Open Access
Add to BibSonomy
Abstract
The effects of ambient pressure on the phenomena in laser welding of pure titanium were observed by X-ray transmission in situ apparatus and high speed camera. The penetration depth increased to 18 mm as the ambient pressure decreased to 0.1 kPa. The depth increment from 100 to 0.1 kPa is nearly 200%, far higher than that of mild steel (60%). Both the backward expansion at keyhole tip and the high-speed spatters could be suppressed by decreasing ambient pressure to 1 kPa or lower. The spatter number decreased at least 4 times as the ambient pressure decreased from 10 to 0.1 kPa. It could be deduced that the melt flow decelerated with decreasing ambient pressure. Relevant mechanisms were discussed by the metallic vapor ejection from keyhole and the melt flow types in molten pool.
© 2017 Optical Society of America
1. Introduction
Recent studies demonstrated that laser welding under subatmospheric pressure could increase weld penetration depth to a similar level with that of electron beam welding [1–7]. Vervaerde et al. [1] carried out CO2 laser welding of pure iron at subatmospheric pressure, and explained the effects of ambient pressure on plasma behavior and penetration depth. Katayama et al. studied laser welding of 304 stainless steel and A5052 aluminum alloy under subatmospheric pressure using 16 kW laser power [2–4]. The penetration depth of 43 mm for stainless steel and 42 mm for aluminum alloy were obtained at 0.1 kPa with the welding speed of 0.3 m min−1, far deeper than that at atmosphere. In fiber laser welding of mild steel under local subatmospheric pressure [8], Luo et al. found the similar phenomenon. Pang et al. explained why the penetration depth increment of laser vacuum welding decreased with increasing welding speed by numeric simulation [9]. Considering the heavy plates containing lead or special protection is not needed to shield X-ray for laser welding [3], which is necessary for electron beam welding, laser vacuum welding would be potential for modern industries.
Titanium (Ti) and its alloys have been widely used in industrial applications, especially the fabrication in aerospace because of high specific weight, excellent corrosion resistance, and high temperature performance [10]. Elmer et al studied laser vacuum welding of pure Ti firstly, and demonstrated the potential of this technique in reducing the porosity and increasing the penetration depth [11]. However, only one subatmospheric pressure was taken into account, which would be not sufficient for future fabrication in aerospace. This paper then aimed to explore the phenomena in laser vacuum welding of pure Ti in detail, which is interest of to deepen the understanding of laser-matter interaction at vacuum.
2. Experimental
The material used was commercial pure Ti. The specimen used was 25 mm in width, 50 mm in length and 20 mm in thickness. The setup was drawn in Fig. 1. Continuous wave disk laser with the maximum laser power of 16 kW and the wavelength of 1030 nm were used. The beam parameter products (BPP) was 8 mm*mrad. The focal distance was 1000 mm. A spot diameter at the focus position was about 500 μm.
During welding, the specimen was fixed in vacuum chamber made of acrylic, which was vacuumized by three rotary pumps. Argon (Ar) was supplied into the chamber as a shielding gas when the ambient pressure decreased to 10 Pa. The chamber pressure was adjusted with Ar flow rate. The welding was performed on bead-on-plate configuration. The laser beam was perpendicular to the specimen. At atmosphere, the shielding gas was at the speed of 0.6 g s−1 through a 16mm-diameter side nozzle declining with the angle of 45°. The welding parameters was listed in Table 1.
Table 1. Welding Parameters
A high-speed video camera was used at the framing rate of 10,000 frames s−1 to observe the behaviors of laser-induced plume and molten pool. A diode laser with the wavelength of 976 nm was used at the power of 30 W for illumination, and an interference filter of 974.5 nm was utilized. An X-ray transmission in situ apparatus was used to observe keyhole geometry at the framing rate of 250 frames s−1, which was presented in previous paper [12]. The specimen size for X-ray observation is shown in Fig. 1. The plasma ionization degree was observed by a spectrograph, which has a measurement area of 500μm-diameter at 2 mm height above specimen surface. A long wave pass filter of less than 750 nm was used to cut off the high order diffracted lights in the wavelength of 800 to 850 nm. All measured wavelengths were calibrated from the references of both a mercury lamp and a spectral irradiance standard lamp. The integration time for spectroscopic measurement was 50 ms.
