Improving weld strength of arc-assisted ultrasonic seam welded Mg/Al joint with Sn interlayer Xiangyu Dai, Hongtao Zhang, Bo Wang, Ang Ji, Jihou Liu, Jicai Feng PII: DOI: Reference:
S0264-1275(16)30219-2 doi: 10.1016/j.matdes.2016.02.095 JMADE 1445
To appear in: Received date: Revised date: Accepted date:
2 December 2015 14 February 2016 17 February 2016
Please cite this article as: Xiangyu Dai, Hongtao Zhang, Bo Wang, Ang Ji, Jihou Liu, Jicai Feng, Improving weld strength of arc-assisted ultrasonic seam welded Mg/Al joint with Sn interlayer, (2016), doi: 10.1016/j.matdes.2016.02.095
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.
ACCEPTED MANUSCRIPT Improving weld strength of arc-assisted ultrasonic seam welded Mg/Al joint with Sn interlayer
IP
T
Xiangyu Daia,b, Hongtao Zhanga,b,*, Bo Wangc, Ang Jib, Jihou Liub, Jicai Fenga,b a
Technology, Harbin 150001, China b
Laboratory of Special Welding Technology of Shandong Province, Harbin Institute
NU
of Technology at Weihai, Weihai 264209, China.
School of Materials Science and Engineering, Kunming University of Science and
MA
c
SC R
State Key Laboratory of Advanced Welding and Joining, Harbin Institute of
Technology, Kunming 650093, China
D
*Corresponding author.
CE P
TE
E-mail address:
[email protected].
Abstract
Welding of Mg and Al alloys is hugely challenging due to the formation of hard
AC
and brittle Mg-Al intermetallic compounds (IMCs). In this research, gas tungsten arc welding (GTAW) assisted hybrid ultrasonic seam welding with Sn interlayer was designed to join Mg to Al alloys. Effects of Sn interlayer and GTAW current on the microstructure and mechanical properties of the joints were investigated in detail. The results indicated that Sn interlayer could restrain the formation of Mg-Al IMCs, which were replaced by Mg2Sn and Sn-based solid solution. The peak load of the joints increased with the GTAW current increasing and then decreased dramatically at higher GTAW current. The maximum peak load of the joints with Sn interlayer was approximately 1.3kN, which about 30% increase over the joints without Sn interlayer. 1
ACCEPTED MANUSCRIPT All the joints failure occurred by the interface mode at the Al/Sn interface and the fracture patterns exhibited entirely brittle fracture mode with cleavage facet feature
IP
T
surface.
SC R
Keywords: Aluminum alloy, Magnesium alloy, Ultrasonic welding, Sn interlayer, Lap shear strength 1. Introduction
NU
The successful joining of dissimilar light materials of Mg and Al alloys is
MA
advantage for producing lightweight structures, pushing forward the project of energy-saving and emission-reduction in the transportation and aerospace
D
industries[1-3]. Unfortunately, traditional fusion welding technologies, such as
TE
resistance spot welding, arc welding and laser welding, are difficult to apply
CE P
successfully to the joining of Mg and Al dissimilar metals[4-6]. The fusion welding between Mg and Al alloys usually results in large amount of brittle Mg-Al
AC
intermetallic compounds (IMCs) formation, which preferentially acts as the source of microcracks to deteriorate the mechanical properties of the joint[7-9]. In order to achieve a good combination of the properties of Mg and Al alloys, the development of reliable joining process between these metals is crucial. Therefore, solid state technologies including friction stir welding and ultrasonic welding, have received much attention as alternative joining technologies for Mg-Al[10-11]. Friction stir welding was a relatively new solid-state joining technique developed and patented by Thomas in The Welding Institute in 1991[12]. In the process of FSW, a rotating shouldered tool with a profiled pin moved between the sheets of pieces to 2
ACCEPTED MANUSCRIPT be joined. As the rotating tool traveled along the weld line, frictional heat was generated between the base material and the tool shoulder. However, this heat was
IP
T
significantly lower than in fusion welding methods, which could help FSW to avoid
SC R
many defects appearing in fusion welding[13]. In consequence, FSW was demonstrably better than traditional welding technique in joining dissimilar material combinations such as Mg-Al. However, the FSW technique had undesirable aspects
NU
including the keyhole left by the tool probe and the reduction of the top sheet
MA
thickness[14].
Ultrasonic welding was another kind of solid state joining technique that used
D
high-frequency ultrasonic vibrations under a modest pressure to induce oscillating
TE
shears between the faying surfaces to produce metallurgical bonds[15-16].The weld
CE P
defects, such as formation of brittle IMCs, high levels of welding distortions and HAZ damage in fusion welding could be typically avoid. In addition, ultrasonic welding
AC
was an attractive point joining technique for light alloy, as it is far more efficient than resistant spot welding, using only 0.6-1.5kJ per weld[17-18]. It was also more efficient than FSW, because the energy was predominantly generated at the weld-line. Recently, a number of researches reported on the joining of dissimilar light metals by ultrasonic spot welding[19-21]. A sound joint without defects could be obtained at usual welding conditions. However, there were surprisingly few researches in joining dissimilar metals by ultrasonic seam welding, even though it possessed higher efficiency and weld continuity characteristics. Conventional ultrasonic seam welding could only be applied to join metal foils or thin plate (< 0.5mm) due to the limitation 3
ACCEPTED MANUSCRIPT of power of the welding system. For joining thicker dissimilar metal sheets, the development of high power ultrasonic welding equipment was a key factor[22].
