ABSTRACT. In this investigation, an at-tempt is made to further the understand-ing of factors influencing the hydrogencontent in duplex stainless steel gas tung-sten arc (GTA) and gas metal arc (GMA)welds as well as to what extent it affectshydrogen-induced cracking susceptibility.

Hydrogen content in GTA and GMAwelds was determined using standard22%Cr duplex and 25%Cr super duplexconsumables. Gas tungsten arc weldswere made using 2.4-mm-diameter con-sumables, argon and argon + 30%heliumshielding gases and heat inputs varyingfrom 0.9 to 2.6 kJ/mm.

Gas metal arc welds were made with1.2-mm-diameter welding wire, argon +2%oxygen and argon + 2%carbon diox-ide shielding gases, and heat inputs rang-ing from 0.8 to 1.3 kJ/mm.

The shielding gas moisture contentwas varied from less than 10 ppm up to1000 ppm. Hydrogen content was ana-lyzed by encapsulating the specimen inan evacuated quartz-glass capsule, heat-ing for 72 h at 400°C and measuring thecontent at ambient temperature using agas chromatograph. An additional hy-drogen determination at 900°C was per-formed also.

Susceptibility to hydrogen crackingwas determined using a three-point bendtest for 24 h, maintaining a 12% strain inspecimens welded under conditionsrated as most critical: high ferrite, highshielding gas moisture (1000 ppm), low

heat input, high hydrogen level and ahigh restraint level.

The results indicated that susceptibil-ity to hydrogen cracking using the GTAor GMA process appears to be limited. Inpractice, maintaining a moisture levelbelow 10 ppm in the shielding gas is ofless importance than the choice of weld-ing parameters. Even a moisture level of1000 ppm in the shielding gas, in com-bination with the correct welding para-meters, will result in a sufficient low hy-drogen content in the weld. Similarly, amoisture level in the shielding gas below10 ppm does not necessarily result in lowhydrogen content in the weld metal. Al-though very high ferrite levels were com-bined with high restrain and high hydro-gen content, none of the GMA and GTAwelds cracked. Susceptibility to hydro-gen cracking is concluded to be limited.

Introduction

Duplex and super duplex stainlesssteels are applied in a variety of indus-tries. It has been demonstrated in prac-

tice that hydrogen cracking in duplexstainless steel weldments may occurwhere hydrogen content and ferrite levelare relatively high and sufficient stress isapplied. However, it remains process andconsumable dependent. In earlier inves-tigations (Refs. 1, 2), authors have re-ported on the sensitivity for hydrogencracking using shielded metal arc weld-ing (SMAW), submerged arc welding(SAW), flux cored arc welding (FCAW)and GTAW consumables. The GTAWprocess, for instance, showed high hy-drogen levels could be found in the weldmetal despite the low ,hydrogen contentof the consumables. Hydrogen crackingwas observed in GTA welds that had over100 FN and 25–30 mL H2/100 g. Processvariables and shielding gas conditionswere indicated to be appropriate for fur-ther investigation.

Among others, Lundin (Ref. 3), Shi-nozaki (Ref. 4) and Ogawa (Ref. 5) re-ported on hydrogen cracking in duplexstainless steel but did not include thesevariables. An attempt to contribute to theunderstanding of factors influencing thehydrogen content in duplex stainlesssteel GTA and GMA weldments, as wellas to what extent it affects the hydrogen-induced cracking susceptibility, was thegoal of this investigation. A specific ques-tion that needs to be answered is whetheran elevated moisture level in the shield-ing gas will increase hydrogen crackingsusceptibility and/or hydrogen level inthe weld metal. In addition, the hydrogendetermination method was evaluated.

