*4.2. The Projected Transfer Zone*

If the arc length were long enough, the droplets would have enough space to drop into the molten pool after detaching from the wire. As a result, short circuits would disappear, and all the droplets would come into the projected transfer zone. The welding formations of Tests 3–6 in the projected transfer zone are displayed in Figure 8. As the arc length increased, the surface formation gradually deteriorated. The specific performances are as follows: the surfaces of Tests 3 and 4 were relatively straight and beautiful. The bead of Test 5 was generally smooth, but the edges of the bead were slightly crooked, with about a 1 mm offset. The surface formation of Test 6 is pretty poor. It can be clearly seen in Test 6 that the bead was uneven, and the edges were severely twisted, being visibly asymmetrical. Dozens of droplets dropped directly outside the molten pool and became large spatters near the bead. In addition, from the perspective of the brown welding fumes covered on the surface of the workpiece, the fumes were almost parallel to the bead on both sides. As the arc grew, there were more and wider fumes.

The relationship of arc length and parameters of cross section formation in the projected transfer zone is shown in Figure 9. With the increase in arc length, the penetration depth and weld reinforcement declined marginally, whereas the weld width rose a little. The greater spread of the plasma arc can cause the energy to be dispersed over a larger area. Figure 10 is the gray-scale image (256 levels) of long and short arc morphology captured at the same peak current time. The average gray levels of arc near molten pool (box region) in Figure 10a,b were 254.82 and 223.25, respectively. As we know, the higher the gray level is, the higher the arc brightness is, and the higher the current density is. Therefore, the current density of shorter arc near molten pool in Figure 10a was higher than that of the longer arc in Figure 10b. It can be also seen in Figure 10 that the arc width will increase with the increase of arc length, and the arc will be more dispersed, contributing to the expansion of heat source radius and heat dissipation of arc. As a result, the penetration depth declined and weld width rise. Additionally, because the speed of wire feed and welding are constant, weld reinforcement declined with the wider weld width.

**Figure 8.** The formation of weld bead in the projected transfer zone.

**Figure 9.** The relationship of arc length and parameters of cross section formation in the projected transfer zone.

**Figure 10.** The gray-scale images of short and long arc morphology captured at the same peak current time: (**a**) Test 3; (**b**) Test 6.

Moreover, in Figure 8, the offset between the deepest point of penetration and the center of weld bead (penetration offset for short) significantly increased with the growth of arc length. The wire was not completely straight when fed out from the contact tube, because it was bent into a disc shape before welding for convenience of transportation and storage. Additionally, during real welding operation, wire inevitably inclines to a certain extent, which is not completely perpendicular to the plane of the workpiece. The end of the wire might be perpendicular to the workpiece, but the inclination of wire would increase with the growth of arc length. In addition, the droplets tend to move along the axis of welding wire in the projected transfer [23]. Therefore, the direction of droplet movement cannot be completely perpendicular to the plane of the workpiece. The rise of arc length increased the distance between wire tip and molten pool. Under the same transverse velocity, droplets could deviate from the center of the weld bead more greatly with the longer arc. Since the heat and mass carried by the droplets deviated from the center were transferred to the molten pool on one side, and the increasing magnetic fields generated by the plasma column likely caused rotation, contributing to an asymmetric driving force within the molten pool, the symmetry of penetration was broken, eventually leading to penetration offset. A similar phenomenon was also observed in Ref. [31]. A more serious deviation of the droplets and a greater penetration offset would be obtained with a longer arc length. Overall, a phenomenon was observed whereby the penetration offset in pulsed GMAW became more sensitive with the increase of arc length.
