Dissimilar joining of aluminum alloy and stainless steel thin sheets by thermally assisted plastic deformation

Journal of Materials Processing Technology 225 (2015) 393–404

Contents lists available at ScienceDirect

Journal of Materials Processing Technology journal homepage: www.elsevier.com/locate/jmatprotec

Dissimilar joining of aluminum alloy and stainless steel thin sheets by thermally assisted plastic deformation Zhequn Huang a,∗ , Jun Yanagimoto b a b

Graduate School of Engineering, The University of Tokyo, Komaba 4-6-1, Meguro, Tokyo 153-8505, Japan Institute of Industrial Science, The University of Tokyo, Komaba 4-6-1, Meguro, Tokyo 153-8505, Japan

a r t i c l e

i n f o

Article history: Received 22 December 2014 Received in revised form 18 June 2015 Accepted 20 June 2015 Available online 24 June 2015 Keywords: Aluminum–steel dissimilar joining Plastic deformation Mechanical anchoring Diffusion bonding Interdiffusion layer

a b s t r a c t Aluminum alloy/steel hybrid components are widely used in different industrial areas because of their high performance. However, the high importance of reducing the thickness of components to realize lightweight products requires the dissimilar joining of Al alloy and steel thin sheets (less than 1 mm in thickness), which is a major challenge using current joining technologies. In this paper, an alternative dissimilar joining process by thermally assisted plastic deformation is proposed for thin metallic sheets. The effects of various parameters on the joining performance were investigated. After exposure to an elevated temperature of 450 ◦ C for 22 s, an optimized joint type was achieved by local plastic deformation using a punch-die pair. This joint type exhibited an average joint efficiency factor of 85.2%, and an average absorption energy of 1.69 kN mm in tensile shear tests, as well as satisfactory joining performance in peel tests. In addition to mechanical anchoring and surface enlargement, atomic interdiffusion at the interface activated by the elevated forming temperature was found to be critical for obtaining high joining quality. The thickness of the Fe-Al interdiffusion layer at the interface was positively correlated with the heat input and the locally distributed plastic strain. This study shows that the proposed dissimilar joining process for Al alloy and steel thin sheets is a promising joining method for Al alloy/steel lightweight structures owing to the excellent joining performance, weight effectiveness, simple operation and long tool lifetime. © 2015 Elsevier B.V. All rights reserved.

1. Introduction The emerging requirements of higher structural performance with lower weight and energy consumption in the aerospace, automobile, electrical, and chemical industries are stimulating the rapid development of lightweight structures. Steel alloy has good creep resistance and formability as well as relatively high strength, while Al alloy possesses low density and excellent corrosion resistance. In recent years, Al-steel hybrid structures have been widely used in lightweight products. Although dissimilar joining between steel and Al alloy has been investigated intensively, achieving satisfactory joining quality remains a technological challenge owing to the distinct differences between the two dissimilar materials, such as thermal expansion coefficient, melting temperature, mechanical properties, etc., as well as the ease of formation of brittle inter-

∗ Corresponding author. Fax: +81 354526204. E-mail addresses: [email protected], [email protected] (Z. Huang). http://dx.doi.org/10.1016/j.jmatprotec.2015.06.023 0924-0136/© 2015 Elsevier B.V. All rights reserved.

metallic compounds (IMCs) at the interface during the welding process. Various thermal welding processes have been proposed for joining Al alloy to steel, such as resistance spot welding, laser welding, laser braze welding and laser rolling welding, which were respectively utilized by Connolly (2007), Torkamany et al. (2010), Sierra et al. (2008) and Ozaki and Kutsuna (2012). However, the existence of microvoids and thick brittle IMCs could not be avoided, which would degrade both the static and fatigue joining performance. The recently developed friction stir welding (FSW) process is another joining process which involves plastic deformation. As reviewed by Mishra and Ma (2005), during FSW process, heat is generated by friction between a rotating tool and the workpieces. The softened and severely plastic-deformed base material is transported from the tool front to the trailing edge and then quickly forged into a joint. FSW has a smaller heat-affected zone and better joining performance than other welding technologies and has been regarded as a promising joining approach for dissimilar sheets. An improved hybrid welding process combining gas tungsten arc welding and FSW was proposed by Bang et al. (2012), which was used to join 3-mm-thick Al alloy to 3-mm-thick SUS304 sheets with a high

394

Z. Huang, J. Yanagimoto / Journal of Materials Processing Technology 225 (2015) 393–404

Table 1 Chemical composition of 0.8-mm-thick A6061P-T6 sheet (wt%). Element

Mg

Si

Fe

Cu

Cr

Zn

Mn

Ti

Al

% wt

0.5–1.2

0.4–0.8

0.7 Max

0.15–0.4

0.04–0.35

0.25 Max

0.15 Max

0.15 Max

Bal.

Table 2 Chemical composition of 0.5-mm-thick SUS304 sheet (wt%). Element

Cr

Ni

Mn

Si

C

P

S

Mo

Fe

% wt.

18.0–20.0

8.0–10.5

2.0 Max

1.0 Max

0.08 Max

0.045 Max

0.03 Max

–

Bal.