IP
T
Unfortunately, the high ultrasonic power welding systems results in high cost and
SC R
inconvenient manipulation of the equipment. An effective solution to this problem was to utilize an additional form of energy. Therefore, in our previous study[23], gas tungsten arc welding (GTAW) was introduced into ultrasonic seam welding for the
NU
first time. The preceding arc with relatively low heat input could preheat the sheet to
MA
enhance the ultrasonic weldability. And reliable joining of Mg and Al sheets with 1mm thickness was achieved without improving the ultrasonic power. Unfortunately,
D
the hard and brittle Mg-Al IMCs such as Al12Mg17 and Al3Mg2 were still formed,
TE
which deteriorated the mechanical properties of the joint. To restrict the formation of
CE P
Mg-Al IMCs, other elements should be introduced into the weld to serve as alloying elements or barrier materials. In most researches[24-26], Zn interlayer was used for
AC
the welding of dissimilar Mg alloys to Al alloys and the lap shear strength of the joints was improved obviously. These studies could be extended to some other interlayers which could interact with Mg and Al. Liu et al.[27] in TIG welding of Mg-to-Al used Sn to improve the wettability of Mg and Al alloys during the welding process. Patel et al.[28] improved the strength of dissimilar joints of Mg-to-Al by using Sn interlayer during ultrasonic spot welding. However, it is still unclear how Sn interlayer would affect the microstructure of arc-assisted ultrasonic seam welded Mg-to-Al joints, whether the intermetallic layer would form and improve the mechanical properties of the joints. 4
ACCEPTED MANUSCRIPT Therefore, in this paper, Sn interlayer was introduced into arc-assisted ultrasonic seam welding to improve the mechanical properties of Mg-to-Al joints by suppressing
IP
T
the formation of brittle Mg-Al IMCs. The selection of Sn interlayer was based on the
SC R
binary phase diagrams and previous works. This study determined if Sn could inhabit the formation of brittle Mg-Al IMCs and instead form high strength bearing, Sn-based IMCs. In addition, the effect of Sn interlayer and GTAW current on the microstructure
NU
and mechanical properties of the joints was investigated in detail.
MA
2. Experimental
The base materials employed in the experience were Mg AZ31B and Al 6061
D
sheets with dimension of 150mm × 60mm × 1mm. 0.3mm thick pure Sn interlayer
TE
was used as interlayer between the Al and Mg alloy plate. Prior to welding, the plates
CE P
and Sn interlayer were degreased and ground. The sheet specimens were clamped between the rotating sonotrode and the anvil in a lap configuration with an overlap
AC
distance of 20mm. The schematic diagram of arc-assisted ultrasonic seam welding system was shown in Fig. 1(a). The K-type thermocouples of 0.5mm diameter were used to measure the temperature of the weld interface, as shown in Fig. 1(b). It mainly consisted of ultrasonic seam welding system and GTAW system. The preceding GTAW torch was attached adjacent to rotating sonotrode and carried out with an angle of 45° adjacent to the joint. The rotating sonotrode tips serrated surface to improve gripping of the lapped sheets. The GTAW assisted hybrid ultrasonic seam welding was carried out at GTAW current ranging from 30 to 50A at a constant of GTAW voltage of 9.5V, welding speed of 8mm/s, pressure of 0.44MPa, amplitude of 5
ACCEPTED MANUSCRIPT 20μm, vibration frequency of 20kHz as listed in Table 1. The distance between arc and sonotrode was approximately 15mm. The welding heat input P was mainly
could be expressed by the following equation[29]:
SC R
Pu k SFAf
IP
T
consisted of arc heat input Pa and ultrasonic heat input Pu. The ultrasonic heat input Pu
(1)
NU
P Pa Pu UI k SFAf
MA
(2)
where U was the GTAW voltage, I was the GTAW current, k was the proportional
D
coefficient, μ was the friction coefficient, S was the contact area, F was the pressure,
TE
A was the amplitude, f was the vibration frequency. Our previous research[23]
CE P
indicated that arc-assisted ultrasonic seam welding of Mg and Al alloy could be achieved and the above parameters was adjusted to optimize the process. And this
AC
work focuses on the effect of Sn interlayer and GTAW current on the microstructure and mechanical properties of the joints. The metallurgical cross-section was observed by scanning electron microscope (SEM) equipped with energy dispersive X-ray spectroscopy (EDS) after standard grinding and polishing procedures. The Vickers hardness was measured along the transverse cross-section of the welded specimen using a load of 200gf and dwell time of 10s. Five 10mm wide specimens perpendicular to welding direction were prepared and subjected to tensile testing machine with a speed of 0.5mm/min at room temperature. In addition, the fracture surface was examined by a SEM equipped with 6
ACCEPTED MANUSCRIPT EDS. X-ray diffraction (XRD) was also carried out on the surface after tensile shear tests to investigate the phase evolution.
