- Email: [email protected]
Removal of oxides and brittle coating constituents at the surface of coated hot-forming 22MnB5 steel for a laser welding process with aluminum alloys
Surface & Coatings Technology 285 (2016) 153–160
Contents lists available at ScienceDirect
Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat
Removal of oxides and brittle coating constituents at the surface of coated hot-forming 22MnB5 steel for a laser welding process with aluminum alloys M. Windmann a,⁎, A. Röttger a, H. Kügler b, W. Theisen a a b
Lehrstuhl Werkstofftechnik, Ruhr-Universität Bochum, 44801 Bochum, Germany BIAS — Bremer Institut für angewandte Strahltechnik GmbH, Klagenfurter Str. 2, 28359, Bremen, Germany
a r t i c l e
i n f o
Article history: Received 23 June 2015 Revised 26 October 2015 Accepted in revised form 21 November 2015 Available online 22 November 2015 Keywords: Surface treatment Laser ablation Sandblasting Intermetallic phases Steel/aluminum laser welding
a b s t r a c t The surface of a press-hardened steel 22MnB5 coated with Al-base (AlSi10Fe3) and Zn-base (ZnNi10) was conditioned by a pulse laser and by sandblasting to remove undesirable oxides and brittle phases. Oxides formed on coating surfaces counteract the wettability of welding filler during a welding or brazing process. Furthermore, welding and brazing joints of 22MnB5 coated with aluminum alloys failed along the brittle intermetallic phases in the coating under a low mechanical load. Treated 22MnB5 surfaces were analyzed microscopically, and the phase compositions were investigated by synchrotron diffraction measurements. It was found that brittle phases could be locally removed by laser ablation; however, high laser energies led to remelting and oxidation of the coating surface. In contrast, sandblasting homogenously removed oxides and brittle intermetallic phases. Surface-treated 22MnB5 steel sheets were joined to AA6016 aluminum sheets by laser welding, and the strength of the weldment was determined by tensile tests. The measured mechanical strength of the aluminum/steel joints was 210–230 MPa. Failure of the weldments under tensile loading occurred within the aluminum sheet, away from the steel surface/welding filler interface if brittle coating components and oxides were removed homogenously. © 2015 Published by Elsevier B.V.
1. Introduction High-strength steel 22MnB5 in the press-hardened condition is used in automotive bodies to comply with the safety requirements (high strength and stiffness) in safety-relevant structures. Aluminum materials can decrease the weight of cars in areas where lower mechanical properties can be tolerated. To utilize the advantages of both materials, suitable joining technologies are required in areas where both materials are used (e.g. underbody of the automotive bodywork). Most of the published investigations about thermal joining of steel/aluminum compounds focused on uncoated or galvanized mild steel sheets. For example, steel and aluminum sheets were joined by shock welding [1], friction welding [2], or laser welding [3]. Especially laser welding was found to be a suitable joining technology for different dissimilar welding operations [3–6]. However, less literature is available about thermal joining (welding/brazing) of coated, high-strength steel 22MnB5 to aluminum alloys. In the fusion welding process of steel and aluminum, generally only the aluminum alloy is melted due to its lower melting temperature. The steel surface remains solid and is wetted with the melted aluminum, which can thus be described as a welding (aluminum) — brazing (steel surface) process [7,8]. Thus, the ⁎ Corresponding author. E-mail address: [email protected] (M. Windmann).
http://dx.doi.org/10.1016/j.surfcoat.2015.11.037 0257-8972/© 2015 Published by Elsevier B.V.
achievable strength of coated 22MnB5/aluminum joints depends on the properties of the steel coating and the adhesion of liquid welding filler on the coated 22MnB5 surface. Sheets of 22MnB5 steel are usually formed and hardened in a presshardening process, in which a material strength of up to 1800 MPa can be achieved in complex formed body structures [9]. During press hardening, the steel sheets are austenitized at 880–950 °C and then transported into a tool where they are formed and quenched in one step [10]. To prevent strong oxidation of the steel sheets during austenitization, a protective Zn-, or Al-base coating is deposited on the steel surface prior to the press-hardening process [11]. During austenitization, distinctive diffusion processes occur at the interface between the steel substrate and the coating, leading to the formation of brittle intermetallic AlxFey phases (Al13Fe4, Al5Fe2, Al2Fe, AlFe) [12] or ZnxFey phases (Fe3Zn10) [13] in the coating. In addition, Al-rich or Znrich oxides of type M2O3 and MO are formed on the coating surface [8]. In the Al-rich coating, the phase of type Al5Fe2 possesses the highest volume fraction in the press-hardened coating and exhibits a low fracture toughness of 1 MPa m0.5 [14]. The formation of brittle intermetallic phases promotes crack initiation and propagation during forming in the press-hardening tools so that the coating microstructure possesses a network of cracks. The brittle Al coating counteracts the braze- and weldability of coated 22MnB5 owing to crack growth along the brittle intermetallic AlxFey phases [15]. For this reason, brittle intermetallic
154
M. Windmann et al. / Surface & Coatings Technology 285 (2016) 153–160
AlxFey phases can be transformed into more ductile phases by adaption of the press-hardening parameters to simultaneously increase the fracture toughness of the coating and decrease the crack density [12, 16]. As an alternative that does not require adaption of the process parameters, the aim of this study is to remove brittle intermetallic phases by means of a pulse laser or sandblasting. In Zn-base coatings, no cracks are formed in the press-hardening process [11]. During welding or brazing with cold-rolled and not additionally heat-treated Zn-coated mild steel sheets, the coating is completely melted and the liquid filler achieves good adhesion to the surface [8,17,18]. In contrast, during press-hardening, the Zn coating completely transforms into α-Fe and Fe3Zn10 [13]. Consequently, the former Zn-base coating is not melted in a brazing process. The melted filler has to be brazed on the oxidized surface of the Zn-base coating on the press-hardened steel. For this reason, the oxides on the Zn coating should be removed to improve the adhesion of a filler in a welding or brazing process. In this study, the surface of Al-base (AlSi10Fe3)- and Zn-base (ZnNi10)-coated and industrially press-hardened 22MnB5 steel was conditioned with a pulse laser and sandblasting to remove undesirable oxides and brittle phases, which should improve the adhesion of a filler on the coating surface. The treated Al-base- and Zn-base-coated 22MnB5 steel sheets were subsequently joined to aluminum alloy AA6016 by laser welding/brazing. Removing brittle coating components and oxides from the coating surfaces should lead to a good bonding strength between the coated steel and the filler, which was evaluated by tensile tests.
this purpose, the cross-section of each sample was ground with 54 and 18 μm abrasives and then polished with a 3 μm diamond suspension and finally with a ¼ μm SiO2 suspension. The chemical composition and oxide content on the coating surface was analyzed by EDS measurements using an acceleration voltage of 20 keV and a working distance of 8.5 mm. EDS measurements enabled to compare the oxygen content at the coating surfaces in the treated and initial condition and thereby discuss the removal of oxides.
2. Experimental
The phase composition of the treated coating surface was analyzed with synchrotron diffraction measurements (λ = 0.45919 Å) at the electron storage ring facility Delta at TU Dortmund (Germany). The synchrotron radiation beam was focused parallel to the specimen and into the treated coating, which was further tilted by an angle of 5°. Debye–Scherrer circle segments (140–155°) were conditioned and integrated with the program Fit2D (ESRF). Phase analysis was performed using the ICDD-JCDPS database PDF-2.
