When Did Welding Start? A Complete History of Welding from 3000 BC to Today

Standing in a modern fabrication shop in Ohio, watching a robot arm lay a flawless MIG bead on a truck frame, it is easy to assume welding is a 20th-century invention. But the answer to “when did welding start” is roughly 5,000 years old — long before electricity, before acetylene, before anyone ever struck an arc.

When Did Welding Start? The Short Answer and Why It Is Complicated

If you typed “when did welding start” into a search bar, you deserve a straight answer before anything else. Here it is: Welding is roughly 5,000 years old. Metalworkers were joining pieces of metal by heating them and hammering them together at least as far back as 3000 BC, in the Bronze Age, long before anyone understood electricity or combustion chemistry.

But here is the honest complication. Ask ten historians for a single founding date and you will get ten defensible answers, because “welding” has meant different things in different centuries. That is not a cop-out. It is the most interesting part of the story.

What Welding Actually Means (And Why the Definition Decides the Date)

The American Welding Society defines welding as the localized coalescence of materials produced by heat or pressure, or both. Read that phrase twice. “Heat or pressure” is the clause that stretches the timeline dramatically.

If welding requires an electric arc, your answer is the early 1880s. If it requires a flame, the answer moves to the 1890s. But if pressure alone counts, then a Bronze Age smith hammering two overlapping gold plates until they fuse into one solid object was welding, and he did it two thousand years before the birth of Christ. Modern engineering has never revoked his claim. The International Agency for Research on Cancer’s monograph on welding fumes states plainly that metal joining by heating and hammering overlapping pieces dates back at least 5,000 years, using the same principle still applied in forge welding and solid-state welding today (IARC/WHO monograph).

So the date you get depends entirely on where you draw the line. That is why you will see “3000 BC,” “1881,” and “1895” circulating online, all technically correct, all answering slightly different questions.

The Discipline Itself Was Young When It Started Keeping Records

Here is a fact that reframes the whole subject: welding is ancient, but welding history is a modern invention. Historian John Lancaster, in his classification of welding’s development, divides the field into five periods — joining in early cultures (5000 BC to 1200 AD), the dawn of science (1700–1875), the appearance of the major processes (1877–1903), welding as a repair technique (1903–1919), and welding for construction and fabrication (1919–1951). That framework comes from his paper “A Brief History of Welding Technology” and it is the skeleton most serious welding histories are built on.

The gap tells you something. For most of its existence, welding knowledge lived in guild workshops and oral tradition, passed from master to apprentice as a commercial secret rather than documented science. When Vannoccio Biringuccio described welding a broken saw blade in Venice in 1540, he was recording a craft secret in print, not publishing a breakthrough.

Then, within barely a century, the discipline exploded. Jefferson’s Welding Encyclopedia covered just three basic processes in its first 1921 edition. By the late 1990s, AWS literature listed more than 90 welding and allied processes — a thirty-fold expansion in under eighty years. That kind of growth is almost unmatched in any trade (American Welding History).

What You Actually Want to Know When You Ask This Question

Search data and forum patterns tell us people asking about welding’s origins usually want four things, and good articles deliver all four:

• A timeline. Not a single date, but a sequence you can follow from antiquity to now.
• Key dates. The handful of years that genuinely changed the practice.
• Named inventors. Petrov, Benardos, Slavyanov, Coffin, Kjellberg, Adams — real people, often omitted from popular lists.
• Context. Why each change mattered, and what problem it solved.

This article delivers all four, drawn from primary patents, the American Welding Society’s own records, and peer-reviewed historical work.

Why a Single Number Misleads You

Consider a concrete example. The oxy-acetylene torch is often called “the first welding process,” but electric arc welding was demonstrated for joining metal before oxy-acetylene became commercially practical. The arc came first; the chemistry that made flames hot enough followed. Meanwhile, forge welding predates both by millennia. Each date answers a different, narrower question.

Lancaster’s periods make the point elegantly. Welding did not begin once. It began, then stalled for centuries as artisanal craft, then ignited repeatedly — first with science, then with processes, then with industry’s demand for something stronger than rivets.

What follows is that full sequence: the Bronze Age goldsmiths, the medieval monks and Renaissance metallurgists, the Russian and French pioneers who struck the first arcs, the two world wars that forced welding from repair shop to shipyard, and the post-war process explosion that gave us MIG, TIG, plasma, laser and friction stir welding. Every date is sourced. Every inventor is named. The answer to “when did welding start” turns out to be five thousand years long — and it is far better story than any single year could ever be.