IP
T
3. Results
SC R
3.1 Weld appearance and cross section
Fig. 2 shows the typical weld appearance of the joint produced at GTAW current of 35A. It could be seen that and the weld exhibited good “knurl pattern” surface
NU
morphology. Weld defects, such as slipping and tool sticking, were not observed. And
MA
the aluminum alloy sheet was not melted during welding. The preceding GTAW with low heat input preheated the aluminum alloy sheet. The rotating sonotrode tips
D
serrated surface to improve gripping of the lapped sheets and prevented slipping
TE
between the sonotrode and the sheet. Previous research[23] indicated that the sheet
CE P
hardness was reduced and the plastic deformation was more extreme due to the addition of GTAW.
AC
Fig. 3 shows the typical SEM micrographs of cross-section of the joints with increasing the GTAW current from 30A to 45A. At a low GTAW current of 30A, the plastic deformation was localized at the weld interface, and the weld join-line remained macroscopically flat. This indicated that the temperature at the weld interface was not yet risen sufficiently, giving rise to inadequate bonding. As the GTAW current increased to 35A and 40A, the weld interface temperature rose and the weld join-line exhibited a wave-like or swirl-like appearance due to the heavy plastic deformation. In addition, Sn interlayer was gradually squeezed out from the weld interface with increasing the GTAW current. As the GTAW current increased to 45A, 7
ACCEPTED MANUSCRIPT Sn interlayer was almost squeezed out from the weld interface. Bhole et al.[30] also observed squeezing of Sn interlayer during dissimilar ultrasonic spot welding of Mg
IP
T
to galvanized and ungalvanized steel. This was because the high weld interface
SC R
temperature along with higher plastic deformation resulted into squeezing the Sn interlayer out from the weld interface. At high GTAW current levels, thinning of the sheets in the weld area were observed, which have been described as the main reason
NU
for the decrease in the joint performance[28]. The joint consisted of two distinct
MA
interfacial reaction zones: Mg/Sn interface and Sn/Al interface. The interfacial reaction zones usually were the weak zone during the joining of dissimilar metals.
D
Thus, the microstructure characteristics of the interfacial reaction zones should be
TE
investigated emphatically in the following sections.
CE P
3.2 Microstructure evolution of interfacial reaction zones: Fig. 4 shows the SEM backscattered electron micrographs of the Al/Sn interface
AC
with different GTAW currents. It could be seen that there was no continuous and stable reaction layer along the Al/Sn interface. EDS line-scan analysis is performed along the Al/Sn interface of the joint at GTAW current of 40A, as shown in Fig. 5. The EDS line-scan curves exhibited a continuous and smooth transition of atomic content from 0 to 100% for Al and Sn element along the Al/Sn interface. The results confirmed the fact that there was no continuous reaction layer along the Al/Sn interface. The absence of reaction layer was due to almost no intersolubility between Al and Sn atoms. According to Al-Sn binary phase diagram in Fig. 6(a), the solubility of Sn and Al was nearly zero. Therefore, almost no metallurgical bonding was 8
ACCEPTED MANUSCRIPT realized along the Al/Sn interface. Fig. 7 shows the XRD profiles of the fracture surface of Al side after the tensile lap shear tests for the joints produced at the GTAW
IP
T
current of 40A. It was indicated that only Sn and Al phases were detected and no
SC R
intermetallic compounds were formed along the Al side, which also confirmed that the correctness of above EDS line-scan analysis. The formation of the joint was mainly due to the plastic deformation and mechanical interlocking. At lower GTAW
NU
current, the plastic deformation was localized at the Al/Sn interface and the weld
MA
join-line remained macroscopically flat. As the GTAW current was increased, the plastically deforming region expanded and the weld join-line exhibited a wave-like or
D
swirl-like appearance owing to the heavy plastic deformation. The wave-like or
TE
swirl-like structure could lead to a mechanical interlocking and therefore enhanced
CE P
the joint strength. The joint strength was a combined effect of waves and micro-bonds along the weld interface[31]. Therefore, the above analysis indicated that the Al/Sn
AC
interface of the joints was the weakest zone, which could be seen later from the failure behavior as well.
The SEM micrographs, acquired in black scattered electrons (BSE) mode, of Mg/Sn interface with different GTAW currents, are illustrated in Fig. 8. At a low GTAW current of 30A, a reaction layer could hardly be observed along the Mg/Sn interface. The absence of the reaction layer was attributed to the insufficient temperature for the chemical reaction and inter-diffusion between Mg and Sn atoms. Besides, the weld join-line still remained macroscopically flat. It could be seen from Fig. 8(b) that at the GTAW current of 35A, a reaction layer was distributed 9
ACCEPTED MANUSCRIPT continuously along the weld interface. The thickness of the reaction layer was approximately 2μm. As the GTAW current increased, the thickness of the reaction
IP
T
layer reached about 4μm for GTAW current of 40A and 8μm for GTAW current of
under the action of heavy plastic deformation.
