Arc characteristics in double-pulsed VP-GTAW for aluminum alloy

Accepted Manuscript Title: Arc Characteristics in Double-Pulsed VP-GTAW for Aluminum Alloy Authors: Wang Yipeng, Qi Bojin, Cong Baoqiang, Yang Mingxuan, Liu Fangjun PII: DOI: Reference:

S0924-0136(17)30201-7 http://dx.doi.org/doi:10.1016/j.jmatprotec.2017.05.027 PROTEC 15235

To appear in:

Journal of Materials Processing Technology

Received date: Revised date: Accepted date:

25-1-2017 20-5-2017 22-5-2017

Please cite this article as: Wang, Yipeng, Qi, Bojin, Cong, Baoqiang, Yang, Mingxuan, Liu, Fangjun, Arc Characteristics in Double-Pulsed VPGTAW for Aluminum Alloy.Journal of Materials Processing Technology http://dx.doi.org/10.1016/j.jmatprotec.2017.05.027 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

Arc Characteristics in Double-Pulsed VP-GTAW for Aluminum Alloy

Wang Yipeng1,2, Qi Bojin1,2,*, Cong Baoqiang1,2, Yang Mingxuan1,2, Liu Fangjun1,2 School of Mechanical Engineering and Automation, Beihang University, 100191, China; MIIT Key Laboratory of Aeronautics Intelligent Manufacturing, Beijing 100191, China

1 2

*

Corresponding author: [email protected] ; Tel.: +86-10-82339961

Abstract: This paper studied the variation in arc profile and arc macroscopic electromagnetic force with double-pulsed variable polarity gas tungsten arc welding (DPVP-GTAW) for aluminum alloy, using a synchronous acquisition and analysis system. The arc shapes were regionalized into arc edge region and arc core region by image processing method based on the gray level of the arc images. The arc characteristic size such as arc electrode end diameter DE, arc workspace end diameter DB and arc length L were measured and analyzed in the arc core region. In variable polarity pulse phase, DE and DB decreased, while L increased when welding current switched from negative polarity to positive polarity. DE, DB and L were all much larger in low-frequency pulse on tp compared with those in low-frequency pulse off tb. The size of arc profile changed periodically between tp and tb throughout the welding process. The arc macroscopic electromagnetic force Fz and arc pressure Pz oscillated in cycles with the variation of arc profile, inducing the refinement of weld zone grain structure.

Keywords: Arc profile; Double pulse; VP-GTAW; Electromagnetic force

1 Introduction Variable polarity gas tungsten arc welding (VP-GTAW) is a very precise and clean welding process that has been extensively used in aeronautics, astronautics and automobile industry to produce high-quality weld joints of aluminum alloy as demonstrated by Pan et al. (2016). Wang et al. (2004) indicated that the welding

2

current with conventional VP-GTAW is typically no more than 250/-250A, making it ideal for welding the thin plates. Workpieces thicker than 3mm require careful processing of V-type or X-type edge groove before welding as described by Lathabai et al. (2008). Balasubramanian et al. (2008) reported that such preparations, combined with multiple passes welding with filler wire, lead to long joint completion time. Cong et al. (2016) investigated that substantial enhancement in welding current would cause high consumption of tungsten electrode, resulting in the mixture of tungsten in weld joint, and further affecting the stability of the welding arc and the joint quality. Kumar et al. (2007) stated that a higher heat input would cause severe grain growth, easily bringing about composition segregation, distortion and cracks of weld joint. In order to realize the high efficiency welding of thick aluminum alloy plates, without significantly increasing the heat input, an innovative double-pulsed VP-GTAW (DPVP-GTAW) technology was developed, in which the variable polarity pulsed current of VP-GTAW was modulated into a relative low-frequency pulse with two stages, i.e., low-frequency pulse on duration tp and low-frequency pulse off duration tb, as shown by Liu et al. (2013). In tp, the amplitude of welding current could be significantly enhanced to obtain the full penetration, while in tb, the welding current is always relatively low for the reduction of heat input, and the consumption of tungsten electrode would be greatly decreased as revealed by Ghosh et al. (2007). Since DPVP-GTAW technology has been gradually applied in industrial production and manufacturing, the welding process of DPVP-GTAW has attracted more and more attention. Liu et al. (2013) showed that the welding arc of DPVP-GTAW is primarily influenced by double pulses simultaneously, including variable polarity pulse and low-frequency pulse. Ghosh et al. (2008) identified that the welding arc plays a critical role in chemical composition, microstructure, mechanical properties, appearance, distortion and defects of the weld joint, as it is the source of heat and force. Therefore, it is necessary to understand the variation in arc profile and arc force with DPVP-GTAW process for aluminum alloy.

