Journal of Materials Processing Technology 210 (2010) 844–851

Contents lists available at ScienceDirect

Journal of Materials Processing Technology

journal homepage: www.elsevier.com/locate/jmatprotec

A newly developed tool without probe for friction stir spot welding and its
performance
Y. Tozaki a , Y. Uematsu b , K. Tokaji b,∗
a
Gifu Prefectural Research Institute for Machinery and Material, 1288 Oze, Seki 501-3265, Japan
b
Department of Mechanical and Systems Engineering, Gifu University, 1-1 Yanagido, Gifu 501-1193, Japan

a r t i c l e i n f o a b s t r a c t

Article history: A newly developed tool for friction stir spot welding (FSSW) has been proposed, which has no probe,
Received 4 November 2009 but a scroll groove on its shoulder surface (scroll tool). By use of this tool, FSSW has been performed on
Received in revised form 12 January 2010 aluminium alloy 6061-T4 sheets and the potential of the tool was discussed in terms of weld structure
Accepted 22 January 2010
and static strength of welds. The experimental observations showed that the scroll tool had comparable
or superior performance to a conventional probe tool. It was confirmed that sound welding could be
achieved without a probe hole, in which the scroll groove played significant roles in the stirring of the
Keywords:
material and the shoulder plunge depth was the important processing variable. The maximum tensile-
Friction stir spot welding
Tool geometry
shear strength of the welds made by the scroll tool was found to be 4.6 kN that was higher than that of
Processing condition the welds made by the probe tool and two different fracture modes, shear fracture and plug fracture,
Weld structure appeared depending on processing condition. The shear fracture took place at smaller shoulder plunge
Tensile-shear strength depths or at shorter tool holding times, while the plug fracture occurred at larger shoulder plunge depths
Fracture mode or at longer tool holding times. It was indicated that the tensile-shear strength and associated fracture
modes were determined by two geometrical parameters in the weld zone.
© 2010 Elsevier B.V. All rights reserved.

1. Introduction for the shoulder diameter of 10 mm. Lin et al. (2008a,b) demon-
strated that the microstructure and the fracture modes of the welds
Lightweight materials such as aluminium and magnesium alloys made by a flat tool were different from those of the welds made by
are increasingly used in transportation industries in which weight a concave tool. Badarinarayan et al. (2008) evaluated the effects of
saving is critical. Therefore, as a joining technique for thin sheets probe profile and probe length on static strength and indicated that
of those alloys, friction stir welding (FSW) developed by TWI in the highest static strength was obtained using a thick probe with a
1991 has received considerable attention, because it offers various probe length of 1.4 mm. Badarinarayan et al. (2009a,b) also argued
advantages such as good retention of baseline mechanical proper- the effect of tool geometry on static strength and hook formation
ties, little distortion, low residual stresses and few weld defects. and showed that the probe geometry significantly affected the hook
Recently, Sakano et al. (2004) have applied FSW to spot welding, formation.
which is called friction stir spot welding (FSSW) or friction spot One of the major drawbacks in FSW and FSSW is that a probe
joining (FSJ). Since FSSW has the same advantages as FSW, this tech- hole remains at the end of the weld line or in the weld cen-
nique is expected for extensive applications to joining of body parts tre. In order to refill the probe hole, Allen and Arbegast (2005)
made of aluminium alloy sheets in automotive industry. and Uematsu et al. (2008) used the double-acting tools for FSSW
A cylindrical tool with a probe has been usually used in FSSW as consisting of outer shoulder and inner probe. In their works, the
well as in FSW. The tool geometry is one of the important process- tools refilled the probe hole successfully, but the systems were
ing variables influencing the weld structure and the performance of rather complicated. Alternatively, Aota and Ikeuchi (2008) per-
friction stir spot welds and thus its effects have been investigated by formed FSSW by a tool without a probe, but a sufficient depth of
several researchers. The authors (Tozaki et al., 2007, 2008) studied the stir zone could not be obtained.
the effects of probe length and shoulder diameter on weld structure In the present study, a newly developed tool without a probe,
and static strength of welds and reported that the highest tensile- but with a scroll groove on its shoulder surface has been proposed.
shear strengths were obtained for the probe length of 3.7 mm and By use of this tool, FSSW was conducted on aluminium alloy sheets
and the weld structure and the static strength of the welds were
evaluated. Based on experimental observations and comparison
∗ Corresponding author. Tel.: +81 58 293 2500; fax: +81 58 293 2491. with a conventional probe tool, the potential of the proposed tool
E-mail addresses: [email protected], [email protected] (K. Tokaji). is discussed.