SC R
45A. In addition, convoluted wave-like or swirl-like structure was observed obviously
EDS analysis has been performed to identify the reaction layer formed along the
NU
Mg/Sn interface of the joints and the results are shown in Fig. 9. It could be seen that
MA
both the Mg and Sn curves across the reaction layer presented a plateau, which indicated that the existence of Mg-Sn IMCs. The atomic ratio of Mg to Sn was
D
approximate 2:1 in reaction layer, suggesting the reaction layer was mainly consisted
TE
of Mg2Sn according to Mg-Sn binary phase diagram in Fig. 6(b). For further separate
CE P
verification of IMCs, XRD analysis is carried out on fracture surface of Mg side after the tensile lap shear tests for the joints produced at the GTAW current of 40A, and the
AC
results are shown in Fig. 10. It could be seen that the phases including Sn, Mg and Mg2Sn were detected, which confirmed that the correctness of above EDS line-scan analysis. Patel et al.[28] and Liu et al.[33] also reported the presence of Mg2Sn phase during Mg-Al dissimilar joining with Sn interlayer. 3.3 Vickers micro-hardness The micro-hardness profiles diagonally across the welded joints for GTAW currents of 35A, 40A and 45A are shown in Fig. 11(a). The insert was the schematic diagram of the micro-hardness profiles diagonally across the welded joints. The brown rectangle was the weld zone of the joints. And the weld zone was mainly 10
ACCEPTED MANUSCRIPT consisted of weld zone in Al side, Al/Sn interface, Sn interlayer, Mg/Sn interface and weld zone in Mg side. It could be indicated that the hardness of the joints increased
IP
T
with increasing GTAW current, where the trend was more obvious on the region of
SC R
welded Al and Mg. The hardness of the Al base metal and Mg base metal was 63-75HV and 62-76HV, respectively; while the hardness of the weld zone in Al side and weld zone in Mg side increased to 100-118HV and 95-115HV, respectively. That
NU
was attributed to the strain hardening in the vicinity of the weld. The Mg/Sn interface
MA
exhibited the maximum hardness values of 124-134HV due to the formation of Mg2Sn phase. Fig. 11(b) shows the micro-hardness profiles of the joints with and
D
without Sn interlayer at GTAW current of 40A. It could be seen that the addition of Sn
TE
interlayer could lead to a significant decrease in the hardness of the weld interface.
CE P
The hardness of Mg-Al IMCs (230-260HV) was much higher compared to the Mg2Sn phase (124-134HV). Therefore, the mechanical properties of the joints with Sn
AC
interlayer were improved.
3.5 Lap shear tensile strength Fig. 12(a) shows the lap shear strength of the joints with Sn interlayer as a function of GTAW current. It could be seen that the peak load of the joints increased with GTAW current until the maximum strength was reached and then decreased dramatically with higher GTAW current. The maximum peak load of the joints was 1.3kN at GTAW current of 45A, which was approximately 52% of Mg AZ31B sheets. At lower GTAW current, the temperature was not yet risen sufficiently to form a sound joint. As the GTAW current increased, the peak load increased in the beginning 11
ACCEPTED MANUSCRIPT owing to the high temperature, which accelerated the diffusion and plastic deformation at weld interface. However, the peak load decreased sharply at GTAW
IP
T
current of 50A due to the squeezing Sn interlayer out from the weld interface.
SC R
Comparison of the maximum lap shear strength of the joints with and without Sn interlayer is shown in Fig. 12(b). The joints with and without Sn interlayer were made with the same welding parameters. It could be concluded that the maximum peak load
NU
of the joints indicated improvements with the addition of Sn interlayer. In our
MA
previous research[23], the maximum peak load of the joints without Sn interlayer was 1.0kN. In this research, the maximum peak load of the joints with Sn interlayer was
D
approximately 1.3kN, which about 30% increase over the sample without Sn
TE
interlayer. The increase in the lap shear strength was attributed to the formation of
CE P
Sn-based solution and Mg2Sn phases, instead of the brittle Mg-Al IMCs. 3.6 Failure behavior
AC
Typical macroscopic photo of the failed lap shear tensile samples with Sn interlayer is shown in Fig. 13(a). All the joints failure occurred by the interface mode at the Al/Sn interface rather than the pull-out mode. The results indicated that the Al/Sn interface of the joints was the weakest zone. This was well in agreement with the previous microstructure evolution analysis in Section 3.2 and 3.4. There was no continuous metallurgical bonding along the Al/Sn interface due to almost no intersolubility between Al and Sn atoms. The formation of bonding along the interface was mainly due to the plastic deformation and mechanical interlocking. Meanwhile, Mg2Sn reaction layer was formed along the Mg/Sn interface of the joints. Therefore, 12
ACCEPTED MANUSCRIPT the micro-crack occurred firstly and then propagated along the Al/Sn interface during the lap shear tensile test.