It was found that there

was not much relevant work done before in the aspect of arc characteristics with DPVP-GTAW. In this study, an investigation was conducted to explore the variation in

3

arc profile and arc macroscopic electromagnetic force with DPVP-GTAW for aluminum alloy by using a synchronous acquisition and analysis system. The work may help in understanding the weld characteristics of DPVP-GTAW, and may be beneficial in using DPVP-GTAW to produce the desired weld quality.

2 Experimental procedure 2.1 Materials The base metal used in this study was AA 2124 aluminum alloy plate with a chemical composition of Cu 4.69, Si 0.13, Mn 0.67, Mg 1.13, Zn 0.07, Fe 0.14, Ni 0.01 and Al balance (all in wt%). The gauge dimension of the workpiece was 200mm×100mm×6mm. The shielding gas adopted was pure argon with a volume fraction of 99.99%, and a cerium tungsten electrode with diameter of 4mm was selected throughout the experiments.

2.2 Welding platform The welding platform employed in the tests is schematically presented in Fig. 1. It can be seen that the welding system was primarily composed of a welding power source, a welding torch, a welding motion mechanism and a synchronous acquisition system which mainly consisted of a high-speed video camera, a voltage sensor, a current sensor, a data acquisition card with maximum sampling frequency 500 kHz and a computer control system. The high-speed video camera (Mega75K) with maximum capturing rate of 7.5 thousand images per second was equipped with a close-focusing macro video lens Navitar ZOOM 7000. A neutral filter and a band pass filter with central wavelength of 670nm and bandwidth of 20nm were used to reduce the intensity of arc. The optical axis of the camera was aligned with the electrode axis, and it was perpendicular to the moving axis. To avoid the differences in the captured arc profile caused by the movement of welding torch, the high-speed camera was set to focus on the fixed region around electrode tip with the unmovable welding torch, and the workpiece clamped to the workbench moved during the welding procedure. The experimental data would be displayed and recorded in the computer control

4

system through the synchronous acquisition system in real-time.

2.3 Welding process parameters The schematic and actual welding current waveform of DPVP-GTAW is shown in Fig. 2. It can be observed that, there were two periodically varying pulses, i.e., the high-frequency variable polarity pulse ranging from 50Hz to 100Hz and the low-frequency pulse with the variation range of 0.5-5Hz. The value of these current parameters could be adjusted on the touchpad in the welding power source. Based on a large number of trials, the process parameters in this study were selected to obtain an excellent weld bead geometry, as shown in Table 1.

3 Results and discussion 3.1 Measurement of arc profile In order to analyze the variation in arc profile of DPVP-GTAW captured by the high-speed video camera, it is necessary to extract the edge of the arc for regionalization and measurement. The arc image processing procedure is shown in Fig. 3. As shown in Fig. 3a, the original arc was taken from the welding video in computer control system; while as shown in Fig. 3b, the original arc was transformed into gray-scale image, where the arc was divided into 256 gray scale levels according to the intensity of the arc, and it could be indicated that the gray scale was larger with higher intensity as illustrated by Yang et al. (2017). Using a median filter, the gray scale image was then denoised, aiming to decrease the noise interference, as illustrated in Fig. 3c. As shown in Fig. 3d, the arc edge region and arc core region were recognized and extracted by demarcating the boundaries with the critical gray scale values of 167 and 242.25, respectively. The yellow arc core region has the highest radiation intensity, largest current density, most concentrated plasma, and most drastic ionization thermal motion as investigated by Zuo et al. (2014). As a result, the arc profile would be measured and analyzed using the arc core region. Arc profile was measured in the arc core region mentioned above, as presented in Fig. 4. The arc profile, which was defined by electrode end diameter DE, workspace

5

end diameter DB and length L, was accurately measured by computerized measurement technique. The dimension of the electrode DR with the diameter of 4mm was taken as the reference to determine the actual size of DE, DB and L. In order to ensure the repeatability and avoid the measuring errors, the average value was calculated using three complete cycles of experimental data.

