The inner characteristics of solidification and crack propagation in laser spot welding of four representative aluminum alloys, A1050, A2024, A5083 and A6061, were firstly observed with the X-ray phase contrast method. Keyhole disappeared within 1 ms after the laser was shut down. The solidification process finished in 2 ms for A1050, 3 ms for A2024, 5 ms for A5083, and 3 ms for A6061, respectively. Longitudinal view area of the molten pool decreased as the thermal conductivity increased, while the average solidification rate increased with increase of the thermal conductivity. Hot crack was observed to propagate from the bottom to the upper surface in the center of spot weld of A2024, A5083, and A6061, which was also the first in situ observation of crack during the welding process. Both the SEM, EBSD and Micro-X-ray computed tomography (CT) results validated that there was a crack propagation in the spot weld, and the mechanism for this crack formation was discussed. This paper provides a better understanding of solidification and crack formation in laser manufacturing.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
Fusion manufacturing, e.g., casting [1,2], additive manufacturing  and welding [4–8], plays a significant role in the production of metal part. The solidification process of molten metal will determine the microstructure [9,10], residual stress , thus final performance of the product . Crack was one of the most serious defects in fusion manufacturing, which was produced under the combined action of stress and microstructure [13–15]. Finally, obtaining a better understanding of solidification and crack formation in fusion manufacturing could contribute to improving the product performance.
Because of the excellent specific strength and corrosion resistance, aluminum has great potential in automobile and aircraft application [16,17]. Laser welding is a widely used fusion joining method for metal due to the high production efficiency and quality [18–20]. Various researchers have focused on the crack formation in welding [21–24]. Chelladurai et al  observed the characteristics of the crack at the cross section of the weld bead at different heat input, and estimated the solidification rate based on the theoretical equation, thus analyzed the relationship between solidification susceptibility and solidification rate. Liu et al  quenched the mushy zone of 5086 Al weld bead to observe the characteristics at the front of the solidification zone, and measured the solidification rate with thermal couples, which contributed to the understanding of the crack susceptibility. High speed camera was a widely used method to build the relationship the relationship between solidification rate and crack susceptibility. In this method, both the solidification rate and the crack propagation on the upper surface of the weld bead could be calculated by the measured continuous images obtained by high speed camera [27,28]. Witzendorff et al.  developed an experimental system to observe the weld spot solidification with high-speed cameras, calculating the relevant solidification parameters in pulsed laser welding of aluminum alloy. Bakir et al.  measured the strain during the welding process with a high-speed camera, and the critical strain required for solidification crack formation was both locally and globally calculated. These researches contributed to better understanding of the mechanism for crack formation. However, current observation of solidification and crack propagation process was mainly focused on the upper surface of the weld bead, the inner characteristics of solidification and crack formation in laser welding have rarely been investigated because of the shield of the surrounding metal. Recently, the X-ray phase-contrast method was reported to have the ability to in situ observe the keyhole, molten pool and porosity during laser welding clearly [30,31], which firstly provided a potential method to quantify the inner characteristics during solidification and crack formation.
Though X-ray has a great advantage in observing the dynamic formation process of porosity and other defects, this technology is still not widely used because of the high requirement for the source of X-ray beam, which could only be provided at several of the famous synchrotron radiation laboratories, e.g. Spring-8 in Japan , DIAMOND in the UK  and SSRL in the USA . The old publications from our lab were measured with the X-ray transmission method, and the observed distribution of the porosity was not very clear because of the limitation in the precision of the facility [34,35]. The experimental results from the DIAMOND source  and SSRL source  were focusing on laser additive manufacturing process, where the porosity and defect distribution were different with those in welding. In this study, the inner characteristics of solidification and crack propagation in laser spot welding of four representative aluminum alloys, A1050, A2024, A5083, and A6061, were observed with the X-ray phase contrast method with the facilities in Spring-8. Effects of material property on the solidification rate and the crack propagation rate were then discussed. Thus, the mechanism for the crack formation in laser spot welding was analyzed based on the observed crack characteristics by SEM, EBSD and CT. This paper provides a better understanding of solidification and crack formation in laser manufacturing.
