Composite Structures 174 (2017) 59–69

Contents lists available at ScienceDirect

Composite Structures
j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / c o m p s t r u c t

A new refill friction spot welding process for

aluminum/polymer composite hybrid structures
Proses pengelasan spot refill friksi baru untuk struktur hybrid aluminium / polimer
komposit
⇑
Hossein Karami Pabandi, Mojtaba Movahedi , Amir Hossein Kokabi
Department of Materials Science and Engineering, Sharif University of Technology, P.O. Box 11365-9466, Azadi Avenue, 14588 Tehran, Iran

a r t i c l e i n f o a b s t r a c t
Article history: A new refill friction spot welding process called Threaded Hole Friction Spot Welding (THFSW) was intro-
Received 16 February 2017
duced to join AA5052 aluminum to short-carbon-fiber-reinforced polypropylene (PP-SCF) composite
Revised 28 March 2017
sheets. The process was based on filling of the pre-threaded hole by melted and re-solidified polymer.
Accepted 20 April 2017
Available online 22 April 2017
The results showed that THFSW was successful to join aluminum to polymer sheets and the hole was
completely filled with melted polymer. Formation of a reaction layer composed mostly of Al, C and O
as well as interlocking between the threaded hole and the re-solidified polymer were recognized as main
Keywords:
Friction spot welding
bonding mechanisms. Maximum shear-tensile strength of the joints reached to 80 percent of the com-
Refill Polymer posite base strength. Moreover, Mechanical strength and fracture energy of the joints increased with
Aluminum enhancement of tool rotational speed. Variation of the joint strength was explored in light of the fracture
Joint strength surface features as well as crystallinity percent of the re-solidified polymer inside the hole.
Fracture energy 2017 Elsevier Ltd. All rights reserved.

1. Introduction
Successful joint between polymers and metallic alloys is a
necessity for the production of light weight polymer/metal hybrid
structures in various industries such as automotive, aeronautics
and shipbuilding industries [1–3]. Together with traditional meth-
ods such as adhesive bonding [4,5] and mechanical fastening [6],
new joining techniques have been patented and developed in
recent years for ‘‘spot” joining of polymers to metallic alloys. Ultra-
sonic staking [7] and Injection Clinching Joining (ICJ) [1,8] are two
methods with relatively the same principles used for spot joining
of polymers and metals. Abibe et al. [1] compared ultrasonic stak-
ing and ICJ methods for joining polyetherimide and AA6082 alu-
minum alloy. Their results indicated that both processes show
similar mechanical properties. However, ICJ had a higher
strength-to-weight ratio due to the presence of hole formed in
the stake. Moreover, ICJ showed longer joining time in comparison
to ultrasonic staking. Abibe et al. [8] also employed ICJ process to
join short-glass-fiber-reinforced polyamide and AA2024 aluminum
alloy. They reported that volume of the rivet head in contact with
the tool system and efficiency of cavity filling are key factors con-
trolling the strength of the joints.
Welding processes have also been used for spot joining of poly-
mers to metals. Friction Stir Spot Welding (FSSW) is a solid state
⇑ Corresponding author.
E-mail address: m_movahedi@sharif.edu (M. Movahedi).
process with high potential for these types of joints. However,
because of leaving an exit-hole at the center of the nugget, the
Fric- tion Spot Joining (FSpJ) process was developed by GKSS of
Ger- many in order to refill the formed exit-hole [9]. FSpJ has
been used successfully for joining of metals to glass and carbon-
fiber- reinforced composites [3,10–13]. Goushegir et al. [3]
utilized FSpJ process for joining AA2024 aluminum and reinforced
poly (pheny- lene sulfide) with 50 vol% carbon fibers (CF-PPS).
They concluded that by enhancement of the tool plunge depth,
the shear fracture load of the joint as well as the extension to
fracture improved due to increase in the joint area at the sheets
interface. Esteves et al. [11] employed FSpJ process for joining
AA6181-T4 aluminum and carbon-fiber-reinforced poly
(phenylene sulfide) composite. Their results showed that tool
rotational speed was the most effec- tive parameter on the joint
strength. In addition, they reported that at high tool rotational
speed (1600 rpm), reduction of the metal viscosity led to
decrease in the heat generated during the process. Therefore, joint
area between the aluminum and polymer sheets at the interface
was reduced.
In the present work, a new friction spot welding process called
‘‘Threaded Hole Friction Spot Welding (THFSW)” was used to join
AA5052 aluminum alloy and short-carbon-fiber-reinforced
polypropylene composite sheets. In this joining process, a pre-
threaded hole in the aluminum sheet is filled by the melted and
re-solidified polymer during the process. Thus, no hole remains
at the joint area after completing the process. In comparison to
ultrasonic staking and ICJ, THFSW does not require an initial

http://dx.doi.org/10.1016/j.compstruct.2017.04.053
0263-8223/ 2017 Elsevier Ltd. All rights reserved.
 H. Karami
H. Karami
Pabandi
Pabandi
et al.
et / al.
Composite
/ Composite
Structures
Structures
174 174
(2017)
(2017)
59–69
59–69
6 6

polymeric stud on the polymer sheet; however, the stud is formed Table 2
in the hole by molten polymer during the process. On the other Chemical composition and mechanical properties of PP-SCF used in this research.

