Journal of Materials Processing Technology 180 (2006) 263–270

Effect of welding parameters on mechanical and microstructural

properties of AA6056 joints produced by
Friction Stir Welding
P. Cavaliere a,∗ , G. Campanile b , F. Panella a , A. Squillace c
a INFM-Department of “Ingegneria dell’Innovazione”, Engineering Faculty, University of Lecce, Lecce I-73100, Italy
b Alenia Aeronautica, Material Processes and NDI Technology Department, Pomigliano d’Arco (NA), Italy
c Department of Materials and Production Engineering, University of Naples “Federico II”, I-80125 Naples, Italy

Received 20 October 2004; received in revised form 17 February 2005; accepted 6 June 2006

Abstract
The effect of processing parameters on mechanical and microstructural properties of AA6056 joints produced by Friction Stir Welding was
analysed in the present study. Different samples obtained by employing rotating speeds of 500, 800 and 1000 rpm and welding speeds of 40, 56
and 80 mm/min were produced. The mechanical properties of the joints were evaluated by means of microhardness (HV) and tensile tests at room
temperature. Fatigue tests on the welds were carried out by using a resonant electro-mechanical testing machine under constant loading control
up to 250 Hz sine wave loading. The low cycle (LCF) and high cycle (HCF) fatigue tests were conducted in the axial total stress-amplitude control
mode with R = σ min /σ max = 0.1, for all the welding and rotating speeds used in the present study. It was observed that the specimens welded at
56 mm/min showed the best behaviour in the low cycle regime. The microstructural evolution of the material was analysed by optical observations
of the welds cross sections.
© 2006 Elsevier B.V. All rights reserved.

Keywords: Friction Stir Welding; Processing parameters; Fatigue

1. Introduction tion between the tool shoulder and the plate top surface, as well
as plastic deformation of the material in contact with the tool
The Friction Stir Welding (FSW) technology is being targeted [6]. The probe is typically slightly shorter than the thickness of
by modern aerospace industry for high performance structural the work-piece and its diameter is typically slight larger than
applications [1]. If compared to traditional welding techniques, the thickness of the work-piece [7]. The FSW process is a solid
FSW strongly reduces the presence of distortions and residual state process and therefore solidification structure is absent in
stresses [2–4]. FSW technology requires a thorough understand- the weld and the problem related to the presence of brittle inter-
ing of the process and consequent mechanical properties of the dendritic and eutectic phases is eliminated [8]. Some aluminium
welds in order to be used in the production of components for alloys can be resistance welded with an extensive surface prepa-
aerospace applications. For this reason, detailed research and ration due to oxide formation. On the other hand, FSW can
qualification work is required [5]. Based on friction heating at be used to join most Al alloys as the surface oxide is not a
the faying surfaces of two sheets to be joined, in the FSW pro- deterrent for the process and therefore no particular cleaning
cess a tool with a specially designed rotating probe travels down operation is needed prior to welding. In FSW the work-piece
the length of contacting metal plates, producing a highly plasti- does not reach the melting point and the mechanical proper-
cally deformed zone through the associated stirring action. The ties of the welded zone (especially when attention is focused on
localized thermo-mechanical affected zone is produced by fric- heat-treatable light alloys) are much higher compared to those
provided by traditional techniques. In fact, the undesirable low
mechanical properties microstructure resulting from melting and
∗ Corresponding author. Tel.: +39 0832 29 7324; fax: +39 0832 325 004. re-solidification is absent in FSW welds leading to improved
E-mail address: [email protected] (P. Cavaliere). mechanical properties, such as ductility and strength in some

0924-0136/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jmatprotec.2006.06.015
264 P. Cavaliere et al. / Journal of Materials Processing Technology 180 (2006) 263–270

alloys [9–11]. In this way, the welds are characterized by low sion loading (R = σ min /σ max = 0.1). All the mechanical tests were performed up
distortion, lower residual stresses and absence of micro defects to failure.
and then of retained products dimensional stability.
The present work is aimed at the evaluation of mechanical 3. Results and discussion
and microstructural behaviour of AA6056 plates obtained by
employing different FSW parameters. 3.1. Friction Stir Welding studies

