Microstructure and Mechanical Properties of Friction Stir Welded AA2024-T3 Aluminum Alloy

7/27/2019 Microstructure and Mechanical Properties of Friction Stir Welded AA2024-T3 Aluminum Alloy

http://slidepdf.com/reader/full/microstructure-and-mechanical-properties-of-friction-stir-welded-aa2024-t3 1/9

Microstructure and Mechanical Properties

of Friction Stir Welded AA2024-T3 Aluminum Alloy

Saad Ahmed Khodir*, Toshiya Shibayanagi and Masaaki Naka

Joining and Welding Research Institute, Osaka University, Ibaraki 567-0047, Japan

AA2024-T3 Aluminum alloy plates of 3 mm thickness were friction stir butt welded at a constant welding speed of 50 mm/min and

rotation speeds of 400, 600, 800, 1000, 1250, and 1500 min1. Effects of rotation speed on microstructures, hardness distributions, and tensile

properties of the joints were investigated. Equiaxed grain size increased with increasingrotation speed till 1000min1 of rotation speed. Increase

of rotation speed more than 1000min1 brought about no significant increase of grain size in the stir zone. Also, increasing rotation speed

resulted in finer and more homogenous distributions of second phase particles in the stir zone. Hardness increased both in the stir zone and

thermo mechanically affected zone as the rotation speed increased and reached to that of base metal. Kissing bond-free joints were fractured at

the heat affected zone on the retreating side and a maximum tensile strength of the joints was 402 MPa which was achieved at 1250 min1 of

rotation speed. The joint efficiency was 88%.

(Received September 30, 2005; Accepted November 25, 2005; Published January 15, 2006)

Keywords: AA2024-T3 aluminum alloy, friction stir butt welding, equiaxed grain size, second phase particles, hardness distributions, tensile

strength

1. Introduction

Friction stir welding is a new solid state welding process

invented for welding aluminum alloys.1) It has been proved to

bring about high quality for high strength Al alloys (2xxx,

7xxx) which are difficult to be welded using conventional

fusion welding.2) In this process a cylindrical tool (consisting

of a shoulder and probe assembly) is rotated with high

rotation speed and plunged into the joint line between the two

pieces to be butt welded together. The frictional heatgenerated by the welding tool makes the surrounding

material softer and allows the tool to move along the joint

line. The softened material starts to flow around probe,

resulting in a transferring of material from the leading edge of

the tool to the trailing side. This easily stirring action by the

rotating tool yields a heavily deformed region in the material.

An FSW joint is known to have three zones such as weld: (i)

intensively deformed zone called the stir zone (SZ), (ii)

thermomechanically affected zone (TMAZ), (iii) heat affect-

ed zone (HAZ).3)

Large number of studies4–13) on age hardening aluminum

alloys such as 2000, 7000, and 6000 series focused onmicrostructure changes and hardness distribution resulted

from FSW, while only a few studies1,13,14) characterized the

tensile properties of FSW joint of AA2024-T3 aluminum

alloy. According to these studies, dynamic recrystallization

takes place in SZ to produce equiaxed fine grain structure

with a tendency to be consisted of high angle grain

boundaries. In TMAZ, the material is also affected by both

plastic plastic deformation and the heat from the process. No

recrystallization occurs in this region, and generally there is a

distinct boundary between TMAZ and SZ especially on the

advancing side. In HAZ, the material has experienced a

thermal cycle which affects mechanical properties in this

region. Particularly no plastic deformation occurs in HAZ.

Hardness profile depends strongly on precipitation behavior

and slightly on the grain size in the weld. Thus dissolution

and growth of the precipitate would result in softening of

joints. Friction stir welding of AA2024-T3 is reported to

reduce the tensile strength to about 5–20% below that of base

metal (BM) depending welding conditions, but there is no

information about the fracture location if the fracture occurs

on the advancing side, the retreating side or in stir zone.

Most of these previous works on AA2024-T3 aluminum

alloy carried out at relatively lower rotation speed and higher

welding speeds. It is necessary to carry out the friction stirwelding in a wider range of rotation speeds, since the

properties of joints depend on many conditions such as

rotation speed of the tool, welding speed, probe geometry,

thickness of the work pieces, and material processed.

The present study aims investigate effects of rotation speed

on microstructure, hardness distribution, and tensile proper-

ties of AA2024-T3 aluminum alloy joints friction stir welded.

