MIG Welding Silicon Bronze: A Complete How-To Guide

A restoration shop spends two days fitting a galvanized rocker panel, then ruins the zinc coating on the first pass with standard mild-steel wire. The weld looks fine — until the rust blooms along the seam six months later. The fix was never a better welder or a hotter arc; it was a different filler metal entirely, one that melts at a lower temperature than the steel underneath it.

Why Silicon Bronze Ends Up in the Strangest Jobs

The rocker panel from the opening story did not fail because the welder lacked skill. It failed because the process was wrong for the job. Standard steel MIG wire needs a molten pool of steel to tie into, and that pool destroys the zinc coating that was protecting the panel in the first place. Silicon bronze takes a different route. It melts and wets the steel while the steel underneath stays solid, so the repair can be made with far less heat and far less damage to the surrounding coating.

That is the short answer to why anyone reaches for this filler. MIG welding silicon bronze is a gas-shielded, semi-automatic process that feeds a copper-silicon wire through a standard MIG gun. The arc melts the wire, the molten bronze spreads across the prepared steel, and the base metal remains below its melting point. The result is properly called braze welding or MIG brazing, not fusion welding. Once you understand that distinction, the process stops looking mysterious. It also explains why it succeeds on thin, coated, and dissimilar materials, and why it punishes operators who treat it like ordinary steel MIG.

The Temperature Gap That Drives Everything

Three different melting points sit at the center of this process, and the gap between them is what makes it useful. Zinc coating melts at roughly 420 °C (788 °F) and vaporizes around 906–910 °C (1,670 °F). Conventional steel MIG welding fuses the base metal at roughly 1,650 °C (3,000 °F), which is far above the point where zinc turns to vapor. Silicon bronze filler sits in between: AWS A5.7 ERCuSi-A melts at approximately 1,866 °F, or about 1,020 °C.

Material or Event	Approximate Temperature
Zinc coating melts	420 °C (788 °F)
Zinc coating vaporizes	906–910 °C (1,670 °F)
ERCuSi-A silicon bronze melts	1,020 °C (1,866 °F)
Steel base metal fuses in conventional MIG	About 1,650 °C (3,000 °F)

Notice that the bronze still melts above zinc’s vaporization point. That matters, because the popular claim that MIG brazing “does not burn the zinc” is too generous. The arc is hot enough to vaporize coating near the joint. What the process actually does is reduce total heat input and direct much of that heat into melting filler rather than melting steel. Less heat spreads into the panel, the heat-affected zone stays narrower, and more of the original coating survives outside the immediate weld seam. On a galvanized body panel, that difference is often the difference between a repair that lasts and a seam that rusts from the inside out.

Why Shops Keep Coming Back to It

The practical appeal goes beyond one rusted rocker panel. Silicon bronze is forgiving on thin sheet because it does not need a molten steel pool. It joins steel to copper, brass, and cast iron when a fusion weld would create brittle intermetallic compounds. It produces less distortion, handles poor fit-up better than many fusion processes, and leaves a bead that can be sanded and painted with relatively little work. Deposits typically test at 50,000 psi minimum tensile strength (345 MPa), according to published ERCuSi-A filler metal data, which is respectable for a lap or fillet joint when the bonded area is adequate.

The failures are just as predictable. Run the machine too hot and the filler becomes weak and crack-prone in its hot-short range. Move too fast or leave contamination on the surface and the bronze sits on top without wetting in. Misjudge the joint as a fusion weld and design it with too little overlap. In every case, the root cause comes back to the same thing: silicon bronze is a brazing process wearing MIG clothing.

What This Guide Covers

The rest of this guide works through the process in the order a shop actually encounters it: the metallurgy behind ERCuSi-A, the welding-versus-brazing terminology, equipment setup with soft wire and dedicated liners, starting parameters, torch technique, base metal combinations, troubleshooting, strength and safety, and finally cost and alternatives. The setup section borrows heavily from the soft-wire playbook, so our aluminum MIG welding guide is useful background if you have never fed a wire that wants to buckle, shave, or nest inside the drive rolls.

