welding duplex

8
Duplex www.stainless-steel-world.net STAINLESS STEEL WORLD DECEMBER 2007 53 DUPLEX Abstract Duplex stainless steels have an ex- tensive, successful, track record in a multitude of corrosive and erosive environments up to 315°C (600°F), while providing high immunity to stress corrosion cracking (SCC). Althoughduplex stainless steels are, in many cases, superior in corrosion resistance and strength compared to 304 and 316 austentic stainless steels, many fabricators continue to have difficulties creating welding procedures that yield repeatable weldments with optimum proper- ties. This paper offers practical weld- ing guidelines to new fabricators who want to archieve high quality, robust stainless steel weldments supplementing API 938-C,. “Use of Duplex Steels in Oil Refining Industry”. Discussion includes the impor- tance of balancing fer- rite to austenite, reducing formation of deleterious intermetallic and nonmetallic phases, measuring fer- rite contents, and suggested welding parameters. Introduction For many engineering applications in the petroleum and refining indus- try, duplex stainless steels (DSS) are the preferred material, combining characteristics of both ferritic and austenitic stainless steel (SS) when welded correctly. When welded in- correctly, the potential to form detri- mental intermetallic phases drastical- ly increases, which could lead to a catastrophic failure. When compar- ing DSS to SS, DSS is more resistant than austenitic SS to stress corrosion cracking (SCC) but not as resistant as ferritic SS; also, DSS toughness is typ- ically superior to that of ferritic SS but not as good as austenitic SS. DSS are two phase alloys based on the iron-chromium-nickel (Fe-Cr-Ni) system. These materials typically comprise approximately equal amounts of body-centered cubic (bcc) ferrite, α-phase and face-cen- tered cubic (fcc) austenite, γ-phase, in their microstructure. It is well documented that maximum corrosion resistance and mechanical Figure 1 Overview - General Duplex Welding Guidelines C o m p o s i t i o n ( a ) , w t % U N S N o . C o m m o n D e s i g n a t i o n C M n S P S i C r N i M o C u W N P R E N ( b ) S u p e r d u p l e x g r a d e s ( P R E N 4 0 ) S32520 UR52N+ 0.03 1.5 0.02 0.035 0.80 24.0-26.0 5.5-8.0 3.0-5.0 0.50-3.00 0.20-0.35 41 S32750 2507 0.03 1.2 0.02 0.035 1.00 24.0-26.0 6.0-8.0 3.0-5.0 0.5 0.24-0.32 41 S32760 Zeron 100 0.03 1.0 0.01 0.03 1.00 24.0-26.0 6.0-8.0 3.0-4.0 0.5-1.0 0.5-1.0 0.30 40 S32906 Safurex 0.03 0.8-1.5 0.03 0.03 0.50 28.0-30.0 5.8-7.5 1.50-2.60 0.80 0.30-0.40 41 S39274 DP3W 0.03 1.0 0.02 0.03 0.80 24.0-26.0 6.0-8.0 2.50-3.50 0.20-0.80 1.50-2.50 0.24-0.32 42 S39277 AF 918 0.025 0.002 0.025 0.80 24.0-26.0 4.5-6.5 3.0-4.0 1.2-2.20 0.80-1.20 0.23-0.33 41 (a) Single values are maximum (b) PREN = %Cr + 3.3x(%Mo + 0.5x%W) + 16x%N I n t e r m e d i a t e - a l l o y g r a d e s ( P R E N 3 2 –3 9 ) S31200 44LN 0.03 2.0 0.03 0.045 1.00 24.0-26.0 5.5-6.5 1.2-2.0 0.14-2.0 33 S31260 DP3 0.03 1.0 0.03 0.03 0.75 24.0-26.0 5.5-7.5 2.5-3.5 0.20-0.80 0.10-0.50 0.10-0.30 38 S31803 2205 0.03 2.0 0.02 0.03 1.00 21.0-23.0 4.5-6.5 2.5-3.5 0.08-0.20 34 S32205 2205+ 0.03 2.0 0.02 0.03 1.00 22.0-23.0 4.5-6.5 3.0-3.5 0.14-0.20 35-36 S32550 255 0.03 1.5 0.03 0.04 1.00 24.0-27.0 4.5-6.5 2.9-3.9 1.5-2.5 0.10-0.25 38 S32900 10RE51 0.06 1.0 0.03 0.04 0.75 23.0-28.0 2.5-5.0 1.0-2.0 33 S32950 7-Mo Plus 0.03 2.0 0.01 0.035 0.60 26.0-29.0 3.5-5.20 1.0-2.5 0.15-0.35 35 L o w - a l l o y g r a d e s ( P R E N < 3 2 ) S31500 3RE60 0.03 0.2-2.0 0.03 0.03 1.4-2.0 18.0-19.0 4.25-5.25 2.5-3.0 0.05-0.10 28 S32001 19D 0.03 4.0-6.0 0.03 0.04 1.00 19.5-21.5 1.0-3.0 0.60 1.00 0.05-0.17 23.6 S32304 2304 0.03 2.5 0.04 0.04 1.00 21.5-24.5 3.0-5.5 0.05-0.60 0.05-0.06 0.05-0.20 25 S32404 UR50 0.04 2.0 0.01 0.30 1.00 20.5-22.5 5.5-8.5 2.0-3.0 1.0-2.0 0.20 31 Table 1 Composition and PREN of wrought DSS covered by UNS designation 2 . *Stainless Steel World had intended to publish this article in two parts. The first part of which was published in the November 2007 edition of the maga- zine on pp. 43, 45, 47, and 49. Unfortunately this contained mistakes resulting from incorrect processing of the material at Stainless Steel World's offices. We would like to express that these mistakes in no way relect upon the the work of the Fluor Corporation. As a result we would like to rectify the situation by publishing the correct version of the entire paper in this December edition of Stainless Steel World. Duplex stainless steel welding: best practices* Barry Messer, Vasile Oprea, Andrew Wright. Fluor Canada Ltd., Canada

