sponsored by the american welding society and the welding

Introduction

Gas metal arc welding (GMAW) is cur-rently one of the most widely used weld-ing methods due to its productivity (as aresult of using an automatically fed con-sumable electrode) (Refs. 1, 2) and its con-venience for mechanized/roboticapplications. The transfer of the meltedwire (electrode) onto the base metal is aprocess referred to as metal transfer. Agood understanding of this metal transferprocess and its mechanism plays a funda-mental role in effectively using/improvingthis welding process for production of bet-ter welds at higher productivity and hasthus been an active area of research anddevelopment in the welding community(Refs. 3–7).

A key issue is that of how the metaltransfer largely depends on the weldingcurrent, which also determines other crit-ical parameters including heat input andarc pressure. An application may require apreferred metal transfer mode that needsto be produced using a particular weldingcurrent while this current may result in aheat input and arc pressure that is notsuitable for this application.

For example, many applications preferthe metal transfer to take place in thespray mode, but it requires a currenthigher than the transition current (Ref. 8)to produce in conventional GMAW. In thismode, the arc pressure is (Ref. 9)

( 1 )where the arc current density

( 2 )and μ0 is the permeability, Ra is the arccurrent radius, R is the equilibrium radiusof the droplet, ∈0 is the amplitude of theperturbation, ω is angular frequency, t isthe time, k is a wave number, z representsthe axial coordinate of a cylindrical coor-dinate system, α is the ratio of arc to theliquid current density, I is the current, andβ is the ratio of arc to the liquid radius(Ref. 9). It can be seen that the arc pres-sure is proportional to the square of thewelding current. Increasing the weldingcurrent thus increases the arc pressure. Anextremely high arc pressure is often not ac-ceptable for many applications.

Pulsing current has been an effectivemethod to achieve the spray transfer at aneeded heat input determined by themean current. However, the peak currentin the pulse still must be higher than thetransition current (Ref. 8). The high arcpressure issue aforementioned still re-mains, and an extremely high arc pressureoften blows the liquid metal away from theweld pool and possibly causes meltthrough (in complete joint penetration ap-plications). Further, the high peak currentitself also increases undesirable fumes(Ref. 10).

The patented surface tension transfer(STT) method (Refs. 11–13), which ad-justs the current waveform reactivelybased on the particular stage during theshort-circuiting transfer process, is an ef-fective method to reduce spatter to a min-

P JR R

R t kza a

a= ( ) − − ∈

( ) ( )

⎡

⎣⎢⎢

⎤

⎦μ

ω02

2 204

2/

cos cos⎥⎥⎥

J I R Ra = −( )+( )α απ β π/ 2 2 21

SUPPLEMENT TO THE WELDING JOURNAL, OCTOBER 2011Sponsored by the American Welding Society and the Welding Research Council

Laser Enhanced Metal Transfer — Part 1: System and Observations

A laser impinging on the droplet in gas metal arc welding applies an auxiliarydetaching force without any significant change in current

BY Y. HUANG and Y. M. ZHANG

KEYWORDS

GMAWLaserMetal TransferDrop GlobularDrop SprayShort CircuitingTransition CurrentWelding Current

Y. HUANG and Y. M. ZHANG ([email protected]) are with the Department ofElectrical and Computer Engineering, Universityof Kentucky, Lexington, Ky.

183-sWELDING JOURNAL

WE

LD

ING

RE

SE

AR

CH

ABSTRACT

Laser-enhanced gas metal arc welding (GMAW) is a recent modification of con-ventional GMAW that applies a relatively low-power laser to the droplet. A systematicseries of experiments were designed and conducted to test this modification. A high-speed camera recorded the metal transfer process during each experiment. The be-haviors of the laser-enhanced metal transfer process observed from high-speed imageswere analyzed using the established physics of metal transfer. The characteristics anduniqueness were identified. In all experiments, the laser was found to affect the metaltransfer process as an additional detaching force that tended to change a short-circuiting transfer to drop globular or drop spray, reduce the diameter of the dropletdetached in drop globular transfer, or decrease the diameter of the droplet such thatthe transfer changed from drop globular to drop spray. In addition, this force also pro-vided an effective method to minimize the wandering of the droplet of a relatively largediameter and thus to control the location where it merged into the weld pool. As a re-sult, the uncontrolled drop globular transfer in conventional GMAW as characterizedby large droplets and poor weld formation was changed to a controlled drop globularprocess with improved droplet directionality and weld formation. The large currentrange associated with drop globular transfer, which required pulsing to change to spraytransfer for practical applications, could now be used without pulsing. Desired heatinput and current/arc pressure waveforms may thus be delivered by GMAW throughlaser enhancement.