2.2. Surface treatment Surface treatment by laser ablation was carried out with the Nd:YAG laser Spectron SL900 and an Arges Racoon 16 scanner system equipped with f-theta lens (Fig. 1). This pulse laser source provided a maximum mean power of 16 W with a wavelength of 1064 nm. In these experiments, we used a spot diameter of 16 μm in the focal plane, a pulse duration of 100 ns, and a beam propagation velocity of 160 mm/s. The pulse frequency was set to 10 kHz, and the mean laser power was varied between 5 and 13 W. Argon was used as shielding gas. Sandblasting was performed manually using a Hessler Strahlboss 140 device with a nozzle diameter of 10 mm, an angle of incidence of 60°, and a distance of 100 mm to the coating surface. Silicon oxide (SiO2: 90–355 μm) was used as the abrasive. The sandblasting time was 10 s per steel sheet (50 × 120 mm) and the air pressure was set to 3 bar. 2.3. Phase analysis
2.1. Materials and microstructural analysis AlSi10Fe3- and ZnNi10-coated sheets of 22MnB5 steel with a size of 50 × 120 mm and a thickness of 1.5 mm were investigated in this study. The AlSi10Fe3 layer (thickness: 25–35 μm) was deposited on the 22MnB5 by hot dipping at a temperature of 675 °C. The ZnNi10 coating (thickness: 24–30 μm) was galvanically deposited on the steel surface. The chemical composition of 22MnB5 steel was measured by optical emission spark spectroscopy and is given in Table 1. In addition, the chemical composition of the AlSi10Fe3 and ZnNi10 coatings was determined in the solidified state by energy dispersive x-ray spectroscopy (EDS) measurements. The coated 22MnB5 sheets were welded with naturally aged (T4) AA6016 aluminum with a thickness of 1.5 mm (see Section 2.4). In addition, fillers of AlSi3Mn (Al-base coating) and ZnAl15 (Zn-base coating) with a wire diameter of 1.2 mm were used. The chemical composition of the AA6016 aluminum alloy and the fillers is also given in Table 1. Microstructural examinations of the cross-sections were performed by optical light microscopy (OM), and examinations of the coating surfaces were performed by scanning electron microscopy (SEM). For
2.4. Welding setup Welding experiments were performed to determine the influence of the surface treatment on the bonding behavior of the coated 22MnB5 steel in a welding/brazing process with aluminum. Thereby, the steel surface remained solid and was wetted with the melted aluminum. Surface-treated 22MnB5 metal sheets (50 × 120 mm) were joined with AA6016 aluminum alloy (50 × 150 mm) in an overlap configuration with AA6016 in the upper position (Fig. 2). Welding was performed using a Nd:YAG laser beam (Trumpf HL4006D — maximum power: 4 kW) and a power supply for inductive preheating (maximum power output of 10 kW). The wire feed angle was adjusted to 25° to the sheet plane. The wire feed of the filler was directed contrary to the welding direction. FontArgen F400 NH flux was applied to the treated
Table 1 Chemical composition of the materials measured by emission spark spectrometry and energy dispersive x-ray spectrometry. Chem. composition [mass%]
C
Si
Al
Mn
Mg
Ni
B
Fe
Zn
Substrate 22MnB5 AlSi10Fe coating ZnNi10 coating AA6016
0.234 – – –
0.289 10.230 – 1.047
0.034 Bal. – Bal.
1.258 – – 0.135
– – – 0.463
– – 10,780 –
0.002 – – –
bal. 2.160 2.080 0.249
– – Bal. –
Filler AlSi3Mn1 ZnAl152
– –
3.300 0.002
Bal. 14.620
1.100 ---
0.010 ---
– ---
– ---
0.190 0.004
– Bal.
1 2
Provided by Drahtwerke Elisental. Provided by Grillo.
M. Windmann et al. / Surface & Coatings Technology 285 (2016) 153–160
155
ZnNi10-coated 22MnB5 steel and the aluminum alloy AA6016 are presented. 3.1. Characterization of the AlSi10Fe3 and ZnNi10 coating after presshardening
Fig. 1. Setup of the laser ablation process.
steel surface and dried before welding. Argon was used as the shielding gas during welding. The welding parameters (Table 2) were chosen on the basis of prior investigations [19]. 2.5. Tensile tests To determine the strength of the welded joints with respect to the surface treatment, tensile tests were performed according to DIN EN ISO 6892. The locations where test specimens were taken are illustrated in Fig. 3. A Z100 Zwick/Roell machine was used for fracture testing of three specimens of each welded joint at a test speed of 0.5 mm/min. The load was measured with a load cell (measuring range: 100+0.01 −0.05 kN), and the elongation was measured with an extensometer (measuring range: 2.5+0.00 −0.03 mm). The measured values were digitized by TestXpert software and mapped as a strength/elongation curve. 3. Results and discussion The microstructure of the Al-base coating (AlSi10Fe3), the Zn-base coating (ZnNi10), and the 22MnB5 steel/coating interface in the presshardened condition is presented first. Subsequently, surface treatment by laser ablation and sandblasting is discussed. Finally, the results of laser welding experiments with the surface-treated AlSi10Fe3- or
Fig. 4a shows the microstructure of the AlSi10Fe3 coating in the press-hardened condition (TAUS = 920 °C, tAUS = 6 min, quenched in the stamping tool). The coating (thickness: 25–35 μm) is characterized by a layered structure and a high crack density. Phases of type AlFe and Al5Fe2 were formed in the press-hardening process due to strong diffusion of iron into the coating at the austenitization temperature of 920 °C. The formation of cracks is attributed to the low fracture toughness of the phase of type Al5Fe2 (KIC = 1 MPa m0.5 [14]), which led to crack initiation and propagation during cooling and forming in the press-hardening tool. The phase formation in the AlSi10Fe3 coating during austenitization in the press-hardening process is discussed in detail in [12]. A thin Al- and Si-rich oxide layer could be identified at the surface (Fig. 4c). An oxygen content of 24 at.% was measured at the surface by EDS. The formation of oxides on the surface of Alcoated iron-base substrates during heat treatment was presented in detail in [20]. Fig. 2b shows the microstructure of the ZnNi10 coating in the presshardened condition (thickness: 24–30 μm). The strong diffusion of iron into the ZnNi10 coating as well as the diffusion of Zn and Ni into the steel substrate during austenitization led to a complete transformation of the coating into α-Fe, in which 8 at.% Ni and 18 at.% Zn were dissolved, as measured by EDS. Phase formation in the ZnNi10 coating during press-hardening is summarized in [13]. In contrast to the AlSi10Fe3 coating, no cracks were formed in the ZnNi10 coating after cooling and quenching in the press-hardening tool. The surface of the ZnNi10 coating was characterized by a 3–5 μm thick oxide layer (Fig. 4d). An oxygen content of 33 at.% was measured at the surface by EDS. Lee et al. found that this layer consists of oxides of type Mn3O4 and ZnO [21]. In this study, the coated steels sheets were joined with an aluminum alloy in a laser welding/brazing process. For a first evaluation of the adhesion of welding fillers AlSi3Mn (for the AlSi10Fe3 coating) and ZnAl15 (for the ZnNi10 coating) on the coating surfaces, brazing experiments were performed. Brazing on the AlSi10Fe3 coating surface led to crack growth and complete separation of the brazed AlSi3Mn layer along the brittle intermetallic coating (Fig. 5a). During cooling after the brazing process, internal stresses were formed due to the different thermal coefficients of 22MnB5, AlSi3Mn, and the intermetallic phases in the coating. These stresses led to crack growth along the brittle intermetallic phases in the coating due to their low fracture toughness. Fig. 5b shows that the surface of the ZnNi10 coating was only poorly wetted with the ZnAl15 filler (contact angle N 90°). Adhesion of the welding filler was probably inhibited by the thick oxide layer on the coating surface. For both coating types, variation of the brazing parameters did not lead to a strong improvement of the wetting characteristic.
Fig. 2. Setup of the Nd:YAG laser-welding process [23].