joint efficiency factor of approximately 93%. However, the penetration and rotation of the pin in the base materials would result in microvoids along the interface and a short tool lifetime. Additionally, micrometer-thick brittle IMCs were also easily formed at the bonding interface during the direct intermixing, which would deteriorate the fatigue joining strength according to Briskham et al. (2006). More importantly, as far as ultrathin metallic sheets are concerned, dissimilar joining by FSW is a major challenge as the penetration of the rotating tool would easily induce severe defects in thin sheets. Conventional mechanical fastening is the main method used to assemble Al alloy and steel sheets. However, it usually induces stress concentration at the connection, decreases the corrosion resistance and requires a predrilled hole and an additional component such as a bolt, screw or rivet. Self-piercing riveting (SPR) and mechanical clinching, which are two typical joining processes by plastic deformation, have been attracting increasing interest in recent years, as reviewed by Mori et al. (2013) and Groche et al. (2014). An SPR joint is achieved through directly piercing sheets with a rivet at room temperature without the need for a predrilled hole. The SPR process has been utilized by Abe et al. (2009) to join Al alloy to steel sheets. However, its applicability is limited to thicker sheets because rupture and fracture are difficult to avoid in the riveting of ultrathin sheets (e.g., 0.5 mm in thickness). Mechanical clinching is a low-noise cold joining process in which the metallic sheets are locally deformed by a rigid punch-die pair without using an additional element or a predrilled hole. Its technical difficulties include forming interlock and avoiding fractures and cracks after deformation. Abe et al. (2012) investigated the applicability of the clinching process to Al alloy and high-strength steel and improved the joining performance by adapting the geometry of the tool and controlling the material flow. Nevertheless, the mechanical clinching process is less effective for very thin sheets owing to their poor ductility and formability at room temperature. Abe et al. (2012) also reported that when the thickness of the lower sheet was less than 1 mm, the crucial interlock could not be successfully formed, which was attributed to the insufficient material flow between the corners of the punch and the die. Therefore, it is necessary to develop more reliable and effective dissimilar joining methods for Al alloy and steel thin sheets (less than 1 mm in thickness) for special lightweight requirements. In this study, we propose a new joining process by thermally assisted plastic deformation that combines mechanical and diffusion bonding techniques for thin metallic sheets, taking 0.8-mm-thick Al alloy and 0.5-mm-thick steel sheets as examples. Different from the conventional clinching technology, the proposed joining process does

not aim at forming an interlock by severe plastic deformation owing to the limited formability of thin metallic sheets. The depth of the die is relatively low and the forming process is stopped at an intermediate stage of the conventional clinching process to avoid the occurrence of fractures and cracks in the joint. After plastic deformation, the two thin sheets are firmly pressed into the die cavity; thus, the relative movement of the sheets is effectively inhibited and the overall contact area is enlarged. In addition, induction heating is introduced to increase the ductility of the thin sheets, provide sufficient material flow and allow atomic interdiffusion between the dissimilar materials. In the present work, the effects of the experimental parameters and tool geometry on the static tensile shear strength and peel strength of a single-lap joint were investigated. The failure modes of different joint types, the main joining mechanisms and the dependence of the joining performance on the forming temperature and plastic strain were also discussed.

2. Experimental procedure 2.1. Materials and experimental setup 0.8-mm-thick precipitate-hardened Al alloy (A6061P-T6) and 0.5-mm-thick stainless steel (SUS304) sheets were used to fabricate dissimilar joints in this study. The chemical compositions of these two metallic sheets are given in Tables 1 and 2, respectively, and their mechanical properties are presented in Table 3. The sheets were cut into pieces of size of 26 mm (W) × 32 mm (L). Neither type of metallic sheet was subjected to any surface treatment. The A6061P-T6 sheet was placed above the SUS304 sheet with an overlapping area of 26 mm (W) × 26 mm (L). The dissimilar joining process by thermally assisted plastic deformation was conducted in air using a 5 t high-precision compression machine. A simplified schematic of the experimental setup for the joining process is shown in Fig. 1(a). A spring was utilized to provide the blank holding force. In the cases of dissimilar joining at elevated temperatures, the specimens were heated by an induction coil unit to target temperatures of 150, 300, 350, 400, and 450 ◦ C at the same heating rate of 15 ◦ C/s using a thermocouplebased sensor-feedback system. After the target temperature was reached and maintained for approximately 10 s, the punch moved downward with a stroke of less than 2.0 mm. The punch was then pulled out of the connection, and the specimen was cooled in air. The heating history and the forming procedure are presented in Fig. 1(b).

Table 3 Mechanical properties of the used metallic sheets. Sheet

Thickness (mm)

Density (g/cm3 )

Young’s modulus (GPa)

Tensile strength (MPa)

Yield strength (MPa)

Elongation (%)

Poisson’s ratio

A6061P-T6 SUS304

0.8 0.5

2.7 8.0

68.9 190

310 520

276 240

12 45

0.33 0.29

Z. Huang, J. Yanagimoto / Journal of Materials Processing Technology 225 (2015) 393–404

395

Fig. 1. (a) Schematic of experimental setup for the dissimilar joining process by thermally assisted plastic deformation. (b) experimental procedure for the joint conducted at an elevated temperature. The forming temperature is ‘T’, the heat holding time is ‘H’ and the forming stroke is ‘S’. A constant heat rate (15◦ C/s) and air cooling were applied for all the cases of joining conducted at elevated temperatures.

Fig. 2. Punch-die pair used in the dissimilar joining process by thermally assisted plastic deformation: (a) dimensions of the punch; (b) dimensions of the outer die; (c) dimensions of the inner die (the diameter of the outer circle of the groove was fixed at 9.6 mm, while the diameter of the inner circle of the groove (ØD) was variable); (d) three groove widths (W1.2, W1.4 and W1.6) were used to study the effect of the geometric parameters of the tool; thus, the ØD values in (c) were 7.2, 6.8 and 6.4 mm, accordingly.

The punch and die with their detailed dimensions are depicted in Fig. 2. The groove in the die increases the overall contact area. In addition, it can accelerate the material flow, as presented by Abe et al. (2012). To investigate the effect of geometrical parameters of the tool, three dies with different groove widths (W = 1.6 mm/1.4 mm/1.2 mm) as well as two punches with different punch corner radii (r(chamfer) = 0.2 mm/0.5 mm) were used. The

modulation of the groove width was realized using a deployable inner die, as shown in Fig. 2(c). In addition to the geometric parameters of the tool, the forming stroke and temperature as well as the heat holding time may also affect the joining performance of the dissimilar joint. To optimize the joining quality, a series of joint types were fabricated with different punch-die combinations and different experimental parameters, as listed in Table 4, where

Table 4 Description of different A6061P-SUS304 joint types fabricated for parametric optimization. Code of joint type

Radius of punch corner: r(chamfer) (mm)

Groove width of die: W (mm)

Stroke: S (mm)

Temperature: T (◦ C)

Heat holding time: H (s)

