Materials and Design 56 (2014) 185–192

Contents lists available at ScienceDirect

Materials and Design

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

Friction stir welding of dissimilar materials between AA6061

and AA7075 Al alloys effects of process parameters
J.F. Guo ⇑, H.C. Chen, C.N. Sun, G. Bi, Z. Sun, J. Wei
Singapore Institute of Manufacturing Technology (SIMTech), 71 Nanyang Drive, Singapore 638075, Singapore

a r t i c l e i n f o a b s t r a c t

Article history: Dissimilar AA6061 and AA7075 alloy have been friction stir welded with a variety of different process
Received 12 August 2013 parameters. In particular, the effects of materials position and welding speed on the material flow, micro-
Accepted 29 October 2013 structure, microhardness distribution and tensile property of the joints were investigated. It was revealed
Available online 9 November 2013
that the material mixing is much more effective when AA6061 alloy was located on the advancing side
and multiple vortexes centers formed vertically in the nugget. Three distinct zones with different extents
Keywords: of materials intercalations were identified and the formation mechanism of the three zones was then dis-
Dissimilar materials joining
cussed. Grain refinement was observed in all three layers across the nugget zone with smaller grains in
Friction stir welding
Microstructure
AA7075 Al layers. All the obtained joints fractured in the heat-affected zone on the AA6061 Al side during
Mechanical properties tensile testing, which corresponds very well to the minimum values in microhardness profiles. It was
Grain refinement found that the tensile strength of the dissimilar joints increases with decreasing heat input. The highest
joint strength was obtained when welding was conducted with highest welding speed and AA6061 Al
plates were fixed on the advancing side. To facilitate the interpretation, the temperature history profiles
in the HAZ and at zones close to TMAZ were also measured using thermocouple and simulated using a
three-dimensional computational model.
Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction components with relatively flat geometry at high productivity.

FSW is most suitable for welding light metal alloys, Al-based metal
Joining of dissimilar materials is very attractive for many appli- matrix composites (MMCs), and dissimilar metals [8,9]. The FSW of
cations as we can use the more costly one only where necessary. In light metals has been extensively studied [8] and the FSW of MMCs
fact, dissimilar joining could be frequently faced in many scenarios and/or dissimilar monolithic metals/alloys has also become an
including automotive, aerospace, electronics and shipbuilding important research topic in recent years [1,3]. High quality FSW
industries, where fusion welding simply is not appropriate due to joints have been obtained between different metals/alloys, MMCs
the large difference of physical and chemical properties between and monolithic alloys [1–3,10–15]. Among these studies, FSW of
the components to be joined [1–3]. Problems including porosity dissimilar materials between Al alloys has been focused by many
formation, solidification cracking, and chemical reaction may arise researchers and engineers from a variety of industries [1,3,16–
during fusion welding of dissimilar materials although sound 20]. It was reported that the temperatures are higher on the
welds may be obtained in some limited cases with special atten- advancing side when welding similar materials [8,21]. It is reason-
tions to the joint design and preparation, process parameters and able to believe that the location of the materials in FSW of dissim-
filler metals [4–6]. In contrast, friction stir welding (FSW) seems ilar alloys could lead to more severe temperature asymmetry. This
to be a very promising technique as it permits welding dissimilar asymmetry in temperature, stress and materials flow between the
materials in solid-state while avoiding the drawbacks of fusion advancing and retreating sides could significantly affect the dis-
welding. similar joint quality and properties, which mainly depends on
Friction stir welding, first invented at TWI in the early nineties the properties of the two materials and the welding parameters
[7], is a relatively new welding technique that has demonstrated such as tool rotation speed, travel speed and positions of the mate-
its great potential in joining materials that are traditionally consid- rials. It was reported that only a portion of material transported
ered to be unweldable or difficult to weld. As a variant process of from the leading edge undergoes chaotic flow and sound joints
friction welding, it provides the opportunity of welding without defects can only be obtained when the weld interface is
on the advancing side [16]. Better materials mixing can be
⇑ Corresponding author. Tel.: +65 6793 2251; fax: +65 6791 6377. achieved when using high rotation speeds in the compromise of re-
E-mail address: jfguo@simtech.a-star.edu.sg (J.F. Guo). duced surface integrity [17,18]. Park et al. [22] and Aval et al. [23]

