*4.1. The Partial Projected Transfer Zone*

The welding formation of Tests 1 and 2 in the partial projected transfer zone are displayed in Figure 4. It can be seen that the surface formation of the weld bead in Test 2 was better than that in Test 1. The edge of bead was crooked, and there are lots of big spatters around the bead in Test 1. In contrast, the edge of the bead was relatively straight, and there were few big spatters in Test 2. The centerline of penetration was useful to measure, but a larger difference can be noted looking at the area of the fusion zone. Although the highlighted profile looks similar in terms of cross section formation, the weld penetration of Test 2 was only 0.1 mm deeper than that of Test 1. Moreover, from the perspective of the brown welding fumes covering the surface of the workpiece, the fumes were obviously fewer and mainly distributed within 1 cm around the weld bead in Test 2, in which the bright steel plate far away from the weld bead was clearly visible. However, the surface of the workpiece was almost covered by fumes, and the regions of bright steel plate were nearly invisible in Test 1.

**Figure 4.** The formation of the weld bead in the partial projected transfer zone.

According to the high-speed photographs of Tests 1 and 2, except for projected transfer, two types of short circuit [24] were observed: normal short circuit (see Figure 5) and instantaneous short circuit (see Figure 6). It is necessary to point out that the short circuit phenomenon in pulsed GMAW is different from that in short circuiting transfer using a constant direct-current (DC) power source. According to Ref. [29], these short circuits in pulsed GMAW, as in Figures 5b and 6b, are also called meso-spray transfer. During meso-spray transfer, the position of necking is between the undetached droplet and solid wire, followed by short circuits. However, during short circuiting transfer using constant DC power source, the position of necking is at the liquid bridge, which is formed after short circuits.

During the normal short circuit (Figure 5), the neck between the undetached droplet and the wire gradually shrank, and the undetached droplet contacted the molten pool at the time P1, with a dramatic fall of voltage. Then, from the next three frames of P1 (3.1–3.3 ms), it can be seen that the bright arc was extinguished because of the short circuit and the minimum position of short circuiting liquid bridge was broken by explosion. A few spatters flied out of the explosive position, with an unstable arc and an abnormal voltage peak.

**Figure 5.** The normal short circuit: (**a**) The welding current and voltage waveform; (**b**) the high-speed photographs of interval AB.

When the arc was reignited, an unstable arc and an abnormal voltage peak were found. This is because at the time of arc reignition after short circuiting, cathode spots were reformed in the center of the weld pool surface. During the process of reforming the arc and cathode spots, the concentration of the cathode spots and the increase of the cathode surface work function led to a rise in the potential gradient across the cathode fall space and the adjoining contraction space. Consequently, the arc voltage became abnormally high despite the short arc length, which can be regarded as a signal of arc reignition after short circuiting [24,30].

A similar phenomenon is also shown in Figure 6. Even though the duration of the short circuit was too short to be captured by the 10,000 fps high-speed camera, the arc behavior and the abnormal voltage peak can indicate an instantaneous short circuit at the time Q1 in Figure 6. During the process of instantaneous short circuit, the undetached droplet contacted the molten pool and was bounced off instantaneously at the time Q1. The high current near the pulse peak current leaded to huge impact force of explosion, which contributed to the split of droplet and various big spatters.

(**b**)

**Figure 6.** The instantaneous short circuit: (**a**) The welding current and voltage waveform; (**b**) the high-speed photographs of interval CD.

To research the cumulative effects of micro droplet transfer on macroscopic weld bead formation, the percentages of various types of droplet transfer were calculated based on each 200 pulses of Tests 1 and 2 (shown in Figure 7). Even through the average arc length in Test 1 (1.5 mm) was only 0.8 mm shorter than that in Test 2 (2.3 mm), their percentages of various types of droplet transfer were totally different. From Test 1 to Test 2, the percentage of projected transfer rise sharply from less than 10% to almost 90%. However, the proportion of the normal short circuits and the instantaneous short circuits fall from 61% and 30% to 9.5% and 2.5%, respectively. Moreover, based on the analysis of Figures 5 and 6, above, a few small spatters resulted from the normal short circuits, while a large

number of spatters (especially the big spatters) resulted from the instantaneous short circuits. The longer arc allowed the droplets more time to gain enough energy to fully liquefy, so that they could impact the molten pool without short circuiting and causing explosions. Therefore, spatters declined obviously when the arc length grew a little from Test 1 to Test 2.

**Figure 7.** The percentages of various types of droplet transfer in the partial projected transfer zone: (**a**) Test 1; (**b**) Test 2.
