Embed Size (px)
Citation preview
ABSTRACT. Laser beam (LB) welding isincreasingly being used in welding ofstructural steels. The thermal cycles as-sociated with laser beam welding aregenerally much faster than those in-volved in conventional arc weldingprocesses, leading to a rather small weldzone, that usually exhibits a high hard-ness for C-Mn structural steels due to theformation of martensite. It is rather diffi-cult to determine the tensile properties ofa laser weld joint area due to the smallsize of the fusion zone. Complete infor-mation on the tensile and fracture tough-ness properties of the fusion zone is es-sential for prequalification and acomplete understanding of the joint per-formance in service, as well as for con-ducting the defect assessment procedurefor such weld joints. Therefore, an ex-perimental investigation on the mechan-ical properties of laser welded jointsusing flat microtensile specimens (0.5mm thick, 2 mm wide) was carried out toestablish a testing procedure to deter-mine the tensile properties of the weldmetal and heat-affected zone (HAZ) ofthe laser beam welds.
In the present work, two similar joints,namely, ferritic-ferritic and austenitic-austenitic and one dissimilar ferritic-austenitic joint were produced with aCO2 laser using 6-mm-thick steel plates.In addition to the testing of flat microten-
sile specimens, the mechanical proper-ties were examined by microhardnesssurvey and conventional transverse andround tensile specimens. The results ofthe microtensile specimens were com-pared with standard round tensile speci-mens, and this clearly showed the suit-ability of the microtensile specimentechnique for such joints. The crack tipopening displacement (CTOD) tests werealso performed to determine the fracturetoughness of the LB welds using three-point bend specimens. The effect ofstrength heterogeneity (mismatching)across the weld joint and at the vicinityof the crack tip on the CTOD fracturetoughness values was also discussed.
Introduction
Steel is a good absorber of the lightwave lengths produced by CO2 andNd:YAG lasers and many steels are read-
ily weldable by this process. A series ofstudies describing the successful use oflaser beam (LB) welding to different steelsin various industrial applications can befound in the literature (Refs. 1–9). How-ever, the chemical composition (particu-larly C, P and S contents as well as car-bon equivalent) of the structural steelssignificantly influences the laser weld-ability of these materials. In modernstructural steels, the carbon content issignificantly reduced and the strength isattained by alloying elements and/orthermal processing during rolling. Thesefine-grained steels are particularly suit-able for the low-heat-input laser weldingprocess to avoid the development of acoarse-grained microstructure in theHAZ region. However, the low heat inputand high cooling rate (high weldingspeed) typical of this process promote theformation of hard and brittle microstruc-tures (i.e., martensite) within the narrowweld and HAZ regions of steels subjectedto solid-state phase transformations.Since the hardness values reached inthese regions are usually well abovethose specified in standards and codes ofconventional arc welds, expensive andtime-consuming qualification proce-dures may be required for some compo-nents (Ref. 10). It has been reported (Ref.11) that the use of the IIW formula for car-bon equivalent (Ceq) is not adequate to as-sess the hardening effect in the fusionzone of the laser welds of C-Mn steels.
The weld formation and quality of LBsteel weldments are usually associatedwith three aspects: porosity, solidifica-
WELDING RESEARCH SUPPLEMENT | 193-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
WELDING RESEARCHSUPPLEMENT TO THE WELDING JOURNAL, JUNE 1999Sponsored by the American Welding Society and the Welding Research Council
Determination of Mechanical and Fracture Properties of Laser Beam Welded Steel Joints
BY G. ÇAM, S. ERIM, Ç. YENI AND M. KOÇAK
Flat microtensile specimens were found suitable for determining the mechanicalproperties of similar and dissimilar laser beamweld joints
KEY WORDS
Laser Beam WeldingLBWCO2 LaserFracture ToughnessFlat MicrotensileHardness TestCrack Tip OpeningC-Mn SteelHAZ
G. ÇAM and M. KOÇAK are with GKSS Re-search Center, Institute of Materials Research,Geesthacht, Germany. S. ERIM and Ç. YENIare with Dokuz Eylül University, MechanicalEngineering Dept., Izmir, Turkey.