0261-3069/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.matdes.2013.10.082
186 J.F. Guo et al. / Materials and Design 56 (2014) 185–192

revealed that more efficient materials mixing was obtained when

the softer AA5xxx aluminum alloy was in the advancing side and
the harder AA6xxx aluminum alloy was in the retreating side.
Slightly higher transverse tensile strengths of joints were obtained
with soft AA5xxx Al alloy on the advancing side [22]. However, the
longitudinal tensile strengths (parallel to FSW direction) of joints
produced with soft AA5xxx Al alloy on the advancing side showed
slightly lower values than the opposite conditions [22]. In another
study on FSW of AA6061 and A356 Al alloys, Lee et al. [24] showed
that higher longitudinal tensile properties of the weld (parallel to
FSW direction) were acquired when a relatively softer material
was fixed at the advancing side because the dominant material
in the weld zone mainly came from the harder retreating side. It Fig. 1. Schematic illustration showing thermocouple locations during temperature
seems the effect of relative position of materials is closely related measurements.

to the materials pair, no simple prediction can be made before

investigation. Sundaram et al. [20] investigated the effects of vari-
ous process parameters including tool pin profile, tool rotational Table 1. All plates were cut before welding into dimensions mea-
speed, and welding speed on mechanical properties of the dissim- suring 300 mm long and 50 mm wide (rolling direction). They were
ilar joints between AA2024 and AA5083 Al alloys. They reported carefully degreased with acetone and dried in air. Butt joints of
that dissimilar joints fabricated using tapered hexagon tool pin these plates were then produced along the longitudinal direction
profile have the highest tensile strength and tensile elongation, using a friction stir welding robot which is capable of generating
whereas the straight cylinder tool pin profile have the lowest ten- 12 kN maximum downward force. The welding tool has a shoulder
sile strength and tensile elongation [20]. However, the increase in with 15 mm diameter and a threaded conical probe with three flats
the tool rotational speed or welding speed led to increase in tensile and 5 mm diameter at probe base. The relative positions of both al-
strength initially and then decrease after reaching a maximum loys were varied to investigate its effects on the materials flow and
value. heat input. Three welding/travel speeds were used at a constant
It is clear that many progresses have been made in FSW of dis- tool rotation speed of 1200 rpm. The detailed welding parameters
similar materials but it has to be understood that study on this to- are listed in Table 2.
pic is far from enough and still in the feasibility stage. Up to now, The temperature profiles in the heat-affected zones (HAZ) were
systematic study on FSW of dissimilar materials between AA6xxx in situ measured using K-type thermocouples with a 0.25 mm
and AA7xxx series Al alloys has never been reported to the best diameter wire. As shown in Fig. 1, four holes (black points) with
of authors’ knowledge [25]. AA6xxx and AA7xxx Al alloys are 2.3 mm diameter were drilled to a depth of 3 mm on each plate.
two series of the most widely used structural materials in automo- Thermocouples were fixed into the holes with short bars of
tive, rail transportation and aerospace industries. Therefore, the 1.6 mm diameter 5356 Al filler wire. The filler wires were then
present study using AA6061 and AA7075 Al alloys as the base mechanically punched to improve the contact between the ther-
materials aims to investigate the effects of various process param- mocouples and the workpiece. The temperature history profiles
eters such as welding speed, positions of materials and the resulted at these locations were recorded at 0.5 s sampling interval using
heat input on the materials flow, microstructures and mechanical a digital data logger with eight channels. The temperature history
properties of the dissimilar joints. profiles in the nugget or zones very close to thermo-mechanically
affected zones (TMAZ) were simulated due to the high temperature
gradient in these zones and the difficulties in accurately position-
2. Experimental procedures ing thermocouples. A transient thermal analysis was carried out
by using commercial finite element software, ANSYS. In this ther-
The base materials used in the study are rolled plates of AA6061 mal analysis, the three-dimensional thermal solid element type,
and AA7075 Al alloy (both are 6.3 mm thick) in T6 temper condi- SOLID 70, was chosen for the model. In total, the model included
tion. The chemical compositions of these two alloys are listed in 38,505 nodes and 30,000 elements. Temperature-dependent

Table 1
Chemical composition of AA6061 and AA7075 alloys (wt.%) [26].

Alloy Zn Mg Cu Mn Si Fe Ti Cr Al
AA6061 0.25 max 0.8–1.2 0.15–0.4 0.15 max 0.4–0.8 0.7 max 0.15 max 0.04–0.35 Bal.
AA7075 5.1–6.1 2.1–2.9 1.2–2.0 0.3 max 0.4 max 0.5 max 0.2 max 0.18–0.28 Bal.

Table 2
Main FSW parameters used in the study.

