Metal Transfer with Force Analysis in Consumable and Nonconsumable Indirect Arc Welding Process

NOVEMBER 2014 / WELDING JOURNAL 431-s Metal Transfer with Force Analysis in Consumable and Nonconsumable Indirect Arc Welding Process At a certain welding current level, there was an optimized wire feed speed to achieve a higher metal transfer rate and smaller droplet size BY J. WANG, Y. HUANG, J. XIAO, J. FENG, C. Y. TIAN, AND J. WANG ABSTRACT This paper proposes an indirect arc welding process that uses a consumable and nonconsumable electrode to establish an arc. Because there is no welding current passing through the workpiece, the heat input is reduced. The consumable electrode is a welding wire and the nonconsumable is a tungsten electrode. The wire feed speed and welding current are two main parameters determining the metal transfer in the process, and their effects are experimentally examined in this paper. It was found that at a certain welding current level, there was an optimized wire feed speed to achieve a higher metal transfer rate and smaller droplet size. When the current increases, the transfer with the optimized feed speed would change from globular to spray transfer. Forces acting on the droplet were analyzed to explain phenomena experimentally observed from the metal transfer process. The arc behavior was also analyzed to better understand this welding process and its metal transfer. KEYWORDS Metal Transfer Force Analysis Indirect Arc Welding Introduction Gas metal arc welding (GMAW) is currently one of the most widely used welding methods due to its productivity (Refs. 1, 2) and convenience for automatic and semiautomatic welding applications. The transfer of the melted wire (electrode) onto the base metal forms a process referred to as metal transfer. Metal transfer plays an essential role in determining the process stability, welding productivity, and so on, plus it has been an active field of research and application in the welding community (Refs. 3 6). The metal transfer depends significantly on the welding current levels. To obtain a free droplet transfer process, a certain level of welding current should be used. However, higher welding current will apply higher heat input into the workpiece, which may be unnecessary or undesirable in certain applications. Hence, research has been conducted to explore welding methods to reduce welding current and achieve stable free metal transfer at the same time. Pulsed gas metal arc welding (GMAW-P) has been proved as an effective method to achieve the desirable free transfer/spray transfer at a needed heat input determined by the average current, but the peak current must be higher than the transition current, which may blow the liquid metal away from the weld pool and cause meltthrough (in complete joint penetration applications) (Refs. 7, 8). Cold metal transfer (CMT), developed at and patented by Fronius, Inc., is a process that utilizes a fast mechanical movement to draw the welding wire back such that the liquid metal can be separated from the wire to transfer the metal at a current much lower than the transition current. As a result, both the total heat input and current levels can be much reduced (Refs. 9, 10). However, the complicated device for the mechanical movement introduces disadvantages affecting its wider acceptance. The patented surface tension transfer (STT) (Refs. 11 13) process is another effective method to reduce spatter to a minimum with low heat input and arc pressure, but its effective range for the average current is restricted by the mandatory need for the particular current waveform/range needed to reduce spatter and may not always be most desirable for certain applications. Zhang et al. proposed a patented method (Refs. 14, 15) to use a peak current much lower than the transition current to produce the desired spray transfer by taking advantage of the momentum of a downward (away from the gun) droplet. Recently, laser-enhanced GMAW, which utilizes a low-power laser to provide an additional detaching force to transfer droplet, was developed to J. WANG (wjtcywkx@gmail.com), J. XIAO, and J. FENG are with the National Key Laboratory of Advanced Welding and Joining, Harbin Institute of Technology, Harbin, China. Y. HUANG is with the Center for Sustaining Manufacturing and Department of Electrical Engineering, University of Kentucky, Lexington, Ky. In addition, WANG, C. Y. TIAN, and J. WANG are with the School of Materials Science and Engineering, Jiamusi University, Jiamusi, China.