3.2 Variation in arc profile Fig. 5 displays a complete welding current cycle (10ms) of variable polarity pulse during the low-frequency pulse off tb, arc images captured per millisecond and their corresponding DE, DB and L, which were measured using the above-mentioned surveying method. In Fig. 5a, it can be seen that the amplitude of welding current Ibn/Ibp was about -170/170A in the duration of negative polarity tHn and positive polarity tHp, respectively. The ratio of tHn to tHp was 2:8, so that two arc images were intercepted in tHn, while eight arc images were intercepted in tHp. Fig. 5b and Fig. 5c illustrate the variation in arc profile throughout the variable polarity pulse period. From 8 to 10ms in tHn, the arc sizes of DE and DB were relatively high, which were more than 5mm and 12mm respectively, while the arc length L was less than 10mm, with arc root remaining at the position of electrode tip. As welding current varied from Ibn to Ibp at 11ms in tHp, the arc constricted with DE and DB decreased to no more than 4mm and 10mm respectively, and L increased to 11mm with arc root climbing to the upper part of the electrode tip. In 12-18ms, the arc profile maintained in a stable state without any obvious change in size. The △DE, △DB and △L in tb were about 1mm, 2mm and 1mm, respectively. In the duration of tHn, with workpiece connected to the cathode, the oxidation films attached to the surface of the aluminum alloy could be removed by the energetic positive ions of the arc plasma, which were accelerated by the voltage drop of the cathode as demonstrated by Sarrafi and Kovacevic (2010). The cathode spots during this process can be recognized in the arc edge region of the arc images at 9ms and 10ms in Fig. 5. After the oxidation films under the cathode spots were cleaned up, the cathode spots would automatically search for the next oxidation films around the weld

6

beam. Therefore, a slight increase in the size of arc profile occurred in tHn. The arc constriction from tHn to tHp resulted in the elevation of arc energy density and arc stiffness with higher arc efficiency. As a consequence, the arc intensity would increase and the melted material would be improved in tHp compared with those in tHn, as shown in Fig. 6. Fig. 7 shows the complete welding current cycle (10ms) of the variable polarity pulse and the variation in the arc profile during the low-frequency pulse on tp, which indicated that DE, DB and L had almost the same changing tendency as those of tb. However, the amplitudes of the welding current Ipn/Ipp was about -360/360A in the duration of the negative polarity tHn and the positive polarity tHp respectively, which were more than two times higher than those of tb. DB and L increased significantly to more than 21mm and 14mm, an increase of 75% and 40% respectively in comparison to those of tb, while there was only a slight change of DE due to the restriction of electrode. The periodical changes of the low-frequency pulsed current could lead to the regular changes in arc profile and arc energy, thereby affecting the thermal and force effects on the welding pool. The appearance of weld surface and weld penetration would thus fluctuate in cycles. In the duration of tp, the energy density of the arc was larger with a higher temperature and penetrating ability. A deep depression of the welding pool occurred with its surface moving down, pushing part of the molten metal to the trail of the welding pool. In the duration of tb, the energy density of the arc was relatively small with a lower temperature and penetrating ability. A shallow depression of the welding pool formed as its surface moved up. The appearance of weld surface with the characteristics of fish scale was thus formed, as shown in Fig. 8a. Correspondingly,

the macrographs

of

their longitudinal section

with

low-frequency pulse of 0.5Hz, 1Hz and 2Hz are shown in Fig. 8b, respectively. It can be found that there was a fluctuation in the weld penetration caused by the low-frequency pulse. The fluctuation decreased with the increase of pulsed frequency, and there was almost no fluctuation when the low-frequency pulse was 2Hz.