2. Experimental procedure FSW is becoming a very effective tool in solving the joining
problems in the aerospace industry where joints high ductility
The material under investigation was a 6056 commercial aluminium alloy and tensile strength are required. In the present work, FSW welds
produced by Pechiney under the form of rolled plates of 4 mm thickness.
of AA6056 sheets were successfully obtained by varying the pro-
200 mm × 80 mm large plates were welded along the rolling direction, and then
subjected to post-weld heat treatment consisting of a former step at 170 ◦ C/6 h, cessing parameters. In Tables 1 and 2, the different parameters,
water quenching and a latter step at 190 ◦ C/13 h and water quenching. Such the recorded forces on the tool and the welded material used in
thermal treatment results in a considerable improvement of the material cor- this work are reported. All these values refer to the regime con-
rosion resistance. Employed rotating speeds of the tools were 500, 800 and ditions. An example of the force evolution with time is shown in
1000 rpm while the welding ones were 40, 56 and 80 mm/min, giving nine dif-
Fig. 2. During the initial stage of welding, higher force values
ferent processing conditions. The machine used for the production of the joints
was instrumented with a Kistler three channel load cell in order to record both act on the material due to tool penetration (FZ ), since the mate-
forces along the tool axis, hereon denoted as FZ , and along the welding direction, rial temperature is still low and consequently its yield strength
hereon denoted as FX , for all the produced welds. Acquisition scan rate changed is high. Only when tool penetration is complete and the travel
as a function of the rotating speed in order to record two time samples per tool motion is not yet started, the softening of material induces a
revolution, in all the examined conditions. The nib (Fig. 1) has a diameter (d) of
drop of the FZ force. FX shows on the contrary an improvement
6.0 mm and is 3.9 mm long (h). A 14 mm diameter shoulder (D) was machined
perpendicular to the tool axis and the tilt angle was set equal to 3◦ . The Vick- with welding speed increase and it varies after a given period
ers hardness profiles of all the welded zones were measured on a cross-section according to the different process parameters, attaining regime
perpendicular to the welding direction using a Vickers indenter with a 200 gf conditions. Fig. 2a and b illustrate the recorded forces in two
load for 15 s. Tensile tests were performed in order to evaluate the mechanical different welding conditions. By focusing the attention on both
properties of the joints obtained in the different welding conditions. The tensile
Tables 1 and 2, it is quite evident how forces raise as the welding
tests were carried out at room temperature using a MTS 810 testing machine
with initial strain rate of 10−3 s−1 . Specimens were sectioned in the perpendic- speed increases.
ular direction along the weld line by employing an electrical discharge machine
(EDM). Some specimens for the microstructural analyses were prepared by 3.2. Mechanical properties
standard metallographic techniques and etched with Keller’s reagent to reveal
the grain structure. A statistical analysis was performed in the nugget based on
light microscopy images. Endurance fatigue tests were performed by a resonant
The tensile properties of the heat-treated 6056 aluminium
electro-mechanical testing machine under constant loading control up to 250 Hz alloys are shown in Table 3. The material room temperature ten-
sine wave loading (TESTRONICTM 50 ± 25 kN, produced by RUMUL, SUI) sile response in all the welding conditions is shown in Fig. 3.
in both low and high regimes. The cyclic fatigue tests were conducted under A strong ductility variation (Fig. 4), calculated as strain to frac-
axial total stress amplitude control mode under fully reversed, push-pull, ten- ture, and yield strength variation (Fig. 5) were observed as a
function of the different welding parameters. The material duc-
tility reaches the highest values at 40 and 56 mm/min, at the

Table 1
Different forces recorded on the tool axis for all the rotating and welding speed
used in the present study
FZ (kN)

500 rpm 800 rpm 1000 rpm

40 4.59 5.52 4.32

Welding speed
56 5.95 6.14 5.19
(mm/min)
80 6.71 6.60 5.62

Table 2
Different forces recorded on the material in a direction perpendicular to the tool
axis for all the rotating and welding speed used in the present study
FX (kN)

500 rpm 800 rpm 1000 rpm

40 1.03 0.99 0.78

Welding speed
56 1.27 1.31 0.90
(mm/min)
80 1.79 1.86 1.50
Fig. 1. Geometry of the tool used in the present study.
 P. Cavaliere et al. / Journal of Materials Processing Technology 180 (2006) 263–270 265

Fig. 2. Forces measured along the tool axis and in the welding direction in two
different welding conditions.

lowest rotating speed (500 rpm); on the contrary, it decreases

strongly as increasing the rotating and the welding speeds. The
highest tensile strength is reached in correspondence of the high-
est rotating speeds (800 and 1000 rpm) at the highest welding
speed used in the present analysis (80 mm/min). The very differ-
ent mechanical behaviour of the FSW joints is also demonstrated
by the strong variation in grain size and distribution observed
by employing optical microscopy.
The fatigue endurance curve of the studied material in the
un-welded condition is shown in Fig. 6.
The S–N curve of the specimens joined at a welding speed of
40 mm/min show the same behaviour in both low and high cycle
regimes (Fig. 7). The specimens welded at a tool rotating speed
of 800 rpm exhibit the highest number of cycles to failure at the
same stress amplitude with respect to the specimens welded at
500 and 1000 rpm.
In Fig. 8 the fatigue endurance curve of the joints obtained
by using a welding speed of 56 mm/min at different rotating
speed is shown. A very similar behaviour is shown in the low
cycle regime by the welds obtained at rotating speeds of 1000 Fig. 3. Tensile response of 6056 aluminium alloys in all the welding conditions
and 800 rpm. In this regime, the sample welded at 500 rpm and used in the present study.