0924-0136/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.jmatprotec.2010.01.015
 Y. Tozaki et al. / Journal of Materials Processing Technology 210 (2010) 844–851 845

Fig. 1. Configuration and photograph of newly developed tool. All dimensions are
in mm.

2. Experimental details
Fig. 2. Schematic illustration showing positions of holes drilled in upper sheet and
anvil for observation of material stirring and for temperature measurement during
2.1. Tool geometries processing, respectively.

The configuration and the photograph of a newly developed tool performed. The copper powders dispersed within the material dur-
are presented in Fig. 1. It should be noted that the tool has no probe, ing processing are clearly discernible by etching after welding.
but a scroll groove of 0.5 mm depth on its shoulder surface, which A hole with 1.5 mm diameter was also drilled at the position of
is referred to as the scroll tool. For comparison with the scroll tool, 2 mm below the bottom surface of the lower sheet in steel plate or
two tools with different geometries were employed. One is a con- anvil (see Fig. 2). A K-thermocouple was inserted into the hole, by
ventional probe tool with a standard metric M3 left-hand threaded which variations in temperature during processing were measured.
probe whose length is 3.7 mm, hereafter denoted as the probe tool.
The probe length is rather long, but the authors (Tozaki et al., 2007)
2.5. Tensile-shear test
have indicated that the highest tensile-shear strength of welds was
obtained for that length among three different probe lengths. The
Tensile-shear tests were carried out in order to evaluate the per-
other is a tool with neither probe nor scroll groove (plain tool). All
formance of welds. The lap-shear specimens according to JIS Z3136
tools have a concave shoulder of 10 mm diameter with the angle
were made by using two 40 mm × 150 mm coupons with the thick-
of concavity of 11.5◦ and are made of high-speed steel (Japanese
ness of 2 mm and a 40 mm × 40 mm overlap area, in the centre of
Industrial Standards (JIS); SKD61).
which FSSW was performed.

2.2. Material
3. Results
The material used is aluminium alloy 6061-T4 plate with the
3.1. Weld structure characterization
thickness of 2 mm whose chemical composition (wt.%) is Mg: 0.6,
Si: 1, Cu: <0.01, Fe: <0.2, Mn: 0.05, Cr: <0.05, Zn: 0.3, Ti: <0.05, Al:
3.1.1. Comparison between probe tool and scroll tool
bal.
Fig. 3 reveals typical examples of the macroscopic weld struc-
tures of a longitudinal section of the welds made by the probe tool
2.3. Processing conditions
and the scroll tool. In case of the probe tool (Fig. 3(a)), a well-

In FSSW, the important processing variables are the tool rota-

tional speed and the tool holding time. In addition to these
variables, the shoulder plunge depth is also believed to be another
variable, particularly for both the scroll tool and the plain tool. Since
the proper processing variables in a mechanical point of view are
different for each tool, the processing conditions employed were
decided based on preliminary test results. For the scroll tool and
the plain tool, the tool rotational speeds of 2000 and 3000 rpm,
the tool holding times of 1–7 s and the shoulder plunge depths of
0.5, 0.7 and 0.9 mm were used. In all cases, the plunge rate was
10 mm/min. For the probe tool, the tool rotational speeds of 2000
and 3000 rpm and the tool holding times of 0.2, 1, 2 and 3 s were
employed, where the plunge rate and the shoulder plunge depth
were 20 mm/min and 0.2 mm, respectively.