3000 BC and Earlier: The Bronze Age, Egyptian Sponge Iron and the First Recorded Welds

To answer “when did welding start” honestly, we have to leave the fabrication shop behind and walk into a royal tomb. In 1920s excavations at the Royal Cemetery of Ur in modern-day Iraq, archaeologists opened the tomb of Queen Pu-abi and found a double-walled gold goblet so finely joined that the seam where the inner vessel meets the outer shell reads as a braze fillet. Nearby, Sumerian swords dating to roughly 3000 BC had been joined by hard soldering. These are not crude accidents of heat. They are deliberate, repeatable joining techniques, practiced by craftsmen who had no word for metallurgy and no theory of temperature, only generations of accumulated results. You can read the full pre-1800 chronology in the pre-1800 welding history timeline.

Egypt, Sponge Iron and the First “Pressure” Weld

The Egyptians took the process somewhere more radical. Lacking plentiful surface iron, they heated iron ore in charcoal fires, reducing it down to a spongy mass of metallic particles, then hammered those particles into a solid mass. Modern metallurgists call this the first recorded example of pressure welding, or solid-phase welding: the metals fuse without ever fully melting. It is a crucial distinction, because it means the oldest known welding technique is not the arc, not the flame, but the hammer.

Evidence of the craft keeps stacking up across the ancient world:

• 1475 BC, Thebes: a wall painting in the tomb of Vizier Rekh-mi-re shows metalworkers brazing, one of the earliest visual depictions of a joining process anywhere.
• ~1000 BC, Egypt: forge-welded iron and bronze objects recovered near the pyramids show hot hammering used as routine production work, not just repair.
• 1000–500 BC, Roscommon, Ireland: small gold boxes called lunulae and related ornaments were pressure-welded by hammering, proving the technique traveled independently across Europe.
• 60 AD, Rome: Pliny the Elder wrote about gold brazing fluxes, documenting the chemical trick of using a cleaning agent to help molten metal wet a surface.

The Iron Pillar of Delhi: A 6-Ton Proof of Skill

If you want one object that settles the ancient welders’ competence argument, it is the Iron Pillar of Delhi. Erected in India between roughly 310 and 400 AD, the pillar stands about 23 to 25 feet high and weighs close to 6 tons, and it was built by blacksmiths forge-welding individual iron billets together, heating them in a charcoal fire and hammering them into one continuous shaft.

What makes it astonishing is not its age but its survival. The weld seams have held for more than 1,600 years in open air, and the pillar is famous for its remarkable resistance to corrosion, a property now attributed to the phosphorus content of the iron and the protective film it forms. For a modern welding engineer, it is a humbling benchmark: a field joint, made without shielding gas, without electrodes, without any of the tools we consider essential, outlasting most of what we build today.

The broader historical survey in Total Materia’s historical development of welding places these artifacts alongside later craft manuals, and the pattern is consistent: ancient joining was empirical, secretive, and far more advanced than the simple “two rocks and a fire” image most people carry.

Why This Era Still Matters

Three ideas from this period never went away. First, heat plus pressure joins metal without melting it, a principle that returned in the 1990s as friction stir welding. Second, flux protects the joint from oxides, the same job today’s slag and shielding gas perform. Third, repeatability separated craft from accident: the Sumerian smith who made the Pu-abi goblet could make another, which means technique was being taught, corrected, and passed down.

So when did welding start? In this stretch of the timeline, the honest answer is around 3000 BC for documented soldered and brazed joints, and perhaps earlier for hammered iron whose evidence has not survived. The starting line is soft, but the direction of travel is not. From hammer and charcoal, the story moves next to the medieval workshop, where written records finally replace archaeology.

From Theophilus to Biringuccio: The Medieval and Renaissance Workshop (11th-16th Century)

After the pressure welds of the Iron Pillar of Delhi and the gold brazing of antiquity, joining methods entered one of the longest quiet periods in their history. For roughly a thousand years between late antiquity and the Renaissance, very little was formally written about joining metals. Knowledge stayed locked in monasteries, passed from master to apprentice, and treated as a guarded craft secret. It was not until the 11th century that these methods began to resurface in written form, setting the stage for the first printed description of welding itself.

The First Written Recipes: Theophilus and Medieval Brazing

The clearest window into medieval metal joining comes from a German monk known only by the pseudonym Theophilus. Writing in the early 11th century, his three-volume treatise ‘De Diversis Artibus’ (On Diverse Arts) is one of the most valuable surviving records of early workshop practice. In its pages, Theophilus recorded detailed recipes for brazing silver, including a flux of potassium tartrate (tartar) and sodium chloride used to join a ~66% silver-copper alloy.

For a modern reader, the significance lies not in the exact formula but in the approach. Theophilus was not theorizing about the atomic structure of metals; he was recording what worked through trial and error. His instructions cover surface preparation, heating, application of the alloy, and cleaning, all hallmarks of a practiced craft. Those empirical methods remained the standard for centuries, passed down orally far more often than they were written onto parchment.

Another quiet shift of the era was the discovery of zinc in 1374. Not directly tied to welding, the new metal would later widen the range of brazing alloys available to European workshops. In the meantime, the core joining methods remained hammer-based pressure work and silver brazing for delicate items.