Huang 10-11_Layout 1 9/9/11 5:25 PM 83

imum with low heat input and arc pres-sure, but its effective range is restricted bythe mandatory need for the particular cur-rent waveform/range and may not be suit-able for other applications that requireother transfer modes and/or different cur-rent waveforms and amperage ranges (typ-ically for higher heat inputs). Zhang, et al.proposed a patented method (Refs. 14, 15)to use a peak current much lower than thetransition current to produce the desiredspray transfer by taking advantage of themomentum of a downward (away from thewelding gun) droplet. Methods based onmechanically assisted droplet transferhave also been proposed/developed toproduce the spray transfer below the tran-sition current (Refs. 16, 17), but the weld-ing gun size and weight are greatlyincreased.

The American Welding Society classi-fies metal transfer into three majortypes/modes: short-circuiting, globular,and spray (Ref. 8). The International In-stitute of Welding (IIW) further classifiesglobular transfer into drop globular andspelled globular (Refs. 18, 19). The metaltransfer process is governed by the forcesexerted on the droplet. In dynamic-forcebalance theory (DFBM) (Refs. 19, 20),five major forces were used to analyze themetal transfer process. Surface tension isthe main retaining force to support thedroplet, while the gravitational force, elec-tromagnetic force, aerodynamic drag

force, and momentum force typically tendto detach the droplet. In the short-circuit-ing transfer mode, the detaching force,mainly the gravitational force, is not largeenough to balance out the retaining force;the droplet would touch the weld pool. Inthis case, the merging of the droplet intothe weld pool is critical in determining theproduction of possible spatter and the for-mation of the weld. For the globular trans-fer mode, as the repelled globular typicallygenerates severe spatter, only the dropglobular transfer may be adopted in appli-cations. In the drop globular transfer, asthe droplet cannot be detached at a rea-sonable small diameter, large and oscillat-ing droplets may be expected inconventional GMAW that cause not onlypotential arc instability/fluctuation, butalso uncontrolled droplet travel directionsthat directly result in the merging of thedroplet with the weld pool at undesired lo-cations to produce poor weld formations.The drop spray transfer mode is usuallycharacterized by uniform droplet diame-ter, regular detachment, directionaldroplet transfer, and it is thus widely usedin the industry.

The metal transfer in GMAW has beentraditionally regarded as a two-stageprocess: first, a droplet forms at the end ofthe solid wire under the arc heating effect;second, the droplet detaches from the endof the welding wire and travels in the arczone. The merging of the droplet into the

weld pool after the travel in the arc-zone isalso a stage in the transfer process but hasnot been much studied. As has been seenabove, the merging is critical as it deter-mines the process stability (short-circuit-ing transfer) or the capability to producegood weld formations (drop globular orspray transfer). To emphasize this, the au-thors add merging as the third stage forthe convenience of analysis in this paper.

Existing methods aforementioned(Refs. 10–17) are “neat” using smart ap-proaches to resolve different issues anddifficulties, but being “neat” also restrictstheir applications in wider ranges. Towardthe development of a more generalmethod, the laser-enhanced GMAW asshown in Fig. 1 has been proposed/devel-oped at the University of Kentucky (Ref.21). It adds a relatively low-power laser toconventional GMAW, and the objective isto provide an auxiliary force to help detachthe droplet at a desired diameter with anydesired current that best suits the applica-tion, including future adaptive control ap-plications where the current needs to beadjusted freely as determined by the con-trol algorithm. It is apparent that laser-en-hanced GMAW is fundamentally differentfrom laser hybrid GMAW (Refs. 22, 23)where a laser beam of substantially highpower aims at the base metal rather thanthe droplet. In a previous paper (Ref. 21)that first documented laser-enhancedGMAW, the laser recoil pressure force wasdemonstrated to be the additional force tohelp detach the droplet. The presentpaper analyzes its metal transfer mecha-nism in order to understand its uniquenessas a desirable process.