tion cracking and high hardness in theHAZ and fusion zone. Pores are formedas a result of dissolved gases or gases aris-ing from contaminated surfaces, trappedprocess gases or evaporation of alloyingelements. In the case of steels, porosityhas been in general associated with lowgrade rimmed steels with oxygen con-tents above 100 ppm especially as thinsheet material, although the literaturealso reports this type of discontinuities onweldments produced in higher steelgrades (Ref. 5). At excessive weld cool-ing rates, the rate of escape of bubbleseventually formed in the fusion zone canbe lower than the rate of solidification re-sulting in various degrees of porosity inthe final weld. A recent study (Ref. 5) ondifferent steel grades, thickness andwelding speeds have shown that gener-ally the porosity level associated withslower welds is higher than those con-nected with faster welding speeds. The
evaluation of the weldment quality may,however, reveal that higher porosity lev-els, irrespective of base plate type, do nothave a particularly detrimental effect onweld joint transverse tensile propertiesdue to high strength overmatching of thefusion zone, which effectively shields thedefective weld zone.
There are no fracture mechanicsbased fracture toughness testing proce-dures available for laser beam weld edjoints despite the wide and inevitable usein modern engineering structures. Thisdiscrepancy is due mainly to the lack ofinformation on the interaction betweenthe base metal and fusion zone, whichhave significantly different tensile prop-erties. Substantial differences in strengthproperties (mismatching) of the basemetal and narrow fusion zone of the LBwelds inevitably occur due to the rapidthermal cycle of the joining process. Forthe CTOD toughness determination, the
effects of specimen geome-try (e.g. weld width, cracksize, notch position, etc.)and the degree of strengthmismatch (mismatch ratio,M = yield strength of fusionzone/yield strength of baseplate) between base metaland weld zone on tough-ness have, therefore, to betaken into account (Refs.12–17).
Almost all LB weldedstructural C-Mn steels ex-hibit a weld metal region ofhigher hardness andstrength (possibly withlower toughness) comparedto the base metal (over-matching) due to the rapidsolidification and singlepass nature of the weldingprocess. The laser weld re-gion with its high hardnessand strength makes it almost
impossible to determine the “intrinsicfracture toughness” properties of theweld region using conventional Charpy-V notch impact (Ref. 18) and CTODtoughness (Ref. 2) testing specimens dueto the crack path deviation towards thesofter base metal as a result of mechani-cal property mismatch between the basemetal and the weld zone. Therefore, bothtest results can only provide informationon the toughness performance of thewhole joint under impact and staticbending loading conditions. They cannotprovide required intrinsic toughnessproperties of the fusion zone due to in-evitable interaction between weld zoneand base plate. The toughness result ob-tained from such a specimen will in-evitably be higher due to tougher basemetal, but the result should not neces-sarily be classified as an “invalid” tough-ness result. An inevitable interaction be-tween lower strength base metal (at a
194-s | JUNE 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 — Three-point bend specimen geometry for fracture toughnesstesting.
Fig. 3 — Macrosections of the joints. A — Similar ferritic; B — similar austenitic; C — dissimilar joints.
Fig. 2 — Schematic showing the extraction and loading of the flat mi-crotensile specimens.
distance of about 1.5 mm to the crack tip)and the crack tip will occur and hencerelax the stress-state (i.e. constraint) at thecrack tip. Plasticity development in thebase metal will subsequently prevent thebrittle fracture initiation in the laser beamweld zone, which contains a martensiticmicrostructure and hence high hardness.The fracture toughness testing procedurefor LBW joints should take this very nat-ural phenomenon into account andhence should not take any artificial mea-sure (by producing larger weld zone orextensive side-grooving) to force the frac-ture process to remain within the weldzone. This may be achieved in laboratoryscale specimens, but fracture behavior ofLBW joints will follow its own naturalcourse during the service.
