Currently, the strength of hot-press-forming (HPF) steel has been increased to produce lightweight vehicles, and in this study, the first arc welding investigation on a 2.0 GPa-strength HPF steel was conducted. HPF steel sheets with thicknesses of 1.1 mm were butt-welded by gas tungsten arc welding without a filler metal, with the heat input controlled in the range of 140–260 J/mm. The weld metal, coarse-grained heat-affected zone (HAZ), and fine-grained HAZ had martensitic microstructures and contained small amounts of bainite only under the highest heat input (260 J/mm). The softened zone was composed of an intercritical HAZ (ICHAZ) (with polygonal ferrite and martensite/bainite) and a subcritical HAZ (with tempered martensite). The ICHAZ softened the most during welding and fractured during tensile testing. The strengths of the arc welds were in the range of 1141–1264 MPa, which were 57–63% strength of the base metal. The ICHAZ softening was revealed to originate from carbon diffusion during the martensite to austenite transformation, which is unavoidable in arc welding and should be considered in the design of a welded structure.
Research Article
Mechanical and microstructural properties of autogenous arc welds of 2 GPa-strength hot-press-forming steel
https://doi.org/10.21203/rs.3.rs-2130179/v1
This work is licensed under a CC BY 4.0 License
Journal Publication
published 15 Jun, 2023
Read the published version in Journal of Materials Engineering and Performance →
Version 1
posted
You are reading this latest preprint version
Hot-press-forming steel
Arc welding
Heat-affected zone
Strength
Microstructure
Softening
Hot-press-forming (HPF) steel, also known as hot-forming steel or press-hardened steel, is a boron-alloyed high-strength steel with a full martensitic microstructure. Considering martensitic steels have limited formability, HPF steel sheets are produced as low-strength pearlite structures and formed above the austenitization temperature. During forming, they are simultaneously quenched in a water-cooled die, causing an austenite-to-martensite transformation. HPF steels with a tensile strength of 1.5 GPa, represented by 22MnB5 steel, have been successfully applied to automotive body components, such as pillars, bumper beams, and impact beams. Moreover, recently steel mill companies have introduced higher-grade HPF steels in the market [1–5].
As welding is the main assembly process for HPF steel sheets, the weldability of 1.5 GPa-grade HPF steels has been evaluated for various welding processes. These include resistance spot welding of overlap joints [6–9]; laser welding of overlap [10, 11], butt [12], and lap–fillet joints [13] and tailor-welded blanks [14–17]; arc welding of butt [18, 19] and lap–fillet joints [20]; and friction stir welding of butt [21] and overlap joints [22].
However, the literature on welding higher-grade HPF steel sheets is limited compared with that on 1.5 GPa-grade HPF steels. Fahlstrom et al. [23] evaluated the laser weldability of 1.8 GPa and 1.9 GPa-tensile-strength steels. Kang and Kim [24] presented the effects of surface oxides during laser welding of 1.8 GPa HPF steel, and Shi et al. [25] demonstrated laser-welded patchwork blanks. Lu et al. [26] conducted the first study on the laser weldability of 2.0 GPa HPF steel, and the authors [27, 28] investigated in detail the heat-affected zone (HAZ) softening and joint strength of laser welds. More recently, laser welds of 1.5/2.0 GPa dissimilar-strength HPF steels were analyzed [29]. However, according to our research, thus far, only laser welding has been investigated for HPF steels above 1.8 GPa.
The HAZ of an HPF-steel-weldment is softened during welding, whereas that of a mild-steel-weldment is hardened owing to the high cooling rate during welding. For partial or full martensitic steels, the degree of HAZ softening increases with increasing the tensile strength of the base metal (BM) [12, 21, 30]. In a previous study on 1.5 GPa-grade HPF steel [12], the tensile strength of a laser-welded specimen was reduced by 15% owing to HAZ softening, even by low-heat-input laser welding. As a relatively high input is applied to HPF steel during arc welding, a more severe loss in tensile strength is expected than in laser welding.
