This document summarizes Mitchell Smith's undergraduate research project on friction stir welding of similar and dissimilar metal alloys. Smith conducted trial welds of aluminum alloys to understand the friction stir welding process and machine settings. Welds were made of similar aluminum alloys and dissimilar aluminum-copper alloys. Challenges included preventing separation of metal pieces during welding. Further tests are needed to characterize the welds and properties of the welded metals, including microstructural analysis, tensile tests, and impact tests. Continued research on welding nano-reinforced metals could expand applications of friction stir welding in fields like nuclear engineering.

1. Smith 1
Friction Stir Welding of Similar and Dissimilar Alloys
Department: Mining and Nuclear Engineering
Submitted by: Mitchell Smith
Advisor: Dr. Carlos H. Castaño

2. Smith 2
Abstract
This project studies the effects of welding similar and dissimilar alloys together and comparing
those welds to the original metals. Nano-reinforced alloys are also researched to see if they
could be welded with the FSW and what would need done to do so. Trial welds were made to
understand the programming and mechanics of the friction-stir-weld machine. Welds were
made on a single aluminum sample and then butt joints were created by welding two samples
of aluminum together. For this project, one welding pin was used and the metal welded was
kept to a thickness of 0.25 inches. The largest problem encountered was the separation of the
metal samples when attempting a butt-joint weld. When the spinning bit was forced into the
metal, the two pieces would separate if not clamped well enough. Welds were visually
inspected to verify that they were properly made. Further tests, both destructive and non-
destructive, are needed to quantify what is happening to the metal in the weld.
Introduction
Friction Stir Welding (FSW) is a modern welding process with unknown possibilities. Welding
processes before FSW needed to melt the metal so two pieces could be combined while in its
liquid state; sometimes a filler metal would be added to help strengthen the weld joint. In
traditional welding, the heat source used to melt metal comes from burning of a gas from an
electrical current. The ability to join similar alloy metals together allows the manufacturing of
much stronger structures when compared to riveting or bolting. When welding, the molten
metal can become very reactive with the oxygen in the air around it. This is why many methods
of welding have a shielding process of some kind. Two types of shielding processes are used
with flux and inert gases. The flux is usually built into the filler metal that is added to the weld.
It melts with the metal and remains on the outside of the weld helping to keep the molten
metal free from contaminants. When gas is used to shield a weld, it may contain one or a
combination of gasses, including, but not limited to Carbon Dioxide (CO2) and Argon. These
gasses do not react with the weld; they displace the air around the molten weld long enough
for it to cool and become solid without any imperfections.
Friction Stir Welding does not melt the metal and therefore does not need any type of
shielding. Since no flux is used, the weld requires no cleaning or grinding after it is finished.
The heat is generated from friction coming from a rotating pin that has been forced into the
material. The amount of heat generated is determined by how fast the pin spins (revolutions
per minute, RPM) and how fast the pin is forced into new material (inches per minute, IPM).
Since the needed heat is generated from friction,
there are no harmful gasses present or bright light
from the arcing of electricity. These are just a few
reasons how FSW is safer and cleaner than
conventional welding.
Figure 1: Gap Formed from Under Clamping
(Reverse Side of Weld)