hand, the same as the ultrasonic staking and ICJ, a hole must be Composition of PP-SCF Mechanical properties of PP-SCF
created in the aluminum sheet before joining process. Further- (wt%)

more, in FSSW a keyhole remains at the joint center due to pin pen- PP-Z30S 92 Hardness, Shore D 72
etration, while the suggested joining process is able to refill the PP-g-MA 5 UTS, MPa 35
threaded hole during the process. In FSpJ, a relatively complex SCF 3
mechanical and electronic system is required for joining. Although,
in THFSW, the joint is produced using a simple cylindrical tool Table 3
without pin. After joining process, mechanisms of bonding, as well Physical propert ies of polypropylene Z30S [15]

as the effects of the tool rotational speed on the mechanical Properties Unit Amount
behav- ior of the joints, were studied. 3
Density g/cm 0.9
Melt flow index (MFI) Dg/min 25
6
2. Materials and experimental procedures Coefficient of thermal expansion (CTE) 10 /K 100–150
Thermal conductivity (solid) W/m.K 0.17–0.22
Thermal conductivity (molten) W/m.K 0.16
2.1. Materials
Softening point C 165

As-received AA5052 aluminum alloy with a thickness of 2 mm

was used as the metallic part of the joint. Chemical composition,
mechanical and physical properties of AA5052 aluminum alloy
used in this study are presented in Table 1. Polymeric sheets used
in this research were short-carbon-fiber-reinforced polypropylene
Z30S composite (PP-SCF) with 2 mm thickness. Dimensions of car-
bon fibers were 3 mm in length and 17 mm in diameter. Chem-
ical composition and mechanical properties of PP-SCF composite
and physical properties of polypropylene Z30S are listed in Tables
2 and 3, respectively. Fig. 1 shows distribution of carbon fibers in
Z30S polypropylene matrix.

2.2. Joining process

Aluminum sheet with a threaded hole and composite sheet,

both with demotions of 70 30 mm2 were used for joining pro-
cess. To produce the threaded hole, aluminum sheet was drilled
with 4 mm diameter and then threaded by M4 reamer (with the
aim of increasing the joint strength through formation of mechan-
ical interlocks). A cylindrical tool from heat treated H13 steel com-
posed of a shoulder with a diameter of 20 mm was employed as
the welding tool. Joining process was performed in lap configura-
tion with an overlap of 30 mm, while aluminum (as the upper
sheet) was in contact with the tool shoulder. Tool rotational speeds
of 500, 1000, 1500 and 2000 rpm and dwell time of 5 s were
selected for joining.
Schematic of different steps of new joining process introduced
in this research (THFSW) is illustrated in Fig. 2. At the first step
of the process, rotating tool is located coaxially with threaded hole
on top of the aluminum sheet surface as shown in Fig. 2-a. In the
next step, the rotating tool is moved downward and the shoulder
penetrates 0.3 mm in the aluminum surface (Fig. 2-b). As a result
of contact between the rotating tool and aluminum sheet,
frictional heat is generated. This heat reaches to the
aluminum/polymer interface via conduction and melts the
polymer at the interface.
Table 1
Chemical composition, mechanical and physical properties of AA5052 used in this
research.
Fig. 1. Distribution of carbon fibers in PP-Z30S matrix.
However, since the polymer is a heat insulation material, melting
of the polymer is limited to the surface of the composite sheet at
the interface. Then, by the downward axial pressure of the tool,
melted polymer flows towards the periphery of the joint location
at the interface. The molten polymer around the hole in aluminum
sheet fills the hole and mechanical interlocks are formed between
polymer and aluminum due to the existence of threads inside the
hole (Fig. 2-c and d). After 5 s dwell time, the tool is retracted
upward (Fig. 2-e), and the joint is completed with the solidification
of the molten polymer inside the hole (Fig. 2-f). It should be noted
that by enhancement of the hole diameter, the required volume of
the molten polymer for filling the hole increases. On the other
hand, raising the shoulder plunging depth leads to increase in
the volume of the molten polymer flowed from the interface of
the base sheets toward the threaded hole. Therefore, larger diam-
eter of the threaded hole needs higher shoulder plunging depth.