In the present study, an emphasis hasalso been given to the establishment ofthe flat microtensile specimen testingprocedure and hence the determinationof the tensile property gradient existing inthe LB fusion zone. The flat microtensilespecimen technique was originally de-veloped for property determination ofHAZ for conventional multipass weldjoints (Ref. 19). Successful applicationsof this technique for thick-section similarand dissimilar electron beam welds (Ref.12) and for the strength determination ofdiffusion welded joints (Refs. 20, 21)were also carried out at the GKSS re-search center. This study is an extensionof these experimental activities andspecifically addresses the developmentand refinement of the testing procedurefor laser beam steel welds, and hence,similar and dissimilar laser beam weldjoints between ferritic and austeniticsteels produced by CO2 laser using 6-mmthick plates were systematically investi-gated.
Experimental Procedure
In this study, austenitic stainless steel(grade 1.4404) and ferritic steel (gradeSt37) were used as base plates. The me-chanical properties of the base plates aregiven in Table 1. Similar and dissimilarsingle-pass full penetration CO2 LB weldson butt joints were produced withoutusing filler metal. Extensive microhard-ness measurements (using 100-g load)were performed across the weld regionsat three different locations, namely at theweld root, mid-section and top part of thejoints. In addition, the HAZ region of theferritic steel was also screened by con-ducting microhardness measurementsparallel to the weld interface.
Nonstandard three-point bend speci-
mens, which were machine notched inthe weld zone, were further fatigue pre-cracked in order to introduce a sharpcrack (a/W = 0.5) as illustrated in Fig. 1,and they were tested at both room tem-perature and –40°C (–40°F). CTOD mea-surements using the GKSS developed d5clip gauges (Refs. 14, 22) were con-ducted on SENB specimens, which al-lowed a direct measurement of the cracktip opening displacement for each spec-imen.
Standard flat transverse-tensile speci-mens were extracted from the weldedplates. Weld reinforcement was ma-
WELDING RESEARCH SUPPLEMENT | 195-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. 4 — Hardness profiles of the joints. A — Schematic illustration of the micro-hardness mea-surement procedure; B — similar ferritic; C — similar austenitic; D — dissimilar joints.
Fig. 5 — Stress-strain curves of LB weld joints determined by testing transverse tensile specimens.
chined before testing. Sets of flat mi-crotensile specimens were also extractedby spark erosion cutting from base met-als, HAZs and weld metals of all thejoints studied as schematically shown inFig. 2. The flat microtensile specimenpreparation was conducted mainly intwo stages: 1) extraction of a pre-shapedblock with laser weld in the middle, 2)cutting out specimens from etched pre-shaped block using a spark erosion cut-ting technique (with 0.1-mm diameterCu wire) parallel to the weld. Due to thesmall size of the microtensile specimens,
loading was introduced using four high-strength round pins at the shoulders ofthe specimens — Fig. 2. All tensile testswere carried out at room temperatureusing a displacement rate of 0.5 mm/minin a screw-driven universal testing ma-chine. After testing, the broken half of theflat microtensile specimens weremounted for microstructural verificationof the specimen location. The fracturesurfaces of selected bend and tensilespecimens were also examined by scan-ning electron microscopy for presence ofporosity.
Results and Discussion
Microstructural Observations
Microstructural examinations of thejoints investigated showed that the weldregions of similar ferritic and dissimilarjoints contained bainite and martensite.Figure 3 shows macrosections of thejoints. Similar ferritic joints displayed aweld metal structure consisting of bainiteand martensite, whereas similaraustenitic welds exhibited an austenite-dendritic (cellular) structure with no evi-dence of martensitic formation as ex-
196-s | JUNE 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. 6 — Stress-strain curves of the flat microtensile specimens ex-tracted from: A — Similar ferritic; B — similar austenitic; C — dis-similar joints.