Experimental System and Conditions

Experimental System Setup

The principle of the proposed laser-enhanced GMAW is shown in Fig. 1. Alaser beam aims at the droplet. The inten-tion is to detach the droplet using the laserrecoil pressure as an auxiliary detaching

OCTOBER 2011, VOL. 90184-s

WE

LD

ING

RE

SE

AR

CH

Fig. 1 — Principle of laser-enhanced GMAW (Ref.21).

Fig. 2 — System installation parameters. A — Installation parameters; B — installation parameters forthe system.

Fig. 3 — Experimental system. A — Olympus high-speed camera; B — GMAW setup and laser combi-nation (the shield board is not shown in the picture).

A

BA

B

Huang 10-11_Layout 1 9/9/11 5:26 PM Page 184

force to compensate for the lack of theelectromagnetic force or gravitationalforce associated with a relatively small am-perage that is needed for a particular ap-plication, rather than provide additionalheat to speed the melting of the wire. Theassociated additional heat from the lasershould be negligible in comparison withthat of the arc used.

Figure 2 shows the specified parame-ters to realize the laser-enhanced GMAWsystem used in this paper. To conduct thelaser-enhanced GMA process in an ex-pected way, parameters need to be set ap-propriately. As shown in Fig. 2, threeparameters should be determined: contacttube-to-workpiece distance d1, angle be-tween laser beam to GMAW welding gunθ, and the distance from the point wherethe laser intersects the wire axis d2. Stan-dards to set these parameters are found inRef. 21. Experimental results suggest thatd1 be set around 20 mm, θ be around 60deg for easy installation at the expense ofreducing system compactness, and d2 beset at a range from 3 to 7 mm.

A high-speed camera was used to cap-ture the video of the welding process foroff-line analysis. Figure 3A shows thehigh-speed camera used that is capable ofrecording the metal transfer at 33,000frames per second. A band-pass filter cen-tered at 810 ±2 nm with full width at halfmaximum 10 ±2 nm was used to observethe process and record the images. All im-ages presented in this study were recordedusing the high-speed camera shown in Fig.

3A with this band-pass filter. The University of Kentucky Welding

Research Laboratory possesses a NuvonyxDiode laser ISL-1000L (Fig. 3B) whosefocal beam dimension is 1 × 14 mm andwavelength is 808 nm. When this laser isused, less than 1⁄14 of the laser beam can beapplied onto the droplet to generate therecoil force to detach the droplet as the di-ameter of the wire is 0.8 mm, and the di-ameter of droplet may be just slightlygreater. For this research, the efficiency ofthe laser was not a primary concern, andthe use of a laser of larger power andlarger focal zone should not affect the ef-fectiveness of the experimental results.

Figure 3B shows the arrangement ofthe laser in relation with the welding gun.In this experimental setup, the laser beamis aligned with the wire. In order to pro-tect the end of the laser from possible con-tamination from fumes, a shielding board(not shown in Fig. 3B) is added betweenthe laser and welding gun, and the laser isprojected through a hole on the shieldingboard to the wire.

Experimental Conditions

A CV (constant voltage) continuouswaveform power supply was used to con-duct experiments. Pure argon was usedand the flow rate was 12 L/min (25.4 ft3/h).The workpiece was mild steel, and experi-ments were done as bead-on-plate at atravel speed of 6.6 mm/s (15.6 in./min).The wire used was ER70S-6 of 0.8 mm

(0.03 in.) diameter. The distance from thecontact tube to the workpiece was 20 mmas aforementioned.

The welding voltage was set at four lev-els: 26, 28, 30, and 32 V. For each voltage,four different wire feed speeds, 250, 300,350, and 400 in./min, were used to producedifferent welding current levels, resultingin 16 sets of experimental conditions. In allexperiments, welding currents were notmore than 135 A, which will generateshort-circuiting or repelled globular trans-fer or non-wire-axis drop globular in con-ventional GMAW. The laser beam wascontinuously applied along the wire (solidand droplet) at four different levels oflaser intensities for each of the 16 experi-mental conditions: 0, 46 W/mm2, 54W/mm2, and 62 W/mm2. There were thus atotal of 64 experiments conducted. Forconvenience, the parameters will be pre-sented as a set (wire feed speed, voltage,laser intensity). Figure 4 shows the meancurrent measured in all experiments. Itcan be seen that all the currents werelower than the transition current, which isapproximately 150 A (Ref. 8) for the wirematerial and diameter. The current in-creases significantly as the voltage settingincreases because of the reduced wire ex-tension. However, the effect of the laseron the current is insignificant, no morethan 5 A.