From Manuscript to Print: Biringuccio’s Landmark 1540 Work

The next major leap came with Vannoccio Biringuccio, an Italian mining engineer and foundryman. In 1540 he published ‘De La Pirotechnia’ in Venice. Often cited by historians as the first printed book to describe welding, it marked a turning point: craft knowledge finally left the cloistered manuscript and entered wider circulation through the printing press.

Biringuccio did not shy away from specifics. His best-known passage details “the secret of welding the fracture of a saw, of a sickle, or of a sword” using low silver, borax, or crushed glass, clamped with hot tongs to bring the broken edges together. This is not brazing in the modern sense of a lower-melting filler merely coating a joint, but a deliberate attempt to unite the base metals through heat and pressure. The method sits firmly in the lineage of forge welding: localized heat, intimate contact, and mechanical force producing coalescence.

By the later Renaissance, other writers echoed similar ideas. Benvenuto Cellini, in his 1568 writings, recorded techniques for brazing fresh silver-copper alloys, showing that these practices had spread well beyond a single workshop. What united these authors was pragmatism. They gave readers usable instructions, yet none attempted to explain why the joint held. Science had simply not caught up with the smith’s hands.

The Word Itself: When ‘Weld’ Entered the English Language

Language lags behind technique, and welding is no exception. According to the etymological record compiled by A History of Welding, the English verb “weld” is first recorded in 1599 as an alteration of “well,” meaning to boil or rise. The noun form did not follow until 1831. That detail reveals something important: for a long time English speakers had no single distinct term for this process, instead saying “soldering,” “brazing,” or “joining” depending on the trade.

The timing is telling. The verb’s first appearance coincides with the spread of printed workshop texts in the late Renaissance, when terms began to standardize as knowledge moved past local dialects. The word dates from the turn of the 17th century; the action it names had already been practised for millennia.

The Long Shadow: Forge Welding’s Modern Legacy

The most enduring lesson of this era is the continuity of the underlying principle. The hot tongs, borax flux, and hammered joint described by Biringuccio descend directly from Bronze Age pressure welds. That is precisely why forge welding is treated as the ancestor of modern solid-state processes like friction stir welding. Both achieve coalescence without fully melting the base metals, relying on heat, pressure, and cleanliness to form a bond.

Blacksmiths of the 16th century could not have imagined a spinning tool plunging into aluminium at thousands of RPM, but they understood the goal: make two separate pieces of metal act as one. Pursued with nothing but fire, flux, and force, that goal bridged antiquity and the Renaissance, and it still echoes in workshops today.

1800-1900: The Electric Arc, Oxy-Acetylene and the Birth of Modern Welding

If the Bronze Age gave us welding’s principle, the 19th century gave it modernity. In a single hundred years, welders gained three things they had never had before: a reliable heat source made of electricity, a fuel gas hot enough to melt steel, and a body of published science explaining why both worked. This is the century that answers “when did welding start” for most people who picture an actual welder at work.

The Science Before the Spark

Antoine Lavoisier’s late-18th-century experiments with oxygen established that the gas supports combustion far more aggressively than ordinary air. That principle is the direct ancestor of oxygen cutting: heat the metal, blast it with pure oxygen, and the metal itself burns away. It took decades to turn the chemistry into a tool, but the idea was on the table before the century began, as the pre-1800 welding history timeline documents.

Then came the electricity. Between 1800 and 1802, Russian academic Professor V.V. Petrov built a working voltaic pile and produced what he called an “electrical arch” — a stable, sustained electric arc between two carbon electrodes. Petrov described its intense light and heat in detail, but he was a scientist, not a fabricator. Nobody joined two pieces of metal with it for another eight decades.

British chemist Humphry Davy demonstrated the carbon arc in 1810, striking it between two carbon pencils. His work is also the reason we say “arc” at all: the glowing band of plasma between the electrodes bows upward, pushed by the rising column of hot air, so it literally looks like an arch. The term stuck, according to a 1927 AIEE historical paper on welding.

Fuel Gases and Resistance Heat

Meanwhile, a parallel track was heating up. Edmund Davy discovered acetylene in 1836, though nobody yet knew what to do with the awkward, smelly gas. James Prescott Joule demonstrated in 1856 that electrical resistance in wires could generate enough heat to melt metal — the seed of resistance welding. Two decades later, American engineer Elihu Thompson turned that principle into machines. Beginning around 1877, and through patents into the 1880s, Thompson became the first engineer to employ contact resistance for practical welding, developing spot, seam, and butt welding processes that factories still use today, as summarized in NASA’s technical report on metals joining.

The Arc Finally Joins Metal

Nikolai Benardos, a Croatian-born inventor working in France, was the first to make Petrov’s curiosity into a trade. His carbon arc process, patented with Stanislaw Olszewski across Paris and London from 1881 to 1885, let a welder strike an arc against a carbon electrode and pool molten metal into a joint. The IARC/WHO monograph on welding records Benardos’s 1882 demonstration as the first recorded welding use of the electric arc.

