Abstract
Keyhole tungsten inert gas (K-TIG) welding is a variant of TIG welding, which can largely improve the weld penetration depth by forming keyholes inside the molten pool during welding. However, K-TIG welding is generally considered unsuitable for aluminum alloys due to their high thermal conductivity. A novel double-pulsed variable polarity TIG (DP-VPTIG) welding process was employed and the stable full penetration keyhole welding of 7 mm-thick AA2219 aluminum alloy was achieved. Keyhole dynamic evolution for DP-VPTIG was investigated based on visual sensing technology. Results indicate that in low-pulsed peak stage of DP-VPTIG process, the keyhole forms under the dominant role of the downward arc pressure against the upward surface tension and hydrostatic pressure acting on the surface of the molten pool, while the keyhole is closed as the upward surface tension and hydrostatic pressure become the dominant role in low-pulsed base stage. The periodic variation of low-frequency pulse in DP-VPTIG process stimulates a periodic keyhole behavior of “opening” and “closing” in the molten pool. The formation of keyhole is beneficial to the increase of weld penetration depth as the arc moves downwards along the keyhole and directly heats the solid metal under the molten pool. The keyhole size decreases with the increase of low-pulsed frequency.
Science Press
Aluminum alloys are widely used in aerospace, automobile and shipbuilding industries due to their excellent fracture toughness, high specific strength, good corrosion resistance and processing properties[1-3]. Welding processing is one of the most important manufacturing methods for the production of aluminium alloys, among which variable polarity tungsten inert gas welding (VPTIG) is extensively adopted owing to its cathode cleaning effect and stable welding process[4,5]. However, the weld penetration depth of TIG welding is shallow, and the arc energy density as well as the welding efficiency are relatively low attributed to its free-diverging arc[6]. The ways to enhance the welding efficiency of TIG mainly include the use of mixed shielding gases, the variation of conventional TIG process and the application of new processes[7]. Keyhole mode TIG, referred as K-TIG, is considered as an effective way to increase the penetration depth[8]. The keyhole during K-TIG welding process is mainly formed by a high current value, normally exceeding 300 A, which is high enough for the arc pressure to penetrate through the liquid molten pool[9]. This process is particularly applied in the joining of mid-thick plates, generally in the range from 6 mm to 12 mm, without edge preparation and additional filler materials[10]. Therefore, the welding efficiency can be largely improved compared with conventional TIG welding process.
K-TIG welding has been successfully applied in the joining of Ti-6Al-4V titanium alloys and commercially pure titanium with thickness up to 13 mm by Lathabai et al[11], 6.35 mm commercially pure zirconium by Lathabai et al[10], mild steels in the range from 6 mm to 12 mm by Lohse et al[12], 6.35 mm carbon steels by Olivares et al[13], 10 mm AISI 316L stainless steels by Feng et al[14], 5.5 mm C-Mn steel by Liu et al[15], and 10.8 mm S32101 duplex stainless steels by Cui et al[16]. However, the materials available for K-TIG welding are mainly low thermal conductivity metals. There are almost no literatures reporting about the K-TIG for aluminium alloys. As Lathabai et al[10] stated, the critical factors for K-TIG are surface tension and thermal conductivity. As to aluminum alloys, the surface tension is quite low and the thermal conductivity is relatively high. When the heat input is increased through simply increasing the arc current, the volume of the weld pool will be greatly enlarged due to the high thermal conductivity, so it is difficult to stabilize the weld pool with a low surface tension when keyhole is formed inside the molten pool.
To improve the stability of K-TIG welding process, a large amount of research works have been carried out. Backing jet flow argon gas was employed in the keyhole welding of C-Mn steel[15]. Experimental results show that this process is beneficial to enhancing the stability of K-TIG welding process. High frequency current (38.6 kHz) was introduced in the K-TIG welding of Q345 steel[17]. It is proved that the current window is significantly expanded and the threshold current is largely reduced, compared with those of the constant current keyhole welding process. The low frequency pulsed current (1~1.5 Hz) is proved to be useful to effectively stabilize the keyhole welding of 304 stainless steel with a wider current window[18]. In general, the current waveform modulation is an effective way to stabilize the K-TIG welding process[19]. In this study, in order to realize the stable keyhole welding of aluminum alloys, the high-frequency current (20~80 kHz) and low-frequency current (0.5~5 Hz) were simultaneously modulated into the conventional variable polarity current, referred as DP-VPTIG. The evolution process of keyhole dynamic behavior in 7 mm AA2219 aluminum alloy was investigated by a CCD-based visual sensing system. The results provide a solid foundation for the application of K-TIG welding process in aluminum alloys.
