J 4 md effect of post weld heat treatments on microstructure and mechanical properties

Materials and Design 43 (2013) 134–143

Contents lists available at SciVerse ScienceDirect

Materials and Design

journal homepage: www.elsevier .com/locate /matdes

Effect of post weld heat treatments on microstructure and mechanical propertiesof friction stir welded joints of Al–Zn–Mg alloy AA7039

Chaitanya Sharma, Dheerendra Kumar Dwivedi ⇑, Pradeep KumarMechanical and Industrial Engineering Department, Indian Institute of Technology, Roorkee, Uttarkhand 247 667, India

a r t i c l e i n f o

Article history:Received 1 April 2012Accepted 13 June 2012Available online 21 June 2012

Keywords:Friction stir weldingPost weld heat treatmentMicrostructureAbnormal grain growthMechanical propertiesAluminum alloy

0261-3069/$ - see front matter � 2012 Elsevier Ltd. Ahttp://dx.doi.org/10.1016/j.matdes.2012.06.018

⇑ Corresponding author. Tel.: +91 1332 285826; faxE-mail address: [email protected] (D.K. Dwiv

a b s t r a c t

Friction stir welded joints of Al–Zn–Mg aluminum alloy AA7039 were given five different post weld heattreatments in order to investigate their effect on microstructure and mechanical properties. In general, allthe applied post weld heat treatments increased the size of a aluminum grains in all zones of friction stirweld joints. Abnormal grain growth was observed in entire zone modified by friction stir welding in caseof solution treated joints with and without artificial aging. The naturally aged joints offered the highestmechanical properties while solution treated joints offered lowest mechanical properties of the joints.Naturally aged joints yielded highest tensile strength (94.9%) and elongation (174.2%) efficiencies whileartificially aged joints yielded highest yield strength efficiency (96.7%). Further, post weld heat treatmentalso affected fracture location and mode of fracture.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Friction stir welding (FSW) is a solid state process in which tem-peratures are less (�425–480 �C) [1] than the melting temperatureof the aluminum alloys being joined but high enough to cause dis-solution/overaging of strengthening particles in WNZ, TMAZ andHAZ leading to the formation of a softened region with degradedmechanical properties generally in heat affected zones [1,2]. There-fore, as a consequence precipitation hardening aluminum alloysjoints generally have weld strength lower than that of the base me-tal [3,4].

A number of techniques are being considered currently for min-imizing the softening and to improve the properties of FSW weldjoints such as optimization of process parameters, post weld heattreatments, in process cooling using external coolants, and underwater or submerged FSW. Post weld heat treatment is a viable op-tion to restore the strength of the joints by modifying the size,shape and distribution of secondary strengthening particles. Lim-ited information is available in open literature on the effect of postweld heat treatment on microstructure and mechanical propertiesof friction stir welded heat treatable aluminum alloys [4–12].Mahoney et al. [4] found that low temperature aging treatment(at 121 �C for 24 h) could not restore the T651 strength and signif-icantly reduced ductility of friction stir weld joints of AA7075T651. Hassan et al. [5] reported that weld nugget zone grain struc-tures were inherently unstable in high strength aluminum alloy

ll rights reserved.

: +91 1332 285665.edi).

AA7010 and abnormal grain growth (AGG) occurred in entireWNZ due to the dissolution of soluble precipitates during solutiontreatment. Sullivan and Robson [6] applied a post weld heat treat-ment to friction stir welded AA7449 joints to obtain Alcan T7 tem-per and reported that post weld heat treatment overaged g0/gprecipitates in WNZ/TMAZ which reduced hardness in WNZ, TMAZand base metal while little change was observed in HAZ hardness.Feng et al. [7] observed AGG in friction stir weld joints of AA2219-O after post weld heat treatment and coarsening degree and tensileproperties increases with increasing solutionizing temperaturefrom 480 to 540 �C. Liu et al. [8] observed significant increase intensile strength and decrease in elongation at fracture after postweld aging (at 165 �C for 18 h) of friction stir welded joints ofAA2219-T6 aluminum alloy. Krishnan [9] reported that hardnessof friction stir weld joints of AA6061 increases with the increasein solutionizing temperature owing to more nucleation site be-cause of higher number of quenched in vacancies and increasedsize of precipitate free zones resulted in brittle fracture duringbend tests. Further, he observed abnormal grain growth after solu-tion treatment at 520–560 �C. Elangovan and Balasubramanian[10] investigated the influence of three different post weld heattreatments i.e. artificial aging, solution treatment (ST) and solutiontreatment followed by artificial aging (STA) on tensile properties ofthe AA6061 FSW joints. Artificial aging was found to be more ben-eficial which increases joint tensile strength by �11% followed bySTA, and ST. Backlund [11] observed increase in WNZ microhard-ness of AA6082-T6 joints after post weld aging treatment while re-verse trend was observed for HAZ microhardness. Sato and Kokawa[12] found that post weld artificial aging and a combination of

Fig. 1. Photograph of the modified vertical milling machine used for friction stirwelding.