Along with resistance spot welding, arc welding is an important welding process in automotive body assembly. In this study, autogenous gas tungsten arc (GTA) welding was conducted on a 2.0 GPa-grade HPF steel. The mechanical and metallurgical characteristics of the welds were investigated with various heat inputs per unit length (HIs). The measured properties were compared with those of laser welds with lower HIs, and the microstructures of the weld metal (WM) and intercritical HAZ (ICHAZ) were examined based on continuous cooling transformation (CCT) diagrams and the HI.
The BM used was Docol® PHS CR 2000 supplied by SAAB AB [31], and its chemical composition is summarized in Table 1. After the HPF process, the BM had the tensile strength, hardness, and elongation of 2.08 GPa, 609 HV, and 3.36%, respectively, and had a full martensite microstructure. The details of the HPF processing and BM properties are available in another study [20].
| Docol® PHS CR 2000 | ||||||||
|---|---|---|---|---|---|---|---|---|
| C | Si | Mn | P | S | Cr | B | Nb | Ti |
| 0.346 | 0.25 | 1.20 | 0.011 | 0.001 | 0.14 | 0.0015 | 0.044 | 0.01 |
Welding specimens were machined to a width of 150 mm, length of 150 mm, and thickness of 1.1 mm. Figure 1 shows that the specimen configuration is a gapless butt joint, along which GTA welding is conducted. A 2% thoriated tungsten rod with a diameter of 2.4 mm was used as the electrode according to the AWS A5.12M/A5.12:2009 standard (specification for tungsten and tungsten–alloy electrodes for arc welding and cutting). The distance between the gas nozzle and the specimen was 5 mm, and the arc length was 3 mm. The work angle was 90° in the cross-sectional direction, and the travel angle was 10° in the forward direction. Argon shielding gas (99.9%) was supplied at a flow rate of 15 L/min.
The welding power source was operated in the direct current electrode negative mode. Table 2 lists the detailed welding conditions adopted to change the HI. The HI was varied from 140 J/mm to 260 J/mm, and the welding speed was fixed at 10 mm/s.
| Specimen No. | Welding current (A) | Welding voltage (V) | Welding speed (mm/s) | Heat input per unit length (J/mm) |
|---|---|---|---|---|
| 1 | 100 | 14 | 10 | 140 |
| 2 | 110 | 14.5 | 10 | 160 |
| 3 | 120 | 15 | 10 | 180 |
| 4 | 130 | 16 | 10 | 208 |
| 5 | 130 | 16 | 5 | 260 |
The mechanical and metallurgical properties of welded specimens were measured. The specimens were polished and etched with 3% natal solution (3 mL HNO3 and 100 mL ethanol). The etched transverse cross-sections were imaged using stereo microscopy, light optical microscopy (LOM), and field-emission scanning electron microscopy (FESEM) to study their microstructures. In the metallurgical analysis, the laser welding specimens used in the authors’ previous study [27] were compared with the arc welded specimens. In the comparative study, the laser power and welding speed were 2 kW and 7 m/min, respectively, and the HI was 17.1 J/mm, much lower than that of arc welding in this study. Details of the laser welding conditions are provided in a previous study [27].
Three square specimens with a width of 12 mm and a length of 100 mm were extracted from the welded specimens for each condition. Static tensile tests were conducted at a head speed of 1 mm/min. The micro-Vickers hardness profile was measured along the specimen centerline of the cross-section of each specimen. Hardness measurements were conducted at an indent interspacing of 0.22 mm and at 81 measuring points with a load and a hold time of 1.96 N and 10 s, respectively.
3.1 Macrosections and hardness profiles for various HIs
Table 3 reports the macrosectional dimensions of the GTA welds produced for various HIs. Each fusion zone is indicated by a yellow dotted line. Noticeably, the bead width increases with increasing HI. As the HI is increased from 140 to 260 J/mm, the bead width in the centerline increases from 3.0 to 4.4 mm. White bands are observed on the cross-sections outside the WM, indicating the location of the ICHAZ [10, 12, 27]. Fig. 2 shows that the hardness of the WM decreases with the increase in the HI. In contrast, the hardness of the supercritical HAZ adjacent to the WM is approximately 600 HV regardless of HI and comparable to the BM hardness of 609 HV. The hardness of the ICHAZ with respect to the HI shows a weak negative correlation with Pearson’s r value (−0.466). In addition, the distance between the ICHAZs increases from 6.2 to 9.2 mm on increasing the HI from 140 to 260 J/mm.
