Industrial Robot: An International Journal

Robotic friction stir welding

George E. Cook Reginald Crawford Denis E. Clark Alvin M. Strauss
Article information:
To cite this document:
George E. Cook Reginald Crawford Denis E. Clark Alvin M. Strauss, (2004),"Robotic friction stir welding", Industrial Robot: An
International Journal, Vol. 31 Iss 1 pp. 55 - 63
Permanent link to this document:
http://dx.doi.org/10.1108/01439910410512000
Downloaded on: 31 January 2016, At: 17:59 (PT)
References: this document contains references to 18 other documents.
To copy this document: permissions@emeraldinsight.com
The fulltext of this document has been downloaded 1430 times since 2006*
Users who downloaded this article also downloaded:
Antoine Bres, Bruno Monsarrat, Laurent Dubourg, Lionel Birglen, Claude Perron, Mohammad Jahazi, Luc Baron,
Downloaded by RMIT University At 17:59 31 January 2016 (PT)

(2010),"Simulation of friction stir welding using industrial robots", Industrial Robot: An International Journal, Vol. 37 Iss 1 pp.
36-50 http://dx.doi.org/10.1108/01439911011009948
Jeroen De Backer, Anna-Karin Christiansson, Jens Oqueka, Gunnar Bolmsjö, (2012),"Investigation of path compensation
methods for robotic friction stir welding", Industrial Robot: An International Journal, Vol. 39 Iss 6 pp. 601-608 http://
dx.doi.org/10.1108/01439911211268813
Yoshihiro Kusuda, (2013),"Honda develops robotized FSW technology to weld steel and aluminum and applied
it to a mass-production vehicle", Industrial Robot: An International Journal, Vol. 40 Iss 3 pp. 208-212 http://
dx.doi.org/10.1108/01439911311309889

Access to this document was granted through an Emerald subscription provided by emerald-srm:393177 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service
information about how to choose which publication to write for and submission guidelines are available for all. Please visit
www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of
more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online
products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication Ethics
(COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation.

*Related content and download information correct at time of download.


Research article
Robotic friction stir
welding
George E. Cook
Reginald Crawford
Denis E. Clark and
Alvin M. Strauss
The authors
George E. Cook is a Professor of Electrical Engineering at
Vanderbilt University, Nashville, Tennessee, USA.
Reginald Crawford is a Graduate Student at Vanderbilt
University, Nashville, Tennessee, USA.
Downloaded by RMIT University At 17:59 31 January 2016 (PT)
Denis E. Clark is with the Industrial and Material
Technologies Department, Idaho National Engineering and
Environmental Laboratory, Idaho Falls, Idaho, USA.
Alvin M. Strauss is a Professor of Mechanical Engineering at
Vanderbilt University, Nashville, Tennessee, USA.
Keywords
Friction welding, Robotics
Abstract
The forces and torques associated with friction stir welding
(FSW) are discussed as they relate to implementation of the
welding process with industrial robots. Experimental results
are presented that support the conclusions drawn from
models developed by others. It is shown that even with
heavy-duty industrial robots with high stiffness, force
feedback is important for successful robotic FSW. Methods of
implementing force feedback are reviewed. Attention is paid
to stability issues that arise with variations in tool rotation
and travel speed. Successful implementations of robotic FSW
are cited.
Electronic access
The Emerald Research Register for this journal is
available at
www.emeraldinsight.com/researchregister
The current issue and full text archive of this journal is
available at
www.emeraldinsight.com/0143-991X.htm
Industrial Robot: An International Journal
Volume 31 · Number 1 · 2004 · pp. 55–63
q Emerald Group Publishing Limited · ISSN 0143-991X
DOI 10.1108/01439910410512000
55
Introduction
Friction stir welding (FSW) is a relatively new
welding process invented by The Welding
Institute (Thomas et al., 1991). In the relatively
short time since its invention, the new welding
process has found potential applications in a
number of industries including: aerospace
(military/civilian aircraft, aircraft parts, fuel
tanks, rockets), land transportation (tailored
blanks, truck bodies, armor plate vehicles,
wheel rims, engine and chassis cradles, fuel
tankers, motorcycle and bicycle frames), railway
(tankers and wagons, container bodies,
underground carriages and trams), shipbuilding
and marine (panels for decks, sides, bulkheads
and floors, helicopter landing pads, offshore
accommodation, hulls and superstructures,
aluminum extrusions), construction (aluminum
bridges, window frames, aluminum pipelines,
heat exchangers, facade panels), electrical
(bus bars, electrical connectors, electric motor
housings, encapsulation of electronics), and gas
(tanks and cylinders).
Unlike fusion welding processes, e.g. arc
welding, electron beam welding, and laser
welding, the FSW process takes place in the
solid phase below the melting point of the
materials being joined. Advantages that have
been cited for the process include: ability to
weld alloys that are difficult to weld by fusion
welding processes, excellent mechanical
properties, low distortion and shrinkage, no
fume, porosity or spatter (frequently associated
with arc welding), energy efficient, and ability to
be used in all positions. Additionally, FSW uses
a non-consumable tool, requires no filler wire,
or gas shielding, and is tolerant of thin oxide
layers.
Because of the large forces and torques
associated with FSW, work pieces must be
rigidly clamped, and, for singled-sided tools, a
backing bar is required. The large forces and
torques also call for machine tool type
equipment for implementation. This is an
Reginald Crawford has recently been awarded an
American Welding Society Fellowship for the
academic year 2003-2004. Information and
photographs pertaining to practical implementations
of robotic friction stir welding were obtained from
Christopher B. Smith and John F. Hinrichs of
Friction Stir Link, Waukesha, WI.
 Robotic friction stir welding Industrial Robot: An International Journal
George E. Cook et al. Volume 31 · Number 1 · 2004 · 55–63