Metal Transfer

The diameter of the detached dropletis obtained from the series of high-speedimages in this study. All images presentedhave the same dimension scale except forthose presented individually. The time in-terval of consecutive images in the sameseries is constant. Figure 5 illustrates thescene in a typical metal transfer image.

Observations

Figure 6A shows a typical metal trans-fer cycle for the experiment conductedusing wire feed speed, voltage, and laserintensity equal to 300 in./min, 30 V, and 0.This is a short-circuiting transfer in which


WE

LD

ING

RE

SE

AR

CH

Fig. 4 — Welding current under different wire feed speeds and different laser powers. A — Wire feed speed at250 in./min; B — wire feed speed at 300 in./min; C — wire feed speed at 350in./min; D — wire feed speed at400 in./min.

B

C D

A

Fig. 5 — Metal transfer image.


the second and third stages of the metaltransfer are combined. From Fig. 4, thecurrent in this experiment was approxi-mately 110 A. In the cycle shown in Fig.6A, the combined detaching force fromthe electromagnetic and gravitationalforce was not sufficient enough to balanceout the retaining force, i.e., the surfacetension that is determined by the surfacetension coefficient and diameter of thewire, before the droplet touched the weldpool. The transfer was short circuiting andspatter was produced. Examination ofrecorded images during this experimentshowed that the short circuiting transferdominated, although the globular transferalso occurred occasionally.

Figure 6B is a typical metal transfercycle from the comparative experimentwith the laser at an intensity of 62 W/mm2.As can be seen, the large droplet did nottouch the base metal before detaching,and there was no spatter produced. This isa free flight transfer type, and it is dropglobular according to IIW classification(Refs. 18, 19). Examination of all imagesshowed that all the metal transfer oc-curred as drop globular. It is apparent thelaser made the difference in changing themetal transfer.

Per Ref. 21, the recoil pressure is themajor force the laser applies to thedroplet. Application of a laser beam to adroplet at an appropriate direction as in


WE

LD

ING

RE

SE

AR

CH

Fig. 6 — Typical metal transfer in comparative experiments with and without laser (300 in./min, 30 V, 62W/mm2). A — Without laser; B — with laser.

Fig. 8 — Typical metal transfer in comparative experiments with and without laser for 250 in./min, 30 V,0 W/mm2, and 250 in./min, 30 V, 62 W/mm2. A — Without laser; B — with laser at 62 W/mm2.

A

B

Fig. 7 — Current waveforms for 300 in./min, 30 V,0 W/mm2, and 300 in./min, 30 V, and 62 W/mm2.A — No laser; B — with laser at 62 W/mm2.

A

B

A

B


this study ensures the recoil pressure to bea detaching force. The added detachingforce from the laser recoil pressure re-duces the need from other sources for thedetaching force. When the electromag-netic force is given, the added detachingforce from the laser recoil pressure re-duces the needed gravitational force tobalance out the surface tension. As a re-sult, the diameter of the droplet neededfor detachment is reduced. If the neededdiameter is reduced sufficiently such thatthe droplet can grow to this diameter be-fore it touches the weld pool, the short-circuiting transfer changes to a free flighttransfer type as observed in Fig. 6B.

For these two comparative experi-ments, the laser does not change the meanwelding current significantly as can beseen from Fig. 4B. However, as the dropletdoes not touch the weld pool, the fluctua-tion of the welding current is reduced ascan be seen in Fig. 7. Further, because thedroplet is detached before touching theweld pool, the average transfer time is re-duced from 183.3 ms without the laser to178.3 ms with the laser. The average di-ameter of droplet decreases from 2.23 mmwithout laser to 1.89 mm with the laser.The laser thus reduced the needed diame-ter (weight) of the droplet for detachmentand changed the metal transfer type.