C. Sharma et al. / Materials and Design 43 (2013) 134–143 135

solution treatment and artificial aging restored the strength of theAA6063-T5 joints to the level of base metal and elongation wasfound to increase with increasing strength after post weld heattreatments.

Literature survey [4–12] suggests that post weld heat treat-ments can efficiently modify the microstructure and improve thetensile properties of FSW joints. However, the published informa-tion on the effect of post weld heat treatments on the microstruc-ture and mechanical properties of the friction stir welded joints ofAA7039 is very scant and only one paper [13] is available publicallyin which solution treatment at 550 �C for 4 h, water quench fol-lowed by artificial aging at 190 �C for 6 h decreases yield strength,ultimate tensile strength but improves percentage elongation.AA7039 is a medium strength noncopper containing aluminum al-loys of 7000 family based on Al–Zn–Mg system, in which Mg com-bine with Zn and form hard intermetallic phases such as Zn2Mg toimprove mechanical properties. This alloy has gained overwhelm-ing acceptance as light weight strong construction material for fab-ricating transportable bridges, girders, armor plates and vehicle formilitary and railway transport systems, storage tank, bicycleframes and high speed trains, ship and boats by welding due totheir desirable properties such as high strength, low density, supe-rior cryogenic properties, excellent corrosion resistant togetherwith ability to gain strength by natural aging [14,15]. Therefore,this study aims to investigate the effect of various post weld heattreatments on microstructure and mechanical properties of frictionstir welded joints of Al–Zn–Mg alloy AA7039.

2. Material and experimental procedures

2.1. Base metals

Five millimeter thick extruded plates of Al–Zn–Mg alloyAA7039-T6 were used as the base metal for this experimentalinvestigation. The chemical composition and mechanical proper-ties of the base metal are given in Table 1.

2.2. Development of FSW joints

The plates were cut and machined to obtain plates of size300 mm long and 50 mm wide as well as to remove natural oxidelayer and other foreign particle from the faying surfaces. The plateswere arranged in square butt joint configuration and were heldfirmly in position, without any gap between faying surfaces ofthe plates using specially fabricated fixture. Friction stir weldingwas performed parallel to plate extrusion direction in single passon modified vertical milling machine as shown in Fig. 1 (HMT In-dia, 5 kW and 635 rpm), using previously optimized parame-ters[16]. The details of welding and tool geometry parametersused for the production of weld joints are enlisted in Table 2.FSW tool as shown in Fig. 2 was made of die steel and had flatshoulder with truncated conical pin having anticlockwise threadof 1 mm pitch.

2.3. Post weld heat treatments

In order to investigate the influence of postweld heat treat-ments on microstructure and mechanical properties of the FSW

Table 1Chemical composition and mechanical properties of AA 7039-T6.

Chemical composition (wt%) Mechanical properti

Al Zn Mg Mn Fe Si Cu Ultimate tensile stre

Bal. 4.69 2.37 0.68 0.69 0.31 0.05 414

joints, five different techniques of heat treatments including natu-ral aging (NA), artificial aging (AA), step aging, solution treatment(ST) and solution treatment and artificial aging (STA or T6) wereselected and joints were designated according to their abbrevia-tions. Natural aging involves room temperature aging of the FSWjoints for more than one year after FSW. Artificial aging was carriedout at 120 �C for a soaking period of 18 h in an electric oven. Stepaging involves pre aging at 100 �C for 8 h and final artificial aging at150 �C for 24 h. Solution heat treatment was carried out at solu-tionizing temperature of 480 �C for a soaking period of 30 min fol-lowed by quenching in water at room temperature. The T6treatment involves a combination of solution treatment at 480 �Cfor a soaking period of 1 h followed by quenching in water at roomtemperature and subsequent artificial aging at 165 �C for 6 h [14].To obtain mechanical properties of the joint in as welded (AW)condition, no post weld heat treatment was applied to specimenbefore tensile test and characterization was performed immedi-ately after FSW within one week.

2.4. Characterization techniques

Weld joints were inspected visually for voids, cracks and othersurface defects. Thereafter, FSW joints were subjected to threepoint bend test to reveal the presence of subsurface defects. FSWjoints passed 90� face and root bend test and no crack wasobserved on external surface subjected to bending. Sub size flattensile specimens were prepared according to ASTM E8 M guide-lines [17]. Computerized universal testing machine (H25 K-S,

es

ngth (MPa) Yield strength (MPa) Elongation (%) Microhardnes (Hv)

328 15.1 135

Table 2Tool dimension and welding parameters used for friction stir welding of AA7039-T6 aluminum alloy.

Tool dimensions Welding parameters

Shoulder diameter Pin diameter (mm) Pin length (mm) Welding speed (mm/min) Rotary speed (rpm) Tool tilt (�)

Top Bottom

16 6 4 4.7 75 635 2.5

All dimensions are in millimeters

Fig. 2. Photograph of the tool used for friction stir welding.