2. Experimental procedures
With X-ray phase-contrast method, an image of inner characteristics of metal during and after welding is obtained based on the observation of interference patterns between diffracted and undiffracted waves . Since X-ray phase-contrast method puts forwards a high requirement for the quality of X-ray beam, the BL22XU beam line at SPring-8 (Super Photon ring-8 GeV, Hyogo, Japan)  was adopted in this experiment. Just similar with the experimental system in our previous publication , the relative position between the welding system and X-ray phase-contrast in situ observation system in the experiment is shown in Fig. 1. During the experiment, the aluminum alloy sample with dimensions of 70 mm × 30 mm × 2 mm was penetrated horizontally through 70 mm × 30 mm surface by X-ray, laser beam was applied to the top surface of the sample (70 mm × 2 mm surface). The aluminum alloy during and after welding was visualized by the X-ray radiation, and images were captured using a high-speed camera with the frequency of 1 kHz . Four representative aluminum alloys were used in the experiment, which were A1050, A2024, A5083 and A6061, respectively. Element composition of each alloy is listed in Table 1 . The laser power, welding time, and defocus distance were 500 W, 50 ms, and −1 mm, respectively. The laser beam was set with an inclination angle of 10° to reduce the reflection. A single-mode-fiber laser with a maximum output power of 500 W was used, and the focal length was 190 mm . The intensity profile of laser beam at the defocus of −1 mm is shown in the top left corner of Fig. 1. The laser beam size at this focus length was around 140 μm. A fan was used in this welding experiment to remove the laser-induced plume . Table 2 shows thermal conductivity of the aluminum alloys [31,36], and Table 3 shows the boiling temperature of elements in aluminum alloy. Because there are significant composition differences between these aluminum alloys, thermal conductivity varies with the grade. After welding experiment, the surface characteristics of the spot weld were observed by the SEM. Microstructure of the weld were observed by SEM and EBSD. The inner characteristics of crack were observed by the Micro-X-ray computed tomography (CT).
3. Results and discussion
Figure 2 shows the dynamic characteristics of molten pool during solidification and cooling observed by X-ray phase contrast method. Figure 2(a), Fig. 2(b), Fig. 2(c) and Fig. 2(d) shows the results for A1050, A2024, A5083 and A6061, respectively. Laser was turned off at t = 0 ms. A clear image of keyhole and molten pool was observed at this time, and keyhole was in the center of the molten pool. At t = 1 ms, keyhole disappeared, and the area of the molten pool decreased slightly. The solidification process finished in 2 ms for A1050, 3 ms for A2024, 5 ms for A5083, 3 ms for A6061, respectively.
Table 4 shows the measured solidification rate and molten pool area for welded aluminum alloys, and the relationships between molten pool area, solidification rate and the thermal conductivity are shown in Fig. 3. The molten pool area represents the area of the molten pool observed by the X-ray phase contrast method at the time of 0 ms, the solidification rate represents the reduction rate of the molten pool depth with the unit of mm/s, and the thermal conductivity of four aluminum alloys is from Table 2. As shown, the molten pool areas for A1050, A2024, A5083 and A6061 were 0.20, 0.31, 0.46 and 0.30 mm2, respectively. And the solidification rates for A1050, A2024, A5083 and A6061 were 277.0, 226.7, 190.1 and 231.2 mm/s, respectively. The molten pool area decreased as the thermal conductivity increased, the reason was that less heat accumulated at the molten pool at higher thermal conductivity. While the solidification rate increased with increase of the thermal conductivity, which is because heat could be dissipated away from the molten pool at higher thermal conductivity situation. Bubbles appeared in the molten pools of A1050, A5083 and A6061, and these bubbles retained in the spot weld with the style of porosity after solidification. As shown in Fig. 2(a) for A1050, bubbles generated from the bottom of the keyhole were trapped in the solid-liquid interface and finally formed porosity. As shown in Fig. 2(b), a narrow and short crack appeared at the bottom of the spot weld of A2024 at t = 4 ms. At t = 5 ms, this crack grew until reaching the upper surface of the spot weld, and the length of the crack also increased. As shown in Fig. 2(c), a crack between the porosity and the dent appeared suddenly in the spot weld of A5083 at t = 8 ms. In the spot weld with porosity, the crack appeared later than that without porosity. It seems that part of the solidification stress could release from the porosity, which delays the appearance of crack. As shown in Fig. 2(d), in A6061, the solid-liquid interface disappeared at t = 4 ms, hot crack occurred and dent was formed on the top of the spot weld. At t = 5 ms, hot cracks grew to the top of the spot weld. There was no crack in the spot weld of A1050, which indicated that A1050 had the lowest crack susceptibility among these four aluminum alloys. To the best of our knowledge, this was the first in situ observation of inner crack formation during welding. For each aluminum alloy, the dent was observed at the top of the spot weld after solidification. The molten metal shrank during solidification, which led to the appearance of dent at the upper surface of the spot weld.