Fig. 7 — Mechanical property variations across the joints (0 represents the weldcenter): A — Similar ferritic; B — similar austenitic; C — dissimilar joints.
pected. Contrary to a similar austeniticjoint, a distinct HAZ development oc-curred in a similar ferritic joint. The HAZin these joints contained a refined fer-rite/pearlite structure and pearlite disso-lution at the base metal sides. A dissimi-lar joint, on the other hand, showed aninhomogeneous weld metal microstruc-ture containing a mixture of ferrite andaustenite solidification structures in vary-ing degrees due to an incomplete mixtureof the molten metals of both sides in theweld pool during solidification. As re-ported earlier for dissimilar LB joints be-tween St37 and austenitic steels (Refs. 2,4), some solidification cracks parallel tothe dendritic growth were observed inthe weld zone of the dissimilar joints,which indicates that sufficient thermalstresses were developed in these weld-ments to separate the grain boundaries inthe cellular structure during solidifica-tion, which was coarser than that in theweld metal of similar austenitic joints.The microstructural aspects of thesejoints were discussed in more detail in anearlier publication (Ref. 4).
Hardness
In order to determine the hardnessprofiles, extensive microhardness mea-surements were conducted at three posi-tions (top, middle and root), Fig. 4A. Themicrohardness results exhibited no sig-nificant difference between these posi-tions for all the weld joints studied, Fig.4, implying that no significant gradient inmechanical properties of laser weldsalong the plate thickness direction is pre-sent. Similar ferritic steel joints displayeda high hardness profile in the weld regionwhile austenitic welds showed almostno change in hardness across the joint.Similar St37 joints exhibited a peakhardness value of about 330 HV, whereassimilar austenitic joints displayed a hard-ness value of about 210 HV in the weldregion — Figs. 4B and 4C, respectively.This is expected due to the lack of bainiteand/or martensite formation in the weldregions of austenitic joints. Dissimilarjoints displayed a hardness peak of about380–400 HV within the weld metal —Fig. 4D, which is slightly higher than thatof similar ferritic weld metal.
Tensile Properties
Standard round tensile specimensfrom the base metal were tested to findout the strength mismatch ratio (M) asgiven in Table 1. The yield strength mis-match ratio, M = YSSt37 / YS1.4404, be-tween the two base metals is found to be0.76. Flat transverse tensile specimenswere also tested to find out the nominal
mechanical proper-ties of the joints. Thespecimens of similarand dissimilar jointscontaining ferriticsteel always failed inthe lower strength fer-ritic base metal. Thestress-strain curvesobtained from flattransverse tensilespecimens showsome differencescompared to theround tensile speci-mens due to the pres-ence of LB weld re-gion at the middle ofthe specimens. Thestress-strain curve ob-tained from the flattransverse tensilespecimens of the dis-similar joint lies be-tween those of thesimilar joints of theconstituent steels, asexpected — Fig. 5.Evidently, these tests,particularly for ferriticjoints where a highhardness profile ex-ists, do not give anyinformation on thelocal mechanicalproperties of the weldregion.
WELDING RESEARCH SUPPLEMENT | 197-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. 8 — Comparison of the stress-strain curves of base metals obtainedby testing standard round and flat microtensile specimens: A — Fer-ritic; B — austenitic steel.
Fig. 9 — Comparison of CTOD values of base metals and laser beam welded joints at room tem-perature. Dissimilar joints exhibit lower toughness levels due to strength mismatch induced con-straint at the crack tip. (Note: a different symbol has been used for each specimen configurationfor clearness).
Therefore, all-weldmetal and all-HAZ flatmicrotensile specimenswere prepared andtested to determine thelocal mechanical prop-erties of the respectiveareas in the weld re-gions. Table 2 summa-rizes the mechanicalproperties of the ferriticbase metal (BM-F),austenitic base metal(BM-A), weld metal(WM), HAZ of ferriticsteel (HAZ-F) and HAZof austenitic steel (HAZ-A) on similar and dissim-ilar joints determinedusing microtensile speci-mens. The ferritic weldmetal region exhibits thehighest strength with ap-proximately 310%strength overmatching asmicrohardness results in-dicated. The full stress-strain curves of thesespecimens are given inFig. 6. The individualvalues of the yieldstrength (YS), tensilestrength (TS) and fracturestrain with respect to thespecimen location arepresented in Fig. 7.