Table 3 56 mm/min of rotation and welding speeds, respectively shows

Tensile properties of the studied alloy a different behaviour, which becomes similar to those exhibited
Alloy Yield strength (MPa) UTS (MPa) Elongation (%) in the high cycle regime. The specimens joined with a welding
6056 T4 240 316 23
speed of 1000 rpm show the best behaviour in the low cycle
regime with respect to the other tool speeds used in the present
266 P. Cavaliere et al. / Journal of Materials Processing Technology 180 (2006) 263–270

Fig. 4. Ductility of the studied alloy as a function of the tool rotating speed for
all the welding speed used in the present study.
Fig. 7. Fatigue endurance curve of the studied alloy for a welding speed of
56 mm/min.

study. The material response changes in the high cycle regime in

which the specimens welded with 500 rpm exhibit higher cycles
to failure at the same stress amplitude.
The fatigue response of the AA6056 Friction Stir Welded
using the highest welding speed is shown in Fig. 9. In this case,
the specimens joined with a rotating speed of 800 rpm show
the best behaviour in the high cycle regime while the worst
behaviour in the high cycle regime is shown by the specimens
welded at 1000 rpm.
For all the welding and rotating speeds used in the present
study, the specimens welded at 56 mm/min show the best
behaviour in the low cycle regime. In Fig. 10 the stress ampli-
tude corresponding to a number of cycles to failure of 105 for
all the welding conditions is shown. The specimens joined with
a welding speed of 56 mm/min showed intermediate values of
Fig. 5. Yield strength variation as a function of the tool rotating speed for all ductility and yield strength, but the most constant one by varying
the welding speed used in the present study. the tool rotating speed.

Fig. 6. Fatigue endurance curve of the studied alloy for a welding speed of Fig. 8. Fatigue endurance curve of the studied alloy for a welding speed of
40 mm/min. 80 mm/min.
 P. Cavaliere et al. / Journal of Materials Processing Technology 180 (2006) 263–270 267

Fig. 9. Stress amplitude variation (at 105 cycles to failure) as a function of the
tool rotation speed for all the used welding speeds.

3.3. Microstructural behaviour

The microhardness profiles of the FSW joints were measured

in the cross sections in order to evaluate the material behaviour
as a function of the different welding parameters. Plots of the
microhardness (Hv1) versus distance from the weld centre are
shown in Fig. 11. The hardness values are higher to one side of
the welds corresponding to the advancing side of the tool. The
profiles appear very uniform for the material joined using the
lowest rotating speed (500 rpm) and the lowest welding speeds,
in particular 40 and 56 mm/min. By using higher rotating and
welding speeds, the material hardness increases in all the exam-
ined conditions and the relative profiles become less uniform
across the weld centre.
The microstructural behaviour of 6056 aluminium alloy
joined by Friction Stir Welding was studied by employing optical
microscopy in all the conditions of welding and rotating speed.
In all the welding conditions, the flow of material inside the

Fig. 11. Microstructure of the centre of the joint AII 1, 4 and 7.

nugget is evidence of substantial plastic stirring during FSW. In

Figs. 12–14 the micrographs of the centre of the nugget zones
of all the studied samples (in cross-sections) are shown. The
microstructure of the materials appears as very fine and equiaxed
grains in all the welding conditions. The mean area of the grains
Fig. 10. Microhardness profiles of the studied joints in all the studied welding was measured by using an image analyser on 100 grains for each
conditions. condition.
268 P. Cavaliere et al. / Journal of Materials Processing Technology 180 (2006) 263–270

Fig. 12. Microstructure of the centre of the joint AII 2, 5 and 8. Fig. 13. Microstructure of the centre of the joint AII 3, 6 and 9.

In such process, the maximum work-piece temperatures are From this point of view, a big effect of rotating speed on
above the solution temperature for the hardening precipitates the maximum hardness levels was observed; for the lower rotat-
and below the melting point of the parent material. ing speed used in the present study, the material reaches lower
The increase in hardness in the nugget zone on the material working temperatures strongly below the solvus one and conse-
suggests that the particles, formed during the previous thermo- quently the microstructure consists of a mixture of coarse and
mechanical operations, redistribute in the structure going into broken particles. By increasing the rotating speed and then the
solid solution and re-precipitation during the welding process material temperature, the volume fraction of coarse second phase
[8,12]. particles is greatly reduced with few remaining soluble particles.
 P. Cavaliere et al. / Journal of Materials Processing Technology 180 (2006) 263–270 269

Fig. 15. Microhardness in the centre of the welds as a function of the average
size of the grains.

