Robots and dials and knobs—oh my!

GMAW power sources have evolved over time

May 29, 2003
By: George J. Williams

In the late 1950s, the U.S. Navy wanted to find a way to join heavy aluminum structural sections used to fabricate motor torpedo boat hulls.

It has been said that necessity is the mother of invention. You be the judge.

In the late 1950s, the U.S. Navy wanted to find a way to join heavy aluminum structural sections used to fabricate motor torpedo boat hulls.

Gas tungsten arc welding (GTAW) offered suitable process capabilities, but it was prohibitively slow in application. At Air Reduction's research laboratory in Murray Hill, N.J., technicians performed basic tests, welding several lengths of 36-inch-long, 1/16-in.-diameter cold aluminum filler wire end to end and then pushing them through a copper tube that was connected to a motor generator set. When technicians used welding-grade argon as a shielding gas, they established an arc on an aluminum plate, and gas metal arc welding (GMAW) was born.

It appeared that the Navy's search had come to an end.

During the next several months, technicians transformed a seemingly successful laboratory investigation into a high-performance fabricating tool.

An early assessment of the components needed to produce a manual welding tool included a sturdy, nonmetallic, pistol grip-type gun; a means for feeding a continuous soft wire; its spooled source; and a nonmetallic guide tube to the gun. A welding power source, a set of variable-speed drive rolls, and a nozzle arrangement to focus the shielding gas would complete the package.

Subsequent refinements to these mechanisms included lubricating the nylon guide tube with powdered graphite, water cooling the welding cable to the gun to increase its current-carrying capacity, and dealing with the welding current pickup problems.

Almost all current pickup problems result in burnback, which happens when the welding arc lengthens to a point at which the molten wire fuses itself into the exit end of the current pickup contact tube. No matter its cause, burnback typically results in wire feed problems that require some of the equipment to be disassembled to clear the wire train, a time-consuming job.

The original power sources used for GMAW were constant-current machines, which led to many application problems. High open-circuit voltage (OCV) often caused arc-starting problems. However, the most persistent problem resulted from varying current pickup locations within the contact tube. The cast in the spooled wire caused these variations.

Cast is a word used to describe the natural and unrestricted diameter that a length of filler wire assumes after it has been cut from the spool. Because this diameter changes with usage, the location of the current pickup varies over the length of the pickup tube. Minor arcingis possible at those locations, which causes wire feed problems.

To correct or lessen the problem, a shallow bend was formed at the exit end of the contact tube, which negated much of the resistance heating by providing a stronger electrical pickup. A short time later, the bend was replaced by a 1/2-in.-long, semicylindrical ceramic insert, which maintained light pressure on the moving wire by holding it in place with a silicon O-ring.

By now much had been learned about GMAW operating parameters. But one area remained difficult to cope with: the burn-off rate, the output welding current and the wire feed speed that would produce a sound welding deposit.

Despite some breakthroughs, a welding power source with certain design features tailored for GMAW still was needed: a direct-current, reverse-polarity machine with a low open-circuit voltage capable of 300 or more amps and sufficient reactance in the secondary to markedly slow down the response time of any minor arc variations.

Finally, GMAW was moving forward. Constant potential (CP) power supplies had taken the guesswork out of setting the burn-off rate. Since the power source would provide sufficient welding current to burn the wire evenly regardless of its feed rate, the welder no longer needed to preselect these values.

What remained was the need to detect a sound weld deposit. Layer- and level-wound spooled wire solved the wire feed problems, and arc instability was virtually a thing of the past.

Meanwhile licensees of the GMAW process had received numerous requests for equipment packages that would allow welders to weld aluminum as thin as 1/8 in. This would necessitate using wire diameters of 1/32 in.

For a brief period it was like starting over.

Wire feed systems had to be redesigned, power sources needed further refinement, and arc starting and stopping became more critical because of the speed at which the wire had to travel. Instant acceleration and deceleration of small-diameter soft wire also presented some interesting problems, which is another story altogether.

During this the late 1950s and early 1960s, all of the old process and application problems were reduced to repeatable and programmable data so that today we enjoy dials, knobs, and robots and comments like, "How can you not love welding?"

Because of a specific need, and because other welding processes had specific shortcomings, the research technicians chose aluminum for their testing. While carbon steel and stainless steel were candidates for GMAW, they had both process and marketing problems.

Of the process problems, porosity and arc instability were the most severe. Using double- and triple-deoxidized filler wire and adding other gases to the basic inert shielding gas overcame most of these problems. However, despite the advantages of feeding hard wire, the cost of consumables and the availability of coated electrodes for both carbon and stainless steel made GMAW a hard sell for many years.

Bear in mind that coated electrodes was a multibillion-dollar industry, and advancements like low-hydrogen and iron-powder additives were helping to preserve it.

Advances in GMAW technology eventually broke down some of the coated electrode competition, but that also is a story for another time.

George J. Williams is business development manager of Trico Metal Products Inc., 2309 Wyandotte Road, Willow Grove, PA 19090, 215-659-2673, fax 215-659-9140,

George J. Williams

Contributing Writer

Related Companies