Similar ferritic joint.Figure 6A illustrates the
stress-strain curves of the microtensilespecimens extracted from a similar fer-ritic joint, and Fig. 7A shows the varia-tion in the mechanical properties of thesame joint. As seen from these figures,the weld metal exhibits the higheststrength level (overmatching, M = 3.1)and the lowest strain value compared tothe BM and HAZ regions.
Similar austenitic joint. For the similaraustenitic joint, the weld metal and HAZexhibit slightly lower strain values withapproximately 110% overmatching(Table 2) in the strength level — Figs. 6Band 7B.
Dissimilar joint. In the dissimilar joint,the weld metal exhibits a high strengthlevel with a limited strain value as seenin Figs. 6C and 7C. The HAZ at the fer-ritic side (HAZ-F) also displays relativelyhigher strength levels and lower strainvalues than the ferritic base metal. TheHAZ at the austenitic side (HAZ-A), onthe other hand, does not exhibit a signif-icant increase in strength, but some de-crease in strain value as compared toaustenitic base metal.
In order to investigate the suitability ofmicrotensile specimen technique to de-termine the mechanical properties oflaser beam welded joints, the stress-strain curves obtained from base metalmicrotensile specimens were comparedwith those obtained from base metalstandard round tensile specimens — Fig.8. As seen from this figure, microtensilespecimens of the base metals exhibitedsimilar stress-strain curves to those ofstandard round tensile specimens of thebase metals, indicating that this tech-nique can be successfully employed todetermine the local mechanical proper-ties of laser beam welded joints.
As demonstrated above, the mechan-ical properties (including full stress-straincurves) of each zone of laser beamwelded similar and dissimilar joints canbe successfully determined using flat mi-crotensile specimens. Complete infor-mation (not only hardness values) on thelocal mechanical properties of laserwelds is often essential for optimizationof the laser welding process and fillermetal development for various alloys, aswell as quality control and fractureanalyses (experimental and numerical) ofthe welds. Development of filler metalcomposition to reduce the hardenabilityof C-Mn steel joints and the softening ofAl-alloy weld joints is currently of greatinterest. In order to make a step forwardin these areas, complete information onthe local properties of real weld jointsshould be generated. The tensile testingtechnique described in this paper canprovide such information.
198-s | JUNE 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. 10 — Comparison of CTOD values of base metals and laser beam welded joints at –40°C.(Note: a different symbol has been used for each specimen configuration for clearness).
Fig. 11 — Crack tip branching and distinct crack path deviationinto lower strength base metal in similar ferritic joint.
CTOD Fracture Toughness
The problem of measuring an “intrin-sic fracture toughness” value (in terms ofstandardized CTOD or J) of mis-matchedlaser weld specimens is very hard toovercome due to the high strength mis-match within a very small region at thevicinity of the crack tip. It is not a simpletask to distinguish the contributions fromboth the base metal (lower strength) andthe narrow LB weld (highly over-matched) at the vicinity of the crack tipto the remotely measured crack mouthopening displacement (CMOD) and loadline displacement (VLL) usually used instandardized CTOD and J estimates.
A local and direct measurement tech-nique (d5 technique) was developed atGKSS Research Center as a measure ofthe CTOD for determining the fracturetoughness and the crack growth resis-tance. It consists of measuring the rela-tive displacement of two gauge points di-rectly at the crack tip using specialdisplacement gauges. The resultingCTOD is called d5 because the gaugelength over which the CTOD is deter-mined amounts to 5 mm. The advantageof this measurement concept is that thed5 type CTOD can be easily measured onany configuration with a surface break-ing crack; no calibration functions are re-quired. Another appealing aspect of thed5 technique is that since it is measuredlocally as a displacement at the locationof interest, it does not have to be inferredfrom remotely measured quantities, likethe J integral or the standardized CTOD.This is of particular importance when thespecimen is mechanically inhomoge-neous, as is the case for highly mis-matched strength in narrow laser welds.The CTOD values of the laser weld jointson the SENB specimens have, therefore,been measured in terms of the CTOD (d5)technique.
