October 21, 2026 · Indumor
Arc welding is like writing with a pencil that wears down as it writes: the tip melts and becomes part of what is written. That idea — that the electrode does not just heat, but melts and fills the joint — took almost a century to appear, and about as long again to become reliable.
The whole story fits in one paragraph. In 1802, Vasily Petrov produced a stable electric arc in St. Petersburg and proposed using it to melt metals; Humphry Davy made it famous in London and gave it its name. Between 1881 and 1882, Nikolay Benardos and Stanisław Olszewski welded with a carbon arc. In 1888, Nikolay Slavyanov in Russia and Charles Coffin in the United States, each on his own, used a metal electrode that melted. In 1904, Oscar Kjellberg founded ESAB in Gothenburg and gave the electrode a coating that shields the weld pool. And in the twentieth century, ships taught two lessons: that a hull could be built without a single rivet, and that the steel had to be tough so a crack would not split it in two.
In the spring of 1802, Petrov built a battery of about 2,100 copper-zinc elements and drew a stable arc between two pieces of carbon. He described it in 1803 and suggested using it for lighting and for melting metals, but he published only in Russian and his work was forgotten for nearly a century. Davy showed a continuous arc at least four inches long in London, and the curved shape of the flame gave it its name.
Eighty years later, according to welding histories, the Frenchman Auguste de Méritens was joining the lead plates of storage batteries with a carbon arc. Benardos and Olszewski went further with the Elektrogefest, named after Hephaestus: a carbon rod in a holder, the workpiece as the other pole, and an arc that could join, cut, drill holes and build up metal. They patented it across much of Europe from 1885 — the first patent was French — and in the United States in 1887, and it was used in railway workshops.
In 1888, at the Perm cannon works, Slavyanov replaced the carbon with a metal rod that melted and filled the joint. He called it “electric casting of metals” and first used it to repair the shaft of a steam engine. That same year, in Detroit, Charles Coffin filed a patent on the same idea, granted in January 1889; the two reached it independently. Shortly before, in 1886, Elihu Thomson had patented another route in Massachusetts: resistance welding, the family that spot welding belongs to.
The trouble with the bare electrode was air: the weld pool picked up oxygen and nitrogen, and the weld came out porous and brittle. Oscar Kjellberg knew the problem well: he had spent some fifteen years working aboard steamships. In 1904 he founded ESAB in Gothenburg and covered the electrode with a coating that, as it melts, forms gas and slag that shield the pool and steady the arc; he patented it in Sweden in 1907 and in Germany in 1908. In England, Arthur Strohmenger and the Quasi-Arc company patented electrodes with a blue asbestos covering, and in Cleveland, around 1911, Lincoln Electric was already selling its first variable-voltage welding machines.
In our shop, even today, the first thing done with a new box of low-hydrogen electrodes is to put it in the oven: the coating that shields the pool also soaks up moisture, and moisture is hydrogen in the weld. More than a hundred and twenty years later, we are still looking after Kjellberg's invention.
In 1920, Cammell Laird launched the Fullagar at Birkenhead, about 150 feet long and without a single rivet: for Lloyd's Register, the world's first fully welded ocean-going ship. In 1924 she ran aground and her bottom was pushed up about 11 inches over some 70 feet. She stayed watertight, reached Liverpool under her own power, and the plating was jacked back into shape instead of being replaced, as a riveted hull would have required.
During the Second World War, the United States built about 2,700 Liberty ships, with welded, prefabricated hulls. The first, the Patrick Henry, took 244 days; by 1944, the average was 42. But the steel of the day turned brittle in the cold, and in a welded hull a crack finds no riveted seam to stop it. On January 16, 1943, the Schenectady, a newly completed T2 tanker, broke in two at its fitting-out dock in Portland, Oregon, with the water at about 39 °F. The investigation left behind what is routine today: requiring toughness from steel with the Charpy test, avoiding corners that concentrate stress and controlling weld quality.
Then came submerged arc welding in the 1930s, TIG in 1941, MIG in 1948 and flux-cored wire in the 1950s. Each one solved something. But the logic is still the pencil that wears down: an arc, a metal that melts and something that shields it from the air.
If a plant shutdown is on the calendar and there is equipment to repair by welding, the procedure and consumables get decided now, not on shutdown day: there is no time to experiment then.
Sources: US patents 363,320 (Benardos and Olszewski), 395,878 (Coffin) and 347,140 and 347,141 (Thomson); Great Soviet Encyclopedia; ESAB; Lloyd's Register Heritage & Education Centre; National Park Service; TWI.
Tell us what needs repairing and we will set the procedure and consumables before the shutdown.
© Indumor 2010–2026 — Industrial steel fabrication, Córdoba, Argentina
