Manufacturing

Arc Welding

Fusing steel with a 6,000°C electric arc

Arc welding is a fusion-joining process that strikes a high-current electric arc — a plasma column burning at roughly 6,000°C — between an electrode and the metal workpiece, melting a small weld pool that fuses two parts into one as it solidifies. A shielding gas or flux keeps atmospheric oxygen and nitrogen out of the molten metal, and filler material is added to fill the joint. The same physics joins skyscraper columns, pressure-vessel seams, ship hulls, pipelines and the chassis under your car. Master three numbers — arc voltage, current and travel speed — and you control fusion, penetration and the strength of the finished joint.

  • Arc temperature~6,000 °C
  • Heat inputQ = η·V·I / v
  • Typical current50–400 A
  • Arc voltage15–35 V
  • Steel melting point~1,500 °C
  • Arc gap1–4 mm

Interactive visualization

Press play, or step through manually. The visualization is yours to drive — try it before reading on.

Open visualization fullscreen ↗

Watch the 60-second explainer

A condensed visual walkthrough — narrated, captioned, under a minute.

How the arc joins metal

Arc welding works by turning electrical energy into a concentrated burst of heat. A power source drives a large current — typically 50 to 400 amperes — through a small air or gas gap between an electrode and the workpiece. Once the gap ionizes, the current flows through a column of plasma: the electric arc. That plasma column burns at around 6,000°C, comfortably above the ~1,500°C melting point of steel, so a small puddle of metal — the weld pool — melts directly beneath the arc. As the welder moves the arc along the joint, the pool trails behind, solidifying into a continuous bead that fuses the two parts into one piece of metal.

Two things must happen for that bead to be sound. First, the molten pool has to be protected from the air. Liquid steel at 1,600°C dissolves oxygen and nitrogen rapidly; both wreck the weld, so a shielding gas (argon, CO₂, or a mix) or a burning flux blankets the pool. Second, the joint usually needs more metal than the base parts can supply, so a filler — a consumable wire or rod — is melted into the pool to build up the bead and bridge the gap.

Heat input — the master variable

Almost every weld property — penetration depth, bead width, microstructure, distortion — traces back to a single quantity: the heat delivered per unit length of weld, called heat input.

Q = η · (V · I) / v

where:
  Q = heat input          (J/m, or kJ/mm)
  η = arc efficiency       (0.6–0.85, process dependent)
  V = arc voltage          (V)
  I = welding current      (A)
  v = travel speed         (m/s)

The numerator V·I is the arc's electrical power in watts; dividing by travel speed converts power into energy deposited per metre. The efficiency η accounts for heat lost to radiation, convection and spatter rather than reaching the workpiece — it is high (~0.8) for stick and MIG, lower (~0.65) for the open TIG arc.

Worked example: a MIG weld at V = 25 V, I = 200 A, travelling at v = 4 mm/s with η = 0.8:

Q = 0.8 × (25 × 200) / 0.004
  = 0.8 × 5,000 / 0.004
  = 4,000 / 0.004
  = 1.0 × 10⁶ J/m
  = 1.0 kJ/mm

One kilojoule per millimetre is a typical structural-steel figure. Push the current higher or slow down, and Q rises: penetration deepens but so does the heat-affected zone and distortion. Run too cold or too fast, and the arc fails to fuse the joint walls — a lack-of-fusion defect that looks fine on the surface but has near-zero strength.

The four major arc processes

"Arc welding" is a family, not a single method. The members differ in their electrode, how filler is supplied, and how the pool is shielded.

ProcessSMAW (Stick)GMAW (MIG)GTAW (TIG)FCAW (Flux-cored)
ElectrodeConsumable flux-coated rodConsumable bare wireNon-consumable tungstenConsumable flux-filled tube
ShieldingFlux → gas + slagBottled gas (Ar/CO₂)Bottled inert gas (Ar)Self-shielded or gas + slag
FillerThe electrode itselfThe wire itselfSeparate hand-fed rodThe tube itself
Deposition rate1–4 kg/h2–8 kg/h0.3–1.5 kg/h3–10 kg/h
Best forField repair, structural, all-positionProduction, automation, mild steelThin stainless, aluminum, aerospaceThick plate, shipyards, outdoors
Wind toleranceGood (slag protects)Poor (gas blows away)PoorGood (self-shielded)
Skill / speedModerate / slowLow / fastHigh / slowModerate / fast

The trade-off is consistent: bottled-gas processes (MIG, TIG) give the cleanest welds but cannot tolerate wind, so they live indoors. Flux processes (stick, flux-cored) carry their own atmosphere in a coating that burns to a protective slag, so they dominate outdoor and field work — pipelines, ironwork, ship hulls.