Upload: sergio-mario-herreros

Post on 30-Nov-2015

66 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Welding Duplex

Duplex

www.stainless-steel-world.net S T A I N L E S S S T E E L W O R L D D E C E M B E R 2 0 0 7 53

DU

PL

EX

AbstractDuplex stainless steels have an ex-tensive, successful, track record in amultitude of corrosive and erosiveenvironments up to 315°C (600°F),while providing high immunity tostress corrosion cracking (SCC).Althoughduplex stainless steels are,in many cases, superior in corrosionresistance and strength compared to304 and 316 austentic stainlesssteels, many fabricators continue tohave difficulties creating weldingprocedures that yield repeatableweldments with optimum proper-ties. This paper offers practical weld-ing guidelines to new fabricatorswho want to archieve high quality,robust stainless steel weldmentssupplementing API 938-C,. “Use ofDuplex Steels in Oil Refining

Industry”. Discussionincludes the impor-tance of balancing fer-rite to austenite, reducing formationof deleterious intermetallic andnonmetallic phases, measuring fer-rite contents, and suggested weldingparameters.

IntroductionFor many engineering applicationsin the petroleum and refining indus-try, duplex stainless steels (DSS) arethe preferred material, combiningcharacteristics of both ferritic andaustenitic stainless steel (SS) whenwelded correctly. When welded in-correctly, the potential to form detri-mental intermetallic phases drastical-ly increases, which could lead to acatastrophic failure. When compar-

ing DSS to SS, DSS is more resistantthan austenitic SS to stress corrosioncracking (SCC) but not as resistant asferritic SS; also, DSS toughness is typ-ically superior to that of ferritic SSbut not as good as austenitic SS. DSS are two phase alloys based onthe iron-chromium-nickel (Fe-Cr-Ni)system. These materials typicallycomprise approximately equalamounts of body-centered cubic(bcc) ferrite, α-phase and face-cen-tered cubic (fcc) austenite, γ-phase, in their microstructure. It iswell documented that maximumcorrosion resistance and mechanical

Figure 1 Overview - General Duplex WeldingGuidelines

Composit ion (a) , wt%

UNS

No.

Common

Designation C Mn S P Si Cr Ni Mo Cu W N

PREN

(b)

Superduplex grades (PREN �40)S32520 UR52N+ 0.03 1.5 0.02 0.035 0.80 24.0-26.0 5.5-8.0 3.0-5.0 0.50-3.00 … 0.20-0.35 41S32750 2507 0.03 1.2 0.02 0.035 1.00 24.0-26.0 6.0-8.0 3.0-5.0 0.5 … 0.24-0.32 �41S32760 Zeron 100 0.03 1.0 0.01 0.03 1.00 24.0-26.0 6.0-8.0 3.0-4.0 0.5-1.0 0.5-1.0 0.30 �40S32906 Safurex 0.03 0.8-1.5 0.03 0.03 0.50 28.0-30.0 5.8-7.5 1.50-2.60 0.80 … 0.30-0.40 �41S39274 DP3W 0.03 1.0 0.02 0.03 0.80 24.0-26.0 6.0-8.0 2.50-3.50 0.20-0.80 1.50-2.50 0.24-0.32 42S39277 AF 918 0.025 … 0.002 0.025 0.80 24.0-26.0 4.5-6.5 3.0-4.0 1.2-2.20 0.80-1.20 0.23-0.33 �41(a) Single values are maximum (b) PREN = %Cr + 3.3x(%Mo + 0.5x%W) + 16x%N

Intermediate-al loy grades (PREN 32–39)

S31200 44LN 0.03 2.0 0.03 0.045 1.00 24.0-26.0 5.5-6.5 1.2-2.0 … … 0.14-2.0 33S31260 DP3 0.03 1.0 0.03 0.03 0.75 24.0-26.0 5.5-7.5 2.5-3.5 0.20-0.80 0.10-0.50 0.10-0.30 38S31803 2205 0.03 2.0 0.02 0.03 1.00 21.0-23.0 4.5-6.5 2.5-3.5 … … 0.08-0.20 34S32205 2205+ 0.03 2.0 0.02 0.03 1.00 22.0-23.0 4.5-6.5 3.0-3.5 … … 0.14-0.20 35-36S32550 255 0.03 1.5 0.03 0.04 1.00 24.0-27.0 4.5-6.5 2.9-3.9 1.5-2.5 … 0.10-0.25 38S32900 10RE51 0.06 1.0 0.03 0.04 0.75 23.0-28.0 2.5-5.0 1.0-2.0 … … … 33S32950 7-Mo Plus 0.03 2.0 0.01 0.035 0.60 26.0-29.0 3.5-5.20 1.0-2.5 … … 0.15-0.35 35

Low-al loy grades (PREN <32)

S31500 3RE60 0.03 0.2-2.0 0.03 0.03 1.4-2.0 18.0-19.0 4.25-5.25 2.5-3.0 … … 0.05-0.10 28S32001 19D 0.03 4.0-6.0 0.03 0.04 1.00 19.5-21.5 1.0-3.0 0.60 1.00 … 0.05-0.17 23.6

S32304 2304 0.03 2.5 0.04 0.04 1.00 21.5-24.5 3.0-5.5 0.05-0.60 0.05-0.06 … 0.05-0.20 25S32404 UR50 0.04 2.0 0.01 0.30 1.00 20.5-22.5 5.5-8.5 2.0-3.0 1.0-2.0 … 0.20 31

Table 1 Composition and PREN of wrought DSS covered by UNS designation2.

*Stainless Steel World had intended to publish this article in two parts. The first part of which was published in the November 2007 edition of the maga-zine on pp. 43, 45, 47, and 49. Unfortunately this contained mistakes resulting from incorrect processing of the material at Stainless Steel World's offices.We would like to express that these mistakes in no way relect upon the the work of the Fluor Corporation. As a result we would like to rectify the situationby publishing the correct version of the entire paper in this December edition of Stainless Steel World.