Fig. 1 Schematic diagram of the experimental configuration. achieve free metal transfer with low welding current (Refs. 16 20). The droplet size and trajectory were well controlled by the laser power intensity. Pulsing current was also combined with this method to reduce the heat input, and the patented droplet oscillating method was also introduced to this process to further control the behavior of the metal transfer (Ref. 21). Double electrode GMAW (DE- GMAW) and double bypass GMAW (DB-GMAW) are arc welding processes that adopted one or two bypass guns to reduce the total heat input to the workpieces (Refs. 4, 5, 22). The current through the main gun was decoupled to two or three parts with only one of them to be passed through the base metal. Pulsing current can be used to further reduce heat input. Development for various low-heatinput arc welding methods suitable for various applications has been a major continuous effort in the welding research community. To contribute to this effort, the consumable and nonconsumable electrodes indirect arc welding (CNC-IAW) was developed to provide an alternative way to reduce heat input in the arc welding process (Refs. 23 26). In the CNC-IAW process, the welding current through the conventional GMAW gun will directly pass to another tungsten torch, considered as a nonconsumable gas tungsten arc (GTA) electrode, back to the power supply. Instead, the current will no longer pass through the base material as in conventional arc welding. The main heat input to the workpiece would be only from the molten droplets. Unfortunately, while this process provides unique and distinguished properties, it lacks understanding at fundamental levels. This paper denotes to enhance understanding the CNC-IAW process from its metal transfer, which fundamentally affects any arc welding process with a consumable electrode. Experimental Setup and Conditions In the CNC-IAW process, compared to the conventional GMAW process, the welding current through the welding wire will not pass into the base material back to the power supply. Instead, a secondary tungsten torch is adopted as the negative terminal that is directly connected to the negative terminal of the power supply. In this case, no welding current would pass through the base material, and the heat input into the workpiece is mainly from the molten droplet whose heat should be relatively small. The total heat input to the workpiece in CNC- IAW will be reduced. Figure 1 shows the schematic configuration of the proposed system for CNC-IAW. The tungsten nonconsumable electrode and welding wire are connected to the power supply s negative and positive terminals, respectively. The indirect arc generated between the two electrodes produces continuously transferred molten droplets. The base metal is independent of the indirect arc and not connected to the welding power supply. The welding current flows from the positive terminal to the welding wire and then to the indirect arc ignited between the two dissimilar electrodes. Through the tungsten electrode, the welding current flows back into the negative terminal of the welding power source. The heat input can be controlled by adjusting the distance between the welding torches and base metal, which is none of polarities of the arc. With the synchronous forward movement of the two torches or movement of the workpiece, the deposited metal solidifies into a weld joining the two members of the workpiece being joined. The metal transfer is imaged using a high-speed camera and backlighting, as can be seen in Fig. 1. Figure 2 shows the main important parameters needed to set up the experimental system to conduct the CNC-IAW process as expected. In this study, the GMAW gun and tungsten torch stayed stationary while the workpiece moved at a constant speed. The high-speed camera was placed at a distance about 1.0 m from the welding torches. The contact tube-to-workpiece distance d 1, which plays an important role in determining process stability, was set at 20 mm. Both torches were set about 30 deg with the perpendicular line such that the angle between the two torches was about 60 deg. The wire extension length was hard to set exactly as it would vary depending on the current level in relation to the wire feed speed. However, the approximate value of d 2 from the cross point, where the wire and tungsten extension intersect as shown in Fig. 2, to the contact tube was about 15 mm. The horizontal distance between the welding wire and tungsten electrode d 3 defined in Fig. 2 was the most important parameter to determine arc stability. From the experimental results, 1 2 mm was found to be the acceptable value for this distance, which ensured arc stability. The welding power source used in this study was a Fronius MW2200, and it was operated in constant current (CC) mode. Pure argon was used as the shielding gas for the welding wire as well as for the tungsten elec- 432-s WELDING JOURNAL / NOVEMBER 2014, VOL. 93