7

3.3 Variation in arc electromagnetic force According to the report of Ando and Hasegawa (1985), the macroscopic electromagnetic force of the arc plasma can be derived by the arc profile as Fz 

 4

I 2ln

RB RE

(1)

where Fz is the axial electromagnetic force, μ is the permeability of the arc atmosphere (μ≈μ0=4π×10-7 H/m), RB=2/DB is the radius of the arc at base metal end, and RE=2/DE is the radius of the arc at electrode end. This expression has been further validated by Cook et al. (1985) and Yang et al. (2013). Therefore, it can be used to quantitatively describe the variation in the arc electromagnetic force of DPVP-GTAW. With the arc characteristic size DE, DB and L described in Fig.5c and Fig.7c, the value of the axial electromagnetic force Fz in response to time alteration in a complete current cycle is revealed in Fig. 9. It can be found that the Fz was more than 0.018N throughout the duration of low-frequency pulse on tp, while it was less than 0.0035N in the low-frequency pulse off tb. Fz increased significantly when welding current was switched from tb to tp. Meanwhile, the ratio of Fz in pulse on to Fz in pulse off remained at the value of more than 6, indicating that the Fz oscillated dramatically and periodically with the low-frequency pulse. In order to show the effect of the axial electromagnetic force Fz on the welding pool, the axial arc pressure Pz was developed by Eq.2. Pz 

Fz  1 R  2 I 2 2 ln B 2  RB 4 RB RE

(2)

The value of Pz in response to time variation is described in Fig. 10. Similarly, the arc pressure Pz enhanced obviously with the welding process changing from tb to tp with respect to both in pulse on and pulse off. The ratio of Pz in pulse on to Pz in pulse off remained at the value between 1.5 and 2. The periodical oscillation of the arc pressure could cause the stirring action on the welding pool, which might enhance the fluidity of the welding pool, thereby resulting in the promotion in heat transfer and mass transfer processes. Meanwhile, the distribution of the internal temperature and composition of the molten pool could be more uniform, which is beneficial in the reduction of the temperature gradient at

8

the solid-liquid interface and the refinement of the grains during the solidification process, as identified by Liu et al. (2012). This assumption could be confirmed by the microstructural observation of the weld zone. Fig. 11 shows the microstructures in the weld zone with conventional VP-GTAW and DPVP-GTAW, respectively. It can be seen that the microstructure of VP-GTAW was primarily composed of coarse dendrite grains with precipitations segregated at the grains boundaries. In the microstructure with DPVP-GTAW, the amount of fine equiaxed grains increased significantly with an obvious reduction of the coarse dendrite grains, and the distribution of the precipitations was more uniform.

4 Conclusions The arc characteristics involving the variation in arc profile and arc macroscopic electromagnetic force in DPVP-GTAW process for aluminum alloy were systematically investigated with conclusions drawn as follows: 1. The arc constriction occurred when the welding current switched from the negative polarity tHn to the positive polarity tHp during the variable polarity pulse phase in DPVP-GTAW, for there was a decrease of the arc electrode end diameter DE and arc workspace end diameter DB and an increase of the arc length L. 2. The size of the arc profile changed periodically between the low-frequency pulse on tp and the low-frequency pulse off tb throughout the welding process, and DE, DB as well as L were all much larger in tp compared with those in tb. 3. The arc macroscopic electromagnetic force Fz and the arc pressure Pz oscillated in cycles with the variation of arc profile, which was beneficial to the refinement of weld zone grain structure.

Acknowledgments: The authors would like to thank all the members in Welding Manufacturing Research Group in Beihang University. The work was supported by the National Natural Science Foundation of China under Grants No. 51675031, 50975015 and 51005011.