Mechanical and physical properties 2.3. Temperature measurement

Composition
(wt%)
6
/K 23.8 [14]
Al Base Yield Strength, MPa 214
Mg 2.34 UTS, MPa 265
Cr 0.20 Elongation,% 13.5
Fe 0.23 Thermal conductivity, W/m.K 138 [14]
Si 0.06 Melting point, C 644 [14]
 H. Karami
H. Karami
Pabandi
Pabandi
et al.
et / al.
Composite
/ Composite
Structures
Structures
174 174
(2017)
(2017)
59–69
59–69
6 6

In order to measure temperature variations during the diameter were embedded into the interface of sheets at a
joining process, K-type thermocouples with 0.5 mm distance of 10, 20 and
30 mm from the joint center. The maximum temperature obtained
 Fig. 2. Schematic of the different steps for the joining process.

2.4. Microscopic observations

The stereomicroscope was used for observation of cross-section

and fracture surface of the joints. Moreover, for investigation of the
joints microstructure and the fracture surfaces, Field Emission
Scanning Electron Microscope (FE-SEM) equipped with Energy Dis-
persive Spectroscopy (EDS) was used. For increasing the quality of
SEM images, a gold layer was coated on the samples surfaces.

2.5. Differential scanning calorimetry analysis

The crystallinity of PP-SCF composite and re-solidified polymer

inside the hole was evaluated by differential scanning calorimetry
(DSC) analysis at the temperature of 250 C for 5 min. The mea-
surements were performed on a 8 mg sample under N2 atmo-
sphere. The specimens were heated and cooled at a constant rate
of 10 C/min. The exothermic crystallization peak was recorded
as a function of temperature.