Duplex stainless steel welding:best practices*Barry Messer, Vasile Oprea, Andrew Wright. Fluor Canada Ltd., Canada

120703_Fluor_v4a 04-12-2007 15:50 Pagina 53

Page 2: Welding Duplex

54 S T A I N L E S S S T E E L W O R L D D E C E M B E R 2 0 0 7 www.stainless-steel-world.net

properties throughout a DSS weld-ment are achieved when the phasebalance of ferrite to austenite is50:50. However, achieving a 50:50phase balance of ferrite to austenite(α→ γ) in a weldment has proven tobe difficult due to many variablessuch as metal chemistry, weldingprocesses, and thermal history of thesteel. Experience coupled with test-ing has shown that DSS have opti-mal corrosion resistance and me-chanical properties when 35 to 60%ferrite content is maintainedthroughout the weldment. Figure 1illustrates the factors that contributein achieving the optimal weld prop-erties.

Many fabricators lack sufficient ex-perience controlling heat input thatachieves a balanced microstructurein DSS weldments. Duplex guide-lines (Figure 1) supplement API 938-C1 and suggest parameters for weld-ing procedure specifications (WPS)that will assist welders achieve theoptimum (α→ γ) balance.

MetallurgyAlloying ElementsFor DSS producers there is no diffi-culty in meeting standard specifica-tions of chemical compositions.Individual steel producers have nar-row target compositions withinASTM/ASME specifications to meetdifferent criteria. DSS are sensitiveto variations in composition, partic-ularly of those elements controllingthe phase balance. The relativelybroad chemical limits permit largevariation in properties. There are three basic categories ofDSS, low-alloy, intermediate alloy,and highly alloyed, or superduplexstainless steel (SDSS) grades,

grouped according to their pittingresistance equivalent number(PREN) with nitrogen and areshown in Table 1. The most widelyused alloys are DSS-grade 2205+ andSDSS-grade 2507.

The remarkable corrosion resistanceand mechanical properties of DSS areattributed to the rich alloy content ofchromium, nickel, molybdenum,and nitrogen that form austenite in aferritic matrix. The combination ofhigh chromium and high molybde-num is a cost-efficient way to achievegood chloride pitting and crevicecorrosion resistance because of thereduced amount of nickel comparedto austenitic SS. The superior attrib-utes of DSS are credited to the inter-actions of alloying elements formingcomplex microstructures. The im-portance of alloying elements is ex-plained in Table 2.

Optimum (α → γ) BalanceFerrite content of DSS will indicatewhether proper welding and/or heattreatment techniques result in corro-sion resistance and mechanicalproperties that fulfil engineering re-quirements. The presence of ferritein DSS imparts the superior chloridestress corrosion cracking (CSCC) re-sistance and high strength. An in-crease of ferrite content causes be-haviour similar to a ferritic SS. Whenthe amount of austenite in DSS in-creases, strength will decrease whilecorrosion resistance and susceptibili-ty to CSCC increases. As a conse-quence, ferrite limits should be spec-ified within a reasonable range andbe used as a control measure. When low temperature impactproperties are required, ferrite con-tent must be carefully controlled.

As the ferrite content exceeds ap-proximately 60%, there will be a no-ticeable decrease in the ductile be-haviour and pitting resistance.Sources indicate there may be a neg-ative effect on ductile behaviourwith ferrite levels below 35%, andreduced resistance to SCC due to achange in the solidification modecausing segregation and precipita-tion of intermetallic phases3.Although it is common to see 30-65% ferrite specified for base andweld metal and 30-70% ferrite HAZ,our experience shows a range be-tween 35-60% ferrite provides opti-mal results. Figure 2 is a theoretical diagram thatillustrates how ferrite content affectsDSS materials. The dotted curverepresents the corrosion rate inchloride containing aqueous envi-

Element

Weight

Percentage (wt %) Elemental Role Alloying Characterist ics

Chromium(Cr)

18 to 30% Ferrite former • Increasing Cr will increase corrosion resistance. • The ferrite content increases with increasing Cr; however, too

much Cr will decrease optimal phase balance.Nickel

(Ni)4 to 8% Austenite former • Ni promotes a change in crystal structure from ferrite to

austenite.• Ni delays the formation of intermetallic phases.

Molybdenum(Mo)

Less than 5% Ferrite former • Enhances pitting corrosion resistance.• Increased tendency to form detrimental intermetallic phases if

Mo content is too high.Nitrogen

(N)Minimum of 0.14% Austenite former • N causes austenite to form from ferrite at elevated temperatures,

allowing for restoration of an acceptable balance of austenite toferrite after a rapid thermal cycle in the HAZ after welding.

• Additions of N increase pitting and crevice corrosion resistanceand strength.

• Delays the formation of intermetallic phases.• Offsets the formation of sigma phase in high Cr, high Mo steels.

Table 2 Importance of alloying elements of DSS.

Figure 2 Corrosion rate and impact energy vs.percent ferrite of DSS.

Figure 3: DSS micrograph (200X)4

120703_Fluor_v4a 04-12-2007 12:32 Pagina 54

Page 3: Welding Duplex

Duplex

www.stainless-steel-world.net S T A I N L E S S S T E E L W O R L D D E C E M B E R 2 0 0 7 55

ronments with respect to percentageof ferrite within the material. Thecorrosion rate is greatest below andrelatively moderate above 35% fer-rite. The solid curve represents im-pact energy at ambient tempera-tures with respect to the percentageof ferrite in DSS. Impact energy is atits greatest magnitude at lower fer-ritic levels right through to approxi-mately 60% ferrite, at which point,the impact energies begin to signifi-cantly decrease.

Precipitation MechanismsOptimum phase balance (α → γ) of aDSS is shown in Figure 3. The lightglobules in the dark body are un-etched austenitic grains within theetched ferritic matrix respectively.