3. Results and discussion
3.1 Weld shape
As shown in Fig. 2, the weld surfaces become smoother with decreasing ambient pressure. All weld surfaces are characterized by silvery white, but visible colors appear in the heat affected zone (HAZ) of the welds at 0.1 and 1 kPa. It is related to the oxygen (O) content in the chamber. The volume of the Ar supplied into the chamber at low pressure is less than that at high pressure. It results in a higher O content in the chamber at low pressure because the remaining air is the same for all the pressures. Thus, the HAZ of the weld at low pressure is easier to contact and absorb the O to form the oxides and change color. This phenomenon also can be demonstrated by EPMA (electron probe micro-analyzer) test, in which the O level on weld surface at 0.1 kPa is 1.5 times than that at 10 kPa.
As shown in Figs. 3 and 4, the weld cross-sections change from inverted triangle shape to nail shape gradually, and the penetration depth increases with decreasing ambient pressure. In addition, the depth increment is inversely proportional to welding speed. For the welding speed of 0.5 m min−1, the penetration depth increases from 6 to 18 mm when decreasing ambient pressure from 100 to 0.1 kPa, about 2 times deeper. For the welding speed of 3 m min−1, this increment is only 1 times. The interesting phenomenon is that the depth increment is much higher than that of steels. The depth increment of mild steel is only 60% deeper from 100 to 0.1 kPa under similar parameters of laser power 5 kW and welding speed 0.5 m min−1 [13,14], far lower than that of pure Ti (200%).
The increase of the penetration depth would be attributed to two reasons. Firstly, the Inverse Bremsstrahlung (IB) effect above the keyhole is suppressed by decreasing ambient pressure. Assuming that the plasma is an ideal gas with its pressure in equilibrium with the ambient pressure, p, the Eq. (1) between the electron density, Ne and the p can be obtained according to Saha equation [1,8].
where, Te is electron temperature. It is expected that the plasma density above keyhole is strongly decreased by decreasing ambient pressure. In fact, as shown in Fig. 5, the spectrogram of the plasma plume shows the spectral intensity of the main line (Ti II 370.44 nm) at 10 kPa is 4.17 × 104 counts, while those at 0.1 and 1 kPa cannot be detected. It well agrees with above theoretical derivation. This change suppresses the plasma shielding effect, transfers more energy to keyhole depths via Fresnel absorption, and thus increases the penetration depth.
Fig. 5 Effects of ambient pressure on spectral properties of plasma plume with welding speed of 0.5 m min−1
Secondly, the low evaporation temperature at low pressure decreases the energy per unit depth necessary to maintain the keyhole. Fabbro et al. gave the following Eq. (2) to computer this power per unit depth, q (in W m−1) of laser vacuum welding [15], which is necessary for moving a cylinder keyhole inside a material with initial temperature (T0).
where, r is keyhole radius, Ts is surface temperature of molten pool (evaporation temperature), v is welding speed, K and k are the heat conductivity and diffusivity coefficient, respectively. The K and k of pure Ti are about 22 W m−1 K−1 and 9.08 × 10−6 m−2 s, respectively [16]. The Ts of pure Ti at 100 and 0.1 kPa is 3258 and 2130 °C, respectively. Using current parameters of v = 0.5 m min−1, T0 = 20 °C and r = 0.25 mm, the calculation shows the q of pure Ti from 100 to 0.1 kPa reduces from 0.198 to 0.128 kW mm−1. Obviously, this reduction is beneficial to increase the penetration depth.In general, the IB effect for the short wavelength (1.06 μm) used in current study is negligible because the wavelength square dependence of the IB absorption coefficient makes it very small [15], although it needs to be considered in fiber laser welding of Al alloys [17]. It can be confirmed that the main reason to increase the penetration depth would be the reduction of q, rather than the IB suppression caused by the decrease of Ne.
The effect of welding speed on penetration depth was explained by Pang et al [9]. According to a modified Clausius–Clapeyron equation, they confirmed that the keyhole front temperature at high welding speed is high enough to cause the pressure applied on melt surface much higher than ambient pressure. It causes the penetration depth dependent of increased surface pressure rather than ambient pressure, and thus decreases the depth increment.
