formation (Ref 2). The car-bide precipitation and other
problems related tothe high chromium content can be resolvedthrough
solution annealing and controlled weld-ing practices such as slower
cooling rates andcontrolled-interpass temperatures (Ref 2).Nitrogen
reduces the partitioning of chromiumbetween the austenite and the
ferrite phases. Italso improves the pitting and crevice
corrosionresistance of the duplex stainless steels (Ref 2).Very
high nickel contents (e.g., 10 wt%) induplex stainless steel weld
metal degrades pit-ting corrosion resistance by diluting the
nitro-gen content in the austenite (Ref 2).
Inuence of Ferrite-Austenite Balance onCorrosion Resistance. The
distribution ofaustenite and ferrite in the weld and HAZ isknown to
affect the corrosion properties and themechanical properties of
duplex stainless steels.Figure 7 shows the effect of the
ferrite-austenitebalance on the pitting resistance of a
duplexstainless steel. To achieve a satisfactory balancein
properties, it is essential that both base metaland weld metal be
of the proper composition.For example, without nickel enrichment in
theller rod, welds can be produced with ferritelevels in excess of
80%. Such microstructureshave very poor ductility and inferior
corrosionresistance. For this reason, autogenous welding(without
the addition of ller metal) is not rec-ommended unless postweld
solution annealingis performed, which is not always practical.
Toachieve a balanced weld microstructure, a lowcarbon content
(approximately 0.02%) and theaddition of nitrogen (0.1 to 0.2%)
should bespecied for the base metal. Low carbon helpsto minimize
the effects of sensitization, and thenitrogen slows the
precipitation kinetics associ-ated with the segregation of chromium
andmolybdenum during the welding operation. Ni-trogen also enhances
the reformation of austen-ite in the HAZ and weld metal during
cooling.
Because these duplex alloys have been usedfor many years,
guidelines relating to austenite-ferrite phase distribution are
available. It hasbeen shown that to ensure resistance to Cl
SCC,
-
Chapter 5: Corrosion of Duplex Stainless Steel Weldments /
107
welds should contain at least 25% ferrite. Tomaintain a good
phase balance for corrosionresistance and mechanical properties
(espe-cially ductility and notch toughness) compara-ble to those of
the base metal, the average ferritecontent of the weld should not
exceed 60%.This means using welding techniques that min-imize weld
dilution, especially in the root pass.Conditions that encourage
mixing of the lower-nickel base metal with the weld metal reduce
theoverall nickel content. Weld metal with a lowernickel content
will have a higher ferrite content,with reduced mechanical and
corrosion proper-ties. Once duplex base metal and welding
con-sumables have been selected, it is then neces-sary to select
joint designs and weld parametersthat will produce welding heat
inputs and cool-ing rates so as to produce a favorable balance
ofaustenite and ferrite in the weld and HAZ.
Researchers have shown that the high-ferritemicrostructures that
develop during welding inlean (low-nickel) base metal and weld
metalcompositions can be altered by adjusting weld-ing heat input
and cooling rate. In these cases, ahigher heat input that produces
a slower coolingrate can be used to advantage by allowing moretime
for ferrite to transform to austenite. Thereare, however, some
practical aspects to considerbefore applying higher heat inputs
indiscrimi-nately. For example, as heat input is increased,base
metal dilution increases. As the amount oflower-nickel base metal
in the weld increases,the overall nickel content of the deposit
de-creases. This increases the potential for moreferrite, with a
resultant loss in impact toughness,ductility, and corrosion
resistance. This wouldbe another case for using an enriched ller
metalcontaining more nickel than the base metal.Grain growth and
the formation of embrittlingphases are two other negative effects
of highheat inputs. When there is uncertainty regardingthe effect
that welding conditions will have oncorrosion performance and
mechanical proper-ties, a corrosion test is advisable.
Effect of Welding on Pitting and SCCResistance. The weld is
usually the part of asystem with reduced corrosion resistance
andlow-temperature toughness, and therefore inmany cases it is the
limiting factor in materialapplication. From a corrosion
standpoint, weld-ing primarily affects pitting corrosion and
ClSCC.