DSS alloys solidify primarily as fer-rite at approximately 1425°C(2597°F) and partially transform toaustenite at lower temperatures by asolid state reaction4. If the coolingrate is rapid, very little ferrite willtransform to austenite resulting inan excessive ferrite phase at room

temperature. Consequently, thecooling rate of duplex welds mustbe slow enough to allow the trans-formation of approximately 50% ofthe ferrite to austenite and, at thesame time, fast enough to preventthe formation of intermetallic phas-es and deleterious microstructures.Unwanted phases may occur duringfabrication when welding differingsection sizes or heavy sections withvery low heat input. The high alloy content and thepresence of a ferritic matrix renderDSS susceptible to embrittlementand loss of mechanical properties,particularly toughness, through pro-longed exposure at elevated temper-atures. As cooling proceeds to lowertemperatures in the range of 475-955°C (887-1750°F) for short periodsof time the precipitation of carbides,nitrides and intermetallic phases, allof which can be detrimental, willoccur. The most notable phases arealpha prime (α′), sigma (σ), chi (χ),and Laves (η) phases. For this rea-son, DSS are generally not used attemperatures above 315°C (600°F).

Cooling provided by thework piece itself is themost effective method ofreducing the time thatthe HAZ is in the temper-ature range formation ofthese intermetallic phas-es. The pseudo binaryphase diagram, Figure 4,is a roadmap of the met-allurgical behaviour ofDSS, and may be used toextrapolate the tempera-tures at which precipita-tion reactions and othercharacteristics occur(Table 3).

Heat Affected Zone(HAZ)The HAZ is the area ofthe base metal that hasits microstructure and

properties altered by inducing in-tensive heat into the metal. TheHAZ should have corrosion resist-ance and impact toughness compa-rable to the base material minimumrequirements. DSS and SDSS exhibita narrow-HAZ, in comparison toaustenitic-SS, due to the low heatinput welding processes and thehigh thermal conductivity of thematerial. Typically an austenitic-SSHAZ is in the order of 500 µm inwidth (approximately 20 grains),whereas a DSS HAZ is often as smallas 50 µm in width (2 grains). Forthis reason, it is extremely difficultto measure the narrow-HAZ of DSSin commercial and industrial set-tings. The morphology of a DSSHAZ is more important than esti-mating (α → γ) values. A low heatinput welding process has sufficientheat to promote the transformationof discontinuous ferrite in the HAZ,and will contribute to the fine grainsize responsible for the increase intoughness of the region (Figure 4).Caution is necessary when using toolow a heat input associated withrapid cooling as a narrow and pre-dominantly ferritic HAZ may beproduced. Sufficient micrographsdemonstrating the presence of dis-continuous ferrite in the HAZ maybe required to ensure a robust weld-ing process. Table 2 indicates the effects of N inDSS; furthermore, N additions to

Duplex Stainless Steel

°C °F

Solidif ication range 1445 to 1385 2633 to 2525Scal ing temperature in air 1000 1832Sigma phase formation 700 to 975 1292 to 1787Carb ide precip itation 450 to 800 842 to 1472475C/885F embr ittlement 350 to 525 662 to 977

Figure 4 Pseudo-binary Fe-Cr-Ni phase at 70% Fe section illustrating areas of detrimental phase formation5.

Table 3 Typical precipitation temperatures for DSS.

Figure 4: Light optical micrograph of a 50mm2205 DSS material.

120703_Fluor_v4a 04-12-2007 12:32 Pagina 55

Page 4: Welding Duplex

Duplex

56 S T A I N L E S S S T E E L W O R L D D E C E M B E R 2 0 0 7 www.stainless-steel-world.net

the shielding gas further support for-mation of austenite during coolingso that the weld and HAZ are moreeasily converted back to the optimalaustenitic to ferritic balance. It is dif-ficult to test the toughness in theHAZ by traditional methods sincethe zone is often not more than oneor a few grain sizes wide4.

DSS General WeldingGuidelinesAcceptable welds do not dependsolely on a welder’s ability to weldDSS; it also depends on a range ofvariables such as base and fillermaterial selection, pre/post-weldcleaning, joint preparation and,most importantly, choice of the mostsuitable welding process for a specif-ic job. Historically, fabricators newto welding DSS spend a significantamount of time fine tuning theirWPS to achieve optimal weldments.The following guidelines are intend-ed to supplement API 938-C. Theguidelines, suggest parameters forweld procedures as well as providingknowledge to create welds with ex-cellent properties, with reasonableproduction, and low repair rates.

WorkmanshipThe most important factors in suc-cessfully welding DSS are qualityworkmanship and welder education.Rewards are significant whenwelders are informed and involvedin the details of the weld proceduresince even the best welder can cre-ate marginal welds with an excel-lent WPS. An informed and proac-tive welder can create successful, re-peatable, welds when there is an un-derstanding of the important rolethe variables play in achieving anoptimum (α → γ) balance in DSS.

EngineeringThe key to obtaining well balancedferrite proportions within the base-metal, weld-metal, and HAZ is toperform Welding ProcedureSpecifications (WPS) and ProcedureQualification Records (PQR) that ad-dress DSS welding issues as well asall requirements and codes for weldjoints and applications. Accordingto code requirements, a WPS andPQR must meet only the minimumrequirements specified in the designcode. The welding of DSS demandsthat additional tests be conducted to

ensure that the weld will be suitablefor the intended service and exhibitthe same physical and corrosion re-sistant qualities as the base metal.In particular, heat inputs should bewell documented in the WPS andPQR so that welders may duplicatethe original during production weld-ing. If the appropriate precautionsand controls are not recognized dur-ing the WPS and PQR developmentstage, production welds can beplagued with problems. In additionto the requirements set forth inASME, Section IX, or the appropriatedesign code for the weldment, thereare a handful of additional tests andcontrols that should be applied tothe WPS and PQR development toensure that the welds produced aremechanically sound, corrosion re-sistant, and repeatable under shopor field conditions. The requirements of ASME SectionIX should be addressed in the devel-opment of WPS and PQR as well asthe tests listed in Table 4.