IP
T
Fig. 13(b) and (c) show the SEM micrographs of fracture surface on the Al side
SC R
of the lap shear tensile failed samples at GTAW current of 35A and 45A, respectively. It could be seen that all the fracture patterns exhibited brittle fracture mode with cleavage facet feature surface. At lower GTAW current of 35A, the temperature was
NU
low and micro-bonded bright spots were observed at the periphery of sonotrode tip
MA
footprint. The bright spots were parallel to the direction of ultrasonic vibration. There were two kinds of regions present on the fracture surface, a bright white region and a
D
dark grey substrate. Quantitative analysis of the chemical composition (spot A to C)
TE
by EDS is listed in Table 2. Based on the EDS results, the bright white region
CE P
consisted of 100 at.% Sn, which suggested that this region mainly contained remains or debris Sn interlayer. The dark grey substrate was composed of 97 at.% Al and 3
AC
at.% Sn, which indicated that this region was Al substrate. As GTAW current increased to 45A, the temperature rose and the micro-bonds increased in density to spread across the faying surface. The EDS result of spot C also indicated that the inter-diffusion of Al and Sn atom was enhanced. 3.7 Weld defects analysis According to the above analysis, a sound ultrasonic welded Mg-Al joint with Sn interlayer could be obtained, if proper GTAW current was selected. As the GTAW current was below 30A, a substantial density of gaps or unbonded areas along the Al/Sn interface could be observed in Fig. 14(a). That was due to lower temperature 13
ACCEPTED MANUSCRIPT that enabled insufficient plastic flow the interface. Only part of the interface was effectively bonded. The results also indicated that the Al/Sn interface of the joints was
IP
T
the weakest zone. However, when the GTAW current reached 50A, the temperature
SC R
was higher than the melting point of Sn interlayer. The Sn interlayer melted and converted into liquid at the weld interface. The molten Sn interlayer was splashed out from the weld interface under the action of ultrasonic vibration. Therefore, the defect
NU
of voids was formed at the weld interface, as shown in Fig. 14(b).
MA
4. Discussion 4.1 Weld temperatures
D
To further understand the welding process and interface evolution, the
TE
temperatures were measured by using 0.5mm diameter K-type thermocouples placed
CE P
carefully at the center of the weld. The peak temperature results are shown in Fig. 15. The peak temperature without GTAW hybrid was 98°C. It could be indicated that the
AC
higher GTAW current resulted in a higher temperature at the center of the weld. The peak temperatures recorded by the thermocouples increased from 121°C for GTAW current of 30A, to 189°C for GTAW current of 40A, to 261°C at the maximum GTAW current of 50A. Patel et al.[21] revealed that the peak temperature was obviously sufficient for mutual diffusion of Al/Sn interface and Sn/Mg interface. Even constitutional liquation was occurred due to the lower melting point of Sn (232°C). 4.2 Formation process of the joint In our previous, GTAW was introduced into ultrasonic seam welding and solid state joining of 1mm thick Al and Mg alloys sheets was accomplished without 14
ACCEPTED MANUSCRIPT improving the ultrasonic power. However, the hard and brittle Al12Mg17 IMC was still formed at the weld interface, which preferentially acted as the source of micro-cracks
IP
T
to deteriorate the mechanical properties of the joint. Therefore, in this research, to
SC R
reduce the Al12Mg17 IMC in the joint, Sn interlayer was introduced into the weld. Based on above analysis, the formation process of the joint could be described as follow. The preceding GTAW could preheat the sheet metal to enhance the ultrasonic
NU
weldability. As static pressure and tangential force were applied by the rotating
MA
sonotrode, the faying surface came into close contact. The local surface oxide layer between contacting asperities began to break down and metallic adhesion took place.