3. Smith 3
The most difficult part of friction stir welding is in the set up. In general, friction stir welders
must maintain large amounts of down force. During the welding of the aluminum, the welder
was left set at the 20,000 pounds of down force. This was more than needed but it did not alter
the quality of the welds. When friction stir welding stainless steel, 40 Kilo Newton’s is required,
which is just under 9000 pounds of force (Friction Stir Welding. 2015, January 1). When joining
two pieces of metal, the large down force can push the two samples apart when the welder
begins if they are not clamped down well enough (see Figure 1). This causes a problem because
FSW does not use a filler metal, so any gap that is formed will cause an imperfection in the
weld. The needed clamping or fixturing for the metal may cause limits on what can be welded
since the welder and the metal must be attached to each other somehow.
Discussion
Friction Stir Welding (FSW), invented by Wayne Thomas at TWI Ltd in 1991, overcomes many of
the problems associated with traditional joining techniques. FSW is a solid-state process which
produces welds of high quality in difficult-to-weld materials such as aluminum, and is fast
becoming the process of choice for manufacturing lightweight transport structures such as
boats, trains and aeroplanes [].FSW has been widely used on all aluminum alloys, especially on
2000, 6000, and 7000 series alloys. One of
the aluminum alloys used for this project
was 5083. The 5000 series alloy has an
increased concentration of magnesium in
the aluminum. The 5083 was available from
the supplier in a size that would require
minimum preparation to friction stir weld.
The first step taken was learning how to
operate the FSW machine located in Fulton
Hall. This was a trial and error learning
process. After reading through the
previously produced manual and making
some trial passes, the correct settings were
established to complete a successful weld. Learning how to
measure and set the angle of the weld pin was important so that the leading side of the pin was
higher than the trailing side. This was a major component of making sure the bit would not
plunge in the weld sample too far (see Figure 3) or completely through and damage the weld
bed that the samples were clamped to. This adjustment, once set, remained at that setting for
all the welds that were done. With the angle of the pin not perpendicular to the metal, the
plunge depth had to be reduced. This measurement was taken with magnetic calipers and a
flat metal surface (see Figure 2). Since the same pin was used for all the welds, this offset also
remained constant for the project. When trying to replicate a process, the more variables that
can be held constant will lead to more consistent results. After research, proper spindle speeds
Figure 3: Plunge
Depth to Deep Figure 2: Measuring Pin
and Shoulder Height

4. Smith 4
and bed speeds were established to reduce any chance of damage to the machine, pin, or weld
sample. The welds were made inside this range with some variation and showed no signs of
imperfection.
With similar alloy welds completed,
the next phase was the joining of
dissimilar alloys. The two dissimilar
alloys were aluminum 5083 and
aluminum 1100. Problems with
welding dissimilar alloys is that
sometimes they do not want to
combine with each other during the
stir process and voids and tunnels are formed (figure 4). These voids can be affected by the
spin speed of the pin and how fast the pin is traveling down the weld. Both speeds affect the
temperature of the metal during the weld. With variation of the speeds the tunneling can be
reduced and eliminated, in some cases. Further research is being done in which aluminum is
being welded to copper and magnesium (Mubiayi, M., & Akinlabi, E. 2013).
Other welds that still need completed are to alloys that have been reinforced with ceramic
nano-particles. Ways to weld these materials are not well known yet. The particles do not
remain homogenized, evenly mixed, in the metal if it is melted. Homogenization is needed to
keep the nano particles distributed evenly throughout the metal. These metals are being
researched for use in many fields of engineering. Oak Ridge National Lab has been researching
how to join oxide dispersion strengthened (ODS) steels, nanostructured ferritic alloys (NFAs),
reduced-activation ferritic/martensitic (RAFM) steels, and dissimilar metal joining between
ODS/NFAs and RAFM steels through friction stir welding technology for use in possible future
fusion reactors (Sokolov, M., & Tan, L. 2103, January 1). Fusion reactors are not the only
application where the nuclear field could benefit from these nano reinforced metals. When
building current fission reactors, and any power plant in general, higher burn temperatures lead
to high efficiencies. Current burn temperatures are limited to what the materials that contain
the burning material can withstand. Materials in the nuclear engineering also have to stand up
to radiation damage along with thermal fatigue. If nano-particle strengthened materials can
maintain their material properties better after irradiation than what is currently available, then
that is another reason nuclear engineers should be researching them.
With all completed welds, tests are needed to verify what the material properties of the metals
are after it has been welded. The highest achievement of any weld process is to have the
welded material to maintain the original properties as it did before the weld. There are many
non-destructive ways to check welds for defects and to see how the weld has formed.
When examining welds the first steps are by visual inspection. This is done to check for
penetration of the weld, possible expansion or separation of weld joint, void formation, and
any other obvious defect. The next step would be to use a microscope to look at cross sections
of the weld. Small samples are cut from the weld; care must be taken to limit heat input so not
Figure 4: Tunneling in Dissimilar Weld Joint