Fig. 2 System installation parameters. trode. The flow rate was about 15 L/min for both torches. The filler metal was 1.0-mm-diameter CuSi3 welding wire, and the base metal was 3.0- mm-thick 30CrMnSi steel. The highspeed camera recorded the welding process for the later analysis at 1000 f/s. When observing the metal transfer, the droplet could not be clearly observed for the influence of the arc light. A continuous xenon lamp could produce the high-intensity white light created by the ionized xenon gas arc. Its intensity is bright enough that the light could penetrate the indirect arc. However, it could not penetrate the droplet. When recording the metal transfer, the xenon lamp was used as the backlighting source to shine on the droplet. The camera could clearly record the projector of the droplet in the arc. When recording the arc behavior, the xenon lamp was not used to filter the disturbance of the arc. The realtime welding voltages and currents were acquired by the data acquisition system shown in Fig. 1. The wire feed speeds and welding current setting values used in the experiments are listed in Table 1. Results and Discussions Effect of Voltage Ampere Characteristic The power supply s voltage-ampere characteristic has a significant effect on the stability of the welding process and metal transfer, and should be discussed first. Hence, the constant voltage (CV) and constant current (CC) modes were compared in the preliminary study for selection of further study. By comparing the welding current and arc voltage waveforms during welding with the two voltage-ampere characteristics, the power supply s effect on this process could be obtained. In particular, when the CV mode was adopted, as shown in Fig. 3A, B, the standard deviations (square roots of the variances) for the welding voltage and current were about 2.2 V and 37.5 A, respectively, for the lower wire feed speed (4.0 m/min). Increasing the wire feed to a higher level (9.0 m/min) resulted in similar results. This result could be explained by the CV mode arc length self-regulation mechanism. In CV mode, the welding current can automatically change with the variation of the arc length. When the arc length increases, the welding current would decrease, and vice versa. Furthermore, the arc will distinguish for the large decreasing welding current as shown in Fig. 3A. When the CC mode was used, as shown in Fig. 3C, D, the standard deviation for the welding voltage was around 0.6 V for lower wire feed speed (5.0 m/min). The fluctuation in the welding current was not significant. When the wire feed speed was increased (also increasing the current level accordingly), similar results were obtained. Hence, to reduce fluctuations in the welding voltage and current during the welding process that directly affect arc stability, the CC mode was chosen to conduct CNC-IAW. Effect of Wire Feed Speed on Metal Transfer Metal transfer plays an essential role in determing the process stability and weld bead formation. The wire feed speed and welding current were identified as two main parameters to affect metal transfer. To explore these two parameters influences, one parameter was set as constant and another to be changed. To this end, the welding current was first set as constant at 110 A. Experiments were conducted with different wire feed speeds at the same welding current to investigate their effect on metal transfer characteristics. Figure 4 shows the typical metal transfer processes under different wire feed speeds for the same welding current of 110 A. Each image series in Fig. 4 shows a complete transfer cycle with different interval times between images. As shown in Fig. 4, different metal transfer phenomena were observed Table 1 Welding Parameters in CNC-IAW Experiments WFS (m/min) Welding Current (A) 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10 90 110 130 150 170 NOVEMBER 2014 / WELDING JOURNAL 433-s

A C D B with different wire feed speeds. When the wire feed speed was at 5.0 m/min, as shown in Fig. 4A, the droplet was usually detached at the position when the wire was above the cross point of the welding wire and tungsten. The metal transfer frequency was about 8 Hz. When the wire feed speed increased to 6.0 m/min, as shown in Fig. 4C, the droplet would be detached below the cross, but the metal transfer rate was still about 8 Hz. However, when the wire feed speed was set at 5.5 m/min, as shown in Fig. 4B, the detaching position was just at the cross position. The metal transfer frequency increased to about 16 Hz. The droplet was easy to detach the tip of the wire compared with the metal transfer in Fig. 4A, C. The main heat input to the workpiece in this process was mainly from the heat in the droplet. This is the main difference between this process and the traditional one. The metal transfer frequency changes as the wire extension length changes such that the wire tip changes its position in the arc zone. It appears that the cross point provides a condition