advantage in which conventional machine tools, Figure 1 The FSW process

such as milling machines, can be used to apply
the process. However, this also leads to a
current drawback, which is that most
applications are presently constrained to
two-dimensional planes by such machine tools.
FSW applications can be expected to expand
substantially when industrial robots are used
more in place of currently employed heavy-duty
machine tool equipment. This will permit
welding three-dimensional contours, with
the robots offering the advantages of greater
flexibility and availability, and relatively low cost.
This paper examines the forces and torques
associated with the FSW process with respect to Plate 1 Friction stir weld in progress with tool in horizontal
considerations necessary for robotic orientation and test plate vertical
implementation. It is shown that force control is
Downloaded by RMIT University At 17:59 31 January 2016 (PT)

an important requirement of robotic FSW.

As FSW technology matures, control
requirements are expected to increase in much
the same way as they have with arc welding
robots. This is expected to be particularly true
as improved means of starting and stopping the
weld are developed.

FSW

The FSW process is shown in Figure 1.

The welding tool is comprised of a shoulder in
contact with the work piece and a pin, which in
butt welds extends almost completely through
the work piece thickness. As the tool rotates,
frictional heat is produced at the contacting
shoulder and pin surfaces causing the work
piece material to soften into the materials plastic
state and flow around the tool pin from the front
to rear, where it consolidates and forges under
the confining pressure of the tool shoulder
regains strength upon cooling to form the weld
(Plate 1). No fusion occurs. Most applications
currently involve the welding of aluminum
alloys, although FSW of steel, titanium, and
dissimilar material combinations has been
demonstrated. All the parameters and forces
and torques cited in this paper refer to
aluminum welding only.
The FSW tool pin may be smooth, threaded,
or take on a number of other configurations
(Thomas and Dolby, 2003) designed to
enhance material flow vertically and provide
improved consolidation of the material at the
56
bottom of the weld. The tool shoulder, whose
diameter is relatively large compared to the pin,
acts to prevent material from being expelled
from the weld zone. This minimizes the
formation of voids in the weld zone. A tool has
also been patented (Campbell et al., 2001) that
utilizes two independently actuated shoulders at
both top and bottom of the weld. With its
independently actuated shoulders, this tool is
able to accommodate varying material
thickness.
The rotational speed of the FSW tool may
vary, with application, from a few hundred
revolutions per minute to several thousand
revolutions per minute. The forces and torques
acting on the tool are shown in Figure 2.
The axial force (Fz) required to counteract the
pressure formed in the flowing plasticized weld
zone may vary from about 1 to 15 kN.
The mechanical power input to the rotating tool
is typically of the order of 1.5-4 kW.
The translational force (Fx) in the direction of
 Robotic friction stir welding Industrial Robot: An International Journal
George E. Cook et al. Volume 31 · Number 1 · 2004 · 55–63

