4 6 8 10
GMAW-P RapidArc STT CMT CMT-P FastROOT
Figure 2.40 – Variation of arc power with actual WFS for all waveforms.
2.4.3. Arc Characteristics
Arc characteristics are generally characterized by the variation of arc length.
Kenney et al. (1998) suggested that arc length provides a traceable view of the physics of the welding process. Although the measurement of arc length is very difficult and often results in errors, several measurements using high speed video images achieve reasonable accuracy. The variation of arc length is a response to the process setting parameter, waveform design and instability phenomena associated with the metal transfer. The comparison of the variation of arc length with WFS level for different waveforms is presented in Figure 2.41.
In general arc length slightly increases with WFS (and consequently with arc current and voltage), but a drop in arc length is observed at high WFS levels for RapidArc, GMAW-P and CMT-P, which are characterized by pulse spray transfer. This phenomenon is a consequence of the differences in the mechanism of metal transfer and will be discussed in the next section. A detailed analysis of the variation of arc length with WFS for RapidArc for higher WFS levels shows that after the transition region of metal transfer, arc length decreases progressively until reaching buried levels (Figure 2.42). It should be noticed that these results were obtained for a specific CTWD of 11mm.
70
Actual WFS [m/min]
0 2 4 6 8 10 12
Arc Length [mm]
0 2 4 6 8
10 GMAW-P
RapidArc STT CMT CMT-P FastROOT
Figure 2.41 – Variation of arc length with actual WFS (measured) for all waveforms.
Actual WFS[m/min]
0 5 10 15 20
Arc Length [mm]
0 2 4 6 8
10 2.5%CO2 97.5%Ar - WFS/TS =16
2.5%CO2 97.5%Ar - WFS/TS =18 1.5%CO2 54%He 44.5%Ar - WFS/TS =16 1.5%CO2 54%He 44.5%Ar - WFS/TS =18
Figure 2.42 – Comparison of the variation of arc length with actual WFS (measured) for RapidArc for different shielding gases and WFS/TS ratios.
The results demonstrate that RapidArc, CMT and CMT-P present significantly high arc length values, around 4mm. GMAW-P presents values around 3mm, while for FastROOT and STT the lowest values are identified, around 2mm. As observed, the most significant variations of arc length are identified for the processes where pulse spray transfer mode is identified.
It is also noticed that, in general, short-circuiting transfer waveforms presents a much lower arc length than spray transfer waveforms. However, CMT is an exception to this observation, and the synchronization between short-circuiting transfer with high arc length levels results from the automatic wire feed speed control system, where wire feeding and retracting controls arc length.
71 The analysis of the effect of different setting parameters on arc length was analysed for all waveforms, respectively the arc length adjusting parameter, CTWD and shielding gas composition.
The variation of arc length with arc length adjusting parameter was compared for all waveforms using a bar graph with the average and RMS variation of several measurements, in five steps, from the lower to the upper limits (Figure 2.43). The results show that variations of arc length are more significant for RapidArc, CMT-P and GMAW-P, and these waveforms are characterized by spray transfer mechanism. For the waveforms characterized by short-circuiting, i.e. CMT, STT and FastROOT, no significant variations of arc length are observed. It is observed that the shielding gas has often a significant effect on arc length, increasing when 1.5%CO2 54%He 44.5%Ar is applied. This effect has been identified in the literature and result from the properties of the gas mixture, in particular the effect of the ionisation potential and density (Eagar 1981) (Kim 1989) (Modenesi 1990).
Arc Length [mm]
0 1 2 3 4 5 6
GMAW-P RapidArc STT CMT CMT-P FastROOT
2.5%CO2 97.5%Ar 1.5%CO2 54%He 44.5%Ar
Figure 2.43 – Comparison of average and RMS variation of arc length for all waveforms, obtained by changing the arc length adjusting parameter (in five steps from the minor to upper limits) for two different shielding gases (WFS set of 6m/min).
It has been pointed out that MIG welding is conventionally operated at constant voltage to provide self-adjusting arc length. This self-adjusting results in small fluctuations in arc voltage. In this case, arc current influences the overall process characteristics, in particular the mode of metal transfer, arc stability control and fusion characteristics.
In Pulsed GMAW spray transfer is obtained by an adjustment of both peak current and time to determine at least the desirable one drop per pulse. In this case the mean arc current, obtained from all parameters together must give a burn-off ratio matching wire feed speed, where arc length can be kept approximately constant. If the pulse time or pulse current is inadequate, the arc becomes unstable with consequent variations of arc length and metal
72
transfer phenomena. This phenomenon will be discussed later but may account for some defects during welding.
It has been pointed out that the increase of arc length is proportional to the increase of arc voltage (American Welding Society 2007). The relationship between arc voltage and arc length was established for all waveforms considering constant wire feed speed (6m/min) and constant arc length correction (nominal values of adjusting arc length parameter). The results demonstrate that arc length in general increases with the increase of arc voltage adjusting parameter was varied in five steps from the lower to upper limits (WFS set of 6m/min).
Arc Voltage [V] varied from low (around 3-4m/min) to high levels (around 8-10m/min) using nominal arc length adjusting parameters for all waveforms.
73 The arc voltage increases with arc length adjusting parameter, but not significantly, for most of the processes evaluated (Figure 2.44). However, this behaviour is not observed for RapidArc, where the variation of trim from lower to upper limits produced a significant increase of the arc length comparable to the increase of arc voltage. Within these results it is found that the variation of arc length is generally smaller, less than 1mm for most of the waveforms. Larger variations are however verified for the RapidArc waveform, about 3mm.
However, the small variations of arc length are enough to generate important variations in arc and waveform characteristics, which can account for significant changes in melting phenomena and resulting bead shape geometry.
It was shown that, in general, the increase of CTWD increases the arc length. However, this behaviour depends on the power source characteristics, waveform design and WFS level applied. The effect of CTWD on arc length was assessed for three levels (11, 13.5 and 16mm); average and RMS variation were calculated and the results presented in bar graphs (Figure 2.45). As observed from the results illustrated in the Figure 2.46, the effect of CTWD on the arc length is much significant for RapidArc and GMAW-P. However, it should be noticed that these variations are applied for 6m/min (WFS set). The variation of arc length associated with the CTWD for RapidArc suggests that this process works in controlled current (CC) mode, while the remaining processes appear to operate in controlled voltage (CV) mode, with self adjusting of stick out when CTWD is varied.
Arc Length [mm]
0 2 4 6 8 10 12
GMAW-P RapidArc STT CMT CMT-P FastROOT
Col 6: --
Figure 2.46 – Comparison of the average and RMS variation of arc length for all waveforms, obtained by changing CTWD (at three levels: 11, 13.5 and 16mm) (WFS set of 6m/min).
From the results presented in Apendix VI is observed that at high WFS levels, arc length varies more significantly (i.e., about 4mm difference from 11mm to 16mm CTWD when WFS set was 10m/min) with CTWD for CMT-P. In contrast, for RapidArc this variation is more significant at lower wire feed speed levels (i.e., 4mm difference between 11 and 16mm CTWD for 4 and 6m/min), rather than at high WFS (2.5mm difference for 8m/min and 1mm difference for 10m/min).
74