Keep one idea in mind as you read. Silicon bronze is not a hotter or colder version of steel MIG. It is a different joining mechanism. That single distinction explains its ability to protect galvanized steel, its tolerance for dissimilar metals, its modest joint strength in the wrong configuration, and nearly every defect covered later in this article.

What Silicon Bronze Actually Is (and What ERCuSi-A Means)

Before you can understand why silicon bronze behaves the way it does under an arc, you need to know what is in the wire. It is a copper alloy, but stopping at the word “bronze” skips the detail that makes the process work.

The filler is classified by the American Welding Society as ERCuSi-A under AWS A5.7, the same document adopted as ASME SFA 5.7. The “ER” prefix means electrode or rod, “Cu” is copper, “Si” is silicon, and “A” marks the first classification in the family. The wrought designation is UNS C65600. Old-timers may still call it Everdur 656, a trade name that outlived the original product line.

The Chemistry, Element by Element

The spec allows ranges, not a fixed recipe, which is why two brands can feel different on the same machine.

Element	Range	What It Does
Copper	Balance	Ductile matrix, carries conductivity
Silicon	2.8–4.0%	Deoxidizer, fluidity, strength
Manganese	1.5% max	Scavenges impurities
Tin	1.0% max	Residual; mild strength
Zinc	1.0% max	Residual scrap; fume source
Iron	0.5% max	Residual; excess hurts ductility
Aluminum	0.01% max	Tightly held; cracking, dross
Lead	0.02% max	Restricted for health

The silicon range is the heart of the alloy. At roughly 3% it does three jobs. It deoxidizes the puddle, tying up oxygen that would otherwise become porosity. It raises fluidity, so the molten filler wets and spreads across steel without needing a puddle of the base metal. And it strengthens and work-hardens the deposit. Push past 4% and the alloy gets brittle and hard to draw into wire. Drop below 2.8% and you lose the wetting behavior the process depends on.

Manufacturers publish full chemistry and property data. The Pinnacle Alloys ERCuSi-A sheet and the Inweld silicon bronze data sheet are useful for comparing brands, and the C656 copper alloy datasheet covers the wrought side of the same family.

The Numbers That Matter

Silicon bronze melts at roughly 1,866 °F (1,019 °C). Mild steel does not melt until about 2,700 °F. That gap is not a small detail. It is the reason you can bond bronze to galvanized sheet without destroying the zinc coating, and why the process is technically braze welding rather than fusion welding.

Typical as-welded properties:

• Tensile strength: about 50,000 psi
• Elongation: roughly 40–65%, depending on form and test method
• Hardness: around 80–100 HB
• Density: about 0.308 lb/in³
• Thermal conductivity: about 21 Btu/(hr·ft·°F)
• Electrical conductivity: roughly 7% IACS

That 50,000 psi figure gets quoted constantly, but in a lap joint the bonded area and joint design dominate the outcome far more than the filler’s tensile number. A thin bronze fillet on a wide overlap can hold more than a fat bead on a narrow one. Filler strength is only one variable.

The low thermal and electrical conductivity are practical facts. Low conductivity keeps heat near the arc instead of racing into the part, which helps on thin sheet but demands disciplined travel speed. The Oxford Alloys silicon bronze reference is a good second source for these values.

Wire Spec vs. Rod Spec: Don’t Buy the Wrong Spool

Two AWS documents describe what sounds like the same alloy:

• AWS A5.7 ERCuSi-A — welding filler, supplied as spooled wire for MIG or bare rod for TIG. This is what a MIG machine takes.
• AWS A5.27 RCuSi-A — brazing and braze welding filler, usually bare cast or drawn rod for torch work.