D
The initial adhesion spots grew to form micro-bonds under the action of materials
TE
plastic deformation. The heat generated at the micro-bonds by the GTAW and friction
CE P
caused the plastic deformation resistance of the micro-bonds to decrease substantially, which was a favoring factor for the ultrasonic welding. Owing to the complex
AC
synergistic effects between the ultrasonic energy and GTAW heat, the temperature of the weld interface rose rapidly (Fig. 15), which resulted in significant inter-diffusion in the micro-bonds. When the inter-diffusion reached a critical value, the nucleation of the Mg2Sn IMC occurred. With increasing GTAW current, the Mg2Sn layer grew along the weld interface and became thicker until a continuous IMC layer was formed. Although the peak temperature was 166°C for GTAW current of 35A, the average thickness of Mg2Sn layer increased to 2μm. The results indicated that the growth rate of Mg2Sn layer in arc-assisted ultrasonic seam welding was much more rapid than that for static growth[18]. When the GTAW current increased to 40A and 45A, the 15
ACCEPTED MANUSCRIPT average thickness of Mg2Sn layer reached about 4μm and 8μm, respectively. In arc-assisted ultrasonic seam welding, under the ultrasonic vibration and GTAW heat,
IP
T
the strain rate and the accumulated stain were very high, even though the welding
SC R
process was short. Hence, severe plastic deformation mainly concentrated at the faying surface. Divinski et al.[34] indicated that during the process of severe plastic deformation, high densities of vacancies and dislocations were formed and a high
NU
density of lattice defects could dramatically enhance the diffusion rate. As the GTAW
MA
current reached 50A, the peak temperature was up to 261°C, higher than the melting point of Sn (232°C). Evidence of solidified microstructure was also observed at the
D
weld interface, as shown in Fig. 13(b). The Sn interlayer squeezed out from the weld
TE
interface and the weld defect of voids was formed due to the action of ultrasonic
CE P
vibration. The addition of Sn interlayer could restrain the extreme reaction between Al and Mg atoms. Hence, the formation of hard and brittle Mg-Al IMCs such as
AC
Al12Mg17 was suppressed. Therefore, the mechanical properties of the joints were improved obviously. 5. Conclusions
Arc-assisted ultrasonic seam welding of Mg and Al alloys was successfully completed using Sn interlayer. The following conclusions could be drawn: (1)The joint consisted of two distinct interface zones: Al/Sn interface and Mg/Sn interface. There was no continuous reaction layer along the Al/Sn interface due to almost no intersolubility between Al and Sn atoms. The bonding was mainly due to the plastic deformation and mechanical interlocking. Meanwhile, a continuous Mg2Sn 16
ACCEPTED MANUSCRIPT layer was formed along the Mg/Sn interface. As the GTAW current increased, the thickness of Mg2Sn layer increased gradually. Therefore, the formation of hard and
IP
T
brittle Mg-Al IMCs was suppressed.
SC R
(2)The maximum peak load of the joints with Sn interlayer was approximately 1.3kN, which about 30% increase over the joints without Sn interlayer. This improvement
brittle Mg-Al IMCs at the interface.
NU
was attributed to the formation of Sn-based soluti on and Mg2Sn phases instead of the
MA
(3)The Al/Sn interface was the weakest zone of the joint. All the joints failure occurred by the interface mode rather than the pull-out mode at the Al/Sn interface.
CE P
References
TE
feature surface.
D
All the fracture patterns exhibited entirely brittle fracture mode with cleavage facet
[1] K. T. Kashyap, S. Murali, K. S. Raman, K. S. S. Murthy, Casting and heat
AC
treatment variables of AI-7Si-Mg alloy, Mater. Sci. Technol., 9 (1993) 189-203. [2] F. Paray, B. Kulunk, J. Gruzleski, Effect of strontium on microstructure and properties of aluminium based extrusion alloy 6061, Mater. Sci. Technol., 12 (1996) 315-322. [3] L.M. Liu, D.X. Ren, F. Liu, A review of dissimilar welding techniques for magnesium alloys to aluminum alloys, Materials, 7 (2014) 3735-3757. [4] P. Penner, L. Liu, A. Gerlich, Y. Zhou, Feasibility study of resistance spot welding of dissimilar Al/Mg combinations with Ni based interlayers, Sci. Technol. Weld. Joi., 18 (2013) 541-550. 17
ACCEPTED MANUSCRIPT [5] H.Y. Wang, L.M. Liu, F. Liu, The characterization investigation of laser-arc-adhesive hybrid welding of Mg to Al joint using Ni interlayer, Mater. Des.,
IP
T
50 (2013) 463-466.
and Al alloy, Mater. Des., 32 (2011) 3730-3735.
SC R
[6] L.M. Liu, D.X. Ren, A novel weld-bonding hybrid process for joining Mg alloy
[7] X.D. Qi, L.M. Liu, Fusion welding of Fe-added lap joints between AZ31B
NU
magnesium alloy and 6061 aluminum alloy by hybrid laser-tungsten inert gas welding
MA
technique, Mater. Des., 33 (2012) 436-443.
[8] H.T. Zhang, X.Y. Dai, J.C. Feng, Joining of aluminum and magnesium via
D
pre-roll-assisted A-TIG welding with Zn interlayer, Mater. Lett., 122 (2014) 49-51.
TE
[9] H.T. Zhang, J.Q. Song, Microstructural evolution of aluminum/magnesium lap
3292-3294.
CE P
joints welded using MIG process with zinc foil as an interlayer, Mater. Lett., 65 (2011)
AC
[10] A. Dorbane, B. Mansoor, G. Ayoub, V.C. Shunmugasamy, A. Imad, Mechanical, microstructural and fracture properties of dissimilar welds produced by friction stir welding of AZ31B and Al6061, Mater. Sci. Eng. A, 651 (2016) 720-733. [11] H.M. Rao, B. Ghaffari, W. Yuan, J.B. Jordon, H. Badarinarayan, Effect of process parameters on microstructure and mechanical behaviors of friction stir linear welded aluminum to magnesium, Mater. Sci. Eng. A, 651 (2016) 27-36. [12] A. Heidarzadeh, T. Saeid, A comparative study of microstructure and mechanical properties between friction stir welded single and double phase brass alloys, Mater. Sci. Eng. A, 649 (2016) 349-358. 18
ACCEPTED MANUSCRIPT [13] J.A. Hamed, Microstructure and residual stress distributions in friction stir welding of dissimilar aluminium alloys, Mater. Des., 87 (2015) 405-413.