to favor detachment. The stability of the metal transfer was improved. When the metal transfer frequency increases, the droplet size will also be reduced. It will benefit the process stability. To understand the peak phenomenon observed on the metal transfer frequency, forces acting on the droplet need to be analyzed. It is well known that in conventional GMAW, the major forces acting on the droplet include the gravitational force, electromagnetic force (Lorentz force), aerodynamic drag force, surface tension, and momentum force (Refs. 27 29). To be simple, the dynamic-force balance theory (DFBM) (Ref. 30) is used in this paper to conduct preliminary analysis on the forces in the CNC-IAW process. The force due to the gravity can be expressed as 4 3 Fg = mdg= πrd ρg 3 (1) where m d is the mass of the droplet, r d is the droplet radius, is the droplet density, and g is the acceleration of the gravity. The surface tension is given as Fσ = 2 πr w σ (2) where r w is the electrode radius while is the surface tension coefficient. The aerodynamic drag force can be expressed as 1 2 Fd = CdAdρpvp 2 (3) where C d is the aerodynamic drag coefficient, A d is the area of the drop seen from above, and p and v p are the density and fluid velocity of the plasma. The momentum force can be expressed as. Fm = ve m (4) d where v e is the wire feed speed and m d is the change of the droplet mass. The electromagnetic force, F em, is given by 1 1 sin 0 2 4 1 = μ ln r d θ I r θ w cos Fem (5) 4π 2 2 + ln 2 (1- cosθ) 1+ cosθ Fig. 3 Effect of welding power voltage ampere characteristic with different wire feed speeds. A 4.0 m/min, 18 V, CV mode; B 9.0 m/min, 18 V, CV mode; C 5.0 m/min, 90 A, CC mode; D 8.5 m/min, 150 A, CC mode.. where 0 is the magnetic permittivity, I is the welding current, and is the half-angle subtended by the arc root at the center of the droplet. The increase of F em accelerates and the detachment is completed rapidly. In the conventional GMAW process, the droplet is not detached when the retaining force F is still sufficient to balance the detaching force F t Ft = Fg + Fd+ Fm+ Fem (6) In Ref. 19, all these forces were approximately calculated and estimated. The results show that the aerodynamic drag force and momentum force were relatively small compared to other forces. In this case, they would be neglected in the later analysis. The main detaching forces, gravitational and electromagnetic, are shown in Fig. 5. The surface tension force is the main retaining force. The aerodynamic drag force and momentum force are not included in Fig. 5 as aforementioned. For the constant current in the CNC-IAW process shown in Fig. 4, when the wire feed speed is low or high, the arc shape would be elongated. In these cases, the main part of electromagnetic force as one of detaching force, F em,h, is not large enough to balance out the main part of the retaining surface tension. The droplet would grow to a larger size to achieve a larger gravitational force to detach from the solid wire. The metal transfer frequency would be lower than the one with the droplet right de- 434-s WELDING JOURNAL / NOVEMBER 2014, VOL. 93

WANG ET AL SUPP NOV 2014_Layout 1 10/15/14 9:25 AM Page 435 WELDING RESEARCH A B C Fig. 5 The main forces acting on the droplet. Fig. 4 Effect of wire feed speed on metal transfer. A 5.0 m/min; B 5.5 m/min; C 6.0 m/min. A B C Fig. 6 Welding current and arc voltage waveforms of different wire feed speeds. taching from the crossing point of the welding wire and tungsten shown in Fig. 4B. The stability of the metal transfer with different wire feed speeds could also be illustrated by the welding current and arc voltage waveform as shown in Fig. 6. The welding current and voltage significantly reflect stability of the metal transfer and arc behavior. The fluctuations in the current and voltage contain important information about the stability of the metal transfer. In the CNC-IAW process with CC mode, the wire feed speed affects the welding voltage to a certain degree. When the wire feed speed was 5.0 m/min, the average voltage was about 25 V, and a relatively large fluctuation in the voltage waveform could be observed. The reason was obvious. As the larger size droplet was needed to increase the gravitational detaching force, and there was an elongated arc, there would be a great opening for the droplet to grow and detach. The root of the indirect arc did not enwrap the whole droplet, which would lead to the larger fluctuation of the arc voltage. When the wire feed speed was increased to 5.5 m/min, the opening between the wire tip and tungsten tip decreased. So, the average welding voltage decreased to about 20 V, and the fluctuation in the voltage wave- form also decreased. As the droplet was detached just beside the tungsten tip, the arc was not so long that it enwrapped the whole droplet although the arc shape was asymmetric. There was no need for a larger gravitational detaching force, and in this case, the droplet was