Relationship between rotation speed and fracture location

was also investigated.

2. Experimental Procedures

AA2024-T3 Aluminum alloy of 3 mm thick plates was

friction stir butt welded using a tool steel (SKD61) with

12 mm diameter shoulder and 4.0 mm diameter threaded pin.

The chemical composition and tensile properties of BM are

shown in Tables 1 and 2, respectively. The tool axis was

tilted by 3 degrees with respect to the vertical axis. The

welding speed was kept constant at 50 mm/min and rotation

speeds were set at 400, 600, 800, 1000, 1250, and 1500

min1. Specimens for microstructures observation were

machined from the welded plates. The observation of

microstructure was performed on a cross-section in the weld

region after mechanical polishing using through grit silicon

papers (up to grade 1500). Final polishing was performed

using 0.3mm alumina suspension and colloidal silica sus-

pension. Surface of the specimens was etched by Keller’s

reagent (1 ml HF, 1.5 ml HCl, 2.5 ml HNO3, and 95 ml H2O).*Ph.D. student, Joining and Welding Research Institute, Osaka University

Materials Transactions, Vol. 47, No. 1 (2006) pp. 185 to 193#2006 The Japan Institute of Metals



Then, microstructures of various regions of the welds were

examined by optical microscopy.

Since the welded alloy is known to show a rapid natural

aging response with time, microhardness measurement wascarried out immediately 4 h after the welding and after 4

months of natural aging for comparison. The measurements

were carried out along the center line of cross section which

is transverse to the welding direction with the internal

spacing of 0.5 mm under the load of 0.98 N for 15 s loading

time. The tensile test was carried out at room temperature at a

constant crosshead speed of 1 mm/min. Figure 1 shows

shape and dimensions of tensile test specimen used. The

tensile properties for each rotation speed were evaluated

using three tensile specimens machined from the same joint.

Temperatures were measured during FSW by utilizing

0.5 mm diameter K-type thermocouple at four positions as

indicated in Fig. 2, the bottom surface of welding center line,

at a distance corresponding to HAZ at the bottom surface in

the advancing side at 1.5 mm depth from bottom surface, and

at a distance of 10 mm either side of the weld center line at

1.5 mm depth below the top surface. Table 3 summarizes the

measurement locations for each welding conditions.

3. Results and Discussions

3.1 Macrostructures of joints

Figure 3 shows macroscopic appearance of cross-section

of joints at rotation speeds of 400 (a), 600 (b), and 1500 (c)

min1. All joints are free from neither porosity nor tunnel

like defects. Kissing bond was observed in the bottom region

of the joints at 400 and 1500 min1, yet it was severer at 400than 1500 min1. This could be attributed to an inadequate

stirring action of the probe to reach the bottom surface.

The size of SZ, TMAZ, and HAZ increase as rotation

speed increases. The weld zone becomes wider towards the

upper surface because of more excessive heating by friction

and plastic flow below the shoulder. In particular, intense

plastic deformation and onion ring were more pronounced on

the advancing side especially at lower rotation speed.

3.2 Microstructures of joints

Microstructures of different regions corresponding to the

locations indicated in Fig. 3(b) are shown in Fig. 4. Weld joint is characterized microstructurally by four distinct zones,

SZ along the weld center line, TMAZ on both sides of SZ,

HAZ which is surrounding the TMAZ, and BM. The

microstructure of BM is shown in Fig. 4(a) represents

elongated grains containing random distribution of second

phase particles. These grains with several hundred microns

long and approximately 70 mm wide are the resulted from the

deformation impozed during rolling.

Microstructure of HAZ shown in Fig. 4(b) is similar to that

of BM. Figure 4(c) represents TMAZ on retreating side,

revealing that the material has been plastically deformed.

Increasing rotation speed resulted in more severe bending of

the grains. No recrystallization occurred in this region, and

there is generally a distinct boundary between the TMAZ and

SZ on advancing side as shown in Fig. 4(e). On the other

hand, on the retreating side of the weld, the boundary

Table 2 Mechanical properties of AA2024-T3 base matal.

Yield stress

(MPa)

Tensile stress

(MPa)

Elongation

(%)

Nominal 295 Min. 440 Min. 15 Min.

