*3.2. Variation in Arc Profile*

In order to better observe and analyze the variation in arc profile collected by the high-speed video camera, it is necessary to divide the arc region for regionalization and measurement. The arc image processing procedure is shown in Figure 6. The original arc image is shown in Figure 6a. Figure 6b is the colored gray-scale image that is transformed from the original arc. According to the intensity of the arc light, the arc can be divided into 256 scale levels. Yang et al. illustrated that the gray scale was larger with higher intensity in the arc gray-scale image, with 90% of the arc intensity as the arc core region [16,17]. The gray-scale image was colored by RGB for the convenience of arc core region [18,19], the red and green region of processed arc image were the core and edge region, respectively. As a result, the arc profile would be measured and analyzed using the arc core region, as presented in Figure 6b. The arc profile, which was defined by arc core length L and arc core width under the nozzle D, was accurately measured by computerized measurement technique.

**Figure 6.** Edge extraction and regionalization of welding arc. (**a**) Original arc image; (**b**) processed arc image.

Figure 7a displays the variation of arc profile in a complete plasma gas cycle of PPG-VPPAW with the 4 Hz plasma gas pulse frequency in EN phase. Figure 8a shows the corresponding L and D, which were measured using the method mentioned above. From 0 to 30 ms, the arc images were relatively stable, the L and D were about 3.76 mm and 1.05 mm, respectively. When the plasma gas was shut off, from 30 ms to 60 ms, the arc constricted rapidly with the L and D decreased to 0 mm. From 60 ms to 80 ms, the core region of the arc almost disappeared because there was no plasma gas supply. When the plasma gas was turned on, from 80 ms to 100 ms, the core region of the arc significantly increased with L and D increased to 3.18 mm and 1.27 mm, respectively. Then, the L had a decrease to 2.07 mm at 110 ms and increased again to 6 mm in 140 ms. After that, the L gradually decreased to 4.09 mm from 140 ms to 170 ms and then maintained in a stable state. In this period, the D only had a slight change due to the restriction of the torch orifice. The above results show that the arc core length had a short decrease during the general rising trend after plasma gas was turned on. These two peaks of arc core length were due to the initial overshoot phenomenon. Once the plasma gas is turned on, the overshoot velocity is higher than the setting value, which causes the subsequent plasma gas to not keep up, leading the L to decrease.

Figure 7b shows the variation of arc profile in a complete plasma gas cycle of PPG-VPPAW with the 20 Hz plasma gas pulse frequency in the EN phase. The corresponding L and D are shown in Figure 8b. It can be seen that the change of PPG-VPPAW arc profile with 4 Hz and 20 Hz plasma gas pulse frequencies have the same tendency. As mentioned before, due to the increase of plasma gas pulse frequency, there is not enough shut-off time of plasma gas to affect the arc profile fully, the L cannot decrease to 0 mm from 8 ms to 18 ms, and the first peak value is also less than that in 4 Hz. In the whole cycle with 20 Hz plasma gas pulse frequency, the core region of the arc always exists, and the D keeps balanced at 1.02 ± 0.17 mm.

**Figure 7.** Comparison of arc images at negative polarity stage. (**a**) PPG-VPPAW (4 Hz); (**b**) PPG-VPPAW (20 Hz). (The color bar of this figure is same as Figure 6).

**Figure 8.** Arc core region size of PPG-VPPAW. (**a**) PPG-VPPAW (4 Hz); (**b**) PPG-VPPAW (20 Hz).

Figure 9 displays the arc profile of EP phase in different VPPAW processes and stages. Figure 9a shows the arc profile in the VPPAW process. Figure 9b,c show shutting-off and turning-on stages, respectively, with 4Hz PPG-VPPAW. Figure 9d,e show shutting-off and turning-on stages, respectively, with 20 Hz PPG-VPPAW. Compared with these arc profiles, it is easy to find that only in the shut-off stage with 4 Hz PPG-VPPAW process, the arc core length L significantly decreased. This phenomenon cannot be found when the plasma gas pulse frequency increased to 20 Hz. That means the influence of plasma gas shutting off in the EP phase of the VPPAW process is much smaller than in the EN phase, especially in a high plasma gas pulse frequency. This is because the arc core region is directly related to the temperature field, which has such a great inertia that the temperature change could not quickly respond to the disturbance. When the plasma gas frequency increases, the effect of pulsed plasma gas on the EP phase is harder to observe.

**Figure 9.** Comparison of arc images at positive polarity stage. (The color bar of this figure is same as Figure 6).