detached at a relatively small size. It would, in turn, reduce the opening and time interval of the voltage variation. The metal transfer was relatively stable. When the wire feed speed increased to about 6.0 m/min, the average voltage increased again. As the arc was elongated again, the arc could not enwrap the whole droplet. As aforementioned, the droplet would grow to a larger size to obtain a larger detaching force. As such, for a fixed welding current level, there was an optimal wire feed speed range to minimize the length of the indirect arc and achieve a stable metal transfer process. By selecting the optimized one, the droplet could be detached at a small size that may benefit the stability of the metal transfer and arc as well as the formation of the weld bead. Effect of Welding Current on Metal Transfer After selecting the optimized wire feed speed for each welding current level, the welding current effect on metal transfer could be conducted. As shown in Fig. 7, it was found that the welding current had a significant influence on the metal transfer modes, frequency, and droplet size. When the wire feed speed and weldnovember 2014 / WELDING JOURNAL 435-s

A B C D E Fig. 7 The effect of different welding currents on metal transfer. A 90 A, 4.5 m/min; B 110 A, 5.5 m/min; C 130 A, 7.0 m/min; D 150 A, 8.0 m/min; E 170 A, 10 m/min. ing current were low, as shown in Fig. 7A, the diameter of the droplet was much larger than that of the welding wire diameter. The corresponding transfer could be considered the globular metal transfer. The transfer frequency was low, which was usually 5 6 Hz. For these experimental conditions, the electromagnetic force, one of the main detaching forces, was small because of low welding current, the droplet would grow to a larger size to gain a large enough gravitational force to balance out the surface tension. As it needed some time interval to achieve a larger-sized droplet, the metal transfer frequency was relatively low. Increasing the wire feed speed and welding current to a higher level (5.5 10.0 m/min and 110 170 A), as shown in Fig. 7B E, the droplet size was reduced. The diameter of the droplet size was almost the same as that of the welding wire, and the metal transfer could be considered as spray mode. As the welding current was higher than previous ones, and it reached a certain level, the electromagnetic force was large enough to provide a detaching force combined with the effect of the gravitational force. A larger droplet size to obtain a large gravitational force was not needed any more. As the time to form a droplet was reduced, the metal transfer frequency also increased. The metal transfer frequency and droplet size are shown in Figs. 8 and 9. In this case, a conclusion could be reached that when increasing wire feed speed and the corresponding welding current, the metal transfer mode would be changed from the globular metal transfer to the projected spray transfer, and then to the stream spray one. The metal transfer frequency also increased with the droplet diameter decrease. When the wire feed speed and corresponding welding current increased, the droplet trajectory also changed gradually. As shown in Fig. 7A, B, the trajectory of the detached droplet was not along with the axis of the welding wire. When the wire feed speed and corresponding current increased, as shown in Fig. 7C E, the droplet was transferred to the weld pool almost along with the feed direction of the welding wire. In fact, as shown in Fig. 5, the gravitational force can be decomposed to two parts: one along the axis of welding wire F g,h, which would detach the droplet, and another vertical to the wire filling direction F g,v, which would cause the oscillation of the droplet. The electromagnetic force could also be decomposed into two parts similarly as the gravitational force, F em,h and F em,v, which tend to cause the droplet to detach and oscillate, respectively. As the direction of aerodynamic drag force and momentum force along with the welding wire feeding direction, the main forces to affect the droplet trajectory would be gravitational force vertical component F g,v and electromagnetic force vertical component F em,v. When the wire feed speed and corresponding welding current were low, a larger-sized droplet was obtained. The gravitational force s vertical component F g,v was larger than the electromagnetic force s vertical component F em,v. In this case, the droplet detaching direction would not be along the wire feed direction, and that tends toward the gravitational force direction. When increasing the welding current, the electromagnetic force s vertical component F em,v would also increase, and it could balance out the gravitational force vertical component F g,v. The droplet would be detached at the direction of the wire axle. Arc Behavior When the droplet formed, detached, and transferred into the weld pool, the arc would be changed correspondingly. Analysis of the arc behavior would benefit the understanding of the metal transfer phenomenon in the CNC-IAW process. As discussed previously, the arc shape has a significant effect on the magnitude of the electromagnetic force. The arc shape changes with the welding current and would influence the droplet detachment position. With the lower wire feed speed, the droplet is detached above the tungsten electrode tip. When 436-s WELDING JOURNAL / NOVEMBER 2014, VOL. 93

