Microstructure and Mechanical Property Change During FSW and GTAW of Al6061 Alloy V. FAHIMPOUR, S.K. SADRNEZHAAD, and F. KARIMZADEH The variation of morphology and mechanical properties of Al6061 automotive aluminum alloy due to friction stir welding (FSW) and gas tungsten arc welding (GTAW) was investigated by optical metallography, scanning electron microscopy, microhardness measurement, X-ray dif- fraction, tensile testing, and fractography. The center-line dendrite emergence and micro- hardness reduction in the heat-affected zone were observed in the GTAW process. Although similar microhardness reduction with respect to the base metal was observed in the FSW samples, higher HVs were obtained for the FSW rather than the GTAW process at almost all heat-affected locations. Ultimate tensile strengths of the FSW and the GTAW samples in the transverse direction were ~0.57 and ~0.35 of the base metal, respectively. Post-weld aging improved the strength, but reduced the ductility of the welding. DOI: 10.1007/s11661-012-1588-4 Ó The Minerals, Metals & Materials Society and ASM International 2013 I. INTRODUCTION RELIABLE joints with high strength, long fatigue life, and substantial fracture resistance are vital to the aerospace and the automobile industries that use alu- minum alloys. Heat treatable aluminum alloys of 2xxx, 6xxx, and 7xxx series are widely used in these indus- tries. [1] This is due to their high strength to weight ratio, good formability, and acceptable corrosion resistance. [2] All parts of a perfect weld need to have the same properties as the base metal. Conventional welding processes suffer from such defects as distortion, gas porosity, residual stress, dendritic brittle structure, lack of fusion, oxide inclusions, hot cracking, and hot tearing of the weld. [2] Reduction of yield stress, ultimate tensile strength, [3] corrosion resistance, [4] and electrical resistiv- ity in the heat-affected zone (HAZ) region [5] are some common consequences of using traditional welding processes. The applied alloy may lose its mechanical strength due to precipitate dissolution [6] and softening heat effects. [7] Gas tungsten arc welding (GTAW) is traditionally used for joining of the aluminum alloys. 6061 aluminum alloys containing magnesium and silicon have low weldability. Microstructure of the GTAW sample usu- ally includes coarse columnar grains similar to the cast samples. According to previous reports, mechanical properties such as tensile [8] and fatigue properties of these alloys have sensibly reduced after pulsed GTAW. Also, according to previous investigations, GTAW has caused significant lowering of metal properties like resistance to corrosion in the welded region of the 6061 aluminum alloy as compared to the base metal. [9] Friction stir welding (FSW) is a solid state joining process applicable to Al alloys. As a result of the friction stirring and the movement of the materials due to the welding, a sever deformation zone forms along the center- line of the weld. [10] The weld microstructure consists, hence, of three regions: (a) stirred (nugget) zone, (b) thermo-mechanically affected zone (TMAZ), and (c) HAZ. [11] According to reports, UTS of the joint by FSW is ~80 pct, while that of GTAW is ~67 pct of the base metal. [12] A linear regression relationship was founded between grain size of the weld nugget of FSW and tensile strength in welded 6061 aluminum samples. [13] As Al alloys are of great significance in the industrial applications, finding a sound welding technique for them is indispensable. A comparison of the FSW and the GTAW, as two common welding techniques for these alloys, on morphology and mechanical properties of the aluminum alloys is the key to the best selection of the appropriate welding method. This paper reports on the effects of the FSW as compared to the GTAW process on the microstructure and the mechanical properties of the Al6061 alloy for better achievement of the desirable welded metal properties. II. MATERIALS AND METHODS Rolled plates of Al6061 alloy in T6 condition of 250 9 50 9 13 mm 3 dimensions, with chemical compo- sitions given in Table I, were butt-joined together perpendicular to the rolling direction by both the FSW and the GTAW methods, separately. Two-sided welding was used in both cases to produce complete bonding of the samples across their thickness. Welding conditions are as listed in Table II. 60 degree beveling was used for V. FAHIMPOUR, Graduate Student, and S.K. SADRNEZHAAD, Professor, are with the Department of Materials Science and Engineering, Sharif University of Technology (SUT), P.O. Box 11365-9466, Tehran, Iran. Contact e-mail: [email protected] F. KARIMZADEH, Associ- ate Professor, is with the Department of Materials Engineering, Isfahan University of Technology (IUT), 84156-83111 Isfahan, Iran. Manuscript submitted June 20, 2012. Article published online January 9, 2013 METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 44A, MAY 2013—2187
Transcript
Microstructure and Mechanical Property Change DuringFSW and GTAW of Al6061 Alloy
V. FAHIMPOUR, S.K. SADRNEZHAAD, and F. KARIMZADEH
The variation of morphology and mechanical properties of Al6061 automotive aluminum alloydue to friction stir welding (FSW) and gas tungsten arc welding (GTAW) was investigated byoptical metallography, scanning electron microscopy, microhardness measurement, X-ray dif-fraction, tensile testing, and fractography. The center-line dendrite emergence and micro-hardness reduction in the heat-affected zone were observed in the GTAW process. Althoughsimilar microhardness reduction with respect to the base metal was observed in the FSWsamples, higher HVs were obtained for the FSW rather than the GTAW process at almost allheat-affected locations. Ultimate tensile strengths of the FSW and the GTAW samples in thetransverse direction were ~0.57 and ~0.35 of the base metal, respectively. Post-weld agingimproved the strength, but reduced the ductility of the welding.