Russia answered quickly. Nikolai Slavyanov developed and patented a metallic (consumable) arc process between 1888 and 1890, feeding a metal rod directly into the arc. That is the family tree of modern stick welding. And in 1890, American C.L. Coffin patented welding in a non-oxidizing gas atmosphere — a shielded arc, decades ahead of its time. Coffin’s patent is the direct ancestor of today’s TIG and MIG processes.

Oxy-Acetylene Arrives

The other half of modern welding came from gas. Frederick Ludwig Roessler discovered calcium carbide in 1895, which made acetylene cheap and practical to produce on demand. The same year, French chemist Le Chatelier introduced the oxyacetylene blowpipe, and bottled oxygen became commercially available through liquid-air distillation. The timing was almost comically perfect: acetylene had been sitting unused for nearly 60 years, waiting for an oxygen supply hot enough to unlock it.

The Brittle Weld Problem

Here is the catch that shaped the next fifty years. Early arc welds, however impressive they looked, were often brittle and unreliable. The arc broke atmospheric nitrogen apart, and the nitrogen dissolved into the molten puddle; oxygen did the same to iron. The result was a weld full of oxides and dissolved gas — strong in a lab sample, prone to cracking in a ship’s hull. Solving that one problem — how to keep air away from the puddle — drove the invention of covered electrodes, gas shielding, flux-cored wire, and every shielding technology that followed.

1900-1939: Covered Electrodes, Automatic Welding and the Founding of the AWS

By the turn of the 20th century, welding had a heat source and a patent or two, but it still had a credibility problem. Arc welds were porous, brittle and full of blowholes, so most engineers trusted a rivet before they trusted a bead. Welders were hired to patch broken castings, not to build bridges. The three decades between 1900 and 1939 changed that completely: they gave the trade a shielding method, a machine that could weld without a human hand, a war that demanded volume, and a professional society to write the rules.

Oscar Kjellberg and the Covered Electrode

The single most important fix was embarrassingly simple. In 1907, Oscar Kjellberg of Sweden, founder of what would become ESAB, took bare iron wire and dipped it into a mixture of carbonates and silicates before striking an arc. As the coating burned, it produced a gas blanket and a layer of slag over the molten puddle, blocking the oxygen and nitrogen that had been tearing early welds apart. According to the welding history timelines maintained by the American Welding Society, that dipping process is the direct ancestor of the stick electrode almost every welder still strikes today.

The payoff came a couple of decades later. As the IARC/WHO monograph on welding processes explains, by about 1930 the modern flux-coated rod had been developed and weld quality was finally adequate for many structural purposes. That gave rise to manual metal arc welding, later renamed shielded metal arc welding, and quietly moved welding from the scrapyard to the structural blueprint.

Warships, Machines and a Welding Committee

Automation arrived before the war did. In 1913, Callahan and Raynor were granted US Patent 1,262,749 for an automatic arc welding machine, proving that a weld could be laid by a mechanism rather than a craftsman’s eye. Then World War I hit shipbuilding like a hammer. The United States needed hundreds of fabricated vessels, and riveting was slow, heavy and labor-hungry. The Emergency Fleet Corporation formed a Welding Committee to figure out whether welding could be trusted for hulls and decks, chaired by Harvard professor Comfort Avery Adams.

The American Welding Society Is Born (March 1919)

With the war over, Adams and his colleagues turned the wartime committee into a permanent organization. The AWS constitution was approved on March 27-28, 1919, with Adams as its first president. In its first year the society grew to 217 members, and the first issue of its Journal appeared in October 1919. From that point on, welding had a body that could define terms, publish procedures and establish the welding symbols that still appear on shop drawings worldwide. As Total Materia’s historical overview of welding notes, standardization is what allowed a craft to become an industry.

Proof of concept came the same year: the 1919 Socony storage tank in Poughkeepsie, New York, was built all-welded and weighed roughly 1,235 tons, a structure no riven seam could have matched for weight or leak-tightness.

The 1920s: Better Control, Hotter Arcs

The decade that followed was a laboratory sprint, much of it at General Electric:

• Automatic regulation: Paul Noble’s 1921 and 1924 patents used arc voltage to control wire feed speed, the same feedback logic modern robot welders rely on.
• Atomic hydrogen welding: Irving Langmuir’s process produced roughly half again as much heat as an oxyacetylene flame, enough to work refractory metals.
• Gas shielding research: Hobart experimented with helium and Devers with argon in the late 1920s, laying the groundwork for the TIG and MIG booms to come.

li>Knowledge in print: the first edition of the Welding Encyclopedia appeared in 1921, giving shops a shared reference instead of passed-down secrets.

Practical safety lagged behind the science. Early operators worked with little more than a handheld shield and leather gloves; the full system of helmets, respirators, fire-resistant clothing and engineered ventilation that we now expect on any job site evolved over the following decades. If you are setting up a home shop or hiring a crew, our guide to welding safety PPE walks through what is actually required today.

