- Email: [email protected]
steel joints made by an alternative friction stir spot welding process
Accepted Manuscript Title: Aluminum/steel joints made by an alternative friction stir spot welding process Author: E. Fereiduni M. Movahedi A.H. Kokabi PII: DOI: Reference:
S0924-0136(15)00194-6 http://dx.doi.org/doi:10.1016/j.jmatprotec.2015.04.028 PROTEC 14399
To appear in:
Journal of Materials Processing Technology
Received date: Revised date: Accepted date:
20-11-2014 13-3-2015 23-4-2015
Please cite this article as: Fereiduni, E., Movahedi, M., Kokabi, A.H.,Aluminum/steel joints made by an alternative friction stir spot welding process, Journal of Materials Processing Technology (2015), http://dx.doi.org/10.1016/j.jmatprotec.2015.04.028 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Highlights 1- Aluminium and steel sheets were joined using an alternative FSSW process.
cr
3- IM layer thickness of 2.3 µm was a critical thickness.
ip t
2- The joint strength first improved and then decreased with increase in dwell time.
Ac ce p
te
d
M
an
us
4- Higher rotational speeds were associated with weaker FSSW joints.
1 Page 1 of 33
Aluminum/steel joints made by an alternative friction stir spot welding process E. Fereiduni, M. Movahedi* and A.H. Kokabi
ip t
Department of Materials Science and Engineering, Sharif University of Technology, P.O. Box 113659466, Azadi Ave., Tehran, Iran
cr
*Corresponding Author, Tel: +98 2166165224, E-mail address: [email protected]
us
Abstract
The effect of the rotational speed and dwell time on the joint interface microstructure and tensile-
an
shear strength of friction stir spot welded Al-5083 aluminum/St-12 steel alloy sheets was investigated. Joining of the sheets was performed using an alternative friction stir spot welding
M
(FSSW) process in which the welding tool tip did not penetrate into the lower steel sheet. Rotational speeds of 900 and 1100 rpm were applied in association with the dwell times of 5 to 15 s to weld the samples. Thermal history was recorded during the joining process. Interfacial
d
microstructure, formation of intermetallic compounds (IMCs) at the joint interface and the
te
fracture locations were studied using stereo, optical and scanning electron microscopy (SEM) equipped with energy dispersive X-ray spectroscopy (EDS). The used alternative process was
Ac ce p
successful to join the sheets. For both of the applied rotational speeds, there were optimum dwell times (10 and 12 s for the rotational speeds of 1100 and 900 rpm, respectively) to reach the maximum failure loads. The increasing trend of the strength as a function of the dwell time was related to the formation of a thin intermetallic (IM) layer at the joint interface. The decreasing trend was attributed to the formation of a relatively thick IM layer at the joint interface as well as the grain growth of aluminum at the exit-hole periphery where the final fracture occurred. The IM reaction layer thickness of 2.3 µm was found to be a critical thickness. Compared to the rotational speed of 1100 rpm, stronger joints were achieved by application of 900 rpm rotational speed.
Keywords: Al-5083; St-12; Friction stir spot welding; Rotational speed; Dwell time; Tensileshear strength 2 Page 2 of 33
1. Introduction The most important process variables in FSSW are the tool rotational speed and dwell time. These parameters determine the mechanical properties of the joints by influencing the generated
ip t
heat, material flow around the pin as well as the weld geometry. In recent years, some
researchers have explored the effect of the tool rotational speed and dwell time on the physical
cr
and mechanical properties of the similar friction stir spot joints of aluminum (Al) and
us
magnesium (Mg) alloys. However, very few studies have been done to join dissimilar Al and steel sheets using this process. Sun et al. (2013) joined 6061-T6 Al alloy and mild steel sheets
an
with the thicknesses of 1 mm using conventional FSSW process. They reported that the optimal rotational speed and dwell time were 700 rpm and 2 s, respectively. They also used three
M
different pin lengths of 1, 1.3 and 1.5 mm to weld the samples. The maximum tensile-shear load
d
of 3200 N was attained for the welds made by the tool with the pin length of 1 mm and at the
te
optimal rotational speed and dwell time. Moreover, their results showed that no obvious intermetallic (IM) layer existed along the Al/steel joint interface, however, an amorphous
Ac ce p
structure was formed adjacent to the joint interface in the Al part of the weld. Chen et al. (2012) successfully joined 6111-T4 Al alloy to DC04 steel sheets using a novel approach called “abrasion circle” FSSW during which a tool is plunged into the upper Al sheet and slightly penetrates into the lower steel sheet and then travels along a circular path. Finally, tool is moved to the center of the circle and extracted. The applied rotational speed in their study was 800 rpm and the dwell time was variable depending on the translation speed. Their results showed that the maximum failure load of 3500 N was achieved at the dwell time of 1.1 s. Characterization of the joint interface by transmission electron microscopy (TEM) revealed that no IM layer was formed at the joint interface. They also compared the failure loads of their samples with those of the
3 Page 3 of 33
joints produced by the conventional FSSW at the same process variables. They reported that their novel approach resulted in the stronger joints than the conventional FSSW (the maximum failure load of 2800 N was achieved at the dwell time of 7 s for the conventional FSSW). Sun
ip t
et al. (2013) and Chen et al. (2012) believe that strong FSSW joints could be attained when the
cr
joint interface is free of intermetallic compounds (IMCs). On the other hand, Bozzi et al. (2010) joined a 1.2 mm thick Al-6061 alloy to a 2 mm thick IF (interstitial free) steel using conventional
us
FSSW process. The applied sets of the rotational speed and penetration depth were 2000 rpm2.5 mm, 3000 rpm- 2.9 mm and 3500 rpm- 2.9 mm and the IM layers with the thicknesses of < 5
an
mµ, 8 mµ and 42 mµ were formed at the joint interface, respectively. They concluded that the
M
presence of an IM layer at the joint interface was necessary to improve the weld strength, but if the IM layer thickness was too thick, cracks would initiate and propagate easily through the IM
d
layer. The optimal IM layer thickness was determined to be 8 mµ at which the maximum failure
te
load of ~4200 N was recorded. In the present work, the effects of the rotational speed and dwell time on the joint strength of dissimilar Al-5083/St-12 joints made by an alternative FSSW
Ac ce p
process in which the tool tip did not penetrate into the lower steel sheet were investigated. The focus was on the formation of IMCs at the joint interface as well as the microstructural changes adjacent to the weld exit-hole and their correlation with the variation of weld strength.
2. Materials and experimental procedure Al-5083 and St-12 alloy sheets with the thicknesses of 3 and 1mm, respectively, were used in the present study as the base materials. Their chemical composition and mechanical properties are presented in Tables 1 and 2, respectively. The annealed St-12 sheet had an average grain size of ~42 µm. The base sheets were cut with the length and width of 100 and 35 mm, respectively.
4 Page 4 of 33
Prior to welding, the faying surfaces of the sheets were wire brushed to develop a suitable surface quality resulting in superior joining of the sheets. The Al-5083 alloy sheet was placed on the St-12 alloy sheet with an overlapped area of 35 35 mm2 and the welds were located at the
cr
ip t
center of the overlapped area. The joint configuration has been illustrated in Fig. 1.
used in the present work Si 0.49 Mn 0.5
Mn 0.47 Si 0.04
Mg 4.45 P,S <0.05
Al Bal. Fe Bal.
M
St-12 O
Fe 0.2 C <0.1
an
Al-5083 H321
us
Table 1. The chemical compositions (in wt.-%) of the Al-5083 H321 Al alloy and St-12 sheets
te
Yield Strength (MPa) 264 193
Ac ce p
Al-5083 H321 St-12 O
Vickers Hardness (HV20Kg) 75.7 73.5
d
Table 2. Mechanical properties of the Al-5083 H321 alloy and St-12 sheets Tensile Strength (MPa) 322 284
Elongation (%) 7 48
5 Page 5 of 33
ip t cr us
an
Fig. 1. Schematic illustration and dimensions of the samples with the associated hole drilled in
M
the steel sheet in order to record the thermal history at the interface of the sheets.
d
As can be seen, a hole with the diameter of 1.5 mm having a distance of 5 mm from the center of
te
the weld was drilled in the lower steel sheet to record the temperature variation at the joint
Ac ce p
‘interface’ during the process using a K-type thermocouple. The tool utilized for FSSW was made of a heat treated H-13 tool steel with a hardness of 52-54 RC. Pin length was selected to be 2.8 mm while the pin and shoulder diameters were 5 and 20 mm, respectively. Rotational speeds of 900 and 1100 rpm were used in association with the dwell times of 5, 7, 10, 12 and 15 s to weld the samples. The primary experiments revealed that the dwell times less than 5 s did not result in the joining of sheets. Therefore, just the samples with the dwell times more than 5 s were considered. After welding, the joint strength of the samples was evaluated by tensile-shear testing on three specimens for each processing condition and the average values were reported. The tests were carried out using an Instron tensile testing machine under the crosshead speed of 5 mm min-1. In the tensile-shear testing procedure, the fixtures were used to reduce the eccentricity 6 Page 6 of 33
of the loading path. Stereo, optical and scanning electron microscopy (SEM) equipped with an energy dispersive X-ray spectroscopy (EDS) were used in order to observe and characterize the Al/Fe IMCs at the joint interface and also to investigate the fracture location. The cross-section
ip t
of the St-12 sheet was etched by the 2% nital etchant solution to observe the microstructure of the steel side. Furthermore, the grain structure of the Al sheet was investigated near the exit-hole.
cr
For this purpose, the Al sides of the welds were electro-etched by the Barker etchant solution
us
(2.5 vol.-% fluoroboric acid in water) and the microstructures were observed by optical
an
microscope under polarized light.
3. Results
M
3.1. Macro/ microstructural characterization of weld
d
Figure 2(a) shows the top view of a FSSWed sample before tensile-shear test. As can be seen,
te
the Al-5083 alloy sheet is placed on top and the weld is located at the center of the overlapping area. Furthermore, a hole with the diameter of 5 mm (equal to the pin diameter) is left at the
Ac ce p
center of the weld as a result of pin penetration. Fig. 2(b) presents a typical cross-section of the weld zone as well as the enlarged micrographs of the selected squares in the different regions of the joint interface.