Actual 316 455 22.8

T = 6

. 0 m m

1 4 . 0 m m

W e l d i n g z o n e

R= 20mm

15mm P= 20.3 mm

L0= 16.8mm

15mm

Fig. 1 Shape and dimension of tensile test specimen.

HAZ

BM

TMAZ

SZ

Adv.Ret.

10.0mm10.0mm

1 . 5 m

m

1 . 5 m m

1 . 5

m m

0.5mm

K-type thermocouple Welding center line K-type thermocouple

K-type thermocouples

Fig. 2 Layout and positions of k-type thermocouples at transverse cross-

section of welded joints.

Table 3 Distances of locations of temperature measurements at the

transverse cross-section of the welded joint.

Rotation speed

(min1)

Measuring temperature distance

from the welding center line

400 0 mm 4 mm — —

600 0 mm 4 mm — —

800 0 m m — 10 m m in adv. side 10 m m in ret. side

1000 0 mm 5 mm — —

1250 0 mm 6 mm — —

1500 0 mm 6 mm 10 mm in adv. side —

Table 1 Chemical composition of AA2024-T3 base metal.

Element Si Fe Cu Mn Mg Cr Zn Ti Ti þ Zr Al

Nominal

(mass%) 0.5 m ax. 0.5 m ax.

3.8 min.

4.9max.

0.3min.

0.9max.

1.2min.

1.8max. 0.1 max. 0.2 max. 0.15 max. 0.2 max. Re.

Actual

(mass%) 0.06 0.08 4.1 0.50 1.4 0.01 0.04 0.02 0.02 Re.

186 S. A. Khodir, T. Shibayanagi and M. Naka



between the TMAZ and SZ is rather unclear as shown in

Fig. 4(c). This is due to different flow behaviors of material

on both sides of the rotating and travelling probe. Severer

metal flow against BM on the advancing side than that on the

other side would be expected since tangential component of

the rotation has the same direction as the travelling direction,

while the two directions are in opposite way on the retreating

side. Thus steeper gradient of plastic strain caused by the

severer deformation mode eventually resulted in rather

distinct boundary between these zones on the advancing side.

Figure 4(d) represents the microstructure of SZ where

microstructure consisted of equiaxed grains with much

smaller size compared to the large elongated grains of BM.

Figure 4(f) represents the microstructure at the bottom

surface of the joint. It contains two different microstructures.

The first is fine grain region at the upper side resulted from

the action of rotating probe and the second is an elongated

microstructure at lower side remaining microstructure of BM

microstructure. No clear kissing bond was observed at this

rotation speed.

Figure 5 shows two types of defects formed under non-

suitable welding conditions. A severe kissing bond indicated

by an arrow in Fig. 5(a) appears at joints welded at 400 min1

and small segments indicated by arrows in Fig. 5(b) are

observed in a tope surface region of the joint fabricated at

1500 mm1. The friction stir welding proceeds in the region

between top and bottom surfaces that are contacting to the

rotating shoulder and the backing plate, respectively. Since

little change of thickness is given in addition by this welding

method, the metal flow around the probe would have a

Fig. 3 Macrostructures of the transverse cross section at constant welding speed 50 mm/min: (a) 400 min1, (b) 600 min1, and (c)

l500min1.

Fig. 4 Optical microstructure at different location corresponding to locations in Fig. 3(b) for 600 min1 rotation speed.

Microstructure and Mechanical Properties of Friction Stir Welded AA2024-T3 Aluminum Alloy 187



similar aspect to that observed in the equal channel angular

extrusion (ECAE). But rather complicated evolution of

microstructure takes place during joining; that is a successive

contact and lamination of each block of materials coming

from both sides of probe. Insufficient joining of these

materials caused by too complex metal flow in the top

surface region and poor contacting of base metals in the weldroot region should bring about the weld defects as observed

in Fig. 5. These defects played as stress concentration in the

joints and affect tensile properties.

Most of the previous works on AA2024-T3 aluminum

alloy friction stir welded6,8,10,15) showed that SZ zone had a

dynamically recrystallized fine grain structure with high

fraction of high angle grain boundaries. Figure 6 shows the

recrystallized grain structure with higher magnification at the

center position of SZ for different rotation speeds. Grain size

increases with increasing rotation speed. In addition, in-

creasing rotation speed resulted in finer and more homoge-

nous distributions of particles in SZ.Quantitative analysis of these micrographs revealed a clear

dependence of grain size on rotation speeds, which is shown

in Fig. 7. Grain size increases linearly from 2.6mm at

400 min1 to 7.1 mm at 1000 min1, while further increase of

rotation speed faster than 1000 min1 did not bring about

significant increase of grain size. A maximum grain size of

7.8mm was obtained at the rotation speed of 1500 min1.