For each weld condition, three deeplynotched (a/W = 0.5) three-point bendspecimens were tested at room tempera-ture (RT) and –40°C, and they all exhib-ited fully ductile fracture behavior. Fer-ritic base metal displayed similar CTODvalues at both testing temperatures, indi-cating that –40°C lies at the upper shelfof ductile-brittle transition for this ferriticsteel grade, whereas austenitic basemetal exhibited slightly lower CTOD val-ues at –40°C, indicating sensitivity oftoughness to testing temperature — Figs.9 and 10.
Similar ferritic joints displayed higherCTOD values than the base metal at RT,which can be attributed to the extensivecrack tip branching and crack path devi-ation into the softer base metal due to ex-
tremely high overmatch-ing of the fusion zone(about 310%, Table 2) —Fig. 11. The very highCTOD values for similarferritic laser weld jointsare obviously not repre-senting the intrinsictoughness properties ofthe weld zone, whichshowed very high hard-ness values and predom-inantly bainitic/marten-sitic microstructure. Ifthe maximum loadCTOD values (dm) are re-ported (as standardCTOD procedure re-quires) the toughnesslevel of laser welds willtherefore be overesti-mated. This is due to theeffect of lower strengthbase metal present nearthe vicinity of the cracktip (laser weld widthbeing approximately 2mm), which relaxes thestress state at the fatiguecrack tip in the middle ofthe laser weld. The ap-plied deformation prin-cipally goes to the lowerstrength base metal partof the specimen, andhence, the critical frac-ture stress for a possiblebrittle fracture at thecrack tip cannot be
WELDING RESEARCH SUPPLEMENT | 199-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. 12 — Crack propagation in dissimilar joint along the weldzone (small arrows indicating solidification cracks). Note crackpath deviation into lower strength ferritic base metal.
reached for ferritic simi-lar welds.
Dissimilar jointsshowed the lowestCTOD values at RT (Fig.9) due to asymmetricplastic zone develop-ment, which comes fromthe strength mismatchbetween ferritic andaustenitic steel base met-als (M = 0.76, Table 1).Figure 12 shows the dis-similar joint crack tipthat displays a one-sideddeformation, and hence,crack growth towardslower strength ferriticsteel (the mismatch be-tween the ferritic basemetal and fusion zone isabout 230%, Table 2). Itis interesting to note thatthe presence of small so-lidification cracks lyingperpendicular to thecrack direction in theweld zone of dissimilarjoints (Fig. 12) did notshow any detrimental ef-fect on the fracturetoughness values, al-though dissimilar jointsexhibited slightly lowerCTOD values than simi-lar joints. This can bedue to the fact that ex-tensive plastic deforma-tion developed at thelower strength ferriticsteel side of the speci-men (shielding effect ofthe overmatched weldzone).
Examination of the fracture surfacesas well as the sectioned specimens ofsimilar austenitic welds after testing at RTexhibited no crack tip branching but stilla slight crack path deviation into the basemetal (overmatching of the fusion zone isabout 110%, Table 2) in the austeniticsimilar joints. The crack propagatesalong the weld region parallel to the weldat both testing temperatures (Fig. 13) il-lustrating that these specimens do notcontain strength mismatch between thelaser weld zone and the austenitic basemetal as hardness results indicated. In thedissimilar ferritic-austenitic joints, thecrack path deviates clearly into the lowerstrength ferritic steel side and propagateswithin the ferritic steel at both testingtemperatures as illustrated in Fig. 12. Onthe other hand, an excessive crack tipbranching in the similar ferritic joints wasobserved, and the cracks always devi-ated into the softer base metal due to ex-tensive plastic zone development withinthe base metal at both testing tempera-tures — Fig. 11.
The fracture behavior of all thesespecimens can be explained with thehelp of strength mismatch information,hardness profiles and the microstructuresdeveloped in the weld regions. Obvi-ously, the ferritic base metal is the weak-est constituent (in terms of tensilestrength) in similar ferritic joints and indissimilar joints into which the crack de-viates. This crack path deviation into thelower strength base metal, which isschematically shown in Fig. 14, duringstandard three-point bend tests with thenotch in the overmatched narrow laserweld zone leads to an experimental dif-ficulty in determining the intrinsic frac-ture toughness properties of thin sectionC-Mn steel laser welds.