7 Page 7 of 33
ip t cr us an M d te Ac ce p
Fig. 2. (a) General view of friction stir spot welded specimen before the tensile-shear testing and (b) typical macroscopic cross-section of FSSW specimen before testing in association with the enlarged micrographs showing different regions of the joint interface.
It is observed that the tool penetration depth was 2.8 mm (equal to the pin length) which resulted in a residual thin Al layer with a thickness of 0.2 mm under the tool pin and on the steel sheet. The upper surface of the St-12 sheet and the lower surface of the Al-5083 alloy sheet after the tensile-shear test are shown in Figs. 3(a) and (b), respectively.
8 Page 8 of 33
ip t cr us an M d te
Ac ce p
Fig. 3. (a) Steel and (b) Al part of the joint after tensile-shear test.
The diameter of the remained Al layer on the St-12 sheet is the same as that of the tool pin (5 mm) revealing that the final fracture occurred inside the Al side of the joint in a position close to the outer circumference of the tool pin where the Al sheet thickness decreased as a result of the tool penetration. The cross-sections of the St-12 and Al-5083 sheets were prepared as shown in Figs. 3(a) and (b) and the associated micrographs are given in Figs. 4(a) and (b), respectively. As can be seen in Fig. 4(b), very fine and equiaxed grains are formed near the exit-hole while
9 Page 9 of 33
coarser and coarser Al grains are being observed with increasing the distance from the weld exit-
Ac ce p
te
d
M
an
us
cr
ip t
hole.
Fig. 4. Macroscopic cross-sections of (a) Steel and (b) Al part of the joint after tensile-shear test. 10 Page 10 of 33
3.2. Tensile-shear strength The average failure loads of the samples welded at different processing conditions are shown in
M
an
us
cr
ip t
Fig. 5.
Ac ce p
te
rpm for various dwell times.
d
Fig. 5. Average fracture loads of the welds processed at the rotational speeds of 900 and 1100
According to the results, for the welding condition of 900 rpm- 5 s, the weld strength was ~2270 N. The strength was improved with the enhancement of the dwell time to 12 s and reached to the maximum failure load of ~4020 N. Then, the joint strength declined to 2870 N with further increasing of the dwell time to 15 s. When the rotational speed of 1100 rpm was used, the weld strength was at its minimum for the dwell time of 5 s (1830 N). The joint strength was then improved with the enhancement of the dwell time to 10 s and attained to the maximum failure load of ~3630 N. Further increasing of the dwell time to 12 and then to 15 s resulted in the decreasing of the weld strength. Accordingly, failure load of ~2360 N was achieved under the welding condition of 1100 rpm- 15 s. Therefore, for both of the applied rotational speeds, failure 11 Page 11 of 33
load first increased and then declined with enhancement of the dwell time. It is noteworthy that when the higher rotational speed was used (1100 rpm), shorter dwell time was needed to achieve
speed from 900 to 1100 rpm led to the decrease in the joint strength.
cr
3.3. Temperature at the joint interface
ip t
the maximum failure load. Furthermore, for a given dwell time, enhancement of the rotational
us
The maximum temperatures generated at the joint interface for different welding conditions are given in Fig. 6. At the rotational speed of 900 rpm, increasing of the dwell time from 5 to 15 s
Ac ce p
te
d
M
an
resulted in the enhancement of the interface maximum temperature from 360 to 410 oC.
Fig. 6. The maximum temperatures measured at the joint interface.
Compared to the rotational speed of 900 rpm, the specimens processed at the rotational speed of 1100 rpm were accompanied by higher peak temperatures of 390 and 420 oC for the dwell times of 5 and 15 s, respectively. Hence, increase in the tool rotational speed and dwell time led to the enhancement of the interface maximum temperature. This could be attributed to greater frictional
12 Page 12 of 33
heating and more severe plastic deformation imposed on the surrounding material around the tool pin when using higher rotational speeds and longer dwell times (Yuan et al., 2011).
present study) is of a crucial importance due to the following reasons:
ip t
Recording of the thermal history during FSSW of dissimilar alloys (such as Al and steel in the
(i) The formation and morphology of the IMCs at the joint interface depend on the
cr
temperature generated in this region.
us
(ii) Temperature has normally a strong effect on the diffusional growth rate and thus the IM layer thickness at the joint interface.
an
(iii) As shown in Figs. 3 and 4, the final fracture occurred from the Al side of the welds where the Al sheet thickness decreased as a result of the tool pin penetration.
M
Therefore, microstructural evolution of this region is of a great importance and is
te
d
mainly dictated by the plastic deformation and temperature.
3.4. Microstructure of the joint interface
Ac ce p
Optical micrographs of the joint interface for the sample welded under the condition of 1100 rpm- 5 s are shown in Figs. 7(a)-(c). The enlarged SEM micrographs of selected regions in Figs. 7(a) and (c) are presented in Figs. 7(d) and (e), respectively.
13 Page 13 of 33
ip t cr us an M d te Ac ce p
Fig. 7. (a, b, c) optical and (d, e, f, g) SEM images of the joint cross-section after failure for the steel part of the specimen with the welding condition of 1100 rpm- 5 s.
Although the low magnification images of the joint interface in Figs. 7(d) and (e) indicated that no IMC was formed, SEM micrographs with higher magnification were taken to confirm this
14 Page 14 of 33
finding (Figs. 7(f) and (g)). As can be seen in Figs. 7(e) and (g), a swirl layered structure was formed as a result of the stirring and mixing effects of the tool pin at the joint interface. According to the short diffusion time of 5 s for this specimen, it seems that enough time has not
ip t
been provided for Al and Fe atoms to interdiffuse through the layers and form Al/Fe IMCs.
Therefore, it is expected that these layers have been consisted of a mechanical mixture of Al and
cr
steel layers. The joint cross sections of the specimen welded under the condition of 1100 rpm- 10
us
s are given in Figs. 8(a)-(i). According to the SEM micrographs shown in Figs. 8(g)-(i), IMCs were formed at the joint interface. As it can be observed in Figs. 8(d) and (f), these IMCs had a
an
non-continuous morphology while some relatively continuous layers were formed at some parts of the joint interface (Figs. 8(e) and (g)). Microstructural observation of the joint interface for
M
this specimen also revealed the formation of the swirl layered structure (Fig. 8(e)). The high
d
magnification SEM images of the peripheral regions of the remained Al layer at the location of
te
final fracture indicated the presence of Al/Fe IMCs (Figs. 8(g) and (i)). The interfacial microstructures of the sample under the welding condition of 900 rpm- 15 s are presented in Fig.
Ac ce p
9. Based on the SEM micrographs shown in Figs. 9(d) and (e), IMCs were formed with a more continuous morphology and higher thicknesses compared to the specimen with the welding condition of 1100 rpm- 10 s.
15 Page 15 of 33
ip t cr us an M d te Ac ce p
Fig. 8. (a, b, c) optical and (d, e, f, g, h, i) SEM images of the joint cross-section after failure for the steel part of the specimen with the welding condition of 1100 rpm- 10 s.
16 Page 16 of 33
ip t cr us an M d te Ac ce p
Fig. 9. (a, b, c) optical and (d, e, f, g, h) SEM images of the joint cross-section after failure for the steel part of the specimen with the welding condition of 900 rpm- 15 s.
Another key finding for the specimen with the welding condition of 900 rpm- 15 s was its different fracture type. Referring to the SEM micrographs shown in Figs. 7 and 8, there was no 17 Page 17 of 33
evidence of crack propagation inside the IMCs at the joint interface of sheets for the welds made with the welding conditions of 1100 rpm- 5 s and 1100 rpm- 10 s. However, based on the SEM micrographs taken from the peripheral regions of the remained Al layer on the steel sheet of the
ip t
sample with the welding condition of 900 rpm- 15 s (Figs. 9(d) and (f)), Al-5083 and St-12
sheets were first separated from each other in a situation that the Al-5083 side of the interface
cr
has moved perpendicular to the applied loading direction (the loading direction was out of
us
paper). Furthermore, IM reaction layers were observed on the separated surface of both St-12 and Al-5083 sheets (Figs. 9(d) and (f)). This observation evidences that the crack propagation
an
occurred inside the IM layer during the tensile-shear testing and then moved to the Al-5083 side of the joint where the thickness was decreased as a result of the pin penetration. The EDS
M
analysis results of the IM reaction layers of the 1100 rpm- 10 s and 900 rpm- 15 s FSSW
Ac ce p
te
d
specimens are given in Fig. 10.
Fig. 10. SEM micrograph in association with the EDS analysis result of the intermetallic layer formed at the joint interface of the sample with the welding condition of (a) 1100 rpm- 10 s before and (b) 900 rpm- 15 s after etching by the 2% nital etchant solution. 18 Page 18 of 33
As can be seen, IMCs contained 77 at.-% Al and 23 at.-% Fe at the joint interface of the 1100 rpm- 10 s sample (Fig. 10(a)) while they were composed of 78 at.-% Al and 22 at.-% Fe for the 900 rpm- 15 s specimen which was very close to the chemical composition of IM layer of the
layers have been formed during the FSSW of both specimens.
ip t
sample with the welding condition of 1100 rpm- 10 s. This indicates that the same type of IM
cr
Referring to Fig. 10(b) showing the etched microstructure of St-12 base metal matrix for the 900
us
rpm- 15 s sample, in comparison with the average grain size of ~42 µm for the St-12 base sheet, noticeably smaller grains are being observed in the steel part adjacent to the joint interface which
an
represents the formation of a structure/ substructure in the steel sheet. Furthermore, the joint interface shows a wavy (serrated) morphology. The observed phenomena may evidence a plastic
M
deformation in the steel side of the joint adjacent to the interface. It is noteworthy that however
d
the pin tip did not penetrate into the steel sheet, this region can undergo plastic deformation
te
probably due to the stirring action of the pin in the surrounding materials and the short distance
Ac ce p
of the pin tip to the joint interface (~0.2 mm).