3.2 Microstructural behavior for various HIs
Fig. 3 shows the cross-sectional LOM images of typical welds, and the WM and the HAZ are noticeable. The detailed morphology of the welds was analyzed using FESEM, and the results are shown in Fig. 4. Figs. 3b and 4a show that the BM has a full martensitic microstructure with a needle-like lath morphology. Fig. 3e shows that the ICHAZ has a ferritic structure, whereas other locations, such as the coarse-grained HAZ (CGHAZ), fine-grained HAZ (FGHAZ), subcritical HAZ (SCHAZ), and BM, are composed of martensitic lath and/or tempered martensite.
No phase transformation is observed in the SCHAZ because the peak temperature is lower than the AC1 temperature; the tempered martensite, including carbides, precipitates primarily along the prior austenite grain and lath boundaries. Owing to the tempering of martensite, the sharpness of the martensite lath is reduced, whereas the packet shape remains (Figs. 3c and d). Within the SCHAZ, the locations closer to the BM (SCHAZ1 in Figs. 3c and 4b) show sharper martensitic laths and fewer tempered carbides than the locations closer to the ICHAZ (SCHAZ2 in Figs. 3d and 4c).
The ICHAZ presents polygonal grains in the LOM image (Fig. 3e); more specifically, it is comprised of polygonal ferrite and martensite/bainite (Fig. 4d). The full martensite of the BM is transformed into polygonal ferrite and austenite at the peak temperature between the AC1 and AC3 temperatures. Subsequently, the austenite is transformed into the nonequilibrium phase of martensite or bainite during cooling. The morphology of the ICHAZ in the GTA welds is significantly different from that in the laser welds [27], and the detailed kinetics are discussed in the next section.
As shown in Figs. 3e–g and 4d–4f, the FGHAZ, CGHAZ, and WM achieve full martensitic structures because their peak temperatures are higher than the AC3 temperature. The CGHAZs and FGHAZs of the HPF steel welds do not remarkably differ, whereas the CGHAZs of the carbon steel welds present a larger prior austenite grain size than the FGHAZs. In the case of HPF steel, the FGHAZ, CGHAZ, and WM regions can become fully austenitic during heating and transform into full martensite during cooling owing to the extremely high hardenability of the HPF steel, confirmed by the hardness profile shown in Fig. 2. In contrast, the ICHAZ and SCHAZ regions cannot reach the full austenization temperature (AC3) during heating, but transform into partial and tempered martensite, respectively, after welding. Consequently, the hardness values of the ICHAZ and the SCHAZ are relatively lower than that of the BM. The lowest hardness is observed in the ICHAZ, comprising a dual-phase (DP) structure of polygonal ferrite and martensite/bainite.
Only the ICHAZ presents an equilibrium phase (polygonal ferrite), accompanied by martensite and bainite as nonequilibrium phases. Therefore, the ICHAZ has a lower surface energy than the other zones [32] and has a relatively lower reaction with the etchant, resulting in the white bands at the cross-sections, as seen in Table 3.
3.3 Tensile properties for various HIs
The tensile strengths of the welds, as shown in Fig. 5, decrease linearly with increasing HI. Fig. 6 shows the fracture locations in the tensile testing, indicating that a tensile fracture occurs near the ICHAZ in all cases. This test is associated with the ICHAZ having the lowest hardness, as shown in the hardness profiles in Fig. 2. The tensile strength presents better linearity (Fig. 5) with the HI than the hardness (Fig. 2f). Pearson’s r value is −0.9568, much better than that between the HI and the hardness (0.466). The relatively low linearity between the HI and the hardness is caused by the spatial resolution of the hardness measurement in the ICHAZ. As shown in Fig. 2, because the ICHAZ is narrow, the hardness is measured at only one or two points within it.
A tensile fracture is initiated in the middle of the specimen, as shown in Figs. 6c, and propagates toward the edges. Simultaneous crack initiation at multiple locations is confirmed by craters, marked as red arrows in the middle (Fig. 6c). Moreover, shear lips and slight necking are observed at the edges with mid-to-edge crack propagation. Therefore, the fracture surface with a cup-and-cone shape has a ductile mode fracture (Fig. 6b). A dimple fracture is confirmed in the more detailed surfaces of the fracture (Fig. 7).