DOI: 10.1007/s11661-012-1588-4� The Minerals, Metals & Materials Society and ASM International 2013
I. INTRODUCTION
RELIABLE joints with high strength, long fatiguelife, and substantial fracture resistance are vital to theaerospace and the automobile industries that use alu-minum alloys. Heat treatable aluminum alloys of 2xxx,6xxx, and 7xxx series are widely used in these indus-tries.[1] This is due to their high strength to weight ratio,good formability, and acceptable corrosion resistance.[2]
All parts of a perfect weld need to have the sameproperties as the base metal. Conventional weldingprocesses suffer from such defects as distortion, gasporosity, residual stress, dendritic brittle structure, lackof fusion, oxide inclusions, hot cracking, and hot tearingof the weld.[2] Reduction of yield stress, ultimate tensilestrength,[3] corrosion resistance,[4] and electrical resistiv-ity in the heat-affected zone (HAZ) region[5] are somecommon consequences of using traditional weldingprocesses. The applied alloy may lose its mechanicalstrength due to precipitate dissolution[6] and softeningheat effects.[7]
Gas tungsten arc welding (GTAW) is traditionallyused for joining of the aluminum alloys. 6061 aluminumalloys containing magnesium and silicon have lowweldability. Microstructure of the GTAW sample usu-ally includes coarse columnar grains similar to the castsamples. According to previous reports, mechanicalproperties such as tensile[8] and fatigue properties ofthese alloys have sensibly reduced after pulsed GTAW.
Also, according to previous investigations, GTAW hascaused significant lowering of metal properties likeresistance to corrosion in the welded region of the 6061aluminum alloy as compared to the base metal.[9]
Friction stir welding (FSW) is a solid state joiningprocess applicable to Al alloys. As a result of the frictionstirring and the movement of the materials due to thewelding, a sever deformation zone forms along the center-line of the weld.[10] The weld microstructure consists,hence, of three regions: (a) stirred (nugget) zone,(b) thermo-mechanically affected zone (TMAZ), and(c) HAZ.[11] According to reports, UTS of the joint byFSW is ~80 pct, while that of GTAW is ~67 pct of thebase metal.[12] A linear regression relationship wasfounded between grain size of the weld nugget of FSWand tensile strength in welded 6061 aluminum samples.[13]
As Al alloys are of great significance in the industrialapplications, finding a sound welding technique forthem is indispensable. A comparison of the FSW andthe GTAW, as two common welding techniques forthese alloys, on morphology and mechanical propertiesof the aluminum alloys is the key to the best selection ofthe appropriate welding method. This paper reports onthe effects of the FSW as compared to the GTAWprocess on the microstructure and the mechanicalproperties of the Al6061 alloy for better achievementof the desirable welded metal properties.