3.5. Al-5083 grain structure
The applied FSSW process not only resulted in the interdiffusion of Al and Fe atoms and formation of Al/Fe IMCs at the joint interface, but also changed the microstructure of the Al5083 base metal near the weld exit-hole. The Al sides of the samples welded at the rotational speed of 1100 rpm with different dwell times of 5, 10 and 15 s were investigated using optical microscopy. The associated microstructures directly adjacent to the weld exit-hole, where the final fracture occurred, are presented in Fig. 11. As the general feature of the micrographs showed, very fine and equiaxed grains have been formed adjacent to the weld exit-hole which is
19 Page 19 of 33
the characteristic of stirred zone (SZ) and can be attributed to the dynamic recrystallization (DRX). DRX occurred as a result of the simultaneous effects of the pin stirring and frictional heat during the FSSW process. Microstructural analysis for the specimen welded under the
ip t
condition of 1100 rpm- 15 s indicated the formation of some relatively coarse grains distributed in a matrix of finer grains which could be caused by abnormal grain growth phenomenon (Fig.
Ac ce p
te
d
M
an
us
cr
11(c)).
Fig. 11. The microstructures of the weld exit-hole periphery for specimens welded at the rotational speed of 1100 rpm for the dwell times of (a) 5, (b) 10 and (c) 15 s. 20 Page 20 of 33
4. Discussion 4.1. Formation of IM reaction layer
ip t
Metallurgical bonds are formed as a result of the atomic diffusion across the joint interface. Al/Fe IMCs will be formed as a consequence of the atomic diffusion if enough energy is
cr
provided at the joint interface. As shown in Fig. 7, welding condition of 1100 rpm- 5 s did not
us
lead to the formation of IMCs at the joint interface which could be due to the short dwell time of 5 s even at the interface maximum temperature of 390 oC (Fig. 6). However, when longer dwell
an
time of 10 s was used at the same rotational speed, IMCs were formed as a non-continuous morphology at the joint interface (Fig. 8). This indicates that at the rotational speed of 1100 rpm
M
and dwell time of 10 s, the conditions for formation of a continuous IM layer have not been
d
provided which led to a non-continuous metallurgical bond at the joint interface. As shown in
te
Fig. 10(a), the average thickness of the IM layer for this specimen was ~2.3 µm. Microstructural analysis of the joint interface for the sample with the welding condition of 900 rpm- 15 s
Ac ce p
revealed the formation of a relatively continuous IM layer with a thickness of ~2.9 µm. The observed increasing trend in the thickness of IM layer at the joint interface from 2.3 to 2.9 µm clearly indicated that compared to the specimen welded at 1100 rpm- 10 s, the processing condition of 900 rpm- 15 s promoted the diffusional growth of the IM layer to a greater thickness. Given the processing conditions presented for these two joints (1100 rpm- 10 s and 900 rpm- 15 s), it is revealed that even though the rotational speed decreased from 1100 to 900 rpm, enhancement of the dwell time from 10 to 15 s increased the thickness of IM layer from 2.3 to 2.9 µm in one hand and changed the layer morphology from non-continuous to continuous on the other hand. Formation of either a continuous or non-continuous IM layer as well as the IM
21 Page 21 of 33
layer thickness at the joint interface is controlled by two major mechanisms of nucleation and growth. Nucleation and growth of IMCs are both affected by the temperature and plastic deformation (stored energy) at the joint interface. Enhancement of the temperature provides the
ip t
activation energy for the formation of IMCs in one hand and facilitates the IMCs diffusional growth on the other hand. The plastic deformation and the related stored energy at the joint
cr
interface during the welding process could also affect the IMCs nucleation. As reported by
us
Springer et al. (2011), variation in the tool rotational and travel speeds in the FSW process influenced the plastic deformation and consequently the nucleation of Al/Fe IMCs at the joint
an
interface. Furthermore, the stored energy at the joint interface may increase the IM layer growth rate as a result of the atomic diffusion through the high energy paths such as grain boundaries
M
and dislocations. In the FSSW process, the welding parameters which influence the IM layer
d
formation are rotational speed and dwell time. Given the IM layers formed at the interface of
te
1100 rpm- 10 s and 900 rpm- 15 s specimens (Figs. 8 and 9), the required temperature and plastic deformation were provided for the nucleation of IMCs. In the present study, as the dwell time
Ac ce p
was increased to 15 s, more diffusion time was provided for the growth of IMCs along and across the joint interface and the non-continuous IMCs became sufficiently dense to form a continuous layer.
The same structure of local and non-continuous Al/Fe IMCs as formed in 1100 rpm -10 s specimens was also observed by Naoi and Kajihara (2007) in solid Al/steel couples. They believed that these IMCs were resulted from the early stage of the interfacial IMC growth.
4.2. Welding parameters/failure load relationship
22 Page 22 of 33
On the basis of the average failure loads of the specimens presented in Fig. 5, for both of the rotational speeds, the joint strength first improved and then declined with the increase in dwell time. As shown in Fig. 7, there was no evidence of IMC formation at the joint interface of the
ip t
sample welded under the condition of 1100 rpm- 5 s. As was already pointed out, such
phenomenon was attributed to the short dwell time of 5 s which did not provide enough time for
cr
Al and Fe atoms to interdiffuse across the joint interface and to form the Al/Fe IMCs. The
us
microstructure of the joint interface for the specimen with the welding condition of 900 rpm- 5 s was not presented in the present paper. However, compared to the interface maximum
an
temperature of 390 oC for the welding condition of 1100 rpm- 5 s, lower interface maximum temperature of 360 oC was recorded for the sample with the processing condition of 900 rpm- 5 s
M
as a result of the lower rotational speed. Accordingly, it is reasonable to expect an IMC free
d
interface for the specimen with the welding condition of 900 rpm- 5 s. Bozzi et al. (2010) believe
te
that if enough IMCs are not formed at the joint interface of Al/steel joints, occurrence of a sudden fracture is more possible because of the sudden change in the chemical composition at
Ac ce p
the joint interface by moving from the Al sheet towards the steel side. Therefore, the low fracture loads for the welds with the processing conditions of 900 rpm– 5s and 1100 rpm- 5 s could be attributed to their IMC free interfaces. As the dwell time was enhanced from 5 to 10 s at the rotational speed of 1100 rpm, a thin IM layer with the thickness of 2.3 µm and a non-continuous morphology was formed at the joint interface and consequently the failure load improved from 1830 to 3632 N which was approximately twice as much. Achievement of the stronger joints under longer dwell times (10 s compared to 5 s) may be attributed to the formation of IMCs at the joint interface which was affected by the following phenomena:
23 Page 23 of 33
(i) More diffusion time was provided for Al and Fe atoms to move through the joint interface and form IMCs (Peng et al., 1999). (ii) The maximum temperature at the joint interface increased with enhancement of the dwell
ip t
time. Consequently, the growth constant of the IM layer increased.
It is noteworthy that compared to the sample with the welding condition of 1100 rpm- 10 s, the
cr
joint strength decreased gradually down to ~2360 N for the specimen with the welding condition
us
of 1100 rpm- 15 s. As the results of the interface maximum temperature in Fig. 6 showed, enhancement of the rotational speed and increase in the dwell time both resulted in the
an
generation of higher interface maximum temperatures. Therefore, enhancement of the dwell time from 10 to 15 s at the rotational speed of 1100 rpm increased the interface maximum
M
temperature. The microstructure of the joint interface for the sample with the welding condition
d
of 1100 rpm- 15 s was not shown in the present study, but as the interfacial microstructural
te
observation of the sample with the same dwell time and lower rotational speed (900 rpm- 15 s) showed, crack propagated inside the IM layer during the tensile testing (Figs. 9(d) and (f)). It is
Ac ce p
further noteworthy that compared to interface maximum temperature of 410 oC for the sample with the welding condition of 900 rpm- 15 s, higher temperature of 420 oC was recorded for the specimen with the welding condition of 1100 rpm- 15 s. This indicates that IM layer with the thickness even more than 2.9 µm could be formed at the joint interface of the sample with the processing condition of 1100 rpm- 15 s. In general, the joint strengths were improved after the formation of a non-continuous IM layer with the thickness less than 2.3 µm and then, decreased significantly when an IM layer with greater thickness and a relatively continuous morphology was formed at the interface of the sheets. The sharp loss of the joint strength for the samples with the dwell times of 15 s can be rationalized due to the formation of relatively thicker IM layers
24 Page 24 of 33
with less toughness at the joint interface in which crack needs less energy to propagate during the tensile-shear test. When the IMCs are formed with a non-continuous morphology at the interface, separation of the bonded sheets needs the propagation of cracks through IMCs as well as the
ip t
aluminum and/or steel base metals adjacent to the joint interface. It is well known that aluminum and steel base materials have greater toughness than the brittle Al/Fe IMCs. This increases the
cr
required energy for the propagation of cracks inside them. On the other hand, the FSSW process
us
enhances the toughness of the SZ in the base metals adjacent to the joint interface due to the formation of fine grained microstructure in this region. As a result, improvement of the Al and
an
steel toughness adjacent to the interface enhances the bond strength. Furthermore, according to the microstructural observations, morphological change of the IM layer from non-continuous to
M
continuous was accompanied by the increase of the thickness from ~2.3 to ~2.9 µm. As reported
d
by Springer (2011), with increase in the thickness of the IM layer at the joint interface of friction
te
stir welded Al/steel joints, the energy-consuming crack-interception at the IMC/steel interface
Ac ce p
became less frequent and consequently the joint strength decreased. Therefore, from the joint strength point of view, it can be concluded that the IM layer thickness of 2.3 µm was a critical thickness. The observed relationship between the Al/Fe IM layer thickness and the joint strength is inconsistent with the results presented by Movahedi et al. (2011) for the roll bonded Al -1100 alloy/St-12 sheets and also the results reported by Tanaka et al. (2009) on the FSW of 7075-T6 Al alloy to mild steel. They mentioned that the formation of IM phases, regardless of their thickness, reduced the bond quality of the joints. On the other hand, Danesh Manesh and Karimi Taheri (2003), Padmanabhan et al. (2008), Zhang et al. (2007) and Mathieu et al. (2006) have stated that the formation of IM layer with the thicknesses less than 10 µm imposed no detrimental effect on the joint strength and even improved the quality of 25 Page 25 of 33
the joint. They have also indicated that the joint strength followed a decreasing variation when the IM layer thickness exceeded 10 µm. In another study, Movahedi et al. (2013) studied the effect of the annealing treatment on the joint strength of FSWed Al-5083/St-12 alloy sheets and
ip t
found the IM layer thickness of 2.6 µm as the critical thickness in which the optimum tensileshear strength was achieved. This thickness is very close to the critical thickness of 2.3 µm in the
cr
present study.
