EFFECT OF WELD PARAMETERS ON

MECHANICAL PROPERTIES OF THE FRICTION

STIR WELDING AA6063-T5
Tran Hung Tra
Lecturer, Materials science Dept., Nha Trang University,Vietnam, Tel(+84) 935272168, Fax: (+84) 58
3831147, [email protected]

Abstract
Effect of tool rotation speed and welding speed on the mechanical properties of the FSW
joint of AA6063-T5 was investigated. The influences of the ratio of rotation speed and
weld speed on the weld thermal cycle, hardness distribution, and tensile strength of the
FSW joint were clarified. The experimental results showed the lower the ratio of rotation
speed and weld speed, the lower weld heat input, the narrower the softened zone and the
higher tensile strength were. In all cases, the tensile fracture located outside the stirred
zone, in the retreating side or advancing side, where the hardness was lowest. It was also
found that the residual stress in and around the welded zone was quite low, in range of ten
percent of base metal yield stress.
Key works: Friction Stir Welding, Thermal Cycle, Hardness, Tensile Properties, Residual
Stress.

Introduction
Friction stir welding (FSW) is emerging as an appropriate alternative technology with high
efficiency due to high-processing speeds. Since the joint can be obtained below the melting
temperature, this method is suitable for joining a number of materials which are extremely
difficult to be welded by conventional fusion techniques [1-3]. Besides welding, friction
stir processing has been developing as a potential metalworking technique for metallic
components such as localized microstructure modification, control of microstructures in
the near-surface layers [4-7]. The basic concept of FSW is remarkably simple. The rotating
shoulder and probe of a non-consumable tool heat and plasticize the surrounding metal and
a solid-state joining is accomplished there. There, the tool serves three primary functions
(1) heating of the workpiece, (2) movement of material to produce the joint, and (3)
containment of the hot metal beneath the tool shoulder [8]. The heating is generated by
friction between the rotation tool (shoulder and probe) and the workpiece and by plastic
deformation of the workpiece. The localized heating softens material around the tool probe
and then combines with the tool shoulder and probe rotation and translation leads to
movement of material from the front to the back of the probe to fill the hole in the tool
wake as the tool moves forward [8]. The tool shoulder restricts metal flow upward. During
FSW process, the welding parameters combined with tool geometry exert a significant
effect on the material flow pattern and temperature distribution, thereby influencing the
microstructural evolution of material, the formation, and mechanical properties of the joint
[1-11]. To broaden the application of FSW, it is necessary to clarify the effect of weld
parameters on the properties of the joint. The aim of this work is to investigate the effect of
tool rotation speed and weld speed on the mechanical properties of the FSW butt-joint of
AA6063-T5 plate.

ASEAN Engineering Journal Part B, Vol 1 No 4, ISSN 2229-127X p.73

Experimental Procedures
The 5 mm AA6063-T5 plates were butt-joined by friction stir welding using the 1.5 kW
NC milling machine. The weld tool geometry applied in this work is a scrolled shoulder
tool and a truncated cone pin with the pin diameter of 5.0 mm at the midle pin length, the
pin heigth of 4.5 mm, and the screw pitch of 1.0 mm. The pin is aligned at a tilt angle of
2.0 deg. in the plane containing pin axis and center weld line (the tilt angle is defined as the
angle between pin axis and the direction perpendicular to the workpieces). The tool tip is
kept at a distance of 0.2 mm from the backing anvil. Various regimes of weld parameters
were performed by combining the tool rotation speed (denoted , revolving/min) and the
weld speed (denoted v, mm/min). The ratio of rotation speed and weld speed (denoted /v,
revolving/mm) investigated is in range from 3.1 to 37.5 revolving/mm. During the welding,
the thermal hysteresis at the end weld center and at the shoulder limit areas in advancing
side and retreating side (1.0 mm far from the shoulder limit line) was monitored by the
thermal couples attached on the weld surface with a computer software interface. After
being fabricated, the microstructure was observed by Scanning Electron Microscope
(SEM). The hardness in and around the welded zone was measured by a diamond
indentation with 50g loading and 10 sec hold time. The tensile strength of the FSW 6063-
T5 was measured by the 3.0 mm thickness plate specimens those were extracted in the
perpendicular to the weld direction from the 5.0 mm FSW plate as shown in Fig. 1. In
these tensile specimens, the weld surface and root surface were removed with the layers of
1.0 mm thickness. The tensile tests were performed by the MTS hydraulic machine at a
constant strain rate of 10-3/s. The residual stress in and around the welded zone of the FSW
joint was measured by X-ray diffraction (XRD) method [12] in which the specimen surface
of specimens was carefully polished by the electrolytic polisher. In X-ray stress
measurement, the stress in the surface layer is in general calculated from the change in the
lattice spacing. To determine residual stress through XRD, the X-ray stress constants K of
BM was determined through six applied load values corresponding to 0, 15, 30, 45, 60, and
75 percent of yield stress in this work. Assuming that the strain in the crystal lattice is
linearly distorted, residual stresses are then calculated by means of the stress constants of
the metal. Here, an X-ray tube with a chrome anode operated at 30 kV and 30 mA was
used to measure the residual stresses. Measurements were collected at a Bragg angle of
139.5 deg. corresponding to diffraction at the {311} planes. The scanning angle of 2 is
from 137 deg. to 142.5 deg. with step 0.05 deg. and preset time (time per measurements) of
10 seconds. At each measured location, the residual stresses was measured in a diffracted
area of 22 mm2 on specimen surface. Two residual stress components were measured;
one perpendicular to weld direction (denoted by transverse) and the other parallel to weld
direction (denoted by longitudinal residual stress).