2.6. Mechanical properties

For evaluation of the mechanical behavior of the joints, shear-
tensile and cross-tension tests were carried out with the cross
head speed of 0.5 mm/min. Dimensions of the sheets and the
overlap area were selected as 70 30 mm2 and 30 30 mm2,
respectively (Fig. 3). The threaded hole was located at the center
of the overlap area. At each set of joining parameters, three
samples were tested for tensile tests and the average of maximum
Fig. 3. Schematic of tensile test specimens and their dimensions: a) Shear-Tensile
forces were reported. specimen and b) Cross-Tension specimen.
The hardness of the composite sheet and re-solidified polymer
inside the hole was measured in Shore D scale to investigate the
effect of temperature changes during the joining process on the
hardness and joint strength.
3. Results and discussion
3.1. Macro-profile of joints
The surface appearance of the joints for different rotational
speeds is presented in Fig. 4. The re-solidified polymer in the hole
and the surface of aluminum affected by the rotating tool are
shown in the figure. As can be seen, surface appearance of the
joints is smooth especially at higher rotational speeds due to more
heat generation and improved flow of the aluminum under the
tool shoulder. Fig. 5 shows the cross-section of the joints and
various macro-structural zones including base materials,
threaded hole and re-solidified polymer inside the threaded hole.
It can be seen that the rotational speed of the tool affected
filling behavior of the threaded hole by the molten polymer. The
higher was the rota- tional speed, the more was the frictional
heat generated during welding. The maximum temperature
experienced at different dis- tances from the joint center (R) are
shown in Fig. 6. It was observed that by enhancement of the tool
rotational speed, the maximum temperature produced during
the process increased. Moreover, the area experiencing
temperatures higher than the melting point of the polymer was
broader. Therefore, the temperature and vol- ume of the molten
polymer at the interface of the base sheets increased.
At the rotational speed of 500 rpm, the volume of the molten
polymer is not enough to fill completely the hole. Fig. 5-a shows
porosity and unfilled zone at the joint center and separation
occurred at the interface of sheets due to lack of adhesion force
in this area. By enhancement of the tool rotational speed, more
heat was generated and as a result, more polymers melted and
entered into the hole. In the cross-section of the joints produced
at 1000, 1500 and 2000 rpm, all areas of the hole were filled com-
pletely with molten polymer and mechanical interlocks were
formed between the threads of the hole and re-solidified polymer.
Furthermore, due to the melting of polymer at the interface of
the aluminum and composite sheets, an adhesion force was
also
Fig. 4. Joint surface appearance at different rotational speeds: a) 500 rpm, b)
1000 rpm, c) 1500 rpm, d) 2000 rpm.
formed at the periphery of the hole [11]. Referring to Fig. 5-d, as
a result of high frictional heat at a rotational speed of 2000 rpm,
the excess melted polymer re-solidified as a convex surface on
the threaded hole. This convexity was formed in effect of the pres-
sure of holding clamps after retracting of the welding tool.
3.2. Reaction layer formation and joining mechanisms
Fig. 7-a shows SEM image from a typical cross-section of the
joint. Magnified SEM image of the interface between aluminum
and re-solidified polymer inside the hole (area ‘‘A” in Fig. 7-a) is
given in Fig. 7-b. It is obvious that a reaction layer was formed with
a thickness of approximately 15 lm at the interface. The term of
‘‘reaction layer” refers to a layer formed between the aluminum
and the molten polymer and atoms of both materials exist in this
layer. It seems that this layer has been formed by erosion mecha-
nism (a well-known phenomenon in the brazing and soldering)
[16] in which due to the collision between molten polymer and
the hole wall at high temperatures, aluminum atoms were sepa-
rated from the aluminum surface and entered into the molten
polymer and formed the reaction layer.
Comparison between Fig. 7-b and c indicate that the thickness
of the reaction layer did not change significantly by variation of
the tool rotational speed. However, the chemical composition of
the layer was affected by the tool rotation speed. Table 4 shows
EDS analysis of the reaction layer for samples welded at the tool
rotational speed of 500 rpm (minimum heat input) and 2000 rpm
(maximum heat input). Enhancement of the tool rotational speed
led to increase in weight percent of Al and slightly decrease in
weight percent of C. In the explanation of this phenomenon, it
can be mentioned that the higher was the temperature of the
mol- ten polymer entered into the hole, the more was the
aluminum surface erosion inside the hole and consequently the