Figures 8–10 are typical metal imagesfrom three additional groups of compara-tive experiments using different wire feedspeeds at 30 V. Because of the changes inthe wire feed speed, the mean currentvaries from experiment to experiment(Fig. 4A, C, D).

First, the typical metal transfer asshown in Figs. 8A and 9A for 250 and 350in./min without the laser was all short-circuiting transfer and a significantamount of spatter was produced. Whenthe laser was applied, as can be seen fromFigs. 8B and 9B, the metal transfer,changed to drop globular transfer andspatter was not found. As the mean weld-ing current did not increase (Fig. 4A andC), it was the authors’ opinion that it wasthe laser recoil pressure that effectivelychanged the type of the metal transfer. Inaddition, the changes in the metal transferresulted in less fluctuating welding currentas shown in Figs. 11 and 12, and the metaltransfer process was thus more stable.

Second, when the wire feed speed in-creased to 400 in./min such that the cur-rent increased, the short-circuitingtransfer no longer dominated. Figure 10Ashows a consecutive transfer processwhere a short-circuiting transfer followeda drop globular transfer. This was typicalin the experiment with 400 in./min withoutthe laser, and different from other experi-ments in the series at the same voltage butlower wire feed speeds where the short-circuiting transfer dominated. The in-

creased mean current was the major rea-son for the frequent occurrence of thedrop globular transfer, but the weldingcurrent fluctuated into relatively low lev-els (Fig. 13A) also produced short-circuit-ing transfers from time to time. When thelaser was introduced, short-circuitingtransfers no longer occurred and transfersbecame totally free flight, as shown in Fig.10B. The droplet diameter became similarto that of the electrode wire and transferwas close to drop spray. As can be seen inFigs. 4D and 13, the mean current and cur-rent levels did not increase by the laser. Itwas the laser recoil pressure that effec-tively changed the metal transfer modefrom a mix of short-circuiting and dropglobular to the drop spray and reduced thefluctuation in the welding current.

Analysis

As has been observed above, the appli-cation of the laser changed the metaltransfer. If the metal transfer in conven-tional GMAW is short circuiting, the ap-plication of the laser at the intensity usedcould change it to the drop globular trans-fer. (The authors believe that it may fur-ther change to the spray transfer as longas the intensity of the laser is sufficient.)When a mix of short-circuiting and globu-lar transfers dominates, it may change to

the drop spray even with the laser inten-sity used. When the drop globular couldbe obtained, the laser reduced the diame-ter of the droplet detached. In all cases,the diameter of the detached droplets wasdecreased as further shown in Fig. 14. Thelaser recoil pressure was identified as themajor cause of these observed changes.

To analyze further, let’s recall that inconventional GMAW, the major sources ofthe detaching force are the gravitational,electromagnetic, aerodynamic drag, andmomentum forces, while the major retain-ing force is the surface tension at the in-terface of the solid wire and liquid droplet(Refs. 19, 20). When the diameter of thewire and material are given, this surfacetension can be considered constant be-cause the temperature at the interfaceaforementioned is the melting point andchanges with neither the welding currentnor the application of the laser. When thewelding current is lower than the transi-tion current such that the current exitsfrom the droplet around its bottom, theelectromagnetic force as a detaching forceis relatively small. The aerodynamic dragforce and momentum force are typicallyrelatively small and often negligible inanalysis such that there is a need for alarge gravitational force to balance out thesurface tension for detachment. In thiscase, as shown in Fig. 6 and Figs. 8–10, the


WEL

DING

RES

EARC

H

Fig. 9 — Typical metal transfer in comparative experiments with and without laser for 350 in./min, 30 V,0 W/mm2, and 350 in./min, 30 V, 62 W/mm2. A — Without laser; B — with laser at 62 W/mm2.

A

B


diameter of the droplet is larger than thatof the wire.