Base Material SelectionBase materials should be deliveredin an acceptable condition since

Test Purpose Additional Details

ASTM A923 Method B or C Determines if any intermetallic phases

or precipitates were formed duringwelding process

Must have predetermined acceptance criteria

Determine if precipitates affect corrosion resistance and mechanicalproperties

Ferrite Readings

(1) Point Count Method Determines % ferrite

WPS and PQR generally uses point count method

Check �–� balance for entire weldment thickness. Optimal ferrite

content between 35-60%

(2)Electromagnetic Measuring Method

Determines % ferrite or FN Used during production welding to verify % ferrite in accordance withWPS and PQR

Sample preparation according to manufacturer’s recommendations

Use for WPS and PQR for comparison with production welds

Device measurements of the weld cap and root should be within 35–60% ferrite content

Hardness survey Check maximum allowable hardness Depends on DSS grade and service environment

Recommended not to exceed HV10 310

Caution: Conversion from Vickers hardness to Rockwell for DSS is notrepresented by ASTM E140. Refer to API-938C, Figure 2, for hardness

conversions.

Impact Testing Toughness measurement Typically associated with minimum design temperature and engineering

recommendations

EN specifies minimum 60J at room temperature

Frequent requirements for parent and weld metal are minimum 45Javerage at -46°C per ASTM A923 Method B. The narrow HAZ

l d t i t t i i l ti

(2)Electromagnetic Measuring Method

Determines % ferrite or FN Used during production welding to verify % ferrite in accordance withWPS and PQR

Sample preparation according to manufacturer’s recommendations

Use for WPS and PQR for comparison with production welds

Device measurements of the weld cap and root should be within 35–60% ferrite content

Hardness survey Check maximum allowable hardness Depends on DSS grade and service environment

Recommended not to exceed HV10 310

Caution: Conversion from Vickers hardness to Rockwell for DSS is notrepresented by ASTM E140. Refer to API-938C, Figure 2, for hardness

conversions.

Impact Testing Toughness measurement Typically associated with minimum design temperature and engineering

recommendations

EN specifies minimum 60J at room temperature

Frequent requirements for parent and weld metal are minimum 45Javerage at -46°C per ASTM A923 Method B. The narrow HAZ

precludes accurate impact measurements in isolation.

Check �–� balance for entire weldment thickness. Optimal ferrite

content between 35-60%

Table 4: Additional details to ASME Section IX in the development of a WPS and PQR.

120703_Fluor_v4a 04-12-2007 12:32 Pagina 56

Page 5: Welding Duplex

58 S T A I N L E S S S T E E L W O R L D D E C E M B E R 2 0 0 7 www.stainless-steel-world.net

corrosion resistance and mechanicalproperties of duplex alloys rely soheavily on the phase balance andabsence of deleterious microstruc-tures. Generally, DSS are producedfrom the mills with a good α→ γbalance that is very close to 50:50;however, it is recommended thattesting be conducted to verify theabsence of harmful intermetalliccompounds or precipitates. ASTMA923 is designed to detect the pres-ence of intermetallic phases in themill product of DSS to the extentthat toughness or corrosion resist-ance is affected significantly6. Itshould be noted that the test meth-ods will not necessarily detect lossesof toughness or corrosion resistanceattributable to other causes6. Goodpractice is to verify %ferrite of thebase metal prior to welding. Dissimilar metal welds (DMW) be-tween DSS and other materials mustbe evaluated case by case. Generally,it is possible to weld all grades ofDSS to DSS, carbon steel (CS), alloysteel, and austenitic SS1. In somecases it is necessary to use butteringtechniques to avoid PWHT of thefinal weld.

Weld Preparation Proper cleaning prior to welding is afoundation for successful DSS welds.Prior to any welding, all weld bevelsshould be examined using liquidpenetrant testing for edge defectsand laminations for all thicknessesgreater than 6 mm (0.24 in), andvisually inspected for nicks, dents,general damage or other surfaceirregularities. It is also important toclean the surrounding weld area,usually 2 inches back from the bev-el, with grinding pads, then wipedwith an approved solvent. Grindingpads, soft packs, wire brushes andothers tools used in austenitic SSfabrication are acceptable for use onDSS. The base metal should also becleaned of rust blooms and embed-ded iron. Additional precautions for DSSshould always be in place to mini-mize iron contamination. Some DSSare susceptible to iron contamina-tion by free iron, whether it be im-pregnated or on the surface. Thecontamination mechanism is a gal-vanic reaction when an electrolyte

is present. Fabricators should havean inspection procedure in place toprevent iron contamination in ac-cordance with ASTM A3807.Measures to protect cleaned surfacesshould be taken as soon as the finalcleaning is completed and should bemaintained during all subsequentfabrication, shipping, inspection,storage, and installation. Finished,cleaned materials should not bestored on the ground or floor, andshould not be permitted to come incontact with asphalt, galvanizedcarbon steel, mercury, zinc, lead,brass, low melting point metals, oralloys that have been in contactwith such materials7. Store DSS ma-terial and equipment on woodskids, pallets or metal surfaces, thatare protected and prevent directcontact with the DSS surface. Keepopenings of pipes, tubes, valves,tanks, pumps, and pressure vessels,etc., capped or sealed at all times ex-cept when in use.

Joint PreparationJoint preparations for DSS are typi-cally common to austenitic SS; how-ever, deviations exceeding 20% ofthe PQR parameters may have sig-nificant detrimental effects on weldconsistency. Experience shows thatDSS can be cut using the plasma-arcprocess, a machine cutter, or grind-ing disc dedicated solely for the useon DSS. If plasma-arc cutting isused, the inside surface must bethoroughly cleaned of all spatter.Sufficient metal should be removedin the bevelling process to eliminateany HAZ that occurred as a result ofthe plasma-arc cutting. Carbon-arcshould not be used for cutting orback gouging.When DSS welds are to be complet-ed from one side, a slightly widergap and a more open angle is prefer-able since duplex filler metals donot penetrate as deeply as austeniticSS filler metals. To achieve a maxi-mum of 50% dilution in the rootpass, a wider gap is needed to allowadditional filler to be deposited. The final surface preparation andconfiguration should be obtained bymachining or grinding. During ma-chining operations, only a cuttingfluid compatible with SS should beused. Any small burrs, nicks, or oth-

er irregularities on the weld bevelshould be repaired, if possible, bylight grinding. Follow necessarycleaning procedures.