IP
T
[14] R.I. Rodriguez, J.B. Jordon, P.G. Allison, T. Rushing, L. Garcia, Microstructure
aluminum alloys, Mater. Des., 83 (2015) 60-65.
SC R
and mechanical properties of dissimilar friction stir welding of 6061-to-7050
[15] C.Y. Zhang, D.L. Chen, A.A. Luo, Joining 5754 automotive aluminum alloy
NU
2-mm-thick sheets using ultrasonic spot welding, Weld. J., 5 (2014) 131-138.
MA
[16] A. Panteli, J.D. Robson, I. Brough, P.B. Prangnell, The effect of high strain rate deformation on intermetallic reaction during ultrasonic welding aluminium to
D
magnesium, Mater. Sci. Eng. A, 556 (2012) 31-42.
TE
[17] M. Shakil, N.H. Tariq, M. Ahmad, M.A. Choudhary, J.I. Akhter, Effect of
CE P
ultrasonic welding parameters on microstructure and mechanical properties of dissimilar joints, Mater. Des. 55 (2014) 263-273.
AC
[18] D. Bakavos, P.B. Prangnell, Mechanisms of joint and microstructure formation in high power ultrasonic spot welding 6111 aluminium automotive sheet, Mater. Sci. Eng. A, 527 (2010) 6320-6334. [19] J.W. Yang, B. Cao, X.C. He, H.S. Luo, Microstructure evolution and mechanical properties of Cu–Al joints by ultrasonic welding, Sci. Technol. Weld. Joi. 19 (2014) 500-504. [20] R. Balasundaram, V.K. Patel, S.D. Bhole, D.L. Chen, Effect of zinc interlayer on ultrasonic spot welded aluminum-to-copper joints, Mater. Sci. Eng. A, 607 (2014) 277-286. 19
ACCEPTED MANUSCRIPT [21] V.K. Patel, S.D. Bhole, D.L. Chen, Microstructure and mechanical properties of dissimilar welded Mg–Al joints by ultrasonic spot welding technique, Sci. Technol.
IP
T
Weld. Joi. 17 (2012) 202-206.
SC R
[22] J.W. Yang, B. Cao, Investigation of resistance heat assisted ultrasonic welding of 6061 aluminum alloys to pure copper, Mater. Des. 74 (2015) 19-24. [23] X. Y. Dai, H. T. Zhang, J. H. Liu, J. C. Feng, Microstructure and properties of
MA
welding, Mater. Des., 77 (2015) 65-71.
NU
Mg/Al joint welded by gas tungsten arc welding-assisted hybrid ultrasonic seam
[24] F. Liu, Z.D. Zhang, L.M. Liu, Microstructure evolution of Al/Mg butt joints
D
welded by gas tungsten arc with Zn filler metal, Mater. Charact., 69 (2010) 84-89.
TE
[25] L.M. Zhao, Z.D. Zhang, Effect of Zn alloy interlayer on interface microstructure
CE P
and strength of diffusion-bonded Mg-Al joints, Scripta Mater. 58 (2008) 283-286. [26] L.M. Liu, F. Liu, M.L. Zhu, Study on Mg/Al weld seam based on Zn-Mg-Al
AC
ternary alloy, Materials, 7 (2014) 1173-1187. [27] X. J. Liu, R. S. Huang, H. Y. Wang, S. H. Liu, Improvement of TIG lap weldability of dissimilar metals of Al and Mg, Sci. Technol. Weld. Joi. 12 (2007) 258-260. [28] V.K. Patel, S.D. Bhole, D.L. Chen, Improving weld strength of magnesium to aluminum dissimilar joints via tin interlayer during ultrasonic spot welding, Sci. Technol. Weld. Joi. 17 (2012) 342-347. [29] Pan JL. Welding handbook. Beijing: China Machine Press; 1995. [30] V.K. Patel, S.D. Bhole, D.L. Chen, Characterization of ultrasonic spot welded 20
ACCEPTED MANUSCRIPT joints of Mg-to-galvanizedand ungalvanized steel with a tin interlayer, J. Mater. Process Technol., 214 (2014) 811-817.
IP
T
[31] C. Doumanidis, Y. Gao, Mechanical modeling of ultrasonic welding, Weld. J. 83
SC R
(2004) 140-146.
[32] P. Villars, A. Prince, Handbook of Binary Alloy Phase Diagrams, ASM International, Metals Park, 1995.