The chemistry is close, but product forms, tolerances, and surface condition differ. A brazing rod is not wound on a spool and will not feed through a liner. Ask for ERCuSi-A MIG wire specifically and check the spool label for the A5.7 classification.

Sizes, Spools, and the Hot-Short Warning

Common MIG diameters run 0.023, 0.030, 0.035, 0.045, and 1/16 in. Small sizes suit thin sheet and low-current work; 0.045 and 1/16 in handle heavier sections. Spools range from 2 lb, 8 in rolls for small jobs and spool guns to 30 lb and larger for production. Buy small first: the wire is soft, it ages on the shelf, and 30 lb is an expensive way to learn you dislike the process.

One behavior shapes every technique in this guide. Datasheets call silicon bronze “hot short”: at elevated temperature it loses ductility and cracks easily. Overheat the deposit and it tears along the bead or at the crater as it cools and contracts. This is not a defect you can weld out with more heat; it is a metallurgical property. The rule that follows is simple: run it as cool as the joint allows.

Coming from aluminum work? The handling challenges will feel familiar: soft wire, sensitive feed, contamination trouble. Our guide to aluminum MIG welding covers the same equipment discipline.

Understand the alloy and the rest of the process becomes logical. The low melting point, the silicon content, hot-short behavior, and braze-welding classification are one fact with several consequences, and every setting later in this article flows from them.

Welding or Brazing? The Word That Trips Everyone Up

Ask five welders whether MIG brazing counts as welding and you will get five confident answers, at least one of them flatly wrong. The confusion is not academic. It decides whether a technician is allowed to touch a structural rail, how a repair gets written up on an estimate, and whether an insurer pays or pushes back. Settle the definition before you strike an arc.

What AWS Actually Means by Braze Welding

The American Welding Society draws the line by temperature, not by appearance. Braze welding joins metals using a filler that melts above 450 °C (840 °F) but below the melting point of the base metals. That is the entire test. Filler below 450 °C is soldering. Base metal melting means fusion welding. Filler above 450 °C while the steel underneath stays solid means braze welding.

There is a second distinction worth knowing. True brazing relies on capillary action to pull filler into a closely fitted joint. Braze welding deposits filler into a groove or fillet, bead by bead, the way a weld is laid down. MIG brazing is the second type: you build a bead rather than wick metal into a hairline gap.

Adhesion, Tinning, and What the Arc Is Really Doing

The temperature gap is the whole story. In MIG brazing the arc runs at roughly 960–1,000 °C (1,760–1,830 °F). A steel fusion arc runs near 1,650 °C (3,000 °F). Silicon bronze filler, ERCuSi-A, melts around 965 °C (1,770 °F), so the arc has just enough energy to melt the wire and barely warm the steel beneath it.

Nothing fuses at the interface. Only a superficial skin of the steel melts, and the bronze wets and tins that surface, forming a thin alloyed layer that does the actual holding. The bulk of the joint is bronze bonded to steel, not steel fused to steel. In a fusion weld the base metal itself becomes liquid and the two edges merge into one continuous structure.

Why a Good-Looking Joint Can Still Fail a Tear-Out Test

This is the part that catches people. A bronze bead can look beautiful, sit perfectly flat, and still peel away under a chisel because the arc was too hot. Push the amperage and the steel surface melts more than superficially. Iron enters the bronze puddle, the filler goes hot short, and the wetting layer turns into a thin, brittle intermetallic film instead of a tough alloyed bond. The bead looks no different. The joint is worthless.

Overheating also burns the zinc on galvanized stock well before it burns the steel, which is the exact damage you were trying to avoid. On a tear-out test, a properly made joint fails by tearing the bronze itself or by pulling a nugget of steel with it. A bad joint separates cleanly at the interface, like a sticker peeling off glass. That difference in failure mode is the most honest test you can run in a small shop.