Welding Process SelectionProcess selection is usually dictatedmore by the availability of consum-ables, economics, and logistic consid-erations than by end desired proper-ties. As well, it is commonly mistak-en that DSS can be welded similarlyto austenitic SS. For DSS, the narrowwelding parameters and specific fillermetals defined in the PQR must beclosely monitored to maintain a bal-anced microstructure. If the balanceis significantly altered and the twophases are no longer in similar pro-portions, then loss of material prop-erties can be significant. Where ex-ceptional low-temperature toughnessis required, gas-shielded processesmay be specified to produce higherweld metal toughness propertiesthan flux-shielded processes2. Gas tungsten arc welding (GTAW),shielded metal arc welding (SMAW),submerged arc welding (SAW), fluxcored arc welding (FCAW), gas metalarc welding (GMAW) and plasma arcwelding (PAW) are commonly usedwith success for most DSS grades.Although all of the common arcwelding processes are suitable forwelding DSS, generally a weldingprocess is selected based on suitabili-ty for the welding environment(whether it is field or shop welding),size, type, orientation of joints, andthe required balance between speedand quality. Once the process hasbeen selected, it can be optimizedfor the specific grade of DSS beingjoined. Autogenous welding, such as elec-tron-beam welding and laser-beamwelding, is not very suitable for thewelding of DSS since this processcreates welds with very high ferritecontent. In these cases, a solution-anneal PWHT can restore an accept-able weld and HAZ phase balanceand remove undesirable precipitatesprovided ideal cooling rates are fol-lowed.

Shielding and Backing GasPure argon (Ar) shielding and back-ing gases can create weldments withsufficient corrosion resistance.

120703_Fluor_v4a 04-12-2007 12:32 Pagina 58

Page 6: Welding Duplex

Duplex

www.stainless-steel-world.net S T A I N L E S S S T E E L W O R L D D E C E M B E R 2 0 0 7 59

However, N loss is not uncommonto a depth of 0.5 mm from the sur-face of the weld. To correct thephase balance and improve corro-sion resistance of the weld, it is ben-eficial to have additions of 1-2% Nin the Ar shielding and 90% N and10% hydrogen (H) in the backinggas. Nitrogen contents above 2% inthe shielding gas can cause degrada-tion of the tungsten electrode forGTAW processes. The addition of Hto the shielding gas is not recom-mended as it may cause H absorp-tion in the weld.Back purging should be maintainedon the joint until at least 6 mm ofweld metal thickness has been de-posited. The oxygen content of theback purged volume should not ex-ceed 0.25% (2500 ppm)1. Since DSS have relatively highchromium contents and relativelylow thermal expansion, an oxidescale appearing as an oxide tint isproduced during welding that istypically thin and difficult to re-move. The appearance and amountof heat-tint produced during weld-ing can be minimized with low lev-els of oxygen (below 0.25%) in theshielding and backing gases, withminimal moisture in the backinggas, and with limiting contaminantson the surface prior to welding.Hydrogen in the Ar backing gas canadversely affect the appearance ofheat tint, and the base metal’s sur-face finish. Shielding gases suitablefor the various gas shielded process-es are listed in Table 5.

Filler Metal SelectionWelding consumables for DSS are

similar in composition to that of thebase material, but the consumablesdo require nitrogen and higher levelsof nickel to ensure an appropriatephase balance in the deposited weldmetal. A nitrogen addition in fillermetals, 0.14-0.20% N, helps to pre-vent the formation of σ phases.Increased addition of Ni promotes achange in crystal structure from fer-rite to austenite and also delays theformations of intermetallic phases. Aweld metal microstructure from afiller composition exactly matchingthat of the base metal would yieldhigh ferrite content, off balancingthe optimal α→ γ required. It is im-portant that the Cr-content of thedeposited filler metal selected pro-vides a close match of the base metal.DSS and SDSS may be welded with aDSS filler metal that is alloyed withhigher amounts of Ni or, alternative-ly, they could be welded with a fullyaustenitized Ni-alloy filler metal. Welding DSS-2205 with ER 2209filler metal is an effective method ofachieving optimal α→ γ phase bal-ance. The ER 2209 filler metal hasequal quantities of Cr-content aswell as 7 to 9% nickel compared tothe 5.5% nickel of the base metal.Table 6 lists consumable specifica-tions for DSS and SDSS accordingAmerican Welding Society (AWS)and British Standard (BS).

Welding SDSS-2507 using ER 2209filler metal is not advised due to un-der-matching chemistry. Some fabri-cators use AWS ER 2553 for weldingSDSS-2507; however this practice isnot recommended due to under-matching Ni-content and unaccept-

ably high addition of copper (Cu).Increased Cr-content and low Ni-content of the filler metal willproduce welds and HAZ with excessferrite upon welding. The increasedCu additions will produce detrimen-tal Cu precipitates (ε-phase) as Cu hasa low solubility in ferrite (<0.2%).