2.4. Observation of material stirring and temperature

measurement

In order to visualize the stirring of the material in the welds

made by the scroll tool and the plain tool, a hole with 1 mm diam-
Fig. 3. Weld structures of longitudinal section made by (a) probe tool under pro-
eter and 0.7 mm depth was drilled in the upper sheet under the
cessing condition of tool rotational speed of 3000 rpm and tool holding time of 3 s
shoulder surface as shown in Fig. 2. Copper powders with the aver- and (b) scroll tool under processing condition of tool rotational speed of 3000 rpm,
age diameter of 0.3 mm were inserted into it and then FSSW was tool holding time of 4 s and shoulder plunge depth of 0.7 mm.
846 Y. Tozaki et al. / Journal of Materials Processing Technology 210 (2010) 844–851

known weld structure can be seen, in which a probe hole remains

at the centre of the weld zone and the lower sheet moves upward
due to the penetration of the probe and the downward material
flow of the upper sheet, resulting in reduction in the thickness of
the upper sheet underneath the shoulder indentation. Grains are
refined inside the solid lines due to recrystallization (stir zone, SZ).
Inside the upward material flow of the lower sheet indicated by
the dotted lines, both sheets are intensely stirred (mixed zone, MZ).
On the other hand, in case of the scroll tool (Fig. 3(b)), there is no
probe hole and onion ring-like patterns emanating from the top
surface of the upper sheet can be seen in both sides of the weld
centre. Compared with the weld structure produced by the probe
tool (Fig. 3(a)), the upward material flow of the lower sheet is less
pronounced under this processing condition. Note that the black
triangular area seen in the weld centre in Fig. 3(b) is due to etching,
not defects.
Based on the macroscopic observations in Fig. 3, the weld struc-
tures are schematically illustrated in Fig. 4. In case of the probe
tool (Fig. 4(a)), the material softened by frictional heating is pushed
down to near the bottom surface and then moves upward outside
the material that flows downward, as indicated by the arrow. The
authors (Tozaki et al., 2007, 2008) have indicated that the thick-
ness of the upper sheet underneath the shoulder indentation, tu , Fig. 4. Schematic illustration of weld structures: (a) probe tool and (b) scroll tool.
and the distance corresponding to the nugget size, dn , were the sig-
nificant geometrical parameters characterizing the weld structure
and determining both the tensile-shear strength and the fracture 3.1.2. Microstructures produced by scroll tool
modes of the welds. In case of the scroll tool (Fig. 4(b)), from the Fig. 5 shows the close-up views of the regions A, B, C and D
onion ring-like patterns seen in Fig. 3(b), the material flow seems marked in Fig. 3(b). Region A has the microstructure consisting
to occur as indicated by the arrows. When the rotated tool contacts of fine equiaxed grains with the average size of 10 ␮m, which is
the top surface of the upper sheet, the material softened by fric- defined as the stir zone (SZ) (Fig. 5(a)). Region B is near the bound-
tional heating is first drawn toward the weld centre by the action ary between the upper and lower sheets. It can be seen that the
of the scroll groove, then the downward flow occurs due to the con- region is still affected by stirring (Fig. 5(b)), because the grain size is
straint of the material at or near the centre where no or less material coarser than in region A, but finer than in the parent metal (region
flow takes place, and subsequently the plasticized material moves D, Fig. 5(d)). Region C is near the boundary between both sheets
toward the outer circumference of the shoulder indentation. With underneath the outer circumference of the shoulder indentation
the penetration of the shoulder surface, the material flow gradually (Fig. 5(c)). The lower sheet moves upward and thus the boundary
expands to the depth direction. has an acute-angled shape.