Pitting corrosion resistance can be affectedby many features of
the welding operation,including:
Localized segregation of alloying elements tothe different
constituent phases in the micro-structure, producing areas lean in
molybde-num and chromium
Incorrect ferrite/austenite phase balance Formation of nitrides
or intermetallic phases Loss of nitrogen from the root pass
Presence of an oxidized surface on the under-
side of the root beadThe extent to which the reduction of
corrosionresistance occurs depends on which of thesefactors are
active and to what degree. Partition-ing of alloying elements
between the austeniteand ferrite occurs in the weld metal,
withchromium, molybdenum, and silicon partition-ing to the ferrite
and carbon, nickel, and nitro-gen to the austenite (Ref 10, 11).
The effect isnot so apparent in as-deposited weld metals, butit
becomes more signicant as a result of reheat-ing a previously
deposited weld pass.
Weld metal and HAZ microstructures withvery high ferrite
contents are also less resistant topitting attack than are balanced
structures. This islargely because predominantly ferritic
structuresare more prone to chromium nitride precipita-tion, which
locally denudes the chromium con-centration and lowers resistance
to pitting attack.
Nitrogen loss in the root pass may reduce weldmetal corrosion
resistance. Up to 20% loss ofnitrogen has been reported for GTA
welds (Ref12), and nitrogen-bearing backing gases havebeen explored
and used in limited applications.
Cleanliness of the root side purge gas mayalso affect pitting
resistance. Figure 8 shows theeffect of reducing oxygen content in
an other-wise pure argon purge gas and its benecialeffect on
pitting resistance. Also shown is theapparent benet of using a
reducing gas (NH10),which would signicantly reduce the tendency
Fig. 8 Plot of pitting temperature versus oxygen content
ofbacking gas for Fe-22Cr-5.5Ni-3Mo-0.15N and Fe-
23Cr-4Ni-0.1N duplex stainless steels tested in 3% NaCl and0.1%
NaCl solutions, respectively, both at anodic potential of+300 mV.
Source: Ref 13
-
108 / Corrosion of Weldments
Fig. 9 Pitting corrosion resistance of base metal relative to
weld metal placed in 6 wt % FeCl3 solution for 24 h duration per
ASTM648 (method A). Source: Ref 14
for oxide formation and leave the underbeadappearance very shiny
(Ref 13).
The net effect on pitting corrosion resistancemay be observed by
applying the ASTM G 48pitting corrosion test to welds and base
materialwith the same PREN value, then assessing thereduction in
critical pitting temperature (that is,the temperature at which
pitting in the ferricchloride solution is rst observed). The
differ-ence is approximately 20 C (35 F), as reportedin Fig. 9,
thereby quantifying the effect ofreduced weld metal properties.
Figure 9 alsoshows that the use of a superduplex stainlesssteel
ller metal with a PREN value of about 40on an
Fe-22Cr-5.5Ni-3Mo-0.15N parent steel(which typically has a PREN
value of about 33to 35) will improve the weld metal pitting
cor-rosion resistance, as assessed by the ASTM G48 test, to
approximately match that of the basematerial.
Resistance to Cl SCC does not appear to beaffected signicantly
by welding per se (Ref15). Nevertheless, welds are likely regions
ofattack for Cl SCC due to the presence of highstresses and the
structural inhomogeneity pres-ent at the weld. If localized pitting
is a necessaryprecursor for Cl SCC, the effects describedabove will
also ultimately affect Cl SCC resis-tance. See also the discussion
on SCC of alloy2205 in the section Corrosion Behavior ofAlloy 2205
Weldments.
Hydrogen-Induced Cracking. Duplexstainless steels can suffer
from weld metalhydrogen cracking, but HAZ cracking has not
been reported in practice and is consideredhighly unlikely to
develop. Hydrogen crackingfrom welding and in-service hydrogen
pickuphas been observed (Ref 16). The duplexmicrostructure provides
a combination of a fer-ritic matrix, where hydrogen diffusion can
befairly rapid, with intergranular and intragranularaustenite,
where the hydrogen diffusion is sig-nicantly slower, thereby acting
as a barrier tohydrogen diffusion. The net effect appears to be
that hydrogen can be trapped within fer-rite grains by the
surrounding austenite, particu-larly where it decorates the prior
ferrite grainboundaries. Due to these characteristics,
low-temperature hydrogen release treatments are noteffective, and
the hydrogen is likely to remain inthe structure for a long period
(Ref 16). Whethercracking actually develops will depend upon
anumber of factors, including the total amount oftrapped hydrogen,
the applied strain, and theamount of ferrite and austenite in the
structure.Weld metal hydrogen content from coveredelectrodes can be
relatively high, and levels up to25 ppm have been reported (Ref
17).