The rotation speed dependence of grain size shown in

Figs. 6 and 7 is consistent with the report by Li et al.,16)

where the flow visualization and residual microstructures

associated with the dissimilar FSW of AA2024 to AA6061

was investigated. Figure 8 depicts a peak temperature at

bottom center position at SZ and HAZ for each rotationspeed. The temperatures increase in SZ and HAZ as rotation

Fig. 5 Defects observed in welded joints with unsuitable welding

conditions: (a) kissing bond at the welding root welded at 400min1,

(b) small segments at the top surface at 1500 min1.

Fig. 6 Optical microstructures of the stir zone: (a) 400 min1 (b) 600min1, (c) l000min1 and (d) 1500min1.




speed increases. There is a larger incremental of temperature

in SZ between 400 and 1000min1 (722.5 K at 400 min1

and 774.0 K at 1000 min1) of rotations speeds than that

between 1000 and 1500 min1 (774.0 K at rotation speed

1000 and 796 K at 1500 min1). Temperature in HAZ

increases gradually from 679 K at 400 min1 to 707K at

1500 min1.

In general, the average grain size would decrease with

decreasing working temperature and increasing working

stain rate.10)

Since the recrystallized temperature decreaseswith increasing strain rate or rotation speeds,6) grain growth

would proceed faster as temperature increases. These

observation of increasing grain size with increasing temper-

ature in SZ is consistent with simple grain growth relations of

the form D2{ Do2 ¼ A expðQ= RT Þðt nÞ:

where D and Do represent the instantaneous (or residual) and

initial dynamically recrystallized grain sizes, respectively, A

is a constant, Q is the corresponding activation energy for

grain growth, R is the gas constant, T is the absolute

temperature, t is the time, and n is a constant usually taken as

unity. The deformation time would be taken to be equal to the

probe diameter divided by traverse speed.9) So the welding

time is also constant, it is apparent that as the process

temperature (T ) increases the value of D2– Do2 will increase.

The previous equation can be applied in high temperature

deformation. Some researchers9,10) working on FSW of other

alloys proved that the dependence of the dynamic recrystal-

lization grain size was found to have the same dependence onthe Zenner–Hollomon parameters as material deformed via

conventional hot working process where the working temper-

ature and strain rate, combined and expressed by the Zenner–

Hollomon parameters.

3.3 Hardness of joints

Figure 9 shows microhardness profiles after 4 months of

natural aging across SZ, TMAZ, HAZ, and BM on a

transverses cross-section of joints welded at different rotation

speeds. Increasing rotation speed has a large influence on

hardness of SZ and TMAZ but slightly effective for HAZ.

Microhardness in SZ and TMAZ increases with increasing

rotation speed and is almost reached to that of BM at arotation speed of 1000 min1. Softened zones exist on both

sides of the HAZ and the value slightly increases as rotation

speed raised. Although increasing rotation speed resulted in

larger grain size in SZ, hardness increased as rotation speed

increased regardless of the coarsening of grains in SZ. This

tendency suggests that significant increase of hardness with

increasing rotation speed is not a function of grain size but a

function of size and distribution of second phase particles and

precipitates.

Grain size is less effective on hardness in SZ comparing

with other factors.17) For example, hardness beyond 10 mm

from the welding center line of the joint welded at 1500min1 is the same as that of BM naturally aged. At about

10 mm from the welding center line, hardness starts to de-

crease and a minimum value is achieved at about 6 mm,

where some precipitates might have coarsened and lost their

coherency due to thermal history. Beyond that point of mini-

mum value, when going to near the welding center line, hard-

ness value starts again to increase and a higher value than that

of base metal was obtained at SZ region. This could be attrib-

uted to increase temperature which may be sufficient to force

larger precipitates to be dissolute and reprecipitate in the

weld during natural aging. Hardness in these regions towards

TMAZ is still lower than that of BM. Unlike coarsening, dis-

solution leads to a subsequent recovery of hardness.18) Much

more precipitates dissolved in TMAZ and SZ during welding

since peak temperature in SZ was 796 K that is higher than

the solution treatment temperature of this alloy.