Fig. 5. Close-up views of each region in weld structure produced by scroll tool shown in Fig. 3(b): (a) region A, (b) region B, (c) region C, and (d) region D.
 Y. Tozaki et al. / Journal of Materials Processing Technology 210 (2010) 844–851 847

Fig. 7. Variations in temperature during processing.

plain tool, no material flow occurred. The other is that much more
time is required for the probe tool to attain a temperature maxi-
mum. In case of the probe tool, the probe pin is first plunged into
the material and then the shoulder surface contacts the top surface
of the upper sheet and penetrates, which may result in less heat
generation for the same time period. In addition, the smaller area
of the shoulder surface than the scroll tool and the plain tool may
be another possible reason.
Fig. 6. Weld structures of longitudinal section produced by (a) scroll tool and (b)
plain tool under processing condition of tool rotational speed of 3000 rpm, tool hold- 3.1.5. Processing variable dependence of weld structure produced
ing time of 4 s and shoulder plunge depth of 0.7 mm. FSSW was performed with by scroll tool
copper powders inserted into hole drilled in upper sheet. Grey areas are etched The macrostructures of the welds made by the scroll tool are
regions in which copper powders are dispersed. (c) and (d) are close-up views of
revealed in Fig. 8 as a function of shoulder plunge depth, where the
region C in (a) and region D in (b), respectively.
tool rotational speed and the tool holding time are 3000 rpm and

3.1.3. Roles of scroll groove on shoulder surface

Fig. 6 reveals the macroscopic weld structures produced by the
scroll tool and the plain tool, where the copper powders were
inserted into the hole drilled in the upper sheet as shown in Fig. 2.
The dark grey areas seen in the weld zone are the etched regions
where the copper powders are dispersed. In case of the scroll tool
(Fig. 6(a)), it can be seen that the material is intensely stirred within
the upper sheet. As described previously, the material flow grad-
ually expands to the depth direction with the penetration of the
shoulder surface, but does not reach the boundary between both
sheets as seen in Fig. 6(c). On the other hand, in case of the plain
tool (Fig. 6(b)), the copper powders stay near the top surface of the
upper sheet, indicating that no downward material flow occurs.
Aota and Ikeuchi (2008) have performed FSSW by use of a convex
tool without a probe and reported that the depth of the material
flow was less than 0.5 mm. Therefore, it is realized that the scroll
groove plays significant roles in the stirring of the material. It should
be noted that the boundary between both sheets is clearly visible
(Fig. 6(b)), suggesting a weak bonding between the surfaces of both
sheets. However, grain refinement and plastic deformation can be
recognized as shown in Fig. 6(d), which may be attributed to both
the frictional heating and the downward force caused by the tool
rotation and the shoulder penetration.

3.1.4. Variations in temperature during processing

Variations in temperature during processing are shown in Fig. 7.
Two particular points should be noticed. One is that there exists no
difference in temperature variation between the scroll tool and the
plain tool under the same processing condition. This may suggest
Fig. 8. Weld structures of longitudinal section for different shoulder plunge depths
that the temperature rise is mainly due to the frictional heating
produced by scroll tool under processing condition of tool rotational speed of
generated by the contact and the penetration of the shoulder sur- 3000 rpm and tool holding time of 4 s: (a) 0.5 mm, (b) 0.7 mm, and (c) 0.9 mm. Arrows
face under the action of a downward force, because in case of the indicate upward material flow of lower sheet.
848 Y. Tozaki et al. / Journal of Materials Processing Technology 210 (2010) 844–851

Fig. 9. Tensile-shear strength of welds made by probe tool as a function of tool

holding time.

4 s, respectively. It should be emphasized that the weld structure

depends significantly on shoulder plunge depth. With increasing
shoulder plunge depth, SZ expands and the upward material flow
of the lower sheet occurs markedly as indicated by the arrows. As
a result, the thickness of the upper sheet underneath the shoulder
indentation, i.e. tu defined in Fig. 4, is reduced. Similar changes in
the weld structure were recognized as the tool rotational speed or
the tool holding time is increased.