The experimental platform of the welding system is dis-played in Fig.1. This system mainly consists of a self-developed DP-VPTIG welding power supply, a screw motion mechanism, a welding robot, a welding torch and its water cooling system, shielding gas and real-time image acquisition system. The acquisition system was primarily composed of a CCD-based area-array camera (BASLER avA1000-120 km) with the sampling rate up to 120 frames per second, a data acquisition card and a computer controlling system. The CCD camera was equipped with an adjustable focal length indu-strial lens (Computer M5018-MP2). In order to filter out the interference caused by the strong arc light, a narrow-band interference filter with a central wavelength of 1064 nm and a bandwidth of 25 nm was installed on the lens. The camera was fixed behind the welding torch in the welding direction and maintained a certain downward tilt angle. It focused on the surface of the work-piece directly below the electrode. During the welding process, the welding torch remained stationary while the work-piece moved with the screw motion mechanism.
Fig.1 Experimental platform of the welding system
The schematic diagram of the current waveform of DP-VPTIG is shown in Fig.2. The conventional variable polarity current was modulated into 0.5~5 Hz ranged low-frequency pulse and high-frequency pulse, in the range of 20 kHz to 80 kHz, superposing in the positive stage. TH is the period of variable polarity current, TL is the period of low-frequency current, tHn is the duration of negative stage, tHp is the duration of positive stage, tb is the duration of low-frequency pulse base stage, tp is the duration of low-frequency pulse peak stage, Ibn is the negative current during tb, Ibp is the positive current during tb, Ipn is the negative current during tp, Ipp is the positive current during tp.
Fig.2 Schematic illustration of DP-VPTIG current waveform
The base metal employed was AA2219 aluminium alloy plate. The work-piece was machined into the dimension of 150 mm×60 mm×7 mm (length×width×thickness). The shielding gas was 99.99% pure argon, and the electrode adopted was cerium tungsten, with 4 mm in diameter and a 30 degree cone angle in the front tip. Since the effect of high-frequency pulsed current on arc behaviour and fluid flow of weld pool have been investigated, it has been proved that the high-frequency pulse can largely constrict the welding arc[20,21], and effectively enhance the weld penetration[22,23]. Therefore, the parameters of high-frequency pulse were set as constant values with the purpose of improving the arc penetrating ability at a relatively low heat input. This study mainly focused on the effect of low-frequency pulse on the keyhole dynamic evolution process. The parameters used were: high-pulsed frequency of 20 kHz, duty cycle of 50%, amplitude of 80 A; TH=100 Hz, tHp:tHn=8:2; TL=0.5 Hz, tp:tb=50%, Ipp/Ipn=360 A, Ibp/Ibn=120 A; shielding gas flow rate of 15 L·min-1, welding speed of 170 mm·min-1, the electrode tip to work-piece distance of 3 mm. The parameters selected above have been experimentally proven to successfully achieve a stable fully penetrated keyhole welding process.
Fig.3 shows the image of weld pool and keyhole entrance captured by the camera. The sampling rate was 60 frames per second, and the resolution was 340×340. In order to quantitatively analyze the dynamic evolution of the keyhole behaviour, the weld pool width Dw and keyhole entrance size Dk were defined as feature sizes of weld pool and keyhole, respectively. A steel ruler was used as the reference size to determine the actual sizes of Dw and Dk. To avoid the measuring errors and ensure the repeatability, three images taken from different periods were used for the calculation of average value.