presence of aluminum atoms in the reaction layer. Moreover, the
presence of oxygen in the reaction layer was probably attributed
to degrada- tion of the molten polymer in effect of reaction with
oxygen [2]. Formation of a reaction layer with 8 wt% of Al was also
mentioned by Shahmiri et al. [16] in FSW of Polypropylene C30S
to AA5052 aluminum alloy.
In addition to the reaction layer, there was also a gap between
the reaction layer and re-solidified polymer some regions inside
the hole (Fig. 7-b and c). It seems that this gap was formed during
the cooling process of the joint due to the large difference between
the coefficients of thermal expansion (CTE) of the base materials
6 1
[1]. The CTE of aluminum and polypropylene are 23.8 10 K
6 1
and 100-150 10 K , respectively [1,15]. Compared to alu-
minum, higher CTE of polypropylene resulted in more shrinkage
of the molten polymer during the cooling cycle of the weld and
thus, a gap was formed between the polymer and the reaction
layer. On the other hand, shrinkage of the molten polymer during
solidification depended on the melt temperature. The higher was
the temperature of the molten polymer, the more was its shrinkage
during cooling. Since the temperature of the molten polymer at
the rotational speed of 2000 rpm was higher (Fig. 6), a wider gap
was expected in this tool rotational speed. As can be seen in Fig. 7-
c, the wide gap was filled with epoxy resin (cold mounting
material) in the preparation stage of the sample for
microstructural studies.
From what mentioned above, the obtained joint between alu-
minum and composite sheets using THFSW process is the result
of two phenomena:
i) Mechanical interlocks between re-solidified polymer and
threads inside the hole,
ii) Formation of reaction layer between aluminum and
polymer.
 Fig. 5. Stereograph images from cross-section of the joints at the rotational speed of: a) 500 rpm, b) 1000 rpm, c) 1500 rpm and d) 2000 rpm.
Fig. 6. The effect of tool rotational speed on the maximum temperature obtained at
different distances from the joint center.
It is crucial of importance that the results of a research carried
out by Khodabakhshi et al. [17] on FSW of aluminum to polymer
sheets proved the formation of Van der Waal’s bonds between
the atoms of aluminum oxide layer on the surface and polymer.
However, formation of a gap between the reaction layer and the
re-solidified polymer inside the hole reduces the effectiveness of
Van der Waal’s bonds and reaction layer for the joint strength. This
mechanism can be helpful for joint strength at the interface of the
base sheets and periphery of the hole (a region with adhesion
force in Fig. 5).
3.3. Distribution of carbon fibers
Fig. 8 presents the carbon fibers distribution in the re-
solidified polymer matrix inside the hole. As can be seen, variation
of the tool rotational speed had no effect on the fibers distribution
in the re- solidified polymer matrix. However, comparison
between Fig. 1 (base composite sheet) and Fig. 8 shows that in
general, the carbon fibers in the re-solidified polymer were
shorter than those in the base composite sheet. This may be due
to the fragmentation of the carbon fibers within the molten
polymer in the effect of their flowing from the interface of the
base sheets toward the threaded hole during the welding
process. On the other hand, the carbon fibers in the base
composite sheet had a relatively preferred orien- tation which is a
well-known phenomenon in the injected polymer matrix
composites [18,19]. However, carbon fibers in the re-
solidified polymer show a random orientation.
3.4. Fracture surface analysis
Fig. 9 shows the fracture surfaces of the samples after cross-
tension and shear-tensile tests. In the samples welded at the tool
rotational speeds of 500, 1000 and 1500 rpm, fracture occurred
at the interface of the composite sheet and re-solidified polymer
inside the hole. Indeed, fracture surface of the joint on the
compos- ite sheet was a circle in which unmelted polymer
surface at the center of this circle was surrounded by the
melted polymer sur- face. The melted surface was formed due to
entry of the melted polymer from the sheets interface toward
inside the hole resulting in surface melting of the primary
polymer at the periphery of the hole and combination of the two
mentioned melts with each other. In other words, the surface of
the composite sheet that is at the base of the hole had no direct
contact with the aluminum sheet. Therefore, an increase in the
temperature of the aluminum sheet
Fig. 7. a) SEM image from a typical cross-section of the joint, b) Reaction layer and the gap formed between aluminum and re-solidified composite in the threaded hole at
the tool rotational speed of 500 rpm (magnified of area ‘‘A”) and c) Reaction layer and the gap refilled by epoxy resin at the tool rotational speed of 2000 rpm.