3.2. Tensile-shear strength

3.2.1. Probe tool

Tensile-shear strength is represented in Fig. 9 as a function of
tool holding time. At the tool rotational speed of 2000 rpm, the
tensile-shear strength increases slightly with increasing tool hold-
ing time, while at the tool rotational speed of 3000 rpm, it does Fig. 10. Tensile-shear strength of welds made by scroll tool as a function of tool
not depend on tool holding time. The tensile-shear strengths at holding time: (a) 2000 rpm and (b) 3000 rpm.
3000 rpm are higher than those at 2000 rpm for the entire tool
holding times studied, where the maximum strength is 3.9 kN at 1 s. slightly on shoulder plunge depth, with the highest strengths at
the shoulder plunge depth of 0.9 mm. It should be noted that the
3.2.2. Scroll tool tensile-shear strengths are considerably low irrespective of shoul-
The relationships between tensile-shear strength and tool hold- der plunge depth and tool holding time compared with the scroll
ing time are shown in Fig. 10. It should be noted that the tool shown in Fig. 10. This may result from weak bonding between
tensile-shear strength is strongly affected by shoulder plunge the two sheets due to insufficient stirring of the material.
depth, but its dependence is different between two tool rota-
tional speeds. At 2000 rpm (Fig. 10(a)), the tensile-shear strength
increases with increasing shoulder plunge depth, and also increases
with increasing tool holding time at the shoulder plunge depths of
0.5 and 0.7 mm, but it does not depend on tool holding time at
the shoulder plunge depth of 0.9 mm, with the maximum strength
of 4.14 kN at 3 s. At 3000 rpm, the tensile-shear strengths are the
highest at the shoulder plunge depth of 0.7 mm for the entire tool
holding times, then 0.9 mm and 0.5 mm in decreasing order. At
the shoulder plunge depth of 0.5 mm, the tensile-shear strength
increases with increasing tool holding time, while at the shoulder
plunge depth of 0.9 mm, it initially increases with increasing tool
holding time, reaches the maximum and then decreases with a fur-
ther increase in tool holding time. The similar dependence on tool
holding time can be seen at the shoulder plunge depth of 0.7 mm,
with the maximum strength of 4.6 kN at 4 s.

3.2.3. Plain tool

Fig. 11 shows tensile-shear strength as a function of tool hold-
ing time. The dependence of tensile-shear strength on tool holding Fig. 11. Tensile-shear strength of welds made by plain tool as a function of tool
time is not recognized, but the tensile-shear strengths depend holding time.
 Y. Tozaki et al. / Journal of Materials Processing Technology 210 (2010) 844–851 849

Fig. 12. Two fracture modes observed under tensile-shear loading in welds made by scroll tool: (a) shear fracture and (b) plug fracture.