Fig.3 Measurement of weld pool width and keyhole size
2.1 Keyhole dynamic evolution process
Fig.4 shows the images captured in the low-pulsed peak stage tp and corresponding weld pool width (Dw) and keyhole size (Dk). It can be seen that both the weld pool width and keyhole size exhibit a trend of increasing first and then gradually becoming stable in the duration of tp. As the heat input is high in tp due to the high current value, the weld pool width Dw monotonically increases from 8.6 mm at the beginning of tp (100 ms) to 13.4 mm at the end of tp (1000 ms). At the time of 100 ms, there is almost no obvious depression on the weld pool surface, meaning that keyhole does not form at the beginning of tp. The weld pool surface is then depressed downwards along the thickness, accompanied with the liquid molten metal pushed to the rear and edge of the weld pool. Simultaneously, a keyhole forms in the center of the weld pool, just below the welding arc. The keyhole size Dk continuously increases from 5.4 mm to 6.8 mm with increasing pulse peak time from 200 ms to 400 ms. Then, the keyhole size gradually becomes stable at around 7 mm.
Fig.4 Keyhole appearance evolution process in low-pulsed peak stage tp (a); weld pool width Dw and keyhole size Dk (b)
The images and corresponding weld pool width (Dw) and keyhole size (Dk) in low-pulsed base stage tb are illustrated in Fig.5. As the welding current switches from pulse peak to pulse base stage, the arc energy is significantly reduced owing to the dramatic reduction in the current value. The arc optical radiation is thus greatly weakened as revealed in the images. Consequently, the heat input to the base metal is drastically decreased, causing the molten pool to cool down and solidify from the rear and edge to the centre. Correspondingly, the weld width Dw decreases continuously from 11.2 mm at the beginning of tb (1100 ms) to 6.9 mm at the end of tb (2000 ms). Meanwhile, the depression of the weld pool surface decreases as the welding arc moves upwards along the thickness. The liquid molten metal flows back from the rear and edges of the weld pool, and fills the keyhole. From 1300 ms, the keyhole is completely closed (Dk=0).
Fig.5 Keyhole appearance evolution process in low-pulsed base stage tb (a); weld pool width Dw and keyhole size Dk (b)
The forces acting on the surface of the weld pool is the key factor in the evolution of keyhole[19]. To deeply understand the keyhole dynamic behavior in DP-VPTIG welding process, it is necessary to investigate the effect of the forces on the keyhole. As seen in Fig.6, the arc pressure Pa-z, hydrostatic pressure Ph and surface tension Ps are the main forces that determine the behavior of keyhole. The arc pressure Pa-z is a downward force that tends to induce the depression of the weld pool surface, while the hydrostatic pressure Ph and the surface tension Ps are upward forces against Pa-z. To maintain the stable existence of the keyhole, the downward force Pa-z should be balanced with the upward forces Ps plus Ph. The equilibrium equation of the three forces can be given as:
Fig.6 Force analysis at the bottom of keyhole
Pa-z can be represented by Eq.(2):
where μ is the permeability of the arc atmosphere, I is the welding current, RE is the radius of the arc near the electrode, and RB is the radius of the arc near the work-piece. From the equation, Pa-z is proportional to the square of the welding current.
Ph can be expressed as:
where ρ is the density of the liquid metal, g is the gravitational acceleration, and z is the depth to the entrance of the keyhole. It is obvious that Ph is proportional to the keyhole depth.
The surface tension Ps follows Young-Laplace equation expressed as
where R is the radius of curvature, and γ is the coefficient of surface tension.
During DP-VPTIG welding process, in low-pulsed peak stage tp, the welding current value is relatively high. The downward force Pa-z is higher than the upward forces Ps plus Ph, so that the downward force plays the dominant role in determining the keyhole dynamic in this stage. The keyhole forms as the weld pool surface moves downwards, and its depth keeps increasing with the duration of tp. Meanwhile, Ps and Ph both increase with the increase of keyhole depth according to Eq.(3) and Eq.(4). As the sum of upward forces Ps+Ph reaches the value equal to Pa-z, Eq.(1) is then established. The keyhole in the weld pool reaches a quasi-stable state in tp. When the current alters to low-pulsed base stage tb, Pa-z will decrease greatly caused by the dramatic reduction of the current value. Therefore, the upward forces Ps and Ph become the dominant role in determining the keyhole dynamic. Therefore, the bottom surface of keyhole moves upwards, accompanied by the decrease of weld pool surface depression. As keyhole depth decreases, the upward forces Ps and Ph both decrease. When the sum of upward forces Ps and Ph is equal to the downward force Pa-z, Eq.(1) is established again. Then, the keyhole behavior reaches another quasi-stable state in tb.