The phenomenon that the depth increment of pure Ti is higher than that of mild steel under the same conditions would be related to the differences in thermal conductivity coefficient. According to Eq. (2), the q is directly proportional to K and (Ts-T0). The K of mild steel (64 W m−1 K−1) is about 3 times higher than pure Ti. The Ts of mild steel is 1818 °C at 0.1 kPa, causing the (Ts-T0) of pure Ti and mild steel are almost similar. These data show that the q of pure Ti at 0.1 kPa is about one third of mild steel. This reduction in the q is so big that pure Ti gets a much deeper penetration depth than mild steel.
3.2 X-ray transmission in situ observation of keyhole geometry
As shown in Fig. 6, the keyhole tip tends to expand backward with increasing ambient pressure. The backward expansion is caused by the ejection of laser-induced plasma generated from keyhole front wall [12]. As mentioned above, the plasma density is directly proportional to ambient pressure. Thus, the high ambient pressure gets a bigger plasma density, and then forms a stronger plasma ejection to result in a more obvious backward expansion at keyhole tip. This backward expansion is usually related to the bubble formation caused by keyhole collapse [18], but no bubbles are observed in current study.
Fig. 6 Effects of ambient pressure on keyhole behaviors, observed by X-ray in situ apparatus, (a) welding speed of 0.5 m min−1, (b) welding speed of 1.0 m min−1.
The keyhole depth and the diameter at 20% depth is presented in Fig. 7, where 3 mm is added to get the real depth according to X-ray transmission specimen. The keyhole depth increases from 11.5 to 18.4 mm with decreasing ambient pressure from 10 to 0.1 kPa at the welding speed of 0.5 m min−1, while it increases from 9.7 to 14.8 mm at the welding speed of 1 m min−1. The keyhole diameter was at the range from 0.65 to 0.88 mm, and tends to decrease with increasing ambient pressure.
Fig. 7 Effects of ambient pressure on keyhole depth (a) and keyhole diameter at 20% depth (b), observed by X-ray in situ apparatus.
3.3 Molten pool and spatter behaviors
As shown in Figs. 8 and 9, both ambient pressure and welding speed have obvious effects on molten pool behavior and spatter distribution. A swelling phenomenon appears on the surface of molten pool, corresponding to spatter ejecting. The higher the swelling, the more the spatter. Decreasing either ambient pressure or welding speed decreases the swelling height, but the effect of welding speed seems bigger. The swelling is weak at all ambient pressures when the welding speed is 0.5 m min−1, while it is strong when the welding speed is 3 m min−1, even at 0.1 kPa. In addition, the interaction trace between laser beam and plasma plume becomes clear at the welding speed of 3 m min−1 and the ambient pressure of 10 kPa.
Fig. 8 Effects of ambient pressure on the behaviors of molten pool and plasma plume at welding speed 0.5 m min−1, (a) 0.1 kPa (b) 1 kPa, (c) 10 kPa
Fig. 9 Effects of ambient pressure on the behaviors of molten pool and plasma plume at welding speed 3 m min−1, (a) 0.1 kPa (b) 1 kPa, (c) 10 kPa
The spatter characterizations are drawn in Fig. 10 according to the data recorded by high speed camera. In terms of the ejecting speed, the spatters can be classified into two types, low-speed spatter with the speed lower than 10 m s−1 and the high-speed spatter with the speed higher than 10 m s−1. The welding at 10 kPa is characterized by a large amount of low-speed spatters and some high-speed spatters. However, when decreasing ambient pressure to 1 kPa or lower, the high-speed spatter is almost suppressed, and the number of low-speed spatter decreases dramatically. That is, the welding at this stage is only characterized by a small amount of low-speed spatters. In addition, the spatter sum decreases with decreasing ambient pressure. The spatter sum at 10 kPa is about 4 times higher than that at 1 kPa or lower when the welding speed is 0.5 m min−1, and is 7.5 times higher when the welding speed is 3 m min−1. The phenomena about the spatter and the swelling pool at the ambient pressure of 10 kPa and the welding speed of 3 m min−1 are similar with that observed in the welding of stainless steel at normal atmosphere [19], which is characterized by high swelling pool and large ejections.