Conditions Materials* Rotation Travel speed Downward Tilt angle (°)

speed (rpm) (mm/s) force (kN)
D1 7075–6061 1200 2 6.6 2.5
D2 7075–6061 1200 3 6.1 2.5
D3 7075–6061 1200 5 7.0 2.5
D4 6061–7075 1200 3 6.0 2.5
D5 6061–7075 1200 5 6.7 2.5
*
The materials on the left were located on advancing side during FSW.
 J.F. Guo et al. / Materials and Design 56 (2014) 185–192 187

material properties of both Al alloys [26], such as thermal conduc-

tivity and density were considered in the model.
The welded samples were transversely sectioned and polished
using conventional mechanical polishing method. Micro-etching
was carried out using Keller’s reagent for 20–150 s at ambient tem-
perature to reveal the grain structures of the welds. Microstruc-
tural characterization was performed using an optical microscope
(OM) and a scanning electron microscope (SEM) equipped with
an energy dispersive spectrometry (EDS). The Vickers hardness
was also measured across the centre of each joint with 100 gf load-
ing at intervals of about 1 mm. Tensile testing was conducted
according to the ASTM E8-04 standard at a test speed of
1 mm min1. Rectangular flat samples with a 50 mm gauge length
and a 12.5 mm width in the reduced section were used.

Fig. 2. Measured and simulated temperature history profiles at 10 mm away from

3. Results and discussions the joint line on AA6061 Al side for condition D5.

3.1. Temperature profiles

definitely influence the thermal profile experienced by the materi-
Friction stir welding is basically a thermo-mechanical process als across the joint. Thus, the temperature history profiles at 8 mm
where the materials experience a thermal cycle and mechanical away from the joint line on AA6061 side were simulated for several
mixing simultaneously. It is well established that the temperature typical welding conditions and are illustrated in Fig. 3. The 8 mm
history determines the microstructure evolution during FSW distance is selected because all joints failed approximately at this
[27,28]. For heat treatable Al alloys, in particular, the local thermal location on AA6061 side as will be discussed later in the article.
hysteresis could strongly influence the distribution of precipitates As shown in Fig. 3, at a constant tool rotation speed, much higher
and make them coarsen, dissolve and re-precipitate. However, due peak temperature is obtained at 2 mm/s travel speed compared to
to the high temperature gradient in the regions close to nugget that of 5 mm/s travel speed, which can be readily understood be-
zone and the difficulties in accurately positioning thermocouples, cause the total friction heat generated during FSW is inversely pro-
it is very difficult to measure the temperature profile during FSW portional to the travel speed. At the same rotation and travel speed,
in the nugget, TMAZ or locations very close to TMAZ where usually slightly higher peak temperature in AA6061 is obtained when
the joints failed during tensile testing. Therefore, only the temper- AA7075 Al was located on the advancing side. This is due to that
ature history profile in HAZ was measured and a computational the material transport is more intense on the advancing side and
model proposed by Riahi et al. [29] was used to simulate the tem- AA7075 alloy has much higher flow stress/yield strength than
perature profiles in these zones. In this model, the local heat (Q) AA6061 alloy does [30].
generated by friction between the tool and the surface of welding
materials was applied as in the following equation: 3.2. Materials flow & microstructure

4p2 Fig. 4 shows macro-views of the cross sections of the friction

Q¼ lNpðr3S þ 3r2P hÞ ð1Þ stir welded joints produced under different conditions. The
3
AA6061 alloy features as darker colors, while AA7075 alloy fea-
where l is the friction coefficient between welding tool and Al tures as light colors because of their different etching response to
plates, N is tool rotation speed (1200 rpm), p is local pressure ap- the Keller’s reagent. It is evident that sound joints of AA6061–
plied by welding tool on the top surface of welding materials (N), AA7075 Al alloys were achieved under all the investigated welding
rS is shoulder radius (7.5 mm), rP is pin radius (2.5 mm) and h is conditions. The material mixing seems to be much more effective
the pin length (6.1 mm). when AA6061 alloy was located on the advancing side, especially
The Al plates used in the study are 300 mm long, 50 mm wide with higher welding speed as seen in Fig. 4b and d. This is mainly
and 6.3 mm thick. Based on the fixture design and lab environment because the higher flow stress of harder AA7075 alloy [30], when
(23 °C), a heat flux of 350 W/m2 K from both top and bottom sur- located on the advancing side, makes it very difficult for AA6061
faces were assumed while a heat flux of 35 W/m2 K was considered
for side surfaces.
Fig. 2 shows the temperature history profiles obtained from
both experiments and simulation at 10 mm away from the joint
line on AA6061 Al side for conditions D5. All simulated tempera-
ture profiles shown below are calculated at 3 mm from the top sur-
face of Al plates since the ones measured by thermocouples are
obtained from the same locations. As seen in Fig. 2, a very close
match between the simulated and experimentally measured tem-
perature profiles are observed, which indicates the two-dimen-
sional model built can be used to effectively predict the thermal
cycles experiences in other zones where the temperature history
profile is difficult/impossible to measure accurately. It was re-
ported that the temperature profile has an asymmetric distribution
during FSW [21], especially in welding of dissimilar materials be-
cause the materials on advancing and retreating sides have differ-
ent flow patterns and the two alloys have very different flow stress.
The relative position of AA6061 and AA7075 alloys would Fig. 3. Simulated temperature history profiles at 8 mm away from the joint line.
188 J.F. Guo et al. / Materials and Design 56 (2014) 185–192

Fig. 4. Macro-views of the cross sections of joints produced under different conditions.

