Waveform analysis for MIG / GMAW transfer modes

Provisional average current and average voltage given

These welds were performed on 5mm steel plate with 1.2mm diameter steel MIG wire and Ar/20%CO2 shielding gas.

Dip transfer

Sample rate 10kHz
(10thousand measurements of V, I and wire-feed-speed recorded every second).

180A 18.7V

Overall view

1105191430P11dip180A18p7V

First sample - 0.1s and 0.01s

1105191430P11dip180A18p7V_23000_24000

1105191430P11dip180A18p7V_23830_23930

Second sample - 0.1s and 0.01s

1105191430P11dip180A18p7V_50000_51000

1105191430P11dip180A18p7V_50690_50790

Fairly much as expected. Assuming zero-volt regions are duration of "dipping"/short-circuiting:

• "short-circuiting" ("dipping") visible - volts go to near-zero for ~1/1000th of second (around +4.5V - the Ohmic V-drop?)
• current rise with steady upward slope during the "dip" (the short-circuit) - due to inductance (or simulation of by inverter power source)
• current fall with steady down-slope when "dip" breaks - also due to inductance (or simulation of)
• voltage sustains at an arc voltage as current stabilises to around 120A for arcing phase
• wire-feed-speed rate bit jagged, presumably due to dipping action

Dip transfer - high current low volts

220A 20V

1105201435P06diphi220A20V

1105201435P06diphi220A20V_30000_31000

1105201435P06diphi220A20V_30180_30280

The higher current, low-voltage dip (higher voltage would give spray transfer at this current).
All comments same as for "normal" dip-transfer in range where only dip transfer is possible at constant voltage (CV).

Spray transfer

Sample rate 10kHz

True values calculated for region of steady weld run:

• true average power 7739W
• average voltage 30.0V, average current 258A

Overall view



0.1s



0.01s



Spray transfer revealed to be exactly as envisaged. The current and voltage hold essentially constant values. The streaming transfer occurs in quasi-static equilibrium, where most other GMAW modes are cyclical.

Pulse - "Lincoln RapidArc" at lower power

18 June 2011
I need to change my entire slant on Lincoln "RapidArc". On understanding how the waveform works, with its "dipping" event, tried the process again and find different range of conditions good. Previously in error trying to find something identical to "pulse transfer". Would be time-consuming to re-write notes. Generally finding good conditions at lower powers than judged best here.

Sample rate of the data-logging is 20kHz (twice as rapidly as for dip and spray).

117A 22.2V

The weld works well, indeed enabling rapid run rates, achieved with no spatter. So the impression is that Lincoln Electric know what they are doing with "RapidArc". The waveform reveals there is a "dipping" event, which can be seen when the voltage goes to near zero. However, occuring at low current, this explains why the event does not cause spatter

1105191535P04pulselo117A22p2V

0.1s and 0.01s

1105191535P04pulselo117A22p2V_6500_8500

1105191535P04pulselo117A22p2V_7000_7200

Even if those quick voltage dips to zero were shorting, they do not result in any spatter.

Pulse - "Lincoln RapidArc" at high power

Sample rate 20kHz

240A 26.6V

1105191340P07pulsehi240A26p6V

0.1s

1105191340P07pulsehi240A26p6V_3100_5100

Both 0.01s sections

1105191340P07pulsehi240A26p6V_4200_4400

1105191340P07pulsehi240A26p6V_10300_10500

Weld recorded as being a good welding condition. The downward transients on the voltage at around the 11000th data point have no duration and caused no spatter or other misbehaviour.

Pulse - "Lincoln RapidArc" at further higher power

Sample rate 20kHz

253A 28.4V

1105191340P05pulsehi253A28p4V

0.1s and 0.01s

1105191340P05pulsehi253A28p4V_8200_10200

1105191340P05pulsehi253A28p4V_8200_8400

Very clean waveform for a weld recorded as being notably very smooth operating during welding.




R D Smith, 24 May 2011, 09 June 2011