136 C. Sharma et al. / Materials and Design 43 (2013) 134–143

Hounsfield) was used for conducting tensile test and bend tests at across head speed of 1 mm/min. Three tensile tests were performedin each condition and average values were used for discussion. AVickers microhardness tester (VHM-002 V Walter UHL, Germany)was employed for measuring the hardness across the joint with aload of 100 g and 30 s dwell time.

FSW joints were polished and etched in Keller’s reagent (2 mlnitric acid, 4 ml hydrofluoric acid, and 94 ml water) for 90 s formacro and microstructural observation using a light optical micro-scope (Leica, Germany). Image analysis of weld micrographs wasdone using Image J 1.37v, image analyzing software to determineaverage grain size of a aluminum present in different zones of fric-tion stir welded joints and base metals. The fracture surfaces of thetensile tested specimens were investigated by a field emissionscanning electron microscope (FE-SEM) (FEI-Quanta 200�) to studythe mode of fracture.

3. Results

3.1. Microstructure

The macrostructure of transverse cross sections of weld jointsare shown in Fig. 3. The weld joints were free from the groove, tun-nel or other defects. FSW modified the microstructure of the basemetal and resulted in the formation of weld nugget zone (WNZ),thermo mechanically affected zone (TMAZ) and heat affected zone(HAZ) (Fig. 3). The weld joints showed trapezoidal WNZ whosedimensions were found similar to tool dimensions. Post weld heattreatments affected the macro and microstructure of FSW weldjoints remarkably. The solution treatment with and without artifi-cial aging resulted in excessive abnormal grain growth in the entireregion modified by FSW, was evident from macrographs shown inFig. 3(b and c). After solution treatment with and without artificial

Fig. 3. Transverse macrostructure of (a) as welded, (b) solution tre

aging different zones i.e. WNZ, TMAZ and HAZ could not bedifferentiated.

The microstructure of FSW joints is shown in Fig. 4. In general,WNZ showed fine grain structure than the base metal, TMAZ andHAZ. The extent of refinement of microstructure was found to de-crease with the increasing distance from the weld center. TMAZlies adjacent to WNZ showed bent and flattened grains becauseof distortion caused by tool stirring. The transition from WNZ toTMAZ is clearly distinguishable. The HAZ exhibited coarser grainstructure than the base metal. The micrographs of WNZs, TMAZsand HAZs of as welded and post weld heat treated weld jointsare shown in Figs. 5–7 respectively. Post weld heat treatments sig-nificantly affected the microstructure and morphology of a alumi-num and precipitates in all zones of FSW joints. Post weld agingnamely natural, artificial and step aging (Fig. 5b–d) coarsened aaluminum grains in WNZs up to about 1.5 times than that in aswelded condition (Table 3).

The different (natural, artificial and step aging) aging processesused in present study showed entirely different influence on mor-phology of strengthening precipitates. Strengthening precipitateswere few and randomly distributed in naturally aged condition(Fig. 5b) while step aging resulted in the reappearance of coarseand agglomerated precipitates along the grain boundaries(Fig. 5d). Solution treatments increased a aluminum grain sizefrom 13.1 lm in as welded condition to about 1588.6 lm and1291.1 lm in solution treatments without artificial aging conditionand solution treatments and artificially aged condition respec-tively. Moreover, no strengthening precipitates were seen aftersolution treatments as they have gone into the solution.

It was observed that width of deformed grain in TMAZs of nat-urally (22.1–108.6 lm) and artificially (18.3–68.5 lm) aged jointswere slightly more than base metal (Fig. 6b and c). Moreover, allthe post weld heat treated joints had width of deformed grainsin TMAZ larger than as welded joints (Fig. 6d). The aging treat-ments resulted in the coarsening of a aluminum grains in TMAZalso. However, the extent of coarsening was marginal. All the agedjoints invariably showed the presence of strengthening precipi-tates of larger size in more numbers than as welded joints(Fig. 6a–d). Natural aging resulted in uniform distribution ofstrengthening precipitates. However, few coarser precipitates werealso observed (Fig. 6b). Moreover, no strengthening precipitateswere seen after solution treatments with and without artificialaging (Fig. 6e and f).

The post weld aging significantly coarsened (1.9–3.1 folds) aaluminum grains in HAZs as compared to as welded joints(Fig. 7b–d). Extent of grain coarsening was maximum (�3.1 times)in artificially aged condition and minimum (�1.9 times) in natu-rally aged condition. Naturally aged joints (Fig. 7b) showed uni-formly distributed finer precipitates than as welded joints.Artificial and step aging resulted in spheroidization of few coarser

ated and (c) solution treated and artificially aged FSW joints.

WNZ

TMAZ

HAZ

BM

200 µm

Fig. 4. Evolution of microstructure in different zones of FSW joints.

C. Sharma et al. / Materials and Design 43 (2013) 134–143 137

strengthening precipitates (Fig. 7c and d). Strengthening precipi-tates in the HAZs of FSW joints in solution treated (ST and STA)conditions were coarsened to greater extent than that of otherPWHTs used in this work. (Fig. 7e and f).