Table 4
tool pin. Therefore, the molten polymer formed at the interface
Effect of tool rotational speed on the composition of the reaction layer.
of the base sheets flows upward with back extrusion into the
Elements Tool rotational speed (rpm) threaded hole just in the effect of the downward pressure of the
500 2000 tool. It should be mentioned that the molten polymer has high flu-
Al 21 38 idity and can easily flow and fill the threads of the hole with the
C 37 35 pressure applied by the welding tool. High fluidity of the molten
O 33 22 polymer is caused by the elevated temperature experienced by it
during the welding process. Fig. 6 confirms that the temperature
of the molten polymer at the distance of 10 mm from the joint
as a result of the frictional heat did not directly melt the polymer
cen- ter is significantly higher than the melting point of
surface at this region. The existence of the unmelted surface in
polypropylene (from 30 C above the melting point at 500 rpm to
the center of the circular base of the threaded hole at the
175 C above the melting point at 2000 rpm). Fig. 10 shows the
interface of the sheets means that re-solidified polymer in the
flow model of the molten polymer inside the threaded hole. The
hole was not joined to the primary polymer at this region.
molten polymer moves upward after entering into the
Therefore, if the re- solidified polymer inside the hole is
threaded hole from the periphery of the hole. However, upward
considered as a small cylinder with a diameter of 4 mm and a
movement of the molten polymer is restricted by the shoulder
height equal to the thickness of alu- minum sheet (2 mm), there
surface. As a result, the mol- ten polymer moves horizontally and
was a crack in the base of this cylinder (interface of the composite
then continues its path down- ward. Therefore, a vortex structure
sheet and re-solidified polymer) due to lack of melting of the
is formed in the re-solidified polymer inside the hole center.
composite sheet. Huang et al. [20] utilized self-riveting friction
By comparing Fig. 9-a to c, it is evident that enhancement of the
stir lap welding for joining of AA6082 to steel with filing the
tool rotational speed reduced the area of the unmelted surface of
prefabricated holes in the bottom steel sheet by plasticized
the polymer. The higher was the temperature of the molten poly-
aluminum. They reported that downward flow of the plasticized
mer, the more was the melting of the polymer surface. Fig. 9-d
aluminum occurred because of two origins: i) with pro- gress of
and e show the fracture surfaces of the samples welded at
the welding tool, aluminum alloy ahead of the pin is plas- ticized
2000 rpm after cross-tension and shear-tensile tests, respectively.
and then, threads of the rotating pin pushes the plasticized metal
It is observed in fracture surface of the shear-tensile sample
downwards; ii) imposed pressure by the pin bottom
(Fig. 9-e) that there is no unmelted surface at the rotational speed
extrudes the plasticized aluminum alloy into the prefabricated
of 2000 rpm. Therefore, the obtained joint failed in cross-tension
hole. Since the tool used for welding in the present work is pin-
test with the exit of the re-solidified polymer from the threaded
less, there is no flow of the molten polymer by the threads of the
 Fig. 8. Carbon fibers distribution in the re-solidified polymer matrix at different rotational speed of: a) 500 rpm, b) 1000 rpm, c) 1500 rpm and d) 2000 rpm.
hole rather than fracture from the interface of the composite
sheet and re-solidified polymer inside the hole (Fig. 9-d).
SEM images of the fracture surface of the joint processed at
500 rpm rotational speed clearly indicate the difference between
the internal and marginal regions of the fracture surface at the
hole base (Fig. 11). The internal region of the joint surface (area
‘‘A” in Fig. 11-a) showed a rough structure, while the structure
of the marginal region was relatively smooth (area ‘‘B” in Fig. 11-
a). The rough structure formed in area ‘‘A” may be the result of
partial plastic deformation during tensile test [21]. In other
words, the region in which the re-solidified polymer and the
initial polymer were combined and joined to each other during
welding undergoes plastic deformation before fracture. However,
there was no evi- dence of plastic deformation in area ‘‘B”. It
seems that this area with brittle fracture characteristics was
formed as a result of crack rapid growth before final fracture. Fig.
11-d to f show the SEM images from the fracture surface of
the sample welded at
2000 rpm after shear-tensile test. As can be seen, the surface
included ductile (region C) and brittle (region D) fracture,
relatively similar to the fracture surface of the cross-tension
sample welded at 500 rpm. However, locations of the brittle and
ductile fracture regions in shear-tensile test are different from
those in cross- tension test, maybe due to difference in loading
mode between two types of the tests.
3.5. Hardness variation of re-solidified polymer
Since fracture in the samples welded at the tool rotational
speed of 500–1500 rpm occurred from the melted and re-
solidified polymer at the interface of the base sheets (as
mentioned
in Section 3.3), the hardness of the re-solidified polymer may play
an important role in the mechanical strength of the joints. Fig. 12-
a gives the hardness variation of the re-solidified polymer inside
the hole versus tool rotational speed. The results showed that
the hardness of re-solidified polymer inside the hole decreased
slightly (about 4–6 in Shore D scale) compared to primary
composite sheet. In order to investigate the reasons for this