By 1939 the trade had a shielded process, an automatic machine, a professional society, a journal and a growing body of codes. What it did not yet have was a crisis severe enough to force every shipyard, tank plant and bridge shop to convert. That crisis arrived six years later.

1939-1945: How World War II Made Welding the Dominant Joining Method

Every previous chapter in this story was a slow burn. The war was an explosion. In roughly six years, welding went from a respected repair technique to the default way of joining metal on Earth, and the reason is brutally simple: you cannot rivet your way through a world war.

Rivets Meet Their Limit

For all the talk of arc welding’s 19th-century invention, the process did not seriously threaten the old order for decades. According to the International Agency for Research on Cancer, riveting remained the principal method of joining metal plates in buildings, bridges, ships, tanks and armaments until the late 1930s. A rivet needs access from both sides of the joint, a riveting crew, a furnace, and time. Welding needed one operator and one side of the plate.

As historian O.W. Boston noted in a study of welding’s wartime role, the emergency pushed the U.S. Emergency Fleet Corporation and later the Wartime Welding Committee to standardize procedures, qualify welders and write the first real rules of the road for the trade (see the welding history records for 1900 onward). That committee work matters more than it sounds: before it, welding was a collection of individual shop practices. After it, welding had procedures, specifications and tests.

The Ships That Broke in Half

The most famous chapter is also the darkest. Early all-welded cargo ships, most notably the Liberty ships, suffered brittle fractures that in some cases split a hull clean through. The metallurgy was unforgiving: hydrogen from the welding arc dissolved into the cooling weld metal and left it notched and brittle, especially in cold Atlantic water. The fix arrived in the early 1940s with basic low-hydrogen manual metal arc electrodes, the subject of L.S. Hobson and O.W. Boston’s 1944 paper on low-hydrogen rod developments (see the published paper). Those electrodes are the direct ancestors of the E7018 rods welders run today.

Tanks, Aircraft and Three Landmark Processes

The war did not just fix welding; it scaled it. Manual metal arc welding became the workhorse of tank production, and submerged arc welding, patented in the United States in 1936, was used extensively for tank manufacture because of its high deposition rates. The technology overview at Total Materia walks through how quickly these processes moved from patent to production line.

Aircraft presented a different problem. Aluminum and magnesium alloy frames could not tolerate the heat and rivet holes of older methods. The answer arrived in 1943: gas tungsten arc welding, now called TIG, introduced as the first successful gas-shielded process, initially used instead of rivets to assemble those lightweight airframes (see the AWS historical record). Suddenly the shielding problem that had dogged arc welding since the 1890s was solved, and the entire family of gas-shielded processes opened up.

The Post-War Verdict

By 1945 the argument was over. As the American Welding Society’s historical archives put it, welding had become the principal means of joining metal throughout manufacturing, shipbuilding and construction. The war did not invent welding. It proved, at industrial scale and under mortal pressure, what welders had known for fifty years: that fusion beats fastening.

That legacy is why the 1940s still shape daily shop practice. Low-hydrogen consumables, qualified welding procedures, operator certification, and the very idea that a weld must be inspected rather than trusted all trace back to those six years. The processes that followed, MIG, TIG refinements, flux-cored wire, were refinements on a question the war had already answered: yes, welded structures can be trusted, if you control the chemistry.

1945-2000: TIG, MIG, Flux-Cored, Plasma, Laser and Friction Stir — The Process Explosion

If the Second World War proved that welding could replace the rivet, the fifty years that followed were about fixing everything that still annoyed welders: brittle welds, sluggish deposition rates, ugly spatter, and the plain fact that a lot of metals simply did not want to be welded in open air. Between 1945 and 2000, the industry went from a handful of experimental processes to a catalog of more than 90 recognized ones. That explosion is why answering “when did welding start” with a single year feels so incomplete.

MIG Arrives: 1948 and the Consumable Electrode

The biggest post-war leap came in 1948, when gas metal arc welding — MIG — was put into use as the first gas-shielded process with a consumable metal electrode. Instead of stopping every 9 inches to change a stick electrode, a machine fed continuous wire mechanically into an arc shielded by argon or helium. Travel speeds jumped, spatter dropped, and thin-sheet welding finally became practical at production rates.

The catch appeared almost immediately. Welders soon tried cheap carbon dioxide instead of expensive inert gas. CO₂ worked, but it reacted with the molten pool and produced porosity and heavy spatter. The fix came in the 1950s: deoxidized wire, formulated to release its gases cleanly as it solidified. That gave welders a distinct, cheaper process we now call MAG welding (metal active gas), which remains one of the workhorses of fabrication. For a practical side-by-side of today’s two most popular arc processes, see our MIG vs TIG comparison.