Polarity and the physics of the arc

A welding arc is asymmetric. In a DC arc, roughly two-thirds of the energy is released where the electrons land — the anode (positive terminal) — and one-third at the cathode. Which way you connect the electrode therefore changes the weld:

  • DCEP (electrode positive, "reverse polarity"): the electrode is the anode. The energetic plasma jet drives deep, narrow penetration into the work and gives a stable arc — the default for stick electrodes like E7018.
  • DCEN (electrode negative, "straight polarity"): the workpiece is the anode and gets the bulk of the heat at its surface, while the electrode melts faster. Penetration is shallower; favored for thin sheet and high deposition.
  • AC: polarity flips 100–120 times per second. Essential for TIG welding aluminum, where the electrode-positive half-cycle blasts apart the tenacious Al₂O₃ oxide film (melting point ~2,050°C) that would otherwise trap inclusions.

The arc also self-regulates in MIG: a longer arc has higher voltage, which burns the wire back faster, shortening it again. This self-correcting arc length is what makes constant-voltage MIG so forgiving and easy to automate.

Weld metallurgy: fusion zone and HAZ

A cross-section through a weld reveals three regions. The fusion zone is metal that fully melted and resolidified — a tiny casting, with columnar grains growing inward from the cold walls toward the centreline. Surrounding it is the heat-affected zone (HAZ): base metal that never melted but was heated high enough to transform its microstructure. Beyond that lies unaffected base metal.

The HAZ is where most weld failures begin. In carbon and low-alloy steel, the weld heats the adjacent metal above its austenitizing temperature (~720–900°C), then the surrounding cold mass quenches it in seconds. Fast cooling can form martensite — hard, brittle, and prone to hydrogen-induced cracking if dissolved hydrogen (from moisture, oil, or a damp electrode) is present. Engineers control this through:

  • Preheat: warming the parts to 100–250°C before welding slows the cooling rate so martensite does not form. Thicker sections and higher carbon equivalents need more preheat.
  • Interpass temperature: keeping multi-pass welds within a window so each bead neither cracks (too cold) nor coarsens the grain (too hot).
  • Low-hydrogen consumables: E7018 rods are baked and stored in ovens to keep their flux bone-dry; damp flux is a classic source of cracking.
  • Post-weld heat treatment (PWHT): stress-relieving thick or critical welds at ~600°C to relax residual stress and temper the HAZ.

Penetration, dilution and joint geometry

Penetration — how deep the fusion zone reaches into the base metal — is set mainly by current and arc force. For full-strength butt welds in plate thicker than a few millimetres, the edges are bevelled into a V-, U- or J-groove so the arc can reach the root. The first pass, the root pass, must fuse the very bottom of the joint; missing it leaves a crack-like lack of penetration that no amount of cap welding fixes.

Dilution is the fraction of the weld bead made up of melted base metal versus added filler — typically 10–50%. It matters when welding dissimilar metals or applying a corrosion-resistant overlay, because base-metal dilution changes the deposited alloy's composition. A stainless overlay diluted by too much carbon steel can lose its corrosion resistance entirely.

Failure modes and defects

  • Porosity. Gas bubbles trapped as the pool freezes, usually from a lost gas shield (wind, low flow, a clogged nozzle) or contaminated metal. The single most common reject. Cure: protect the pool and clean the joint.
  • Lack of fusion / lack of penetration. The weld metal did not melt into the joint walls or did not reach the root. Caused by too-low heat input, too-fast travel, or poor technique. Invisible from the surface; found only by ultrasonic or radiographic testing.
  • Hydrogen (cold) cracking. Delayed cracks in the HAZ appearing hours after welding, from the combination of hydrogen, martensite and residual stress. Prevented by low-hydrogen consumables and preheat.
  • Solidification (hot) cracking. The weld centreline tears while still mushy, when low-melting impurities (sulphur, phosphorus) segregate to the last liquid. Controlled by clean base metal and bead shape.
  • Undercut. A groove melted into the base metal at the weld toe that is not filled, leaving a stress raiser. From excessive current or a poor angle.
  • Slag inclusions. In stick and flux-cored welds, slag from a previous pass not cleaned off gets trapped in the next bead.
  • Distortion and residual stress. The weld shrinks as it cools, pulling the parts out of alignment and locking in tensile residual stress that can drive fatigue and cracking. Managed with balanced sequencing, fixturing and tack welds.

