WELDING RESEARCH -s 253 WELDING JOURNAL ABSTRACT. A three-dimensional numeri- cal simulation of heat and fluid flow and phase change in pulsed gas metal arc weld (GMAW-P) deposits was used to study hump formation and its suppression by a hy- brid laser welding process. The simulation results allowed definition of the events lead- ing to the formation of a humped bead. In the initial stage of hump formation, which occurred at relatively high (in terms of Pé- clet number) travel speed, a thin elongated molten bead was formed and pinching due to capillary instability resulted in a dramat- ically reduced cross section. Solidification then divided the molten pool into front and back sections, guaranteeing hump forma- tion. Conditions for formation of a hump at the beginning of a weld bead and formation of subsequent humps were different, but the events leading to hump formation were the same for both. Experimental and simula- tion results for the hybrid process showed that a defocused laser beam located in front of the GMA weld pool could decrease the size of the initial bead hump and prevent formation of subsequent humps. This was the case if a laser melt pool of sufficient width was created in front of the weld pool to make the broadened weld bead stable from a capillary point of view. Conversely, simulations showed that hump formation was not prevented by the hybrid process if the laser beam did not sufficiently widen the weld bead. Introduction The formation of weld bead humps often limits the maximum achievable welding travel speed and hence process productivity. The term “humps” refers to the undulations of the weld deposit seen, for example, in the pulsed gas metal arc weld (GMAW-P) bead images in Fig. 1. Because of its role in limiting welding pro- ductivity, past investigations have sought to understand the physical phenomena in- volved in bead hump formation and to de- velop process modifications that suppress it. In this work, hump formation and its prevention by a second heat source are studied by numerical simulation. A thor- ough review of prior literature has re- cently been published (Ref. 1). For brevity, the most relevant work will be summarized here. In the earliest experimental studies of weld bead humping in GMAW, Bradstreet (Ref. 2) identified the humping defect as one of welding defects that occurred at high travel speeds. Based on weld images, capillary instability as well as weld metal flow and wetting were identified as factors in hump formation. Capillary instability has been recognized by other researchers as a fundamental mechanism for bead hump formation. Lord Rayleigh’s analysis (Ref. 3) showed that a liquid cylinder with radius r having periodic axisymmetric per- turbations of wavelength λ whose ampli- tude exceeds a critical value λ/2πr = >1 are unstable and identified wavelengths having the greatest growth rate. The capillary stability of liquids and molten materials deposited on solid sub- strates has also been analyzed (Refs. 4–6 ) and results show that capillary instability can lead to the formation of undulations. The stability of a long molten bead on a flat substrate with parallel, immobile con- tact lines is determined by the contact angle. If the internally measured contact angle is greater than π/2, the molten bead is unstable to height perturbations with longitudinal wavelength λ greater than a critical value, which then leads to the pe- riodic separation of a molten bead. The critical wavelength for instability de- creases as wetting angle increases and be- comes equal to the wavelength predicted by Rayleigh stability analysis in the limit where the deposit does not wet the sub- strate at all. Deposits with fixed contact lines and contact angles less than π/2 are stable. It is also mentioned that heat trans- fer and nonisothermal wetting, for exam- ple as analyzed by Schiaffino and Sonin (Ref. 5), are important to wetting, spread- ing and final shape of molten material de- posited on a cool substrate. Other researchers dealing specifically with weld bead humping (Refs. 7–9) have noted that relatively rapid weld pool fluid flows are observed in images of GTAW and GMAW weld bead hump formation and have proposed explanations for humping in terms of these molten metal flows. The work of Nguyen et al. (Ref. 9) thoroughly characterized the strong back- ward fluid flow observed during the GMAW process. The momentum of trans- ferred filler metal droplets is identified as a source of the observed weld pool metal flow. Mendez and Eagar (Ref. 8) also identified the strong backward fluid flow as a dominant factor in hump formation for high-current GTA welding, but pres- sure and plasma “jet” flow at the arc anode were identified as the main driving forces that caused the backward flow. Analysis has shown that this same high flow rate of arc plasma over the pool sur- face of GTA welds may cause Kelvin- Helmholtz instability of the surface, lead- ing to hump formation (Ref. 10). It must SUPPLEMENT TO THE WELDING JOURNAL, SEPTEMBER 2007 Sponsored by the American Welding Society and the Welding Research Council KEYWORDS Hump Formation Hybrid Process Laser Beam Welding Numerical Simulation Pulsed Gas Metal Arc Weld Pool Simulation Study of a Hybrid Process for the Prevention of Weld Bead Hump Formation A numerical simulation shows how formation of bead humps in high-speed GMAW is prevented by additional laser heat input BY M. H. CHO AND D. F. FARSON M. H. CHO AND D. F. FARSON are with The Ohio State University Welding Engineering Pro- gram, Columbus Ohio.