NU
[33] Z. Wang, H.Y. Wang, L.M. Liu, Study on low temperature brazing of
MA
magnesium alloy to aluminum alloy using Sn-xZn solders, Mater. Des. 39 (2012) 14-19.
D
[34] S.V. Divinski, G. Reglitz, H. Rösner, Y. Estrin, G. Wilde, Ultra-fast diffusion
TE
channels in pure Ni severely deformed by equal-channel angular pressing, Acta
AC
CE P
Mater., 59 (2011) 1974–1985.
21
ACCEPTED MANUSCRIPT Figures Captions Fig. 1 Schematic diagram of (a)arc-assisted hybrid ultrasonic seam welding system;
IP
T
(b)thermal couple position for temperature measurement
SC R
Fig. 2 Weld appearance of the joint produced at GTAW current of 35A: (a)front; (b)back
Fig. 3 Typical SEM micrographs of cross-section of the joints at GTAW current of
NU
(a)30A; (b)35A; (c)40A; (d)45A
MA
Fig. 4 SEM micrographs of Al/Sn interface at GTAW current of (a)30A; (b)35A; (c)40A; (d)45A
D
Fig. 5 EDS line scan results of the line indicated in Fig.3 (c)
TE
Fig. 6 The binary phase diagram: (a)Al-Sn phase diagram; (b)Mg-Sn phase
CE P
diagram[32]
Fig. 7 X-ray diffraction of fracture surface on Al side at GTAW current of 40A
AC
Fig. 8 SEM micrographs of Mg/Sn interface at GTAW current of (a)30A; (b)35A; (c)40A; (d)45A
Fig. 9 EDS line scan results of the line indicated in Fig.7 (c) Fig. 10 X-ray diffraction of fracture surface on Mg side at GTAW current of 40A Fig. 11 Micro-hardness profile obtained across the cross-section of the following samples: (a) at different GTAW currents; (b) at GTAW current of 40A Fig. 12 (a)Lap shear strength of the joints with Sn interlayer as a function of GTAW current; (b)comparison of the maximum lap shear strength of the joints with and without Sn interlayer 22
ACCEPTED MANUSCRIPT Fig. 13 Typical fractured welded joints: (a)macroscopic images of failed lap shear tensile samples; SEM images of fracture surface on Al side at GTAW of (b)35A,
SC R
Fig. 14 Weld defects: (a)unbonded weld; (b)voids
IP
T
(c)45A and EDS box analysis in the magnified fracture surface area.
Fig. 15 Peak temperature as a function of GTAW currents
Table 1 Optimized welding parameters
NU
Tables Captions
AC
CE P
TE
D
MA
Table 2 Chemical compositions of regions indicated in Fig. 13 (at.%)
23
SC R
IP
T
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
Fig. 1
24
AC
Fig. 2
CE P
TE
D
MA
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
25
AC
Fig. 3
CE P
TE
D
MA
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
26
AC
Fig. 4
CE P
TE
D
MA
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
27
TE
D
MA
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
AC
CE P
Fig. 5
28
SC R
IP
T
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
Fig. 6
29
D
MA
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
AC
CE P
TE
Fig. 7
30
AC
Fig. 8
CE P
TE
D
MA
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
31
TE
D
MA
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
AC
CE P
Fig. 9
32
TE
D
MA
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
AC
CE P
Fig. 10
33
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
Fig. 11
34
SC R
IP
T
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
Fig. 12
35
AC
CE P
TE
D
MA
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
Fig. 13
36
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
Fig. 14
37
TE
D
MA
NU
SC R
IP
T
ACCEPTED MANUSCRIPT
AC
CE P
Fig. 15
38
ACCEPTED MANUSCRIPT Table 1 Optimized welding parameters Vibration frequency(kHz) 20 20 20 20 20
CE P
TE
D
MA
NU
SC R
IP
T
Amplitude (μm) 20 20 20 20 20
AC
S1 S2 S3 S4 S5
GTAW GTAW Welding Pressure current(A) voltage(V) speed(mm/s) (MPa) 30 9.5 8 0.44 35 9.5 8 0.44 40 9.5 8 0.44 45 9.5 8 0.44 50 9.5 8 0.44
39
ACCEPTED MANUSCRIPT Table 2 Chemical compositions of regions indicated in Fig. 12 (at.%) Mg 5
Sn 100 3 65
Possible phases Sn Al Sn, Al
T
Al 97 30
AC
CE P
TE
D
MA
NU
SC R
IP
Spot A B C
40
ACCEPTED MANUSCRIPT
AC
CE P
TE
D
MA
NU
SC R
IP
T
Graphical abstract
41
ACCEPTED MANUSCRIPT Highlights 1. Mg and Al alloys were successfully joined by arc-assisted ultrasonic seam welding.
IP
T
2. Sn interlayer was used to restrain the formation of brittle Mg-Al IMCs.
SC R
3. Effects of GTAW current on microstructure and mechanical properties were studies. 4. The maximum peak load with Sn interlayer was 1.3kN, which was about 30%
AC
CE P
TE
D
MA
NU
increase.
42