us
According to the failure loads of 1830 and 2870 N for samples with the welding conditions of
an
1100 rpm- 5 s and 900 rpm- 15 s, respectively, it can be concluded that the formation of a relatively thick Al/Fe IM layer with the thickness of 2.9 µm at the joint interface may be
M
beneficial for improving the joint strength compared to the condition in which no IMC has been formed.
d
As previously pointed out, the maximum failure loads attained for the rotational speeds of 1100
te
and 900 rpm were related to the dwell times of 10 and 12 s, respectively. As it is clear, shorter
Ac ce p
dwell time was needed to reach the optimum tensile-shear strength when a higher rotational speed was applied. In fact, as shown in Fig. 6, increase in the rotational speed resulted in the generation of higher interface maximum temperature and decreased the time required for the formation of an IM layer with the critical thickness at the joint interface. Therefore, at higher rotational speeds, shorter dwell times are needed to obtain the critical IM layer thickness at which the optimum tensile-shear strengths are achieved. As shown in Fig. 5, for a certain dwell time, the tensile-shear strengths of the joints welded at rotational speed of 1100 rpm were lower than those welded at the rotational speed of 900 rpm. The observed finding is in agreement with the results reported by Lathabai et al. (2006), Tozaki et al. (2007) for the similar FSSW of Al alloy sheets. They reported that both the cross-tension 26 Page 26 of 33
and tensile-shear strengths were decreased with increasing the tool rotational speed. As the rotational speed increases, the heat input will increase which subsequently gives rise to the grain growth and the joint strength decreases. As Fig. 11 shows, the final fracture occurred from the
ip t
SZ in the periphery of the exit-hole. Therefore, the microstructural observation of this region is of a great importance since it determines the load required for the final fracture to occur. The
cr
enlarged micrographs of the exit-hole periphery in Fig. 11 revealed the formation of some
us
noticeably coarse grains distributed in the fined-grain matrix of SZ for the specimen with the welding condition of 1100 rpm- 15 s.
an
From what mentioned above, it can be concluded that the sharp loss of the joint strength for the specimens with longer dwell times could not be completely attributed to the formation of a
M
relatively thick and continuous IM layer at the joint interface, but also the grain growth
Ac ce p
5. Conclusions
te
strength.
d
phenomenon adjacent to the weld exit-hole would have also a detrimental effect on the joint
For both of the applied rotational speeds of 900 and 1100 rpm, the joint strength first improved and then declined with the increase in the dwell time. Increasing trend of the tensile-shear strength as a function of the dwell time was attributed to the formation of a non-continuous IM layer with the thickness less than 2.3 µm at the joint interface. Decreasing trend was also related to the formation of a relatively continuous IM layer with the thickness of 2.9 µm and the growth of Al grains adjacent to the weld exit-hole.
27 Page 27 of 33
By enhancement of the rotational speed from 900 to 1100 rpm, the required dwell time to achieve the optimum tensile-shear strengths decreased from 12 to 10 s. For a given dwell time, the specimens welded at the rotational speed of 1100 rpm were
cr
ip t
accompanied by lower tensile-shear strengths.
Acknowledgments
us
The authors wish to thank the research board of Sharif University of Technology for the financial
an
support and the provision of the research facilities used in this work.
M
References
Bozzi, S., Helbert-Etter, A.L., Baudin, T., Criquic, B., Kerbiguet, J.G., 2010. Intermetallic
te
d
compounds in Al 6016/IF-steel friction stir spot welds. Mater. Sci. Eng. A 527, 4505-4509.
Ac ce p
Chen, Y.C., Gholinia, A., Prangnell, P.B., 2012. Interface structure and bonding in abrasion circle friction stir spot welding: A novel approach for rapid welding aluminium alloy to steel automotive sheet. Mater. Chem. Phys. 134, 459-463. Danesh Manesh, H., Karimi Taheri, A., 2003. The effect of annealing treatment on mechanical properties of aluminum clad steel sheet. Mater. Des. 24, 617-622. Lathabai, S., Painter, M.J., Cantin, G.M.D., Tyagi, V.K., 2006. Friction spot joining of an extruded Al-Mg-Si alloy. Scripta Mater. 55, 899-902.
28 Page 28 of 33
Mathieu, A., Matteï, S., Deschamps, A., Martin, B., Grevey, D., 2006. Temperature control in laser brazing of a steel/aluminium assembly using thermographic measurements. NDT & E Int.
ip t
39, 272-276. Merzoug, M., Mazari, M., Berrahal, L., Imad, A., 2010, Parametric studies of the process of
cr
friction spot stir welding of aluminium 6060-T5 alloys. Mater. Des. 31, 3023-3028.
us
Movahedi, M., Kokabi, A.H., Seyed Reihani, S.M., 2011. Investigation on the bond strength of
an
Al-1100/St-12 roll bonded sheets, optimization and characterization. Mater. Des. 32, 3143-3149. Movahedi, M., Kokabi, A.H., Seyed Reihani, S.M., Cheng, W.J., Wang, C.J., 2013. Effect of
M
annealing treatment on joint strength of aluminum/steel friction stir lap weld. Mater. Des. 44, 487-492.
d
Naoi. D., Kajihara, M., 2007. Growth behavior of Fe2Al5 during reactive diffusion between Fe
te
and Al at solid-state temperatures. Mater. Sci. Eng. A 459, 375-382.
Ac ce p
Padmanabhan, R., Oliveira, M.C., Menezes, L.F., 2008. Deep drawing of aluminium-steel tailorwelded blanks. Mater. Des. 29, 154-160. Peng, X.K., Wuhrer, R., Heness. G., Yeung, W.Y., 1999. On the interface development and fracture behavior of roll bonded copper/aluminium metal laminates. J. Mater. Sci. 34, 20292038.
Springer, H., 2011. Fundamental research into the role of intermetallic phases in joining of aluminium alloys to steel. PhD Thesis. Ruhr-University Bochum, Bochum, Germany.
29 Page 29 of 33
Springer, H., Kostka, A., dos Santos, J.F., Raabe, D, 2011. Influence of intermetallic phases and Kirkendall-porosity on the mechanical properties of joints between steel and aluminium alloys.
ip t
Mater. Sci. Eng. A 528, 4630-4642. Sun, Y.F., Fujii, H., Takaki, N., Okitsu, Y., 2013. Microstructure and mechanical properties of
cr
dissimilar Al alloy/steel joints prepared by a flat spot friction stir welding technique. Mater. Des.
us
47, 350-357.
Tanaka, T., Morishige, T., Hirata, T., 2009. Comprehensive analysis of joint strength for
an
dissimilar friction stir welds of mild steel to aluminum alloys. Scripta Mater. 61, 756-759.
M
Tozaki, Y., Uematsu, Y., Tokaji, K., 2007. Effect of tool geometry on microstructure and static strength in friction stir spot welded aluminium alloys. Int. J. Mach. Tools Manuf. 47, 2230-2236.
d
Yuan, W., Mishra, R.S., Webb, S., Chen, Y. L., Carlson, B., Herling, D.R., Grant, G.J., 2011.
te
Effect of tool design and process parameters on properties of Al alloy 6016 friction stir spot
Ac ce p
welds. J. Mater. Proc. Technol. 211, 972-977. Zhang, H.T., Feng, J.C., He, P., Hackl, H., 2007. Interfacial microstructure and mechanical properties of aluminium-zinc-coated steel joints made by a modified metal inert gas weldingbrazing process. Mater. Charact. 58, 588-592.
30 Page 30 of 33
Figure Captions Fig. 1. Schematic illustration and dimensions of the samples with the associated hole drilled in
ip t
the steel sheet in order to record the thermal history at the interface of the sheets. Fig. 2. (a) General view of friction stir spot welded specimen before the tensile-shear testing and
cr
(b) typical macroscopic cross-section of FSSW specimen before testing in association with the
us
enlarged micrographs showing different regions of the joint interface. Fig. 3. (a) Steel and (b) Al part of the joint after tensile-shear test.
an
Fig. 4. Macroscopic cross-sections of (a) Steel and (b) Al part of the joint after tensile-shear test.
M
Fig. 5. Average fracture loads of the welds processed at the rotational speeds of 900 and 1100 rpm for various dwell times.
te
d
Fig. 6. The maximum temperatures measured at the joint interface. Fig. 7. (a, b, c) optical and (d, e, f, g) SEM images of the joint cross-section after failure for the
Ac ce p
steel part of the specimen with the welding condition of 1100 rpm- 5 s. Fig. 8. (a, b, c) optical and (d, e, f, g, h, i) SEM images of the joint cross-section after failure for the steel part of the specimen with the welding condition of 1100 rpm- 10 s. Fig. 9. (a, b, c) optical and (d, e, f, g, h) SEM images of the joint cross-section after failure for the steel part of the specimen with the welding condition of 900 rpm- 15 s. Fig. 10. SEM micrograph in association with the EDS analysis result of the intermetallic layer formed at the joint interface of the sample with the welding condition of (a) 1100 rpm- 10 s before and (b) 900 rpm- 15 s after etching by the 2% Nital etchant solution. 31 Page 31 of 33
Fig. 11. The microstructures of the weld exit-hole periphery for specimens welded at the
Ac ce p
te
d
M
an
us
cr
ip t
rotational speed of 1100 rpm for the dwell times of (a) 5, (b) 10 and (c) 15 s.