W1.6–S1.7–RT W1.6–S1.7–T150–H22 W1.6–S1.7–T300–H22 W1.6–S1.7–T350–H22 W1.6–S1.7–T400–H22 W1.6–S1.7–T450–H22 W1.6–S1.7–T450–H28 W1.6–S1.8–T300–H22 W1.6–S1.8–T350–H22 W1.4–S1.7–T400–H22 W1.2–S1.7–T400–H22

0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.5 0.5

1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.6 1.4 1.2

1.7 1.7 1.7 1.7 1.7 1.7 1.7 1.8 1.8 1.7 1.7

RT 150 300 350 400 450 450 300 350 400 400

0 22 22 22 22 22 28 22 22 22 22

396

Z. Huang, J. Yanagimoto / Journal of Materials Processing Technology 225 (2015) 393–404

Fig. 3. (a) Schematic of two primary parameters W and S; (b) top view of eight typical dissimilar joints obtained by the proposed dissimilar joining method: W1.6–S1.7–RT, W1.6–S1.7–T150–H22, W1.6–S1.7–T300–H22, W1.6–S1.7–T350–H22, W1.6–S1.7–T400–H22, W1.6–S1.7–T450–H22, W1.6–S1.8–T300–H22 and W1.6–S1.8–T350–H22.

‘W’, ‘S’, ‘T’ and ‘H’ denote the groove width in the die, the forming stroke, the forming temperature and the heat holding time, respectively.

2.2. Evaluation tests and characterization methods An optical microscope (Keyence VHX-S15C) was used to observe the cross-sectional shapes of the fabricated joints. The tensile shear strength and peel force of each joint type were evaluated using a Shimadzu AG-50kNG Instron test machine at room temperature. The crosshead speed in the tensile shear test was 1 mm/min and the gauge length was 28 mm. The peel test was performed by applying separation loading perpendicular to the joint with a traverse speed of 5 mm/min. The Vickers hardness values of typical joints were measured using a Shimadzu-HMV digital microhardness tester with 100 g load and 10 s dwell time. To study the joining mechanism, the A6061P-T6/SUS304 interface and failure surface of a typical joint were observed after the tensile shear test using a field-emission scanning electron microscope (FE-SEM: JSM-7100F) equipped with energy-dispersive X-ray spectroscopy (EDS), and TSL-OIM electron backscatter diffraction (EBSD) systems. Kernel average misorientation (KAM) mapping was utilized to analyze the plastic strain distribution on the base SUS304 material across the joint, and thereby indirectly evaluate the distribution of normal pressure at the joining interface.

3. Results 3.1. Evaluation of joining quality by cross-sectional observation In this study, to avoid excessive thinning of the upper sheet and the occurrence of cracks and fractures in the joint, the forming process was stopped at an intermediate stage of the conventional clinching process. No interlock was formed for all the fabricated joint types listed in Table 4. A schematic indicating the two primary parameters (W and S) is shown in Fig. 3(a). The appearances of the eight typical joints under different experimental conditions (W1.6–S1.7–RT, W1.6–S1.7–T150–H22, W1.6–S1.7–T300–H22, W1.6–S1.7–T350–H22, W1.6–S1.7–T400–H22, W1.6–S1.8–T300–H22, and W1.6–S1.7–T450–H22, W1.6–S1.8–T350–H22) are shown in Fig. 3(b). For the joint types fabricated at temperatures lower than 350 ◦ C, the dissimilar sheets separated immediately after forming. In contrast, the two dissimilar sheets bonded together when the forming temperature was increased to 350 ◦ C or higher. The cross-sectional shapes of the W1.6–S1.7–RT, W1.6–S1.7– T300–H22, W1.6–S1.8–T300–H22 and W1.6–S1.8–T350–H22 joints were further examined using an optical microscope, as shown in Fig. 4. Debonding between the Al alloy and stainless steel was observed in the W1.6–S1.7–RT and W1.6–S1.7–T300–H22 joints, as shown in Fig. 4(a) and (b). Moreover, brittle fracture occurred in the neck region of the W1.6-S1.7-RT joint, owing to

Fig. 4. Cross-sectional shapes of the joints with defects: (a) W1.6–S1.7–RT joint; (b) W1.6–S1.7–T300–H22 joint; (c) W1.6–S1.8–T300–H22 joint; (d) W1.6–S1.8–T350–H22 joint.

Z. Huang, J. Yanagimoto / Journal of Materials Processing Technology 225 (2015) 393–404

397

Fig. 5. (a–d) Cross-sectional shapes of the successfully bonded joints (W1.6–1.7–T350-H22, W1.6–S1.7–T400–H22, W1.6–S1.7–T450–H22 and W1.6–S1.7–T450–H28, respectively); (b), (e), and (f) cross-sectional shapes of the bonded joints using three different punch-die combinations (W1.6–S1.7–T400–H22, W1.4-S1.7–T400–H22 and W1.2–S1.7–T400–H22, respectively).  a ,  b , and  c are the maximum slope angles in the neck regions of the three joints, which are 78.7, 50.2, and 59.5◦ , respectively.