Conclusions
The results of this work provide thefollowing conclusions:
1) The microstructures of CO2 laserwelded C-Mn steel contain large propor-tions of bainite/martensite in the weld re-gion due to rapid cooling involved. Anextreme hardness increase in the weld re-gions of ferritic similar joints and ferritic-austenitic dissimilar joints were ob-served, while there is no significanthardness increase in the weld region ofthe austenitic steel similar joints.
2) All the transverse tensile specimensof the joints containing ferritic con-stituent failed at the lower strength ferriticbase metal sides due to the strength mis-match effect. Even the presence of somesolidification cracks in dissimilar jointsdid not change the fracture location, due
200-s | MONTH 1997
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 — Crack propagation in a similar austenitic joint. There isno distinct crack path deviation but the crack propagates in thebase metal along the weld zone.
Fig. 14 — Schematic showing extensive plastic zone development in the soft base metal of simi-lar ferritic steel laser beam weld during fracture toughness testing. The crack path deviation in-evitably occurs towards the lower strength base metal and the toughness result obtained repre-sents the base metal in thin section laser welds of C-Mn steels.
to the strength overmatching of the weldregion.
3) Microtensile specimens extractedfrom the weld metal of similar ferriticjoints displayed significantly higherstrength values (M = ~3.1) and markedlylower strain values than the base metalspecimens, due to the presence of bai-nite/martensite in the weld zone. The all-HAZ specimens also exhibited higherstrength values and lower strain valuesthan those of the base metal.
4) Similar austenitic joints exhibitedno significant property variation acrossthe weld zone as indicated by a hardnessprofile. The microtensile specimens ex-tracted from the weld zone of the dis-similar joint exhibited high strength andlow strain values. The all-HAZ speci-mens extracted from the HAZ at the fer-ritic side also displayed higher strengthand lower strain values than the ferriticbase metal.
5) CTOD fracture toughness testingdoes not provide “intrinsic toughnessvalue” for thin section laser beam weldsowing to crack tip branching and/orcrack path deviation towards the lowerstrength base metal side (mismatch).
6) The CTOD values obtained for sim-ilar and dissimilar joints demostrate thetoughness trends, which can be ex-plained with strength mismatch and mi-crostructure aspects.
7) Finally, an application of the flatmicrotensile specimen technique to laserwelded joints was demonstrated. Byusing this technique, it is possible to de-termine the local mechanical propertiesof the joints, which can be correlated tothe respective microstructure and hard-ness.
Acknowledgments
This work is a part of the German-Turkish joint project financed by Interna-tional Bureau, KFA-Jülich and TürkiyeBilimsel Teknik Arastirma Kurumu(TÜBITAK), Turkey. The authors wouldlike to thank both organizations for theirfinancial support. They also would like tothank Mr. S. Riekehr and Mr. H. Mackelfor their technical assistance in testing.
References
1. Rippl, P. 1994. Laser beam welding withrobots in the automotive industry. 5th Euro-pean Conference on Laser Treatment of Mate-rials, ECLAT’94, pp. 151–164, Bremen, Ger-many.
2. Yeni, Ç., Erim, S., Çam, G., and Koçak,M. 1996. Microstructural features and fracturebehaviour of laser welded similar and dissim-ilar steel joints. Int. Welding Technology ’96Symposium, pp. 235–247, Istanbul, Turkey,Gedik Education Foundation.
3. Çam, G., Riekehr, S., and Koçak, M.1997. Determination of mechanical proper-ties of laser welded steel joints with microten-sile specimens. The ASM International Euro-pean Conference on Welding and JoiningScience and Technology, pp. 63–79, Madrid,Spain, ASM International, IIW Doc. SC X-F-055-97.
4. Çam, G., Yeni, Ç., Erim, S., Ventzke, V.,and Koçak, M. 1998. Investigation into prop-erties of laser welded similar and dissimilarsteel joints. Journal of Science and Technologyof Welding and Joining 3(3).