3.3. Fracture morphology under tensile-shear loading the shoulder plunge depth of 0.7 mm. The upper sheet partially
fails in tensile mode, but the shear fracture takes place through
In case of the probe tool, two different fracture modes appeared the boundary between both sheets. Fig. 13(b) is the plug fracture
as indicated in Fig. 9. One is the shear fracture of the nugget and the that occurred at the larger shoulder plunge depth of 0.9 mm than in
other is the mixed fracture that is the mixture of tensile fracture of Fig. 13(a). It can be seen that the weld fails at the upper sheet along
the upper sheet underneath the shoulder indentation and the shear the outer circumference of the shoulder indentation. The plug frac-
fracture of the nugget. The authors (Tozaki et al., 2007, 2008) have ture occurs as sufficient stirring of the material is achieved or the
pointed out that the tensile-shear strength and associated frac- actual nugget is sufficiently developed, i.e. at the faster tool rota-
ture modes were determined by two geometrical parameters in the tional speed or longer tool holding times or larger shoulder plunge
weld zone, i.e. tu and dn defined in Fig. 4. At 2000 rpm, the tu and
dn values become thinner and larger, respectively, with increasing
tool holding time. Therefore, the tensile-shear strength increases
with an increase in tool holding time due to increased dn values,
but fracture originates from the upper sheet underneath the shoul-
der indentation in tensile mode due to reduction in its thickness.
The shear fracture only occurs at the shortest tool holding time of
0.2 s due to insufficient stirring of the material. On the other hand, at
3000 rpm, since sufficient stirring of the material can be achieved,
the dn values become large enough to resist the shear fracture of
the nugget, resulting in higher tensile-shear strengths irrespective
of tool holding time. Fracture also originates from the upper sheet
underneath the shoulder indentation.
Also in case of the scroll tool, two different fracture modes, shear
fracture and plug fracture, were observed. Typical examples of such
fractures are represented in Fig. 12 that shows the top views of the
upper and lower sheets. Fig. 12(a) is the shear fracture in which the
boundary between both sheets is sheared-off, while Fig. 12(b) is the
plug fracture in which the nugget drops off due to tensile fracture
of the upper sheet along the outer circumference of the shoulder
indentation. At 2000 rpm, the shear fracture was seen under all the
processing conditions. At 3000 rpm, the shear fracture occurred for
the shoulder plunge depth of 0.5 mm and below 5 s for the shoulder
plunge depth of 0.7 mm, while the plug fracture took place at longer
tool holding times for the shoulder plunge depth of 0.7 mm and at
all tool holding times for the shoulder plunge depth of 0.9 mm. With
increasing shoulder plunge depth or tool holding time, the dn value
increases, but the tu value significantly decreases as seen in Fig. 8,
which becomes critical for final fracture.
Fracture appearances of a longitudinal section of the welds Fig. 13. Macroscopic appearances of tensile-shear fractures in welds made by scroll
tool: (a) shear fracture (tool rotational speed of 3000 rpm, tool holding time of 4 s
joined by the scroll tool are revealed in Fig. 13. Fig. 13(a) is the
and shoulder plunge depth of 0.7 mm), (b) plug fracture (tool rotational speed of
shear fracture that occurred under the processing condition of the 3000 rpm, tool holding time of 4 s and shoulder plunge depth of 0.9 mm). Arrows
tool rotational speed of 3000 rpm, the tool holding time of 4 s and indicate the loading direction.
850 Y. Tozaki et al. / Journal of Materials Processing Technology 210 (2010) 844–851

Fig. 14. SEM micrographs showing fracture surface of weld made by scroll tool under processing condition of tool rotational speed of 3000 rpm, tool holding time of 4 s and
shoulder plunge depth of 0.7 mm: (a) overview and (b) close-up view of rectangular region in (a).

Fig. 15. SEM micrographs showing fracture surface of weld made by scroll tool under processing condition of tool rotational speed of 3000 rpm, tool holding time of 1 s and
shoulder plunge depth of 0.7 mm: (a) overview and (b) close-up view of rectangular region in (a).

depths, because under those processing conditions, the tu values tensile-shear strength was obtained under this processing condi-
become significantly small. tion. However, there exist lesser dimples near the nugget centre
than in the circumference. Fig. 15 reveals the fracture surface of
3.4. Fracture surface appearance of nugget a weld joined by the scroll tool at the shorter tool holding time
of 1 s than in Fig. 14. Some traces in the circumferential direction
Fracture surface appearances of the nugget are shown in are also recognized (Fig. 15(a)), but there are some regions where
Figs. 14–16. Fig. 14 is the shear fracture surface of a weld made by only a few dimples can be seen (Fig. 15(b)), indicating that the
the scroll tool under the processing condition of the tool rotational actual nugget is not fully developed compared with that shown
speed of 3000 rpm, the tool holding time of 4 s and the shoulder in Fig. 14. Fig. 16 is the fracture surface of a weld joined by the
plunge depth of 0.7 mm. Some traces in the circumferential direc- plain tool under the processing condition of the tool rotational
tion and a ridge along the outer circumference of the nugget can be speed of 3000 rpm, the tool holding time of 4 s and the shoul-
seen (Fig. 14(a)). The former suggests that the effect of the material der plunge depth of 0.7 mm. Some traces in the circumferential
stirring extends to the boundary between both sheets and the latter direction are discernible as indicated by the arrow (Fig. 16(b)).
results from the upward material flow of the lower sheet under- This suggests that the effect of the tool rotation extends to the
neath the shoulder indentation. It should be noted that elongated boundary between both sheets. However, since no material flow
dimples are extensively seen (Fig. 14(b)), indicating well bond- of the upper sheet occurred, the actual nugget is insufficiently
ing between the upper and lower sheets. In fact, the maximum developed.