II. MATERIALS AND METHODS
Rolled plates of Al6061 alloy in T6 condition of250 9 50 9 13 mm3 dimensions, with chemical compo-sitions given in Table I, were butt-joined togetherperpendicular to the rolling direction by both the FSWand the GTAW methods, separately. Two-sided weldingwas used in both cases to produce complete bonding ofthe samples across their thickness. Welding conditionsare as listed in Table II. 60 degree beveling was used for
V. FAHIMPOUR, Graduate Student, and S.K. SADRNEZHAAD,Professor, arewith theDepartment ofMaterials Science andEngineering,Sharif University of Technology (SUT), P.O. Box 11365-9466, Tehran,Iran. Contact e-mail: [email protected] F. KARIMZADEH, Associ-ate Professor, is with the Department of Materials Engineering, IsfahanUniversity of Technology (IUT), 84156-83111 Isfahan, Iran.
Manuscript submitted June 20, 2012.Article published online January 9, 2013
METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 44A, MAY 2013—2187
preparation treatment of the plate edges before GTAW.The welding procedures used were according to theGTAW technique traditionally used for Al 6061 alloysin industry[14] and the optimum FSW conditionsobtained from previous investigations by the sameauthors.[14]
A H13 steel tool with a 7-mm pin diameter was usedfor the FSW process.[15] To increase hardness and wearresistance, the following heat cycle was applied to theFSW tool according to ASM standard with somemodifications: (1) austenitization at 1323 K (1050 �C)for 1 hour, (2) water quenching to the room tempera-ture, and (3) tempering at 723 K (450 �C) for 20 min-utes.[16] As a result, the hardness of the tool increasedfrom 48 to 63 HRC. For improvement of the strength ofthe welded samples, T6 heat treatment (the cycle shownin Figure 1) was applied to them.[17]
The welded samples were characterized by optical andscanning electron microscopy (SEM, Philips XL30). TheGTAW samples were etched using Keller’s etchant(composed of 2 mL HF, 3 mL HCl, 5 mL HNO3, and190 mL H2O).[18] Electroetching was also carried out byplacing the FSW samples as anode, an aluminum plateas cathode, and etchant containing 40 mL HNO3,30 mL HCl, 2.5 mL HF, and 42.5 mL H2O as electro-lyte. The electric potential was 10 V. Both electrolytecomposition and electroetching voltage were obtainedfor best metallographic results by a trial and errormethod of compositions given in ASM handbook[19]
also reported in a previous paper.[14] The lineal interceptprocedure was used to measure the average grainboundary sizes. XRD analysis was used for investiga-tion of different phases in base and welded samplesbefore and after T6 heat treatment.
A strain rate of 0.1 s�1 was used for tensile testing ofthe sub-size samples which were made according to theASTM-E8M standard.[18] Both longitudinal and trans-verse (with respect to the weld line) strengths were
measured. Average values of at least three specimenswere considered. A 50-g load was used to determine thevariation of the microhardness in the cross section of theweld. SEM micrographs of the fracture surfaces of thesamples in both longitudinal and transverse directionswere compared after tensile tests to determine effects ofdifferent welding procedures on rupture behavior of thesamples.
III. RESULTS AND DISCUSSION
A. Microstructure
Because of the cast-like solidification effects, thestructure of the GTAW adjacent to the fusion-line ofthe joined couple was dendritic. At the center-line of thewelds, many equiaxed grains were observed. This wassimilar to what was reported by previous investiga-tors.[20] Microstructures of the base metal, HAZ, par-tially melted zone (PMZ), and the dendritic region of aGTAW are depicted in Figure 2. Figure 3 illustrates themicrostructural regions of a typical FSW specimen. Thisfigure indicates that heat effects (due to severe plasticdeformation) result in grain refinement of the nuggetzone via dynamic recrystallization.Average grain size of the base metal is 90 ± 37 lm.
The nugget grain sizes are ~2 ± 1.5 lm. The thermo-mechanically affected zone (TMAZ) of the sample haslarger grains (~30 ± 14 lm), meaning that recrystalli-zation has occurred in the TMAZ regions, but due toinsufficient deformation strain, the recrystallization isless than nugget zone.[10] The HAZ region retains thesame microstructure as the parent alloy because there isno mechanical effect and the dissolution of the precip-itates due to lower-heating is small.[21]
The FSW process causes up to 673 K (400 �C)of temperature rise in the sample, which facilitates
Table I. Chemical Composition of the Base and the Filler Metals Expressed in Weight Percent
Table II. Conditions of the FSW and the GTAW ProcessesUsed in this Research
FSW Condition GTAW Condition
Tool material H13 filler material 4043*Pin diameter 7 mm diameter of
filler1.6 mm
Shoulderdiameter
20 mm feeding rate 300 mm/min
Pin height 7 mm welding current 200 ARotatingvelocity
800 rpm voltage 35 V
Advancingspeed
200 mm/min advancingspeed
160 mm/min
*4043 filler metal composition is given in Table I. Fig. 1—T6 heat treatment program performed on samples after theirwelding procedure.