More specifically, when the wire feedspeed is low, such as 250 to 300 in./min,the droplet needs to grow to acquire a suf-ficient mass to produce a sufficient gravi-tational force to balance out the surfacetension. However, before this large massis obtained, the droplet touches the weldpool because of the relatively slow growth(due to the relatively small current and archeat). The metal transfer is dominated bythe short-circuiting transfer. When thewire feed speed/welding current increases,for example to 350 in./min, such that thewelding current and electromagnetic forceincreases, the mass needed to balance outthe surface reduces. However, if this re-duced mass needed is still not achieved be-

fore the droplet touches the weld pool, thetransfer will still be short circuiting. Inlaser-enhanced GMAW, the laser recoilpressure is added to the detaching force,and the needed mass is reduced. If theneeded mass is produced before thedroplet touches the weld pool, the metaltransfer would change from short circuit-ing to drop globular or effectively reducethe diameter of the droplet detached. Asshown in Fig. 14, all the diameters ofdroplet in laser-enhanced GMAW aresmaller than their respective counterpartsin conventional GMAW. As long as thereis a large enough laser recoil pressure(laser intensity), the drop globular and,the authors believe, drop spray would beobtained. To verify the latter, a larger in-tensity laser is needed.

Further, when the wire feed speed in-creases, such as to 400 in./min, the transferwill be dominated by a mix of globular andshort-circuiting transfer in conventionalGMAW. For the laser intensity applied,the short-circuiting transfer in conven-tional GMAW will change to drop globu-lar in laser-enhanced GMAW. The dropglobular in conventional GMAW could re-main or change to the drop spray. In bothcases, the diameter of droplet detached re-duces in laser-enhanced GMAW, and thedrop spray occurs when the diameter re-duces to a level close to that of the wire.

Controlled Drop Globular Transfer

In the laser-enhanced GMAW experi-ments conducted in this study, drop glob-ular was a major metal transfer mode. Theauthors found the drop globular transferwith an enhancement from a laser behavesdifferently from those without a laser en-hancement in conventional GMAW.

Figure 15 shows two images in a cycle ofdrop globular transfer in conventionalGMAW at 400 in./min, 30 V, and 0 W/mm2.Figure 15A is the image of the dropletshortly before its detachment. It is foundthat the center of the sphere of the dropletis not exactly along the axis of the wire. Infact, as long as the droplet is not detached,the center of the sphere oscillates, as shownin Fig. 10A. The trajectory of the detacheddroplet is thus not fixed. It is not fixed alongthe axis of the wire and may change fromcycle to cycle. As a result, the transverse lo-cation where the detached droplet mergeswith the weld pool is not fixed and maychange from cycle to cycle. The images inFig. 15A and B demonstrate this uncontrol-lability of the droplet merging location as-sociated with a droplet globular transfer inconventional GMAW. This type of dropglobular is referred to as uncontrolled dropglobular transfer in this study.

Figure 16 are counterpart images, ofthose in Fig. 15, under (400 in./min, 30V, 62W/mm2). They show the droplet shortly be-fore its detachment and merging into the


WE

LD

ING

RE

SE

AR

CH

A B

Fig. 10 — Typical metal transfer in comparative experiments with and without laser for 400 in./min, 30 V, 0 W/mm2, and 400 in./min, 30 V, 62 W/mm2. A —Without laser; B — with laser at 62 W/mm2.

Fig. 11 — Current waveforms for 250 in./min, 30 V, 0 W/mm2, and 250 in./min, 30 V, 62 W/mm2. A —No laser; B — with laser at 62 W/mm2.

Fig. 12 — Current waveforms for 350 in./min, 30 V, 0 W/mm2, and 350 in./min, 30 V, 62 W/mm2. A —No laser; B — with laser at 62 W/mm2.

A

A B

B


weld pool in drop globular transfer with alaser enhancement. It is apparent that thislaser-enhanced drop globular transfer dif-fers from its counterpart in conventionalGMAW. The center of the droplet sphere isapproximately along the axis of the wire.There is indeed a slight deviation of thiscenter from the axis, but observation andanalysis of images in different cycles show:1) its magnitude and direction are both con-sistent in different cycles; and 2) this slightconsistent deviation is away from the direc-tion of laser application. It is apparent thatthis deviation is caused by the laser recoilpressure. Because the droplet is approxi-mately along the axis and the slight devia-tion is consistent in magnitude anddirection, the transverse location of themerging is also consistent, slightly awayfrom the axis of the wire. As can be seen, theapplication of the laser brings certain con-trols to the drop globular transfer, and the

resultant drop globular becomes a con-trolled drop globular.