Fig. 16. SEM micrographs showing fracture surface of weld made by plain tool under processing condition of tool rotational speed of 3000 rpm, tool holding time of 4 s and
shoulder plunge depth of 0.7 mm: (a) overview and (b) close-up view of rectangular region in (a).
 Y. Tozaki et al. / Journal of Materials Processing Technology 210 (2010) 844–851 851

4. Discussion 5. Conclusions

The newly developed tool or the scroll tool is based on simple A newly developed tool for friction stir spot welding (FSSW)
design concepts. In order to avoid a probe hole in the weld cen- has been proposed, which has no probe, but a scroll groove on
tre, the tool has no probe. However, in the absence of probe, the its shoulder surface (scroll tool). By use of this tool, FSSW was
material stirring that is required for well bonding is not expected performed on aluminium alloy 6061-T4 sheets and the potential
and thus a scroll groove is formed on the shoulder surface. By use of the tool was discussed in terms of weld structure and static
of this tool, FSSW has been performed under various processing strength of welds. The experimental observations showed that the
conditions and consequently it was confirmed that sound weld- scroll tool had comparable or superior performance to a conven-
ing could be achieved without a probe hole. In case of the scroll tional probe tool. It was confirmed that sound welding could be
tool, in addition to the tool rotational speed and the tool holding achieved without a probe hole, in which the scroll groove played
time, the shoulder plunge depth was another important process- significant roles in the stirring of the material and the shoulder
ing variable. With increasing those variables, the material stirring plunge depth was the important processing variable. The maxi-
and resulting material flow occurred more intensely and exten- mum tensile-shear strength of the welds made by the scroll tool
sively within the upper sheet. Therefore, when the shoulder plunge was found to be 4.6 kN that was higher than that of the welds made
depth is increased, the effect of the material stirring extends to the by the probe tool. Two different fracture modes, shear fracture and
boundary between the upper and lower sheets, thus both sheets plug fracture, were observed depending on processing condition. It
are believed to be joined due to the contact of fresh surfaces under was indicated that the tensile-shear strength and associated frac-
the action of the downward force. ture modes were determined by two geometrical parameters in
Similar to the probe tool, changes in two geometrical parameters the weld zone, i.e. the actual nugget size and the thickness of the
in the weld zone with the processing condition were recognized, i.e. upper sheet underneath the shoulder indentation, either of which
the actual nugget size and the thickness of the upper sheet under- became critical for final fracture. Under the processing conditions
neath the shoulder indentation. The static strength and associated in which the plug fracture occurred, the tensile-shear strength was
fracture modes of the welds were determined by the magnitudes reduced.
of both parameters. As shown in Fig. 10(b), the maximum tensile-
shear strength of the welds joined by the scroll tool was 4.6 kN that References
was higher than that of the welds joined by the probe tool (see
Fig. 9). This may be primarily attributed to larger actual nugget Allen, C.D., Arbegast, W.J., 2005. Evaluation of friction spot welds in aluminum alloys.
In: SAE International, 2005-01-1252, pp. 107–113.
size. With an increase in the processing variables such as tool rota- Aota, K., Ikeuchi, K., 2008. Development of friction stir spot welding using rotat-