the poor formability of the thin Al alloy sheet at room temperature. When the forming temperature was increased to 300 ◦ C, the fracture in the upper Al alloy sheet was avoided as the elevated temperature effectively improved the ductility of the Al alloy sheet. However, when the forming stroke was increased from 1.7 to 1.8 mm, even at elevated temperatures of 300 and 350 ◦ C, some defects such as fractures and cracks were observed in the W1.6–S1.8–T300–H22 and W1.6–S1.8–T350–H22 joints, as shown in Fig. 4(c) and (d), respectively. It appeared that the large forming stroke (1.8 mm) was beyond the forming limit of the thin metallic sheets. cross-sectional shapes of the The W1.6–S1.7–T350–H22, W1.6–S1.7–T400–H22, W1.6–S1.7–T 450–H22 and W1.6–S1.7–T450–H28 joints are shown in Fig. 5(a–d). The thicknesses of the Al alloy in the groove regions of these four joints were 0.94, 1.00, 1.03 and 1.03 mm, respectively, which indicated that the increase in the forming temperature effectively assisted the plastic deformation of both the Al alloy and stainless steel thin sheets. All the four joints were bonded together without any visible gaps, fractures or cracks. Nevertheless, as shown in Fig. 5 (d), distortion of the Al alloy sheet clearly occurred in the upper part of the neck region in the W1.6–S1.7–T450–H28 joint. The Al alloy in this joint surrounding the punch was backwardextruded outside during the forming process and slightly dragged out by the stripping force during the removal of the punch, leading to the occurrence of distortion. At the forming temperature of 450 ◦ C, it was observed that the striping process became difficult when the heat holding time was increased to 28 s. This may be attributed to the increased sticking coefficient of the Al alloy. Flitta and Sheppard (2003) reported that during aluminum extrusion at elevated temperatures ranging from 300 to 450 ◦ C, the friction coefficient changed from sliding friction to nearly sticking friction. Moreover, Schikorra et al. (2007) observed almost perfect sticking between an AA6060 billet and extrusion containers after heating to 450 ◦ C for approximately 100 s. Therefore, to prevent distortion in the joint, excessive heat input should be avoided in this joining process. The geometrical parameters of the punch-die pair should also be considered. Three punch-die combinations were used to fabricate different joint types, with the forming stroke, forming temperature and heat holding time fixed at 1.7 mm, 400 ◦ C and 22 s, respectively. The cross-sectional shapes of the obtained W1.6–S1.7–T400–H22, W1.4–S1.7–T400–H22 and W1.2–S1.7–T400–H22 joints are shown

in Fig. 5 (b),(e) and (f), respectively. It was found that excessive thinning of the upper sheet and crack formation in the neck region of the joint were effectively suppressed with a larger punch corner radius. However, the smaller capacity of the groove and the reduced compression stress from the punch with a corner radius of 0.5 mm resulted in insufficient filling of the upper Al alloy sheet into the die, larger springback of the lower SUS304 sheet and a gap in the shoulder region of the joint, reducing the overall contact area between the upper and lower sheets. Moreover, the maximum slope angle in the neck region () also changed with the groove width and the punch corner radius. For the W1.6–S1.7–T400–H22, W1.4–S1.7–T400–H22 and W1.2–S1.7–T400–H22 joints, the corresponding  a ,  b and  c are 78.7, 50.2, and 59.5◦ , respectively. The reduced slope angle in the neck region is expected to decrease the slip resistance of the joint to the separation load during the tensile shear test. According to Figs. 4 and 5, it can be concluded that the crosssectional quality of the dissimilar joint strongly depends on the forming stroke (S), the groove capacity in the die being related to the groove width (W), as well as the supplied thermal energy which is related to both the forming temperature (T) and the heat holding time (H). A schematic showing the dependence of the deformation in the joint on these critical process parameters is presented in Fig. 6, which is expected to serve as a design guide for this joining process. “1”, “0” and “−1” denote excessive, appropriate, and insufficient values, respectively. 3.2. Evaluation of joining quality by static tensile shear test A static tensile shear test at room temperature was first utilized to evaluate the effects of the forming temperature, the heat holding time and the geometry of the tool on the joining performance of the joints. The results of the tensile shear tests for the W1.6–S1.7–T350–H22, W1.6–S1.7–T400–H22, W1.6–S1.7–T450–H22 and W1.6–S1.7–T450–H28 joint types are shown in Fig. 7(a). The ultimate tensile shear load of the W1.6–S1.7–T350–H22 joint was much smaller than those of the joint types fabricated at 400 and 450 ◦ C. After reaching the peak tensile shear load, the upper and lower sheets in the W1.6–S1.7–T350–H22 joint type separated rapidly with little plastic deformation on the Al alloy side. Both the W1.6–S1.7–T400–H22 and W1.6–S1.7–T450–H22 joint types exhibited ultimate tensile shear load of almost 2.0 kN, and the latter showed bet-

398

Z. Huang, J. Yanagimoto / Journal of Materials Processing Technology 225 (2015) 393–404

Fig. 6. Schematic of the dependence of the deformation in the joint on the critical process parameters including the forming stroke, the groove capacity in the die and the supplied thermal energy.

3.0

(b)

W1.6-S1.7-T350-H22 W1.6-S1.7-T400-H22 W1.6-S1.7-T450-H22 W1.6-S1.7-T450-H28

2.5

Tensile shear load (kN)

Tensile shear load (kN)

(a)

2.0 1.5 1.0 0.5 0.0

3.0

W1.6-S1.7-T400-H22 W1.4-S1.7-T400-H22 W1.2-S1.7-T400-H22

2.5 2.0 1.5 1.0 0.5 0.0

0.0

0.5

1.0

1.5

2.0

2.5

Displacement (mm)

0.0

0.5

1.0

1.5

2.0

2.5

Displacement (mm)

Fig. 7. (a) Tensile shear load/displacement curves of the W1.6–S1.7–T350–H22, W1.6–S1.7–T400–H22, W1.6–S1.7–T450–H22 and W1.6–S1.7–T450–H28 joint types; (b) tensile shear load/displacement curves of the W1.6–S1.7–T4000–H22, W1.4–S1.7–T400–H22, and W1.2–S1.7–T400–H22 joint types.

ter joining stability. For both the W1.6–S1.7–T400–H22, and W1.6–S1.7–T450–H22 joint types, the tensile shear test curves gradually descended after reaching the ultimate tensile shear load, indicating the tearing of the Al alloy around the neck region of the joints. Nevertheless, further extending the heat holding time from 22 to 28 s at 450 ◦ C severely softened the Al alloy in the neck region of the W1.6–S1.7–T450–H28 joint, resulting in the remarkable degradation of the mechanical properties of the Al alloy and thus the ultimate tensile shear load of the joint. To investigate the effect of the groove width and the punch corner radius on the joining performance, tensile shear tests were also performed on the W1.4–S1.7–T400–H22, and W1.2–S1.7–T400–H22 joint types, as shown in Fig. 7(b). The results clearly showed that the ultimate tensile shear loads of the W1.4–S1.7–I400–H22, and W1.2–S1.7–T400–H22 joint types were smaller than that of the W1.6–S1.7–T400–H22 joint type. This experimentally confirmed that the imperfect material flow in the