phenomenon, DSC anal- ysis was carried out on the primary
composite and re-solidified polymer inside the hole. Fig. 12-b
presents DSC curve of the pri- mary composite sheet and the
effect of the tool rotational speed on the crystallinity percent of
the re-solidified polymer. Given Fig. 12, by enhancement of the
tool rotational speed, the hardness and crystallinity percent of
the re-solidified polymer decreased. This may be related to
molecular weight reduction of the resolidi- fied polymer, since
the molten polymer was affected by the fric- tional heat during
the process [21,22]. The more was the tool rotational speed,
the higher was the temperature of the molten polymer.
3.6. Strength and fracture energy of joints
3.6.1. Cross-tension test
Fig. 13 shows the force-extension curves and the effect of tool
rotational speed on the fracture force and fracture energy of the
joints in cross-tension tests. As can be seen, the fracture force
increased by enhancement of the tool rotational speed. In this join-
ing process, two factors compete with each other in order to deter-
mine the fracture force by enhancement of the tool rotational
speed:
Fig. 9. Fracture surface of the polymeric composite sheet; variation in the area of the unmelted surface and fracture location at different rotational speed of: a) 500 rpm, b)
1000 rpm, c) 1500 rpm, d) 2000 rpm (cross-tension) and e) 2000 rpm (shear-tensile).
Fig. 10. Flow model of the molten polymer inside the hole.
1) Increase in fracture force due to the increase in the load
bearing area caused by the reduction of the unmelted sur-
face of the polymer (as mentioned in Section 3.4).
2) Decrease in the fracture force due to the reduction of the
hardness at the fracture location, i.e. re-solidified polymer
(as mentioned in Section 3.5).
Results of cross-tension tests confirm that increase in the load
bearing area overcomes hardness reduction of the re-solidified
polymer and thus, fracture force improves with enhancement of
the tool rotational speed. During tensile test, the unmelted surface
of the polymer at the base of the hole (which mentioned in Sec-
tion 3.5) not only reduces the load bearing area, but also acts as
a stress concentration location and consequently decreases the
joint strength. As mentioned before, the samples joined at the tool
rotational speed of 500–1500 rpm failed from the interface of re-
solidified polymer and primary composite sheet. By increasing
the tool rotational speed, the polymer surface at the base of the
hole was melted more and the joint strength improved. At the tool
rotational speed of 2000 rpm, there was no unmelted surface or
this surface was very small. Therefore, failure occurred by the exit
of the re-solidified polymer from the threaded hole rather than
fracture from the interface of the composite sheet and re-
solidified polymer and the fracture load increased to higher values.
It is worth mentioning that distribution of carbon fibers in the
threaded hole can also be effective on the joint strength. However,
as discussed in Section 3.3, distribution of carbon fibers did not
change with variation of the tool rotational speed. Therefore, from
this point of view, enhancement of the tool rotational speed has no
effect on the joint strength.
Given the curves in Fig. 13-a, due to a decrease in the crack
length (i.e. unmelted surface) at the interface of the primary
Fig. 11. SEM image from fracture surfaces: a) two different areas on the fracture surface of the sample welded at 500 rpm, b) ductile fracture surface (area A), c) brittle
fracture surface (area B), d) two different areas on the fracture surface of the sample welded at 2000 rpm, e) ductile fracture surface (area C) and f) brittle fracture surface
(area D).
Fig. 12. a) Hardness and crystallinity percent of the primary and re-solidified
composite inside the hole versus the tool rotational speed and b) DSC curve of the
primary composite sheet.
composite sheet and re-solidified polymer, the extension of the
joints at the maximum fracture load increased by enhancement
of the tool rotational speed.
Fig. 13-b shows fracture energy of the joints calculated by mea-
suring the area under the load-extension curve until maximum
fracture force obtained from cross-tension test. It is obvious that
enhancement of the tool rotational speed led to increasing the frac-
ture energy due to the simultaneous improvement of the fracture
force and extension of the joints.
3.6.2. Shear-tensile test
The effect of tool rotational speed on the fracture force and
energy in shear-tensile tests is shown in Fig. 14-a. All samples in
this test failed from the interface of the primary composite sheet
and the re-solidified polymer. According to the results, the rela-
tionship between the fracture force and energy with tool rotational
speed in shear-tensile test is similar to that of in cross-tension
test, i.e., by enhancement of the tool rotational speed, both
fracture force and energy increased. This may be explained
again by an increase in the load bearing area in the joints due
to a decrease in the unmelted surface of the polymer at the base
of the threaded hole. The maximum fracture load was 350 N
obtained at the tool rotational speed of 2000 rpm. If the joint area
is considered as a cir- cle with a diameter of 4 mm (base of the
threaded hole), the max- imum shear strength of the joint may be
calculated as 28 MPa, i.e.,
80% of the composite sheet strength.
Fig. 14-b shows that the fracture energy in the cross-tension
test was significantly higher than that in shear-tensile test.
Although the fracture loads of the joints in cross-tension test
were relatively lower due to opening-mode of fracture [23],
extensions