Fig. 5. Images show the complex materials flow patterns (onion rings) on the top advancing and retreating sides: (a and b), and the multiple vortexes in the nugget center: (c
and d).

alloy to penetrate into the nugget. In FSW, the materials on retreat- authors’ knowledge. The formation mechanism of the multiple vor-
ing side are much less intensely transported around the tool before texes is believed to be related to both of the flats numbers and
deposited in the wake of the weld when compared with the ones threads pitch of the tool. The complex materials flow patterns
on advancing side. This is in accordance with the results obtained (onion rings) on the top advancing and retreating sides, and the
by Park et al. [22] in FSW of dissimilar AA5051-H32 and AA6061- multiple vortexes in the nugget center are illustrated in Fig. 5.
T6 alloys. They concluded that the materials were more properly Three distinct sub-layers can be observed in the onion rings of
mixed when the AA5052-H32 aluminum alloy was in the advanc- the dissimilar welds of AA6061–AA7075 Al alloys. These sub-
ing side than the case of the AA6061-T6 aluminum alloy in the layers were then further identified as: (a) 6061 alloy sub-layer
advancing side. (spectrum 1), (b) AA7075 alloy sub-layer (spectrum 2), and (c)
As shown in Fig. 4, onion ring typically seen in FSW was clearly mixed sub-layer of the two alloys (spectrum 3) by EDS under
observed in all the joints thank to the different etching responses SEM (Fig. 6). The formations of AA6061 and AA7075 alloy sub-
of the two alloys. It is interesting to notice that a series of vortex layers are quite straight forward; while the formation of the mixed
centers could be seen vertically in the nugget center especially sub-layer could be attributed to that the plasticized materials con-
when the AA6061 alloy was located on the advancing side. The tained in the spaces adjacent to the flats may have experienced
multiple vortexes feature observed in the present study is quite intense extrusion turbulences and have enough time to be well
unique since the onion ring layers reported in literature usually be- mixed before finally deposited to the wake of the weld. The
long to only one single vortex center [22,25,31]. Such multiple vor- thicknesses of these onion ring sub-layers were measured at
texes compilation has never been reported before to the best of around 30–100 lm, which are much smaller compared to the
 J.F. Guo et al. / Materials and Design 56 (2014) 185–192 189

Fig. 6. SEM image and EDS spectrums showing the chemical compositions of the three distinct layers formed in FSW.

Fig. 7. Optical micrographs showing the grain structures of different alloy sub-layers in the nugget zones of the dissimilar welds and the two base metals.

advance per revolution (100–250 lm) defined as traveling distance only reveal the material flow at one location in the weld. In
of the welding tool per revolution [8]. The much thinner onion ring contrast, the use of dissimilar materials may illustrate the material
sub-layers are resulted from that three materials flow waves flow pattern across the whole weld. Nonetheless, the material flow
instead of one have formed during FSW because the tool used in pattern in dissimilar materials joining might be slightly different
the study has a probe with three flats. Thus, the thicknesses of from that in FSW of similar materials. The above results also indi-
the onion ring sub-layers are roughly one-third of the advance cate that use of FSW tools with multiple flats and threads can be an
per revolution. effective way to achieve seamless bonding during dissimilar mate-
It is evident that use of dissimilar alloys/materials can be a very rials joining. However, it might be difficult to add such features for
effective alternative method to investigate the material flow tools when joining material pairs involving hard alloys such as
behavior during FSW. Other methods including inserted copper steels and super nickel alloys.
foil, plated surfaces, and composite markers have been used to Fig. 7 shows the grain structures of different alloy sub-layers in
study the material flow during FSW [8]. However, those methods the nugget zones of the dissimilar welds and the two base metals.
190 J.F. Guo et al. / Materials and Design 56 (2014) 185–192