Fig. 10 Effects of ambient pressure on spatter ejecting distribution at different welding speed, (a) 0.5 m min−1 (b) 1.0 m min−1, (c) 3.0 m min−1
Usually, the high-speed spatter is caused by instable molten pool, while the slow-speed spatter is caused by melt flow [20, 21]. Above spatter characteristics denote that the molten pool is stabilized, and the melt flow speed is decreased by decreasing ambient pressure to 1 kPa or lower. The molten pool stability is dependent on the metallic vapor ejection from keyhole. The stronger the ejection, the bigger the recoil force, and the more instable the molten pool. As discussed above, decreasing ambient pressure weakens the metallic vapor ejection by decreasing plasma density, and then stabilizes the molten pool and reduces the spatter number. On the other hand, Kawahito et al. claimed that there are two types of melt flow circulation in the upper and lower parts of the molten pool at the rear of the keyhole during laser welding, as shown in Fig. 11(a). The surface swelling of molten pool is caused by the upper flow circulation that is accelerated by upward aerodynamic drag force driven by metallic vapor ejection from keyhole [2]. During laser vacuum welding, this upper drag force decreases due to the weakened metallic vapor ejection. It decelerates the upper flow circulation, and decreases the swelling height. When increasing the ambient pressure to 10 kPa and the welding speed to 3 m min−1, the surface pressure applied on the melt surface becomes much higher than ambient pressure, so the molten pool is more like that at atmosphere. At this stage, as shown in Figs. 9(c) and 11(b), the upper melt flow becomes a one-way flow, and disperses as the spatters. Both of the changes decrease the process stability and increase the spatters.
Fig. 11 Schematic drawings of melt flow in molten pool of laser welding with 6 kW power, observed by Ref [18], (a) welding speed of 3 m min−1, (b) welding speed of 9 m min−1.
4. Conclusion
In order to deepen the understanding of laser-matter interaction at vacuum, the phenomena in laser welding of pure Ti were observed by X-ray transmission in situ apparatus and high speed camera. The following conclusions can be drawn.
- (1) The penetration depth is up to 18 mm at the laser power of 6 kW and the welding speed of 0.5 m min−1 when the ambient pressure decreases to 0.1 kPa. The depth increment from 100 to 0.1 kPa is 200%, far higher than that of mild steel (about 60%).
- (2) The backward expansion at keyhole tip is observed, but without bubble formation. The degree of the keyhole backward expansion decreases with decreasing ambient pressure. The keyhole diameter is at the range of 0.65 to 0.88 mm, and tends to decrease with increasing ambient pressure.
- (3) A swelling phenomenon corresponding to spatter ejecting is observed on the surface of molten pool, the degree of which decreases with decreasing either ambient pressure or welding speed. The high-speed spatter is almost totally suppressed when decreasing the ambient pressure to 1 kPa or lower. The sum of the spatter at 1 kPa or lower is at least 4 times less than that at 10 kPa.
- (4) The observed phenomena demonstrated that the molten pool was stabilized and the melt flow speed was decreased by decreasing ambient pressure to 1 kPa or lower. It was attributed to the weakened metallic vapor ejection and the decelerated melt flow in upper molten pool.
Funding
National Natural Science Foundation of China (No. 51475183, and 51275186).
References and links
1. A. Verwaerde, R. Fabbro, and G. Deshors, “Experimental study of continuous CO2 laser welding at subatmospheric pressures,” J. Appl. Phys. 78(5), 2981–2984 (1995). [CrossRef]
2. S. Katayama, Y. Kobayashi, M. Mizutani, and A. Matsunawa, “Effect of vacuum on penetration and defects in laser welding,” J. Laser Appl. 13(5), 187–192 (2001). [CrossRef]
3. S. Katayama, A. Yohei, M. Mizutani, and Y. Kawahito, “Development of deep penetration welding technology with high brightness laser under vacuum,” Phys. Procedia 12, 75–80 (2011). [CrossRef]
4. S. Katayama, Y. Kawahito, and M. Mizutani, “Latest progress in performance and understanding of laser welding,” Phys. Procedia 39, 8–16 (2012). [CrossRef]
5. U. Reisgen, S. Olschok, and S. Longerich, “Laser beam welding in vacuum–a process variation in comparison with electron beam welding,” in Proceedings of International Congress on Applications of Lasers and Electro-Optics, (Laser Institute of America, 2010), paper 1304.