Shielding Gas

One of the potential sources for weld

WELDING RESEARCH SUPPLEMENT | 7-s

RE

SE

AR

CH

/DE

VE

LO

PM

EN

T/R

ES

EA

RC

H/D

EV

EL

OP

ME

NT

/RE

SE

AR

CH

/DE

VE

LO

PM

EN

T/R

ES

EA

RC

H/D

EV

EL

OP

ME

NT

How to Control Hydrogen Level in (Super)Duplex Stainless Steel Weldments Using the

GTAW or GMAW Process

BY V. v. d. MEE, H. MEELKER and R. v. d. SCHELDE

Welding parameters seem to be the determining factor for controlling hydrogen levels even when high shielding gas moisture, high ferrite level and high restraint

conditions were induced

KEY WORDS

(Super) Duplex Stainless SteelGTAWGMAWShielding GasHydrogenMoistureCracking

V. v. d. MEE is with The Lincoln Electric Com-pany, Cleveland, Ohio. H. MEELKER is withLincoln Smitweld bv, The Netherlands. R. v. d.SCHELDE is with AGA Gas bv, The Nether-lands.

metal hydrogen is moisture in the shield-ing gas. As most welding procedures pre-scribe limits for moisture level in theshielding gas, many of the shieldinggases for arc welding are supplied with aguaranteed maximum moisture level of10 ppm. In Fig. 1, the relationship ofmoisture content and dew point temper-ature of gases is plotted. The moisturecontent of a gas is usually expressed asdew point temperature. This dew pointtemperature is the temperature wherewater in the gas starts to condensate. Re-sults from an investigation by Gandy andFindlan (Ref. 6) on the influence of mois-ture in the shielding gas process on dif-fusible hydrogen content of GTA-welded, low-alloyed weld metal are alsoplotted in Fig. 1. This figure, however,shows that an increase in shielding gasmoisture from 10 ppm (–60°C dew point) to 1000 ppm (–20°C dew point) willhave a marginal effect on the diffusiblehydrogen content.

For the gas supplier, the specifiedmoisture level is “as delivered” in cylin-ders or in liquid form, which may varysubstantially from the one measured atthe gas cup of the welding torch. Conta-mination, due to leakage at regulatorsand/or fittings, but mainly hoses, is a well-known phenomena as published by Far-ish (Ref. 7) and Brodén (Ref. 8); however,it is not always recognized in practice.

In case the partial pressure of a gas(e.g., water vapor) outside of the gas hoseis higher than on the inside, the gas will,depending of the hose material, diffusethrough the hose wall and contaminatethe shielding gas. Enrichment of theshielding gas in this way was prevented byusing special gas hoses, almost non-per-meable to moisture. As in practice alltypes of hoses can be used, and usually noflushing times are specified, a range up to

1000 ppm H2O in theshielding gas was usedin this investigation.

The AGA shieldinggases with certifiedmoisture content usedfor this investigationare in Table 1.

For purging the gassupply system in be-tween and prior to thevarious tests, dry ni-trogen with a 4-ppmmoisture level wasused. In addition tothe “as delivered”condition of the shielding gas, moisturelevels of 500 and 1000 ppm, controlledand measured during the experiments,were investigated. These levels were cre-ated by bubbling part of the shielding gasthrough a water container. Moisture levelwas measured using a calibrated hy-grometer with a sensor suitable for a–75°/20°C dew point temperature rangeand controlled by adjusting the flow ofdry and moistened gas. The system forcontrolling the moisture level in the gasis illustrated in Fig. 2. During all tests, cli-matological conditions were monitored.

Consumables

In order to produce welds with awider range of ferrite levels, both duplexand super duplex grades were tested. InTable 2, the chemical composition of thebase metal and weld metal is stated. Forthe GTA process, 2.4-mm-diameter25%Cr super duplex (Z100M) and22%Cr duplex (4462) filler metals wereused. For the GMA process, weldingwires of 1.2-mm-diameter 25%Cr superduplex (Z100M) and 22%Cr duplex(4462) were chosen.

The total hydrogen content in the con-sumables and base metal used, as mea-sured using a Ströhlein H2500 analyzer,is stated in Table 3.

Welding Variables

Welds were made under varying con-ditions to evaluate their susceptibility tohydrogen-induced cracking.

Welding parameters for the GTAwelds were 150 A, 15 V, direct currentelectrode negative (DCEN) and 8 L/mingas flow. The heat input was varied by

8-s | JANUARY 1999

RE

SE

AR

CH

/DE

VE

LO

PM

EN

T/R

ES

EA

RC

H/D

EV

EL

OP

ME

NT

/RE

SE

AR

CH

/DE

VE

LO

PM

EN

T/R

ES

EA

RC

H/D

EV

EL

OP

ME

NT

Fig. 1 — Relationship of the shielding gas moisture content, dew pointtemperature and diffusible hydrogen content.