Flux-Cored Wire: Strength in a Tube

Late in the 1950s, someone asked a simple question: what if the flux from a stick electrode were packed inside a hollow wire instead of coated on the outside? The answer was the flux-cored tubular electrode, and it spread fast. Self-shielded flux-cored wires, which need no external gas at all, gained popularity across the USA, USSR and Japan during the 1960s. The appeal was blunt: high deposition rates, deep penetration, and the ability to weld outdoors in wind that would blow away a shielding gas. Shipyards, structural steel crews and pipeline crews took to it immediately.

Plasma, Electron Beam and Laser

By the late 1960s, TIG had produced a demanding offspring: plasma arc welding, which constricts the arc through a small nozzle to produce a hotter, tighter, more stable column capable of cutting and welding much faster. Around the same period, two energy sources that belonged to physics laboratories walked into the fabrication shop. Electron beam welding, developed in the late 1950s, delivered a deeply penetrating beam under vacuum for nuclear and aerospace work. Laser beam welding followed once high-power lasers existed, eventually enabling the clean, fast, low-heat welds now routine in car body lines. The American Welding Society’s historical records document how quickly these processes moved from laboratory demonstration to shop floor.

Friction Stir Welding: 1991 and the 5,000-Year Loop

Then, in 1991, The Welding Institute (TWI) in the United Kingdom patented something that sounded almost anachronistic: friction stir welding. No arc, no flame, no melting. A rotating tool is forced into the joint and dragged along it, and the friction plus mechanical stirring fuses the metal in a solid state. It was a deliberate return to the Bronze Age principle of joining metal with heat and pressure alone, and it solved problems no arc could, especially welding aluminum alloys for aerospace and high-speed rail.

The Post-War Process Lineup at a Glance

Here is the part listicles tend to leave out: the parade of shiny new processes did not retire the old ones. Despite everything invented since, the large majority of welding worldwide still involves MMA, MAG, TIG, flux-cored and submerged arc — and MMA in particular has been the dominant process since the 1930s, a fact documented in Lancaster’s classic work on welding physics. New processes carve out niches; they rarely dethrone a workhorse that a welder can carry in one hand, strike anywhere, and fix with a hammer.

Welding Today: The Trade by the Numbers (Jobs, Wages and the Workforce Gap)

Five thousand years of history are fascinating, but they do not pay the bills. If you are weighing a welding program, hiring a fabricator, or simply wondering whether this trade still has a future, the timeline above needs to land on something concrete: what the work pays, how many people do it, and where the gaps are.

What the Government Says: BLS Figures for 2025

The U.S. Bureau of Labor Statistics groups welders, cutters, solderers and brazers into a single occupation, and its numbers are the most reliable baseline available. The full breakdown is published in the Occupational Outlook Handbook:

• Median annual wage: $53,750 in 2025, roughly $25.84 per hour.
• Total jobs: 437,700 positions across the United States.
• Projected growth: 2% between 2025 and 2035, which the BLS classifies as slower than average.
• Annual openings: about 40,300 every year, driven mostly by workers retiring or changing occupations rather than by brand-new jobs.
• Employment base: 63% of workers are employed in manufacturing.
• Entry-level education: typically a high school diploma plus moderate-term on-the-job training.

Read that 2% growth figure carefully. It is not a dying trade; it is a mature one where automation trims headcount in some shops while retirements open seats in others. A welder with a stack of certifications in a high-demand process will have a very different experience from someone with a single entry-level skill.

The Industry’s Own Count: A Wider Definition

The welding profession counts itself differently, and the numbers are larger. Data from WeldingWorkforceData.com, endorsed by the American Welding Society, estimates roughly 771,000 welding professionals working in the U.S. in 2025, with 320,500 new welding professionals needed by 2029 — an average of about 80,000 per year. Meanwhile, roughly 157,000 welders are nearing retirement.

Where the Welders Are (And Where They Are Not)

Roughly 60% of the U.S. welding workforce is concentrated in three regions: the South, the Southeast, and the Great Lakes. That concentration makes sense — shipyards, automotive plants, pipelines, and heavy equipment manufacturing cluster there. For job seekers, it means geography matters as much as skill. For shop owners outside those corridors, competing for welders is genuinely harder, and wages reflect it.

The Image Problem Holding the Trade Back

Why is there a shortage at all, when 40,300 openings a year exist? The AWS Foundation points to perception: welding still fights the reputation of being “dark, dirty, dangerous, and dead-end.” That description belonged to an earlier era. Modern shops run better ventilation, better PPE, robotic cells that handle the nastiest repetitive work, and career lattices that lead into inspection, engineering technology, and supervision. The work has changed faster than the stereotype has.

The honest summary: welding is no longer a growth occupation by headcount, but it remains one by necessity. Robots can lay beads, yet someone must program them, inspect them, repair what breaks, and build what cannot be batch-produced. That is why, five millennia after a Bronze Age smith hammered two ends together, the trade is still hiring.