Fig. 13. a) Force-extension curves of cross-tension tests related to the specimens
with maximum fracture force at different tool rotational speeds and b) effect of the
tool rotational speed on the average cross-tension force and fracture energy of the
joints.
of the joints at the maximum fracture force in cross-tension tests
(5–7 mm) were noticeably higher than that in shear-tensile test
(less than 1 mm). The higher extension of the joints in cross-
tension test was attributed to the elastic bending and deformation
of the composite base sheet in the tensile direction during cross-
tension test. Since the fracture energy is directly related to the
area under the force-extension curve up to maximum fracture
load, fracture energy in the cross-tension test was obtained
higher than that in shear-tensile test.
4. Conclusion
In this research, a new refill friction spot welding process
named Threaded Hole Friction Spot Welding (THFSW) was used
to join AA5052 aluminum alloy and short-carbon-fiber-reinforced
polypropylene composite sheets. The microstructure and mechan-
ical properties of the joints were investigated and the following
results were obtained:
1) The results indicated that THFSW was used successfully in
order to join aluminum and polymer sheets and the hole
refilled completely by the molten polymer at appropriate
tool rotational speeds (higher than 1000 rpm).
2) A reaction layer composed mostly of Al, C and O was formed
between aluminum and re-solidified polymer inside the
hole. While enhancement of the tool rotational speed did
not affect significantly the reaction layer thickness, the
weight percent of Al in this layer increased.
Fig. 14. a) Effect of tool rotational speed on the average shear-tensile load and
fracture energy of the joints and b) comparison of fracture energy of the joints in
shear-tensile and cross-tension tests.
3) Mechanical interlocks between threads and re-solidified
polymer inside the hole and formation of the reaction layer
between aluminum and polymer were recognized as two
main bonding mechanisms in this process. Formation of a
gap between the reaction layer and the re-solidified polymer
inside the hole reduced the effectiveness of Van der Waal’s
bonds and reaction layer for the joint strength.
4) The samples joined at the rotational speed of 500, 1000 and
1500 rpm failed from the interface the composite sheet
and the re-solidified polymer inside the hole. However, at
the rotational speed of 2000 rpm, the joint failed by the exit
of the re-solidified polymer from the threaded hole. The
frac- ture surface was composed of two distinct regions of
melted and unmelted polymer. The internal part of the
melted region showed evidence of plastic deformation
during a ten- sile test, while at the marginal region,
brittle fracture occurred due to the rapid growth of the
crack.
5) Two factors of load bearing area and the hardness of the
fracture location competed with each other for determina-
tion of the fracture force. The strength and fracture energy
of the joints improved by enhancement of the tool rotational
speed due to a decrease in the unmelted surface of the poly-
mer and increase in load bearing area. The maximum shear-
tensile strength of the joints reached to 80 percent of the
composite base strength at the tool rotational speed of
2000 rpm. Fracture energy in cross-tension test was signifi-
cantly higher than that in shear-tensile test.
References
[1] Abibe AB, Sônego M, Dos Santos JF, Canto LB, Amancio-Filho ST. On the
feasibility of a friction-based staking joining method for polymer-metal hybrid
structures. Mater Des 2016;92:632–42.
[2] Seong MS, Kim TH, Nguyen KH, Kweon JH, Choi JH. A parametric study on the
failure of bonded single-lap joints of carbon composite and aluminum.
Compos
Struct 2008;86:135–45.
[3] Goushegir SM, Dos Santos JF, Amancio-Filho ST. Friction spot joining of
aluminum AA2024/carbon-fiber reinforced poly (phenylene sulfide) composite
single lap joints: microstructure and mechanical performance. Mater Des
2014;54:196–206.
[4] Teixeira de Freitas S, Sinke J. Failure analysis of adhesively-bonded metal-skin-
to-composite-stiffener: effect of temperature and cyclic loading. Compos
Struct 2017;166:27–37.
[5] Kang SG, Kim MG, Kim CG. Evaluation of cryogenic performance of adhesives
using composite–aluminum double-lap joints. Compos Struct 2007;78:440–6.
[6] Lambiase F, Ko D-C. Two-steps clinching of aluminum and carbon fiber
reinforced polymer sheets. Compos Struct 2017;164:180–8.
[7] Amancio Filho ST, Dos Santos JF. Joining of polymers and polymer–metal
hybrid structures: recent developments and trends. Polym Eng Sci 2009;49
(8):1461–76.
[8] Abibe AB, Amancio-Filho ST, Dos Santos JF, Hage E. Mechanical and failure
behaviour of hybrid polymer-metal staked joints. Mater Des 2013;46:338–47.
[9] Schilling C, Dos Santos JF. Method and device for joining at least two adjoining
work pieces by friction welding. US Patent No. 6722556 B2; 2004.
[10] Martinsen K, Hu SJ, Carlson BE. Joining of dissimilar materials. CIRP Ann –
Manuf Technol 2015;64(2):679–99 .
[11] Esteves JV, Goushegir SM, Dos Santos JF, Canto LB, Hage E, Amancio-Filho ST.
Friction spot joining of aluminum AA6181-T4 and carbon fiber-reinforced
poly (phenylene sulfide): effects of process parameters on the microstructure
and mechanical strength. Mater Des 2015;66:437–45.
[12] Ratanathavorn W. Hybrid joining of aluminum to thermoplastics with friction
stir welding (M.Sc. thesis). Stockholm: Department of Materials Science and
Engineering, KTH-Royal Institute of Technology; 2012.
[13] André NM, Goushegir SM, Dos Santos JF, Canto LB, Amancio-Filho ST. Friction
Spot Joining of aluminum alloy 2024–T3 and carbon-fiber-reinforced poly
(phenylene sulfide) laminate with additional PPS film interlayer:
microstructure, mechanical strength and failure mechanisms. Compos B Eng
2016;94:197–208.
[14] ASM International Handbook Committee. Metals Handbook: Vol. 2, Properties
and selection-nonferrous alloys and pure metals. tenth ed. New York: ASM
International Handbook; 1990. pp. 3-215.
[15] Hindle C. Polypropylene (PP). http://www.bpf.co.uk/plastipedia/polymers/pp.
aspx; [accessed on 14 February 2017].
[16] Shahmiri H, Movahedi M, Kokabi AH. Friction stir lap joining of aluminium
alloy to polypropylene sheets. Sci Technol Weld Joining 2016;22(2):120–6.
[17] Khodabakhshi F, Haghshenas M, Chen J, Shalchi Amirkhiz B, Li J, Gerlich AP.
Bonding mechanism and interface characterisation during dissimilar friction
stir welding of an aluminium/polymer bi-material joint. Sci Technol Weld
Joining 2016, Article in press.
[18] Mahajan C, Cormier D. 3D printing of carbon fiber composites with
preferentially aligned fibers. Proceedings of the 2015 Industrial and Systems
Engineering Research Conference, Nashville, 30 May–2 June 2015. p. 2953–62.
[19] Hannesschläger C, Revol V, Plank B, Salaberger D, Kastner J. Fibre structure
characterisation of injection moulded short fibre-reinforced polymers by X-ray
scatter dark field tomography. Case Studi Nondestr Test Eval 2015;3:34–41.
[20] Huang Y, Wang J, Wan L, Meng X, Liu H, Li H. Self-riveting friction stir lap
welding of aluminum alloy to steel. Mater Lett 2016;185:181–4.
[21] Nagatsuka K, Yoshida S, Tsuchiya A, Nakata K. Direct joining of carbon-fiber–
reinforced plastic to an aluminum alloy using friction lap joining. Compos B
Eng 2015;73:82–8.
[22] Kiss Z, Czigány T. Applicability of friction stir welding in polymeric materials.
Period Polytech Mech Eng 2007;51(1):15–8.
[23] Hertzberg RW. Deformation and fracture mechanics of engineering
materials. New York: Wiley; 1989. p. 261–88.