W1.4–S1.7–T400–H22 and W1.2–S1.7–T400–H22 joints was detrimental to their joining performance. The effects of the forming temperature and the maximum slope angle in the neck region () on the ratio of the tensile strength of the joint to the base A6061P-T6 and the tensile shear absorption energy are summarized in Fig. 8. The W1.6–S1.7–T450–H22 joint type was the optimal one with an average ratio of the tensile strength to the base A6061P-T6 of as high as 85.2% and an average tensile shear absorption energy of 1.69 kN mm. Such high and stable joining quality was achieved for the 0.8-mm-thick Al alloy and 0.5mm-thick steel sheets by a simple die-forging process in air without employing any shielding gas, flux brazing or surface pretreatment. Compared with friction stir spot welding, the proposed joining process by thermally assisted plastic deformation has marked benefits in terms of its flexible applicability to ultrathin sheets, repeatability of joining quality, dwell time to separation load in the tensile shear test, operation simplicity and tool lifetime.

Z. Huang, J. Yanagimoto / Journal of Materials Processing Technology 225 (2015) 393–404

(b)

Groove width W=1.6 mm, θ=78.7⁰

Forming temperature T=400 ⁰C Ratio of tensile strength of joint to the base A6061P-T6

3.0 (450, 85.2)

80

2.5

(400, 81.8)

2.0 60

(450, 1.69)

1.5 (350, 42.9)

40

1.0

(400, 0.95)

20

0.5 (350, 0.18)

0.0

0 300

350

400

450

500

Forming temperature T (⁰⁰C)

Tensile shear absorption energy

Ratio of tensile strength of joint to the base A6061P-T6 (%)

Ratio of tensile strength of joint to the base A6061P-T6 (%)

Tensile shear absorption energy

100

Tensile shear absorption energy (kN·mm)

Ratio of tensile strength of joint to the base A6061P-T6

3.0

100

2.5

80 (78.7, 81.8)

(59.5, 61.6)

60

2.0

(50.2, 52.0)

1.5 40

(50.2, 0.90)

(78.7, 0.95) 1.0 (59.5, 1.09)

20

0.5 0.0

0 40

50

60

70

80

90

Tensile shear absorption energy ((kN·mm)

(a)

399

Slope angle in the neck of joint θ (⁰⁰)

Fig. 8. (a) Variations in the ratio of the tensile strength of the joint to the base A6061P-T6 and the tensile shear absorption energy of the joint with the forming temperature; (b) variations in the ratio of the tensile strength of the joint to the base A6061P-T6 and the tensile shear absorption energy of the joint with the slope angle in the neck region of the joint.

Fig. 9. (a–e) Failure surfaces of the representative W1.6–S1.7–T350–H22, W1.6–S1.7–T400–H22, W1.6–S1.7–T450–H22, and W1.6–S1.7–T450–H28 joints after the tensile shear test; (f) schematics of the failure modes after the tensile shear test.

The failure surfaces of the W1.6–S1.7–T350–H22, W1.6–S1.7– T400–H22, W1.6–S1.7–T450–H22, and W1.6–S1.7–T450–H28 joint types after the tensile shear test are shown in Fig. 9(a)–(e), respectively. The failure modes are summarized in Fig. 9(f-1)–(f-3). At a lower forming temperature (350 ◦ C), the failure mode of the W1.6–S1.7–T350–H22 joint was ductile button pull-out, as shown in Fig. 9(a) and (f-1). When the forming temperature was increased

to 400 ◦ C, the W1.6–S1.7–T4000–H22 joints exhibited two failure modes (modes A and B), as shown in Fig. 9(b) and (c), respectively. Schematics of these two failure modes are depicted in Fig. 9(f2). The majority of the W1.6–S1.7–T400–H22 joints exhibited the button pull-out and neck cracking co-existent failure mode, while a small number of the W1.6–S1.7–I400–H22 joints showed the ductile button pull-out mode with a certain amount of plastic defor-

400

Z. Huang, J. Yanagimoto / Journal of Materials Processing Technology 225 (2015) 393–404

W1.6–S1.7–T450–H22 joint types were considerably superior to the W1.6–S1.7–T350–H22 joint type in terms of the ultimate peel force and displacement. In addition, after reaching the peak value, the peel test curves of the W1.6–S1.7–T400–H22 joint type exhibited gradual decreases, while the curves of the W1.6–S1.7–T450–H22 joint type rapidly declined. According to the results of the peel tests, the W1.6–S1.7–T400–H22 joint type absorbed more failure energy than the W1.6–S1.7–T450–H22 joint type before the total separation of the dissimilar components. 4. Discussion 4.1. Mechanisms of joining by thermally assisted plastic deformation

Fig. 10. (a) Image of a T-shaped specimen prepared for the peel test; (b) image of a typical W1.6–S1.7–T4000–H22 joint during the peel test; (c) failure surfaces of the failed W1.6–S1.7–T350–H22, W1.6–S1.7–T400–H22 and W1.6–S1.7–T450–H22 joints after the peel tests.