With tp providing the sufficient arc force to penetrate through the weld pool, and tb significantly reducing the heat input, the stable keyhole mode welding of aluminium alloy is successfully achieved. The periodic variation in low-frequency pulsed current stimulates a periodic keyhole behavior of “opening” and “closing” inside the weld pool.
2.2 Effect of keyhole on weld penetration
According to the force analysis, the low-pulsed peak current Ipp plays a critical role in determining the formation of keyhole through affecting the maximum arc force during tp. To investigate the effect of keyhole on weld penetration, trials were carried out with Ipp ranging from 210 A to 360 A. As observed in the above section, the weld pool and keyhole appearance at 1000 ms are sustained in a stable state in tp. Here, the keyhole images were all captured at 1000 ms with different Ipp.
Fig.7 reveals the appearances of keyhole, the cross-section profile of the weld bead and corresponding characteristic sizes with increasing Ipp. As Ipp increases from 210 A to 360 A, the weld pool width Dw increases monotonously from 8 mm to 13.5 mm due to a higher heat input. Almost no keyhole forms when Ipp is less than 260 A. This is mainly because the downward arc force generated by this current level is insufficient to overcome the upward surface tension. Namely, a threshold current value, around 260 A in this condition, is needed for the formation of keyhole. As Ipp becomes higher, the depressed deformation of the weld pool surface is significantly increased, with the formation of a deep penetrated keyhole inside the weld pool. Meanwhile, the keyhole size Dk increases from 2.1 mm to 5.8 mm. The weld width Wweld is basically the same as the weld pool width Dw. The weld depth Dweld is 3.8 mm when Ipp is 260 A and is 7 mm when Ipp is 310 A. It means that there is a saltation in weld depth when the current increases from 260 A to 310 A, which is consistent with the keyhole size Dk. The ratio of D/W shows that the welding penetrating ability increases greatly when Ipp is higher than 260 A.
Fig.7 Effect of Ipp on keyhole and weld formation: (a) appearance of weld pool and cross-section; (b) weld pool width Dw and keyhole size Dk (Ipp=210, 260, 310, 360 A, Ibp=120 A, fL=2 Hz, duty cycle=50%)
From the results, the weld penetration depth Dweld is simultaneously affected by the welding current Ipp and keyhole size Dk. To better understand the relationship among Dweld, Ipp and Dk, regression analysis was performed, taking Dweld as the response of Dk and the square of Ipp. The regression equation can be expressed by:
From the results, the weld penetration depth is greatly affected by the formation of keyhole in addition to the welding current. The formation of keyhole inside the weld pool is beneficial to the increase of the weld penetration depth. As welding arc moves downwards along the keyhole, the solid base metal under the weld pool can be directly heated by the welding arc, rather than the heat conduction from the weld pool. The welding efficiency will be largely enhanced after the formation of keyhole in the weld pool.
2.3 Effect of low-pulsed frequency on keyhole formation
To understand the effect of low-pulsed frequency on keyhole formation, trials were undertaken with changing fL from 0.5 Hz to 5 Hz. Fig.8 shows the appearance of weld pool and keyhole during tp. It can be seen that under the condition of the same current value, the appearance of weld pool and keyhole differs largely at different fL. As fL increases, the weld pool width Dw presents a downward trend, decreasing from 13.4 mm at 0.5 Hz to 10.1 mm at 5 Hz. The keyhole size Dk exhibits a similar trend as Dw does. It decreases from 7.1 mm at 0.5 Hz to 4.1 mm at 5 Hz. The keyhole size Dk shows an approximately linear changing trend with fL, following the fitted Eq.(7). This is mainly because the pulse period decreases as the pulse frequency increases, leading to a decrease in residence time of pulse peak current in a complete pulse period.