In this section, the microstructures of the WM and ICHAZ of GTA welding with higher HIs and laser welding with lower HIs are discussed. The WM, CGHAZ, and FGHAZ showed fully martensitic microstructures because the peak temperature exceeded the AC3 temperature, whereas the peak temperature of the ICHAZ lay between the AC1 and AC3 temperatures. In particular, the morphologies of the arc welds were compared with those of the laser welds with a much lower HI of 17.1 J/mm.
Figure 8 presents the microstructures of the WMs of laser welding with an HI of 17.1 J/mm and GTA welding with varying HIs between 160–260 J/mm. The WM morphologies of the laser and arc welds with the minimum HI (specimen No. 2) are similar despite a significant difference in the HI. As the HI increases in arc welding, the WM is primarily martensitic and has lower bainite-containing carbide precipitation (Fig. 8d). The increase in the HI decreases the cooling rate, which increases the time for carbon diffusion during the austenite-to-martensite transformation. At a magnification of ×10,000, the lath width of the martensite increases as the HI increases. The weakening of the martensite morphology is consistent with the decrease in the hardness with increasing the HI. (Fig. 2).
From the analytical solution for the temperature distribution of a thin sheet by a moving heat source [33], the cooling time from 800 °C to 500 °C can be calculated by the following equation [34]:
where q is the actual power of the heat source, v is the welding speed, λ is the thermal conductivity, ρ is the density, c is the specific heat, d is the thickness of the sheet, and T0 is the room temperature in degree Celsius. The actual power of the heat source can be calculated by multiplying the welding power and the welding efficiency, η.
The cooling rate for each case is calculated using the material properties and parameters listed in Table 4, and the calculated values are listed in Table 5. The mill maker recommends a cooling rate of at least 60 K/s for the hot-press-forming process to achieve full design hardness (Fig. 9) [31]. The cooling rates for Figs. 8a and 8b are higher than 60 K/S, explaining the identical WM morphology for both cases despite the large difference in the cooling rate. In the highest HI case, the cooling rate is reduced to 10.3 K/s. Therefore, bainite formation in Fig. 8d can be explained by the cooling rate and the CCT diagram (Fig. 9).
Table 4 Material properties and parameters for cooling rate calculation [35]
|
Parameter |
Value |
|
Welding efficiency, η |
0.5 |
|
Thermal conductivity, λ (W·m-1·K-1 ) |
45 |
|
Density, ρ (kg·m-3) |
7200 |
|
Specific heat, c (J·kg-1) |
703.4 |
|
Thickness of sheet, d (m) |
0.0012 |
|
Room temperature, T0 (°C) |
25 |
Table 5 Cooling rates calculated for various HIs
|
Specimen |
Heat input per unit length (J/mm) |
T8/5 (s) |
Cooling rate (K/s) |
|
Laser |
17.1 |
0.049 |
6100 |
|
Arc No.2 |
160 |
4.27 |
70.3 |
|
Arc No. 4 |
208 |
7.26 |
41.3 |
|
Arc No. 5 |
260 |
29.03 |
10.3 |
In the welds of partial or full martensitic steel with a strength >780 MPa, softening is inevitable in the ICHAZ [36]. Consequently, the width of the ICHAZ [37] and the amount of softening determine the strength of the welded joint and the fracture location during tensile testing [30].
The ICHAZ width was measured at five points along the depth direction for the laser welds and arc welding specimen no. 2, with the boxplot shown in Fig. 10. The ICHAZ width with GTA welding is 380 μm, approximately 13 times larger than that in laser welding (29 μm). In both cases, the tensile specimens fracture in the ICHAZ. In the laser welding of 800 MPa-grade DP steel [37], HAZ softening did not affect the fracture location and tensile strength when the HI is below 42 J/mm, despite the ICHAZ width being much larger than the ICHAZ width of the laser welds (29 μm), as shown in Fig. 10. Furthermore, in the case of 1180 MPa-grade transformation-induced plasticity (TRIP) steel, the HI criterion for determining the ICHAZ fracture is 80 J/mm [38]. The ICHAZ hardness of 2.0 GPa-grade HPF steel is similar to that of 800 MPa-grade DP steel and 1180 MPa-grade TRIP steel. However, BM fracture is difficult despite the narrow ICHAZ because the strength difference between the ICHAZ and the BM is larger than that of lower-grade steels.