5. Smith 5
to alter the metal from extra heat input. These samples are prepared through a sanding and
polishing process to ensure that the different parts of the weld can be seen.
This figure is a general
representation of what would
be seen from a cross section of
the weld under five to ten times
magnification. The four main
areas of the weld are
represented here (see Figure 5).
Increasing the magnification to
twenty and up to one hundred
times magnification along with the use
of scanning electron microscopes (SEM), grain boundaries can be identified in the metal. These
are important because the grain boundaries are what form between the crystallites of the
metal as it cools. This type of inspection would be the best way to see how dissimilar alloys mix
together after being friction stir welded. The images below are grain boundaries from
aluminum 5083 (left) and copper (right). The structures can be difficult to identify but they are
different. This microstructure/nanostructure that develops can be used to see how the
material was formed or altered. The grain boundaries form when the crystallites are not
aligned properly; the larger the areas between boundaries, the more molecules are arranged
uniformly. Some metals after forging and welding processes, go through various heat
treatments to try to relieve internal stresses and to strengthen the metal.
Figure 7: Grain Structure of Pure Copper
Welds are often checked with X-ray (Radiographic) and Ultrasonic inspection. Both of these
methods allow visualization of the interior of the weld. Microscopic cracks, holes, and other
voids become visible with these tests. A good weld will have consistent characteristics
throughout the weld; the variations in the images are the imperfections. This non-destructive
test can save money and time by testing materials before any destructive testing is done. Even
the smallest imperfection inside the weld can drastically affect how it will perform in a stress
250 Microns
Figure 5: Animated Crossection of a Friction Stir Weld
Figure 6: Grain Structure of Aluminum 5083

6. Smith 6
test. If a weld is found to have a defect then it can be discarded and a new one produced
before any more testing is completed.
A tensile test is a stress test that is used to see how far a
material will stretch before fracturing. The material goes
through plastic deformation which is when the bonds between
atoms that make up the material are stretched past their
rebound point and are unable to return to their previous
configuration(Elastic/Plastic Deformation, n.d.). The difference
in lengths is the amount of plastic deformation that the
material can withstand before fracturing (see Figure 8). A
fracture is a crack the forms when the material is stretched and
propagates until the material splits apart (Kopeliovich, D. 2014,
March 5). With the friction stir welder not melting the metal it maintains more of its tensile
property. A simple test would be to see if a friction-stir-welded sample and an unwelded
sample would stretch the same amount. If they had similar results the next step would be to
examine the crystal structure and the grain boundaries at the surface of the fracture.
Another test is the Charpy test, which is an impact test to see how
much energy is absorbed by the test sample when it fractures
(Friction Stir Welding. 2015, January 1). This test helps to determine
the toughness of a material. Toughness is the ability to absorb
energy and plastically deform without fracturing. Materials that
break into two pieces have a low toughness and material that bends
without completely breaking has a high toughness. Testing both
parent materials and comparing them to the friction-stir-welded
sample would show if any toughness was gained or lost from the
weld.
As material research continues to grow in all fields of study, metal
production and metal joining processes will remain an important
field so that technology may continue to grow. Friction stir welding
is a proven and needed way to join the metals of today and the
future. Friction stir welding allows for materials that were once very challenging to weld to be
welded faster and with higher quality than before. Materials that were thought to be
unweldable are beginning to be joined with friction-stir-welding methods. The more scientists
and engineers are able to understand and use different and new materials, the further and
faster technology will be able to grow.
Figure 8: Tensile Test
Figure 9: Charpy Test