The Rules That Actually Govern Your Repair

In automotive structural repair, whether MIG brazing is permitted is decided by the OEM procedure for that specific vehicle, not by what worked on someone else’s truck. I-CAR guidance is blunt on this point: MIG brazing should only be performed in areas specified by an OEM procedure.

The history is instructive. As reported in I-CAR’s 2004 MIG brazing publication, the process was not recommended for structural parts on vehicles sold in North America at that time, while several European manufacturers were already specifying it. Vehicles cited in that guidance included the 2004 Holden Vectra C, the 2004 Volkswagen Golf, PSA Peugeot Citroën models, and certain Toyota panel joints.

Two cautions follow. That snapshot is more than twenty years old, and OEM body construction has moved steadily toward MIG brazing on high-strength and galvanized steels precisely because it keeps heat input low. Guidance also differs by region and model year, sometimes within the same nameplate. A technique legal on a European-built panel may be prohibited on the North American version of the same car.

Call it braze welding on the invoice, call it MIG brazing in conversation, but call the OEM first when structure is involved. The terminology argument is worth winning because the repair decision hangs on it.

Setting Up: Equipment, Drive Rolls, Liners, Gas, and Torch

Silicon bronze setup usually fails in the feeder long before it fails at the arc. ERCuSi-A is soft, has low columnar strength, and deforms under pressure that mild steel wire simply shrugs off. Treat the drive path as your first welding parameter. Get it wrong and you will spend an hour chasing porosity and arc instability without ever suspecting the wire was the problem.

Treat the Wire Like Aluminum Wire

Filler manufacturers are consistent on this point: copper-alloy wire wants soft or formed drive rolls and a low-friction liner. I-CAR goes further in its MIG brazing guidance, advising techs to handle CuSi3 exactly like aluminum wire — no abrasions on the wire surface, smooth half-round drive rolls, and a plastic or Teflon liner rather than a steel coil conduit. You can read that guidance directly in the I-CAR MIG brazing article.

What happens with the wrong hardware is predictable. Knurled or V-groove rolls bite into the wire and shave bronze flakes into the liner. Those flakes migrate to the contact tip, choke the arc, and show up later as erratic feed or a burnback. Set drive-roll tension just high enough to feed without slipping, then stop. If you can flatten the wire between your fingers after it exits the rolls, the tension is too high.

• Drive rolls: smooth half-round or U-groove, sized to the wire diameter
• Liner: Teflon, plastic-graphite, or carbon-fiber — never a steel liner that has run steel wire
• Tension: light. Soft wire deforms before it slips
• Wire diameter: 0.8 mm (0.030 in) is the most forgiving for thin sheet; 1.0 mm and 1.2 mm feed more reliably on thicker work

If you have set up aluminum before, the logic will feel familiar — the same principles govern feed reliability on soft wire, as covered in this aluminum MIG welding guide.

Spool Guns, Push-Pull, and Contact Tips

On 0.8 mm bronze, a short gun lead is your friend. Long conventional torches add friction and kinking, so spool guns and push-pull setups earn their place on thin wire and awkward bodywork positions. Keep the contact tip matched to wire size or one size larger if feed is marginal, and change it whenever the bore shows oval wear — bronze is unforgiving of a sloppy tip.

Stickout is commonly listed around 1/2 in for MIG brazing and is far less critical than with steel wire. You are not plunging into a molten pool, so a slightly longer standoff mostly affects wetting and shielding coverage rather than penetration.

Clean the Feeder Before You Switch

Steel residue is contamination, not just clutter. Graphite, mill scale dust, and metal shavings left in a shared liner or torch will be dragged into the bronze and appear as dark specks and porosity in the bead. Fit a dedicated liner and torch, blow out the conduit, and inspect the inlet guide. Five minutes here saves a rework job.

Polarity and Shielding Gas

GMAW silicon bronze runs on DCEP — electrode positive — in spray transfer. Reverse the polarity and the arc behaviour changes completely; the process is designed around the wire being the positive pole.