-1400 600 800 1000 1200 1400 1600

1

2

3

4

5

6

7

8

9

10

Rotation speed, r / min

G r a i n s

i z e , d /

m

Fig. 7 Relationship between rotation speed and grain size of stir zone.

Rotation speed, r / min-1

400 600 800 1000 1200 1400 1600

640

660

680

700

720

740

760

780

800

820

HAZ

SZ

P e a k t e m p e r a t u r e , T /

K

Fig. 8 Relationship between rotation speeds and peak temperature in SZ

and HAZ.

Distance from weld zone center line, l /mm-12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12

M i c r o h a r d n

e s s ( H v 9 8 0 . 7 m N )

80

90

100

110

120

130

140

150

160

RS: 400min-1

RS: 600min-1

RS: 1000min-1

RS: 1500min-1

Pin diameter

Shoulder diameter Ret.Adv.

Fig. 9 Hardness profiles at mid thickness transverse to welding directions

of the weld joint for different rotation speeds.




Figure 10 shows hardness profiles after 4 h and 4 months

of natural aging for the cases of 400 min1 (a) and

1500 min1 (b). In case of 400 min1 as shown in Fig. 10(a),

four hours of natural aging resulted in a slight increase of

hardness both in SZ and TMAZ but no significant increase

occurred both in HAZ and BM. Prolonged aging time of 4months brought about further increase of hardness in the

length of shoulder diameter and higher value of increment

was observed in the region comparable to the length of probe

diameter, but the value is still lower than that of BM. On the

other hand, Fig. 10(b) for the case of 1500 min1 represents a

significant increase of hardness in SZ and TMAZ after

natural aging of both 4 h and 4 months. Particularly 4-months

aging made SZ slightly harder than BM. HAZ showed less

increment of hardness and no change occurred in BM.

Hardness after 4 h of natural aging was higher both in SZ and

TMAZ at higher rotation speed than that at lower one.

This significant increase of hardness could be attributedmainly to the second phase particles size and distributions

(especially in SZ) and dislocation density (especially in

TMAZ), where increasing rotation speed mean increasing

strain. After four month of natural aging for the joint welded

at 400min1, the hardness increased at 5.5 mm from the

welding center line and maximum value increased by 8–

10 Hv in SZ and TMAZ, while at higher rotation speed such

as 1500 min1, hardness increased at 8 mm from welding

center line and maximum value increased by 15–20 Hv in SZ

and TMAZ as shown in Fig. 10. Some research works on the

other precipitation hardening alloys reported that SZ contains

a high density of dislocations,19) while other research

reported a low dislocation density in stir zone.3) Su et al.5)

observed high density of dislocation in many grains suggest-

ing that plastic deformation might have been introduced both

during FSW even after dynamic recrystallization and during

the cooling from the elevated weld temperature to ambient

temperature. Also it is clear that hardness partially recovered

in the softened HAZ in all welded joints after four months,

indicating that not all precipitates coarsened due to overaging

but some of these precipitates dissolved during joining and

precipitated again during the natural aging.

During FSW the temperature history and the deformationrate vary depending on the region through the joint. For

precipitation hardening aluminum alloys, these parameters

strongly influence the precipitation distribution that controls

the weld properties. Precipitations play dominant role such as

either a hardening agent at small sizes or detrimental to the

mechanical properties at large sizes. Hence the T3 treatment

brings about three possible changes in aging process in SZ,

TMAZ and HAZ where the initial precipitates structure

evolves: dissolution, over aging, and artificial aging. Hugh et

al.18) described softening behaviors of 2024-T3 Al alloy by

quenching after isothermal treatment at temperatures from

473 to 773 K for different holding times. Different behaviorsof precipitates take place at each temperature as follows;

– at 773 K, full strength recovery was observed for all the

hold time,

– at 673 K, partial recovery occurs, to a progressively

greater extent with increasing hold time,

– at 623 K the reverse is true—partial strength recovery

decreases as hold time increases.