9

References Ando, K., Hasegawa, M., 1985. Welding arc phenomena. China Machine Press, Beijing, pp. 249– 250. Balasubramanian, V., Ravisankar, V., Madhusudhan, R.G., 2008. Effect of pulsed current welding on fatigue behavior of high strength aluminum alloy joints. Mater. Des. 29, 492-500. Cong, B., Ouyang, R., Qi, B., Ding, J., 2016. Influence of cold metal transfer process and its heat input on weld bead geometry and porosity of aluminum-copper welds. Rare Metal Mat. Eng. 45(3), 606-611. Cook, G.E., Eassa, E.H., 1985. The effect of high-frequency pulsing of a welding arc. IEEE T. Ind. Appl. A-2(5), 1294-1299. Ghosh, P.K., Dorn, L., Hubner, M.V., Goyal, K., 2007. Arc characteristics and behavior of metal transfer in pulsed current GMA welding of aluminum alloy. J. Mater. Process. Technol. 209, 163-175. Ghosh, P.K., Dorn, L., Kulkarni, S., Hofmann, F., 2008. Arc characteristics and behavior of metal transfer in pulsed current GMA welding of stainless steel. J. Mater. Process. Technol. 188, 1-13. Kumar, T., Balasubramanian, V., Sanavullah, M.Y., 2007. Influences of pulsed current tungsten inert gas welding parameters on the tensile properties of AA 6061 aluminum alloy. Mater. Des. 28, 2080-2092. Lathabai, S., Jarvis, B.L., Barton, K.J., 2008. Keyhole gas tungsten arc welding of commercially pure zirconium. Sci. Technol. Weld. Joi. 13(6), 573-581. Liu, A., Tang, X., Lu, F., 2013. Study on welding process and prosperities of AA5754 Al-alloy welded by double pulsed gas metal arc welding. Mater. Des. 50, 149-155. Liu, A., Tang, X., Lu, F., 2013. Arc profile characteristics of Al alloy in double-pulsed GMAW. Int. J. Adv. Manuf. Technol. 65, 1-7. Liu, Y., Wang, W., Xie, J., 2012. Microstructure and mechanical properties of aluminum 5083 weldments by gas tungsten arc and gas metal arc welding. Mat. Sci. Eng. 549, 7-13. Pan, J., Hu, S., Yang, L., Wang, D., 2016. Investigation of molten pool behavior and weld bead formation in VP-GTAW by numerical modelling. Mater. Des. 111, 600-607. Sarrafi, R., Kovacevic, R., 2010. Cathodic cleaning of oxides from aluminum surface by variable-polarity arc. Weld. J. 89, 1-10. Wang, H., Jiang, W., Quyang, J., Kovacevic, R., 2004. Rapid prototyping of 4043 Al-alloy parts by VP-GTAW. J. Mater. Process. Technol. 148(1), 93-102. Yang, M., Qi, B., Cong, B., Liu, F., Yang, Z., Chu, P., 2013. Study on electromagnetic force of arc plasma with by ultrahigh frequency pulsed GTAW of Ti-6Al-4V. IEEE T. Plasma. Sci. 41(9), 2561-2568. Yang, M., Zheng, H., Li, L., 2017. Arc shapes characteristics with ultra-high-frequency pulsed arc welding. Appl. Sci. 7(1), 45. Zuo, Z., Guo, J., Liu, J., 2014. Effect of area arc distribution on diamond nucleation. J. Synth. Cryst. 43(10), 2515-2521.

10

power source synchronous controller voltage sensor ON OFF touch pad

welding torch

data acquisiton card

current sensor neutral filter

400

arc

Current/A

200

0

-200

-400

0

10

20

30

40

50

60

70

80

90

100 110

120

Time/ms

shielding gas

high speed video camera

band pass filter

weld pool base metal computer control system

Fig. 1 Schematic diagram of experimental platform I/A

TL

tp

Ipp

tb 400

Ibp 0 Ibn

tHp tHn

Ipn

TH

Current/A

200

t/ms

(a) Schematic waveform of DPVP-GTAW

0

-200

-400

600

0

10

20

30

40

50

60

70

80

90

100 110

120

Time/ms

500

(c) Transition condition from tp to tb

400 300 200

Current/A

400

100

Current/A

200

0

-100 -200

0

-200

-400

-300

0

-400 -500 0.0

10

20

30

40

50

60

70

80

90

100 110 120 130

Time/ms 0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

(d) Transition condition from tb to tp

Time/s

(b) Actual current waveform of DPVP-GTAW

TL-period of low-frequency pulse; tp-duration of low-frequency pulse on; tb-duration of low-frequency pulse off; Ipp-positive polarity current during tp; Ipn-negative polarity current during tp; Ibp-positive polarity current during tb; Ibn-negative polarity current during tb; TH-period of variable polarity pulse; tHp-duration of positive polarity; tHn-duration of negative polarity Fig. 2 Waveform of DPVP-GTAW