Figure 4 shows the characteristics of spot weld observed by Micro-X-ray computed tomography (CT). The front view is shown in Fig. 4(a), and the right view is shown in Fig. 4(b). The experimental results showed that there was a crack surface in the center of the spot weld zone, which validated that the cracks progressed in two dimensions. Since laser spot welding is performed on a thin plate for X-ray phase contrast observation, it is presumed that the constraint stress in the longitudinal direction is large than that of transverse direction and the crack has propagated in one direction. Figure 5 shows the surface characteristics of the crack in the spot weld of A5083 observed by SEM. Figure 5(a) shows the overall view of the spot weld. Diameter of the spot weld was 1.25 mm, and a crack appeared in the center. Figure 5(b) shows the enlarged view of the crack zone b in Fig. 5(a). A crack with irregular shape went through the center of the spot weld, and this crack penetrated into the spot weld, which led to the formation of the crack surface observed by CT shown in Fig. 4. Figure 5(c) shows the characteristics of the crack. Dendrite structure was observed on the surface of the crack, which provided a proof for hot crack.
Figure 6 shows the microstructure of the spot weld of A5083 observed by SEM and EBSD. Figure 6(a) shows the overall front view of the cross-section observed by SEM. A dent was formed at the top of the spot weld because of the shrinkage during the solidification process. Crack was observed directly beneath the dent. Figure 6(b), 6(c), and 6(d) shows inverse pole figure in the overall view of the cross-section, the enlarged view of crack zone and the enlarged view of the spot weld boundary zone observed by EBSD, respectively. Large columnar grains with epitaxial growth from the boundary of the spot weld were clearly observed, and small equiaxed grains appeared in the final solidification zone of the spot weld. In fact, columnar grains were easily observed in the weld bead of laser welding because of the oriented solidification characteristics . The columnar grain growth shifted to equiaxed grain due to the rapid increase of the nucleation and growth rate of the equiaxed grain in the final solidification zone, which was caused by the decrease of the temperature gradient. Hot crack, which was observed in the equiaxed grain zone, started at the boundary of the columnar grain, and propagated towards the upper surface of the spot weld. Hot crack in the weld bead of laser welded aluminum has been reported by various previous researchers [24,29], the generation of solidification crack is influenced by both the solidification rate and crack susceptibility of the material [28,37].
Figure 7 shows the schematic for solidification and crack formation on the upper surface of the spot weld based on the results of A5083. Figure 7(a) shows the schematic of the solidification process. At this stage, the grain grew towards the center of the spot weld, which led to the formation of the columnar grain. Figure 7(b) shows the final status of the solidified spot weld. At the terminal stage of the solidification process, the nucleation speed of crystal in the center of the spot weld was higher than the growth speed of the columnar grain, which led to the formation of equiaxed grain in the center of the spot weld. Figure 7(c) shows the crack at the initial stage. During the cooling process, molten films owing to microscopic segregation were produced at the grain boundaries inside the welded part, which made the boundaries the weakest position in the spot weld. Besides, the spot weld zone had a trend to shrink during the cooling process, which would induced the inner stress in the spot weld zone. As a result, hot cracks propagated from inside the welded part to surface following the solidification process. As shown in Fig. 7(c), in the circle direction had the trend to separate the grain, which led to the propagation of the crack from inside the welded part to surface. Figure 7(d) shows the final status of the cracked spot weld. Under the action of the stress, the crack propagated along the grain boundaries. The crack propagation direction varies with the constraint stress due to the difference in welded joint and structure. Thus, the irregular crack was produced in the spot weld zone, which was observed by both the Micro-X-ray computed tomography (Fig. 4) and SEM (Fig. 5). Mass composition of Al in A1050 is higher than 99.5% (as shown in Table 1), and the content of other elements is very low, which reduced the molten films at the grain boundaries produced during the solidification process. This could be the reason for the low crack susceptibility of A1050 compared to those of A2024, A5083 and A6061.
Figure 8 indicates schematic illustration for hot crack propagation of right view based on the results of A5083(Fig. 4(b)). The hot crack propagated not three-dimensionally but two-dimensionally from above porosity to the upper surface in the center of spot weld due to the large constraint stress in the longitudinal direction. From X-ray phase contrast images, the propagation rate achieved higher than 200 mm/s, which is close to solidification rate. Moreover, the crack propagated along the boundaries of equiaxed grain from inside of spot weld to the surface. These first in situ observation results provide an in-depth understanding of crack formation in laser manufacturing.