Welding SDSS-2507 with a 25Cr-9Ni-4Mo filler metal according toEN specification (Table 6), or choos-ing a Ni-alloy filler metal in replace-ment of DSS filler metal is suggested.Fully austenitic Ni-alloy filler metalsprovide improved pitting corrosionresistance as well as immunity fromSCC to the weldment. Successfulwelding has been reported with vari-ous Ni-alloy filler metals such asE/ER NiCrMo-14, E/ER NiCrMo-10,and E/ER NiCrMo-3. It is importantto note that caution is needed whilewelding DSS with Ni-alloy filler met-als that contain niobium (Nb). Forexample, E/ER NiCrMo-3, as theweldment is more susceptible tostress cracking when welded withhigh heat inputs, forming Nb-ni-tride and Nb-carbonitride precipi-tates. A low heat input is recom-mended in this case. Dissimilar metal weld (DMW) fillerselection should preserve the lowcarbon content of DSS to achieve aphase balance that is mechanicallytough and strength and corrosionresistance superior to at least one ofthe dissimilar metals1. DSS filler isgenerally used for DMW joints be-tween DSS and CS or austenitic SS1;however, a case-by-case review isnecessary for process conditions anddesign parameters. For furtheranalysis, refer to NACE CorrosionPaper No. 07568 “Selection ofDissimilar Metal Welds in SevereEnvironments for Today’sPetrochemical Plants.”

Thermal History Preheat and InterpassTemperaturesDSS should be free of moisture priorto welding. Preheating to a maxi-mum of 95°C (203°F) may be appliedto remove moisture; however, weld-ments should be allowed to cool toroom temperature prior to welding.Generally, preheating is not recom-mended immediately prior to weld-

Welding Process Gas TypesGTAW 99.996%Ar, Ar+2%N2, Ar+5%N2

GMAW Ar+1%O2, Ar+30%He+1%O2, Ar+2%CO2, Ar+15%He+2%CO2

FCAW Ar+1% O2, Ar+20%CO2, Ar+2%CO2

PAW 99.996%Ar

Table 5 Shielding gas general recommendations.

Alloy Process AWS Specification BS EN SpecificationDSS GTAW ER 2209 W 22 9 3 N L

GMAW ER 2209 G 22 9 3 N LSMAW E 2209-15/16/17 E 22 9 3 N L B/RFCAW E 2209 t0-1/4, E 2209 t1-1/4 T 22 9 3 N L R/PSAW ER2209+flux S 22 9 3 N L R/P

SDSS GTAW - W 25 9 4 N LGMAW - G 25 9 4 N L SMAW - E 25 9 4 N L B/RFCAW - -SAW - S 25 9 4 N L + flux

Table 6 DSS and SDSS consumable specifications.

120703_Fluor_v4a 04-12-2007 12:32 Pagina 59

Page 7: Welding Duplex

Duplex

www.stainless-steel-world.net S T A I N L E S S S T E E L W O R L D D E C E M B E R 2 0 0 7 61

ing as it may negatively affect cool-ing rates required to achieve optimalphase balance. There may be in-stances, for example thick-wall tothin-wall, where weldments may re-quire a small degree of preheating toprevent too rapid a cooling rate.These instances should be reviewedby a welding engineer on a case bycase basis. The maximum recommended inter-pass temperature should not exceed200°C (392°F) for DSS alloys and150°C (302°F) for SDSS alloysthroughout welding operations2.Excessive interpass temperatures cancause embrittlement and low impactvalues in the root region. Take tem-perature readings on the base metalsurface, directly adjacent to the weldgroove1. Note that wall thicknessesgreater than 25 mm will have highertemperatures deep in the weld thanthe temperature measured on thesurface. In these situations, harmfulphases could form over prolongedtime.

Heat InputHeat input significantly affects thecooling rate that results in the finalmicrostructure. Too low a heat inputwill often result in a weld that is pre-dominantly ferritic and will not havethe same characteristics as the basemetal. Often a moderately low pre-heat (below 75°C (167°F)) combinedwith low heat input can achieve opti-mum phase balance without the for-mation of excessive ferrite.Conversely, high a heat input resultsin slow cooling rates that increasethe risk of the formation of inter-metallic compounds or precipitatesin the weld and HAZ. Choose a heatinput with an appropriate intermedi-ate cooling rate that favors austenitere-formation at high temperaturesand retards sigma formation at lowertemperature1. For DSS, a relativelylow heat input, in the range of 0.5-2.5 kJ/mm and for SDSS, 0.2-1.5KJ/mm has been successfully used.

Post-Weld Heat Treatment(PWHT) and CleaningPWHT of DSS is generally not re-quired but may be necessary whendetrimental amounts of intermetal-lic phases have formed. Typically, afull solution anneal followed by a

water quench should be considered.When correction heat treatmentsare not possible, cut out and re-placement should be considered. Post weld cleaning is required tomaintain the corrosion resistance.Removal of any heat tint helps toachieve maximum heat resistance.Grinding to clean bright metal is ef-fective. Blasting might be effective,but depending on the scale and theblasting medium it may not be aseffective as grinding.

Testing for OptimumProperties At present, experimental methodsare not available that give absolutemeasurements of the ferrite contentin weld metal, either destructivelyor non-destructively. This situationhas led to the development and use,internationally, of the concept of a“ferrite number” (FN). The FN is adescription of the ferrite content ofa weld metal determined using astandardized procedure. FN is ap-proximately equivalent to theweight percentage ferrite content,particularly at low FN values. A number of methods exist to meas-ure the ferrite of a given sample al-though, historically, there is a greatdeal of confusion over whichmethod to apply and how to inter-pret the results. The most commonmethods to measure ferrite contentof DSS are point count method(PCM) and an electromagneticmeasuring device (EMD), both gov-erned by ASTM E562. The PCM, amore time consuming but moreaccurate method, is best to verifythe consistency of balanced α→ γduring WPS and PQR developmentbecause it is a destructive test. The EMD is a preferred method forverifying the austenite to ferriteratio of production welds. The fer-rite content determined by thesemethods is arbitrary and is not nec-essarily the true or absolute ferritecontent throughout the weldmentthickness. The relative percent accu-racy is based on unbiased statisticalestimations described in ASTM E562.

Point Count Method (PCM)The PCM is a systematic, manualcounting procedure for statisticallyestimating the volume fraction of an

identifiable constituent or phase.This method requires a mounted andpolished planar cross section of theweld specimen, a method that can-not be applied to production welds,but can be applied to WPS and PQR.The PCM verifies α→ γ balance inthe PQR test coupons to provide as-surance that the WPS will yield opti-mum phase balance during produc-tion welding. Tests are recommendedfor the following locations:• In the parent metal, one measure-

ment on each side of the weld atmid thickness (total 2).