156
M. Windmann et al. / Surface & Coatings Technology 285 (2016) 153–160
Table 2 Welding parameters. Coating
Preheating temperature [°C]
Laser energy [kJ/cm]
Welding velocity [m/min]
Velocity of filler [m/min]
AlSi10Fe ZnNi10
490 320
3.3 1.5
0.5 0.2
2.6 1.3
Brazing experiments indicated that the surface of Al- or Zn-coated 22MnB5 steel sheets have to be treated after press-hardening to increase their ability for a fusion welding/brazing process with aluminum alloys. Thus, the coating surfaces were treated by laser ablation and sandblasting to remove oxides and brittle coating components, thus improving the adhesion on the 22MnB5 surface.
coating components of the AlSi10Fe3 coating. On the one hand, if the laser energy was too low, it interacted insufficiently with the coating surface, thus only an incomplete removal of the coating occurred. If the laser energy was too high, it led to partial or full remelting of the coating, which is associated with a strong oxidation of the coating.
3.2. Laser ablation of AlSi10Fe3 and ZnNi10 coatings on 22MnB5
3.2.2. ZnNi10 coating Fig. 7 shows the cross-section and top view of the press-hardened ZnNi10 coating after pulsed laser ablation. Irrespective of the used laser power, the oxidized coating components could not be removed by spalling due to thermal stresses, as described previously for the AlSi10Fe3 coating. The ZnNi10 coating was characterized by bodycentered α-Fe in the press-hardened condition (see Section 3.1). As a result, the thermal expansion coefficients of the coating and the martensitic 22MnB5 steel did not significantly differ, for which reason cracks and delaminations due to thermoshock were not formed. Treatment with the pulse laser only led to the formation of new oxides on the surface (≥ 6.5 W). The amount of resulting oxides increased with increasing laser power and was always above the measured value in the press-hardened condition (33 at.%). As found for the AlSi coating, the ZnNi coating was also remelted at high laser powers (≥13 W). The high temperature in the melted coating led to evaporation of zinc and strong oxidation of the surface. A maximal oxygen content of 42 at.% was measured at the surface by EDS. Hence, the removal of oxidized ZnNi10 coating components without formation of new oxides was not possible by pulse laser ablation.
3.2.1. AlSi10Fe3 coating Fig. 6 shows the cross-section and the top view of the AlSi10Fe3 coating after laser treatment with different laser powers. A laser power of up to 5 W did not promote ablation of the coating components. A thin oxide layer could still be identified on the coating surface. The oxygen content at the surface did not differ significantly from the measured value in the press-hardened condition (24 at.%). Thus, the structure of the coating was identical to the press-hardened condition. At higher laser powers (≥ 6.5 W), coating particles of phases Al5Fe2 and AlFe were locally removed from the coating surface. The coating thickness decreased to 7–20 μm after surface treatment with a laser power of 6.5 W (initial condition: 24–30 μm). Material ablation of the coating surface can be explained by thermoshock. At the surface, the coating was rapidly heated by the focused laser and then cooled by heat transfer into the steel substrate. Thermal stresses formed as a result of differences in the thermal coefficients and in the Young's moduli of the intermetallic phases in the coating and the steel substrate. If the stresses exceed the material strength, this leads to crack growth and delamination of coating components across the brittle intermetallic phases. The amount of ablated brittle coating components increased with increasing laser power which resulted in thinner coating thicknesses after laser ablation. However, a coating layer with a thickness of 5–10 μm remained at the surface for all used laser powers. Furthermore, a laser power higher than 13 W led to remelting of the coating and a strong oxidation of the surface. EDS measurements confirmed the formation of Al-, Fe-, and Si-rich oxides. A high oxygen content of 41 at.% was measured at the coating surface. In this study, no laser power was identified that enabled complete removal of the brittle
Fig. 3. 22MnB5/AA6016 welded joints and positions of the specimens for tensile tests.
Fig. 4. Top row: OM micrographs of the cross-section of a) AlSi10Fe3- and b) ZnNi10-coated 22MnB5. Bottom row: SEM micrographs of the coating surfaces (top view) of a) AlSi10Fe3 and b) ZnNi10 after press-hardening.
M. Windmann et al. / Surface & Coatings Technology 285 (2016) 153–160
157
Fig. 5. a) OM micrograph of the cross-section of AlSi3Mn brazed on AlS10Fe3-coated 22MnB5, b) OM micrograph of the cross-section of ZnAl15 brazed on ZnNi10-coated 22MnB5.
3.3. Sandblasting of AlSi10Fe3 and ZnNi10 coatings on 22MnB5 3.3.1. AlSi10Fe3 coating As indicated by Fig. 8, the brittle AlSi10Fe3 coating could be almost completely removed by sandblasting. Synchrotron diffraction measurements confirmed that the surface of the AlSi10Fe3-coated 22MnB5 was free of intermetallic phases after sandblasting (Fig. 9). Oxides were also almost completely removed from the surface (4 at.% O). The surface was characterized by α-Fe and AlFe, in which up to 40 at.% aluminum was measured by EDS. During sandblasting, the brittle and crack-initiated coating was removed layer-wise as a result of a strong mechanical interaction between the accelerated grit and the brittle Al5Fe2 and AlFe coating components. Fig. 10 shows the structure of the coating surface after a short sandblasting time of 3 s. The sandblasting abrasives broke out the brittle coating components along the cracks formed during the press-hardening process (Fig. 4). A sandblasting time of
10 s was necessary (for the sandblasting setup and specimen size in this study, see Section 2) to completely remove all brittle coating components and to form a crack and oxide-free surface. 3.3.2. ZnNi10 coating Sandblasting also enabled removal of the oxide layer from the ZnNi10 surface (4 at.% O), as indicated by Figs. 8 and 9. In contrast to the AlSi10Fe3 coating, ZnNi10 coating particles were not removed. This can be explained by the absence of brittle phases and cracks in the ZnNi10 coating (α-Fe) after press-hardening. 3.4. Welding experiments and tensile testing As described above, only sandblasting enabled complete removal of brittle coating particles and oxides of Al-base- and Zn-base-coated HSLA steel surfaces. Thus, welding experiments (brazing on the steel side)
Fig. 6. Top row: OM micrographs of the cross-section of AlSi10Fe3-coated 22MnB5 after laser ablation with a mean power of a) 5 W, b) 6.5 W, and c) 13 W. Bottom row: SEM micrographs of the coating surfaces (top view) after laser ablation with a mean power of d) 5 W, e) 6.5 W, and f) 13 W.
158
M. Windmann et al. / Surface & Coatings Technology 285 (2016) 153–160
Fig. 7. Top row: OM micrographs of the cross-section of ZnNi10-coated 22MnB5 after laser ablation with a mean power of a) 5 W, b) 6.5 W, and c) 13 W. Bottom row: SEM micrographs of the coating surfaces (top view) after laser ablation with a mean power of d) 5 W, e) 6.5 W, and f) 13 W.
were performed using sandblasted steel sheets. Fig. 11 shows that the surface treatment promoted good adhesion of the filler to the 22MnB5 steel surface. At the interface of the sandblasted AlSi10Fe3 coating and
Fig. 8. Top row: OM micrographs of the cross-section of a) AlSi10Fe3- and b) ZnNi10-coated 22MnB5. Bottom row: SEM micrographs of the coating surfaces (top view) after sandblasting of c) AlSi10Fe3 and d) ZnNi10.
the AlSi3Mn filler, no cracks were formed, in contrast to the presshardened condition (Fig. 3). During the brazing process on the steel surface, diffusion of iron into the melted aluminum and aluminum into the steel substrate led to the formation of intermetallic Al–Fe phases at the steel/aluminum interface. The formation of brittle intermetallic Al–Fe phases promotes crack formation and propagation, accompanied by a brittle fracture of the welded joint during cooling if a critical thickness of 10 μm is exceeded [22]. However, the aluminum-saturated 22MnB5 surface (up to 40 at.% Al) inhibited strong aluminum diffusion into the steel substrate. The thickness of the intermetallic Al–Fe layer was below 10 μm along the length of the steel/aluminum interface. Phase
Fig. 9. Diffraction patterns of the AlSi10Fe3 and ZnNi10 coatings in the press-hardened and sandblasted condition.