E

S

qu

ec

iax

ond

ed

ar

re

y

A

cry

Mg

bn

sta

Zn2

or

lli

pr

ma

zed

eci

lly

αα

pit

gr

Al

ate

ow

Gr

s

n α

ain

Al

s

Grai

50

50

500

ns

µ

µ

µ

a

c

e

m

m

m

Fig. 5. Effect of post weld heat treatments on weld nugget grain structu

3.2. Microhardness

The post weld heat treatments radically affect the microhard-ness distribution and each heat treatment technique has entirelydifferent influence on the microhardness profile of the joint. Themicrohardness profiles for post weld heat treated and as weldedjoints are shown in Fig. 8 (measured at the mid of the transverseplane of the joints). The average microhardness of WNZs and HAZsof the post weld heat treated and as welded joints are summarizedin Table 3. The as welded joints showed softened region compris-ing WNZ and HAZ and the extent of softening was more in HAZand were in accordance to previous findings [18,19]. The averagemicrohardness in WNZ and HAZ of as welded joints were lower(115.3 and 107.6 Hv) than the base metal (135 Hv). The naturalaging and artificial aging treatment prominently hardened theweld joints in all zones as evident from their higher microhardnessthan the as welded joints. The measured values of microhardnessin WNZ (131.7 Hv) and HAZ (128.3 Hv) of naturally aged jointswere significantly higher than the as welded joints and was highestamong the post weld heat treated joints. The artificial aging wasnot as effective as the natural aging though it also increasedmicrohardness.

The effect of step aging was found different in different zones. Itretarded the microhardness in the WNZ and increased the same in

5

50

50 µ

00

µ

µm

µ

m

b

d

f

m

m

re of FSW joints (a) AW, (b) NA, (c) AA, (d) SA, (e) ST, and (f) STA.

50 µm

50 µm 50 µm

500 µm 500 µm

50 µm

a b

c d

e f

Secondary MgZn2 precipitates

Bent and flattened αα Al Grains

Spherical secondary MgZn2 precipitates

Abnormally grown α Al Grains

Fig. 6. Effect of post weld heat treatments on thermo mechanically affected zone grain structure of FSW joints (a) AW, (b) NA, (c) AA, (d) SA, (e) ST, and (f) STA.

138 C. Sharma et al. / Materials and Design 43 (2013) 134–143

the HAZ. The solution treatment with and without artificial agingboth drastically reduced the microhardness than the as weldedjoints across the entire transverse cross section of the joints. Thisin turn resulted in almost flat microhardness profile except inWNZ where a low microhardness region can be seen (Fig. 8). Thesolution treated joints showed lowest microhardness in all thezones of the joints.

The hardness maxima were located in the TMAZs on the retreat-ing side for the as welded, naturally aged, artificially aged and stepaged joints while no discernible hardness maxima was observed inFSW joints in ST and STA condition. The peak value of hardnessmaxima was 159 Hv for naturally aged joints. Post weld heat treat-ments prominently affected the locations of hardness minima andshifted same from advancing side HAZ in as welded condition toTMAZ–HAZ interface on advancing side of naturally and artificiallyaged condition. Moreover, the hardness minima for step aged, solu-tion treated joints with and without artificial aging were located inthe WNZ.

It is important to note that natural aging is most beneficial postweld aging treatment and improves joints microhardness signifi-cantly. The solution treated joints with and without artificial agingdeteriorated the joints microhardness sharply due to the formationof coarse (302.4 lm–3100 lm) grain structure.

3.3. Tensile properties

The Tensile properties of the FSW joints are dependent onmicrostructure which in turn depends on welding conditions, basemetal initial temper conditions and post weld treatments. The ten-sile properties of the joints in post weld heat treated and as weldedconditions are summarized in Table 4 and Fig. 9 shows correspond-ing engineering strain stress diagrams for the same. The post weldheat treatments significantly influenced the tensile properties ofFSW joints as evident from Table 4. The natural aging significantlyimproved the tensile properties of the FSW joints followed by arti-ficial aging than in as welded condition while others (SA, ST, andSTA) deteriorated the same. The tensile, yield strength and % elon-gation of FSW joints in naturally aged condition were 392.8 MPa,292.6 MPa and 21.4% respectively and that in as welded conditionwere 354.4, 217.3 MPa and 26.3%, respectively. The natural agingshowed 10.9%, 34.7% and 22.9% increase in tensile, yield strengthand % elongation of FSW joints than that in as welded condition.The solution treated joints offered worst tensile propertiesfollowed by solution treated and artificially aged joints. The ulti-mate tensile, yield strength and % elongation of FSW joints in solu-tion treated condition were 253.8 MPa, 99.6 MPa and 17.9%respectively which were significantly (28.3%, 54.2% and 16.4%