• In the HAZ on each side of theweld, in the region of the rootpass (total 2). Sufficient micro-graphs may be required to ensurediscontinuous ferrite along thenarrow HAZ.

• In the weld metal, 3 measure-ments near to the vertical centerline of the weld, one in the cap,one in the root, and one at midthickness (total 3).

Electromagnetic MeasuringDevice (EMD)An EMD provides fast and easy read-ings without any significant prepa-ration or the requirement to sectionwelds and verifies that ferrite valuesdo not deviate significantly fromvalues in the WPS and PQR. TheFischer Feritscope® MP30E is an ex-ample of an EMD that measuresboth ferrite numbers (FN) and per-cent ferrite according to the mag-netic induction method. This devicecan measure from 0.1 to 110 FN and0.1 to 80% ferrite8. Production fer-rite tests should be performed at in-tervals of 3 tests every 5 feet ofweld; one in each of the base metaland weld metal. All single pointmeasurements from this methodshould be within the specifiedrange. Surface preparation shouldfollow the manufacturers recom-mendations. The ferrite content ofthe narrow HAZ cannot be accurate-ly measured by this method.

ConclusionsGood corrosion resistance and me-chanical properties of DSS are the re-sult of well crafted WPS/PQR thatdefine heat inputs and cooling ratesto achieve weldments with optimumferrite to austenite balance. The

120703_Fluor_v4a 04-12-2007 12:32 Pagina 61

Page 8: Welding Duplex

Duplex

www.stainless-steel-world.net S T A I N L E S S S T E E L W O R L D D E C E M B E R 2 0 0 7 63

presence of ferrite in DSS impartsthe superior CSCC resistance andhigh strength. On the other hand,austenite in DSS provides the highaqueous corrosion resistance andlow temperature impact toughness. The recommended phase balance ofDSS and SDSS should contain 40-60% ferrite in the base metal and35-60% ferrite in the weld metal.Special consideration must be givento the narrow-HAZ. It is essential tohave a discontinuous fine grain mi-crostructure of ferrite and austenitein the narrow region. DSS shows rather complex precipita-tion behaviour due to the highamount of alloying elements. Theformation of carbides, nitrides andintermetallic phases, which can bedetrimental, will begin to form inshort periods of time as cooling pro-ceeds to lower temperatures in therange of 475-955°C (887-1750°F).For this reason DSS should not beused at temperatures above 300°C(570°F). Optimum DSS welds depend onmultiple factors such as engineeringdesign, material selection, pre/post-weld cleaning, joint preparation andmost importantly, choice of a suit-able welding process. Guidelinessupplement API 938C and providesrecommendations for WPS and PQRdevelopment to achieve an opti-mum ferrite to austenite balanceduring welding.

References1. API Technical Report 938-C. 2005. Use of Duplex Stainless Steels in the Oil Refining Industry:

American Petroleum Institute: First Edition. Washington, D.C.

2. Davis, J.R. Corrosion of Weldments: ASM International® (2006): Materials Park, OH 44073-002.

3. Gooch, T.G., Leonard, A.J., and Gunn, R.N. 2000. Hydrogen cracking of ferritic-austenitic stain-less steel weld metal (DA2_058): Stainless Steel World: Page 354-357.

4. Sieurin, H., and Sandström, R. Austenite Reformation in the heat-affected zone of duplex stainlesssteel 2205: Material Science Engineering A 418 (2006) 250-256.

5. Pohl, M., Storz, O., Glogowski, T. Effect of intermetallic precipitations on the properties of duplexstainless steel. Material Science Engineering 58 (2007) 65-71.

6. ASTM© Standards A 923, Standard Test Methods for Detecting Detrimental Intermetallic Phase inWrought Duplex Austenitic/Ferritic Stainless Steels.

7. ASTM© Standards A 380, Standard Practice for Cleaning, Descaling, and Passivation of StainlessSteel Parts, Equipment, and Systems.

8. BS 6787:1987, ISO 8249:1985. Determination of ferrite number in austenitic weld metal depositedby covered Cr-Ni steel electrodes: British Standard.

Barry Messer isTechnical Director andSenior Fellow forMaterials and WeldingEngineering with FluorCanada Ltd. and a di-rector with theCanadian Welding Bureau. Barry hasover 30 years experience in metallur-gy, welding, and NDE development andmaterial selection. He is regularly in-volved in the analysis and mitigation offabrication and in-service failures forthe chemical, petroleum, power, andmining [email protected]

About the authors

Vasile Oprea is aSenior Metallurgicaland Welding Engineerwith Fluor Canada Ltd.He has over 25 yearsexperience in materialselection, welding, heattreatment, NDE, and failure analysis.

Andrew Wright is ametallurgical Engineerwith Fluor Canada Ltd.He provides weldingand metallurgical sup-port for piping andequipment fabrication.Andrew is currently involved in high al-loy welding issues on internationalpetrochemical projects.

Heat-Resistant Stainless Steel TubesWe offer one of the widest ranges in Europe

Ourgrades:

1.4713TP 409 1.4720TP 405 1.4724TP 430 1.4742TP 446 1.4749TP 446 1.4762

More than

20 years of Competence

·SCHENK·H

EAT-RESISTANT TUB

ES

TP 327 1.4821TP 309 1.4828TP 314 1.4841TP 310s 1.4845TP 330 1.4864Alloy 800 1.4876

TP 32/27 1.4877TP 321 H 1.4878Alloy 600 2.4816Alloy 601 2.4851Alloy 625 2.4856Alloy 825 2.4858

Schenk Stahl GmbHP.O. Box 27 03 38, D-40526 Düsseldorf, Germany

Tel: +49 2131 230-37, Fax: +49 2131 [email protected], www.schenk-stahl.de

120703_Fluor_v4a 04-12-2007 12:32 Pagina 63