M. Windmann et al. / Surface & Coatings Technology 285 (2016) 153–160
159
Fig. 12. Specimen of the steel/aluminum welding joints after tensile tests; laser welded/ brazed after sand-blasting of the coating surface.
22MnB5/filler interface, but rather in the aluminum AA6016 alloy in tensile tests. The tensile strength of all tested specimens was 210–230 MPa. This value meets the tensile strength of the Al-base material AA6016 in the T4 condition. Thus, sandblasting of the AlSi10Fe3 and ZnNi10 coating surfaces increased the bonding strength at the coating/filler interface above the tensile strength of the AA6016 alloy. 4. Conclusions Fig. 10. SEM micrograph of the AlSi10Fe3 coating surface (top view) after sandblasting for 3 s.
formation at the interface between 22MnB5 steel and AlSi3Mn filler was described in detail in [19]. In the tensile tests, the welded joints did not fail at the 22MnB5/filler interface, but rather in the aluminum AA6016 alloy (Fig. 12). The adhesion of the welding filler on the Zn-base coating surface was greatly improved by sandblasting (Fig. 11b) compared to the initial state (oxidized coating surface) after press-hardening (Fig. 5). As indicated by Fig. 11d, oxides were no longer found at the interface between the sandblasted ZnNi10 coating and the ZnAl15 filler. Metallurgical bonding was primarily accomplished by formation of intermetallic Al–Fe phases due to the high aluminum content in the ZnAl15 filler. For these intermetallic phases, a stoichiometry near to the phase of type Al5Fe2 was measured by EDS. Pure zinc precipitates were formed between the intermetallic Al–Fe phases. Cracks were not found along the steel/filler interface. The welded joints also did not fail at the
The surfaces of Al-base and Zn-base coatings on press-hardening steel 22MnB5 were treated by laser ablation and sandblasting to remove brittle coating components and oxides from the coating surface. This improved the adhesion of the welding fillers on the coated steel for the laser welding/brazing process with aluminum alloy AA6016. The following key results were found: - AlSi10Fe3- and ZnNi10-coated and press-hardened steel 22MnB5 had a poor wettability with the welding filler in the initial state - Laser ablation was not able to completely remove brittle coating components and oxides of the AlSi10Fe3 and ZnNi10 coating surfaces. Brittle coating particles could be locally removed from the coating surface. Use of high laser powers led to remelting and formation of new oxides on the coating surface. - Brittle coating components and oxides could be almost completely removed by sandblasting. - Removal of brittle coating components from the AlSi10Fe3 surface occurred layer-wise.
Fig. 11. a) OM micrograph of the cross-section of a laser beam welding/brazing joint between AlSi10Fe3-coated 22MnB5 and AA6016 (filler: AlSi3Mn). b) OM micrograph of the crosssection of a laser beam welding/brazing joint between ZnNi10-coated 22MnB5 and AA6016 (filler: ZnAl15). c) SEM micrograph of the interface between sandblasted 22MnB5 + AlSi10Fe3 and AlSi3Mn filler. d) SEM micrograph of the interface between sandblasted 22MnB5 + ZnNi10 and ZnAl15 filler.
160
M. Windmann et al. / Surface & Coatings Technology 285 (2016) 153–160
- The wettability of the sandblasted coating surfaces was greatly improved compared to the initial state after press-hardening. - Laser welding/brazing joints of sandblasted Al- or Zn-coated 22MnB5 with the aluminum alloy AA6016 did not fail at the interface between the coating surface and filler, but rather in the aluminum alloy in tensile tests (tensile strength: 210–230 MPa).
Acknowledgments We gratefully acknowledge financial support from the BMBF (Bundesministerium für Bildung und Forschung) within the project “Hybrides Fügen von Multimaterialsystemen für Kraftfahrzeuge” (03X3032). We thank our project partners Volkswagen AG, ThyssenKrupp Steel, Waldaschaff Automotive GmbH, Drahtwerke Elisental, Grillo and Steremat Induktion GmbH (all Germany) for their support with the specimens and the welding setup. In addition, we thank the Delta electron storage ring in Dortmund (Germany) for their great support and help with the synchrotron measurements. References [1] H. Date, Y. Sato, M. Naka, J. Mater. Sci. Lett. 12 (1993) 626–628. [2] M. Sahin, Int. J. Adv. Manuf. Technol. 41 (2009) 487–497. [3] M. Schimek, A. Springer, S. Kaierle, D. Kracht, V. Wesling, Phys. Procedia 39 (2012) 43–50.
[4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23]
M.J. Torkamany, F. Malek Ghaini, R. Poursalehi, Mater. Des. 53 (2014) 915–920. C. Tan, W. He, X. Gong, L. Li, J. Feng, Mater. Des. 78 (2015) 51–62. J. Sun, Q. Yan, W. Gao, J. Huang, Mater. Des. 83 (2015) 120–128. M.J. Zhang, G.Y. Chen, Y. Zhang, K.R. Wu, Mater. Des. 45 (2013) 24–30. G. Qin, Z. Lei, Y. Su, B. Fu, X. Meng, S. Lin, J. Mater. Process. Technol. 214 (2014) 2684–2692. T. Manzenreiter, M. Rosner, T. Kurz, G. Brugger, R. Kelsch, D. Hartmann, A. Sommer, BHM 157 (2012) 97–101. A. Barcellona, D. Palmeri, Metall. Mater. Trans. A 40 (2009) 1160–1174. D.W. Fan, B.C. De Cooman, Steel Res. Int. 83 (No.5) (2012) 412–433. M. Windmann, A. Röttger, W. Theisen, Surf. Coat. Technol. 226 (2013) 130–139. J. Kondratiuk, P. Kuhn, E. Labrenz, C. Bischoff, Surf. Coat. Technol. 205 (2011) 4141–4153. U. Köster, W. Liu, H. Liebertz, M. Michel, J. Non-Cryst. Solids 153/154 (1993) 446–452. C. Kim, M.J. Kang, Y.D. Park, Procedia Eng. 10 (2011) 2226–2231. M. Windmann, A. Röttger, W. Theisen, Surf. Coat. Technol. 246 (2014) 17–25. P. Peyre, G. Sierra, F. Deschaux-Beaume, D. Stuart, G. Fras, Mater. Sci. Eng. A 444 (2007) 327–338. C. Dharmendra, K.P. Rao, J. Wilden, S. Reich, Mater. Sci. Eng. A 528 (2011) 1497–1503. M. Windmann, H. Kügler, A. Röttger, W. Theisen, F. Vollertsen, J. Mater. Process. Technol. 217 (2015) 88–95. Y. Kitajima, S. Hayashi, T. Nishimoto, T. Narita, S. Ukai, Oxid. Met. 73 (2010) 375–388. C.W. Lee, W.S. Choi, Y.R. Cho, B.C. De Cooman, ISIJ Int. 54 (10) (2014) 2364–2368. C. Radscheit, Laserstrahlfügen von Aluminium und Stahl, BIAS Verlag, 1997. Reprinted from Journal of Materials Processing Technology, 217, M. Windmann, A. Röttger, H. Kügler, W. Theisen, F. Vollertsen, Laser beam welding of aluminum to Al-base coated high-strength steel 22MnB5/Experimental, p. 90, Copyright (2015), with permission from Elsevier
Contents lists available at ScienceDirect
Surface & Coatings Technology journal homepage: www.elsevier.com/locate/surfcoat
Removal of oxides and brittle coating constituents at the surface of coated hot-forming 22MnB5 steel for a laser welding process with aluminum alloys M. Windmann a,⁎, A. Röttger a, H. Kügler b, W. Theisen a a b
Lehrstuhl Werkstofftechnik, Ruhr-Universität Bochum, 44801 Bochum, Germany BIAS — Bremer Institut für angewandte Strahltechnik GmbH, Klagenfurter Str. 2, 28359, Bremen, Germany
a r t i c l e
i n f o
Article history: Received 23 June 2015 Revised 26 October 2015 Accepted in revised form 21 November 2015 Available online 22 November 2015 Keywords: Surface treatment Laser ablation Sandblasting Intermetallic phases Steel/aluminum laser welding
a b s t r a c t The surface of a press-hardened steel 22MnB5 coated with Al-base (AlSi10Fe3) and Zn-base (ZnNi10) was conditioned by a pulse laser and by sandblasting to remove undesirable oxides and brittle phases. Oxides formed on coating surfaces counteract the wettability of welding filler during a welding or brazing process. Furthermore, welding and brazing joints of 22MnB5 coated with aluminum alloys failed along the brittle intermetallic phases in the coating under a low mechanical load. Treated 22MnB5 surfaces were analyzed microscopically, and the phase compositions were investigated by synchrotron diffraction measurements. It was found that brittle phases could be locally removed by laser ablation; however, high laser energies led to remelting and oxidation of the coating surface. In contrast, sandblasting homogenously removed oxides and brittle intermetallic phases. Surface-treated 22MnB5 steel sheets were joined to AA6016 aluminum sheets by laser welding, and the strength of the weldment was determined by tensile tests. The measured mechanical strength of the aluminum/steel joints was 210–230 MPa. Failure of the weldments under tensile loading occurred within the aluminum sheet, away from the steel surface/welding filler interface if brittle coating components and oxides were removed homogenously. © 2015 Published by Elsevier B.V.