Fig. 2 — System for controlling the moisture level in the shieldinggas.

using different travel speeds. For theGMA welds, welding parameters were180–260 A, 26–31 V, direct current elec-trode positive (DCEP), 38 cm/min weld-ing speed and 15 L/min gas flow. All threelevels of moisture in the shielding gas (assupplied, 500 ppm and 1000 ppm) wereused. The system was purged with dry ni-trogen during intervening periods.

During welding, atmospheric condi-tions were monitored and were withinthe 16–19°C temperature and 40–60%relative humidity range. All welding wasmanual.

For GTA welding, a water-cooledtorch was used with a 11.5-mm-diametergas cup, with provisions for a laminar gasflow. The GMA gun had a 16-mm-diam-eter conic gas cup.

Test Program

In order to be able to study the rela-tionship between process variables andhydrogen content, specimens werewelded according to ISO 3690/lIW II-1155-91 standards with the exceptionthat all welds were made on 22%Cr du-plex base metal. The AWS or ISO stan-dards for ferritic mild steel are not ap-propriate for measuring the hydrogencontent in (super) duplex stainless steelweldments, as can be explained fromphase transformations of duplex stainlesssteels (Ref. 10).

Based on previous experiments by theauthors (Refs. 1, 2), Fekken (Ref. 9) andLundin (Ref. 3), an extraction in a quartzcapsule for 72 h at 400°C was appliedinitially for determination of hydrogencontent in (super) duplex stainless steel.A recent group-sponsored project at TWI(Ref. 11) indicated that significant levelsof hydrogen might remain in the sampleafter encapsulation; therefore, all sam-

ples were analyzed at 900°C after hydro-gen extraction for 72 h at 400°C.

Hydrogen content in GTA and GMAwelds was determined using both duplexand super duplex consumables. Gastungsten arc filler metal of a 2.4-mm di-ameter was used with argon 4.0 andargon + 30% helium shielding gas. Twodifferent welding speeds resulted in heatinputs varying from 0.8 to 2.6 kJ/mm. For1.2-mm-diameter GMA welding wires,argon + 2% oxygen and argon + 2% car-bon dioxide were used as shieldinggases, and welding was performed withheat inputs ranging from 0.8 through 1.3kJ/mm.

The weld metal ferrite content was de-termined for both hydrogen test samplesand the restrained welding specimenusing Magne Gage according toANSI/AWS 4.2-91.

The eight most severe conditions (asindicated in Table 5) were tested forcracking susceptibility using a three-point bend test as shown in Fig. 3. Herea longitudinal bead-on-plate weld wasmade on a 16-mm duplex stainless steelbase plate, which, immediately aftercooling in air, was placed in the bend testunit for 24 h, where it was bent to a 20-deg angle using a 50-mm former diame-ter. This configuration results in approxi-mately 12% strain on the top surface ofthe weld (calculated), which was main-tained for 24 h.

After this load, the specimens wereexamined using the liquid penetranttechnique. When no cracks were identi-fied, a further bend of the specimen to 90deg, which is equivalent to approxi-mately 27% (calculated) strain on the topsurface of the weld, was applied. The testwelds were inspected microscopicallyfor the presence of cracks. Figure 4 illus-

trates a rutile SMA welded specimen bentto 90 deg.

Results

In an earlier program (Ref. 2), wheretest conditions were identical to this in-vestigation, cracking data were pre-sented for shielded metal arc, flux coredarc, gas tungsten arc and submerged arcwelds as shown in Fig. 5. In that investi-

WELDING RESEARCH SUPPLEMENT | 9-s

RE

SE

AR

CH

/DE

VE

LO

PM

EN

T/R

ES

EA

RC

H/D

EV

EL

OP

ME

NT

/RE

SE

AR

CH

/DE

VE

LO

PM

EN

T/R

ES

EA

RC

H/D

EV

EL

OP

ME

NT

Fig. 3 — Schematic bend test setup with po-sition former, support and weld bead.

gation, particular high hydrogen levelswere measured in GTA duplex weldmetal, which in combination with highferrite levels resulted in cracking. A sum-mary of results showing hydrogen crack-ing were indicated by shaded areas. Hy-drogen cracking using Ar + 2%H2 as ashielding gas was the basis for the inves-tigation reported here.