Real-world numbers

  • Arc plasma temperature: ~5,000–6,500°C; the workpiece surface under it reaches ~1,600–1,800°C.
  • Steel melts at about 1,500°C, aluminum at 660°C, copper at 1,085°C.
  • Typical structural heat input: 0.8–2.0 kJ/mm; thin sheet may run under 0.3 kJ/mm.
  • Open-circuit voltage of a stick power source is ~50–90 V; it collapses to a 20–30 V arc voltage once the arc strikes.
  • Shielding gas flow: 10–20 L/min for MIG; too little invites porosity, too much causes turbulence that sucks in air.
  • Cooling rate in the HAZ of a thick plate can exceed 100°C/s — fast enough to form martensite in unprepared carbon steel.

Frequently asked questions

How hot is an arc-welding arc?

The plasma column of a welding arc reaches roughly 5,000–6,500°C (about 6,000°C is the figure usually quoted), far above steel's melting point of about 1,500°C. The arc itself is an ionized gas conducting tens to hundreds of amperes across a 1–4 mm gap. Only a thin layer of the workpiece actually melts; the surrounding metal stays solid and conducts the heat away rapidly, which is why penetration is shallow and controllable.

Why does arc welding need shielding gas or flux?

Molten steel reacts greedily with atmospheric oxygen and nitrogen. Oxygen forms iron oxide and burns out carbon and alloying elements; nitrogen dissolves into the pool and forms brittle nitrides and porosity as it cools. Shielding gas (argon, CO₂, or mixes in MIG/TIG) or a flux coating (which burns to release a CO₂/CO blanket and forms a slag in stick and flux-cored welding) displaces air from the weld pool so it solidifies clean and strong. Lose the shield — by welding in wind above about 8 km/h, for example — and you get porosity, the most common weld defect.

What is the difference between MIG, TIG and stick welding?

All three are arc processes but differ in electrode and shielding. Stick (SMAW) uses a consumable flux-coated rod that both carries current and supplies filler; the flux makes slag. MIG (GMAW) feeds a continuous bare wire that is both electrode and filler, shielded by a bottled gas — fast and easy to automate. TIG (GTAW) uses a non-consumable tungsten electrode under argon, with filler added separately by hand; it is the slowest but cleanest, used for thin stainless, aluminum and aerospace work.

What is the heat-affected zone (HAZ)?

The heat-affected zone is the band of base metal next to the fused weld that got hot enough to change its microstructure but not hot enough to melt. In steel it sees rapid heating and quenching by the cold surrounding metal, which can form hard, brittle martensite and drive hydrogen cracking. Controlling cooling rate via preheat, interpass temperature and heat input keeps the HAZ tough; this is why thick or high-carbon steel is preheated before welding.

How do you calculate welding heat input?

Heat input per unit length is Q = η · (V · I) / v, where V is arc voltage, I is current, v is travel speed, and η is the arc efficiency (about 0.7–0.85 for stick and MIG, 0.6–0.7 for TIG). For example, 25 V × 200 A at 4 mm/s travel with η = 0.8 gives Q = 0.8 × 5000 / 0.004 = 1.0 MJ/m, or 1.0 kJ/mm. Heat input is the master variable: too low and the weld lacks fusion, too high and you grow a coarse, weak HAZ and risk burn-through.

Why is DC electrode-positive used for deep penetration?

In a DC arc, roughly two-thirds of the heat is liberated at the positive terminal (the anode) because that is where high-energy electrons land. With direct-current electrode-positive (DCEP, reverse polarity) the electrode is positive, concentrating energy in the arc and driving a hot plasma jet that gives deep, narrow penetration into the work. DCEN (electrode negative) puts more heat into the workpiece surface and melts the electrode faster but penetrates less; AC is used for aluminum TIG to break up the tenacious oxide film each half-cycle.