tional speed, tool holding time or shoulder plunge depth, the actual ing tool without probe and its application to low carbon steel plates. Quarterly
nugget size increases, while the thickness of the upper sheet under- Journal of the Japan Welding Society 26, 54–60 (in Japanese).
Badarinarayan, H., Yang, Q., Hunt, F., 2008. Effect of pin geometry on static strength
neath the shoulder indentation reduces. The shear strength of the of friction stir spot welds. In: SAE International, 2008-01-0147.
nugget increases monotonously with increasing those variables, Badarinarayan, H., Yang, Q., Zhu, S., 2009a. Effect of tool geometry on static strength
but the reduced thickness of the upper sheet becomes critical for of friction stir spot-welded aluminum alloy. International Journal of Machine
Tools & Manufacture 49, 142–148.
final fracture and thus the fracture mode transition from the shear
Badarinarayan, H., Shi, Y., Li, X., Okamoto, K., 2009b. Effect of tool geometry on
fracture to the plug fracture occurs, resulting in decreased tensile- hook formation and static strength of friction stir spot welded aluminum
shear strength. As can be seen in Fig. 10, the optimum shoulder 5754-O sheets. International Journal of Machine Tools & Manufacture 49,
plunge depth was 0.9 mm at 2000 rpm, while 0.7 mm at 3000 rpm. 814–823.
Lin, P.-C., Pan, J., Pan, T., 2008a. Failure modes and fatigue life estimations of
This can also be explained in the same manner. At 2000 rpm, spot friction welds in lap-shear specimens of aluminum 6111-T4 sheets.
since the material stirring is less remarkable than at 3000 rpm, Part 1. Welds made by a concave tool. International Journal of Fatigue 30,
much deeper plunge depth is needed in order to achieve suffi- 74–89.
Lin, P.-C., Pan, J., Pan, T., 2008b. Failure modes and fatigue life estimations of spot
cient stirring of the material. Therefore, with increasing shoulder friction welds in lap-shear specimens of aluminum 6111-T4 sheets. Part 2. Welds
plunge depth, the actual nugget size increases, leading to increased made by a flat tool. International Journal of Fatigue 30, 90–105.
tensile-shear strength. At 3000 rpm, intense material stirring is also Sakano, R., Murakami, K., Yamashita, K., Hyoe, T., Fujimoto, M., Inuzuka, M., Nagao,
Y., Kashiki, K., 2004. Development of spot FSW robot system for automobile body
achieved at the shoulder plunge depth of 0.9 mm, but the thickness members. In: Proceedings of the 3rd International Symposium of Friction Stir
of the upper sheet is significantly reduced, thus the tensile-shear Welding, Kobe, Japan, CD-ROM.
strengths decrease compared with those at the shoulder plunge Tozaki, Y., Uematsu, Y., Tokaji, K., 2007. Effect of tool geometry on microstructure
and static strength in friction stir spot welded aluminium alloys. International
depth of 0.7 mm. Journal of Machine Tools & Manufacture 47, 2230–2236.
In the present study, various advantages of the newly developed Tozaki, Y., Uematsu, Y., Tokaji, K., 2008. Effect of tool shoulder diameter on mechani-
scroll tool have been indicated in terms of the weld structure and cal properties of friction stir spot welded joints. Transactions of the Japan Society
of Mechanical Engineers, Series A 74, 268–274 (in Japanese).
the performance of the welds. The scroll tool is expected to be useful
Uematsu, Y., Tokaji, K., Tozaki, Y., Kurita, T., Murata, S., 2008. Effect of re-filling probe
for future FSSW process, but some geometrical modifications of the hole on tensile failure and fatigue behaviour of friction stir spot welded joints
tool such as the shape and the size of a scroll groove should be in Al–Mg–Si alloy. International Journal of Fatigue 30, 1956–1966.
further considered.