Results of hydrogen measurements forduplex and super duplex stainless steelweldments, using GTA and GMA con-sumables and shielding gas moisture lev-els up to 1000 ppm, are listed in Table 4,along with weld metal ferrite level andapplicable heat input.

In Figures 6 and 7, the effect of mois-ture content in the shielding gas on hy-drogen level in weld metal is illustratedfor GTA welds with duplex and super du-plex consumables using argon 4.0 andargon + 30% helium gas mixtures. Thenumbers plotted are the total of 400°C

and 900°C measurements. It appearsthat, in general, there is no significant dif-ference in weld metal hydrogen levelsdespite an increased moisture level in theshielding gas up to 1000 ppm for most ofthe trials. Only the welding wire (25%Crsuper duplex) used with the low heatinput showed a significant increase in hy-drogen level.

The data in this study showed morevariation for hydrogen measurementsthan mild steel weldments. This is be-lieved to be caused mainly by weldingparameters. The influence of weldingspeed (heat input) on hydrogen level is il-lustrated in Figs. 8 and 9. A longer arctime for the same weld length, which isreflected in a higher heat input, showeda significant reduction in weld metal hy-drogen.

Gas metal arc weldments were madeusing duplex and super duplex consum-ables, Ar + 2%O2 and Ar + 2%CO2

shielding gases and different current lev-els. The moisture content of the shieldinggas was found to affect the weld metalhydrogen level only slightly, as illustratedin Figs. 10 and 11. Increased hydrogenlevels were found to be associated withlower current levels — Figs. 12, 13. Thehydrogen level was found to be a func-tion of heat input (through welding speedand current) for both the GTA and GMAwelds made in this study. The practicalsignificance of these numbers should beinterpreted in comparison to standardwelding conditions being 1.0–2.5 kJ/mmfor 1.2-mm GMA wire and 0.8–2.0kJ/mm for 2.4-mm GTA filler metal.

Susceptibility to cracking was deter-mined for a set of conditions believed tobe most susceptible, which were highferrite, high shielding gas moisture (ap-proximately 1000 ppm), low current(GMAW), high welding speed/low heatinput (GTAW), high deposited hydrogenlevel and an increased restraint level.

10-s | JANUARY 1999

RE

SE

AR

CH

/DE

VE

LO

PM

EN

T/R

ES

EA

RC

H/D

EV

EL

OP

ME

NT

/RE

SE

AR

CH

/DE

VE

LO

PM

EN

T/R

ES

EA

RC

H/D

EV

EL

OP

ME

NT

Fig. 4 — Rutile-coated SMAW specimen bent to 90 deg.

Fig. 5 — Cracking data with several processes from previous investigations(Refs. 1, 2).

Fig. 6 — Effect of moisture content in Ar 4.0 shielding gas on hydrogencontent for GTA welds.

HD

M (

mL/

100

g)hy

drog

en c

onte

nt (

mL/

100

g)

Table 5 shows a summary of these vari-ables. None of the investigated weldsshowed any (micro) cracks, even afterbending the specimen to a 90-deg angle,which is equivalent to approximately27% plastic strain at the top surface in theweld. An overview of the results is illus-trated in Figs. 14 and 15.

Discussion

Comparison tests, performed in severallaboratories over the last years, demon-strated a good relationship between vac-uum hot extraction (VHE) and the encap-

sulation technique in quartz, followed byan extraction for 72 h at 400°C.

A recent publication out of a group-sponsored project at TWI (Ref. 11), how-ever, indicated that not all hydrogen inthe weld metal might evolve using theencapsulation technique. In that publi-cation, an additional hot extractionmethod at 900°C, following the encap-sulation test method, gave an extra vol-ume of hydrogen. The hydrogen releasefrom a sample in a capsule was con-cluded to be hindered by the develop-ment of a positive hydrogen partial pres-sure within the capsule.