7. Smith 7
Acknowledgements
I would like to give credit and thank all the individuals who helped me through this project.
Although my progress was not what I had originally hoped and planned for, I have learned a lot
about the experimental and research process.
Dr. Castaño:
Thank you for allowing me to step into this OURE project that was already established. This
opportunity has given me some real life experience on conducting research outside the
classroom. You have provided me with this opportunity and the lessons I learned from it will be
invaluable as I further my education to become a Nuclear Engineer. Your vast interests in
material science have opened my eyes to some of what is currently being researched and
developed.
Gretchen Smith:
To my wife, thank you for encouraging me to take on this project and to continue to try new
things. You have done your part to keep me on track at school and have been willing to help
when I needed you too. Thank you for the support you have given me since I started my
engineering studies.
Michael Hall and Charlie Moore:
To the both of you, thanks for helping me do the dirty work of cleaning and preparing a place to
work and for assisting me with the welding itself. I am glad I was able to meet you guys and
work with you. I have always been able to work much better in a group than by myself and
your presence allowed me to do my best.
Dedie Wilson:
Thank you for allowing me to step into this project and to continue where it left off. I
appreciate that you take part in providing undergraduate students a chance to do experiments
and research outside the confinements of the classroom. Thank you for always having such
quick responses to emailed questions that I have had throughout this project.

8. Smith 8
References
Casati, R., & Vedani, M. (2014). Metal Matrix Composites Reinforced by Nano-Particles—A
Review. Metals, (4), 65-83.
Elastic/Plastic Deformation. (n.d.). Retrieved April 1, 2015, from https://www.nde-
ed.org/EducationResources/CommunityCollege/Materials/Structure/deformation.htm
Feng, Z., Yu, Z., Hoelzer, D., Sokolov, M., & Tan, L. (2103, January 1). Friction Stir Welding of ODS
Steels and Advanced Ferritic Structural Steels. Retrieved April 1, 2015, from
http://web.ornl.gov/sci/physical_sciences_directorate/mst/fusionreactor/pdf/Vol.54/2.5_Feng.
pdf
Friction Stir Welding. (2015, January 1). Retrieved April 1, 2015, from http://www.twi-
global.com/capabilities/joining-technologies/friction-processes/friction-stir-welding/
Kopeliovich, D. (2014, March 5). Materials Engineering. Retrieved April 1, 2015, from
http://www.substech.com/dokuwiki/doku.php?id=fracture_toughness
Miao, P., Odette, G., et al (2007). The microstructure and strength properties of MA957
nanostructured ferritic alloy joints produced by friction stir and electro-spark deposition
welding. Journal of Nuclear Materials, 1197-1202. Retrieved April 1, 2015, from
http://www.ncnr.nist.gov/programs/sans/pdf/publications/0598.pdf
Morishige, T., Kawaguchi, A., Tsujikawa, M., Hino, M., Hirata, T., & Higashi, K. (2008). Dissimilar
Welding of Al and Mg Alloys by FSW. Materials Transactions, 49(5), 1129-1131.
Mubiayi, M., & Akinlabi, E. (2013). Friction Stir Welding of Dissimilar Materials between
Aluminium Alloys and Copper - An Overview. Proceedings of the World Congress on
Engineering, 3(WCE 2013).
Figure 5: Kallee, S., & Nicholas, D. (2003, January 1). Friction Stir Welding at TWI. Retrieved April
1, 2015, from http://materialteknologi.hig.no/Lettvektdesign/joining methods/joining-welding-
friction stir weld.htm
Figure 6: AluMATTER Grain Size. (2010, January 1). Retrieved April 1, 2015, from
http://aluminium.matter.org.uk/content/html/eng/default.asp?catid=208&pageid=214441660
5
Figure 7: Resources: Standards & Properties - Copper & Copper Alloy Microstructures: Coppers.
(2015, January 1). Retrieved April 1, 2015, from
http://www.copper.org/resources/properties/microstructure/coppers.html

9. Smith 9
Figure 8 and 9: Friction Stir Welding. (2015, January 1). Retrieved April 1, 2015, from
http://www.twi-global.com/capabilities/joining-technologies/friction-processes/friction-stir-
welding/