Fig. 11 shows the ICHAZ microstructures for various HIs. Although the laser welds with an HI of 17.1 J/mm and arc welding specimen No. 2 with an HI of 160 J/mm show similar WM microstructures of full martensite (Fig. 8), the ICHAZ microstructures are clearly different from each other (Figs. 11a and c). The ICHAZs of the arc welds consist of polygonal ferrite, martensite, and bainite, whereas those of the laser welds include martensite and lath ferrite. The WM microstructure is determined using the cooling rate and the CCT diagram. In comparison, the ICHAZ microstructure is determined by the martensite to austenite transformation, in addition to the cooling rate. A small amount of bainite is observed in the ICHAZ of the arc welds, and the cooling rate is approximately 40 K/s (Fig. 9). The microstructure of the ferrite–martensite transformation in the arc ICHAZ is completely different from that of martensite in the BM, suggesting that sufficient carbon is diffused and the grains are restructured during the martensite to austenite transformation in the arc ICHAZ. The kinetics of this process will be discussed in a future study.
This study conducted the first investigation on an arc-welded HPF steel with a strength of 2.0 GPa. A detailed analysis of the mechanical and microstructural properties of the autogenous GTA welds produced by various HIs (140–260 J/mm) was performed and the following conclusions were drawn:
1. In the arc welds, the WM, CGHAZ, and FGHAZ had fully martensite microstructures, while the tempered ICHAZ and SCHAZ contained partial martensite with tempered martensite. The ICHAZ had the lowest hardness among the weldments, ranging between 277 HV and 344 HV under the HIs applied in this study.
2. The tensile strength of the arc welds decreased from 1264 to 1141 MPa as HI increased from 140 to 260 J/mm. Fracture occurred in the ICHAZ, the most softened zone, and it was initiated in the ductile mode at the mid-thickness of the specimen and propagated toward the edge.
3. The hardness values of the CGHAZ and FGHAZ were similar to that of the BM, while the WM had a slightly lower hardness than the BM. The WM primarily had a martensite structure, and a small amount of bainite formation was observed at the highest HI of 260 J/mm, as confirmed by the CCT diagram.
4. The ICHAZ of a GTA weld showed a DP structure of polygonal ferrite and martensite/bainite, unlike those of the laser welds composed of martensite and lath ferrite. The width of the arc ICHAZ (380 μm) was significantly larger than that of the laser ICHAZ (29 μm).
- H. Karbasian and A. E. Tekkaya, J. Mater. Process. Technol. 210, 2103–2118 (2010).
- M.-G. Lee and S.-J. Kim, Met. Mater. Int. 18, 425–432 (2012).
- H. Kim, D. Kim, K. Ahn, D. Yoo, H.-S. Son, G.-S. Kim and K. Chung, Met. Mater. Int. 21, 874–890 (2015).
- K.-G. Chin, C.-Y. Kang, J. Park and S. Lee, Met. Mater. Int. 24, 489–495 (2018).
- S.-M. Yun, H. Gwon, J. Oh and S.-J. Kim, Met. Mater. Int., Online (2022) 10.1007/s12540-021-01161-w
- M. J. Kang and C. H. Kim, Mater. Sci. Forum 695, 202–205 (2011).
- H. Fujimoto, M. Yasuyama, H. Ueda, R. Ueji and H. Fujii, Weld. Int. 31, 681–691 (2017).
- B. V. Feujofack Kemda, N. Barka, M. Jahazi and D. Osmani, Met. Mater. Int. 27, 1205–1223 (2021).
- H. You and C. Kim, J. Weld. Joining 39, 658–665 (2021).
- C. Kim, M. Kang and Y. Park, Proc. Eng. 10, 2226–2231 (2011).
- H. You, S. Kang, M. Y. Kang, S., S. Hyun and C. Kim, J. Weld. Joining 39, 36–44 (2021).
- C.-H. Kim, J.-K. Choi, M.-J. Kang and Y.-D. Park, J. Achiev. Mater. Manuf. Eng. 39, 79–86 (2010).