1. Introduction High-strength steel 22MnB5 in the press-hardened condition is used in automotive bodies to comply with the safety requirements (high strength and stiffness) in safety-relevant structures. Aluminum materials can decrease the weight of cars in areas where lower mechanical properties can be tolerated. To utilize the advantages of both materials, suitable joining technologies are required in areas where both materials are used (e.g. underbody of the automotive bodywork). Most of the published investigations about thermal joining of steel/aluminum compounds focused on uncoated or galvanized mild steel sheets. For example, steel and aluminum sheets were joined by shock welding [1], friction welding [2], or laser welding [3]. Especially laser welding was found to be a suitable joining technology for different dissimilar welding operations [3–6]. However, less literature is available about thermal joining (welding/brazing) of coated, high-strength steel 22MnB5 to aluminum alloys. In the fusion welding process of steel and aluminum, generally only the aluminum alloy is melted due to its lower melting temperature. The steel surface remains solid and is wetted with the melted aluminum, which can thus be described as a welding (aluminum) — brazing (steel surface) process [7,8]. Thus, the ⁎ Corresponding author. E-mail address: [email protected] (M. Windmann).
http://dx.doi.org/10.1016/j.surfcoat.2015.11.037 0257-8972/© 2015 Published by Elsevier B.V.
achievable strength of coated 22MnB5/aluminum joints depends on the properties of the steel coating and the adhesion of liquid welding filler on the coated 22MnB5 surface. Sheets of 22MnB5 steel are usually formed and hardened in a presshardening process, in which a material strength of up to 1800 MPa can be achieved in complex formed body structures [9]. During press hardening, the steel sheets are austenitized at 880–950 °C and then transported into a tool where they are formed and quenched in one step [10]. To prevent strong oxidation of the steel sheets during austenitization, a protective Zn-, or Al-base coating is deposited on the steel surface prior to the press-hardening process [11]. During austenitization, distinctive diffusion processes occur at the interface between the steel substrate and the coating, leading to the formation of brittle intermetallic AlxFey phases (Al13Fe4, Al5Fe2, Al2Fe, AlFe) [12] or ZnxFey phases (Fe3Zn10) [13] in the coating. In addition, Al-rich or Znrich oxides of type M2O3 and MO are formed on the coating surface [8]. In the Al-rich coating, the phase of type Al5Fe2 possesses the highest volume fraction in the press-hardened coating and exhibits a low fracture toughness of 1 MPa m0.5 [14]. The formation of brittle intermetallic phases promotes crack initiation and propagation during forming in the press-hardening tools so that the coating microstructure possesses a network of cracks. The brittle Al coating counteracts the braze- and weldability of coated 22MnB5 owing to crack growth along the brittle intermetallic AlxFey phases [15]. For this reason, brittle intermetallic
154
M. Windmann et al. / Surface & Coatings Technology 285 (2016) 153–160
AlxFey phases can be transformed into more ductile phases by adaption of the press-hardening parameters to simultaneously increase the fracture toughness of the coating and decrease the crack density [12, 16]. As an alternative that does not require adaption of the process parameters, the aim of this study is to remove brittle intermetallic phases by means of a pulse laser or sandblasting. In Zn-base coatings, no cracks are formed in the press-hardening process [11]. During welding or brazing with cold-rolled and not additionally heat-treated Zn-coated mild steel sheets, the coating is completely melted and the liquid filler achieves good adhesion to the surface [8,17,18]. In contrast, during press-hardening, the Zn coating completely transforms into α-Fe and Fe3Zn10 [13]. Consequently, the former Zn-base coating is not melted in a brazing process. The melted filler has to be brazed on the oxidized surface of the Zn-base coating on the press-hardened steel. For this reason, the oxides on the Zn coating should be removed to improve the adhesion of a filler in a welding or brazing process. In this study, the surface of Al-base (AlSi10Fe3)- and Zn-base (ZnNi10)-coated and industrially press-hardened 22MnB5 steel was conditioned with a pulse laser and sandblasting to remove undesirable oxides and brittle phases, which should improve the adhesion of a filler on the coating surface. The treated Al-base- and Zn-base-coated 22MnB5 steel sheets were subsequently joined to aluminum alloy AA6016 by laser welding/brazing. Removing brittle coating components and oxides from the coating surfaces should lead to a good bonding strength between the coated steel and the filler, which was evaluated by tensile tests.
this purpose, the cross-section of each sample was ground with 54 and 18 μm abrasives and then polished with a 3 μm diamond suspension and finally with a ¼ μm SiO2 suspension. The chemical composition and oxide content on the coating surface was analyzed by EDS measurements using an acceleration voltage of 20 keV and a working distance of 8.5 mm. EDS measurements enabled to compare the oxygen content at the coating surfaces in the treated and initial condition and thereby discuss the removal of oxides.
2. Experimental
The phase composition of the treated coating surface was analyzed with synchrotron diffraction measurements (λ = 0.45919 Å) at the electron storage ring facility Delta at TU Dortmund (Germany). The synchrotron radiation beam was focused parallel to the specimen and into the treated coating, which was further tilted by an angle of 5°. Debye–Scherrer circle segments (140–155°) were conditioned and integrated with the program Fit2D (ESRF). Phase analysis was performed using the ICDD-JCDPS database PDF-2.