Fig. 16. Variation of microhardness in the center of the nugget zone as a function
of grain size for all the rotation speeds analysed.

the nugget zone has a strong influence on the fatigue behaviour

of the studied Friction Stir Welded aluminium alloy.
The microhardness measured in the centre of the welds was
correlated to the average grain size in the same point (Fig. 16).
The behaviour of microhardness with grain size shows that
the strong influence on mechanical behaviour of the different
specimens is due to the microstructural evolution obtained by
employing the different welding parameters.
Fig. 14. Average grain size variation as a function of the tool rotation for all the
used welding speeds.
4. Conclusions

The grain size variation was plotted versus tool rotation speed The microstructural and mechanical behaviour of 6056 alu-
as reported in Fig. 15. On one side, finer grain size in the nugget minium alloy Friction Stir Welded by using three different
region is likely to enhance ductility, on the other side it was welding speeds (40, 56 and 80 mm/min) and three different
observed a very similar behaviour between the grain size varia- tool rotation speeds (1000, 800 and 500 rpm) was studied in
tion in all the studied welding conditions and the variation of the this work. All the forces on the tool and the welded material
stress amplitude at 105 cycles to failure reported in Fig. 10. It were recorded. The tensile tests performed at room temperature
can be concluded that the grain size evolution of the material in showed that the material ductility reaches the highest values
270 P. Cavaliere et al. / Journal of Materials Processing Technology 180 (2006) 263–270

for 40 and 56 mm/min welding speed and the lowest rotating [3] R. John, K.V. Jata, K. Sadananda, Residual stress effects on near threshold
speed (500 rpm), decreasing strongly as increasing the rotating fatigue crack growth in friction stir welded aerospace alloys, Int. J. Fatigue
speed and the welding speed. The highest tensile strength is 25 (2003) 939–948.
[4] K.V. Jata, K.K. Sankaran, J. Ruschau, Friction stir welding effects on
reached in correspondence of the higher rotating speeds (800 microstructure and fatigue of aluminum alloy 7050-T7451, Metall. Mater.
and 1000 rpm) for the highest welding speed used in the present Trans. 31A (2000) 2181–2192.
analysis (80 mm/min). The fatigue endurance curves showed [5] W.D. Lockwood, B. Tomaz, A.P. Reynolds, Mechanical response of friction
very different response of the material as a function of the differ- stir welded AA2024: experiment and modelling, Mater. Sci. Eng. A323
ent processing parameters. The plots of the microhardness (Hv1) (2002) 348–353.
[6] M. Guerra, C. Schmidt, J.C. McClure, L.E. Murr, A.C. Nunes, Flow
versus distance from the weld centre appear very uniform for the patterns during friction stir welding, Mater. Char. 49 (2003) 95–
material joined using the lowest rotating speed (500 rpm) and 101.
the lowest welding speeds, in particular for 40 and 56 mm/min. [7] P. Ulysse, Three-dimensional modeling of the friction stir welding process,
By using higher rotating and welding speeds, the hardness of Int. J. Mach. Tools Manuf. 42 (2002) 1549–1557.
the material reaches higher values in all the conditions and the [8] C.G. Rhodes, M.W. Mahoney, W.H. Bingel, Effects of friction stir weld-
ing on microstructure of 7075 aluminium, Scripta Mater. 36 (1997) 69–
profiles become less uniform across the weld centre. 75.
The microstructure of the materials appears as very fine and [9] Y.S. Sato, M. Urata, H. Kokawa, K. Ikeda, Hall–Petch relationship in fric-
equiaxed grains in all the welding conditions. tion stir welds of equal channel angular-pressed aluminium alloys, Mater.
Sci. Eng. A354 (2003) 298–305.
References [10] P.B. Berbon, W.H. Bingel, R.S. Mishra, C.C. Bampton, M.W. Mahoney,
Friction stir processing: a tool to homogenize nanocomposites aluminum
alloys, Scripta Mater. 44 (2001) 61–66.
[1] W.M. Thomas, E.D. Nicholas, J.C. Needam, M.G. Murch, P. Templesmith, [11] W.B. Lee, Y.M. Yeon, S.B. Jung, The improvement of mechanical prop-
C.J. Dawes, GB Patent Application No. 9125978.8, December 1991 and erties of friction stir welded A356 Al alloy, Mater. Sci. Eng. A355 (2003)
US Patent No. 5460317, October 1995. 154–159.
[2] G. Bussu, P.E. Irving, The role of residual stress and heat affected zone [12] L.E. Murr, G. Liu, J.C. McClure, A TEM study of precipitation and related
properties on fatigue crack propagation in friction stir welded 2024-T351 microstructures in friction-stir-welded 6061 aluminium.
aluminium joints, Int. J. Fatigue 25 (2003) 77–88.