mation in the upper Al alloy sheet. The pulled out part of the Al alloy sheet around the neck region and the warpage of the failed joint shown in Fig. 9(b) resulted from the bending moment created during the tensile shear test. The deformation in the Al alloy shown in Fig. 9(c) indicated that the W1.6–S1.7–T400–H22 joint with failure mode B had better bonding quality than the W1.6–S1.7–T350–H22 joint. In general, the coexistence of the stronger failure mode A and the relatively weak failure mode B led to the unstable joining performance in the tensile shear test of the W1.6–S1.7–T400–H22 joint type shown in Fig. 7(a). When the forming temperature was further increased to 450 ◦ C, the failure mode was neck cracking for both the W1.6–S1.7–T450–H22 and W1.6–S1.7–T450–H28 joint types, as shown in Fig. 9(d),(e) and (f-3). 3.3. Evaluation of joining quality by static peel test Static peel tests were also carried out to further evaluate the joining performance of the W1.6–S1.7–T350–H22, W1.6–S1.7–T400–H22 and W1.6–S1.7–T450–H22 joint types. Fig. 10(a) shows a typical T-shaped specimen prepared for the peel test, and Fig. 10(b) shows an image of a representative joint during the peel test. The peel load was applied perpendicular to the joining surface, and the cross-head speed was 5 mm/min. To mitigate the instability of the results, five peel tests were performed for each joint type. The failure surfaces for the three joint types after the peel tests are shown in Fig. 10(c). The failure mode for the W1.6–S1.7–T350–H22 joint type was the ductile button pullout mode, while the failure mode for the W1.6–S1.7–T400–H22 joint type was the gradually separated button pull-out mode with more severe deformation of the sheets, demonstrating its superior interfacial bonding quality to that of the W1.6–S1.7–T350–H22 joint type. For the W1.6–S1.7–T450–H22 joint type, the cracks originated and propagated along the neck ring of the joint and the dominant failure mode was neck cracking. The peel force-displacement curves obtained from the peel tests are shown in Fig. 11. It is clear that the W1.6–S1.7–T400–H22, and

As discussed above, when the forming temperature was 300 ◦ C or lower, the thin sheets could not be bonded together, while the joints fabricated at 400 and 450 ◦ C demonstrated sound joining performance in both the tensile shear test and the peel test. This distinct difference in the joining performance may be attributed to diffusion bonding between the dissimilar materials. With the aim of revealing the diffusion phenomena of the main elements at the A6061P-T6/SUS304 interface, EDS line scan was performed for a typical W1.6–S1.7–T400–H22 joint. The results of the EDS line scan along the red line across the interface in the neck region of the W1.6–S1.7–T400–H22 joint at 30,000 × magnification are presented in Fig. 12(a). An approximately 800-nm-thick interdiffusion layer could be observed at the interface, which was mainly composed of the elements Al, Fe and Cr. The smooth variations of the Al, Fe and Cr indicate good atomic interdiffusion at the interface. Here, the formation of the thin interdiffusion layer at the interface in the neck region is one of the main reasons for the good joining performance realized by the proposed dissimilar joining process. In addition to diffusion bonding, the deformed pit and groove in the joint allowed surface enlargement and increased slipping resistance to tensile shear loading. This can be considered as a macroanchoring effect acting on the joint, as schematically illustrated in Fig. 12(b). Moreover, an irregular and wavy interface was observed in the neck region of the W1.6–S1.7–T400–H22 joint, as shown in the SEM images in Fig. 12(c). According to Bay (1983), the main bonding mechanisms for cold pressure welding were the fracture of the oxide layer and the local thinning of the contaminant film, which were induced under a large normal pressure and shear stress. These bonding mechanisms for cold pressure welding are also applicable to the joining process presented here. The severe plastic deformation in the neck region combined with the induction heating allowed the Al alloy to be firmly pressed into the microscale pits on the surface of the lower SUS304 sheet. The intimate contact enabled anchoring at the microscale and further improved the joining performance of the fabricated joint. 4.2. Effect of forming temperature on the joining performance In this study, the heat input improved the formability of the base materials and promoted atomic interdiffusion at the interface of the joint. However, it may also change the mechanical properties of the base materials. Vickers microhardness profiles measured from the Al alloy side were used to characterize the strength distributions of the Al alloy across the W1.6–S1.7–T350–I22, W1.6–S1.7–T400–H22 and W1.6–S1.7–T450–H22 joints after sufficient natural aging time, as shown in Fig. 13. A schematic diagram indicating the measured points (A6061P-T6 side of the corresponding joint) in the Vickers hardness tests is presented in Fig. 13(a). After forming at RT, 350, 400 and 450 ◦ C, the hardness values of the upper A6061-T6 sheet

Z. Huang, J. Yanagimoto / Journal of Materials Processing Technology 225 (2015) 393–404

401

Fig. 11. (a–c) Peel force/displacement curves of the W1.6–S1.7–T350–H22, W1.6–S1.7–T400–H22 and W1.6–S1.7–T450–H22 joint types.

Fig. 12. (a) EDS line scan across the interface of the neck region of the W1.6–S1.7–T400–H22 joint; (b) schematic of the macroanchoring effect of the deformed pit and groove in the joint; (c) microanchoring at the interface of the W1.6–S1.7–T400–H22 joint observed by SEM.

Fig. 13. (a) Schematic diagram of the measured points (A6061P-T6 side of the corresponding joint) in the Vickers hardness tests; (b) Vickers hardness profiles of the W1.6–S1.7–T350–H22, W1.6–S1.7–T400–H22, and W1.6–S1.7–T450–H22 joints.

402

Z. Huang, J. Yanagimoto / Journal of Materials Processing Technology 225 (2015) 393–404

Fig. 14. Fractured surfaces of the Al alloy remaining on the SUS304 side after the tensile shear test. (a) and (b) Point A and Point B in the failed W1.6–S1.7–T400–H22 joint (shown in Fig. 9(b)); (c) and (d) Point A and Point B in the failed W1.6–S1.7–T450–H22 joint (shown in Fig. 9(d)).