Figure 2 Forces and torques associated with FSW Figure 3 Force feedback control implemented as an outer force control loop
around the ordinary position control system of a robot manipulator

travel may vary from a very low value (or

position (with respect to the wrist frame) of the
negative value when the rotating tool actually
robot is modified as required to maintain the
pulls itself along by the friction generated along
desired axial force set by the outer control loop.
the returning side) to about 1 kN, and the
This approach is attractive because it does not
transverse force (Fy), acting toward the
require access to the basic position control loop
Downloaded by RMIT University At 17:59 31 January 2016 (PT)

advancing side of the pin at an angle of 908 to

of the robot. Stability of this scheme will depend
the direction of travel, varies typically from
largely on the indentation characteristic of the
about 0.2 to 2 kN.
rotating tool as it acts against the plasticized weld
For robotic implementations of FSW, the tool
zone material. Most force control schemes
rotation is provided by a separate tool rotation
assume a linear elastic environment. However,
head attached to the robot’s end plate. Both
in FSW the tool/work piece environment is
electric and hydraulic drives have been used for
non-linear, non-elastic, and a function of the
the rotation head. Beyond simple positioning of
welding parameters, e.g. tool rotation speed and
the weld tool, the major requirement of the robot
travel speed. This has not been found to be a
is to provide the necessary axial force (Fz).
major problem, provided that the force control
loop is made inactive during the start and stop
Force control portions of the weld. As will be discussed in a
later section, the plunge force at weld start, for
As stated earlier, a machine used for FSW may example, may rise initially to three to five times
be called upon to deliver and consistently the weld value.
maintain an axial force of 1-15 kN, depending The control scheme shown in Figure 3
on the tool dimensions and welding parameters. assumes an independent force sensing means at
With a precision heavy-duty machine tool, such the weld head. An approach that does not
as a milling machine, this can be done by simply require direct sensing of the axial force is shown
moving the rotating tool into the work piece in Figure 4. Here, use is made of the Jacobian
until the desired axial force is obtained, and relationship between actuator torques and force
then initiating the traverse to make the weld. given by (Craig, 1989),
With the rigidity and precision of the machine
tool, the axial force will be reasonably
Figure 4 Force feedback control with the force derived from the actuator
maintained over the extent of the weld. With a
torques of a manipulator arm
robotic implementation, one cannot assume
such rigidity. At the forces required, there is
usually enough compliance in the manipulator
arm that force feedback becomes a necessity.
A force control scheme that has been
successfully used for this application is shown in
Figure 3. An “outer” force control loop is closed
around the “inner” position control loop of the
robot manipulator (De Schutter and
Van Brussel, 1988). The programmed z-axis
57
 Robotic friction stir welding
George E. Cook et al.
F ¼ J t; ð1Þ
where F is a 6 £ 1 Cartesian force-moment
vector acting at the end-effector, J is the 6 £ 6
manipulator Jacobian, and t is a 6 £ 1 vector of
actuator torques. This approach has been used
(Smith, 2000) and demonstrated to work
adequately at rotation speeds of 1,500 rpm, or
greater, however, the update time was limited to
2 Hz because of the computational burden of
computing the manipulator Jacobian.
Forces and torques
Since FSW was patented (Thomas et al., 1991),
a number of models have been developed to
enhance our understanding of the process.
Downloaded by RMIT University At 17:59 31 January 2016 (PT)
Recently models based on numerical solutions
have been presented by Seidel and Reynolds
(2003) and Ulysse (2002). Seidel and Reynolds
present a two-dimensional model based on fluid
mechanics theory. Ulysse presents a three-
dimensional model based on visco-plastic
modeling theories. These models have been
shown to correlate well with the experimental
results of material flow, temperature profiles,
and force/torque measurements. Earlier, Nunes
et al. (2000) proposed a “plug” model in which
material is transported around the pin as it
translates through the work piece by means of a
“last in first out” wiping action. While the
Nunes et al. model does not contain the
modeling detail of either of the numerically-
based models, its simple form leads to a good
qualitative understanding of parameter
relationships as needed to establish the best
operating conditions for robotic welding, and it
is reviewed below.
The Nunes et al. (2000) model represents the
metal in the vicinity of the rotating tool as a
superposition of two flows:
(1) a primary flow consisting of a plug of metal
rotating with the pin tool and shearing over
a cylindrical surface (assuming a
cylindrically shaped pin); and
(2) a secondary flow driven by threads on the
pin (assuming a threaded pin design) and
resembling a vortex ring around the pin.
The metal transport around the pin takes place
by a wiping action on the shearing surface of the
plug (Nunes, 2001).
58
Industrial Robot: An International Journal
Volume 31 · Number 1 · 2004 · 55–63
The flows associated with the rotating plug
model (Nunes et al., 2000) are shown in
Figure 5. The thickness of the inner volume of
the rotating plug is shown greatly exaggerated
relative to the experimental observations
(Colligan, 1999). Hence, for purposes of
developing a simplified understanding of the
forces and torque acting on the rotating tool, the
rotating plug may be assumed to have the same
dimensions as the pin-tool itself. Furthermore,
the secondary flow may be neglected for a
first-order analysis.
With these assumptions, the torque acting on
the pin-tool is computed as a function of the
forces acting between the rotating plug shearing
with respect to the adjacent metal. If we let the
rotating plug be taken approximately as the tool/
work piece contact surface and include the
shoulder contact as well as contact at the bottom
of the pin, the torque on the plug rotating
against boundary shear stress s is approximately
given by (Nunes et al., 2000),
Torque ¼ Torque on shoulder
þ Torque on pin sides
þ Torque on pin bottom
Z Z r
R
M¼ 2
2p r s dr þ 2p r t s þ 2p r 2 s dr
r 0