The solidification and crack propagation process in laser spot welding were firstly observed with the X-ray imaging technology. The conclusions were as follows:
- 1) The area of the molten pool decreased as the thermal conductivity increased, while the average solidification rate increased with increase of the thermal conductivity. The solidification process finished in 2 ms for A1050, 3 ms for A2024, 5 ms for A5083, and 3 ms for A6061 respectively. The solidification rates for A1050, A2024, A5083 and A6061 were 277.0, 226.7, 190.1, and 231.2 mm/s, respectively.
- 2) Hot crack was observed to propagate from the bottom to the upper surface in the center of spot weld of A2024, A5083, and A6061. Both the SEM and Micro-X-ray computed tomography (CT) results validated that there was a crack surface in the spot weld. The crack propagation finished within 2 ms in the spot weld of A2024 and A6061, while within 1 ms in the spot weld of A5083. Porosity in the spot weld delayed the appearance of crack.
- 3) Irregular crack surface was observed by CT and SEM. Crack propagated along the grain boundary under the action of welding stress. A1050 had lower crack susceptibility than A2024, A5083, and A6061.
Grant-in-Aid for Scientific Research (C) (17K06818), Japan; Grant supported by the Ministry of Education, Culture, Sports, Science, and Technology (2014A-E19, 2014B-E17), Japan; JAEA’s beamline at SPring-8 (2014A3782, 2014B3781), Japan.
References and links
1. J. J. Baumberg, “Breaking the mould: Casting on the nanometre scale,” Nat. Mater. 5(1), 2–5 (2006). [CrossRef]
2. E. Guo, A. Phillion, B. Cai, S. Shuai, D. Kazantsev, T. Jing, and P. D. Lee, “Dendritic evolution during coarsening of Mg-Zn alloys via 4D synchrotron tomography,” Acta Mater. 123, 373–382 (2017). [CrossRef]
3. T. D. Roy, H. Wei, J. Zuback, T. Mukherjee, J. Elmer, J. Milewski, A. Beese, A. Wilson-Heid, A. De, and W. Zhang, “Additive manufacturing of metallic components–Process, structure and properties,” Prog. Mater. Sci. 92, 112–224 (2017).
6. J. Zou, N. Ha, R. Xiao, Q. Wu, and Q. Zhang, “Interaction between the laser beam and keyhole wall during high power fiber laser keyhole welding,” Opt. Express 25(15), 17650–17656 (2017). [CrossRef] [PubMed]
7. M. Zhang, G. Chen, Y. Zhou, and S. Li, “Direct observation of keyhole characteristics in deep penetration laser welding with a 10 kW fiber laser,” Opt. Express 21(17), 19997–20004 (2013). [CrossRef] [PubMed]
9. H. Wang, Y. Zhang, and S. Li, “Laser welding of laminated electrical steels,” J. Mater. Process. Technol. 230, 99–108 (2016). [CrossRef]
10. E. Guo, S. Shuai, D. Kazantsev, S. Karagadde, A. Phillion, T. Jing, W. Li, and P. D. Lee, “The influence of nanoparticles on dendritic grain growth in Mg alloys,” Acta Mater. 152, 127–137 (2018). [CrossRef]
11. Y. Zhang, H. Wang, K. Chen, and S. Li, “Comparison of laser and TIG welding of laminated electrical steels,” J. Mater. Process. Technol. 247, 55–63 (2017). [CrossRef]
12. H. Huang, N. Ma, T. Hashimoto, and H. Murakawa, “Welding deformation and residual stresses in arc welded lap joints by modified iterative analysis,” Sci. Technol. Weld. Join. 20(7), 571–577 (2015). [CrossRef]
13. W. Savage, E. Nippes, and J. Varsik, “Hot-Cracking Susceptibility of 3004 Aluminum,” Weld. J. 58, 1–9 (1979).
14. I. Farup, J.-M. Drezet, and M. Rappaz, “In situ observation of hot tearing formation in succinonitrile-acetone,” Acta Mater. 49(7), 1261–1269 (2001). [CrossRef]