Fig.8 Effect of low-pulsed frequency on keyhole formation: (a) appearance of weld pool and keyhole; (b) weld pool width Dw and keyhole size Dk (Ipp=360 A, Ibp=120 A, fL=0.5~5 Hz, duty cycle=50%)
1) During DP-VPTIG keyhole mode welding, the keyhole forms in low-pulsed peak stage tp due to the dominant role of downward arc force Pa-z. The keyhole is closed in low-pulsed base stage tb as the upward forces surface tension Ps and hydrostatic pressure Ph become the dominant role. The periodic variation in low-frequency pulsed current stimulates a periodic keyhole behavior of “opening” and “closing” in the weld pool.
2) The threshold value of low-pulsed peak current Ipp for the formation of keyhole is about 260 A. The formation of keyhole is beneficial to the increase of the weld penetration depth as the arc moves downwards along the keyhole, thus directly heating the solid base metal under the weld pool.
3) The weld pool width Dw and keyhole size Dk both decrease with the increase of low-pulsed frequency fL.
References
1 Cong B Q, Ding J L, Williams S et al. International Journal of Advanced Manufacturing Technology[J], 2013, 76: 1593 [Baidu Scholar]
2 Zhang P, Chen M H, Xie L S et al. Rare Metal Materials and Engineering[J], 2020, 49(3): 819 [Baidu Scholar]
3 Li Q, Wang G Q, Dong M H et al. Rare Metal Materials and Engineering[J], 2021, 50(5): 1649 [Baidu Scholar]
4 Wang Y P, Qi B J, Cong B Q et al. Journal of Materials Processing Technology[J], 2017, 249: 89 [Baidu Scholar]
5 Wang L L, Wei J H, Wang Z M et al. International Journal of Advanced Manufacturing Technology[J], 2018, 95: 2421 [Baidu Scholar]
6 Howse D S, Lucas W. Science and Technology of Welding and Joining[J], 2013, 5(3): 189 [Baidu Scholar]
7 Jerzy N. Welding International[J], 2014, 28(1): 45 [Baidu Scholar]
8 Jarvis B L, Ahmed N U. Science and Technology of Welding and Joining[J], 2000, 5(1): 1 [Baidu Scholar]
9 Liu Z M, Chen S Y, Liu S et al. International Journal of Heat and Mass Transfer[J], 2018, 123: 54 [Baidu Scholar]
10 Lathabai S, Jarvis B L, Barton K J. Science and Technology of Welding and Joining[J], 2008, 13(6): 573 [Baidu Scholar]
11 Lathabai S, Jarvis B L, Barton K J. Materials Science & Engineering A[J], 2001, 299: 81 [Baidu Scholar]
12 Lohse M, Fussel U, Schuster H. Welding in the World[J], 2013, 57: 735 [Baidu Scholar]
13 Olivares E A G, Silva R H G, Dutra J C. Welding International[J], 2015, 20(3): 262 [Baidu Scholar]
14 Feng Y Q, Luo Z, Liu Z M et al. Materials & Design[J], 2015, 85: 24 [Baidu Scholar]
15 Liu Z M, Fang Y X, Qiu J Y et al. Journal of Materials Processing Technology[J], 2017, 250: 132 [Baidu Scholar]
16 Cui S W, Shi Y H, Sun K. Materials Science & Engineering A[J], 2018, 709: 214 [Baidu Scholar]
17 Fang Y X, Liu Z M, Cui S L et al. Journal of Materials Processing Technology[J], 2017, 250: 280 [Baidu Scholar]
18 Cui S L, Liu Z M, Fang Y X et al. Journal of Materials Processing Technology[J], 2017, 243: 217 [Baidu Scholar]
19 Wang Y P, Qi B J, Cong B Q et al. Journal of Manufacturing Processes[J], 2018, 34: 179 [Baidu Scholar]
20 Yang M X, Li L, Qi B J et al. Science and Technology of Welding and Joining[J], 2017, 22(7): 580 [Baidu Scholar]
21 Yang M X, Zheng H, Qi B J et al. Journal of Materials Processing Technology[J], 2017, 243: 9 [Baidu Scholar]
22 Wang Y P, Cong B Q, Qi B J et al. Journal of Materials Processing Technology[J], 2019, 266: 255 [Baidu Scholar]
23 Wang Y P, Qi B J, Cong B Q et al. Rare Metal Materials and Engineering[J], 2018, 47(11): 3522 [Baidu Scholar]