Shielding gas is normally 100% argon at roughly 25–30 cfh. Argon gives a stable, controllable spray and quiet wetting on thin steel. For thicker sections, a 75% argon / 25% helium mix raises heat input without forcing you to push amperage through a soft wire. Some sources note that argon with a small CO2 addition (around 2%) can improve arc stability, but pure argon remains the default and is what most published parameters assume — see the ERCuSi-A datasheet for baseline figures.

Gas	Typical use	Notes
100% argon	Default, thin sheet, body panels	25–30 cfh, most stable low-current arc
75% Ar / 25% He	Thicker sections, copper base metal	More heat without extra amperage
Ar + ~2% CO2	Occasional arc-stability aid	Not the default; test on scrap first

Power Source: Why the Low End Matters

Ignore the headline top-end amperage on the spec sheet. MIG brazing lives between roughly 45 and 65 amps, and a machine that cannot hold a stable arc down there is useless for this work no matter what it claims at 250 amps. Look for low-end stability, fine voltage steps, and good arc-start behaviour.

Synergic programs are a genuine advantage: you set wire diameter and material thickness, and the machine tracks voltage with wire-feed speed so you can adjust heat with one knob. Pulsed modes help on thin galvanized sheet by lowering average heat input while keeping the arc directional, which reduces zinc burn-off and distortion.

Dialing In: Real Starting Parameters by Wire Size and Thickness

Every silicon bronze job begins with the same optimistic mistake: copying a parameter chart from a filler datasheet onto a machine that is about to run a 0.8 mm galvanized panel. The numbers are real, but they were never meant to be a universal recipe. They describe the filler’s comfortable working range on heavier sections. Thin sheet lives somewhere else entirely.

So use published data as a reference point, then let the puddle tell you which way to move. That is the whole skill of dialing in this process.

Pick the Wire Diameter First

Wire diameter does most of the preliminary work. For automotive sheet and light fabrication, 0.030 in and 0.035 in are the everyday choices. Step up to 0.045 in for heavier steel, thicker copper, and larger fillets, and reserve 1/16 in for thick copper base metal or serious build-up where heat input is less of a concern.

Published GMAW parameters for ERCuSi-A with 100% argon shielding give a useful map of the filler’s range:

Wire diameter	Amperage	Voltage	Wire-feed speed
0.030 in	130–150 A	21–23 V	460–500 ipm
0.035 in	145–185 A	23–25 V	400–440 ipm
0.045 in	195–215 A	26–28 V	280–310 ipm
1/16 in	260–280 A	27–30 V	150–210 ipm

Those figures come from published ERCuSi-A GMAW parameters. Notice how high the wire-feed speeds are for the small diameters. Silicon bronze is a soft wire, and it takes a lot of feed to deliver modest current.

Read the Puddle, Not Just the Display

At a fixed wire diameter, wire-feed speed controls how much metal enters the puddle, and amperage follows. Voltage controls arc length and how widely the bead spreads. Increase wire-feed speed without enough voltage and the wire stubs; add voltage without feed speed and the arc wanders, the bead flattens, and the steel underneath gets hotter than it should.

Travel speed is the third variable, and it is the one most people ignore. A slow pass increases heat input even at low amperage because the arc sits in one place. If the toes are wetting but the panel is starting to blue or the zinc is burning away over a wide band, move faster before you touch the voltage.

The Low-Current Regime for Galvanized Sheet

Thin coated steel is where the process earns its reputation, and where it is easiest to ruin. The goal is to wet the bronze onto the steel without melting through the zinc on the far side of the joint. Collision repair guidance on MIG brazing stresses low heat input for exactly this reason. Runs in the 45–65 amp range, paired with a short, stable arc, keep the steel below its melting point and limit coating damage.

This is also why a machine’s low-end stability matters more than its top-end amperage. An arc that cuts out or stutters at 50 amps is useless here, no matter how impressive the label on the front panel says it is.

