10 A Novel High-Efficiency Keyhole Tungsten Inert Gas …
339
Fig. 10.21 Weld pool oscillation process during K-TIG welding [31]
with the conversion of the molten metal from a convex to a concave surface. The
oscillation process is shown in Fig. 10.21.
10.3.2 Acoustic Signal Characteristics
Figure 10.11 shows the collected arc sound signal. Two dominating noises exist in
the monitored arc sound signals. First is the DC bias voltage due to the zero drift of
signal conditioner and the DC coupling circuit. The second is the ambient noise from
KUKA controller, the cooling water source and cooling fans of the welding power
source, etc.
The power spectrum of the noise and arc sound signal is shown in Fig. 10.22.
The energy of noise mainly focuses on 0–8 kHz, and it encounters with another peak
in 16 kHz. It accounts for the ambient noise of arc sound signal during most of the
frequency bands. A conclusion is able to be drawn that the dominating noise in arc
sound signal is the environmental noise illustrated in Fig. 10.22. A simple passband
filter could not work well and denoise the arc acoustic signal.
Fig. 10.22 Power spectrum of arc acoustic signal and noise [34]
339
Fig. 10.21 Weld pool oscillation process during K-TIG welding [31]
with the conversion of the molten metal from a convex to a concave surface. The
oscillation process is shown in Fig. 10.21.
10.3.2 Acoustic Signal Characteristics
Figure 10.11 shows the collected arc sound signal. Two dominating noises exist in
the monitored arc sound signals. First is the DC bias voltage due to the zero drift of
signal conditioner and the DC coupling circuit. The second is the ambient noise from
KUKA controller, the cooling water source and cooling fans of the welding power
source, etc.
The power spectrum of the noise and arc sound signal is shown in Fig. 10.22.
The energy of noise mainly focuses on 0–8 kHz, and it encounters with another peak
in 16 kHz. It accounts for the ambient noise of arc sound signal during most of the
frequency bands. A conclusion is able to be drawn that the dominating noise in arc
sound signal is the environmental noise illustrated in Fig. 10.22. A simple passband
filter could not work well and denoise the arc acoustic signal.
Fig. 10.22 Power spectrum of arc acoustic signal and noise [34]