Since the encapsulation techniquehas been used in a number of investiga-tions and is the basis for several publica-tions worldwide, this project used the en-capsulation technique, followed by hotextraction at 900°C.

Hydrogen present in the base metaland the welding consumables previouslywas identified as the first potential sourcefor weld metal hydrogen. This hydrogenlevel in consumables may vary from oneheat to another. However, the initial hy-drogen level of the consumable ap-peared to be of less importance. BothGTA and GMA weldments, regardless of

WELDING RESEARCH SUPPLEMENT | 11-s

RE

SE

AR

CH

/DE

VE

LO

PM

EN

T/R

ES

EA

RC

H/D

EV

EL

OP

ME

NT

/RE

SE

AR

CH

/DE

VE

LO

PM

EN

T/R

ES

EA

RC

H/D

EV

EL

OP

ME

NT

Fig. 7 — Effect of moisture content in Ar + 30%He shielding gas onhydrogen content for GTA welds.

Fig. 8 — Effect of heat input (welding speed) on hydrogen content forGTAW with Ar 4.0.

hydr

ogen

con

tent

(m

L/10

0 g)

hydr

ogen

con

tent

(m

L/10

0 g)

12-s | JANUARY 1999

RE

SE

AR

CH

/DE

VE

LO

PM

EN

T/R

ES

EA

RC

H/D

EV

EL

OP

ME

NT

/RE

SE

AR

CH

/DE

VE

LO

PM

EN

T/R

ES

EA

RC

H/D

EV

EL

OP

ME

NT

Fig. 9 — Effect of heat input (welding speed) on hydrogen content forGTAW with Ar + 30%He.

Fig. 10 — Effect of moisture content in Ar + 2%O2 shielding gas on hy-drogen content for GMA welds.

Fig. 11 — Effect of moisture content in Ar + 2%CO2 shielding gas onhydrogen content for GMA welds.

Fig. 12 — Effect of heat input on hydrogen content for GMAW in Ar+ 2%O2.

hydr

ogen

con

tent

(m

L/10

0 g)

hydr

ogen

con

tent

(m

L/10

0 g)

hydr

ogen

con

tent

(m

L/10

0 g)

hydr

ogen

con

tent

(m

L/10

0 g)

type of consumable or shielding gas,tended to go to a 5–10 mL H2/100 g levelwith appropriate welding procedures.

One of the objectives was to determinethe effect of elevated moisture levels in theshielding gas. Today, in most procedures,the maximum allowable moisture contentin shielding gas is 10 ppm.

In real life, however, moisture contentof the gas leaving the gas cup at the timeof welding has measured up to 50–100ppm. Should gas purging not be suffi-cient, or if moisture diffusion through thehoses has occurred, 375 ppm H2O in theshielding gas is not rare, as reported byFarish (Ref. 7) and AGA (Ref. 12), illus-trated in Fig. 16. That is why, in this study,moisture levels in the shielding gas up to1000 ppm were applied. It appearedhowever that the hydrogen content didnot significantly increase with increasedshielding gas moisture.

This investigationalso examined the ef-fect of weldingprocess parameterson weld metal hydro-gen content. It wasfound that parameterselection had a sig-nificant effect on hy-drogen content. Itwas found in this in-vestigation that mini-mum heat inputs of 2kJ/mm for the 2.4-mm GTA filler metaland 1.3 kJ/mm (250A) for the 1.2-mmGMA welding wiresachieved hydrogenlevels not exceeding10 mL/100 g in theweld metal.

WELDING RESEARCH SUPPLEMENT | 13-s

RE

SE

AR

CH

/DE

VE

LO

PM

EN

T/R

ES

EA

RC

H/D

EV

EL

OP

ME

NT

/RE

SE

AR

CH

/DE

VE

LO

PM

EN

T/R

ES

EA

RC

H/D

EV

EL

OP

ME

NT

Fig. 13 — Effect of heat input on hydrogen content for GMAW in Ar +2%CO2.

Fig. 14 — Cracking susceptibility for GMA welds.

Fig. 15 — Cracking susceptibility for GTA welds. Fig. 16 — Effect of flushing time on moisture content in the shieldinggas.