far from the plastically deformed zone were 93.4, 69.6, 60.8, and 54.7Hv, respectively, as shown by the dashed horizontal lines in Fig. 13(b). Overall, the hardness of the Al alloy was inversely proportional to the forming temperature, and the hardness values in the plastically deformed zone were lower than those in the undeformed zone. According to Sato et al. (1999), the softening and loss of strength of Al-Mg-Si alloy was mainly attributed to the heat input, because of the dissolution of strengthening needle-shaped ␤ (Mg2 Si) precipitates and the coarsening of non-strengthening rod-shaped ␤ (Mg2 Si) precipitates. A higher annealing temperature or greater heat input during the joining process would dissolve more initial strengthening ␤ precipitates in the Al alloy, markedly degrading the hardness and strength of the Al alloy. Moreover, a similar trend in the hardness profile was observed in the different observed regions for all the three joints except for region C (groove region). For both the W1.6–S1.7–T350–H22 and W1.6–S1.7–T400–H22 joints, the hardness values in region C were slightly lower than those in the other regions. Nevertheless, the decrease in Vickers hardness in the groove region was absent in the W1.6–S1.7–T450–H22 joint. This was because the Al alloy and SUS304 sheets were adequately filled into the groove in the die at a higher forming temperature (450 ◦ C), the compression pressure from the die could work harden the Al alloy in the groove region. Therefore, the Al alloy in the W1.6–S1.7–T450–H22 joint showed a different hardness profile in region C. The fractured surfaces of the Al alloy remaining on the SUS304 side in the failed W1.6–S1.7–T450–H22 and W1.6–S1.7–T400–H22 joints after the tensile shear tests were further observed by SEM, as shown in Fig. 14. Fig. 14(a) and (b), respectively, correspond to Points A and B shown in Fig. 9(b), while Fig. 14(c) and (d), respectively, correspond to Points A and B shown in Fig. 9(d). For both joints, during the tensile shear test, the tensile shear stress was concentrated at Point A, inducing crack initiation and propagation along the neck ring to Point B, whereas the bending moment caused extra peel stress at Point B. The failure morphologies at Point A showed the dominant ductile fracture mode with deep dimples and fibrils in the direction of pulling, while those at Point B displayed the main ductile fracture mode with many dimples as well as minor brittle cleavage facets. Furthermore, it was observed that

the dimples in the fractured surface of the W1.6–S1.7–I450–H22 joint were deeper than those in the W1.6–S1.7–T400–H22 joint. The increased ductility of the base Al alloy was consistent with the larger displacement of the W1.6–S1.7–T450–H22 joint upon tensile shear loading shown in Fig. 7. Nevertheless, as shown in Fig. 11, the W1.6–S1.7–T450–H22 joint exhibited lower failure energy in the peel test than the W1.6–S1.7–T400–H22 joint, which was due to the more severe degradation of the strength of the Al alloy caused by the higher forming temperature (450 ◦ C). Theoretically, a higher forming temperature (450 ◦ C) could improve the material flow and accelerate the atomic interdiffusion at the interface. Nevertheless, from the viewpoint of reducing the deterioration of the mechanical properties of the sheets, a lower forming temperature (400 ◦ C) is preferable. Further systematic investigation of the forming temperature in the range of 400–450 ◦ C is necessary to obtain a balance between the resulting joining performance and the residual mechanical properties of the base materials. 4.3. Effect of plastic strain distribution in the joint on the interdiffusion layer In the FSW process for Al alloy and steel sheets, both Watanabe et al. (2006) and Coelho et al. (2012) reported that the frictional heat and severe plastic deformation allow iron to interact with the aluminum to form IMCs at the interface of the FSW joint. In this study, the dissimilar components were also subjected to an elevated temperature and a large forming pressure during the bonding process. The distribution of plastic strain throughout the entire joint may affect the formation of the interdiffusion layer, and thus, lead to variation in the joining performance. Here, the three joints (shown in Fig. 5(b),(e) and (f)) fabricated at the same forming temperature of 400 ◦ C with the same heat holding time of 22 s were compared to rule out the effect of the heat input and only focus on the effect of plastic deformation. Fig. 15 shows the results of EDS characterization in different regions of the same W1.4–S1.7–T400–H22 joint, revealing that the interdiffusion layer was much thinner in the bottom region of the joint than that in the neck region. For the W1.4–S1.7–T400–H22 and W1.2–S1.7–T400–H22 joints, even in the neck region, the interdif-

Z. Huang, J. Yanagimoto / Journal of Materials Processing Technology 225 (2015) 393–404

403

Fig. 15. (a) Backscattered electron image of the A6061P-T6/SUS304 interface in the neck region of the W1.6–S1.7–T400–H22 joint at 10,000 × magnification and EPMA mapping of the element distribution across the interface; (b) backscattered electron image of the A6061P-T6/SUS304 interface in the bottom region of the W1.6–S1.7–T400–H22 joint at 10,000 × magnification and EPMA mapping of the element distribution across the interface.

Fig. 16. IPF and KAM maps of the SUS304 sheets in different regions of the fabricated joints. The length of each map equals the thickness of the SUS304 sheet after the corresponding plastic deformation. (a–c) bottom, groove and neck regions of the W1.6–S1.7–T400–H22 joint; (d) neck region of the W1.4–S1.7–T400–H22 joint; (e) neck region of the W1.2–S1.7–T400–H22 joint. The step size in EBSD is 0.8 ␮m. The gradation in color from blue to red corresponds to degrees of misorientation of 0–5◦ .

fusion layer was very thin (similar to the results shown in Fig. 15 (b)). The much thinner interdiffusion layers in the neck regions of the W1.4–S1.7–T400–H22 and W1.2–S1.7–T400–H22 joints may be the main reason for their inferior tensile shear joining performance compared with that of the W1.6–S1.7–T400–H22 joint. To further clarify the relationship between the thickness of the interdiffusion layer and the plastic strain, EBSD analysis was utilized to directly visualize the distribution of plastic strain throughout the joint, as shown in Fig. 16. Each mapping covered the entire thickness of the SUS304 sheet in the corresponding region of the joint. The top zone of each map, representing the SUS304 material adjacent to the joining interface, was employed to evaluate the amount of plastic deformation and the normal pressure at the interface. Specifically, inverse pole figure (IPF) and kernel average misorientation (KAM) maps for the bottom, groove and neck regions of the same W1.6–S1.7–T400–H22 joint are shown in Fig. 16(a–c), respectively. Among the three observed regions, the grains in the neck region were severely stretched and elongated parallel to the interface with the largest degree of KAM, indicating the largest plastic strain in the neck region. In contrast,

the material in the bottom region was subjected to little plastic strain and had almost unchanged grain sizes and the smallest degree of KAM. Moreover, the IPF and KAM maps in Fig. 16(d) and (e) show that the neck regions of the W1.4–S1.7–T400–H22, and W1.2–S1.7–T400–H22 joints have larger grain sizes and smaller degrees of KAM than the neck region of the W1.6–S1.7–T400–H22 joint, indicating less plastic strain. Overall, among the five regions characterized by EBSD analysis, the plastic strain in the neck region of the W1.6–S1.7–T400–H22 joint was the largest, which corresponded to the thickest interdiffusion layer as demonstrated by the aforementioned EDS characterization. Therefore, in this joining process, the thickness of the interdiffusion layer at the interface of the joint is positively correlated with the locally distributed plastic strain. 5. Conclusion In the present work, a dissimilar joining process by thermally assisted plastic deformation was employed for 0.8-mm-thick A6061P-T6 and 0.5-mm-thick SUS304 thin sheets. The proposed