2pR3 r2t
¼ 1þ3 3 s ð2Þ
3 R
where M is the torque, r the pin-tool radius,
R the shoulder radius, t the pin-tool depth, and
s the shear flow stress.
The mechanical power input to the tool is
essentially given by,
P ¼ MV ð3Þ
where V is the angular velocity of pin-tool.
Figure 5 Depiction of flows associated with rotating plug
model (Nunes et al., 2000) of FSW
 Robotic friction stir welding Industrial Robot: An International Journal
George E. Cook et al. Volume 31 · Number 1 · 2004 · 55–63

There are other mechanical power inputs, Figure 6 Tool design used in experiments
principal among which is the translational
power, but these are of a lower order of
magnitude and can be neglected. Substituting
the torque given in equation (2) into equation
(3), the power input can be expressed as,


2pR3 r2t
P¼ 1þ3 3 sV ð4Þ
3 R

The mechanical power input is balanced by the

9.53 mm, and the pin diameter and length are
thermal loss Qh and the plastic work Qp. Qp may
4.06 and 2.79 mm, respectively. The
be neglected in comparison to Qh (Nunes et al.,
experiments were conducted on a Willis vertical
2000) as it typically represents less than
CNC milling machine modified for FSW. The
10 percent of the total work. Qh is comprised of
forces and torques acting on the tool pin were
the following components:
measured with a Kistler Model 9124B rotating
(1) conduction loss to the work piece,
cutting force dynamometer. The samples used
(2) conduction loss to the tool,
for the experiments were 305 mm in length by
Downloaded by RMIT University At 17:59 31 January 2016 (PT)