2.2. Surface treatment Surface treatment by laser ablation was carried out with the Nd:YAG laser Spectron SL900 and an Arges Racoon 16 scanner system equipped with f-theta lens (Fig. 1). This pulse laser source provided a maximum mean power of 16 W with a wavelength of 1064 nm. In these experiments, we used a spot diameter of 16 μm in the focal plane, a pulse duration of 100 ns, and a beam propagation velocity of 160 mm/s. The pulse frequency was set to 10 kHz, and the mean laser power was varied between 5 and 13 W. Argon was used as shielding gas. Sandblasting was performed manually using a Hessler Strahlboss 140 device with a nozzle diameter of 10 mm, an angle of incidence of 60°, and a distance of 100 mm to the coating surface. Silicon oxide (SiO2: 90–355 μm) was used as the abrasive. The sandblasting time was 10 s per steel sheet (50 × 120 mm) and the air pressure was set to 3 bar. 2.3. Phase analysis
2.1. Materials and microstructural analysis AlSi10Fe3- and ZnNi10-coated sheets of 22MnB5 steel with a size of 50 × 120 mm and a thickness of 1.5 mm were investigated in this study. The AlSi10Fe3 layer (thickness: 25–35 μm) was deposited on the 22MnB5 by hot dipping at a temperature of 675 °C. The ZnNi10 coating (thickness: 24–30 μm) was galvanically deposited on the steel surface. The chemical composition of 22MnB5 steel was measured by optical emission spark spectroscopy and is given in Table 1. In addition, the chemical composition of the AlSi10Fe3 and ZnNi10 coatings was determined in the solidified state by energy dispersive x-ray spectroscopy (EDS) measurements. The coated 22MnB5 sheets were welded with naturally aged (T4) AA6016 aluminum with a thickness of 1.5 mm (see Section 2.4). In addition, fillers of AlSi3Mn (Al-base coating) and ZnAl15 (Zn-base coating) with a wire diameter of 1.2 mm were used. The chemical composition of the AA6016 aluminum alloy and the fillers is also given in Table 1. Microstructural examinations of the cross-sections were performed by optical light microscopy (OM), and examinations of the coating surfaces were performed by scanning electron microscopy (SEM). For
2.4. Welding setup Welding experiments were performed to determine the influence of the surface treatment on the bonding behavior of the coated 22MnB5 steel in a welding/brazing process with aluminum. Thereby, the steel surface remained solid and was wetted with the melted aluminum. Surface-treated 22MnB5 metal sheets (50 × 120 mm) were joined with AA6016 aluminum alloy (50 × 150 mm) in an overlap configuration with AA6016 in the upper position (Fig. 2). Welding was performed using a Nd:YAG laser beam (Trumpf HL4006D — maximum power: 4 kW) and a power supply for inductive preheating (maximum power output of 10 kW). The wire feed angle was adjusted to 25° to the sheet plane. The wire feed of the filler was directed contrary to the welding direction. FontArgen F400 NH flux was applied to the treated
Table 1 Chemical composition of the materials measured by emission spark spectrometry and energy dispersive x-ray spectrometry. Chem. composition [mass%]
C
Si
Al
Mn
Mg
Ni
B
Fe
Zn
Substrate 22MnB5 AlSi10Fe coating ZnNi10 coating AA6016
0.234 – – –
0.289 10.230 – 1.047
0.034 Bal. – Bal.
1.258 – – 0.135
– – – 0.463
– – 10,780 –
0.002 – – –
bal. 2.160 2.080 0.249
– – Bal. –
Filler AlSi3Mn1 ZnAl152
– –
3.300 0.002
Bal. 14.620
1.100 ---
0.010 ---
– ---
– ---
0.190 0.004
– Bal.
1 2
Provided by Drahtwerke Elisental. Provided by Grillo.
M. Windmann et al. / Surface & Coatings Technology 285 (2016) 153–160
155
ZnNi10-coated 22MnB5 steel and the aluminum alloy AA6016 are presented. 3.1. Characterization of the AlSi10Fe3 and ZnNi10 coating after presshardening
Fig. 1. Setup of the laser ablation process.
steel surface and dried before welding. Argon was used as the shielding gas during welding. The welding parameters (Table 2) were chosen on the basis of prior investigations [19]. 2.5. Tensile tests To determine the strength of the welded joints with respect to the surface treatment, tensile tests were performed according to DIN EN ISO 6892. The locations where test specimens were taken are illustrated in Fig. 3. A Z100 Zwick/Roell machine was used for fracture testing of three specimens of each welded joint at a test speed of 0.5 mm/min. The load was measured with a load cell (measuring range: 100+0.01 −0.05 kN), and the elongation was measured with an extensometer (measuring range: 2.5+0.00 −0.03 mm). The measured values were digitized by TestXpert software and mapped as a strength/elongation curve. 3. Results and discussion The microstructure of the Al-base coating (AlSi10Fe3), the Zn-base coating (ZnNi10), and the 22MnB5 steel/coating interface in the presshardened condition is presented first. Subsequently, surface treatment by laser ablation and sandblasting is discussed. Finally, the results of laser welding experiments with the surface-treated AlSi10Fe3- or
Fig. 4a shows the microstructure of the AlSi10Fe3 coating in the press-hardened condition (TAUS = 920 °C, tAUS = 6 min, quenched in the stamping tool). The coating (thickness: 25–35 μm) is characterized by a layered structure and a high crack density. Phases of type AlFe and Al5Fe2 were formed in the press-hardening process due to strong diffusion of iron into the coating at the austenitization temperature of 920 °C. The formation of cracks is attributed to the low fracture toughness of the phase of type Al5Fe2 (KIC = 1 MPa m0.5 [14]), which led to crack initiation and propagation during cooling and forming in the press-hardening tool. The phase formation in the AlSi10Fe3 coating during austenitization in the press-hardening process is discussed in detail in [12]. A thin Al- and Si-rich oxide layer could be identified at the surface (Fig. 4c). An oxygen content of 24 at.% was measured at the surface by EDS. The formation of oxides on the surface of Alcoated iron-base substrates during heat treatment was presented in detail in [20]. Fig. 2b shows the microstructure of the ZnNi10 coating in the presshardened condition (thickness: 24–30 μm). The strong diffusion of iron into the ZnNi10 coating as well as the diffusion of Zn and Ni into the steel substrate during austenitization led to a complete transformation of the coating into α-Fe, in which 8 at.% Ni and 18 at.% Zn were dissolved, as measured by EDS. Phase formation in the ZnNi10 coating during press-hardening is summarized in [13]. In contrast to the AlSi10Fe3 coating, no cracks were formed in the ZnNi10 coating after cooling and quenching in the press-hardening tool. The surface of the ZnNi10 coating was characterized by a 3–5 μm thick oxide layer (Fig. 4d). An oxygen content of 33 at.% was measured at the surface by EDS. Lee et al. found that this layer consists of oxides of type Mn3O4 and ZnO [21]. In this study, the coated steels sheets were joined with an aluminum alloy in a laser welding/brazing process. For a first evaluation of the adhesion of welding fillers AlSi3Mn (for the AlSi10Fe3 coating) and ZnAl15 (for the ZnNi10 coating) on the coating surfaces, brazing experiments were performed. Brazing on the AlSi10Fe3 coating surface led to crack growth and complete separation of the brazed AlSi3Mn layer along the brittle intermetallic coating (Fig. 5a). During cooling after the brazing process, internal stresses were formed due to the different thermal coefficients of 22MnB5, AlSi3Mn, and the intermetallic phases in the coating. These stresses led to crack growth along the brittle intermetallic phases in the coating due to their low fracture toughness. Fig. 5b shows that the surface of the ZnNi10 coating was only poorly wetted with the ZnAl15 filler (contact angle N 90°). Adhesion of the welding filler was probably inhibited by the thick oxide layer on the coating surface. For both coating types, variation of the brazing parameters did not lead to a strong improvement of the wetting characteristic.
Fig. 2. Setup of the Nd:YAG laser-welding process [23].