According to them, the total hardness recovered in SZ was

due to mainly by natural aging because temperature in SZ

was more than 773 K. But the total amount of recovered

hardness in SZ cannot attributed only to the natural aging

since the cooling rate in FSW is low compared to the critical

cooling rate in quenching after the isothermal heating. Thecritical cooling rate is defined as the cooling rate where full

hardness recovery takes place at room temperature after

cooling. The amount of precipitation occurring during

cooling to room temperature controls the amount of

subsequent hardening since any further precipitation reaction

needs super-saturated solute remaining after cooling.

The hardness measured at 10 mm from the weld center line

of joints welded at all rotation speeds is the same as that of

BM. The temperature measured at that point for the case of

1500 min1 is equal to about 635 K, indicating that there is

neither overaging nor dissolution below that temperature.

Only artificial ageing occurred below that temperature.Overaging occurred at the temperature more than 635 K.

Precipitates start to dissolve at around 673 K in the region

about 8 mm from the welding center line and the phenom-

enon is accelerated with increasing temperature. Dissolving

of precipitates at that region resulted in hardness recovery

due to natural aging at room temperature as shown in

Fig. 10(b).

But partial recovery of hardness in the softened area of

HAZ (especially from 5 to 8 mm from the welding center line

for example in the case of 1500 min1) clearly observed after

the natural aging longer than 3 months. This may be

explained as follows. Precipitation process is explained in

terms of nucleation and growth. Driving force for the

nucleation strongly depends on the degree of super-satura-

tion, and growth kinetics is mainly affected by diffusion of

constituents that are affected by vacancies. In addition,

M i c r o h

a r d n e s s ( H v 9 8 0 . 7 m N )

80

90

100

110

120

130

140

150

160

After 4 hours

After 4 months

Ret.Adv.(a)

-12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12

M i c r o h a r d n e s s ( H v 9 8 0 . 7 m N )

80

90

100

110

120

130

140

150

160

After 4 Hours

After 4 months

(b)

Distance from weld zone center line, l /mm

Shoulder diameter

Pin diameter

Fig. 10 Hardness profiles after 4 h from welding and naturally aged after 4

months at mid thickness transverse to welding directions of the weld joint.

(a) 400min1, (b) 1500min1.




nucleation sites are given by dislocations, vacancies and

other heterogeneous regions in lattice. The degree of super-

saturation in HAZ would rather decreased than that in SZ

since temperature increase in HAZ is less than that in SZ.

Therefore this difference of chemical driving force should

have played an important role in the difference of driving

force for the nucleation resulting in the difference of timerequired for the recovery of hardness observed.

Besides the role of chemical driving force, lattice defects

such as dislocations and point defects introduced during FSW

should be taken into account as nucleation sites. Vacancies

also play dominant role in diffusion process of solute atoms

to nucleation site and growing precipitates. Since HAZ

experienced a lower temperature than SZ during FSW,

dislocations would have larger chance to remain resulting in

giving more nucleation sites. Equilibrium amount of vacan-

cies is larger at higher temperature, and larger cooling rate

brings about larger amount of quench-in vacancies. Thus

HAZ, having lower maximum temperature and cooling ratethan SZ, would contain less amounts of the defects and

eventually give less nucleation sites and diffusivity of

constituents.

The temperature history during FSW affects the chemical

driving force, nucleation sites and diffusivity in terms of

many factors such as degree of super-saturation, dislocation

density and amounts of vacancies remained after cooling

down to room temperature. As a combined effect based on

these fundamental factors, different recovery process of

hardness proceeded as observed in the present joints.

The slight increase of hardness in HAZ with increasing

rotation speed as shown in Fig. 9 could be explained by the

relatively increasing heating or cooling rate during welding.Figure 11 shows temperature change with time at a point in

HAZ during FSW at 400 and 1500 min1 of rotations speeds.

Heating and cooling rates could be calculated from temper-

ature (T /K) changes with time (t /s) in temperature range

from 635 K to peak temperature corresponding for each

rotation speed as shown in Fig. 11. This temperatures range

was selected according to the temperature measured at

10 mm from the welding center line, where coarsening of

precipitates starts after 635 K. For 400 min1, heating and

cooling rates were 7.8 and 6.0 K/s respectively. While

heating and cooling rates at 1500 min1 were 15.6 and

10.2 K/s, respectively. So the amount of precipitates coars-

ened during heating decreased with increasing heating ratesince increasing heating rate reduce the time for precipitates

to grow and hence leads to hardness increase in HAZ. Also

increasing cooling rate after welding increases the amount of

supersaturated solute which will be available for further

precipitation reaction at room temperature.