Drop globular is seldom used in indus-try (Ref. 8), and its unfixed droplet trajec-tory in a natural/uncontrolled form may bethe major reason. However, there is a lackof effective solutions in literature. In laser-enhanced GMAW, the trajectory of thedroplet is controlled by the laser recoilpressure, and the merging of the dropletin drop globular transfer becomes con-trollable. It is the laser that made the dropglobular become controllable in this study.Laser enhanced drop globular transfer isa controlled drop globular transfer, but itis possible that a controlled drop globularmay also be achieved using other means.

Compared to the uncontrolled dropglobular process, a controlled drop globu-lar process produces welds more consis-tently because of the controlled/consistentdroplet trajectory and transverse merging

location. As can be seen in Fig. 17A, a typ-ical weld produced by uncontrolled dropglobular lacks control on the transverse di-rection. Spatter is also found becausesome droplets may merge into the weldpool at its edges (Ref. 8). Rough weld sur-faces are also found because of the unfixedpositions where large droplets merge intothe weld pool. In laser-enhanced GMAW,all issues, weld direction inconsistency,spatter, and rough weld surfaces, are re-solved by the controlled/consistent droplettrajectory, controlled/consistent/appropri-ate merging location, and reduced dropletsize, as shown in Fig. 17B. Quality weldsmay thus be produced by the laser-en-hanced GMAW at a controlled drop glob-ular transfer, and drop globular thus maybecome a valid process for applicationswhere the current requires desired wave-forms or need to be below the transitioncurrent.


WE

LD

ING

RE

SE

AR

CH

Fig. 13 — Current waveforms for 400 in./min, 30 V, 0 W/mm2, and 400 in./min, 30 V, 62 W/mm2. A — Nolaser; B — with laser at 62 W/mm2.

Fig. 17 — Typical surface appearance in comparative experiments with and without laser at 400 in./min, 30 V, and 62 W/mm2. A — Without laser; B — withlaser at 62 W/mm2.

Fig. 15 — Images of uncontrolled drop globular process. A — Before detach-ment; B — merging into the weld pool.

Fig. 14 — Droplet sizes at 30 V and different wirefeed speeds.

Fig. 16 — Images of controlled drop globular process. A — Before detach-ment; B — merging into the weld pool.

A

A

A

AB

B

B

B


Conclusions

• An experimental system has been es-tablished, and a series of 64 sets of exper-iments have been designed and conductedto symmetrically study the laser-enhancedGMAW.

• The laser aiming at the droplet inlaser-enhanced GMAW applies an auxil-iary detaching force on the droplet with-out a significant change in the current.

• Free flight transfers could be suc-cessfully produced at continuous currentsfrom 90 to 135 A with a 0.8-mm-diametersteel wire without spatter.

• Laser-enhanced metal transferprocess is also governed by the establishedphysics of metal transfer except for there isa need to include the additional detachingforce generated by the laser.

• If the metal transfer is short-circuit-ing transfer in conventional GMAW, laser-enhanced GMAW may change it to dropglobular transfer. If conventional andlaser-enhanced GMAW both producedrop globular, the latter reduces the di-ameter of the droplet. If the metal transferis short-circuiting or drop globular trans-fer in conventional GMAW, laser-enhanced GMAW may become the dropspray. The established physics of metaltransfer can explain all these changes bycounting the additional detach force in-troduced by the laser.

• Controlled drop globular transfer inlaser-enhanced GMAW offers desirablemetal transfer characteristics that benefitthe formation of quality welds.

• Controlled drop globular transfer ex-tends the capability of the productiveGMAW process into the range of constantcurrent that conventionally produces un-desirable drop transfer, which is not mostsuitable for practical use.

• Laser enhancement provides an ef-fective method to achieve a controlleddrop globular transfer and to enableGMAW to use a constant current in an in-

creased range to meet the requirementsfor different applications.

Acknowledgment

This work was funded by the NationalScience Foundation under grant CMMI-0825956 entitled Control of Metal Trans-fer at Given Arc Variables.

References

1. Sadler, H. 1999. A look at the fundamen-tals of gas arc metal welding. Welding Journal78(5): 45–50.

2. Thomsen, J. S. 2006. Control of pulsedgas metal arc welding. International Journal ofModelling, 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 froma gas metal arc welding electrode. J. Physics D:Appl. Physics 31(1): 107–123.