(3) conduction loss to the anvil, and

76.2 mm wide and 6.4 mm thick. All welds were
(4) convection heat loss as the tool moves from
made bead on plate. The material was 6061-
hotter to colder metal.
T651. Further details of the experimental set up
Each of these is a direct function of the shear can be found in the work of Cook et al. (2003).
zone temperature, which varies only a small The spindle power is plotted in Figure 7
amount over a wide range of operating versus the tool rotation speed with travel speed
conditions (Nunes et al., 2000; Song et al., as a parameter. While the power varies, the
2003). Therefore, the power input remains variation is relatively small for the 150 percent
approximately constant over a wide range of change in tool rotation speed. The power
spindle speeds. remains more nearly constant at the lower travel
If the power input is approximately constant, speed, changing only 23 percent for the large
equation (4) shows that an increase in spindle change in tool rotation speed.
speed V must be balanced by a decrease in shear The axial force (Fz) is plotted in Figure 8
flow stress s. All the forces (axial, translational, versus the tool rotation speed with travel speed
and transverse) and the input torque as a parameter. As predicted by the models,
(equation (1)) vary directly as the shear stress
(Nunes et al., 2000). Therefore, Nunes et al. Figure 7 Spindle power vs tool rotation speed
(2000) conclude that a reduction in the critical
axial force could be achieved by an increase in
the spindle speed. Ulysse (2002) reached the
same conclusion based on his three-dimensional
visco-plastic numerical model. It has not yet
been established, however, how far one can go
in load reduction by increasing the spindle
rotation rate. However, as will be seen later in
the experimental results, early experiments
certainly support these modeling observations.
To test the modeling results, a matrix of
experiments were conducted with spindle
rotation speeds of 1,000, 1,500, 2,000, and
2,500 rpm, and travel speeds of 3.4, 6.8, and
13.6 mm/s. Each experiment was run twice and
the results were averaged. The tool used is
shown in Figure 6. The shoulder diameter is
59
 Robotic friction stir welding
George E. Cook et al.
Figure 8 Axial force vs tool rotation speed
Downloaded by RMIT University At 17:59 31 January 2016 (PT)
the axial force decreases as the spindle speed
increases. Based on the travel speeds of 3.4 and
6.8 mm/s, the axial force appears to be
converging to about 3 kN as the rotation speed
increases. Perhaps the higher travel speed of
13.6 mm/s is converging similarly, but data at
higher rotation speeds would be required to
confirm this. It is noted that the simplified
model of Nunes et al. (2000) only shows that
a reduction in axial force may be achieved by
an increase in spindle speed. It does not show
the limiting range of this relationship.
The axial force (Fz) versus travel speed is
shown in Figure 9. For all cases of tool rotation
Figure 9 Axial force vs travel speed
60
Industrial Robot: An International Journal
Volume 31 · Number 1 · 2004 · 55–63
speed, the axial force increases as the travel
speed increases.
The conclusions that can be drawn from
these results is that the axial force requirements
of a robotic FSW implementation can be
reduced by the use of a high tool rotation speed.
A relatively low travel speed is also required.
The axial force requirements will be different for
other tool geometries, so the only general
conclusion that can be reached for robotic
implementations is that high tool rotation
speeds and relatively low travel speeds should
result in the best conditions for minimum axial
force requirements on the robot.
Indentation characteristic
As part of the experimental effort to determine
the tool/work piece indentation characteristic,
experiments were conducted in which the axial
position was changed (in a direction vertically
into the plate) in sudden 0.05 mm steps and the
corresponding change in the axial force was
measured. The results of one such test is shown
in Figure 10. This figure shows the axial force vs
change in axial position for a weld made with the
tool design shown in Figure 6. The tool rotation
speed was 700 rpm and the travel speed was
0.85 mm/s The axial force change shown
resulted from a 0.05 mm change in axial
position. It can be seen that the force changes
quickly from about 2.85 to 3.5 kN and then
Figure 10 Axial force change for step change in position
 Robotic friction stir welding
George E. Cook et al.
settles back to nearly its starting value in about
14 s. This is due to the tool (both the pin and
shoulder) simply plunging into the material as
the position is increased, resulting in excess flash
displaced by the shoulder, but no significant
increase in force other than in the initial
transient. With respect to robotic force feedback
control, the transient characteristic should not
present a problem as long as the update period of
the controller is sufficiently short. The transient
could be expected to result in stability problems,
however, with an update period measured in
seconds. It is noted that this was the case for the
previously cited scheme of using the Jacobian
relation between the actuator torques and
Cartesian forces in lieu of direct force sensing
when the update time was 2 Hz (Smith, 2000).
Downloaded by RMIT University At 17:59 31 January 2016 (PT)
Another observation that can be drawn from
Figure 10 is that the plasticized weld zone of the
work piece represents a rather stiff environment
for the force-controlled robot. The change in
axial force per change in axial position is
approximately 13 kN/mm for the tool geometry
and parameters shown.
In many current FSW applications, the
beginning and end of the weld are made on start
and stop tabs that may be removed later. This
permits the weld to be started in a pre-drilled
hole for the pin, and it allows a hole at the end of
the weld where the pin is withdrawn. Thus,
force control is unnecessary during the start and
stop periods. As FSW techniques evolve,
however, we can expect an interest in
controlling the weld during the start and stop
portions, as well as the weld itself. For example,
one technology that has been invented is a
retractable pin tool (Ding and Oelgoetz, 1999).
This tool permits the pin extension to be
adjusted during welding to accommodate
varying thickness materials, and it may be
gradually withdrawn at the end of the weld to
eliminate or reduce the ending hole. As this
technology is perfected, it might become
desirable to control the force throughout the
ending period.
During the plunge of the pin into the material
at weld start (assuming a predrilled starting hole
is not used), the axial force may rise initially to
three to five times the weld value depending on
the specific weld parameters, tool shape, and the
plunge rate (Cook et al., 2003). With such a
wide variation in the tool/work piece
61
Industrial Robot: An International Journal
Volume 31 · Number 1 · 2004 · 55–63
indentation characteristic, stability issues may
arise in the force feedback control, if force
control is attempted during the weld startup.
This might call for more advanced control
methods including adaptive control to
accommodate the widely varying