32 Page 32 of 33
Table Captions Table 1. The chemical compositions (in wt.-%) of the Al-5083 H321 Al alloy and St-12 sheets
ip t
used in the present work
Ac ce p
te
d
M
an
us
cr
Table 2. Mechanical properties of the Al-5083 H321 alloy and St-12 sheets
33 Page 33 of 33
S0924-0136(15)00194-6 http://dx.doi.org/doi:10.1016/j.jmatprotec.2015.04.028 PROTEC 14399
To appear in:
Journal of Materials Processing Technology
Received date: Revised date: Accepted date:
20-11-2014 13-3-2015 23-4-2015
Please cite this article as: Fereiduni, E., Movahedi, M., Kokabi, A.H.,Aluminum/steel joints made by an alternative friction stir spot welding process, Journal of Materials Processing Technology (2015), http://dx.doi.org/10.1016/j.jmatprotec.2015.04.028 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Highlights 1- Aluminium and steel sheets were joined using an alternative FSSW process.
cr
3- IM layer thickness of 2.3 µm was a critical thickness.
ip t
2- The joint strength first improved and then decreased with increase in dwell time.
Ac ce p
te
d
M
an
us
4- Higher rotational speeds were associated with weaker FSSW joints.
1 Page 1 of 33
Aluminum/steel joints made by an alternative friction stir spot welding process E. Fereiduni, M. Movahedi* and A.H. Kokabi
ip t
Department of Materials Science and Engineering, Sharif University of Technology, P.O. Box 113659466, Azadi Ave., Tehran, Iran
cr
*Corresponding Author, Tel: +98 2166165224, E-mail address: [email protected]
us
Abstract
The effect of the rotational speed and dwell time on the joint interface microstructure and tensile-
an
shear strength of friction stir spot welded Al-5083 aluminum/St-12 steel alloy sheets was investigated. Joining of the sheets was performed using an alternative friction stir spot welding
M
(FSSW) process in which the welding tool tip did not penetrate into the lower steel sheet. Rotational speeds of 900 and 1100 rpm were applied in association with the dwell times of 5 to 15 s to weld the samples. Thermal history was recorded during the joining process. Interfacial
d
microstructure, formation of intermetallic compounds (IMCs) at the joint interface and the
te
fracture locations were studied using stereo, optical and scanning electron microscopy (SEM) equipped with energy dispersive X-ray spectroscopy (EDS). The used alternative process was
Ac ce p
successful to join the sheets. For both of the applied rotational speeds, there were optimum dwell times (10 and 12 s for the rotational speeds of 1100 and 900 rpm, respectively) to reach the maximum failure loads. The increasing trend of the strength as a function of the dwell time was related to the formation of a thin intermetallic (IM) layer at the joint interface. The decreasing trend was attributed to the formation of a relatively thick IM layer at the joint interface as well as the grain growth of aluminum at the exit-hole periphery where the final fracture occurred. The IM reaction layer thickness of 2.3 µm was found to be a critical thickness. Compared to the rotational speed of 1100 rpm, stronger joints were achieved by application of 900 rpm rotational speed.
Keywords: Al-5083; St-12; Friction stir spot welding; Rotational speed; Dwell time; Tensileshear strength 2 Page 2 of 33
1. Introduction The most important process variables in FSSW are the tool rotational speed and dwell time. These parameters determine the mechanical properties of the joints by influencing the generated
ip t
heat, material flow around the pin as well as the weld geometry. In recent years, some
researchers have explored the effect of the tool rotational speed and dwell time on the physical
cr
and mechanical properties of the similar friction stir spot joints of aluminum (Al) and
us
magnesium (Mg) alloys. However, very few studies have been done to join dissimilar Al and steel sheets using this process. Sun et al. (2013) joined 6061-T6 Al alloy and mild steel sheets
an
with the thicknesses of 1 mm using conventional FSSW process. They reported that the optimal rotational speed and dwell time were 700 rpm and 2 s, respectively. They also used three
M
different pin lengths of 1, 1.3 and 1.5 mm to weld the samples. The maximum tensile-shear load
d
of 3200 N was attained for the welds made by the tool with the pin length of 1 mm and at the
te
optimal rotational speed and dwell time. Moreover, their results showed that no obvious intermetallic (IM) layer existed along the Al/steel joint interface, however, an amorphous
Ac ce p
structure was formed adjacent to the joint interface in the Al part of the weld. Chen et al. (2012) successfully joined 6111-T4 Al alloy to DC04 steel sheets using a novel approach called “abrasion circle” FSSW during which a tool is plunged into the upper Al sheet and slightly penetrates into the lower steel sheet and then travels along a circular path. Finally, tool is moved to the center of the circle and extracted. The applied rotational speed in their study was 800 rpm and the dwell time was variable depending on the translation speed. Their results showed that the maximum failure load of 3500 N was achieved at the dwell time of 1.1 s. Characterization of the joint interface by transmission electron microscopy (TEM) revealed that no IM layer was formed at the joint interface. They also compared the failure loads of their samples with those of the
3 Page 3 of 33
joints produced by the conventional FSSW at the same process variables. They reported that their novel approach resulted in the stronger joints than the conventional FSSW (the maximum failure load of 2800 N was achieved at the dwell time of 7 s for the conventional FSSW). Sun
ip t
et al. (2013) and Chen et al. (2012) believe that strong FSSW joints could be attained when the
cr
joint interface is free of intermetallic compounds (IMCs). On the other hand, Bozzi et al. (2010) joined a 1.2 mm thick Al-6061 alloy to a 2 mm thick IF (interstitial free) steel using conventional
us
FSSW process. The applied sets of the rotational speed and penetration depth were 2000 rpm2.5 mm, 3000 rpm- 2.9 mm and 3500 rpm- 2.9 mm and the IM layers with the thicknesses of < 5
an
mµ, 8 mµ and 42 mµ were formed at the joint interface, respectively. They concluded that the
M
presence of an IM layer at the joint interface was necessary to improve the weld strength, but if the IM layer thickness was too thick, cracks would initiate and propagate easily through the IM
d
layer. The optimal IM layer thickness was determined to be 8 mµ at which the maximum failure
te
load of ~4200 N was recorded. In the present work, the effects of the rotational speed and dwell time on the joint strength of dissimilar Al-5083/St-12 joints made by an alternative FSSW
Ac ce p
process in which the tool tip did not penetrate into the lower steel sheet were investigated. The focus was on the formation of IMCs at the joint interface as well as the microstructural changes adjacent to the weld exit-hole and their correlation with the variation of weld strength.
2. Materials and experimental procedure Al-5083 and St-12 alloy sheets with the thicknesses of 3 and 1mm, respectively, were used in the present study as the base materials. Their chemical composition and mechanical properties are presented in Tables 1 and 2, respectively. The annealed St-12 sheet had an average grain size of ~42 µm. The base sheets were cut with the length and width of 100 and 35 mm, respectively.
4 Page 4 of 33
Prior to welding, the faying surfaces of the sheets were wire brushed to develop a suitable surface quality resulting in superior joining of the sheets. The Al-5083 alloy sheet was placed on the St-12 alloy sheet with an overlapped area of 35 35 mm2 and the welds were located at the
cr
ip t
center of the overlapped area. The joint configuration has been illustrated in Fig. 1.
used in the present work Si 0.49 Mn 0.5
Mn 0.47 Si 0.04
Mg 4.45 P,S <0.05
Al Bal. Fe Bal.
M
St-12 O
Fe 0.2 C <0.1
an
Al-5083 H321
us
Table 1. The chemical compositions (in wt.-%) of the Al-5083 H321 Al alloy and St-12 sheets
te
Yield Strength (MPa) 264 193
Ac ce p
Al-5083 H321 St-12 O
Vickers Hardness (HV20Kg) 75.7 73.5
d
Table 2. Mechanical properties of the Al-5083 H321 alloy and St-12 sheets Tensile Strength (MPa) 322 284
Elongation (%) 7 48
5 Page 5 of 33
ip t cr us
an
Fig. 1. Schematic illustration and dimensions of the samples with the associated hole drilled in
M
the steel sheet in order to record the thermal history at the interface of the sheets.
d
As can be seen, a hole with the diameter of 1.5 mm having a distance of 5 mm from the center of
te
the weld was drilled in the lower steel sheet to record the temperature variation at the joint
Ac ce p
‘interface’ during the process using a K-type thermocouple. The tool utilized for FSSW was made of a heat treated H-13 tool steel with a hardness of 52-54 RC. Pin length was selected to be 2.8 mm while the pin and shoulder diameters were 5 and 20 mm, respectively. Rotational speeds of 900 and 1100 rpm were used in association with the dwell times of 5, 7, 10, 12 and 15 s to weld the samples. The primary experiments revealed that the dwell times less than 5 s did not result in the joining of sheets. Therefore, just the samples with the dwell times more than 5 s were considered. After welding, the joint strength of the samples was evaluated by tensile-shear testing on three specimens for each processing condition and the average values were reported. The tests were carried out using an Instron tensile testing machine under the crosshead speed of 5 mm min-1. In the tensile-shear testing procedure, the fixtures were used to reduce the eccentricity 6 Page 6 of 33
of the loading path. Stereo, optical and scanning electron microscopy (SEM) equipped with an energy dispersive X-ray spectroscopy (EDS) were used in order to observe and characterize the Al/Fe IMCs at the joint interface and also to investigate the fracture location. The cross-section
ip t
of the St-12 sheet was etched by the 2% nital etchant solution to observe the microstructure of the steel side. Furthermore, the grain structure of the Al sheet was investigated near the exit-hole.
cr
For this purpose, the Al sides of the welds were electro-etched by the Barker etchant solution
us
(2.5 vol.-% fluoroboric acid in water) and the microstructures were observed by optical
an
microscope under polarized light.
3. Results
M
3.1. Macro/ microstructural characterization of weld
d
Figure 2(a) shows the top view of a FSSWed sample before tensile-shear test. As can be seen,
te
the Al-5083 alloy sheet is placed on top and the weld is located at the center of the overlapping area. Furthermore, a hole with the diameter of 5 mm (equal to the pin diameter) is left at the
Ac ce p
center of the weld as a result of pin penetration. Fig. 2(b) presents a typical cross-section of the weld zone as well as the enlarged micrographs of the selected squares in the different regions of the joint interface.