Fig. 8 Metal transfer frequency in CNC IAW. Fig. 9 Droplet size in CNC IAW. A B C Fig. 10 Arc shape in CNC IAW. A Y shape; B heart shape; C pencil shape. increasing the wire feed speed to a higher level, the droplet is detached under the tungsten electrode tip. When using the medium and optimized wire feed speed, the droplet is detached in the vicinity of the tungsten electrode tip, as illustrated in Fig. 10. When the detachment occurs above the tungsten tip, as shown in Fig. 10A, the arc shape is characterized as a Y shape, and the arc size is large and its brightness is not intensive enough. The distance between the two electrodes is large, and the arc looks to be elongated. As analyzed in the previous section, the electromagnetic force was not large in this situation. When the droplet was detached under the tungsten tip as shown in Fig. 10C, the arc shape was characterized as a pencil shape. The elongated arc may be the reason for the reduced arc brightness. When the detachment occurred in the vicinity of the tungsten tip as shown in Fig. 10B, the arc shape was considered a heart shape. The distance between the welding wire tip and tungsten tip was approximately minimized. This position is considered to be optimal. The arc shape gradually changed in the metal transfer process, as shown in Fig. 11. The periodic change of the arc shape was just correlated with different stages of droplet detachment. During the initial period (0 9 ms), the droplet began to form at the tip of the welding wire, and the arc evidently did not change. When the droplet grew (9 21 ms), since the arc always searched for the shortest channel for the minimal energy consumption, the position of the arc root on the droplet changed gradually. When the droplet grew to a certain size (21 30 ms), the arc was squeezed between the consumable and nonconsumable electrode. The elongated arc could be observed. In the last stage after the droplet detached (30 33 ms), the elongated arc went back to its original shape. The arc had an intrinsic self-regulation mechanism when the arc length was short (Ref. 31). The arc intrinsic regulation was closely related to the instantaneous wire melting rate, which would fluctuate depending on the different heating positions of the arc (Ref. 32). In the CNC-IAW process, the instantaneous wire feed speed could fluctuate within a small range because of the characteristics of the wire feeder. It could lead to arc length fluctuations. When the instantaneous wire feed speed was lower than the setting one, the arc length would increase, as shown in Fig. 12A. The indirect arc could not enwrap the whole pending droplet before detaching, and the instantaneous melting rate of welding wire decreased a little. In this case, more solid wire would appear in the arc zone to maintain a stable arc length. If the instantaneous wire feed was higher than the setting one, as shown in Fig. 12C, the melting speed would increase and then the length of the solid wire would reduce. The arc would go back to the previous balancing position, as shown in Fig. 12B. As mentioned previously, the distance between the consumable and nonconsumable electrodes was only 1 2 mm. The arc length in CNC-IAW was much shorter than that in conventional GMAW. In this case, the arc intrinsic self-regulation ability would be obvious and important to maintaining a certain arc length. It could benefit the stability of the metal transfer process, and thus of the welding process. Conclusions An indirect arc welding process with consumable and nonconsumable electrodes was proposed, and parameters that determine a stable process were identified; For a given welding current, there was an optimal wire feed speed that could maintain a stable metal transfer process for higher metal transfer frequency and smaller droplet size; The welding current mainly determines the metal transfer mode, droplet size, and metal transfer frequency; The forces acting on the droplet in the proposed CNC-IAW were analyzed to explain and understand the metal transfer phenomena observed; The arc shape had a significant effect on the droplet transfer, and it was related to the wire feed speed and welding current. NOVEMBER 2014 / WELDING JOURNAL 437-s