4. Shi, Y., Liu, X., Zhang, Y. M., and John-son, M. 2008. Analysis of metal transfer andcorrelated influences in dual-bypass GMAW ofaluminum. Welding Journal 87(9): 229-s to 236-s.

5. Li, K., and Zhang, Y. M. 2007. Metaltransfer in double-electrode gas metal arc weld-ing. Journal of Manufacturing Science and Engi-neering-Transactions of the ASME 129(6):991–999.

6. Zhang, Y. M., and Liguo, E. 2000. Nu-merical analysis of dynamic growth of dropletin GMAW. Journal of Mechanical EngineeringScience 214(10): 1247–1258.

7. Kim, Y. S., and Eager, T. W. 1993. Analy-sis of metal transfer in gas metal arc welding.Welding Journal 72(6): 269-s to 278-s.

8. O’Brien, R. L. ed. 1991. Welding Hand-book Vol. 2: Welding Processes. 8th edition,Miami, Fla., American Welding Society.

9. Rhee, S., and Kannatey-Asibu, E. 1991.Analysis of arc pressure effect on metal transferin gas-metal arc welding. Journal of AppliedPhysics 70(9): 5068–5075.

10. Wang, G., Huang, P. G., and Zhang, Y.

M. 2004. Numerical analysis of metal transferin gas metal arc welding under modified pulsedcurrent conditions. Metall. Mater. Trans. B 35(5):857–866.

11. Parks, J. M., and Stava, E. K. 1991. Ap-paratus and method of short circuiting arc weld-ing. U.S. Patent #5,003,154.

12. Stava, E. K. 1992. System and method ofshort circuiting arc welding. U.S. Patent,#5,148,001.

13. Stava, E. K. 1993. A new, low-spatter arcwelding machine. Welding Journal 72(1): 25–29.

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 mon-itoring excited droplet oscillation. Welding Jour-nal 77(9): 388-s to 395-s.

16. Wu, Y., and Kovacevic, R. 2002. Me-chanically assisted droplet transfer process ingas metal arc welding. Journal of EngineeringManufacturing 216: 555–564.

17. Yang, S. Y. 1998. Projected droplettransfer control with additional mechanicalforces (AMF) in MIG/MAG welding process.PhD dissertation, Harbin Institute of Technol-ogy.

18. Ponomarev, V., Scotti, A., Silvinskiy, A.,and Al-Erhayem, O. 2003. Atlas of MIG/MAGwelding metal transfer modes. IIW Doc. XII-1771 to 1775-03, Bucharest.

19. Lancaster, J. F. 1984. The Physics ofWelding. Pergamon Press, Oxford, UK.

20. Choi, J. H., Lee, J., and Yoo, C. D. 2001.Dynamic force balance model for metal transferanalysis in arc welding. Journal of Physics D: Ap-plied Physics 34: 2658–2664.

21. Huang, Y., and Zhang, Y. M. 2010.Laser enhanced GMAW. Welding Journal 89(9):181-s to 188-s.

22. Engstrom, H., Nilsson, K., Flinkfeldt, J.,Nilsson, T., Skirfors, A., and Gustavsson, B.2001. Laser hybrid welding of high strengthsteels. Proc. ICALEO: 125–134.

23. Bagger, C., and Olsen, F. O. 2005. Re-view of laser hybrid welding. Journal of LaserApplications 17(1): 1–14.


WE

LD

ING

RE

SE

AR

CH

Authors: Submit Research Papers Online

Peer review of research papers is now managed through an online system using Editorial Man-ager software. Papers can be submitted into the system directly from the Welding Journal pageon the AWS Web site (www.aws.org) by clicking on “submit papers.” You can also access the newsite directly at www.editorialmanager.com/wj/. Follow the instructions to register or login, and makesure your information is up to date. This online system streamlines the review process, and makesit easier to submit papers and track their progress.

By publishing in the Welding Journal, more than 66,000 members will receive the results of yourresearch. Additionally, your full paper is posted on the American Welding Society Web site forFREE access around the globe. There are no page charges, and articles are published in full color.By far, the most people, at the least cost, will recognize your research when you publish in theworld-respected Welding Journal.


sponsored by the american welding society and the welding

Documents