6. S. Yang, J. Wang, B. Carlson, and J. Zhang, “Vacuum-assisted laser welding of zinc-coated steels in a gap-free lap joint configuration,” Weld. J. 92(7), 197–204 (2013).
7. U. Reisgen, S. Olschok, S. Jakobs, and M. Mücke, “Welding with the Laser Beam in Vacuum,” Laser Tech. J. 12(2), 42–46 (2015). [CrossRef]
8. Y. Luo, X. Tang, and F. Lu, “Experimental study on deep penetrated laser welding under local subatmospheric pressure,” Int. J. Adv. Manuf. Technol. 73(5), 699–706 (2014). [CrossRef]
9. S. Pang, K. Hirano, R. Fabbro, and T. Jiang, “Explanation of penetration depth variation during laser welding under variable ambient pressure,” J. Laser Appl. 27(2), 022007 (2015). [CrossRef]
10. R. Boyer, “An overview on the use of titanium in the aerospace industry,” Mater. Sci. Eng. A 213(1–2), 103–114 (1996). [CrossRef]
11. J. W. Elmer, J. Vaja, and H. D. Carlton, “The effect of reduced pressure on laser keyhole weld porosity and weld geometry in commercial pure titanium and nickel,” Weld. J. 95, 419–430 (2016).
12. Y. Kawahito, M. Mizutani, and S. Katayama, “Elucidation of high-power fibre laser welding phenomena of stainless steel and effect of factors on weld geometry,” J. Phys. D Appl. Phys. 40(19), 5854–5859 (2007). [CrossRef]
13. C. Börner, K. Dilger, V. Rominger, T. Harrer, T. Krüssel, and T. Löwer, “Influence of ambient pressure on spattering and weld seam quality in laser beam welding with the solid-state laser,” in Proceedings of International Congress on Applications of Lasers and Electro-Optics, (Laser Institute of America, 2011) paper 1604.
14. C. Börner, T. Krüssel, and K. Dilger, “Process characteristics of laser beam welding at reduced ambient pressure,” Proc. SPIE 8603, 86030M (2013). [CrossRef]
15. R. Fabbro, K. Hirano, and S. Pang, “Analysis of the physical processes occurring during deep penetration laser welding under reduced pressure,” J. Laser Appl. 28(2), 022427 (2016). [CrossRef]
16. F. Cardarelli, Materials Handbook: A Concise Desktop Reference (Springer Science & Business Media, 2014).
17. M. Gao, C. Chen, M. Hu, L. B. Guo, Z. M. Wang, and X. Y. Zeng, “Characteristics of plasma plume in fiber laser welding of aluminum alloy,” Appl. Surf. Sci. 326, 181–186 (2015). [CrossRef]
18. H. Nakamura, Y. Kawahito, K. Nishimoto, and S. Katayama, “Elucidation of melt flows and spatter formation mechanisms during high power laser welding of pure titanium,” J. Laser Appl. 27(3), 032012 (2015). [CrossRef]
19. Y. Kawahito, Y. Uemura, Y. Doi, M. Mizutani, K. Nishimoto, H. Kawakami, M. Tanaka, H. Fujii, K. Nakata, and S. Katayama, “Elucidation of the effect of welding speed on melt flows in high-brightness and high-power laser welding of stainless steel on basis of three-dimensional X-ray transmission in situ observation,” Weld. Int. 31(3), 206–213 (2017). [CrossRef]
20. A. F. H. Kaplan and J. Powell, “Spatter in laser welding,” J. Laser Appl. 23(3), 032005 (2011). [CrossRef]
21. M. Zhang, G. Chen, Y. Zhou, S. Li, and H. Deng, “Observation of spatter formation mechanisms in high-power fiber laser welding of thick plate,” Appl. Surf. Sci. 280, 868–875 (2013). [CrossRef]