3.4 Tensile properties of joints

Figure 12 shows the tensile properties of joints welded at

different rotation speeds. Tensile strength, yield strength, and

elongation increased with increasing rotation speeds but still

lower than those of base metal, especially the elongation of

the joints is much lower than that of base metal and its

maximum value was about 12.4% at 1250 min1 (about half of base metal). This could be attributed to the coarsening of

the precipitates in HAZ.20) The maximum tensile strength of

the joints was 402 MPa which was achieved at 1250 min1 of

rotation speed and the joint efficiency was 88.3%. The lowest

tensile properties were obtained for the joint welded at

400 min1 due to lower hardness and severe kissing bond at

the root of weld as shown in Fig. 5(a). Also, due to kissing

bond at 1500 min1 and small segments observed in a top

surface at 1500 min1 Fig. 5(b), the tensile properties of

joints slightly decreased less than those of joint welded at

1250 min1.

Figure 13 shows the changes in fracture location of tensiletested specimens welded at three kinds of rotation speeds.

Fracture occurs in SZ for the specimen welded at a rotation

speed of 400 min1 as shown in (a). In this case, crack

propagated almost center position of SZ and seems to be in a

brittle manner. As rotation speed raised to 1500 min1, the

crack changed to propagate along the border between SZ and

TMAZ on the retreating side as shown in (b). On the other

hand, when the joints are fabricated at rotation speed from

600 to 1250 min1 where no defects exist, crack propagated

in HAZ on the retreating side as seen in (c) for the case of

1250 min1.

Tensile properties and fracture location of FSW joints are,

to large extent, depend on the welding defects and hardness

distributions of the joints, of which in return are function of

the welding parameters. In addition, the friction-stir-welded

joint is known as a heterogeneous composite and its different

Welding time, t / s10 20 30 40 50 60 70 80 90 100

P e a k t e m p e r a t u r e , T / K

350

400

450

500

550

600

650

700

750

400min-1

1500min-1

Heating or cooling rate

= tan = T(K) / t (s)

T / ( K )

t / (s)

Fig. 11 Temperature change by time at the HAZ during FSW of 400,

1500min1 rotations speeds.

Rotation speed, r / min-1

200 400 600 800 1000 1200 1400 1600

S t r e n g t h ,

/ M P a

240

270

300

330

360

390

420

450

E l o

n g a t i o n ( % )

2

4

6

8

10

12

14

16

18

20

Yield Strength

Tensile strength

Elongation

Fig. 12 Relationship between rotation speed and tensile properties of the

welded joints.




component parts, including their interfaces, possess different

mechanical properties. When joints are free from defects,

their tensile properties are only affected by the microhardness

distributions.21) As seen in Fig. 9, increasing tensile strength

of joints welded at rotation speeds from 600 to 1250 min1

could be attributed to a slight increase of hardness in the

HAZ. When a tensile load is applied to the joint, the stress

and strain concentration takes place in the lowest-strength

part or region in the joint, and consequently the joint is

fractured in this region. Thus the improvement of strength in

HAZ resulted in the increase of joint strength.As mentioned above, when the joints are free from defects,

the joints are fractured on the retreating side instead of the

advancing side. Temperatures measurements at 10 mm from

both sides of welding center line at 800 min1 of rotation

speed were 584 K on retreating side and 577 K on the

advancing side. Higher temperature resulted in softening

retreating side and hence fracture occurred in this side. This

result sufficiently indicates that the tensile strength on the

advancing side is higher than that in the other side. The

advancing side shows slightly higher hardness than the other

side. Similar results of fracture locations were obtained in the

joints of 6061-T6 aluminum alloy where fracture occurred onthe retreating side.21)

4. Conclusions

Effects of rotation speed on microstructures, hardness

distributions, and tensile properties of AA2024-T3 were

investigated and the following results were obtained;

(1) An increase of equiaxed grain size in the SZ with

increasing rotation speed was observed till 1000 min1

of rotation speed. Increasing rotation speed faster than

1000 min1 did not bring about significant increase of

grain size in the SZ. Also, increasing rotation speed

resulted in finer and homogenous distributions of

particles in the SZ.