2188—VOLUME 44A, MAY 2013 METALLURGICAL AND MATERIALS TRANSACTIONS A
dissolution of precipitates as observed by previousauthors.[21] The sizes of the precipitates are muchsmaller in the nugget zone than in the base metal region(Figure 4). Smaller precipitates indicate sensitivity of themicrostructure to the thermomechanical impressionsand the amount of heat released. This sensitivity reduceswith increase of distance from the center-line. Thenumber of precipitates in the nugget zone is higher thanin the base metal, but their volume fraction is lower.This is due to their dissolution resulting from theheating process. Figure 5 shows microstructure of (a)GTAW and (b) FSW samples after T6 heat treatment.The dendritic structure of the melted zone is observablein the GTAW-T6 specimen, while Nugget and TMAZhave the prevailing microstructures of the FSW-T6specimen. Figure 5(b) indicates the effect of T6 heattreatment on precipitation of larger intermetallics withingrains of the base metal, as observed by a comparisonwith Figures 2 and 3.
B. Composition
The compositional change due to welding and heattreatment is studied by XRD analysis of the weldedsamples. Figure 6 compares X-ray patterns of basemetal, FSW, GTAW, FSW-T6, and GTAW-T6 speci-mens. As can be seen, base metal contains small amountof Mg2Si precipitates with Al and Al2Fe as neighboringagents. Similar conditions exist in the FSW and GTAWsamples. Peaks of Si are also observable in the GTAWand GTAW-T6 samples, while no such peaks are seen inthe other specimens. These peaks are due to 4043 fillermetal used in the GTAW process. A comparison showsthat the effect of T6 heat treatment on GTAW samplesis more significant than on the other samples. Heattreatment increases the tendency of formation of FeSi2phase in GTAW specimens.
C. Microhardness
Microhardness profiles of the FSW and the GTAWsamples for areas close to the weld line are compared inFigure 7 along the transverse cross section of the welds.Both curves follow a similar trend. Microhardnessdecreases from the base metal to the HAZ region andincreases toward the center-line zone are observed. TheFSW microhardness profile shows a maximum of about71 HV at the small grain size region of the center-linezone. Microhardness decline from the center-line out-ward to ±7.5 mm of Figure 7 shows the effect of grainsize enlargement. Microhardness of the parent phaseincreases at further distances due to diminution of theFSW heating effect and dissolution of precipitates. Themicrohardness in the base material exceeds 91 HV. Allthese behaviors can be explained by consideration ofchanges in the grain size, precipitate dissolution, anddistribution during the welding process (Figures 3 and 4).
Fig. 2—Microstructures of the base metal, PMZ, HAZ, and the weld metal of a typical GTAW sample.
Fig. 3—Optical microstructure of a typical FSW sample.
METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 44A, MAY 2013—2189
Microhardness of the nugget zone is smaller than thebase metal due to higher precipitate dissolution near theFSW axis, but larger than the TMAZ zone because ofsmaller grain size and bigger thermomechanical effectsdue to heating up and deformation in the nugget zone ascompared to the TMAZ regions of the FSW specimens.