“environment”. Similar advances were required
in arc welding as starting and stopping
technology matured with these processes.
For example, in gas tungsten arc welding, it is
sometimes desired to maintain automatic
voltage control throughout the weld, including
the start and stop portions. During the startup,
however, the current may be ramped from a very
low value to the weld value. In this case, the
“environment” parameter is the arc voltage vs
arc length characteristic, which changes
substantially in a non-linear manner as a
function of current. As a result, stability
problems have frequently arisen in critical arc
welds during the startup and end of the weld.
Adaptive automatic voltage control systems
have been created to overcome these critical
start/stop issues in arc welding (Koseeyaporn
et al., 2000), and similar control advancements
may become necessary in FSW, as the
technology matures.
Robotic implementations
Provided that the welding procedures do not
call for widely varying operating parameters,
stable robust force feedback control can be
readily implemented with properly sized
industrial robots. For example, Smith et al.
(2003) have reported a robotic FSW system
based on the use of an ABB IRB 7600
articulated robot arm, as shown in Plate 2. This
robot has a 500 kg payload. The spindle and
motor drive for rotating the pin tool are attached
to the end plate of the manipulator. The spindle
is designed to accept the forces (both axial and
radial) required for FSW, and it additionally
contains force sensing for force feedback
control. Smith et al. (2003) have shown that the
system is capable of maintaining constant axial
force even under conditions of a programmed
3 mm change in component surface height over
the length of the weld.
A different robot design that offers increased
rigidity is shown in Plate 3. This is an ABB IRB
940 tricept type robot. Its parallel design
 Robotic friction stir welding
George E. Cook et al.
Plate 2 Robotic FSW system (courtesy of Friction Stir Link, Wankesha, WI)
Downloaded by RMIT University At 17:59 31 January 2016 (PT)
Plate 3 Parallel robot designed for high stiffness
(ABB Limited, 2003)
provides an extremely stiff structure that
is capable of a vertical force of 1,300 kg
(ABB Limited, 2003). The high stiffness of this
arm design is derived from the use of three
parallel linear actuators connected to a center
tube in the middle with three additional axes on
the wrist. The working envelope of this type
robot is less than the articulated serial arm, but
62
Industrial Robot: An International Journal
Volume 31 · Number 1 · 2004 · 55–63
for applications where this is not a problem, it
provides excellent rigidity for stable force
feedback control. Strombeck et al. (2000) have
reported the use of a parallel robot for robotic
FSW with force feedback control. They report
excellent results with the system.
Conclusions
Robotic FSW requires force feedback control
to compensate for compliance in the robot
arm as well as compliance in the fixturing.
For most current applications, particularly
those that do not require force control during
the startup and end of the weld, the force
feedback control can be expected to operate
stably and robustly. The axial force
requirements needed to counteract the pressure
formed in the flowing plasticized weld zone
may vary for aluminum from about 1 to 15 kN.
The axial force requirements imposed on the
robot can be reduced by operating at high
tool rotation speed and lower travel speed.
Both articulated manipulator arms and arms
with a parallel design may be used provided
that they are structurally capable of providing
the necessary force. For applications permitting
a relatively limited working envelope, the
parallel design is capable of providing the
greatest arm rigidity.
Robotic FSW systems are available that have
been demonstrated to be capable of welding
three-dimensional contours in various positions
with excellent weld control through the use of
force feedback (Smith et al., 2003; Strombeck
et al., 2000). As advances are made in the
technology relating to starting and ending the
weld, advances in control methods may be
required in much the same way that they have
been required in arc welding.
References
ABB Limited (2003), Zurich, Switzerland. http://www.abb.
com/robots
Campbell, C.L., Fullen, M.S. and Skinner, M.J. (2001), US.
Patent no. 6,199,745.
Colligan, K. (1999), “Material flow behavior during
friction stir welding”, Welding Journal, Vol. 78 No. 7,
pp. 229-s-37s.
Cook, G.E., Smartt, H.B., Mitchell, J.E., Strauss, A.M. and
Crawford, R. (2003), “Controlling robotic friction stir
welding”, Welding Journal, Vol. 82 No. 6, pp. 28-34.
 Robotic friction stir welding
George E. Cook et al.
Craig, J.J. (1989), Introduction to Robotics, Addison-Wesley,
New York, USA, p. 179.
De Schutter, J. and Van Brussel, H. (1988), “Compliant robot
motion, parts I-II”, International Journal of Robotics
Research, Vol. 7 No. 4, pp. 3-33.
Ding, R.J. and Oelgoetz, P.A. (1999), U.S. Patent
no. 5,893,507.
Koseeyaporn, P., Cook, G.E. and Strauss, A.M. (2000),
“Adaptive voltage control in fusion arc welding”, IEEE
Transactions on Industry Applications, Vol. 36 No. 5,
pp. 1300-7.
Nunes, A.C. (2001), “Wiping metal transfer in friction stir
welding”, TMS Annual Meeting, New Orleans.
Nunes, A.C., Bernstein, E.L. and McClure, J.C. (2000),
“A rotating plug model for friction stir welding”,
Presented at 81st American Welding Society Annual
Convention.
Seidel, T.U. and Reynolds, A.P. (2003), “Two-dimensional
friction stir welding process model based on fluid
mechanics”, Science and Technology of Welding and
Joining, Vol. 8 No. 3, pp. 175-83.
Downloaded by RMIT University At 17:59 31 January 2016 (PT)
Smith, C.B. (2000), “Robotic friction stir welding using a
standard industrial robot”, 2nd Friction Stir Welding
International Symposium, CD-ROM.
Smith, C.B., Hinrichs, J.F. and Crusan, W.A. (2003),
“Robotic friction stir welding: the state-of-the-art”,
63
Industrial Robot: An International Journal
Volume 31 · Number 1 · 2004 · 55–63
4th Friction Stir Welding International Symposium,
CD-ROM.
Song, M., Kovacevic, R., Ouyang, J. and Valant, M. (2003),
“A detailed three-dimensional transient heat transfer
model for friction stir welding”, in David, S.A.,
DebRoy, T., Lippold, J.C., Smartt, H.B. and Vitek, J.M.
(Eds), Proc. 6th International Conf. on Trends in
Welding Research, ASM International, pp. 212-7.
Strombeck, A.V., Schilling, C. and Santos, J.F.D. (2000),
“Robotic friction stir welding – tool technology and
applications”, 2nd Friction Stir Welding International
Symposium, CD-ROM.
Thomas, W.M. and Dolby, R.E. (2003), “Friction stir welding
developments”, in David, S.A., DebRoy, T., Lippold, J.C.,
Smartt, H.B. and Vitek, J.M. (Eds), Proc. 6th
International Conf. on Trends in Welding Research,
ASM International, pp. 203-11.
Thomas, W.M., Nicholas, E.D., Needham, J.C., Church, M.G.,
Templesmith, P. and Dawes, C.J. (1991), International
Patent Application no. PCT/GB92/02203 and GB Patent
Application no. 9125978.9.
Ulysse, P. (2002), “Three-dimensional modeling of
the friction stir-welding process”, International
Journal of Machine Tools and Manufacture, Vol. 42,
pp. 1549-57.
 This article has been cited by:

1. Nuno Mendes, Pedro Neto, Altino Loureiro, António Paulo Moreira. 2016. Machines and control systems for friction stir welding:
A review. Materials & Design 90, 256-265. [CrossRef]
2. Jiafeng Wu, Rui Zhang, Guangxin Yang. 2015. Design and experiment verification of a new heavy friction-stir-weld robot for
large-scale complex surface structures. Industrial Robot: An International Journal 42:4, 332-338. [Abstract] [Full Text] [PDF]
3. Martin Kadlec, Roman Růžek, Lucie Nováková. 2015. Mechanical behaviour of AA 7475 friction stir welds with the kissing bond
defect. International Journal of Fatigue 74, 7-19. [CrossRef]
4. K. Wang, F. Léonard, G. Abba. 2015. Dynamic Model Identification of Axial Force in Robotic Friction Stir WeldingÕ. IFAC-
PapersOnLine 48, 1936-1941. [CrossRef]
5. M. Palpacelli, G. Palmieri, L. Carbonari, M. Callegari. 2014. Experimental identification of the static model of the HPKM Tricept
industrial robot. Advanced Robotics 28, 1291-1304. [CrossRef]
6. N. Mendes, P. Neto, M. A. Simão, A. Loureiro, J. N. Pires. 2014. A novel friction stir welding robotic platform: welding polymeric
materials. The International Journal of Advanced Manufacturing Technology . [CrossRef]
7. Mumin Sahin. 2014. Characterization of properties in friction-welded austenitic-stainless steel and aluminium joints. Industrial
Lubrication and Tribology 66:2, 260-271. [Abstract] [Full Text] [PDF]
8. Kathryn J. Dharmaraj, Chase D. Cox, Alvin M. Strauss, George E. Cook. 2014. Ultrasonic thermometry for friction stir spot
welding. Measurement 49, 226-235. [CrossRef]
Downloaded by RMIT University At 17:59 31 January 2016 (PT)