Fig. 17 — Hydrogen cracks in ferrite crystals at 200X in duplex stain-less steel weldments (122 FN).

hydr

ogen

con

tent

(m

L/10

0 g)

hydr

ogen

con

tent

(m

L/10

0 g)

hydr

ogen

con

tent

(m

L/10

0 g)

Sensitivity for cracking in duplexstainless steel weldments using the bendtest has been proven in earlier investiga-tions by the authors (Refs. 1, 2) andFekken (Ref. 9). These investigations re-ported cracks for GTA welds using Ar +2%H2 shielding gas, SA welds with alu-minate-rutile fluxes, FCA welds with ru-tile flux cored wires and SMA welds withrutile-coated electrodes. An overviewhas been given in Fig. 5. A typical exam-ple of cracks in weld metal after defor-mation (bend specimen) from the previ-ously mentioned reports is in Fig. 17. Thisfigure shows hydrogen cracks in the fer-ritic crystals at a magnification of 200X induplex stainless steel weld metal with122 FN.

Microstructure and hydrogen contentof the weld metal are important for crack-ing susceptibility. When the ferrite con-tent in duplex stainless steel increases,hydrogen-assisted cracking becomesmore relevant. Austenitic-ferritic duplexstainless steels solidify from liquid as fer-rite. When temperature decreases, thisferrite becomes unstable relative to otherphases. Desirable ferrite transformspartly to austenite but intermetallic com-pounds can be formed as well. Depend-ing on composition and cooling rate, car-bides and nitrides may precipitate in theferrite phase. During arc welding, cool-ing rates from the fully ferritic high tem-peratures are relatively rapid, so equilib-rium is not fully reached and more ferritethan desired will be formed. Hydrogen-induced cracking has occurred in prac-tice with high ferrite levels in the weldmetal (Ref. 13). In duplex stainless steels,austenite and ferrite have diffusion coef-ficients for mono-atomic hydrogen thatdiffers about 2 or 3 orders of magnitude,being the lowest in austenite. The austen-ite in a duplex structure acts as a barrierto hydrogen escape (Ref. 10). As austen-ite forms first at the grain boundaries ofthe original ferrite and envelopes the pri-mary ferrite, the hydrogen present istrapped within the ferrite and hydrogendiffusion is limited. When plastic defor-mation occurs in this situation, active hy-drogen at the deformed austenite/ferriteinterface may cause hydrogen cracking.

The lower hydrogen levels, measuredwith higher heat input, might be due to alonger time for the weld metal being inthe delta ferrite phase, giving the hydro-gen time to diffuse out of the weld metal.In the past, hydrogen cracking has beennoticed in cases where hydrogen wasadded to the shielding gas (Ref. 14) andboth restraint intensity and weld-metalferrite level were high. However, the ad-dition of hydrogen to the shielding gas forimproved welding efficiency or betterweld appearance through increase of arc

energy density is currently not commonpractice in welding (super) duplex stain-less steels.

As shown in Fig. 14, high hydrogenlevels over 15 mL/100 g deposited weldmetal were present in GMA weldmentswith high ferrite. Ferrite numbers of 81and 82 for duplex and 105 and 106 forsuper duplex did not result in crackingwhen a strain ranging from 21 to 28%was applied. For GTA weldments (Fig.15), hydrogen levels up to 40 mL/100 gdeposited weld metal and ferrite num-bers of 91–114 for duplex and 126–135for super duplex gave no cracks, evenwith a strain ranging from 25 to 28%.Theoretically, 140 FN (being 100% fer-rite) is the maximum one can measure.Based upon welding procedure and con-sumables, the measured ferrite numbersare not rare. They are, however, severefor hydrogen cracking susceptibility. Forthese severe test conditions, the shieldinggas moisture level was 1000 ppm.

Conclusions

Susceptibility to hydrogen cracking,using the GTA or GMA process with stan-dard duplex and super duplex weldingconsumables appears to be limited. Al-though very high ferrite levels were com-bined with high restraint and high hy-drogen content, none of the GMA andGTA welds cracked.

In practice, a guaranteed moisturelevel below 10 ppm in the shielding gasis of less importance than applied weld-ing parameters for these conditions. Evena moisture level of 1000 ppm in theshielding gas, in combination with cor-rect welding parameters, will result inlow hydrogen content. A maximum of 10ppm moisture in shielding gas, however,will not necessarily result in low weld-metal hydrogen content.