Common Myths, Mistakes and Misconceptions About Welding’s History

If you search “when did welding start” and skim the first few listicles, you will meet a remarkable number of confidently stated wrong answers. Many of these claims repeat each other so often that they feel like fact. Let’s clear the slate and dismantle the six biggest myths, plus the oldest confusion of all: welding versus brazing.

Myth 1: “Welding Was Invented in 1836 by Edmund Davy”

Edmund Davy did make a landmark discovery in 1836 — he identified acetylene as a gas. That is a chemistry milestone, not a welding process. Acetylene alone burns hot, but it does nothing useful for metal joining until oxygen is mixed with it, and that practical combination did not arrive until the 1890s. Credit Davy for the fuel, not the process.

Myth 2: “Arc Welding Was Invented in America”

This one gets repeated constantly, but the pioneering arc work was decidedly international and heavily Eastern European. Russian physicist V.V. Petrov described the stable electric arc in 1800–1802, and Russian engineers Nikolai Benardos (working in Paris/London) and Nikolai Slavyanov patented carbon and metallic arc welding processes in the 1880s, with Slavyanov’s work published in 1892. American innovation came later and real — Elihu Thompson’s resistance welding, C.L. Coffin’s shielding work — but the roots of arc welding are Russian, French and British. The American Welding Society’s own historical records credit these inventors explicitly.

Myth 3: “Forge Welding Is Obsolete”

Far from dead, forge welding is the direct ancestor of one of the newest processes in the industry. Forge welding is a solid-state process: heat and pressure join metals without melting them. Friction stir welding, patented by The Welding Institute in the UK in 1991, uses a rotating tool to generate frictional heat and mechanically stir the joint together below the melting point. The 5,000-year-old Bronze Age principle never left; it simply came back with a machine tool attached.

Myth 4: “Oxy-Acetylene Came Before Arc Welding”

The timeline says otherwise. Arc welding was demonstrated as a practical joining method in the early 1880s (Benardos and Olszewski’s 1881 patents), while oxy-acetylene torches became commercially viable only after Frederick Ludwig Roessler’s 1895 discovery of calcium carbide, which made acetylene production economical. Le Chatelier proposed the oxyacetylene blowpipe that same year. The two technologies matured around the same era, but the arc got there first.

Myth 5: “Rivets Were Replaced Quickly Once Welding Arrived”

Not even close. Riveting held its dominant position in shipbuilding and heavy construction until the late 1930s. Welding had to earn trust through catastrophic early failures of all-welded vessels, improved electrode coatings, and finally the pressures of World War II. Adoption of any joining technology is measured in decades, not years.

Myth 6: “Welding Hasn’t Changed Since the 1950s”

The core physics of SMAW hasn’t changed, but everything around it has. Robotic welding cells now run automotive lines without a human hand on the torch, laser-hybrid processes combine arc and beam energy, and AI-assisted vision systems inspect weld quality in real time. If you stopped learning in 1950, you would be unemployable in a modern fabrication shop.

The Temperature Rule That Separates the Three

• Soldering: filler melts below about 450°C (840°F).
• Brazing: filler melts above 450°C, but base metal stays below its melting point.
• Fusion welding: the base metal itself is melted and coalesced.

Early arc welds were often brittle precisely because the surrounding air introduced oxygen and nitrogen into the molten pool, as documented in IARC/WHO’s occupational exposure review. That metallurgical reality is why shielding gas and flux coatings took fifty years to perfect.

Good history isn’t about gatekeeping; it’s about giving credit where the record shows it — Russian physicists, Venetian metalworkers, and Bronze Age artisans alike.

The Future of Welding: Robotics, Automation and What 5,000 Years Teach Us

If you have followed this article from Queen Pu-abi’s gold goblet to the arc lights of the 1880s, you have already noticed the pattern. Welding has never stood still. So the honest way to finish the question “when did welding start” is to ask where it is going next.

Robots on the Line, Humans at the Helm

Walk into any automotive body shop today and you will see hundreds of robot arms laying identical beads in perfect rhythm. Robotic welding is no longer the frontier; it is the baseline for mass production. What has changed more recently is the job description of the person standing next to the robot.

Welders are increasingly moving into hybrid roles that combine hands-on skill with programming, calibration, and oversight of automated cells. A technician who can read a weld procedure specification, diagnose a drifting torch angle, and re-teach a robot path is worth more than someone who can only run a bead by hand. The American Welding Society described exactly this shift in its October 2025 issue of Welding Digest, “Where Are the Welders?” — the trade is not disappearing, it is upgrading.

What the Numbers Say About Demand

The employment picture is genuinely mixed, and it is worth reading honestly. According to the U.S. Bureau of Labor Statistics Occupational Outlook Handbook, welders, cutters, solderers, and brazers held about 437,700 jobs in 2025 with a median pay of $53,750 per year, and the agency projects roughly 40,300 openings annually through 2035. BLS explicitly notes that automation in manufacturing may limit overall employment growth, while the nation’s aging infrastructure will keep requiring welders to rebuild bridges, highways, and buildings. Robots build cars; they do not yet climb under a rusted overpass.