Figure 7 indicates that the microhardness of the FSWsample in different regions is higher than the GTAWspecimen. This is due to the difference in compositionsof the welded zone of the FSW and the GTAWmetal (Figure 6), lower heat release, smaller grain sizes(Figure 3), homogenous tiny precipitates (Figure 4), andthe thermomechanical impressions remaining in theFSW joint. Because of the lower FSW temperature,precipitate dissolution during the FSW is lower than theGTAW process. The different composition in weldregion due to use of 4043 filler metal was the mostimportant reason, while dissolution of some precipitates
and size increase were some others during the GTAWtechnique, which made the microhardness of heat-affected zone lower than the base metal.Microhardness of the welding line of the FSW sample
is different at different distances from lateral surfacesperpendicular to longitudinal axes of the samples. Themaximum value of 71 HV appearing in middle of thetwo surfaces is due to overlap of the two-side weldingand higher plastic deformation resulting from twice pintransverse. This value reduces to 56 HV and thenincreases to approximately 65 HV near the externalsurfaces perpendicular to the longitudinal axes of thesamples. Microhardness of post-heat-treated samples ofboth welding techniques is relatively more than the basemetal. This means that a good recovery was done formicrohardness with T6 heat treatment. In fact, afterwelding, large amounts of alloying elements are dis-solved in the Al matrix due to temperature increase
Fig. 4—SEM micrograph of different regions in the FSW specimen. A few precipitates are highlighted on the figure.
Fig. 5—Microstructures of (a) GTAW and (b) FSW after T6 heat treatment.
2190—VOLUME 44A, MAY 2013 METALLURGICAL AND MATERIALS TRANSACTIONS A
during the welding process. T6 post-welding heattreatment improved the weld with reprecipitating theseelements and strengthening the weld.[22]
D. Tensile Strength
Results of the tensile experiments for the base metal,the FSW, and the GTAW specimens before and after T6heating cycle are presented in Table III. Strength andductility of the base metals are significantly higher thanthe longitudinal and transverse values of the FSW andthe GTAW samples before T6 heat treatment, except the
FSW longitudinal sample that has higher ductility thanthe base metal. This seems due to the cover-up of allmicrostructural regions like the base material, the HAZ,the TMAZ (for FSW sample), and the center-linesectors in the transverse direction which causes failureof the samples from the least strong part of the FSW orGTAW samples. According to Table III, least strengthsof the FSW and GTAW samples occur in the HAZ–TMAZ interface and HAZ region, respectively. Acomparison with diagrams shown in Figure 7 indicatesthat the microhardness of the weakest zone of the FSWsample is greater than that of the GTAW weldedsample, justifying the greater fracture strength of thetransverse tensile specimens of the FSW specimen ascompared to the GTAW samples. Despite the coarsedendritic structure of the GTAW metal, the FSWnugget zone has fine and equiaxed grains with smallprecipitates, which results in greater tensile strength andductility.In the case of the longitudinal FSW results, because of
the sample direction along the center-line welding zoneand the smaller structural precipitates which providelower crack nucleation sites than the base metal,elongation increased in comparison with the base alloy.In order to recover the lost tensile strength of the
welding zones, a T6 post-weld aging treatment wasconducted on the welded samples. The aging treatmentresulted in considerable recovery of the tensile strengthsin both FSW and GTAW specimens. It reduced,however, the ductility of the samples (Table III).Strength enhancement and ductility reduction of thepost-weld treated samples can be attributed to theincrease in the volume fraction of the hard precipitatesas well as their buildup in the grain boundaries of thespecimens as indicated by previous investigators.[23] On
Fig. 6—X-ray diffraction of base, FSW, GTAW, FSW-T6, andGTAW-T6 samples.
Fig. 7—Microhardness profiles of the FSW and the GTAW welded samples. Notice 50 HV shifting of the data for prevention of theirinterference.
METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 44A, MAY 2013—2191
the other hand, the ductility of FSW-T6 samples is nearthree times higher than the GTAW-T6 samples due todifferent alloy compositions in the welded regions.
E. Fractography
Figure 8 exhibits the SEM micrograph of the basemetal fracture surface after tensile test. As can be seen,the fracture surface is of dimpled kind showing ductilefracture behavior.
Figure 9(a) shows the fracture surface in the longitu-dinal direction of the sample (FSW-L). As is seen, thefracture surface is of a ductile dimpled kind, identical tothat of the base sample shown in Figure 8. The dimpleshave a similar depth indicating relatively equal elonga-tion consistency. Figure 9(b) illustrates the fracturesurface in transverse direction of the sample (FSW-T)which consists of a non-uniform dispersion of dimples ofdifferent sizes. As is seen, the depth of the dimples is lessthan the base sample which, indeed, is a confirmationfor elongation reduction from 19 pct of the base metalto 12 pct for FSW-T sample.