156
M. Windmann et al. / Surface & Coatings Technology 285 (2016) 153–160
Table 2 Welding parameters. Coating
Preheating temperature [°C]
Laser energy [kJ/cm]
Welding velocity [m/min]
Velocity of filler [m/min]
AlSi10Fe ZnNi10
490 320
3.3 1.5
0.5 0.2
2.6 1.3
Brazing experiments indicated that the surface of Al- or Zn-coated 22MnB5 steel sheets have to be treated after press-hardening to increase their ability for a fusion welding/brazing process with aluminum alloys. Thus, the coating surfaces were treated by laser ablation and sandblasting to remove oxides and brittle coating components, thus improving the adhesion on the 22MnB5 surface.
coating components of the AlSi10Fe3 coating. On the one hand, if the laser energy was too low, it interacted insufficiently with the coating surface, thus only an incomplete removal of the coating occurred. If the laser energy was too high, it led to partial or full remelting of the coating, which is associated with a strong oxidation of the coating.
3.2. Laser ablation of AlSi10Fe3 and ZnNi10 coatings on 22MnB5
3.2.2. ZnNi10 coating Fig. 7 shows the cross-section and top view of the press-hardened ZnNi10 coating after pulsed laser ablation. Irrespective of the used laser power, the oxidized coating components could not be removed by spalling due to thermal stresses, as described previously for the AlSi10Fe3 coating. The ZnNi10 coating was characterized by bodycentered α-Fe in the press-hardened condition (see Section 3.1). As a result, the thermal expansion coefficients of the coating and the martensitic 22MnB5 steel did not significantly differ, for which reason cracks and delaminations due to thermoshock were not formed. Treatment with the pulse laser only led to the formation of new oxides on the surface (≥ 6.5 W). The amount of resulting oxides increased with increasing laser power and was always above the measured value in the press-hardened condition (33 at.%). As found for the AlSi coating, the ZnNi coating was also remelted at high laser powers (≥13 W). The high temperature in the melted coating led to evaporation of zinc and strong oxidation of the surface. A maximal oxygen content of 42 at.% was measured at the surface by EDS. Hence, the removal of oxidized ZnNi10 coating components without formation of new oxides was not possible by pulse laser ablation.
3.2.1. AlSi10Fe3 coating Fig. 6 shows the cross-section and the top view of the AlSi10Fe3 coating after laser treatment with different laser powers. A laser power of up to 5 W did not promote ablation of the coating components. A thin oxide layer could still be identified on the coating surface. The oxygen content at the surface did not differ significantly from the measured value in the press-hardened condition (24 at.%). Thus, the structure of the coating was identical to the press-hardened condition. At higher laser powers (≥ 6.5 W), coating particles of phases Al5Fe2 and AlFe were locally removed from the coating surface. The coating thickness decreased to 7–20 μm after surface treatment with a laser power of 6.5 W (initial condition: 24–30 μm). Material ablation of the coating surface can be explained by thermoshock. At the surface, the coating was rapidly heated by the focused laser and then cooled by heat transfer into the steel substrate. Thermal stresses formed as a result of differences in the thermal coefficients and in the Young's moduli of the intermetallic phases in the coating and the steel substrate. If the stresses exceed the material strength, this leads to crack growth and delamination of coating components across the brittle intermetallic phases. The amount of ablated brittle coating components increased with increasing laser power which resulted in thinner coating thicknesses after laser ablation. However, a coating layer with a thickness of 5–10 μm remained at the surface for all used laser powers. Furthermore, a laser power higher than 13 W led to remelting of the coating and a strong oxidation of the surface. EDS measurements confirmed the formation of Al-, Fe-, and Si-rich oxides. A high oxygen content of 41 at.% was measured at the coating surface. In this study, no laser power was identified that enabled complete removal of the brittle
Fig. 3. 22MnB5/AA6016 welded joints and positions of the specimens for tensile tests.
Fig. 4. Top row: OM micrographs of the cross-section of a) AlSi10Fe3- and b) ZnNi10-coated 22MnB5. Bottom row: SEM micrographs of the coating surfaces (top view) of a) AlSi10Fe3 and b) ZnNi10 after press-hardening.
M. Windmann et al. / Surface & Coatings Technology 285 (2016) 153–160
157
Fig. 5. a) OM micrograph of the cross-section of AlSi3Mn brazed on AlS10Fe3-coated 22MnB5, b) OM micrograph of the cross-section of ZnAl15 brazed on ZnNi10-coated 22MnB5.
3.3. Sandblasting of AlSi10Fe3 and ZnNi10 coatings on 22MnB5 3.3.1. AlSi10Fe3 coating As indicated by Fig. 8, the brittle AlSi10Fe3 coating could be almost completely removed by sandblasting. Synchrotron diffraction measurements confirmed that the surface of the AlSi10Fe3-coated 22MnB5 was free of intermetallic phases after sandblasting (Fig. 9). Oxides were also almost completely removed from the surface (4 at.% O). The surface was characterized by α-Fe and AlFe, in which up to 40 at.% aluminum was measured by EDS. During sandblasting, the brittle and crack-initiated coating was removed layer-wise as a result of a strong mechanical interaction between the accelerated grit and the brittle Al5Fe2 and AlFe coating components. Fig. 10 shows the structure of the coating surface after a short sandblasting time of 3 s. The sandblasting abrasives broke out the brittle coating components along the cracks formed during the press-hardening process (Fig. 4). A sandblasting time of
10 s was necessary (for the sandblasting setup and specimen size in this study, see Section 2) to completely remove all brittle coating components and to form a crack and oxide-free surface. 3.3.2. ZnNi10 coating Sandblasting also enabled removal of the oxide layer from the ZnNi10 surface (4 at.% O), as indicated by Figs. 8 and 9. In contrast to the AlSi10Fe3 coating, ZnNi10 coating particles were not removed. This can be explained by the absence of brittle phases and cracks in the ZnNi10 coating (α-Fe) after press-hardening. 3.4. Welding experiments and tensile testing As described above, only sandblasting enabled complete removal of brittle coating particles and oxides of Al-base- and Zn-base-coated HSLA steel surfaces. Thus, welding experiments (brazing on the steel side)
Fig. 6. Top row: OM micrographs of the cross-section of AlSi10Fe3-coated 22MnB5 after laser ablation with a mean power of a) 5 W, b) 6.5 W, and c) 13 W. Bottom row: SEM micrographs of the coating surfaces (top view) after laser ablation with a mean power of d) 5 W, e) 6.5 W, and f) 13 W.
158
M. Windmann et al. / Surface & Coatings Technology 285 (2016) 153–160
Fig. 7. Top row: OM micrographs of the cross-section of ZnNi10-coated 22MnB5 after laser ablation with a mean power of a) 5 W, b) 6.5 W, and c) 13 W. Bottom row: SEM micrographs of the coating surfaces (top view) after laser ablation with a mean power of d) 5 W, e) 6.5 W, and f) 13 W.
were performed using sandblasted steel sheets. Fig. 11 shows that the surface treatment promoted good adhesion of the filler to the 22MnB5 steel surface. At the interface of the sandblasted AlSi10Fe3 coating and
Fig. 8. Top row: OM micrographs of the cross-section of a) AlSi10Fe3- and b) ZnNi10-coated 22MnB5. Bottom row: SEM micrographs of the coating surfaces (top view) after sandblasting of c) AlSi10Fe3 and d) ZnNi10.
the AlSi3Mn filler, no cracks were formed, in contrast to the presshardened condition (Fig. 3). During the brazing process on the steel surface, diffusion of iron into the melted aluminum and aluminum into the steel substrate led to the formation of intermetallic Al–Fe phases at the steel/aluminum interface. The formation of brittle intermetallic Al–Fe phases promotes crack formation and propagation, accompanied by a brittle fracture of the welded joint during cooling if a critical thickness of 10 μm is exceeded [22]. However, the aluminum-saturated 22MnB5 surface (up to 40 at.% Al) inhibited strong aluminum diffusion into the steel substrate. The thickness of the intermetallic Al–Fe layer was below 10 μm along the length of the steel/aluminum interface. Phase
Fig. 9. Diffraction patterns of the AlSi10Fe3 and ZnNi10 coatings in the press-hardened and sandblasted condition.