O

Sp

ver

he

Co

ag

ric

ar

ed

C

al s

sen

thi

oar

ec

ed

Co

ck

se

on

αα

ars

ro

ned

dar

Al

en

d li

M

y

Gr

ed

ke M

gZ

Mg

ain

α

Mg

n2

Zn

s

Al

gZ

pr

2 p

Gra

n2 p

eci

rec

ins

re

pit

ipi

cip

ate

tat

5

ita

s

es

5

0

50 µ

tes

0 µ

µm

m

a

c

e

m

Spheri

S

cal

eco

se

nd

con

Abn

ar

da

or

y

ry M

ma

MgZ

Mg

lly

n2

gZ

gr

pr

n2 p

ow

eci

re

n α

pit

cip

A

5

5

5

ate

ita

l G

0

0

0

s

tes

rai

µm

µm

µm

ns

b

d

f

Fig. 7. Effect of post weld heat treatments on heat affected zone grain structure of FSW joints (a) AW, (b) NA, (c) AA, (d) SA, (e) ST, and (f) STA.

Table 3Influence of post weld heat treatments on a aluminum grains in different zones of joints.

Material/joint condition Size of a Al grains (lm) Microhardness (Hv)

WNZ TMAZ HAZ WNZ HAZ Maxima Minima

Base metal 44.3 135As welded 13.1 37.3 75.7 115.3 107.6 139 89Natural aging 18.3 49.1 86.9 131.7 128.3 159 112Artificial aging 19.2 43.6 136.9 125.5 121.6 148 106Step aging 19.1 41.8 91.8 109.7 123 137 102Solution treatment 1588.6 – – 101.2 92.7 99 82Solution treatment & artificially aged 1291.1 – – 102.8 104.9 109 92

C. Sharma et al. / Materials and Design 43 (2013) 134–143 139

respectively) lower than that in as welded condition. FSW joints innaturally and artificially aged condition exhibited % elongationhigher than that in as welded condition. The step aged and solutiontreated FSW joints exhibited significantly lower % elongation thanthat in as welded condition.

Step aged joints showed 13.4% elongation and the same waslowest among the various post weld heat treatment techniquesused in this work. Further, post weld heat treatments reduced yieldstrength to a greater extent (3.3–66.5%) than the tensile strength(5.1–38.7%). Natural aging resulted in greatest improvement intensile strength (10.9%) and % elongation (22.9%) while artificial

aging offered greatest improvement in yield strength (45.8%) ascompared to as welded joint.

The joint efficiency is defined as the ratio of a tensile property ofFSW joint to that of unwelded base metal. The different joint effi-ciencies of the FSW joints in post weld heat treated and as weldedconditions can be seen in Fig. 10. The tensile, yield strength andelongation efficiencies of FSW joints were 85.6%, 66.3% and141.7% respectively in as welded condition. The maximum tensilestrength and elongation efficiencies of 94.9% and 174.2% were ob-tained in naturally aged condition while artificial aging offeredmaximum yield strength efficiency about 96.7%. The solution

-15 -10 -5 0 5 10 15

80

90

100

110

120

130

140

150

160AS

RS

HAZ

TMAZ

WNZ

Mic

roha

rdne

ss (H

v)

Distance from weld centre (mm)

AW NA AA SA ST STA

Fig. 8. Variation of microhardness across as welded and post weld heat treated FSWjoints.

0 2 4 6 8 10 12 14 16 18 20 22 24 26 280

50

100

150

200

250

300

350

400

450

Engi

neer

ing

Stre

ss (M

Pa)

Engineering Strain (%)

BM AW NA AA SA ST STA

Fig. 9. Engineering stress strain diagrams for as welded and post weld heat treatedFSW joints.

AW NA AA SA ST STA

60

80

100

120

140

160

180

200

Join

t effi

cien

cy (%

)

Joint conditions

TSE YSE EE

Fig. 10. Influence of post weld heat treatments on joint efficiencies.

140 C. Sharma et al. / Materials and Design 43 (2013) 134–143

treated joints showed lowest tensile (61.3%) and yield strength(33.5%) efficiencies while step aged joints offered minimum elon-gation efficiency (88.7%).

In general, during tensile test fracture of FSW joints took placefrom minimum hardness region of the joints. The location of min-imum hardness region is found to vary from HAZ, TMAZ–HAZinterface to WNZ depending upon the condition of the joints. Nat-urally and artificially aged joints fractured from minimum hard-ness region of TMAZ–HAZ interface (112, 106 Hv respectively) onadvancing side. It was observed that during tensile test crack initi-ated at TMAZ from the top side and propagate towards the HAZ un-til fracture of the joint. The tool foot prints i.e. semicircular marksleft by tool on the top side of the joints provide sites for easy initi-ation of microcracks. The step aged joints fractured from WNZ onretreating side while solution treated joints with and without arti-ficial aging fractured from WNZ on advancing side of the joints. It isworthwhile to mention that as welded joints fractured from HAZon advancing side. Thus it can be noted that the post weld heattreatment shifted the fracture locations from HAZ in as weldedcondition to TMAZ–HAZ interface or WNZ in post weld heat treatedcondition.