Obviously, both AA6061 and AA7075 alloys have experienced

dynamic recrystallization. For AA7075 alloy (Fig. 7a–c), the grain
structure has been refined from large mm-scale elongated ones
to very fine equiaxed ones with less than 6 lm equivalent diame-
ters. The grain size decreased from 6 lm to 3 lm when the
welding speed increased from 3 mm/s to 5 mm/s. This can be
attributed to that slow welding speed resulted in more heat input
and longer time of grain growth for the materials in the nugget
zones. As shown in Fig. 3, the time span above 200 °C at 2 mm/s
is much longer than at 5 mm/s. For AA6061 alloy, the equiaxed
grain structure has been refined from around 50 lm to less than
10 lm. Similar as in AA7075 sub-layer, the grain size decreased
from 10 lm to 5 lm when the welding speed increased from
3 mm/s to 5 mm/s. It has to be noted that the grain size of
AA7075 alloy sub-layer is much smaller than that of AA6061
sub-layer in the same weld (Fig. 7b vs. Fig. 7e, Fig. 7c vs. Fig. 7f).
The grain size difference was also noted by Srinivasan et al. in a
corrosion study on FSW of dissimilar AA7075 and AA6056 alloys
[25]. They attributed this difference to the smaller initial grain size Fig. 8. Vicker’s microhardness profiles of the cross-section of dissimilar joints.

of the AA7075 alloy used in their study. However, the grain size
difference cannot be explained by their initial grain sizes in the
the arrows marked in Fig. 8, the hardness minima are observed
present study since the AA7075 alloy has much larger initial grain
in the HAZ on the AA6061 side regardless of the relative materials
size. At such micro-scale, the temperature history at adjacent
position or the applied process parameters. In fact, all joints failed
AA6061 and AA7075 sub-layers should be almost identical. There-
on the AA6061 side in HAZ regions very close to the TMAZ in ten-
fore, such difference is probably due to that AA7075 alloy has
sile testing, where the minimum hardness is located. As described
much higher alloying elements content than AA6061 alloy does
in the previous part (Fig. 3), the peak temperatures at 8 mm away
(see Table 1: 2% vs. 10%). All second phase particles and inter-
from the joint line (approximately where the hardness minimum is
metallics could provide much more nucleation sites and retard
located) reached temperatures higher than 365 °C, which are still
grain growth during recrystallization process, and thus result in fi-
lower than the solution heat treatment temperature of AA6061 al-
ner grain structure. It was reported that second phase particles
loy (530 °C) [38]. The decrease of microhardness in HAZ is due to
could provide nuclei during annealing and recrystallization pro-
the disappearance of G.P. zones and coarsening of strengthening
cesses [32]. In 7xxx series of Al alloys, zirconium or chromium is
precipitates. Thus, less severe precipitate coarsening could occur
added as grain refiner and to retard recrystallization during heat
in the HAZ of the joints produced with lower heat input and hence
treatment. Thus, all second phase particle in AA7075 alloy includ-
the hardness profile of the joints produced with the lowest heat in-
ing N-MgZn2 precipitates, Al3Zr dispersoids and Al–Cu–Fe constitu-
put (condition D5) has the highest minimum value as illustrated in
ent phase particles will provide nucleation sites (larger than
Fig. 8.
0.5 lm) [33] and pinning effect for recrystallization. In contrast,
the content of such second phase particles is much lower in
AA6061 alloy compared to that in AA7075 alloy. Detailed discus- 3.4. Tensile properties
sion on the particle-stimulated nucleation and particle pinning ef-
fect is beyond the scope of the present study and can be found Table 3 shows the tensile properties of the base AA6061 alloy
elsewhere in literature [33–36]. and the dissimilar joints between AA6061and AA7075 alloys. Only
the tensile properties of AA6061 alloy are shown in the table
3.3. Microhardness because all joints failed on the AA6061 side at positions in HAZ re-
gions where the hardness minima are located. As reported previ-
Fig. 8 shows the Vicker’s microhardness profiles of the cross- ously in dissimilar FSW of other Al alloys, the friction stir welded
section of the dissimilar joints produced under different conditions joints usually fractured at locations in HAZ on the weaker material
(D2–D5). In general, both AA6061 and AA7075 alloys have exhib- side [18,31,39]. However, those investigations did not study the ef-
ited certain microhardness decrease in the weld compared to their fect of material position on the tensile properties. The failure loca-
corresponding base metals (both materials are in T6 temper condi- tions in HAZ also indicate seamless bonding has been achieved
tion). This is mainly because of the coarsening, dissolution and rep- between dissimilar AA6061and AA7075 alloys under all investi-
recipitation of strengthening precipitates caused by FSW thermal gated welding conditions. As shown in Table 3, all the joints exhibit
cycle [27,37] although grain structure refinement may also have very good tensile properties with ultimate tensile strength (UTS)
some minor contributions. In the HAZ, the slightly lower hardness higher than 215 MPa and elongation higher than 6%. The highest
is due to the disappearance of Guinier–Preston (G.P.) zones and UTS value is 32% higher than required for FSW of AA6061 alloy at
coarsening of strengthening precipitates, which is similar to an T6 condition in the AWS standard (186 MPa): AWS D17.3/
over-ageing process. In the TMAZ, more severe coarsening and pos- D17.3M:200X [40]. The UTS increases with the increase of welding
sibly complete dissolution of precipitates occurred due to a similar speed when the location of AA6061/AA7075 alloy is fixed. At a con-
effect of the solution treatment. In the nugget, some reprecipita- stant welding speed, the UTS are slightly higher when AA6061 alloy
tion after complete dissolution may have taken place since the was located on the advancing side. The increases of UTS are mainly
material experienced higher temperature in this region. The transi- because FSW at higher welding speed or with AA6061 on the
tion of microhardness in the nugget from AA6061 to AA7075 is advancing side induced less heat input and thus less severe precip-
more gradual (conditions D2, D4 and D5) when AA6061 alloy itate coarsening in the HAZ. As illustrated in Fig. 3, the peak temper-
was located on the advancing side or lower welding speed is used. ature at 8 mm away from the joint line decreases from 469 °C to
This indicates again more effective material mixing is obtained un- 378 °C with the increase of welding speed from 2 mm/s to 5 mm/s
der these welding conditions (Fig. 4a, c and d). In all the cases, as when the AA6061 alloy is fixed on the retreating side. At the same
 J.F. Guo et al. / Materials and Design 56 (2014) 185–192 191