5. Brown, P. M., and Bird, J. 1996. Thelaser welding of Q1N, HSLA80 and NSS550.IIW Doc. IX-679-96
6. Kristensen, J. K., and Borgreen, K. 1994.Mechanical properties and microstructures oflaser welded constructional steels. 5th Euro-pean Conference on Laser Treatment of Mate-rials, ECLAT’94, pp. 226–239, Bremen, Ger-many.
7. Sun, Z., and Moisio, T. 1994. Effect ofprocessing parameters on laser welded dis-similar joints. Welding Journal 73(4): 63-s to70-s.
8. Strychor, R., Moon, D. W., and Met-zbower, E. A. 1984. Microstructure of ASTMA-36 steel laser beam weldments. JOM, pp.59–61
9. Metzbower, E. A. 1990. Laser beamwelding: thermal profiles and HAZ hardness.Welding Journal 69(7): 272-s to 278-s.
10. Dilthey, U., Dobner, M., Ghandehari,A., Lüder, F., and Träger, G. 1996. Entwick-lung, Stand und Perspektiven der Strahltech-nik, pp. 1–14, Konferenz Strahltechnik, Halle,Germany.
11. Kalla, G., Funk, M., et al. 1993. Heavysection laser beam welding of fine-grainedstructural steels. LASER Technology and Ap-plications, pp. 319, H. Kohler (ed.), Vulkan-Verlag, Essen, Germany.
12. Koçak, M., and Junghans, E. 1994.Fracture toughness properties of similar anddissimilar electron beam welds. 2nd EuropeanConference on Joining Technology, Florence,Italy.
13. Koçak, M. and Schwalbe, K.-H. 1994.Fracture of weld joints: mismatch effect.Neuentwicklungen im Konstruktiven Inge-nieurbau, pp. 137–155, Ed. H. Saal and Ö.Bucak, Univ. of Karlsruhe, Germany.
14. Schwalbe, K.-H., and Koçak, M. 1992.Fracture mechanics of weldments-propertiesand applications to components. 3rd Int. Conf.on Trends in Welding Research, pp. 479–494,Gatlinburgh, Tenn.
15. Koçak, M., Es-Souni, M., Chen, L., andSchwalbe, K.-H. 1989. Microstructure andweld metal matching effects on heat affectedzone toughness. 8th Int. Conf. OMAE-ASME,pp. 623–633, The Hague, Netherlands.
16. Kirk, M. T., and Doods, R. H. 1992. Theinfluence of weld strength mis-match on cracktip constraint in single edge notch bend spec-imens. University of Illinois Report, UILU-ENG-92-2005.
17. Cao, H. C. and Evans, A. G. 1994. Anexperimental study of the fracture resistance ofbi-material interfaces. Mechanics of Materials7: 295–304.
18. Kristensen, J. K. 1996. Procedure qual-ification, process monitoring, NDT and adap-tive welding control. Int. Conf. on Explotationof Laser Processing in Shipyards and StructuralSteelworks, Glasgow, U.K.
19. Klausnitzer, E. 1976. Entwicklungeines Verfahrens zur Entnahme und Prüfungvon Mikro-Flachzugproben aus der WEZ vonSchweißverbindungen. Materialprüfung 18(11): 411–416
20. Çam, G., Bohm, K.-H., Müllauer, J.,and Koçak, M. 1996. The fracture behavior ofdiffusion-bonded duplex gamma TiAl. JOM,pp. 66–68.
21. Çam, G., Müllauer, J., and Koçak, M.1997. Diffusion bonding of two-phase g-TiAlalloys with duplex microstructure. Journal ofScience and Technology of Welding and Join-ing 2(5): 213–219.
22. Koçak, M., and Denys, R. 1994. CTODand wide plate testing of welds with particu-lar emphasis on mis-match welded joints.10th European Conf. on Fracture, ECF 10,Structural Integrity, Experiments, Models, Ap-plications, pp. 97–120, Edited by K.-H.Schwalbe and C. Berger, Engineering Materi-als Advisory Services (EMAS) Ltd.
WELDING RESEARCH SUPPLEMENT | 201-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
