GMAW waveforms obtained using AMV4000 "weld monitor"

Abbreviations

• WFS = wire feed speed
• TS = travel speed - rate weld progresses
• GMAW = Gas Metal Arc Welding (MIG, MAGS)
• WERC = the Welding Engineering Research Centre at Cranfield University

The activities in general

Had AMV4000 logging at 4000 times per second the current and voltage at that instant.

True average arc power is average of instantaneous powers. That is, instantaneous current x instantaneous volts.

In maths, that's
P_avg = sum[i=1->i=n](V_i * I_i)/n

"CMT" fairly high WFS

WFS=6m/min (TS=1m/min) true avg. arc power = 2800W

• overall trace
• 3 voltage cycles
• peak of middle cycle

Shows in reality spraying - concurs work Pickin in WERC - 3 spikes are 3 spray droplets detaching. Not how Fronius would have you believe "CMT" works throughout the range

"CMT" high WFS

WFS=7m/min (TS=2m/min) true avg. arc power = 3800W
If is behaving like previous, those 8 spikes mean 8 spray droplets per cycle. Still must be getting a dipping action though - only way can have current but no voltage (see all-datalog plot). However, looks to be dominatingly a spray-transfer process. Realistically, has to be. It'll "spray" if it's above the spray-transition Volts and Amps...

CMT in low WFS "true CMT" range

WFS=3.5m/min (TS=1m/min) true avg. arc power is 870W

There's 1kW of laser power, spot-size 5mm, 4mm arc-to-laser axis, arc leading, going onto this weld. Varying laser power doesn't appear to make a difference, but will declare it. On this 5mm thick sample, doesn't make a fully-formed weld - so no "zero laser". Just telling you for what it's worth...

Model "CMT" behaviour, by the looks of it.

Pulse high WFS

WFS=7m/min (TS=2m/min) true avg. arc power = 3400W

Note that for this high WFS the GMAW-Pulse arc power is less than the "CMT" arc power...

Pulse low WFS

WFS=3.5m/min (TS=1m/min) true avg. arc power is 1570W

As per CMT 3.5m/min, there's 1kW of laser power, spot-size 5mm, 4mm arc-to-laser axis, arc leading, going onto the weld. With this arc heat and 1kW laser On 5mm thick sample, does make fully-formed weld. No "zero laser". No effect of changing laser power, but again, declaring what is...

You can see in the "full-weld" plot that there's shorting - where the red Volts trace dives to Zero - the vertical red lines. Particularly around "5000". So the "Pulse" isn't fully correctly set-up Pulse-mode. The shorting gave spatter, as you'd expect.

Second plot during shorting, third plot during smooth Pulse operation.

Done alright here?

R. D. Smith, 25 April 2011