Table 3
Tensile properties of the dissimilar joints between AA6061and AA7075 alloys.

Conditions Material on Adv. side UTS (MPa) YS (MPa) e (%) Joint efficiency* Failure location
6061_T6 – 310 276 12 – –
D1 7075 215 ± 2 160 ± 5 7±1 69 HAZ_6061
D2 7075 221 ± 4 152 ± 3 7±0 71 HAZ_6061
D3 7075 235 ± 3 168 ± 3 6±1 76 HAZ_6061
D4 6061 228 ± 2 159 ± 3 8±1 74 HAZ_6061
D5 6061 245 ± 3 177 ± 6 6±1 79 HAZ_6061
*
Joint efficiency defined as the ratio of the UTS of the joint and the base metal at T6 condition.

Fig. 9. Fractured surfaces of the tensile tested specimens under SEM for conditions D2 and D5. (a and c): condition D1, (b and d): condition D5.

welding speed at 5 mm/s, the peak temperature decreases slightly ups. Consequently, the microvoids grow, coalesce, and eventually
from 378 °C to 365 °C when the location of AA6061 alloy was form a continuous fracture surface as the strain increases during
changed from the advancing side to the retreating side. A compar- tensile testing [41]. Therefore, the reason for the shallower dimples
ison of the tensile data and the microhardness profiles reveals the in condition D5 is probably because more severe precipitate coars-
tendency of UTS corresponds very well with the tendency of hard- ening occurred during FSW and thus less closely spaced b-Mg2Si
ness minima, which demonstrates the precipitate evolution caused phase particles has formed in the HAZ of the joints produced under
by FSW thermal cycle determines both the microhardness distribu- condition D1. Second phase particles including incoherent b-Mg2Si
tion and tensile property of the joints. phase and various Al–Fe–Si intermetallics in AA6061 alloy could
effectively provide nucleation sites for microvoids during fracture
process. Fig. 9c and d illustrate some examples of such nucleation
3.5. Fractured surfaces sites at the bottom of dimples or at the top of ridges. However,
quantitative measurement of the amount of such second phase
Fig. 9 shows the fractured surfaces of the tensile tested speci- particles is unrealistic since most of the particles are usually hid-
mens under SEM for conditions D1 and D5. These two conditions den at the bottom of the dimples and have similar color as the sur-
were selected for fractographic analysis because heat input is the rounding Al matrix on fractured surfaces under SEM.
main factor determining the tensile properties of joints and the dif-
ference of heat input between these two conditions is the most
marginal. As seen in Fig. 9a and b, the fractured surfaces are char-
acterized as a large amount of equiaxed dimples with different 4. Conclusions
size. Compared to condition D1, the fractured surface of the joints
produced under condition D5 shows shallower dimples. As we Dissimilar AA6061 and AA7075 alloys have been friction stir
know, in such dimple rupture mode, overload is the principal cause welded with a variety of different process parameters. At a con-
of fracture and the failure is governed by coalescence of microv- stant tool rotation speed of 1200 rpm, the effects of materials posi-
oids. The microvoids may nucleate at regions adjacent to second tion and welding speed on materials flow, microstructures,
phase particles, inclusions, grain boundaries, and dislocation pile- microhardness distributions and tensile properties of the joints
192 J.F. Guo et al. / Materials and Design 56 (2014) 185–192