3.4. Fractography

SEM study of tensile fracture surfaces was done to investigatethe mode of fracture and to understand the effect of post weld heattreatment on the mode of failure of FSW joints (Fig. 11a–f). Neck-ing was observed for as welded joints and flat fracture surface wasinclined at �45� to the loading axis. Fracture surface of as weldedjoints invariably showed dimples of varying size and shapeuniformly distributed over the surface along with secondary cracks(Fig. 11a). Dimples were deep and fewer, contain secondary

Table 4Tensile properties of as welded and post weld heat treated FSW joints.

Joint condition Tensile strength(MPa)

Yield strength(MPa)

Elongatio(%)

Base metal 414 328 15.1As welded 354.4 217.3 21.4Natural aging 392.8 292.6 26.3Artificial aging 381.5 317.3 25.5Step aging 308.9 224.4 13.4Solution treatment 253.8 99.6 17.9Solution treatment & Artificially

Aged305.1 243 15.7

precipitates. It is believed that fracture of precipitates triggeredthe fracture. The mode of fracture was ductile as evident from dim-pled fracture surface. These observations are in agreement withhigh elongation efficiency of the joint (141.7%).

Naturally aged joints exhibited fracture surface covered withdeeper and larger dimples as well as few flat regions (Fig. 11b).Moreover, some deep pits with relatively featureless surface canbe seen on the fractured surface. The fracture mode was also duc-tile for naturally aged joints which showed highest elongation effi-ciency (174.2%). Similar behavior was observed for artificially agedjoints which exhibited dimpled fracture surface (Fig. 11c). The

n Tensile strengthefficiency (%)

Yield strengthefficiency (%)

Elongationefficiency (%)

– – –85.6 66.3 141.794.9 89.2 174.292.2 96.7 168.980.3 68.4 88.761.3 33.5 118.573.7 74.1 103.9

a b

c d

e f

Pores

Dimples

Secondary cracks

Dimples

Tearing ridges

Flat feature less regions

Layered ridges

Sheared dimples

Fig. 11. Fracture surface of FSW joints (a) AW, (b) NA, (c) AA, (d) SA, (e) ST, and (f) STA.

C. Sharma et al. / Materials and Design 43 (2013) 134–143 141

fracture surfaces of step aged joints were covered with fine dim-ples and large featureless regions with randomly scattered visiblepores (Fig. 11d). The mode of fracture was brittle or locally ductileand is in agreement with low (88.7%) elongation efficiency of theFSW joints in step aged condition. The fracture surfaces of solutiontreated joints showed shallow dimples of random shape and sizealong with feature less region (Fig. 11e). Moreover, presence offew elongated dimples suggested shearing of the dimples duringtensile test. The mix ductile–brittle mode of fracture was observedfor solution treated joints. The fracture surface of solution treatedand artificially aged joints were covered with layered ridges andunderwent mix ductile–brittle mode of fracture (Fig. 11f).

4. Discussion

Based on the results of experimental work it can be noted thatpost weld heat treatments had significant influence on the

microstructure and mechanical properties of FSW joints and follow-ing inferences can be made: (1) all the post weld aging processescoarsened the size of a aluminum grains in all zones of the jointsthan that of as welded FSW joints, (2) ST and STA resulted in verylarge sized a aluminum grains, (3) size and distribution of strength-ening precipitates was significantly different in different zones ofFSW joints for all the post weld heat treatments evaluated in thiswork, (4) strengthening precipitates were fine and uniformly dis-tributed in case of naturally aged joints compared to other joints.This may be the reason of higher microhardness and superior tensileproperties of naturally aged joints than in as welded FSW joints. Ful-ler et al. [20] reported that natural aging increased weld microhard-ness of friction stir welded joints of AA7075 alloy. Further, theyobserved that with the increasing natural aging time some of theprimary GP(I) zones are replaced by a larger volume fraction ofthe higher stability GP(II) zones and smaller volume fraction of fineg0 precipitates thereby producing strongest microstructure with

142 C. Sharma et al. / Materials and Design 43 (2013) 134–143

highest microhardness of WNZ. The presence of high concentrationvacancies and solute supersaturation promotes the rapid nucleationand growth of GP zones resulting in substantial gain in WNZ microh-ardness. The gain in HAZ microhardness can be attributed to in-crease in volume fraction of GP(II) zones and g0 precipitates incase of NA, AA and SA. Therefore, naturally aged FSW joints showedmechanical properties substantially higher than the as welded FSWjoints. During FSW weld nugget was solutionized owing to high(420–480 �C) process temperatures [1,4] which resulted in the dis-solution of MgZn2 precipitates in the WNZ. This may be the reasonof low microhardness of FSW joints in as welded condition. Whilethe softening of HAZ of FSW joints in as welded condition is due tocoarsening of MgZn2 precipitates as well as due to coarsening of aaluminum grain owing to severe reversion of MgZn2 precipitates[3,4,16]. Microhardness results are not in agreement with a alumi-num grain size as the same is also greatly affected by the distributionof strengthening precipitates (size, shape and volume) in case ofprecipitation hardening aluminum alloys [21]. The increasedmicrohardness of the TMAZ may be attributed to the increased vol-ume fraction of dissolved precipitates which favors the reprecipita-tion of the strengthening precipitates during natural aging. Thenaturally aged joints fractured from the TMAZ–HAZ interface onadvancing side. Sato and Kokawa [12] also found similar trend of re-sults for friction stir welded joints of AA6063 and reported that ten-sile strength, yield strength and % elongation all increased afterartificial aging at 175 �C for 12 h.

