25 July 1997


In the first of a two-part series, ESABs Welding Process Centre superintendent Mick Andrews helps Andrew Pearce shed some light on MIG set-up

MIG sets are a bit like the Roman god Janus: Two-faced. Looking one way is an inviting "easy-to-use-me" smile. Pointing the other is a more mysterious side, labelled "setting up". For while MIG is one of the simplest welding processes to learn in terms of doing the job, twiddling the knobs to optimise plant performance is something else again.

Most farms work mainly with steels, using relatively simple sets and solid filler wire. So its these well centre on. But before delving into setting, its vital to sort out the process options and detail where each is best used.

The operator can often choose how filler wire gets to the weld pool. The mode used – either dip transfer or spray transfer – has big implications for weld pool control, fusion and penetration. Whether there actually is a choice depends on the sets maximum output, the filler wire diameter and the shield gas.

Where the work involves thin sections, positional (ie, not flat) work or gap filling, dip transfer is the mode to use. In this theres not enough voltage for the arc to be self-sustaining, so the weld pool is relatively small and fast-freezing. Heat input is relatively low so distortion can be too, and metal is only transferred during contact between filler wire and the weld pool – never in free flight across the arc.

Fig 1 shows dip transfer happening. Filler wire comes out of the torchs contact tip just faster than the arc can burn it off, so travels down and touches the work. The resulting electrical short heats the wire until a length melts into the weld pool, leaving behind a temporary arc. The process repeats 20- 200times/sec, generating MIGs characteristic crackle.

As voltage and current (wire speed) are increased, the way filler wire metal is transferred to the weld pool changes. More voltage allows a self-sustaining arc, and dip transfer shifts to globular transfer. In this, large blobs of molten wire move across the arc in free flight (Fig 2). Droplets are the same width or wider than wire diameter and break off randomly, changing the smooth rip of arc noise into a rougher popping.

Although heat release is greater as the arc is "on" continuously, globular transfer is not good news. Theres not enough magnetic energy to direct those big droplets to the weld pool, so either they miss it altogether or whack home like a hippo hitting a paddling pool. Either way, weld metal is lost from the joint and spatter goes everywhere.

There is no off/on transition between dip and globular transfers, just a blending from one to the other. When the changeover comes depends on the set, the conditions and the shield gas, so the boundary is not fixed. Whenever it comes in, globular transfer is unwelcome. If youre using an argon/CO2 shield gas mix, sidestepping it is a matter of spotting whats happening and tuning it out – see later. But if youre using pure CO2 you cant avoid globular transfer. Penetration will be good but the higher the current goes, the worse the spatter will get and the rougher the weld will look.

Spray transfer is the third way to freight filler metal to the weld pool, though its available only at higher arc energies. With smaller wires, thats on sets delivering at least 160Amp and using shield gas with minimum 80% argon; with thicker wires, at higher energies still.

Using high voltage and high wire feed speed together produces globular transfer. But by winding back the wire feed speed a little, the wire can be burnt off in a more orderly fashion. Much smaller droplets are pinched off the wires end, flying across the arc in a fine spray (Fig 3).

The high-current arc delivers heat in a wider fan. Fusion and penetration improve, suiting spray transfer to heavier steel sections. The downside is that the hot, fluid weld pool limits spray to flat (or very gently inclined) work on steels.

Noise changes too. As globular transfer turns into spray, its stutter fades to a smooth, low hum. Light and heat output both go up noticeably.

Whats the plan?

First off, decide which metal transfer mode is right for the job – dip for thinner stuff and positional work, spray for heavy sections if welding position allows. Then set the set.

The notion is to balance voltage and wire speed to produce a particular transfer mode, then tweak these settings for the work in hand. Several things are important.

There is no spot-on right setting. Instead, theres a narrow band of good welding conditions to find and move around in.

At the borders of good welding conditions lie transition zones into worse conditions. Steer clear of these.

The calibrations on welding set controls seldom tie in with any specific voltage or wire speed. Theyre just markings, useful as waypoints during setting and for subsequent storage in the operators memory.

Although there are tables laying down the minimum and maximum currents usable with each wire size, these are not much good unless the set has read-outs for volts and amps. Most farm-level sets dont, so unless you rig up your own meters you must fly by the seat of your pants. Table 1 has values for those who can use them.

How do you know when to change wire size? There is no hard rule – just economics and plant capability. As joint size goes up, more wire is needed to fill it. Shifting to a fatter wire delivers the extra metal while keeping feed speed down and forward travel rate up. Bigger wire can support the higher currents needed for heavy work, without moving to globular transfer and its spatter, porosity and tatty finish. At the other end of the scale, the lower usable limit for a given wire comes when the voltage needed to burn it off blows holes in the work.