A comparison of longitudinal with transverse fracturesurfaces of FSW andGTAW samples indicates shallowerand less uniform dimple distribution in both longitudinaland transverse directions of the GTAW (Figures 9(a)through (d)). Non-uniformity and shallowness of the
dimples in the transverse direction are much higher thanin the longitudinal path. Some cleavages are alsoobserved in both cases. It can, therefore, be stated thatquasi-cleavage fracture takes place in both samples.While the contribution of the cleavage is more in theGTAW-T sample, the contribution of the dimple and thecleavage fractures is equal in the GTAW-L samples. Thiscauses a higher tendency toward brittle fracture in theGTAW.SEM micrographs of longitudinal and transverse
directions of the FSW-T6 samples are shown inFigure 10. A comparison of the micrographs ofFigures 9(a) and (b) with Figure 10 indicates decreasingof the length and the depth of the dimples due to T6 heattreatment. Cleavage, which is interpreted as brittlefracture, is also apparent in the figure. It can, therefore,be said that both ductile (due to existence of dimples) andbrittle (due to appearance of cleavage) fractures occur inthe FSW-T6 samples. This implies the occurrence of somekind of quasi-cleavage fracture. Considering the lowerpercentage of the dimples with respect to the cleavage, itmay be concluded that the tendency toward brittlefracture is higher in these samples. Reduction in elonga-tion of the samples due to the heat treatment seems,therefore, logical. Post-weld T6 heat treatment strength-ens the FSW samples with reformation of precipitates inthe matrix. These precipitates act, however, as places of
Table III. Results of Tensile Tests for Base Metal and the Welded Samples
Material YP (MPa) UTS (MPa) Elongation (pct) Position of Fracture
Fig. 8—SEM micrograph of the fracture surface of the base metal at (a) low and (b) high magnifications.
2192—VOLUME 44A, MAY 2013 METALLURGICAL AND MATERIALS TRANSACTIONS A
crack initiation and decrease the ductility of the sample,especially when collected together.
F. Structure–Property Relationship
The effect of structure on properties of FSW can beexplained by comparison of microhardness data withchanges of grain size and precipitates distribution due towelding as is observable from Figures 3, 4, and 7.Microhardness of the nugget zone is smaller than thebase metal due to higher precipitate dissolution near theFSW axis, but larger than the TMAZ zone because ofsmaller grain size and larger thermomechanical effects(by heating and deformation) in the nugget zone ascompared to the TMAZ region of the FSW specimens.
The morphology of the precipitates in differentregions of the FSW sample is as illustrated in Figure 4.Some precipitates are highlighted with their averagesizes being given in the figure. Size, distribution, andmorphology of the precipitates before and after weldingby the FSW method in different regions are compared inTable IV. The SEM micrograph of FSW weld after T6heat treatment is also shown in Figure 11.Precipitates in the nugget zone have a nearly homog-
enous distribution (Figure 4) with relatively small size(2.5 ± 0.9 lm). This is due to the violent stirring of thisregion during FSW operation. Precipitates of TMAZregion are larger in size and less homogeneous(in comparison with the nugget zone). Precipitates ofthe HAZ region are even larger (in comparison with
Fig. 9—SEM micrograph of the fracture surface of (a) FSW-L, (b) FSW-T, (c) GTAW-L, and (d) GTAW-T samples.
Fig. 10—SEM micrograph of the fracture surface of the FSW-T6 sample in (a) longitudinal and (b) transverse directions.
METALLURGICAL AND MATERIALS TRANSACTIONS A VOLUME 44A, MAY 2013—2193
the nugget and TMAZ regions) with a much largerdeviation.
The SEM micrograph of the FSW welding zone afterT6 heat treatment is shown in Figure 11. The presenceof too many precipitates of relatively large size withsharp edges ends up with greater hardness and strengthin T6 heat-treated samples. These samples have lesselongation due to their precipitate sharp edges whichfacilitate failure initiation during tension.
The microhardness of the heat-treated samples forboth welding techniques is greater than the base metal.Precipitation hardening is the major reason for thisimprovement. T6 heat treatment results in reprecipita-tion of the elements dissolved during welding into thebase alloy.[22] T6 heat treatment results in recovery oftensile strengths of both FSW and GTAW specimens,but reduces their ductility as is seen in Table III.Strength enhancement and ductility reduction of thepost-weld heat-treated samples are because of theincreasing of the volume fraction of the hard precipi-tates as well as their buildup in the grain boundaries ofthe specimens. This has also been reported by previousinvestigators.[22] The threefold higher ductility of FSW-T6 samples with respect to the GTAW-T6 can beattributed to the difference in alloy compositions of thewelded regions.