11

Gray-scale image

Original arc

(b) Gray-scale transformation

(a) Original arc

Edge extraction

Median filter

Arc edge

Arc core region

(c) Median filter

(d) Edge extraction and regionalization

Fig. 3 Edge extraction and regionalization of welding arc

Original arc

Processed arc

DR

4mm (DR) DE

Arc edge Arc edge

L

Arc core region Arc core region (a) Original arc

Fig. 4 Schematic of arc profile measurement

DB (b) Processed arc

12

400

15

DE

14 300

Positive duation

Negative duation

DB

13 12

200

L

L

11 10

Size/mm

Current/A

100 0

-100

9

DB

8

Negative duration

7

Positive duration

6 -200

5

DE

4

-300

3 -400

2 8

9

10

11

12

13

14

15

16

17

18

19

9

8

Time/ms

10

11

12

13

14

15

16

17

18

19

Time/ms

(c) Arc size

(a) Current waveform 9ms

10ms

Negative duration

11ms

12ms

13ms

14ms

Positive duration 15ms

16ms

17ms

18ms

(b) Arc images

Fig. 5 Arc profile in low-frequency pulse off

(a) Negative polarity duration

(b) Positive polarity duration

Fig. 6 Arc profile and appearance of welding pool in tb (captured through a band pass filter with central wavelength of 960nm and bandwidth of 20nm)

13

28

600

400

Positive duation

Negative duation

L

22 20

200

18

Size/mm

100

16

0

Negative duration

Positive duration

L

14

-100

12

-200

10

-300

8

-400

6

-500 -600 78

DB

DB

24

300

Current/A

DE

26

500

DE

4 79

80

81

82

83

84

85

86

87

88

2 78

79

80

81

82

83

84

85

86

Time/ms

Time/ms

(a) Current waveform 79ms

(c) Arc size 80ms

Negative duration

Positive duration

81ms

82ms

83ms

84ms

85ms

86ms

87ms

88ms

(b) Arc images

Fig. 7 Arc profile in low-frequency pulse on

(a) Morphology of weld surface (Ipp/Ipn=250A/-250A, Ibp/ Ibn=100A/-100A)

(b) Macrographs of longitudinal section (Ipp/Ipn=250A/-250A, Ibp/ Ibn=100A/-100A) Fig. 8 Morphology of weld surface and macrographs of longitudinal section

87

88

89

14

0.028

12

0.026

Fz in pulse on

0.024

Fz in pulse off Ratio Fz

0.022

11 10

0.020

9

Fz in pulse on

FZ /N

0.016 0.014

8

Fz in pulse on / Fz in pulse off

0.012 0.010

7

Nagtive duration

0.008

Positive duration

0.006

Ratio Fz

0.018

6

Fz in pulse off

0.004

5

0.002 0

1

2

3

4

5

6

7

8

9

10

4 11

Time/ms

Fig. 9 Axial electromagnetic force Fz

80 75

Nagtive duration

70

Pzin pulse on

Positive duration

65

Pz in pulse off Ratio Pz

Pz in pulse on

60

5.0 4.5 4.0

55

Pz /Pa

45 3.0

40 35

Pz in pulse off

30

Ratio Pz

3.5

50

2.5

25

Pz in pulse on / Pz in pulse off

20

2.0

15 1.5

10 5 0

1

2

3

4

5

6

7

8

9

10

11

Time/ms

Fig. 10 Axial arc pressure Pz

(a) Conventional VP-GTAW Fig. 11 Microstructure of weld zone

(b) DPVP-GTAW

15

Table 1 Welding process parameters Parameter

Value

Unit

Argon flow rate

15

L·min-1

Welding velocity

170

mm·min-1

Electrode extension length

5

mm

Arc image capture rate

1000

s-1

Signal sampling rate

100

kHz

fL=1/TL

2

Hz

tp : tb

3:7

-

fH=1/TH

100

Hz

tHp : tHn

8:2

-

Ipp/Ipn

360/-360

A

Ibp/Ibn

170/-170

A