7 Page 7 of 33
ip t cr us an M d te Ac ce p
Fig. 2. (a) General view of friction stir spot welded specimen before the tensile-shear testing and (b) typical macroscopic cross-section of FSSW specimen before testing in association with the enlarged micrographs showing different regions of the joint interface.
It is observed that the tool penetration depth was 2.8 mm (equal to the pin length) which resulted in a residual thin Al layer with a thickness of 0.2 mm under the tool pin and on the steel sheet. The upper surface of the St-12 sheet and the lower surface of the Al-5083 alloy sheet after the tensile-shear test are shown in Figs. 3(a) and (b), respectively.
8 Page 8 of 33
ip t cr us an M d te
Ac ce p
Fig. 3. (a) Steel and (b) Al part of the joint after tensile-shear test.
The diameter of the remained Al layer on the St-12 sheet is the same as that of the tool pin (5 mm) revealing that the final fracture occurred inside the Al side of the joint in a position close to the outer circumference of the tool pin where the Al sheet thickness decreased as a result of the tool penetration. The cross-sections of the St-12 and Al-5083 sheets were prepared as shown in Figs. 3(a) and (b) and the associated micrographs are given in Figs. 4(a) and (b), respectively. As can be seen in Fig. 4(b), very fine and equiaxed grains are formed near the exit-hole while
9 Page 9 of 33
coarser and coarser Al grains are being observed with increasing the distance from the weld exit-
Ac ce p
te
d
M
an
us
cr
ip t
hole.
Fig. 4. Macroscopic cross-sections of (a) Steel and (b) Al part of the joint after tensile-shear test. 10 Page 10 of 33
3.2. Tensile-shear strength The average failure loads of the samples welded at different processing conditions are shown in
M
an
us
cr
ip t
Fig. 5.
Ac ce p
te
rpm for various dwell times.
d
Fig. 5. Average fracture loads of the welds processed at the rotational speeds of 900 and 1100
According to the results, for the welding condition of 900 rpm- 5 s, the weld strength was ~2270 N. The strength was improved with the enhancement of the dwell time to 12 s and reached to the maximum failure load of ~4020 N. Then, the joint strength declined to 2870 N with further increasing of the dwell time to 15 s. When the rotational speed of 1100 rpm was used, the weld strength was at its minimum for the dwell time of 5 s (1830 N). The joint strength was then improved with the enhancement of the dwell time to 10 s and attained to the maximum failure load of ~3630 N. Further increasing of the dwell time to 12 and then to 15 s resulted in the decreasing of the weld strength. Accordingly, failure load of ~2360 N was achieved under the welding condition of 1100 rpm- 15 s. Therefore, for both of the applied rotational speeds, failure 11 Page 11 of 33
load first increased and then declined with enhancement of the dwell time. It is noteworthy that when the higher rotational speed was used (1100 rpm), shorter dwell time was needed to achieve
speed from 900 to 1100 rpm led to the decrease in the joint strength.
cr
3.3. Temperature at the joint interface
ip t
the maximum failure load. Furthermore, for a given dwell time, enhancement of the rotational
us
The maximum temperatures generated at the joint interface for different welding conditions are given in Fig. 6. At the rotational speed of 900 rpm, increasing of the dwell time from 5 to 15 s
Ac ce p
te
d
M
an
resulted in the enhancement of the interface maximum temperature from 360 to 410 oC.
Fig. 6. The maximum temperatures measured at the joint interface.
Compared to the rotational speed of 900 rpm, the specimens processed at the rotational speed of 1100 rpm were accompanied by higher peak temperatures of 390 and 420 oC for the dwell times of 5 and 15 s, respectively. Hence, increase in the tool rotational speed and dwell time led to the enhancement of the interface maximum temperature. This could be attributed to greater frictional
12 Page 12 of 33
heating and more severe plastic deformation imposed on the surrounding material around the tool pin when using higher rotational speeds and longer dwell times (Yuan et al., 2011).
present study) is of a crucial importance due to the following reasons:
ip t
Recording of the thermal history during FSSW of dissimilar alloys (such as Al and steel in the
(i) The formation and morphology of the IMCs at the joint interface depend on the
cr
temperature generated in this region.
us
(ii) Temperature has normally a strong effect on the diffusional growth rate and thus the IM layer thickness at the joint interface.
an
(iii) As shown in Figs. 3 and 4, the final fracture occurred from the Al side of the welds where the Al sheet thickness decreased as a result of the tool pin penetration.
M
Therefore, microstructural evolution of this region is of a great importance and is
te
d
mainly dictated by the plastic deformation and temperature.
3.4. Microstructure of the joint interface
Ac ce p
Optical micrographs of the joint interface for the sample welded under the condition of 1100 rpm- 5 s are shown in Figs. 7(a)-(c). The enlarged SEM micrographs of selected regions in Figs. 7(a) and (c) are presented in Figs. 7(d) and (e), respectively.
13 Page 13 of 33
ip t cr us an M d te Ac ce p
Fig. 7. (a, b, c) optical and (d, e, f, g) SEM images of the joint cross-section after failure for the steel part of the specimen with the welding condition of 1100 rpm- 5 s.
Although the low magnification images of the joint interface in Figs. 7(d) and (e) indicated that no IMC was formed, SEM micrographs with higher magnification were taken to confirm this
14 Page 14 of 33
finding (Figs. 7(f) and (g)). As can be seen in Figs. 7(e) and (g), a swirl layered structure was formed as a result of the stirring and mixing effects of the tool pin at the joint interface. According to the short diffusion time of 5 s for this specimen, it seems that enough time has not
ip t
been provided for Al and Fe atoms to interdiffuse through the layers and form Al/Fe IMCs.
Therefore, it is expected that these layers have been consisted of a mechanical mixture of Al and
cr
steel layers. The joint cross sections of the specimen welded under the condition of 1100 rpm- 10
us
s are given in Figs. 8(a)-(i). According to the SEM micrographs shown in Figs. 8(g)-(i), IMCs were formed at the joint interface. As it can be observed in Figs. 8(d) and (f), these IMCs had a
an
non-continuous morphology while some relatively continuous layers were formed at some parts of the joint interface (Figs. 8(e) and (g)). Microstructural observation of the joint interface for
M
this specimen also revealed the formation of the swirl layered structure (Fig. 8(e)). The high
d
magnification SEM images of the peripheral regions of the remained Al layer at the location of
te
final fracture indicated the presence of Al/Fe IMCs (Figs. 8(g) and (i)). The interfacial microstructures of the sample under the welding condition of 900 rpm- 15 s are presented in Fig.
Ac ce p
9. Based on the SEM micrographs shown in Figs. 9(d) and (e), IMCs were formed with a more continuous morphology and higher thicknesses compared to the specimen with the welding condition of 1100 rpm- 10 s.
15 Page 15 of 33
ip t cr us an M d te Ac ce p
Fig. 8. (a, b, c) optical and (d, e, f, g, h, i) SEM images of the joint cross-section after failure for the steel part of the specimen with the welding condition of 1100 rpm- 10 s.
16 Page 16 of 33
ip t cr us an M d te Ac ce p
Fig. 9. (a, b, c) optical and (d, e, f, g, h) SEM images of the joint cross-section after failure for the steel part of the specimen with the welding condition of 900 rpm- 15 s.
Another key finding for the specimen with the welding condition of 900 rpm- 15 s was its different fracture type. Referring to the SEM micrographs shown in Figs. 7 and 8, there was no 17 Page 17 of 33
evidence of crack propagation inside the IMCs at the joint interface of sheets for the welds made with the welding conditions of 1100 rpm- 5 s and 1100 rpm- 10 s. However, based on the SEM micrographs taken from the peripheral regions of the remained Al layer on the steel sheet of the
ip t
sample with the welding condition of 900 rpm- 15 s (Figs. 9(d) and (f)), Al-5083 and St-12
sheets were first separated from each other in a situation that the Al-5083 side of the interface
cr
has moved perpendicular to the applied loading direction (the loading direction was out of
us
paper). Furthermore, IM reaction layers were observed on the separated surface of both St-12 and Al-5083 sheets (Figs. 9(d) and (f)). This observation evidences that the crack propagation
an
occurred inside the IM layer during the tensile-shear testing and then moved to the Al-5083 side of the joint where the thickness was decreased as a result of the pin penetration. The EDS
M
analysis results of the IM reaction layers of the 1100 rpm- 10 s and 900 rpm- 15 s FSSW
Ac ce p
te
d
specimens are given in Fig. 10.
Fig. 10. SEM micrograph in association with the EDS analysis result of the intermetallic layer formed at the joint interface of the sample with the welding condition of (a) 1100 rpm- 10 s before and (b) 900 rpm- 15 s after etching by the 2% nital etchant solution. 18 Page 18 of 33
As can be seen, IMCs contained 77 at.-% Al and 23 at.-% Fe at the joint interface of the 1100 rpm- 10 s sample (Fig. 10(a)) while they were composed of 78 at.-% Al and 22 at.-% Fe for the 900 rpm- 15 s specimen which was very close to the chemical composition of IM layer of the
layers have been formed during the FSSW of both specimens.
ip t
sample with the welding condition of 1100 rpm- 10 s. This indicates that the same type of IM
cr
Referring to Fig. 10(b) showing the etched microstructure of St-12 base metal matrix for the 900
us
rpm- 15 s sample, in comparison with the average grain size of ~42 µm for the St-12 base sheet, noticeably smaller grains are being observed in the steel part adjacent to the joint interface which
an
represents the formation of a structure/ substructure in the steel sheet. Furthermore, the joint interface shows a wavy (serrated) morphology. The observed phenomena may evidence a plastic
M
deformation in the steel side of the joint adjacent to the interface. It is noteworthy that however
d
the pin tip did not penetrate into the steel sheet, this region can undergo plastic deformation
te
probably due to the stirring action of the pin in the surrounding materials and the short distance
Ac ce p
of the pin tip to the joint interface (~0.2 mm).