15. S. Kou, “A criterion for cracking during solidification,” Acta Mater. 88, 366–374 (2015). [CrossRef]
17. J. R. Davis, “Aluminum and Aluminum Alloys,” ASM International, 351–416 (2013).
18. S. Katayama, H. Nagayama, M. Mizutani, and Y. Kawahito, “Fibre laser welding of aluminium alloy,” Weld. Int. 23(10), 744–752 (2009). [CrossRef]
20. L. Wang, M. Gao, C. Zhang, and X. Zeng, “Effect of beam oscillating pattern on weld characterization of laser welding of AA6061-T6 aluminum alloy,” Mater. Des. 108, 707–717 (2016). [CrossRef]
21. J. Wang, H.-P. Wang, X. Wang, H. Cui, and F. Lu, “Statistical analysis of process parameters to eliminate hot cracking of fiber laser welded aluminum alloy,” Opt. Laser Technol. 66, 15–21 (2015). [CrossRef]
22. G. Agarwal, M. Amirthalingam, S. Moon, R. Dippenaar, I. Richardson, and M. Hermans, “Experimental evidence of liquid feeding during solidification of a steel,” Scr. Mater. 146, 105–109 (2018). [CrossRef]
23. J. Liu and S. Kou, “Crack susceptibility of binary aluminum alloys during solidification,” Acta Mater. 110, 84–94 (2016). [CrossRef]
24. P. Von Witzendorff, S. Kaierle, O. Suttmann, and L. Overmeyer, “In situ observation of solidification conditions in pulsed laser welding of AL6082 aluminum alloys to evaluate their impact on hot cracking susceptibility,” Metall. Mater. Trans., A Phys. Metall. Mater. Sci. 46(4), 1678–1688 (2015). [CrossRef]
25. J. Liu, H. P. Duarte, and S. Kou, “Evidence of back diffusion reducing cracking during solidification,” Acta Mater. 122, 47–59 (2017). [CrossRef]
26. A. Chelladurai, K. Gopal, S. Murugan, S. Albert, S. Venugopal, and T. Jayakumar, “Effect of energy transfer modes on solidification cracking in pulsed laser welding,” Sci. Technol. Weld. Join. 20(7), 578–584 (2015). [CrossRef]
27. P. von Witzendorff, S. Kaierle, O. Suttmann, and L. Overmeyer, “Monitoring of solidification crack propagation mechanism in pulsed laser welding of 6082 aluminum,” in High-Power Laser Materials Processing: Lasers, Beam Delivery, Diagnostics, and Applications V (International Society for Optics and Photonics, 2016), p. 97410H.
28. M. Kang, H. N. Han, and C. Kim, “Microstructure and Solidification Crack Susceptibility of Al 6014 Molten Alloy Subjected to a Spatially Oscillated Laser Beam,” Materials (Basel) 11(4), 648 (2018). [CrossRef] [PubMed]
29. N. Bakir, A. Gumenyuk, and M. Rethmeier, “Investigation of solidification cracking susceptibility during laser beam welding using an in-situ observation technique,” Sci. Technol. Weld. Join. 23, 1–7 (2017).
30. M. Miyagi, Y. Kawahito, H. Kawakami, and T. Shoubu, “Dynamics of solid-liquid interface and porosity formation determined through x-ray phase-contrast in laser welding of pure Al,” J. Mater. Process. Technol. 250, 9–15 (2017). [CrossRef]
31. Y. Kawahito and H. Wang, “In-situ observation of gap filling in laser butt welding,” Scr. Mater. 154, 73–77 (2018). [CrossRef]
32. C. L. A. Leung, S. Marussi, R. C. Atwood, M. Towrie, P. J. Withers, and P. D. Lee, “In situ X-ray imaging of defect and molten pool dynamics in laser additive manufacturing,” Nat. Commun. 9(1), 1355 (2018). [CrossRef] [PubMed]
33. C. Zhao, K. Fezzaa, R. W. Cunningham, H. Wen, F. De Carlo, L. Chen, A. D. Rollett, and T. Sun, “Real-time monitoring of laser powder bed fusion process using high-speed X-ray imaging and diffraction,” Sci. Rep. 7(1), 3602 (2017). [CrossRef] [PubMed]
34. N. Seto, S. Katayama, and A. Matsunawa, “Porosity formation mechanism and suppression procedure in laser welding of aluminium alloys,” Weld. Int. 15(3), 191–202 (2001). [CrossRef]
35. A. Matsunawa, M. Mizutani, S. Katayama, and N. Seto, “Porosity formation mechanism and its prevention in laser welding,” Weld. Int. 17(6), 431–437 (2003). [CrossRef]
36. A. M. Handbook, “Properties of wrought aluminum and aluminum alloys,” ASM International Hand Book Committee 2, 62–122 (1990).
37. C. Zhang, H. Zhang, L. Wang, M. Gao, and X. Zeng, “Microcracking and mechanical properties in laser-arc hybrid welding of wrought Al-6Cu aluminum alloy,” Metall. Mater. Trans., in press (2018).