Fig. 11 Periodic change in the indirect arc shape (150 A, 8.0 m/min). A B C Fig. 12 The schematic of an arc intrinsic self regulation mechanism. A Lower instantaneous welding wire; B normal instantaneous welding wire; C higher instantaneous welding wire. Acknowledgments This work was supported by the National Natural Science Foundation of China (51005101), Jiamusi University Talent Cultivation Project (RC2010-028), and State Key Laboratory of Advanced Welding Production Technology Project (AWJ-M13-04). References 1. Sadler, H. 1999. A look at the fundamentals of gas arc metal welding. Welding Journal 78(5): 45 50. 2. Thomsen, J. S. 2006. Control of pulsed gas metal arc welding. International Journal of Modelling, Identification, and Control 1(2): 115 125. 3. Jones, L. A., Eagar, T. W., and Lang, J. H. 1998. A dynamic model of drops detaching from a gas metal arc welding electrode. J. Physics D: Appl. Physics 31(1): 107 123. 4. Shi, Y., Liu, X., Zhang, Y., and Johnson, M. 2008. Analysis of metal transfer and correlated influences in dual-bypass GMAW of aluminum. Welding Journal 87(9): 229-s to 236-s. 5. Li, K., and Zhang, Y. M. 2007. Metal transfer in double-electrode gas metal arc welding. Journal of Manufacturing Science and Engineering-Transactions of the ASME 129(6): 991 999. 6. Zhang, Y. M., and Liguo, E. 2000. Numerical analysis of dynamic growth of droplet in GMAW. Journal of Mechanical Engineering Science 214(10): 1247 1258. 7. O Brien, R. L. 1991. Vol. 2: Welding processes. Welding Handbook 8th edition. Miami, Fla.: American Welding Society. 8. Wang, G., Huang, P. G., and Zhang, Y. M. 2004. Numerical analysis of metal transfer in gas metal arc welding under modified pulsed current conditions. Metall. Mater. Trans. B 35(5): 857 866. 9. Himmelbauer, K. 2005. The CMTprocess A revolution in welding technology. IIW Doc XII-1875-05, 20 27, IIW. 10. The CMT Process A revolution in materials-joining technology. 2004. Brighton: Fronius USA LLC. 11. Parks, J. M., and Stava, E. K. 1991. Apparatus and method of short circuiting arc welding. U.S. Patent #5,003,154. 12. Stava, E. K. 1993. A new, low-spatter arc welding machine. Welding Journal 72(1): 25 29. 13. Stava, E. K. 2006. Short circuit arc welder and method of controlling same. U.S. Patent #7,109,439. 14. Zhang, Y. M., and Liguo, E. 1999. Method and system for gas metal arc welding. U.S. Patent #6,008,470. 15. Zhang, Y. M., Liguo, E., and Kovacevic, R. 1998. Active metal transfer control by monitoring excited droplet oscillation. Welding Journal 77(9): 388-s to 395-s. 16. Huang, Y., and Zhang, Y. M. 2010. Laser enhanced GMAW. Welding Journal 89(1): 181-s to 188-s. 17. Huang, Y., and Zhang, Y. M. 2011. Laser enhanced metal transfer: Part I system and observations. Welding Journal 90(10): 183-s to 190-s. 18. Huang, Y., and Zhang, Y. M. 2011. Laser enhanced metal transfer: Part II analysis and influence factors. Welding Journal 90(11): 205-s to 210-s. 19. Huang, Y., and Zhang, Y. M. 2012. Nonlinear modeling of dynamic metal transfer process in laser enhanced GMAW. Welding Journal 91(5): 140-s to 148-s. 20. Wang, X., Huang, Y., and Zhang, Y. M. 2013. Droplet transfer model for laser enhanced GMAW. International Journal of Advanced Manufacturing Technology 64(1): 207 217. 21. Xiao, J., Zhang, G. J., Chen, S. J., Wu, L., and Zhang, Y. M. 2013. Active droplet oscillation excited by optimized waveform. Welding Journal 92(7): 205-s to 217-s. 22. Li, K. H. Sept. 2007. Double-electrode gas metal arc welding: Process, modeling and control. PhD dissertation, University of Kentucky. 23. Wang, J., Cao, J., and Feng, J. C. 2010. Microstructure and mechanical performance of depositing CuSi 3 alloy onto 30CrMnSi steel plate by the novel consumable and nonconsumable electrodes indirect arc welding. Material and Design 31(7): 2253 to 2258. 24. Wang, J., Wang, Y. X., and Feng, J. C. 2009. A study on consumable aided tungsten indirect arc welding. China Welding 18(3): 27 31. 25. Wang, J., Feng, J. C., and Zhang, H. T. 2009. TIG-MIG indirect arc welding process. Transactions of the China Welding Institution 30(2): 145 148. 26. Wang, J., Wu, D., Liao, P., Tian, C. Y., Li, M. Q., and Feng, J. C. 2013. Metal transfer and arc behaviour of novel consumable and non-consumable electrode indirect arc droplet welding. Science and Technology of Welding & Joining 18(3): 261 270. 27. Watkins, A. D., Smartt, H. B., and Johnson, J. A. 1992. A dynamic model of droplet growth and detachment in GMAW. International Trends in Welding Science and Technology. Gatlinburg, Tenn.; June 1 5. pp. 993 997. 28. Kim, Y.-S., and Eagar, T. W. 1993. Analysis of metal transfer in gas metal arc welding. Welding Journal 72(6): 269-s to 278-s. 29. Lancaster, J. F. 1984. The Physics of Welding. Pergamon Press, Oxford, England. 30. Choi, J. H., Lee, J., and Yoo, C. D. 2001. Dynamic force balance model for metal transfer analysis in arc welding. Journal of Physics D: Applied Physics 34: 2658 2664. 31. Wen, Y. M., Huang, S. S., and Liu, G. X. 2011. Testing and analysis of high-speed camera for droplet transition. Advanced Materials Research 271: 79 83. 32. Harwig, D. D., Dierksheide, J. E., Yapp, D., and Blackman, S. 2006. Arc behavior and melting rate in the VP-GMAW process. Welding Journal 85(3): 52 62. 438-s WELDING JOURNAL / NOVEMBER 2014, VOL. 93