The problem of weld metal hydrogen crack-ing in practice must
not be overstated. Thereported incidences of hydrogen cracking
induplex stainless steels have been restricted tocases in which the
alloy has been heavily coldworked or weld metals have seen high
levels ofrestraint or possessed very high ferrite contentsin
combination with very high hydrogen levelsas a result of poor
control of covered electrodesor the use of hydrogen-containing
shielding gas.
-
Chapter 5: Corrosion of Duplex Stainless Steel Weldments /
109
Table 5 Chemical compositions of alloy 2205 specimens tested and
ller metals used in Ref 19Element, %
Specimen size and conguration C Si Mn P S Cr Ni Mo Cu N
Parent metals
48.1 mm (1.89 in.) OD, 3.8 mm(0.149 in.) wall tube
0.015 0.37 1.54 0.024 0.003 21.84 5.63 2.95 0.09 0.15
88.9 mm (3.5 in.) OD, 3.6 mm(0.142 in.) wall tube
0.017 0.28 1.51 0.025 0.003 21.90 5.17 2.97 0.09 0.15
110 mm (4.3 in.) OD, 8 mm (0.31in.) wall tube
0.027 0.34 1.57 0.027 0.003 21.96 5.62 2.98 0.09 0.13
213 mm (8.4 in.) OD, 18 mm (0.7in.) wall tube
0.017 0.28 1.50 0.026 0.003 21.85 5.77 2.98 0.10 0.15
20 mm (3/4 in.) plate 0.019 0.39 1.80 0.032 0.003 22.62 5.81
2.84 . . . 0.13Filler metals1.2 mm (0.047 in.) diam wire 0.011 0.48
1.61 0.016 0.003 22.50 8.00 2.95 0.07 0.131.6 mm (0.063 in.) diam
rod 0.011 0.48 1.61 0.016 0.003 22.50 8.00 2.95 0.07 0.133.2 mm
(0.125 in.) diam wire 0.011 0.48 1.61 0.016 0.003 22.50 8.00 2.95
0.07 0.133.25 mm (0.127 in.) diam covered
electrode0.020 1.01 0.82 0.024 0.011 23.1 10.4 3.06 . . .
0.13
4.0 mm (0.16 in.) diam coveredelectrode
0.016 0.94 0.78 0.015 0.011 23.0 10.5 3.13 . . . 0.11
Indeed, other studies have shown how resistantduplex stainless
steel weld metals are to hydro-gen cracking, even with consumables
intention-ally humidied (Ref 18), and that hydrogen-containing
backing gases can be employedwithout producing cracking. There is
no doubtan effect of hydrogen on the ductility of duplexstainless
steel, and to avoid fabrication-relatedcracking problems,
high-hydrogen-potentialwelding processes, such as SMAW, should
becontrolled by careful storage and use of elec-trodes, and by
ensuring that the weld metalphase balance is within acceptable
limits.
Corrosion Behavior of Alloy 2205Weldments
The inuence of different welding conditionson various material
properties of alloy 2205(UNS S31803, Fe-22Cr-5.5Ni-3.0Mo-0.15N)has
been studied (Ref 19). Chemical composi-tions of test materials are
given in Table 5, andthe results of the investigation are detailed
in thefollowing sections.
Intergranular Corrosion. Despite the useof very high arc
energies (0.5 to 6 kJ/mm, or 13to 150 kJ/in.) in combination with
multipasswelding, the Strauss test (ASTM A 262, prac-tice E) failed
to uncover any signs of sensitiza-tion after bending through 180.