3.5. Al-5083 grain structure
The applied FSSW process not only resulted in the interdiffusion of Al and Fe atoms and formation of Al/Fe IMCs at the joint interface, but also changed the microstructure of the Al5083 base metal near the weld exit-hole. The Al sides of the samples welded at the rotational speed of 1100 rpm with different dwell times of 5, 10 and 15 s were investigated using optical microscopy. The associated microstructures directly adjacent to the weld exit-hole, where the final fracture occurred, are presented in Fig. 11. As the general feature of the micrographs showed, very fine and equiaxed grains have been formed adjacent to the weld exit-hole which is
19 Page 19 of 33
the characteristic of stirred zone (SZ) and can be attributed to the dynamic recrystallization (DRX). DRX occurred as a result of the simultaneous effects of the pin stirring and frictional heat during the FSSW process. Microstructural analysis for the specimen welded under the
ip t
condition of 1100 rpm- 15 s indicated the formation of some relatively coarse grains distributed in a matrix of finer grains which could be caused by abnormal grain growth phenomenon (Fig.
Ac ce p
te
d
M
an
us
cr
11(c)).
Fig. 11. The microstructures of the weld exit-hole periphery for specimens welded at the rotational speed of 1100 rpm for the dwell times of (a) 5, (b) 10 and (c) 15 s. 20 Page 20 of 33
4. Discussion 4.1. Formation of IM reaction layer
ip t
Metallurgical bonds are formed as a result of the atomic diffusion across the joint interface. Al/Fe IMCs will be formed as a consequence of the atomic diffusion if enough energy is
cr
provided at the joint interface. As shown in Fig. 7, welding condition of 1100 rpm- 5 s did not
us
lead to the formation of IMCs at the joint interface which could be due to the short dwell time of 5 s even at the interface maximum temperature of 390 oC (Fig. 6). However, when longer dwell
an
time of 10 s was used at the same rotational speed, IMCs were formed as a non-continuous morphology at the joint interface (Fig. 8). This indicates that at the rotational speed of 1100 rpm
M
and dwell time of 10 s, the conditions for formation of a continuous IM layer have not been
d
provided which led to a non-continuous metallurgical bond at the joint interface. As shown in
te
Fig. 10(a), the average thickness of the IM layer for this specimen was ~2.3 µm. Microstructural analysis of the joint interface for the sample with the welding condition of 900 rpm- 15 s
Ac ce p
revealed the formation of a relatively continuous IM layer with a thickness of ~2.9 µm. The observed increasing trend in the thickness of IM layer at the joint interface from 2.3 to 2.9 µm clearly indicated that compared to the specimen welded at 1100 rpm- 10 s, the processing condition of 900 rpm- 15 s promoted the diffusional growth of the IM layer to a greater thickness. Given the processing conditions presented for these two joints (1100 rpm- 10 s and 900 rpm- 15 s), it is revealed that even though the rotational speed decreased from 1100 to 900 rpm, enhancement of the dwell time from 10 to 15 s increased the thickness of IM layer from 2.3 to 2.9 µm in one hand and changed the layer morphology from non-continuous to continuous on the other hand. Formation of either a continuous or non-continuous IM layer as well as the IM
21 Page 21 of 33
layer thickness at the joint interface is controlled by two major mechanisms of nucleation and growth. Nucleation and growth of IMCs are both affected by the temperature and plastic deformation (stored energy) at the joint interface. Enhancement of the temperature provides the
ip t
activation energy for the formation of IMCs in one hand and facilitates the IMCs diffusional growth on the other hand. The plastic deformation and the related stored energy at the joint
cr
interface during the welding process could also affect the IMCs nucleation. As reported by
us
Springer et al. (2011), variation in the tool rotational and travel speeds in the FSW process influenced the plastic deformation and consequently the nucleation of Al/Fe IMCs at the joint
an
interface. Furthermore, the stored energy at the joint interface may increase the IM layer growth rate as a result of the atomic diffusion through the high energy paths such as grain boundaries
M
and dislocations. In the FSSW process, the welding parameters which influence the IM layer
d
formation are rotational speed and dwell time. Given the IM layers formed at the interface of
te
1100 rpm- 10 s and 900 rpm- 15 s specimens (Figs. 8 and 9), the required temperature and plastic deformation were provided for the nucleation of IMCs. In the present study, as the dwell time
Ac ce p
was increased to 15 s, more diffusion time was provided for the growth of IMCs along and across the joint interface and the non-continuous IMCs became sufficiently dense to form a continuous layer.
The same structure of local and non-continuous Al/Fe IMCs as formed in 1100 rpm -10 s specimens was also observed by Naoi and Kajihara (2007) in solid Al/steel couples. They believed that these IMCs were resulted from the early stage of the interfacial IMC growth.
4.2. Welding parameters/failure load relationship
22 Page 22 of 33
On the basis of the average failure loads of the specimens presented in Fig. 5, for both of the rotational speeds, the joint strength first improved and then declined with the increase in dwell time. As shown in Fig. 7, there was no evidence of IMC formation at the joint interface of the
ip t
sample welded under the condition of 1100 rpm- 5 s. As was already pointed out, such
phenomenon was attributed to the short dwell time of 5 s which did not provide enough time for
cr
Al and Fe atoms to interdiffuse across the joint interface and to form the Al/Fe IMCs. The
us
microstructure of the joint interface for the specimen with the welding condition of 900 rpm- 5 s was not presented in the present paper. However, compared to the interface maximum
an
temperature of 390 oC for the welding condition of 1100 rpm- 5 s, lower interface maximum temperature of 360 oC was recorded for the sample with the processing condition of 900 rpm- 5 s
M
as a result of the lower rotational speed. Accordingly, it is reasonable to expect an IMC free
d
interface for the specimen with the welding condition of 900 rpm- 5 s. Bozzi et al. (2010) believe
te
that if enough IMCs are not formed at the joint interface of Al/steel joints, occurrence of a sudden fracture is more possible because of the sudden change in the chemical composition at
Ac ce p
the joint interface by moving from the Al sheet towards the steel side. Therefore, the low fracture loads for the welds with the processing conditions of 900 rpm– 5s and 1100 rpm- 5 s could be attributed to their IMC free interfaces. As the dwell time was enhanced from 5 to 10 s at the rotational speed of 1100 rpm, a thin IM layer with the thickness of 2.3 µm and a non-continuous morphology was formed at the joint interface and consequently the failure load improved from 1830 to 3632 N which was approximately twice as much. Achievement of the stronger joints under longer dwell times (10 s compared to 5 s) may be attributed to the formation of IMCs at the joint interface which was affected by the following phenomena:
23 Page 23 of 33
(i) More diffusion time was provided for Al and Fe atoms to move through the joint interface and form IMCs (Peng et al., 1999). (ii) The maximum temperature at the joint interface increased with enhancement of the dwell
ip t
time. Consequently, the growth constant of the IM layer increased.
It is noteworthy that compared to the sample with the welding condition of 1100 rpm- 10 s, the
cr
joint strength decreased gradually down to ~2360 N for the specimen with the welding condition
us
of 1100 rpm- 15 s. As the results of the interface maximum temperature in Fig. 6 showed, enhancement of the rotational speed and increase in the dwell time both resulted in the
an
generation of higher interface maximum temperatures. Therefore, enhancement of the dwell time from 10 to 15 s at the rotational speed of 1100 rpm increased the interface maximum
M
temperature. The microstructure of the joint interface for the sample with the welding condition
d
of 1100 rpm- 15 s was not shown in the present study, but as the interfacial microstructural
te
observation of the sample with the same dwell time and lower rotational speed (900 rpm- 15 s) showed, crack propagated inside the IM layer during the tensile testing (Figs. 9(d) and (f)). It is
Ac ce p
further noteworthy that compared to interface maximum temperature of 410 oC for the sample with the welding condition of 900 rpm- 15 s, higher temperature of 420 oC was recorded for the specimen with the welding condition of 1100 rpm- 15 s. This indicates that IM layer with the thickness even more than 2.9 µm could be formed at the joint interface of the sample with the processing condition of 1100 rpm- 15 s. In general, the joint strengths were improved after the formation of a non-continuous IM layer with the thickness less than 2.3 µm and then, decreased significantly when an IM layer with greater thickness and a relatively continuous morphology was formed at the interface of the sheets. The sharp loss of the joint strength for the samples with the dwell times of 15 s can be rationalized due to the formation of relatively thicker IM layers
24 Page 24 of 33
with less toughness at the joint interface in which crack needs less energy to propagate during the tensile-shear test. When the IMCs are formed with a non-continuous morphology at the interface, separation of the bonded sheets needs the propagation of cracks through IMCs as well as the
ip t
aluminum and/or steel base metals adjacent to the joint interface. It is well known that aluminum and steel base materials have greater toughness than the brittle Al/Fe IMCs. This increases the
cr
required energy for the propagation of cracks inside them. On the other hand, the FSSW process
us
enhances the toughness of the SZ in the base metals adjacent to the joint interface due to the formation of fine grained microstructure in this region. As a result, improvement of the Al and
an
steel toughness adjacent to the interface enhances the bond strength. Furthermore, according to the microstructural observations, morphological change of the IM layer from non-continuous to
M
continuous was accompanied by the increase of the thickness from ~2.3 to ~2.9 µm. As reported
d
by Springer (2011), with increase in the thickness of the IM layer at the joint interface of friction
te
stir welded Al/steel joints, the energy-consuming crack-interception at the IMC/steel interface
Ac ce p
became less frequent and consequently the joint strength decreased. Therefore, from the joint strength point of view, it can be concluded that the IM layer thickness of 2.3 µm was a critical thickness. The observed relationship between the Al/Fe IM layer thickness and the joint strength is inconsistent with the results presented by Movahedi et al. (2011) for the roll bonded Al -1100 alloy/St-12 sheets and also the results reported by Tanaka et al. (2009) on the FSW of 7075-T6 Al alloy to mild steel. They mentioned that the formation of IM phases, regardless of their thickness, reduced the bond quality of the joints. On the other hand, Danesh Manesh and Karimi Taheri (2003), Padmanabhan et al. (2008), Zhang et al. (2007) and Mathieu et al. (2006) have stated that the formation of IM layer with the thicknesses less than 10 µm imposed no detrimental effect on the joint strength and even improved the quality of 25 Page 25 of 33
the joint. They have also indicated that the joint strength followed a decreasing variation when the IM layer thickness exceeded 10 µm. In another study, Movahedi et al. (2013) studied the effect of the annealing treatment on the joint strength of FSWed Al-5083/St-12 alloy sheets and
ip t
found the IM layer thickness of 2.6 µm as the critical thickness in which the optimum tensileshear strength was achieved. This thickness is very close to the critical thickness of 2.3 µm in the
cr
present study.