Besides compositional change, dissolution of precip-itates and grain size increase during GTAW all result inlowering of the microhardness of the weld. The greater
ductility of FSW-T6 (near three times the ductility ofGTAW-T6) is also due to change of compositions in thewelded region as is observable from XRD peaks ofFigure 6. Structure–property correlations for FSW andGTAW methods are comparatively summarized inTable IV.
IV. PRACTICAL RECOMMENDATIONS
This investigation contributes to the best selection ofthe appropriate welding method of 6061 Al alloys. Thebenefits of using FSW, as a more advanced technique,for alleviation of the significant shortcomings of thetraditional GTAWmethod are described throughout thepaper. Achievable recoveries are as follows:
– Fusion region with defects in the traditional GTAWmethod improves with substitution of the solid stateFSW technique.
– FSW shows mechanical properties (hardness, tensilestrength, UTS, and elongation) superior to theGTAW.
– T6 heat treatment helps more enhancements of theUTS.
A recommendation can, thus, be made for the use ofFSW for 6061 aluminum alloys’ specific applicationslike in aerospace, while GTAW is more appropriate forordinary less expensive samples.
Table IV. Precipitates in Different Regions of FSW Sample Before and After T6
Sample Average size (lm) Distribution Morphology
FSWNugget 2.5 ± 0.9 homogenous small and relatively sphericalTMAZ 2.7 ± 1.6 homogenous small and large sphericalHAZ 4.0 ± 3.8 relatively homogenous relatively large and flake-likeBase 3.1 ± 2.0 non-homogenous relatively spherical
Fig. 11—SEM micrograph of FSW region after T6 heat treatment.
2194—VOLUME 44A, MAY 2013 METALLURGICAL AND MATERIALS TRANSACTIONS A
V. CONCLUSIONS
Aluminum 6061 alloy sheets of 13 mm thickness weresuccessfully joined together by (a) FSW and (b) GTAWprocedures. Microstructure, precipitate compositions,mechanical properties, and fracture behavior of theweldedsheets were scrutinized for exploring the effect of size andshape of the grains as well as formation and dissolution ofthe precipitates due to welding and heat treatment ondifferent regions and alternatively welded samples.
The GTAW resulted in formation of dendritic struc-ture adjacent to the center-line of the weld, while theFSW process reduced the grain size of the specimensclose to the center-line regions. The average size of thegrains decreased from ~90 lm in the base metal to~2 lm in the nugget zone, while grain diameter in theTMAZ was about 30 lm. The maximum value of thehardness in the center-line region of the FSW samplewas ~71 HV and the minimum hardness in the HAZ/TMAZ interface zone was ~32 HV. Correspondingvalues for the GTAW process were ~55 HV in the weldcenter-line and ~27 HV in the heat-affected zone.
Yield stress and ultimate tensile strength of the FSWand the GTAW specimens were less than the parentmetal, while their ductility reduced in all samples exceptin the longitudinal FSW specimens, which showedhigher elongation. Post-weld aging recovered a largeportion of the welding zone strength, but reduced theductility of the welded samples. Mechanical strengthand elongation of the FSW specimens were greater thanthe GTAW samples. The shallower and less uniformlydistributed dimples were the main cause of lowerelongation in the GTAW samples.
ACKNOWLEDGMENTS
The authors would like to thank Mr. S.H. Sonbolestan(MSc student in the Sharif University of Technology)and Mr. M.H. Tahmasebi (PhD candidate in theIsfahan University of Technology) for their help in run-ning the tests and the Sharif University of Technologyand Isfahan University of Technology for allowingresearch work to be conducted in their laboratories.
REFERENCES1. M. Jariyaboon, A.J. Davenport, R. Ambat, B.J. Connolly, S.W.
Williams, and D.A. Price: Corros. Sci., 2007, vol. 49, pp. 877–909.2. G. Mathers: The Welding of Aluminum and Its Alloys, Woodhead