were investigated. Based on the above results and discussion, the [12] Sato YS, Park SHC, Michiuchi M, Kokawa H. Constitutional liquation during
dissimilar friction stir welding of Al and Mg alloys. Scripta Mater
following conclusions can be drawn accordingly.
2004;50:1233–6.
[13] Sharifitabar M, Nami H. Microstructures of dissimilar friction stir welded
(1) The material mixing is much more effective when AA6061 joints between 2024-T4 aluminium alloy and Al/Mg2Si metal matrix cast
alloy was located on the advancing side and multiple vor- composite. Compos: Part B 2001;42:2004–12.
[14] Pan C, Hu L, Li Z, North TH. Microstructural features of dissimilar MMC/AISI
texes centers formed vertically in the nugget center. 304 stainless steel friction joints. J Mater Sci 1996;31:3667–74.
(2) The onion ring observed in dissimilar welds of AA6061– [15] Rodrigues DM, Loureiro A, Leitao C, Leal RM, Chaparro BM, Vilaça P. Influence
AA7075 Al alloys consist of three distinct sub-layers: (a) of friction stir welding parameters on the microstructural and mechanical
properties of AA 6016-T4 thin welds. Mater Des 2009;30:1913–21.
AA6061 alloy sub-layer, (b) AA7075 alloy sub-layer, and (c) [16] Kumar K, Kailas SV. Positional dependence of material flow in friction stir
mixed sub-layer of the two alloys. The thicknesses of these welding: analysis of joint line remnant and its relevance to dissimilar metal
onion ring sub-layers were measured at around 30– welding. Sci Technol Weld Joining 2010;15:305–11.
[17] Steuwer A, Peel MJ, Withers PJ. Dissimilar friction stir welds in AA5083–
100 lm which is roughly one-third of the advance per AA6082: the effect of process parameters on residual stress. Mater Sci Eng A
revolution. 2006;441:187–96.
(3) Both AA6061 and AA7075 alloys have experienced dynamic [18] Silva AAMd, Arruti E, Janeiro G, Aldanondo E, Alvarez P, Echeverria A. Material
flow and mechanical behaviour of dissimilar AA2024-T3 and AA7075-T6
recrystallization and the grain size in both alloys decreases aluminium alloys friction stir welds. Mater Des 2011;32:2021–7.
significantly with the increase of welding speed. The grain [19] Jun T-S, Dragnevski K, Korsunsky AM. Microstructure, residual strain, and
size of AA7075 alloy sub-layer is much smaller than that of eigenstrain analysis of dissimilar friction stir welds. Mater Des
2010;31:S121–5.
AA6061 sub-layer in the same weld.
[20] Sundaram NS, Murugan N. Tensile behavior of dissimilar friction stir welded
(4) Both AA6061 and AA7075 alloys have exhibited certain joints of aluminium alloys. Mater Des 2010;31:4184–93.
microhardness decrease in the weld compared to their cor- [21] Nandan R, DebRoy T, Bhadeshia HKDH. Recent advances in friction-stir
responding base metals (both materials are in T6 temper welding – process, weldment structure and properties. Prog Mater Sci
2008;53:980–1023.
condition). The minimum hardness values are observed in [22] Park S-K, Hong S-T, Park J-H, Park K-Y, Kwon Y-J, Son H-J. Effect of material
the HAZ on the AA6061 side regardless of the relative mate- locations on properties of friction stir welding joints of dissimilar aluminium
rials position or the applied process parameters. The hard- alloys. Sci Technol Weld Joining 2010;15:331–6.
[23] Aval HJ, Serajzadeh S, Kokabi AH. Thermo-mechanical and microstructural
ness profile of the joints produced with the lowest heat issues in dissimilar friction stir welding of AA5086–AA6061. J Mater Sci
input (condition D5) has the highest minimum value. 2011;46:3258–68.
(5) All the joints failed at positions in HAZ on the AA6061 side [24] Lee WB, Yeon YM, Jung SB. The mechanical properties related to the dominant
microstructure in the weld zone of dissimilar formed Al alloy joints by friction
where the minimum hardness is located and exhibited very stir welding. J Mater Sci 2003;38:4183–91.
good tensile strengths and ductility. The UTS of joints [25] Srinivasan PB, Dietzel W, Zettler R, Santos JFd, Sivan V. Effects of inhibitors on
increases with the decrease of heat input induced by friction corrosion behaviour of dissimilar aluminium alloy friction stir weldment.
Corros Eng Sci Technol 2007;42:161–7.
and corresponds very well to the minima in the microhard- [26] ASM Handbook Committee. Properties and selection: nonferrous alloys and