Its possible to come at good welding conditions from two directions. The precise way is to set wire speed to a given value and then find the right voltage to burn it off, which is ideal where all the variables are known.

But in the suck-it-and-see farm workshop its easier to do things the other way about. Pick a voltage that delivers the heat you want, then find the wire speed range that suits it.

Heres a way to drop into dip transfer. Until you know your set inside out, tune up on a bit of clean scrap steel of the same thickness as the job. Its a low arc-energy mode, so pick a voltage towards the lower end of the scale and set wire speed to about one-third of maximum. Adjust voltage until youre happy with fusion and penetration – see next for judgment criteria.

Fig 4 shows a series of runs on 2mm plate, made with 0.8mm wire in a 165A Murex Tradesmig set. The only thing changed between each run was voltage; wire speed was left alone. Each run corresponds to one voltage step on the set, starting low and working up. See how bead shape and spatter level change?

In MIG, voltage controls arc length – the more volts you dial in, the longer and wider the arc grows. In runs 1 and 2 voltage is too low. The bead is thin, uneven, humped and shows little sign of fusing with the parent plate as heat is lacking. Theres a lot of spatter from weld metal being physically splashed from the pool, and the operator could feel "stubbing" through the torch. This judder sets in as, without enough voltage to burn it off, incoming wire jabs down through the weld pool and into the plate.

With increasing voltage, runs 3-6 blossom into good welding conditions . Now theres enough energy to burn off the wire, so the bead is regular and fuses to the plate. With each step up in voltage, the arc lengthens and widens a little so the bead flattens. By runs 4 and 5 dip transfer is in full swing, the arc is sizzling nicely and spatter has died clean away.

By run 7 the party is nearly over. With this much voltage around, the arc is long and burning well back up the wire. The wider arc produces a wider melt zone. And could you but hear it, your ears would tell you that the rate of sizzle is slowing right down.

In run 8 (maximum voltage) the arc is burning right back to the contact tip in a series of fuzzy-sounding "whooshes". Its so hot that the melt zone is wider than the filler could fill, leaving undercut at the edges. Globular transfer has started so spatter is back, and the whole area has a powdery brown coat.

This bad-to-good-to-bad-condition sequence happens with any wire diameter as voltage at the set is wound up. Fig 5 reinforces the point. Turning over the plate shows penetration; see how it rises and widens with voltage.

Sorry to break off here, but its time. Next month well look at the other half of the equation – wire speed – and set up for spray transfer.

Dip transfer. Wire feeds downward just faster than the low-voltage arc can burn it off, so touches the pool. The resulting short circuit (centre) heats the wire until it fuses into the pool, leaving behind a short-lived arc (far right and left). Metal never moves in free flight across the arc, spatter is low and the fast-freezing weld pool is good for thin sheet, positional work and gap filling.

MIG: Not too hard once the plant is set.

Table 1: Wire sizes, currents and voltages

Wire diameter Current, (mm)min/max (A)





Typical operating ranges;

dip transfer 16-20V

globular transfer 22-28V

spray transfer 24-32V

Note: These values are approximate and will vary with equipment, shield and welding position.

Fig 4 above: The voltage effect. In runs 1 and 2, voltage is too low; the weld bead is narrow and fusion poor. Increasing voltage through runs 3-6 produces good welding conditions; the bead is widening and flattening, fusion is fine and spatter disappears. Run 3 shows cool dip transfer, run 6 hot dip transfer. By runs 7 and 8 there was not enough wire to satisfy the lengthening arc, which burnt progressively back towards the contact tip. The bead is wide and very shallow, undercut appears and spatter comes back as globular transfer creeps in.

Fig 5 right: The reverse of Fig 4. Moving up from low voltage (bottom) to high shows increasing penetration.

Globular transfer. When voltage and wire feed speed (current) are both high enough, a self-sustaining arc appears. But energy in it is not high enough to produce true spray transfer or channel molten filler metal straight to the weld pool. The result? Blobs of filler fall through the arc into the weld pool, or are pushed sideways and out. Spatter is high, weld finish is poor.

Spray transfer. Higher arc energies pinch off filler wire in much smaller droplets, which are taken down through the arc by magnetic field. Spatter is low, the arc long and wide, the weld pool hot and liquid. Penetration and fusion are potentially high.

MIG plant voltage controls arc length and width – set it first. Set control markings seldom correspond to specific values; just use them as a guide.