The results ofHuey tests (ASTM A 262, practice C) on sub-
merged-arc welds showed that the corrosionrate increased
slightly with arc energy in thestudied range of 0.5 to 6.0 kJ/mm
(13 to 150kJ/in.). For comparison, the corrosion rate forparent
metal typically varies between 0.15 and1.0 mm/yr (6 and 40
mils/yr), depending on sur-face nish and heat treatment cycle.
Similar results were obtained in Huey tests ofspecimens from
bead-on-tube welds producedby GTAW welding. In this case, the
corrosionrate had a tendency to increase slightly with arcenergy up
to 3 kJ/mm (75 kJ/in.).
Pitting tests were conducted in 10% ferricchloride (Fe Cl3) at
25 and 30 C (75 and 85 F) inaccordance with ASTM G 48. Results of
tests onsubmerged-arc test welds did not indicate anysignicant
change in pitting resistance when thearc energy was increased from
1.5 to 6 kJ/mm (38to 150 kJ/in.). Pitting occurred along the
bound-ary between two adjacent weld beads. Attackwas caused by slag
entrapment in the weld; there-fore, removal of slag is
important.
Gas tungsten arc weld test specimens (arcenergies from 0.5 to 3
kJ/mm, or 13 to 75 kJ/in.)showed a marked improvement in pitting
resis-tance with increasing arc energy. In order forduplicate
specimens to pass the FeCl3 test at 30C (85 F), 3 kJ/mm (75 kJ/in.)
of arc energywas required. At 25 C (75 F), at least 2 kJ/mm(50
kJ/in.) was required to achieve immunity.Welds made autogenously
(no nickel enrich-ment) were somewhat inferior; improvementswere
achieved by using higher arc energies.
-
110 / Corrosion of Weldments
Fig. 12 Preferential attack of the continuous austenite phasein
an autogenous gas tungsten arc weld in Ferralium
alloy 255. Crevice corrosion test was performed in synthetic
sea-water according to ASTM D 1141 at 100 C (212 F). Etched with50%
HNO3. 100
Fig. 11 Preferential corrosion of the ferrite phase in the
weldmetal of Ferralium alloy 255 gas tungsten arc welds
in 10% FeCl3 at room temperature. Base metal was 3.2 mm (1/8
in.) thick.
For comparison with a different alloy, Fig. 10shows the effect
of heat input on the corrosionresistance of Ferralium alloy 255
(UNS S32550,Fe-25.5Cr-5.5Ni-3.0Mo-0.17N) welds madeautogenously and
tested on FeCl3 at 15 C (60F). Preferential corrosion of the
ferrite phase isshown in Fig. 11. In a different test,
Ferralium
alloy 255 was welded autogenously and testedin a neutral
chloride solution according toASTM D 1141 at 60 to 100 C (140 to
212 F).In this case, preferential attack of the austenitephase was
observed. An example is shown inFig. 12. Similar results would be
expected foralloy 2205.
A study of the alloy 2205 weld microstructures(Ref 19) revealed
why high arc energies werefound to be benecial to pitting
resistance. Manyinvestigations have indicated that the presence
ofchromium nitrides in the ferrite phase lowers theresistance to
pitting of the weld metal and theHAZ in duplex stainless steels. In
this study, bothweld metal and HAZ produced by low arc ener-gies
contained an appreciable amount ofchromium nitride (Cr2N) (Fig.
13). The nitrideprecipitation vanished when an arc energy of 3
kJ/mm (75 kJ/in.) was used (Fig. 14).
The results of FeCl3 tests on submerged-arcwelds showed that all
top weld surfaces passedthe test at 30 C (85 F) without pitting
attack,irrespective of arc energy in the range of 2 to 6kJ/mm (50
to 150 kJ/in.). Surprisingly, the weldmetal on the root side, which
was the rst to bedeposited, did not pass the same test
temperature.
The deteriorating effect of high arc energieson the pitting
resistance of the weld metal on theroot side was unexpected.
Potentiostatic testscarried out in 3% sodium chloride (NaCl) at
400mV versus saturated calomel electrode (SCE)conrmed these ndings.