Industry-side data points to tighter supply than that modest growth rate suggests. Figures hosted at WeldingWorkforceData.com, endorsed by the AWS, estimate around 771,000 U.S. welding professionals and a projected need for roughly 320,500 new welders by 2029 — about 80,000 a year — against a wave of retirements.

New Frontiers, Old Physics

The technical frontier is moving in several directions at once. Laser-hybrid processes combine a laser’s deep, narrow penetration with the gap-bridging tolerance of a gas metal arc, and they are showing up in shipbuilding and vehicle assembly. Friction stir welding, patented at the UK’s Welding Institute in 1991, is now routine in aerospace and marine fabrication — a process that joins metal without melting it, which is precisely the Bronze Age principle we opened this article with. Construction, historically one of the least mechanized sectors, is next in line for automation as firms experiment with track-mounted and handheld welding cobots on site.

The Real Bottleneck: People

Every forecast converges on one constraint — the pipeline. The American Welding Society’s own foundation has been blunt about the image problem: welding is still widely perceived as dark, dirty, dangerous, and a dead-end, which keeps technical-school enrollment below what the industry needs. That perception is out of date by about a century, and rehabilitating it is now a genuine economic priority.

Actionable Advice: Students, Hobbyists and Buyers

• Students: Learn manual MIG and stick first, then add robotics and inspection. Welders who can program and certify work are the ones getting hired.
• Hobbyists: Start with a quality machine and proper PPE, and practice on cheap mild steel before buying exotic filler.
• Buyers: Ask a fabricator for their weld procedure specification and certifications. A shop that documents its procedures is a shop that understands its own history.

So, when did welding start? It did not start once. It started again and again — around 3000 BC with a hammer, in 1881 with an arc, in 1907 with a coated electrode, in 1943 with argon, and in 1991 with a spinning tool that never melts anything. Each restart followed the same three-step rhythm: a new heat source appears, air ruins the weld, and someone invents a fix. Five thousand years on, that rhythm is still the best answer we have.

For the full documentation behind the dates and claims in this series, see the historical records maintained by the AWS Welding History Committee and the timeline compiled at Total Materia.

Frequently Asked Questions

When did welding start?

Welding started about 5,000 years ago. The earliest recorded welding is Egyptian ‘pressure’ or solid-phase welding around 3000 BC, where sponge iron produced from iron ore was welded together by hammering. Modern fusion welding using electricity and gas began much later, in the 1880s to 1890s.

Who invented welding?

No single person invented welding, because the earliest forge welding was developed collectively by ancient smiths. For modern welding, the key pioneers are Nikolai Bernardos (carbon arc welding, 1881-1885), Nikolai Slavyanov (metallic arc welding, 1888-1890), Elihu Thompson (resistance welding, 1877), and Oscar Kjellberg (the covered electrode, ~1907).

What is the difference between welding, brazing and soldering?

All three join metals, but at different temperatures and with different filler metals. Welding melts the base metal itself; brazing uses a filler metal that melts above about 450°C but below the base metal’s melting point; soldering uses a filler that melts below about 450°C. Ancient Sumerians were hard-soldering and brazing around 3000 BC.

When was arc welding invented?

The electric arc was discovered around 1800-1802 by Russian academic V.V. Petrov, and its first recorded use for welding was by Nikolai Bernardos in 1882. Bernardos and Stanislaw Olszewski patented the carbon arc welding process in 1885, and Slavyanov’s metallic arc followed around 1888-1890.

When was oxy-acetylene welding invented?

Acetylene was discovered by Edmund Davy in 1836, but the practical oxyacetylene blowpipe came later. Le Chatelier introduced the oxyacetylene torch in 1895, the same era that bottled oxygen became commercially available through liquid air liquefaction and calcium carbide made acetylene production practical.

Why did early arc welds fail or crack?

Early arc welds were often brittle because oxygen and nitrogen from the surrounding air reacted with the molten weld pool. Nitrogen absorption caused embrittlement, and porosity was common. The solution came in stages: covered electrodes (~1907), gas shielding in the 1920s-1940s, and low-hydrogen electrodes developed in the early 1940s.

When was MIG and TIG welding invented?

TIG (gas tungsten arc welding) was introduced in 1943 as the first successful gas-shielded welding process, originally used for aluminum and magnesium aircraft frames. MIG (gas metal arc welding) followed in 1948 as the first gas-shielded process using a consumable wire electrode.

Is welding still a good career in 2026?

Yes. BLS reports a 2025 median wage of $53,750 for welders, cutters, solderers and brazers, with about 40,300 openings projected each year through 2035. AWS-endorsed data estimates 320,500 new welding professionals will be needed by 2029, driven largely by retirements and infrastructure demand.