9. B.T. Gibson, D.H. Lammlein, T.J. Prater, W.R. Longhurst, C.D. Cox, M.C. Ballun, K.J. Dharmaraj, G.E. Cook, A.M. Strauss.
2014. Friction stir welding: Process, automation, and control. Journal of Manufacturing Processes 16, 56-73. [CrossRef]
10. Jeroen De Backer, Gunnar Bolmsjö, Anna-Karin Christiansson. 2014. Temperature control of robotic friction stir welding using
the thermoelectric effect. The International Journal of Advanced Manufacturing Technology 70, 375-383. [CrossRef]
11. Lingaraju Dumpala, D. Lokanadham. 2014. Low Cost Friction Stir Welding of Aluminium Nanocomposite – A Review. Procedia
Materials Science 6, 1761-1769. [CrossRef]
12. T.S. Hong, M. Ghobakhloo, W. KhaksarRobotic Welding Technology 77-99. [CrossRef]
13. Nuno Mendes, Pedro Neto, J. Norberto Pires, Altino Loureiro. 2013. Discretization and fitting of nominal data for autonomous
robots. Expert Systems with Applications 40, 1143-1151. [CrossRef]
14. Jeroen De Backer, Anna‐Karin Christiansson, Jens Oqueka, Gunnar Bolmsjö. 2012. Investigation of path compensation methods
for robotic friction stir welding. Industrial Robot: An International Journal 39:6, 601-608. [Abstract] [Full Text] [PDF]
15. Tyler A. Davis, Yung C. Shin, Bin Yao. 2011. Observer-Based Adaptive Robust Control of Friction Stir Welding Axial Force.
IEEE/ASME Transactions on Mechatronics 16, 1032-1039. [CrossRef]
16. Jakob Hilgert, Jorge F. dos Santos, Norbert HuberInvestigation of the Material Shear Layer in Bobbin Tool Friction Stir Welding
187-193. [CrossRef]
17. William R. Longhurst, Alvin M. Strauss, George E. Cook. 2011. The Identification of the Key Enablers for Force Control of
Robotic Friction Stir Welding. Journal of Manufacturing Science and Engineering 133, 031008. [CrossRef]
18. Tyler A. Davis, Yung C. Shin, Bin YaoObserver-based adaptive robust control of friction stir welding axial force 1162-1167.
[CrossRef]
19. W R Longhurst, A M Strauss, G E Cook, C D Cox, C E Hendricks, B T Gibson, Y S Dawant. 2010. Investigation of force-
controlled friction stir welding for manufacturing and automation. Proceedings of the Institution of Mechanical Engineers, Part B:
Journal of Engineering Manufacture 224, 937-949. [CrossRef]
20. William R. Longhurst, Alvin M. Strauss, George E. Cook. 2010. Enabling Automation of Friction Stir Welding: The Modulation
of Weld Seam Input Energy by Traverse Speed Force Control. Journal of Dynamic Systems, Measurement, and Control 132,
041002. [CrossRef]
21. T. Zappia, C. Smith, K. Colligan, H. Ostersehlte, S.W. KalleeFriction stir welding equipment 73-117. [CrossRef]
22. Satish Perivilli, John Peddieson, Jie Cui. 2009. Friction Stir Welding Heat Transfer: Quasisteady Modeling and Its Validation.
Journal of Manufacturing Science and Engineering 131, 011007. [CrossRef]
23. M. Soron. 2008. Towards Multidimensionality and Flexibility in FSW using an Industrial Robot System. Welding in the World
52, 54-59. [CrossRef]
 24. Paul Fleming, David Lammlein, D. Wilkes, Katherine Fleming, Thomas Bloodworth, George Cook, Al Strauss, David DeLapp,
Thomas Lienert, Matthew Bement, Tracie Prater. 2008. In‐process gap detection in friction stir welding. Sensor Review 28:1,
62-67. [Abstract] [Full Text] [PDF]
25. Xin Zhao, Prabhanjana Kalya, Robert G. Landers, K. KrishnamurthyDesign and Implementation of a Nonlinear Axial Force
Controller for Friction Stir Welding Processes 5553-5558. [CrossRef]
26. Mikael Soron, Ivan KalaykovA Robot Prototype for Friction Stir Welding 1-5. [CrossRef]
27. R. Crawford, G. E. Cook, A. M. Strauss, D. A. Hartman, M. A. Stremler. 2006. Experimental defect analysis and force prediction
simulation of high weld pitch friction stir welding. Science and Technology of Welding and Joining 11, 657-665. [CrossRef]
28. Therman Analysis of HP3, a Penetrometer to Measure the Planetary Surface Heat Flow . [CrossRef]
Downloaded by RMIT University At 17:59 31 January 2016 (PT)