Microexamination ofthe entire joint disclosed the presence
ofextremely ne austenite precipitates, particu-larly in the second
weld bead (Fig. 15) but also
Fig. 10 Effect of welding heat input on the corrosion
resis-tance of autogenous gas tungsten arc welds in Fer-
ralium alloy 255 in 10% FeCl3 at 10 C (40 F). The base metalwas
25 mm (1 in.) thick. Source: Ref 20
-
Chapter 5: Corrosion of Duplex Stainless Steel Weldments /
111
Fig. 14 Microstructure of bead-on-tube weld made by auto-genous
GTAW with an arc energy of 3 kJ/mm (76
kJ/in.). Virtually no chromium nitrides are present, which
resultsin adequate pitting resistance. 200. Source: Ref 19
Fig. 15 Microstructure of the second weld bead of a
sub-merged-arc weld joint in 200 mm (3/4 in.) duplex
stainless steel plate. The extremely ne austenite precipitate
wasformed as a result of reheating from the subsequent weld
pass,which used an arc energy of 6 kJ/mm (150 kJ/in.). 1000.
Source:Ref 19
Fig. 13 Microstructure of bead-on-tube weld made by auto-genous
GTAW with an arc energy of 0.5 kJ/mm (13
kJ/in.). Note the abundance of chromium nitrides in the
ferritephase. See also Fig. 14. 200. Source: Ref 19
in the rst or root side bead. The higher the arcenergy, the more
austenite of this kind was pres-ent in the rst two weld beads.
Thus, nitridesgive rise to negative effects on the pitting
resis-tance, as do ne austenite precipitates that werepresumably
reformed at as low a temperature asapproximately 800 C (1470
F).
Therefore, the resistance of alloy 2205 to pit-ting corrosion is
dependent on several factors.First, Cr2N precipitation in the
coarse ferritegrains upon rapid cooling from temperaturesabove
approximately 1200 C (2190 F) causesthe most severe impairment to
pitting resis-tance. This statement is supported by a greatnumber
of FeCl3 tests as well as by potentiosta-tic pitting tests.
Generally, it seems difcult toavoid Cr2N precipitation in welded
joints com-pletely, particularly in the HAZ, the structure ofwhich
can be controlled only by the weld ther-mal cycle. From this point
of view, it appearsadvisable to employ as high an arc energy
aspractical in each weld pass. In this way, thecooling rate will be
slower (but not slow enough
-
112 / Corrosion of Weldments
Table 6 Corrosion resistance of Ferralium alloy 255 weldments
using various nickel-base alloyllers and weld techniques3.2 mm
(0.125 in.) plates tested in 10% FeCl3 for 120 h
Critical pitting temperature
Gas tungsten arc
Filler metal C F C F C F
Hastelloy alloy G-3 3035 8595(a) 30 85(a) 3035 8595(b)IN-112 30
85(a) . . . . . . 3540 95105(b)Hastelloy alloy C-276 . . . . . . .
. . . . . 2530 7585(a)Hastelloy alloy C-22 30 85(a) . . . . . .
35-40 95105(a)(a) Haz. (b) HAZ plus weld metal
Submerged arcGas metal arc
to encounter 475 C, or 885 F, embrittlement),and the reformation
of austenite will clearlydominate over the precipitation of
Cr2N.
In addition, if there were no restriction onmaximum interpass
temperature, the heat pro-duced by previous weld passes could be
used todecrease the cooling rate further in the criticaltemperature
range above approximately 1000C (1830 F). Preliminary tests with
preheatedworkpieces have shown the signicance of tem-perature in
suppressing Cr2N precipitation. Cur-rently, the maximum recommended
interpasstemperature for alloy 2205 is 150 C (300 F).This
temperature limit does not appear to becritical, and it is
suggested that this limit couldbe increased to 300 C (570 F). The
maximumrecommended interpass temperature for Ferral-ium alloy 255
is 200 C (390 F). Excessivegrain growth as a result of too much
heat inputmust also be considered to avoid loss of ductil-ity and
impact toughness.
Second, the ne austenite precipitates foundin the reheated
ferrite when high arc energiesand multipass welding were combined
are com-monly referred to as 2 the literature. The harm-ful inuence
of 2 on the pitting resistance hasbeen noted with isothermally aged
specimens,but as far as is known, it has never beenobserved in
connection with welding. It is felthowever, that 2 is less
detrimental to pittingthan Cr2N. Moreover, 2 formation is
believedto be benecial to mechanical properties, suchas impact
strength and ductility.