ness profiles. The highest UTS achieved in condition D5 special-purpose materials. Materials Park, OH: ASM International; 1991.
reached 245 MPa which is 32% higher than required in the [27] Sato YS, Kokawa H, Enomoto M, Jogan S. Microstructural evolution of 6063
aluminium during friction-stir welding. Metall Mater Trans A
AWS standard on FSW.
1999;30:2429–37.
(6) The fractured surfaces of tensile tested specimens are char- [28] Frigaard Ø, Grong Ø, Midling OT. A process model for friction stir welding of
acterized as a large amount of equiaxed dimples with differ- age hardening aluminum alloys. Metall Mater Trans A 2001;32A:1189–200.
ent size. Shallower dimples were observed in the fractured [29] Riahi M, Nazari H. Analysis of transient temperature and residual thermal
stresses in friction stir welding of aluminum alloy 6061-T6 via numerical
surface of the joints produced under lower heat input. Sec- simulation. Int J Adv Manuf Technol 2011:143–52.
ond phase particles including incoherent b-Mg2Si and vari- [30] Akeret R. Extrusion of semifinished products in aluminum alloys. In: Bauser M,
ous Al–Fe–Si intermetallics could effectively provide Sauer G, Siegert K, editors. Extrusion. Materials Park, OH 44073-0002: ASM
International; 2006. p. 207–26.
nucleation sites for microvoids during fracture process. [31] Amancio-Filho ST, Sheikhi S, Santos JFD, Bolfarini C. Preliminary study on the
microstructure and mechanical properties of dissimilar friction stir welds in
References aircraft aluminium alloys 2024-T351 and 6056-T4. J Mater Process Technol
2008;206:132–42.
[32] Hutchinson B. Nucleation of recrystallization. Scr Metall Mater
[1] DebRoy T, Bhadeshia H. Friction stir welding of dissimilar alloys – a
1992;27:1471–5.
perspective. Sci Technol Weld Joining 2010;15:266–70.
[33] McNelley TR, Swaminathan S, Su JQ. Recrystallization mechanisms during
[2] Guo J, Gougeon P, Chen X-G. Microstructure evolution and mechanical
friction stir welding/processing of aluminum alloys. Scripta Mater
properties of dissimilar friction stir welded joints between AA1100-B4C
2008;58:349–54.
MMC and AA6063 alloy. Mater Sci Eng A 2012;553:149–56.
[34] Humphreys FJ. Local lattice rotations at 2nd phase particle in deformed metals.
[3] Murr LE. A review of FSW research on dissimilar metal and alloy systems. J
Acta Mater 1979;27:1801–14.
Mater Eng Perform 2010;19:1071–89.
[35] Humphreys FJ, Hatherly M. Recrystallization and related annealing
[4] Ellis MBD. Joining of aluminium based metal matrix composites. Int Mater Rev
phenomena. 2nd ed. Elsevier Ltd.; 2004.
1997;41:41–58.
[36] Hassan KhAA, Norman AF, Price DA, Prangnell PB. Stability of nugget zone
[5] Çam G, Koçak M. Progress in joining of advanced materials. Int Mater Rev
grain structures in high strength Al alloy friction stir welds during solution
1998;43:1–39.
treatment. Acta Mater 2003;51:1923–36.
[6] Guo J, Gougeon P, Chen X-G. Study on laser welding of AA1100-16 vol.% B4C
[37] Su J-Q, Nelson TW, Mishra R, Mahoney M. Microstructural investigation of
metal-matrix composites. Compos: Part B 2012;43:2400–8.
friction stir welded 7050-T651 aluminium. Acta Mater 2003;51:713–29.
[7] Thomas WM, Nicholas ED, Needham JC, Murch MG, Templesmith P, Dawes CJ.
[38] ASM Handbook Committee. Heat treating. Materials Park, OH: ASM
Friction-stir butt welding. G.B. Patent 9125978.8, UK: 1991.
International; 1991.
[8] Mishra RS, Ma ZY. Friction stir welding and processing. Mater Sci Eng R
[39] Koilraj M, Sundareswaran V, Vijayan S, Rao SRK. Friction stir welding of
2005;50:1–78.
dissimilar aluminum alloys AA2219 to AA5083 – Optimization of process
[9] Threadgill PL, Leonard AJ, Shercliff HR, Withers PJ. Friction stir welding of
parameters using Taguchi technique. Mater Des 2012;42:1–7.
aluminium alloys. Int Mater Rev 2009;54:49–93.
[40] AWS D17 Committee on Welding in the Aircraft and Aerospace Industries.
[10] Chen T. Process parameters study on FSW joint of dissimilar metals for
Specification for friction stir welding of aluminum alloys for aerospace
aluminum-steel. J Mater Sci 2009;44:2573–80.
hardware. Miami, Florida: American Welding Society; 2010. p. 60.
[11] Mofid MA, Abdollah-zadeh A, Ghaini FM. The effect of water cooling during
[41] Kerlins V. Modes of fracture. In: ASM Handbook Committee Editors.
dissimilar friction stir welding of Al alloy to Mg alloy. Mater Des
Fractography. Materials Park, OH: ASM, International; 1991.
2012;36:161–7.