Both the solution treatments (ST and STA) resulted in the for-mation of few big a aluminum grains (302.4–3100 lm) in the en-tire region modified by FSW (refer Fig. 3b and c) which may beattributed to abnormal grain growth. Moreover, no strengtheningprecipitates were observed in the WNZ confirming their dissolu-tion into a aluminum matrix during solution treatments (referFig. 5e and f). Abnormal grain growth, dissolution of MgZn2 precip-itates into a aluminum matrix during solution treatment loweredthe microhardness in the WNZ while presence of coarsened precip-itates in relatively small numbers along with few thick rods likeprecipitates lowered the microhardness in the HAZ (refer Fig. 7eand f). This may be the reason for inferior tensile properties ofthe solution treated joints. The extent of loss of hardness was morein the WNZ and fracture during tensile testing occurred from WNZfor both the solution treated joints. Attallah and Salem [21] re-ported that abnormal grain growth deteriorated the tensilestrength of friction stir welded joints of AA2095. Further, Elango-van and Balasubramanian [10] found that solution treatment withand without artificial aging deteriorated the tensile properties offriction stir welded joints of AA6061-T6.

Abnormal grain growth was also reported by many authors incase of post weld heat treatment of friction stir weld joints of alu-minum alloys [5,9,21,22]. Friction stir welded joints of aluminumalloys invariably contains very fine equiaxed recrystallized grainswith high angle grain boundaries (>15�), have features of severelydeformed microstructure. Moreover, the size of these grains werefound to vary from top to bottom of the WNZ, from advancing toretreating side of the joints and bottom of the WNZ has been re-ported to contain high density of sub boundaries and less deformedunrecrystalized microstructure [5,22,23]. All these factors lead toinherently unstable microstructure of friction stir welded jointsin WNZ at high temperatures such as those encountered duringpost weld solution treatments. Abnormal grain growth occurredwhen the normal grain growth of the matrix grains suppressedand driving force is the desire to attain more stable state by reduc-ing grain boundary energy [5,22,24]. The thermodynamic forces,reduction in pinning forces, anisotropy in grain boundary energyand mobility are the main factors which provoke abnormal graingrowth [24] and can be minimized or eliminated effectively bymanipulating FSW process parameters; alloy chemistry [25].

5. Conclusion

Many post weld heat treatment techniques were applied to aswelded FSW joints of AA7039 alloy in order to improve theirmechanical properties by modifying the microstructure. Conclu-sions drawn from this experimental study are summarized below.

(1) Post weld natural, artificial and step aging coarsened a alu-minum grains in all the zones of FSW joints. The extent ofcoarsening was more in HAZs than WNZs. Natural agingresulted in homogeneous distribution of fine strengtheningprecipitates while artificial and step aging resulted inagglomerated spherical strengthening precipitates. Solutiontreatment with and without artificial aging resulted inabnormal grain growth across the entire region modifiedby FSW, strengthening precipitates were dissolved in WNZsand severely coarsened in HAZs.

(2) The natural aging improved the tensile properties (10.9%,34.7% and 22.9% increase in tensile, yield strength and %elongation) of the FSW joints followed by artificial agingthan in as welded condition while others (SA, ST, and STA)deteriorated the same. Therefore, natural aging is most ben-eficial post weld aging treatment followed by artificial agingto improve mechanical properties of FSW joints.

(3) Post weld heat treatments adversely affected yield strength(3.3–66.5%) more than the tensile strength (5.1–38.7%) offriction stir weld joints.

(4) The post weld heat treatment changed the fracture locationfrom HAZ in as welded condition to TMAZ–HAZ interface fornaturally and artificially aged joints, or WNZ for step agedand solution treated joints with and without artificial aging.

(5) The mode of fracture was ductile for naturally and artificiallyaged and as welded joints while step aged joints exhibitedbrittle mode of fracture. Solution treated joint with andwithout artificial aging showed ductile–brittle mode offracture.

Acknowledgements

The authors are thankful to Dr.N.K. Jain of Indian Institute ofTechnology, Indore for providing support in carrying out microh-ardness characterization of welded joints. Authors are also gratefulto DST, Govt. of India for providing financial support through grantno. SR/S3/MERC/005/2009 for carrying out this work under projectentitled ‘structural instability in friction stir welded joints of alu-minum alloys and their effect on mechanical properties’’. Mr. Cha-itanya Sharma, Research scholar kindly acknowledges the MHRD,Govt. of India for awarding fellowship.

References

[1] Rhodes CG, Mahoney MW, Bingel WH. Effects of FSW on microstructure of7075 Al. Scripta Mater 1997;36:69–75.