Hydrogen in the weld metal appearedto be more dependent on welding pro-cedure rather than moisture content ofthe shielding gases investigated, or by hy-drogen initially present in GTA or GMAconsumables.

For the GTA process, a minimum heatinput of 2 kJ/mm was preferred for 2.4-mm filler metal to get low-deposited hy-drogen in the weld metal. For the GMAprocess, a 1.2-mm welding wire at 250 Aand a heat input of 1.3 kJ/mm gave thesame low hydrogen result.

The lower hydrogen levels measuredat the higher heat inputs might be due toa longer time the weld metal was in thedelta-ferrite phase, giving the hydrogentime to diffuse. The main reason for a dif-ference in hydrogen level was the heatinput, which was a result of changingwelding speed for GTA welds and currentlevel for GMA welds.

The type of shielding gas, which in-fluences the arc temperature, did not sig-

nificantly affect the hydrogen level in theweld metal.

The encapsulation technique for mea-surement of hydrogen content in superduplex stainless weldments appeared tobe inadequate without a subsequent hotextraction at 900°C. On average, another12% was measured with an additionalhot extraction cycle at 900°C.

Acknowledgment

The authors would like to recognizeJ. de Ruiter for his dedicated work on hy-drogen analysis.

References

1. Mee, van der V., Meelker, H., and Nas-sau, van L. 1994. How to avoid hydrogencracking in (super) duplex stainless steel weld-ments. Proc. 4th International Conference onDuplex Stainless Steels. Glasgow, Scotland,Paper 119. Glasgow. Woodhead PublishingLtd.

2. Mee, van der V., and Meelker, H. 1995.Waterstofscheuren in (super) duplex roestvaststaallasmetaal. Lastechniek April, pp. 9–12.

3. Lundin, C. D., et al. 1991 thru 1993.Measurement of diffusible hydrogen contentand hydrogen effects on the cracking potentialof duplex stainless steel weldments. Phase re-ports, Welding Research Council Progress Re-ports.

4. Shinozaki, K., Ke, L., and North, T. 1992.Hydrogen cracking in duplex stainless steel weldmetal. Welding Journal 71 (11): 387-s to 396-s.

5. Ogawa, K., and Miura, M. 1991. Hy-drogen cracking in duplex stainless steel weld-ments. Welding International (9): 691–696.

6. Gandy, W., and Findlan, S. 1993. Dif-fusible hydrogen in low alloy steel GTA welds.EPRI program.

7. Farish, E. 1994. Atmospheric contami-nation of TIG welding gas hoses — causes andcures. Welding & Metal Fabrication, July, pp.300, 301.

8. Brodén, G., and Tåje, R. 1990. Impuri-ties in gas supply systems for laser gases. In-ternal document REMW-9091.

9. Fekken, U., Nassau, van L., and Verweij,M. 1986. Hydrogen induced cracking inaustenitic ferritic duplex stainless steel. Proc.2nd International Conference on DuplexStainless Steels. Paper 26, The Hague.

10. Kotecki, D. J. 1991. Phase transforma-tions and weldability of duplex stainless steels.IIW document H-C-874-91.

11. Gunn, R. N. 1997. Hydrogen crackingof duplex stainless steel multipass weldments.Technical note from Group Sponsored Project5669 TWI.

12. AGA. 1992. Facts about laser gas supplysystems. Internal document AGA AB, Sweden.

13. Mee, van der V. 1997. Hydrogencrackining of duplex stainless steel. Internalreport, Lincoln Smitweld, The Netherlands.

14. Hilkes, J. 1985. Investigation of crack-phenomena in GTA welding DSS (W NR1.4462) using Ar or Ar+5%H2 shielding gas,

14-s | JANUARY 1999

RE

SE

AR

CH

/DE

VE

LO

PM

EN

T/R

ES

EA

RC

H/D

EV

EL

OP

ME

NT

/RE

SE

AR

CH

/DE

VE

LO

PM

EN

T/R

ES

EA

RC

H/D

EV

EL

OP

ME

NT