404

Z. Huang, J. Yanagimoto / Journal of Materials Processing Technology 225 (2015) 393–404

joining process, which exhibits excellent joining properties as well as high stability and productivity, is an alternative joining method for assembling Al-steel lightweight hybrid products. The conclusions drawn from this work are given below: 1 The effects of the forming stroke, the forming temperature, the heat holding time, the groove width in the die and the punch corner radius on the joining performance were investigated. In the forming temperature range of 350–450 ◦ C, the two dissimilar sheets could be successfully bonded. Among the various joint types, the W1.6–S1.7–T450–H22 joint exhibited the highest joint efficiency factor of 85.2% with the largest displacement and absorption energy, while the W1.6–S1.7–T400–H22 joint absorbed the highest failure energy in the peel test. 2 The excellent joining performance of this dissimilar joining process was attributed to the good atomic interdiffusion, the microanchoring effect at the interface, the macroanchoring effect of the deformed pit and groove in the joint and the overall surface enlargement. 3 Increasing the forming temperature has a twofold effect in the joining process. On the one hand, it significantly improves the formability of the base materials and accelerates atomic interdiffusion at the interface, which are beneficial for the joining performance. On the other hand, a high forming temperature degrades the mechanical properties of the base materials, which is detrimental to the resulting joining performance. 4 The interdiffusion layer thickness is positively correlated with the amount of locally distributed plastic strain at the interface of the fabricated joint. A suitable heating history and a large amount of plastic deformation result in the formation of a continuous 800nm-thick interdiffusion layer mainly composed of Al, Fe and Cr in the neck region of the W1.6–S1.7–T400–H22 joint. Acknowledgments This work was financially supported by a Grant-in-Aid for Scientific Research (A) (Contract No. 22246093) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan.

References Abe, Y., Kato, T., Mori, K., 2009. Self-piercing riveting of high tensile strength steel and aluminium alloy sheets using conventional rivet and die. J. Mater. Process. Technol. 209, 3914–3922. Abe, Y., Mori, K., Kato, T., 2012. Joining of high strength steel and aluminium alloy sheets by mechanical clinching with dies for control of metal flow. J. Mater. Process. Technol. 212, 884–889. Bang, H.S., Bang, H.S., Jeon, G.H., Oh, I.H., Ro, C.S., 2012. Gas tungsten arc welding assisted hybrid friction stir welding of dissimilar materials Al6061-T6 aluminum alloy and STS304 stainless steel. Mater. Des. 37, 48–55. Bay, N., 1983. Mechanisms producing metallic bonds in cold welding. Weld. Res. Suppl., 137–142. Briskham, P., Blundell, N., Han, L., Hewitt, R., Young, K., Boomer, D., 2006. Comparision of self-piece riveting, resistance spot welding and spot friction welding for aluminum automotive sheet. SAE Int. 01, 0774. Coelho, R.S., Kostka, A., dos Santos, J.F., Kaysser-Pyzalla, A., 2012. Fiction-stir dissimilar welding of aluminium alloy to high strength steels: mechanical properties and their relation to microstructure. Mater. Sci. Eng. A 556, 175–183. Connolly, C., 2007. Friction spot joining in aluminium car bodies. Ind. Robot 34 (1), 17–20. Flitta, I., Sheppard, T., 2003. Nature of friction in extrusion process and its effect on material flow. Mater. Sci. Technol. 19, 837–846. Groche, P., Wohletz, S., Brenneis, M., Pabst, C., Resch, F., 2014. Joining by forming–a review on joint mechanisms, applications and future trends. J. Mater. Process. Technol. 214 (10), 1972–1994. Mishra, R.S., Ma, Z.Y., 2005. Friction stir welding and processing. Mater. Sci. Eng. R 50, 1–78. Mori, K., Bay, N., Fratini, L., Micari, F., Tekkaya, A.E., 2013. Joining by plastic deformation. CIRP Ann. – Manuf. Technol. 62 (2), 25–28. Ozaki, H., Kutsuna, M., 2012. Dissimilar metal joining of zinc coated steel and aluminium alloy by laser roll welding. Intech, http://dx.doi.org/10.5772/48242 Sato, Y.S., Kokawa, H., Enomoto, M., Jogan, S., 1999. Microstructural evolution of 6063 aluminum during friction-stir welding. Metall. Mater. Trans. A 30A, 2429–2437. Schikorra, M., Donati, L., Tomesani, L., Kleiner, M., 2007. The role of friction in the extrusion of AA6060 aluminium alloy process analysis and monitoring. J. Mater. Process. Technol. 191 (1–3), 288–292. Sierra, G., Peyre, P., Beaume, F.D., Stuart, D., Fras, G., 2008. Steel to aluminium braze welding by laser process with Al-12 Si filler wire. Sci. Technol. Weld. Join. 13 (5), 430–437. Torkamany, M.J., Tahamatan, S., Sabbaghzadeh, J., 2010. Dissimilar welding of carbon steel to 5754 aluminium alloy by Nd YAG pulsed laser. Mater. Des. 31, 458–465. Watanabe, T., Takayama, H., Yanagisawa, A., 2006. Joining of aluminium alloy to steel by friction stir welding. J. Mater. Process. Technol. 178, 342–349.