Experimental Results and Discussion

Several combined regimes of rotation speed and welding speed were performed for
fabricating the joints to study the effect of rotation speed and weld speed on the
mechanical properties of the FSW joint of aluminum 6063-T5. The typical microstructure
of a FSW 6063-T5 (fabricated at /v = 3.1 revolving/mm) characterized by the dynamic
recystallization was seen in Fig. 2. In general, the grain sizes in the stirred zone and base
metal were about 15m and 70m in diameter, respectively (compare Fig. 2(c) with Fig.
2(d)). In the stirred zone near the interface area (Fig. 2(b)), the microstructure displays an
inhomogeneous feature. In the stirred zone, the grain size in region near interface area
(region II in Fig. 2) seems to be finer than that in the middle of stirred zone (region III in
Fig. 2); this might associates with the fact that the weld temperature in region II is higher
than that in region III [13]. Outside the stirred zone, the grain is comparable to that of base

ASEAN Engineering Journal Part B, Vol 1 No 4, ISSN 2229-127X p.74

metal (BM), see Fig. 2(a). The fine grain size in the stirred zone is mostly independent to
both rotation speed and weld speed. In almost all of the fabricated joints, the formation of
zigzag-line pattern (usually called “lazy S”) in the stirred zone was found to be inevitable
excepting for the FSW fabricated at high value of /v, i.e. /v = 37.5 revolving/mm. It
should be noted that, in the case of free zigzag-line pattern FSW, the weld temperature is
highest and the weld surface is severely turbulent deformation.
The temperature distribution within and around the stirred zone directly influences the
microstructure of the welds, such as grain size, grain boundary character, coarsening and
dissolution of precipitates, and resultant mechanical properties of the welds [9-11,14]. It is
important to obtain the temperature distribution in and around the welded zone during
FSW. However, be noted that direct temperature measurements within the stirred zone are
difficult due to the intense plastic deformation produced by the rotation and translation of
the pin tool. The results of weld temperature measured at the advancing side, the retreating
side and at the weld end center are summarized in Fig. 3-5. The dependence of the peak
weld temperatures on weld parameters are shown in Figs. 3&4. Both the rotation speed and
the weld speed influence the peak weld temperature there the weld temperature is
increased with the increase of rotation speed and/or with the decrease of weld speed, see
Fig. 3. In all cases, the maximum weld temperature was found at the weld center and the
weld temperatures in advancing side and in retreating side seem to be comparable, see Fig.
4. Furthermore, the results in Fig. 4 showed the peak weld temperatures are proportional
correlation to the ratio /. Furthermore, Fig. 4 also indicates the weld temperature is
proportional to the weld input energy. From this view, it is reasonable to choose the ratio
of tool rotation and the weld speed, /, as a weld parameter covering both tool rotation
speed and weld speed, and their interaction. Another important aspect seen from Fig. 4 is
the peak weld temperatures are higher than that of dissolution point and remarkable lower
than melting point of this alloy, the dash-lines in Fig. 4 (the dissolution point of AA6063-
T5 was referred in the literature [14]). The influence of weld parameters on the thermal
cycles is shown in Fig. 5. It is obviously that weld speed influences significantly the
heating rate whereas the rotation speed has a minor effect.
The hardness distributions measured at the middle-line in the cross sections are shown
in Fig. 6 as a function of weld parameter, /v. In general, a softened feature in and around
the welded zone is observed in all FSWs. Here, the higher /v the higher weld temperature
(see Fig. 4) and the wider softened zone (see Fig. 6) in the FSW were seen. The width of
soft zone increases with the increase of /v. The softening appearing in and around the
welded zone could be relating to the dissolution and/or coarsening of precipitates in this
alloy [14]. It was also found that, in all cases, the lowest hardness in the cross section of
the FSW is placed at the heat affected zone (HAZ) in the advancing side and/or retreating
side, outside the stirred zone (see Fig. 7). The fact that the hardness in the stirred zone is
higher than that in HAZ might be associated with high density of grain boundaries in the
stirred zone or Hall Petch effect [15].