us
According to the failure loads of 1830 and 2870 N for samples with the welding conditions of
an
1100 rpm- 5 s and 900 rpm- 15 s, respectively, it can be concluded that the formation of a relatively thick Al/Fe IM layer with the thickness of 2.9 µm at the joint interface may be
M
beneficial for improving the joint strength compared to the condition in which no IMC has been formed.
d
As previously pointed out, the maximum failure loads attained for the rotational speeds of 1100
te
and 900 rpm were related to the dwell times of 10 and 12 s, respectively. As it is clear, shorter
Ac ce p
dwell time was needed to reach the optimum tensile-shear strength when a higher rotational speed was applied. In fact, as shown in Fig. 6, increase in the rotational speed resulted in the generation of higher interface maximum temperature and decreased the time required for the formation of an IM layer with the critical thickness at the joint interface. Therefore, at higher rotational speeds, shorter dwell times are needed to obtain the critical IM layer thickness at which the optimum tensile-shear strengths are achieved. As shown in Fig. 5, for a certain dwell time, the tensile-shear strengths of the joints welded at rotational speed of 1100 rpm were lower than those welded at the rotational speed of 900 rpm. The observed finding is in agreement with the results reported by Lathabai et al. (2006), Tozaki et al. (2007) for the similar FSSW of Al alloy sheets. They reported that both the cross-tension 26 Page 26 of 33
and tensile-shear strengths were decreased with increasing the tool rotational speed. As the rotational speed increases, the heat input will increase which subsequently gives rise to the grain growth and the joint strength decreases. As Fig. 11 shows, the final fracture occurred from the
ip t
SZ in the periphery of the exit-hole. Therefore, the microstructural observation of this region is of a great importance since it determines the load required for the final fracture to occur. The
cr
enlarged micrographs of the exit-hole periphery in Fig. 11 revealed the formation of some
us
noticeably coarse grains distributed in the fined-grain matrix of SZ for the specimen with the welding condition of 1100 rpm- 15 s.
an
From what mentioned above, it can be concluded that the sharp loss of the joint strength for the specimens with longer dwell times could not be completely attributed to the formation of a
M
relatively thick and continuous IM layer at the joint interface, but also the grain growth
Ac ce p
5. Conclusions
te
strength.
d
phenomenon adjacent to the weld exit-hole would have also a detrimental effect on the joint
For both of the applied rotational speeds of 900 and 1100 rpm, the joint strength first improved and then declined with the increase in the dwell time. Increasing trend of the tensile-shear strength as a function of the dwell time was attributed to the formation of a non-continuous IM layer with the thickness less than 2.3 µm at the joint interface. Decreasing trend was also related to the formation of a relatively continuous IM layer with the thickness of 2.9 µm and the growth of Al grains adjacent to the weld exit-hole.
27 Page 27 of 33
By enhancement of the rotational speed from 900 to 1100 rpm, the required dwell time to achieve the optimum tensile-shear strengths decreased from 12 to 10 s. For a given dwell time, the specimens welded at the rotational speed of 1100 rpm were
cr
ip t
accompanied by lower tensile-shear strengths.
Acknowledgments
us
The authors wish to thank the research board of Sharif University of Technology for the financial
an
support and the provision of the research facilities used in this work.
M
References
Bozzi, S., Helbert-Etter, A.L., Baudin, T., Criquic, B., Kerbiguet, J.G., 2010. Intermetallic
te
d
compounds in Al 6016/IF-steel friction stir spot welds. Mater. Sci. Eng. A 527, 4505-4509.
Ac ce p
Chen, Y.C., Gholinia, A., Prangnell, P.B., 2012. Interface structure and bonding in abrasion circle friction stir spot welding: A novel approach for rapid welding aluminium alloy to steel automotive sheet. Mater. Chem. Phys. 134, 459-463. Danesh Manesh, H., Karimi Taheri, A., 2003. The effect of annealing treatment on mechanical properties of aluminum clad steel sheet. Mater. Des. 24, 617-622. Lathabai, S., Painter, M.J., Cantin, G.M.D., Tyagi, V.K., 2006. Friction spot joining of an extruded Al-Mg-Si alloy. Scripta Mater. 55, 899-902.
28 Page 28 of 33
Mathieu, A., Matteï, S., Deschamps, A., Martin, B., Grevey, D., 2006. Temperature control in laser brazing of a steel/aluminium assembly using thermographic measurements. NDT & E Int.
ip t
39, 272-276. Merzoug, M., Mazari, M., Berrahal, L., Imad, A., 2010, Parametric studies of the process of
cr
friction spot stir welding of aluminium 6060-T5 alloys. Mater. Des. 31, 3023-3028.
us
Movahedi, M., Kokabi, A.H., Seyed Reihani, S.M., 2011. Investigation on the bond strength of
an
Al-1100/St-12 roll bonded sheets, optimization and characterization. Mater. Des. 32, 3143-3149. Movahedi, M., Kokabi, A.H., Seyed Reihani, S.M., Cheng, W.J., Wang, C.J., 2013. Effect of
M
annealing treatment on joint strength of aluminum/steel friction stir lap weld. Mater. Des. 44, 487-492.
d
Naoi. D., Kajihara, M., 2007. Growth behavior of Fe2Al5 during reactive diffusion between Fe
te
and Al at solid-state temperatures. Mater. Sci. Eng. A 459, 375-382.
Ac ce p
Padmanabhan, R., Oliveira, M.C., Menezes, L.F., 2008. Deep drawing of aluminium-steel tailorwelded blanks. Mater. Des. 29, 154-160. Peng, X.K., Wuhrer, R., Heness. G., Yeung, W.Y., 1999. On the interface development and fracture behavior of roll bonded copper/aluminium metal laminates. J. Mater. Sci. 34, 20292038.
Springer, H., 2011. Fundamental research into the role of intermetallic phases in joining of aluminium alloys to steel. PhD Thesis. Ruhr-University Bochum, Bochum, Germany.
29 Page 29 of 33
Springer, H., Kostka, A., dos Santos, J.F., Raabe, D, 2011. Influence of intermetallic phases and Kirkendall-porosity on the mechanical properties of joints between steel and aluminium alloys.
ip t
Mater. Sci. Eng. A 528, 4630-4642. Sun, Y.F., Fujii, H., Takaki, N., Okitsu, Y., 2013. Microstructure and mechanical properties of
cr
dissimilar Al alloy/steel joints prepared by a flat spot friction stir welding technique. Mater. Des.
us
47, 350-357.
Tanaka, T., Morishige, T., Hirata, T., 2009. Comprehensive analysis of joint strength for
an
dissimilar friction stir welds of mild steel to aluminum alloys. Scripta Mater. 61, 756-759.
M
Tozaki, Y., Uematsu, Y., Tokaji, K., 2007. Effect of tool geometry on microstructure and static strength in friction stir spot welded aluminium alloys. Int. J. Mach. Tools Manuf. 47, 2230-2236.
d
Yuan, W., Mishra, R.S., Webb, S., Chen, Y. L., Carlson, B., Herling, D.R., Grant, G.J., 2011.
te
Effect of tool design and process parameters on properties of Al alloy 6016 friction stir spot
Ac ce p
welds. J. Mater. Proc. Technol. 211, 972-977. Zhang, H.T., Feng, J.C., He, P., Hackl, H., 2007. Interfacial microstructure and mechanical properties of aluminium-zinc-coated steel joints made by a modified metal inert gas weldingbrazing process. Mater. Charact. 58, 588-592.
30 Page 30 of 33
Figure Captions Fig. 1. Schematic illustration and dimensions of the samples with the associated hole drilled in
ip t
the steel sheet in order to record the thermal history at the interface of the sheets. Fig. 2. (a) General view of friction stir spot welded specimen before the tensile-shear testing and
cr
(b) typical macroscopic cross-section of FSSW specimen before testing in association with the
us
enlarged micrographs showing different regions of the joint interface. Fig. 3. (a) Steel and (b) Al part of the joint after tensile-shear test.
an
Fig. 4. Macroscopic cross-sections of (a) Steel and (b) Al part of the joint after tensile-shear test.
M
Fig. 5. Average fracture loads of the welds processed at the rotational speeds of 900 and 1100 rpm for various dwell times.
te
d
Fig. 6. The maximum temperatures measured at the joint interface. Fig. 7. (a, b, c) optical and (d, e, f, g) SEM images of the joint cross-section after failure for the
Ac ce p
steel part of the specimen with the welding condition of 1100 rpm- 5 s. Fig. 8. (a, b, c) optical and (d, e, f, g, h, i) SEM images of the joint cross-section after failure for the steel part of the specimen with the welding condition of 1100 rpm- 10 s. Fig. 9. (a, b, c) optical and (d, e, f, g, h) SEM images of the joint cross-section after failure for the steel part of the specimen with the welding condition of 900 rpm- 15 s. Fig. 10. SEM micrograph in association with the EDS analysis result of the intermetallic layer formed at the joint interface of the sample with the welding condition of (a) 1100 rpm- 10 s before and (b) 900 rpm- 15 s after etching by the 2% Nital etchant solution. 31 Page 31 of 33
Fig. 11. The microstructures of the weld exit-hole periphery for specimens welded at the
Ac ce p
te
d
M
an
us
cr
ip t
rotational speed of 1100 rpm for the dwell times of (a) 5, (b) 10 and (c) 15 s.
32 Page 32 of 33
Table Captions Table 1. The chemical compositions (in wt.-%) of the Al-5083 H321 Al alloy and St-12 sheets
ip t
used in the present work
Ac ce p
te
d
M
an
us
cr
Table 2. Mechanical properties of the Al-5083 H321 alloy and St-12 sheets
33 Page 33 of 33