A third factor that lowers pitting resistance isoxide scale.
Where possible, all surface oxidesshould be removed by mechanical
means or,preferably, by pickling. Root surfaces (in pipe),however,
are generally inaccessible, and pittingresistance must rely on the
protection from thebacking gas during GTAW. It is therefore
advis-
able to follow the current recommendation forstainless steels,
which is a maximum of 25 ppmoxygen in the root backing gas.
Stress-Corrosion Cracking. The SCC resis-tance of alloy 2205 in
aerated, concentratedchloride solutions is very good. The effect
ofwelding on the SCC resistance is negligiblefrom a practical point
of view. The thresholdstress for various welds, as well as for
unweldedparent metal in the calcium chloride (CaCl2)test, is as
high as 90% of the tensile strength atthe testing temperature. This
is far above allconceivable design limits.
Also, in environments containing both hydro-gen sulde (H2S) and
chlorides, the resistanceof welds is almost as high as for the
parentmetal. In this type of environment, however, itis important
to avoid too high a ferrite content inweld metal and the HAZ. For
normal welding ofjoints, the resulting ferrite- contents should
notcause any problems. For weld repair situations,however, care
should be taken so that extremelyhigh ferrite contents (>75%)
are avoided. Topreserve the high degree of resistance to SCC,the
ferrite content should not be less than 25%.
Another reason to avoid coarse weld micro-structures (generated
by excessive weldingheat) is the resultant nonuniform plastic
ow,which can locally increase stresses and inducepreferential
corrosion and cracking effects.
Use of High-Alloy Filler Metals. In criticalpitting or crevice
corrosion applications, the pit-ting resistance of the weld metal
can be en-hanced by the use of high Ni-Cr-Mo alloy llermetals. The
corrosion resistance of such weld-ments in Ferralium alloy 255 is
shown in Table6. For the same weld technique, it can be seenthat
using high-alloy llers does improve corro-sion resistance. If
high-alloy llers are used, theweld metal will have better corrosion
resistance
-
Chapter 5: Corrosion of Duplex Stainless Steel Weldments /
113
than the HAZ and the fusion line. Therefore,again, proper
selection of welding techniquecan improve the corrosion resistance
of theweldments.
ACKNOWLEDGMENTS
Portions of this chapter were adapted from:
D.N. Noble, Selection of Wrought DuplexStainless Steels,
Welding, Brazing, and Sol-dering, Vol 6, ASM Handbook, ASM
Interna-tional, 1993, p 471481
K. F. Krysiak, et al., Corrosion of Weld-ments, Corrosion, Vol
13, Metals Handbook,9th ed., ASM International, 1987, p 344368
REFERENCES
1. Wrought Duplex Stainless Steels, AlloyDigest Source Book:
Stainless Steels, J.R.Davis, Ed., ASM International, 2000, p
317356
2. D.K. Aldun, K.M. Makhamreh, and D.J.Morrison, Corrosion
Resistance of Second-Generation Duplex StainlessSteel Weldments in
NACE Solution,International Trends in Welding Scienceand
Technology, Proc. 3rd Int. Conf.Trends in Welding Research, S.A.
Davidand J.M. Vitek, Ed., ASM International,1992, p 661666
3. T.G. Gooch, Corrosion Resistance ofWelds in Duplex Stainless
Steels, DuplexStainless Steels 91, Vol 1, J. Charles andS.