ASEAN Engineering Journal Part B, Vol 1 No 4, ISSN 2229-127X p.75

 15

20.9

0
R1

45 100

10

Fig. 1. Geometry of tensile specimen

(i) (ii) (iii)

Stirred zone 2 mm

(a) (b)

50m 50m
(c) (d)

50m 50m

Fig. 2. Microstructure in the cross section of FSW No.5; (a) region (i), (b) region (ii), (c)
region (iii), and (d) base metal.

ASEAN Engineering Journal Part B, Vol 1 No 4, ISSN 2229-127X p.76

 500

Maximum weld temperature, C

o
400

300

200
Rotation speed of 1080 rpm
Rotation speed of 1875 rpm
100
0 100 200 300 400
Weld speed, mm/min

Fig. 3. Effect of rotation speed and weld speed on the maximum temperature in the FSWs.

Advancing side
700 Retreating side
Weld center
Peak weld temperature, oC

Melting point
600

500

400

300

Dissolution point
200
2 10 50
Ratio of rotation and weld speed, rev/mm

Fig. 4. Relation of peak weld temperatures with the ratio of tool rotation speed and weld
speed (/v).
500
 = 1080/50 (rev/mm)
400  = 1080/350 (rev/mm)
Temperature, oC

= 1875/50 (rev/mm)

 = 1875/350 (rev/mm)
300

200

100

0
0 100 200 300 400
Time, sec

Fig. 5. Effect of weld parameters on the thermal cycle at the end weld center of the FSWs.

ASEAN Engineering Journal Part B, Vol 1 No 4, ISSN 2229-127X p.77

 Measured location

100

80

Hardness, Hv
60

40
 = 37.5 rev/mm
20  = 21.6 rev/mm
 = 3.1 rev/mm
0
-15 -10 -5 0 5 10 15
Distance from weld center, mm

Fig. 6. Hardness in the cross section of the FSW under different weld conditions.

Fig. 7. Cross sections and the tensile fracture locations in FSW No.3 (/v = 7.2 rev./mm)
and FSW No.4 (/v = 5.4 rev./mm) (Advacing side and retreating side are in the left side
and right side, respectively).

Fig. 8. Hardness distribution in cross section of FSW No.3

Fig. 9. Hardness distribution in cross section of FSW No.4.

ASEAN Engineering Journal Part B, Vol 1 No 4, ISSN 2229-127X p.78

 1.0

Normalized to base metal

0.8

0.6

0.4

0.2
Tensile strength
Elongation
0.0
2 10 50
Ratio of rotation and weld speed, rev/mm

Fig. 10. Elongation and tensile strength of FSWs normalized to these of BM under various
weld conditions.

260 520

Maximum weld temperature, C

o
Tensile strength
Maximum weld temperature
Tensile strength, MPa

240 480

220 440

200 400

180 360

160 320

2 10 50
Ratio of rotation and weld speed, rev/mm

Fig. 11. Relation of tensile strength and weld temperature with the ratio of rotation speed
and weld speed.

20 Transverse residual stress

Longitudinal residual stress
Residual stress, MPa

Transverse
0

Longitudinal
-20

-40
Advancing side Retreating side

-10 -5 0 5 10
Distance from weld center, mm

Fig. 12. Residual stress in and around the welded zone.

ASEAN Engineering Journal Part B, Vol 1 No 4, ISSN 2229-127X p.79

 The tensile strength of the FSW 6063-T5 was carried out by the MTS hydraulic
machine at a constant strain rate of 10-3/s by means of the specimen shown in Fig. 1. The
tensile failures of all FSW specimens took place at the HAZ in the advancing side or in the
retreating side as seen in Fig. 7. It should be noted that all tensile failures were occurred
outside the stirred zone. The tensile fracture locations were well agreement with the lowest
hardness locations as seen in Fig. 8 and Fig. 9 (the locations of solid-lines).
The results of tensile strength and elongation under various weld conditions,
normalized to these of BM, are presented in Fig. 10. In all weld conditions, the tensile
property of the FSW was lower than that of BM. Generally, the tensile strength and
elongation of the FSWs are about 75% and 55% that of BM, respectively. The influence of
weld parameters on tensile strength of the FSW can be seen in Fig. 11. Here, the tensile
strength is reduced when the weld parameter /v is increased. This reduction of tensile
strength could be concerned with the higher heat input during the welding process as seen
in Fig. 11 that must be resulting from the dissolution and/or coarsening of precipitates in
and around the welded zone of this alloy (see the dissolution point in Fig. 4). It is
necessary to note that the FSW fabrication at low value of /v could lead to the incomplete
joint, i.e., /v = 2.5 rev/mm.
Finally, the residual stress in and around of the welded zone of a FSW (fabricated at
/v = 3.1 rev/mm) was examined by X-ray diffraction technique. The measured results of
the transverse and longitudinal residual stress are shown in Fig. 12 in which each value is
averaged from three measurements. The result in Fig. 12 approved that the FSW could be
obtained with quite low residual stress. Here, both longitudinal and transverse residual
stresses were in range of about ten percent of yield stress of BM.