M. Windmann et al. / Surface & Coatings Technology 285 (2016) 153–160
159
Fig. 12. Specimen of the steel/aluminum welding joints after tensile tests; laser welded/ brazed after sand-blasting of the coating surface.
22MnB5/filler interface, but rather in the aluminum AA6016 alloy in tensile tests. The tensile strength of all tested specimens was 210–230 MPa. This value meets the tensile strength of the Al-base material AA6016 in the T4 condition. Thus, sandblasting of the AlSi10Fe3 and ZnNi10 coating surfaces increased the bonding strength at the coating/filler interface above the tensile strength of the AA6016 alloy. 4. Conclusions Fig. 10. SEM micrograph of the AlSi10Fe3 coating surface (top view) after sandblasting for 3 s.
formation at the interface between 22MnB5 steel and AlSi3Mn filler was described in detail in [19]. In the tensile tests, the welded joints did not fail at the 22MnB5/filler interface, but rather in the aluminum AA6016 alloy (Fig. 12). The adhesion of the welding filler on the Zn-base coating surface was greatly improved by sandblasting (Fig. 11b) compared to the initial state (oxidized coating surface) after press-hardening (Fig. 5). As indicated by Fig. 11d, oxides were no longer found at the interface between the sandblasted ZnNi10 coating and the ZnAl15 filler. Metallurgical bonding was primarily accomplished by formation of intermetallic Al–Fe phases due to the high aluminum content in the ZnAl15 filler. For these intermetallic phases, a stoichiometry near to the phase of type Al5Fe2 was measured by EDS. Pure zinc precipitates were formed between the intermetallic Al–Fe phases. Cracks were not found along the steel/filler interface. The welded joints also did not fail at the
The surfaces of Al-base and Zn-base coatings on press-hardening steel 22MnB5 were treated by laser ablation and sandblasting to remove brittle coating components and oxides from the coating surface. This improved the adhesion of the welding fillers on the coated steel for the laser welding/brazing process with aluminum alloy AA6016. The following key results were found: - AlSi10Fe3- and ZnNi10-coated and press-hardened steel 22MnB5 had a poor wettability with the welding filler in the initial state - Laser ablation was not able to completely remove brittle coating components and oxides of the AlSi10Fe3 and ZnNi10 coating surfaces. Brittle coating particles could be locally removed from the coating surface. Use of high laser powers led to remelting and formation of new oxides on the coating surface. - Brittle coating components and oxides could be almost completely removed by sandblasting. - Removal of brittle coating components from the AlSi10Fe3 surface occurred layer-wise.
Fig. 11. a) OM micrograph of the cross-section of a laser beam welding/brazing joint between AlSi10Fe3-coated 22MnB5 and AA6016 (filler: AlSi3Mn). b) OM micrograph of the crosssection of a laser beam welding/brazing joint between ZnNi10-coated 22MnB5 and AA6016 (filler: ZnAl15). c) SEM micrograph of the interface between sandblasted 22MnB5 + AlSi10Fe3 and AlSi3Mn filler. d) SEM micrograph of the interface between sandblasted 22MnB5 + ZnNi10 and ZnAl15 filler.
160
M. Windmann et al. / Surface & Coatings Technology 285 (2016) 153–160
- The wettability of the sandblasted coating surfaces was greatly improved compared to the initial state after press-hardening. - Laser welding/brazing joints of sandblasted Al- or Zn-coated 22MnB5 with the aluminum alloy AA6016 did not fail at the interface between the coating surface and filler, but rather in the aluminum alloy in tensile tests (tensile strength: 210–230 MPa).
Acknowledgments We gratefully acknowledge financial support from the BMBF (Bundesministerium für Bildung und Forschung) within the project “Hybrides Fügen von Multimaterialsystemen für Kraftfahrzeuge” (03X3032). We thank our project partners Volkswagen AG, ThyssenKrupp Steel, Waldaschaff Automotive GmbH, Drahtwerke Elisental, Grillo and Steremat Induktion GmbH (all Germany) for their support with the specimens and the welding setup. In addition, we thank the Delta electron storage ring in Dortmund (Germany) for their great support and help with the synchrotron measurements. References [1] H. Date, Y. Sato, M. Naka, J. Mater. Sci. Lett. 12 (1993) 626–628. [2] M. Sahin, Int. J. Adv. Manuf. Technol. 41 (2009) 487–497. [3] M. Schimek, A. Springer, S. Kaierle, D. Kracht, V. Wesling, Phys. Procedia 39 (2012) 43–50.
[4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23]
M.J. Torkamany, F. Malek Ghaini, R. Poursalehi, Mater. Des. 53 (2014) 915–920. C. Tan, W. He, X. Gong, L. Li, J. Feng, Mater. Des. 78 (2015) 51–62. J. Sun, Q. Yan, W. Gao, J. Huang, Mater. Des. 83 (2015) 120–128. M.J. Zhang, G.Y. Chen, Y. Zhang, K.R. Wu, Mater. Des. 45 (2013) 24–30. G. Qin, Z. Lei, Y. Su, B. Fu, X. Meng, S. Lin, J. Mater. Process. Technol. 214 (2014) 2684–2692. T. Manzenreiter, M. Rosner, T. Kurz, G. Brugger, R. Kelsch, D. Hartmann, A. Sommer, BHM 157 (2012) 97–101. A. Barcellona, D. Palmeri, Metall. Mater. Trans. A 40 (2009) 1160–1174. D.W. Fan, B.C. De Cooman, Steel Res. Int. 83 (No.5) (2012) 412–433. M. Windmann, A. Röttger, W. Theisen, Surf. Coat. Technol. 226 (2013) 130–139. J. Kondratiuk, P. Kuhn, E. Labrenz, C. Bischoff, Surf. Coat. Technol. 205 (2011) 4141–4153. U. Köster, W. Liu, H. Liebertz, M. Michel, J. Non-Cryst. Solids 153/154 (1993) 446–452. C. Kim, M.J. Kang, Y.D. Park, Procedia Eng. 10 (2011) 2226–2231. M. Windmann, A. Röttger, W. Theisen, Surf. Coat. Technol. 246 (2014) 17–25. P. Peyre, G. Sierra, F. Deschaux-Beaume, D. Stuart, G. Fras, Mater. Sci. Eng. A 444 (2007) 327–338. C. Dharmendra, K.P. Rao, J. Wilden, S. Reich, Mater. Sci. Eng. A 528 (2011) 1497–1503. M. Windmann, H. Kügler, A. Röttger, W. Theisen, F. Vollertsen, J. Mater. Process. Technol. 217 (2015) 88–95. Y. Kitajima, S. Hayashi, T. Nishimoto, T. Narita, S. Ukai, Oxid. Met. 73 (2010) 375–388. C.W. Lee, W.S. Choi, Y.R. Cho, B.C. De Cooman, ISIJ Int. 54 (10) (2014) 2364–2368. C. Radscheit, Laserstrahlfügen von Aluminium und Stahl, BIAS Verlag, 1997. Reprinted from Journal of Materials Processing Technology, 217, M. Windmann, A. Röttger, H. Kügler, W. Theisen, F. Vollertsen, Laser beam welding of aluminum to Al-base coated high-strength steel 22MnB5/Experimental, p. 90, Copyright (2015), with permission from Elsevier