Bernhardsson, Ed., Les ditions dePhysique, Oct 1991, p 325346
4. J. Charles, Super Duplex Stainless Steels:Structure and
Properties, Duplex Stain-less Steels 91, Vol 1, J. Charles and
S.Bernhardsson, Ed., Les ditions dePhysique, Oct 1991, p 151168
5. J. Harston, E. Hutchins, and S. Sweeney,The Development and
Construction ofDuplex Stainless Steel Pipelines for UseOffshore in
the Southern North Sea, Proc.3rd Int. Conf. Welding and
Performanceof Pipelines, The Welding Institute, 1986
6. D.J. Kotecki, Heat Treatment of DuplexStainless Steel Weld
Metals, Weld. J.,Nov 1989, p 431s441s
7. L. Odegarg and S.-A. Fager, Choice ofRight Filler Metal for
Joining SandvikSAF 2507 to a Super Austenitic Stainless
6 Mo Steel, ASM Materials Week 87,ASM International, Oct 1987, p
441449
8. R.N. Gunn, The Inuence of Compositionand Microstructure on
the CorrosionBehaviour of Commercial Duplex Alloys,Recent
Developments in the Joining ofStainless Steels and High Alloys,
EdisonWelding Institute, Columbus, Ohio, Oct1992
9. How to Weld 2205, Avesta WeldingTrade Literature
10. T. Ogawa and T. Koseki, Effect of Com-position Proles on
Metallurgy and Cor-rosion Behaviour of Duplex StainlessSteel Weld
Metals, Weld. J., Vol 88 (No.5), May 1989, p 181s191s
11. R.A. Walker and T.G. Gooch, PittingResistance of Weld Metal
for 22Cr-5NiFerritic-Austenitic Stainless Steel, Br.Corros. J., Vol
26 (No. 1), 1991, p 5159
12. T.G. Gooch, Weldability of Duplex Fer-ritic-Austenitic
Stainless Steels, DuplexStainless Steels, R.A. Lula, Ed.,
AmericanSociety for Metals, 1983, p 573602
13. L. Odegard and S.-A. Fager, The RootSide Pitting Corrosion
Resistance ofStainless Steel Welds, Sandvik SteelWeld. Rep., No. 1,
1990
14. R. Dolling, V. Neubert, and P. Knoll, TheCorrosion Behaviour
of Super DuplexSteel Cast Alloys with a PREN>41,Duplex Stainless
Steels Conf. Proc., Vol2, Les ditions de Physique, Les UlisCedex,
France, Oct 1991, p 13411351
15. T.G. Gooch, Welding in the World, Vol24 (No. 7/8), 1986, p
148167
16. P. Sentance, The Brae Field, Stainl. SteelEur., Vol 4 (No.
20), Sept 1992, p 3841
17. C.D. Lundin, K. Kikuchi, and K. K. Kahn,Measurement of
Diffusible HydrogenContent and Hydrogen Effects on theCracking
Potential of Duplex StainlessSteel Weldments, Phase 1 Report,
TheWelding Research Council, March 1991
18. R.A. Walker and T.G. Gooch, HydrogenCracking of Welds in
Duplex StainlessSteel, Corrosion, Vol 47 (No. 8), Sept1991, p
10531063
19. B. Lundquist, P. Norberg, and K. Olsson,Inuence of Different
Welding Condi-tions on Mechanical Properties and Cor-rosion
Resistance of Sandvik SAF 2205(UNS S31803), Paper 10, presented
atthe Duplex Stainless Steels 86 Confer-ence, the Hague,
Netherlands, Oct 1986
-
114 / Corrosion of Weldments
20. N. Sridhar, L.H. Flasche, and J. Kolts,Effect of Welding
Parameters on Local-ized Corrosion of a Duplex Stainless
Steel,Mater. Perform., Dec 1984, p 52 55
SELECTED REFERENCES
R. Bradshaw, R.A. Cottis, and M.J. Scho-eld, Stress Corrosion
Cracking of DuplexStainless Steel Weldments in Sour Condi-tions,
Mater. Perform., April 1996, p 6570
R.M. Davison and J.D. Redmond, PracticalGuide to Using Duplex
Stainless Steels,Mater. Perform., Jan 1990, p 5762
B. Gissler and B. Bouldin, Corrosion Perfor-mance of Welded
Super Duplex StainlessSteel with Multiple Re-weld/Re-work
WeldRepairs in Aqueous H2S and FeCl3 Environ-ments, Paper No.
04143, presented at Corro-sion 2004, NACE International
H. Inoue, T. Ogawa, and R. Matsuhashi, Cor-rosion Behavior of
Duplex Stainless SteelWelds in Highly Concentrated Sulfuric
Acid,International Trends in Welding Science andTechnology, Proc.
3rd Int. Conf. Trends inWelding Research, S.A. David and J.M.Vitek,
Ed., ASM International, 1992, p667671