- M. Kang, C. Kim and J. Lee, J. Laser Appl. 24, 022004 (2012).
- Y.-G. An, C.-Y. Kang, Y.-S. Kim, C.-H. Kim and T.-K. Han, J. Weld. Joining 29, 90–98 (2011).
- M. Kang and C. Kim, J. Laser Appl. 26, 032007 (2014).
- M. Kang, C. Kim and S. Bae, Int. J. Automot. Technol. 16, 279–283 (2015).
- M. Kang, Y.-M. Kim and C. Kim, J. Mater. Process. Technol. 228, 137–144 (2016).
- G. K. Ahiale and Y.-J. Oh, Mater. Sci. Eng. A 597, 342–348 (2014).
- T. Pfeifer and M. S. Węglowski, Adv. Manuf. Sci. Technol. 38, 71–80 (2014).
- M. Kuo Gas Metal Arc Welding of Advanced High Strength Steels, AUTO/STEEL Partnership, Detroit, MI, USA (2019)
- H. You, M. Kang, S. Yi, S. Hyun and C. Kim, Appl. Sci. 11, 5728 (2021).
- M. Kang, J. Yoon and C. Kim, Int. J. Adv. Manuf. Technol. 106, 1671–1681 (2020).
- K. Fahlström, K.-A. Persson, J. K. Larsson and E. V. Ferrer, J. Laser Appl. 28, 022426 (2016).
- M. Kang and C. Kim, J. Laser Appl. 30, 032415 (2018).
- D. Shi, K. Watanabe, J. Naito, K. Funada and K. Yasui, Thin-Walled Struct. 170, 108523 (2022).
- L. Lu, Z. Liang, J. Yang, Q. Sun, T. Zhu and X. Wang, J. Mater. Res. Technol. 9, 13147–13152 (2020).
- K. Kim, N. Kang, M. Kang and C. Kim, Appl. Sci. 11, 5774 (2021).
- K. Kim, N. Kang, M. Kang and C. Kim, J. Laser Appl. 34, 012012 (2022).
- Z. Zhang, X. Wang, Q. Sun, B. Yang, L. Xiong and H. Liu, Opt. Laser Technol. 150, 107933 (2022).
- M. Kang, I.-H. Jeon, H. Han and C. Kim, Met. 8, 365 (2018).
- Press hardening steels (PHS) for complex shapes https://www.ssab.com/en/brands-and-products/docol/automotive-steel-grades/press-hardening-steel
- T. Hanamura, F. Yin and K. Nagai, ISIJ Int. 44, 610–617 (2004).
- D. Rosenthal, Trans. ASME 68, 849–866 (1946).
- K. Easterling, Introduction to the physical metallurgy of welding, Butterworth-Heinemann, Oxford, UK (1992).
- C. H. Kim, W. Zhang and T. DebRoy, J. Appl. Phys. 94, 2667–2679 (2003).
- J.-K. Choi, M.-J. Kang, J.-H. Cho and C.-H. Kim, J. Weld. Joining 27, 1–4 (2009).
- X.-N. Wang, Q. Sun, Z. Zheng and H.-S. Di, Mater. Sci. Eng. A 699, 18–25 (2017).
- S. Nam, H. W. Lee, C. Kim, C.-Y. Jung and Y.-M. Kim, J. Weld. Joining 39, 505–512 (2021).
Funding & Acknowledgements
This study was supported by the Technology Innovation Program Materials and Components Development Program (Grant No. 20011427), funded by the Korean Ministry of Trade, Industry, and Energy (MOTIE). We also acknowledge the financial & technical support provided by Korea Institute of Industrial Technology (Grant No. EO-22-050).
Conflict of interest
The authors declare that they have no conflict of interest.
Author Contributions
Conceptualization: Kwangsoo Kim, Cheolhee Kim; Methodology: Sanghoon Kang, Minjung Kang; Investigation: Kwangsoo Kim, Sanghoon Kang, Minjung Kang; Writing - original draft preparation: Kwangsoo Kim, Hanji Park; Writing - review and editing: Cheolhee Kim, Namhyun Kang; Funding acquisition: Minjung Kang, Namhyun Kang; Supervision: Namhyun Kang, Cheolhee Kim
Table 3 is available in the supplementary files section.