[2] Su JQ, Nelson TW, Mishra RS, Mahoney MW. Microstructural investigations offriction stir welded 7050–T651 aluminium. Acta Mater 2003;51(3):713–29.

[3] Jata KV, Sankaran KK, Ruschau J. Friction stir welding effects on microstructureand fatigue of Al alloy 7050–T7451. Metall Mater Trans A 2000;31:2181–92.

[4] Mahoney MW, Rhodes CG, Flintoff JG, Spurling RA, Bingel WH. Properties offriction stir welded 7075 T651 Al. Metall Mater Trans A 1998;29:1955–64.

[5] Hassan KhAA, Norma AF, Price DA, Prangnell PB. Stability of nugget zone grainstructures in high strength Al-alloy friction stir welds during solutiontreatment. Acta Mater 2003;51(7):1923–36.

[6] Sullivan A, Robson JD. Microstructural properties of friction stir welded andpost weld heated 7449 aluminum alloy thick plate. Mater Sci Eng A2008;478:351–60.

[7] Feng JC, Chen YC, Liu HJ. Effect of post weld heat treatment on microstructureand mechanical properties of friction stir welded joints of 2219-O aluminumalloy. Mater Sci Technol 2006;22(1):86–90.

C. Sharma et al. / Materials and Design 43 (2013) 134–143 143

[8] Liu HJ, Chen YC, Feng JC. Effect of heat treatment on tensile properties offriction stir welded joints of 2219–T6 aluminium alloy. Mater Sci Technol2006;22(2):237–41.

[9] Krishnan KN. The effect of post weld heat treatment on the properties of 6061friction stir welded joints. J Mater Sci 2002;37:473–80.

[10] Elangovan K, Balasubramaniam V. Influences of post-weld heat treatment ontensile properties of friction stir-welded AA6061 aluminum alloy joints. MaterCharacterization 2008;59:1168–77.

[11] Backlund J, Norlin A, Andersson A. Friction stir welding-weld properties andmanufacturing techniques. In: Proceedings of the 7th international conferenceon joints in aluminium. INALCO98; 1998. p. 171.

[12] Sato YS, Kokawa H. Distribution of tensile property and microstructure infriction stir weld of 6063 aluminum. Metall Mater Trans A 2001;32:3023–31.

[13] Singh RKR, Sharma Chaitanya, Dwivedi DK, Mehta NK, Kumar P. Themicrostructure and mechanical properties of friction stir welded Al–Zn–Mgalloy in as welded and heat treated conditions. Mater Des 2011;32:682–7.

[14] ASM handbook volume 2. Properties and selection: nonferrous alloys andspecial-purpose materials. Properties of wrought aluminum and aluminumAlloys. ASM international Material Park; 1992. p. 434.

[15] Lua ZJ, Evans WJ, Parkera JD, Birley S. Simulation of microstructure andliquation cracking in 7017 aluminium alloy. Mater Sci Eng A 1996;220(1–2):1–7.

[16] Sharma Chaitanya, Dwivedi DK, Kumar P. Effect of welding parameters onmicrostructure and mechanical properties of friction stir welded joints ofAA7039 aluminium alloy. Mater Des 2012;36:379–90.

[17] ASTM E8/E8M-09. Standard test methods for tension testing of metallicmaterials. Pennsylvania (USA): ASTM International; December 2009.

[18] Sharma Chaitanya, Dwivedi DK, Kumar P. Friction stir welding of Al–Zn–Mgalloy AA7039. TMS 141 annual meeting. Orlando, Florida, USA; March 11–15,2012. p. 503–7.

[19] Hassan Kh AA, Prangnell PB, Norman AF, Price DA, Williams SW. Effect ofwelding parameters on nugget zone microstructure & properties in highstrength Al alloy friction stir welds. Sci Technol Weld Join 2003;8(4):257–68.

[20] Fuller CB, Mahoney MW, Calabrese M, Micona L. Evolution of microstructureand mechanical properties in naturally aged 7050 and 7075 Al friction stirwelds. Mater Sci Eng A 2010;527(9):2233–40.

[21] Attallah MM, Salem HG. Friction stir welding parameters: a tool for controllingabnormal grain growth during subsequent heat treatment. Mater Sci Eng A2005;39:51–9.

[22] Sato YS, Watanabe H, Kokawa H. Grain growth phenomena in friction stirwelded 1100 Al during post weld heat treatment. Sci Technol Weld Join2007;12(4):318–23.

[23] Frigaard Ø, Grong Ø, Midling OT. A process model for friction stir weldingof age hardening aluminum alloys. Metall Mater Trans A 2001;32:1189–200.

[24] Charit I, Mishra RS, Mahoney MW. Multi-sheet structures in 7475 aluminumby friction stir welding in concert with post-weld super plastic forming.Scripta Mater 2002;47:631–6.

[25] Charit I, Mishra RS. Abnormal grain growth in friction stir processed alloys.Scripta Mater 2008;58:367–71.