Conclusions
The FSW of aluminum alloy 6063-T5 was fabricated and the effect of weld parameters on
its thermal cycles, hardness, and tensile properties was investigated. The heat input was
found to be proportional to the ratio of tool rotation speed and weld speed, / . The lower
the ratio of rotation speed and weld speed, the higher tensile strength were. The tensile
fracture located outside the stirred zone, in the retreating side or advancing side, where the
hardness was lowest. The kissing bond defects seem to be inevitable in all most all of cases
excepting for a high value of / (i.e., 37.5 rev/min) there the weld surface was deformed
turbulently. The FSW joint could be obtained with quite low residual stress.

Acknowledgements
The author is grateful to Prof. Masakazu Okazaki for his guidance and experimental
equipment support. The author also thanks to Prof. Kenji Suzuki for his support in the
residual stress measurement.

References
[1] M. Gene, The welding of aluminium and its alloys, Woodhead Publishing Ltd,
Cambridge England; 2002.
[2] Tran Hung Tra, Seino M., Sakaguchi M., Okazaki M., “Fatigue crack propagation
behavior relevant to inhomogeneous microstructure of friction stir welding AA6063-
T5”, Journal of solid mechanics and materials engineering, 4 (6), 840-848, 2010.
[3] M. Sakaguchi, A. Sano, Tran Hung Tra, M. Okazaki, M. Sekihara, “Low cycle and
thermal-mechanical fatigue of friction welded dissimilar superalloys joint”, Journal of
solid mechanics and materials engineering, 2 (12), 1508-1516, 2008.

ASEAN Engineering Journal Part B, Vol 1 No 4, ISSN 2229-127X p.80

[4] ZY. Ma, “Friction Stir Processing Technology: A Review”, Metallurgical and
Materials Transactions; 37A: 642-658, 2008.
[5] T. Morishige, M. Tsujikawa, M. Hino, T. Hirata, S. Oki, K. Higashi, “Microstructural
modification of cast Mg alloys by friction stir processing”, International journal of cast
metals research, 21(1-4), pp.109-113, 2008.
[6] J.-Q. Su, T.W. Nelson, C.J. Sterling, “Grain refinement of aluminum alloys by friction
stir processing”, Philosophical Magazine, 86(1), pp.1-24, 2006.
[7] Y. Morisada, H. Fujii, T. Mizuno, G. Abe, T. Nagaoka, M. Fukusumi, “Modification of
nitride layer on cold-work tool steel by laser melting and friction stir processing”,
Surface and Coatings Technology, 204(3), pp.386-390, 2009.
[8] R.S. Mishra, M.W. Mahoney editors, Friction Stir Welding and Processing, ASM
International, 2007.
[9] D.P. Field, T.W. Nelson, Y. Hovanski, K.V. Jata, “Heterogeneity of crystallographic
texture in friction stir welds of aluminum”, Metallurgical and Materials Transactions A:
Physical Metallurgy and Materials Science, 32(11), pp.2869-2877, 2001.
[10] R.W. Ponda, J.F. Bingert, “Precipitation and grain refinement in a 2195 Al friction
stir weld”, Metallurgical and Materials Transactions A: Physical Metallurgy and
Materials Science, 37 pp.3593-3604, 2006.
[11] A. Oosterkamp, L.D. Oosterkamp, A. Nordeide, “Kissing bond' phenomena in
solid-state welds of aluminum alloys”, Welding Journal (Miami Fla), 83(8), pp.225-S,
2004.
[12] Standard Method for X-Ray Stress Measurement, The Society of Materials Science
of Japan 2005.
[13] W. Tang, X. Guo, J.C. McClure, L.E.Murr, and A. Nunes, “Heat Input and
Temperature Distribution in Friction Stir Welding”, J. Mater. Process. Manuf. Sci., Vol
7, pp 163-172, 1998.
[14] Y.S. Sato, H. Kokawa, M. Enomoto, S. Jogan, T. “Hashimoto, Precipitation
sequence in friction stir weld of 6063 aluminum during aging”, Metallurgical and
materials transactions A, 30 pp.3125-3130, 1999.
[15] W.D. Callister, Materials Science and Engineering - An Introduction, John Wiley
& Sons, Inc., 2007, pp. 188-198.

ASEAN Engineering Journal Part B, Vol 1 No 4, ISSN 2229-127X p.81