(2) Hardness increased in the SZ and TMAZ as the rotation

speed increased. This was due to finer and homogenous

distributions of particles in the SZ and aging of fine

precipitates.

(3) Hardness minima existed in the HAZ and slightly

increased as the rotation speed increased. After four

months natural aging, the hardness values partially

recovered in HAZ but still lowered than that of BM.

(4) The kissing bond-free joints were fractured at the HAZ

on the retreating side, while the defect containing joints

fractured at SZ or border between SZ and TMAZ on the

retreating side. The maximum tensile strength was

402.0 MPa which was achieved at 1250min1 of rotation speed. The joint efficiency was 88.3%.

Acknowledgements

This work was supported by Grant-in-Aid for scientific

research B (project No. 17360353) and Grant-in-Aid for

cooperative research project of nationwide joint-use Re-

search Institutes on Development Base of Joining Technol-

ogy for new Metallic Glasses and Inorganic Materials from

the Ministry of Education, Culture, Sports, Science and

Technology, Japan.

REFERENCES

1) Y. J. Chao, Y. Wang and K. W. Miller: Weld. J. 81 (2001) 196–200.

2) C. J. Dawes and W. M. Thomas: Weld. J. 75 (1996) 41–45.

3) C. G. Rhodes, M. W. Mahoney, W. H. Bingel, R. A. Spurling and C. C.

Bampton: Scr. Mater. 36 (1997) 69–75.

4) O. V. Flores, C. Kennedy, L. E. Murr, D. Brown, S. Pappu, M. N. Brook

and J. C. McClure: Scr. Mater. 38 (1998) 703–708.

5) J.-Q. Su, T. W. Nelson, R. Mishra and M. Mahoney: Acta Mater. 51

(2003) 713–729.

6) S. Benavides, Y. Li, L. E. Murr, D. Brown and J. C. McClure: Scr.

Mater. 41 (1999) 809–815.

7) M. A. Sutton, B. Yang, A. P. Reynolds and J. Yan: Mater. Sci. Eng. A

364 (2004) 66–74.8) M. A. Sutton, B. Yang, A. P. Reynolds and R. Taylor: Mater. Sci. Eng.

A 323 (2002) 160–166.

9) K. V. Jata and S. L. Semiatin: Scr. Mater. 43 (2000) 743–749.

10) C. I. Chang, C. J. Lee and J. C. Huang: Scr. Mater. 51 (2004) 509–514.

Fig. 13 Optical micrograph of ruptured tensile test specimens.




11) C. G. Rhodes, M. W. Mahoney, W. H. Bingel and M. Calabrese: Scr.

Mater. 48 (2003) 1451–1455.

12) M. J. Jones, P. Heurtier, C. Desrayaud, F. Montheillet, D. Allehaux and

J. H. Driver: Scr. Mater. 52 (2005) 693–697.

13) G. Biallas, R. Braun, C. Dalle Donne, G. Staniek and W. A. Kaysser:

Proceedings of the 1st International Symposium on Friction Stir

Welding, (The Welding Institute, Cambridge, UK) Thousand Oaks

California, (1999) 14–16 June.14) R. Zettle, S. Lomolino, J. Dossantos, T. Donath, F. Beckmann, T.

Lippman and D. Lohwasser: Proceeding of IIW Pre-Assembly Meeting

on FSW (Nagoya, Japan, 2004) 65–70.

15) A. F. Norman, I. Brough and P. B. Prangnel: Mater. Sci. Forum 331–

337 (2000) 1713–1718.

16) Ying Li, L. E. Murr and J. C. McClure: Mater. Sci. Eng. A 271 (1999)

213–223.

17) H. G. Salem: Scr. Mater. 49 (2003) 1103–1110.

18) H. R. Shercliff, J. R. Michael, A. Taylor and T. L. Dickerson:

Mecanique & Industries 6 (2005) 25–35.

19) K. V. Jata, K. K. Sankaran and J. J. Ruschau: Metall. Mater. Trans. A

31 (2000) 2181.20) D. S. Thompson: Metall. Mater. Trans. A 6 (1975) 671–683.

21) H. J. Liu, H. Fujii, M. Maeda and K. Nogi: J. Mater. Process. Technol.

142 (2003) 692–696.


Microstructure and Mechanical Properties of Friction Stir Welded AA2024-T3 Aluminum Alloy

Documents