Why a Crowned Bead Is Correct

A finished bronze bead should sit slightly proud, with a rounded crown and smooth, bonded toes. It will not look like a fused steel weld, and it is not supposed to. That raised profile is the filler sitting on a wetted interface, not a deep fusion zone.

Filler datasheets warn that turning up the heat to flatten the bead defeats the whole purpose of low-heat MIG brazing. A wide, washed-out, nearly flat bead is usually a sign of excess voltage, too much current, or travel that is far too slow. Do not “fix” a crowned bead with more voltage. Fix your expectations instead.

Preheat and Interpass: Steel Versus Copper

Steel sheet generally needs no preheat. Copper is a different story. It conducts heat away from the joint so quickly that a cold start struggles to wet, so a controlled preheat is normal practice for copper base metals. Follow the procedure for the alloy and thickness rather than guessing.

Interpass temperature is commonly capped at 150 °F (66 °C), with some sources specifying a 100 °C (212 °F) maximum depending on thickness. Those caps exist to protect the low-heat advantage and to avoid overheating the filler and the base metal. Check the relevant filler datasheet for the exact limit on your product, and use a temperature crayon or contact thermometer instead of your hand.

Short Arc, Pulse, or Spray?

Short-circuit transfer is the default for thin sheet because it keeps the arc short and the heat input low. Pulsed equipment is reported to give the best results because it delivers one molten droplet per pulse with lower average heat input, which makes wetting more consistent on galvanized steel. Spray transfer belongs on heavier sections; it is too hot and too fluid for thin panels. If you have synergic or pulsed programs, they are worth learning, much like the pulse controls covered in this aluminum MIG welding guide.

Start conservative, change one variable at a time, and judge the result by bead profile, toe wetting, and coating damage. The numbers get you close. The puddle closes the deal.

Technique: Torch Angle, Push vs Pull, Bead Shape, and Distortion Control

Setup is done. What decides whether the joint holds is what your hands do. With silicon bronze you are not melting steel; you are persuading a low-melting filler to wet a solid surface.

Work Angle and Travel Angle

On a lap joint, set the torch about 45° from vertical and 60° away from the direction of travel. The work angle splits heat between the top edge and the panel beneath it; the travel angle leans the wire back over deposited metal and keeps the nozzle clear of the unwelded seam.

Angling away from travel also lets zinc vapor escape. Galvanized coating vaporizes near 907 °C, below steel’s melting range and right in the arc zone. Shroud the joint and that vapor must bubble up through molten bronze; trapped bubbles become porosity. On a fillet, steepen to 40–45° into the corner so both legs wet — but do not add voltage to force bronze into a tight crevice. It wets outward instead.

Push, Don’t Pull

Default to pushing. The arc preheats plate ahead of the puddle and burns off zinc before the bronze arrives: clean steel, immediate wetting, no vapor underneath. Pulling reverses the order, so vaporization happens under the bead and drives gas up through it. The result is porosity, a rougher bead, and an unstable arc. Most “bronze MIG is spattery” complaints come from dragging the torch.

Bead Shape: Crowned Is Correct

New welders see a bead 2–3 mm proud and add voltage to flatten it — the most common technique error with this filler. Bronze does not fuse into the base metal, so there is no fusion line to sink into; it wets across the surface and surface tension holds it up. Datasheets call for stringers and a small pool because silicon bronze is hot short: it loses ductility at temperature and cracks if cooked. A crown means the pool stayed small and the heat stayed controlled. Chase it flat and you get a wide, porous, cracked deposit.

Fit-Up, Root Gap, and Tacking

Wetting, not penetration, carries the load, so fit-up matters more than in fusion welding.

• Small root opening, about 0.5–1.0 mm. It gives the bronze somewhere to wick, increasing bonded area; a zero-gap lap leaves filler sitting on top.
• Clean, bare metal. Strip zinc from the joint zone with a stripping disc or flap wheel, then degrease. You want bronze bonding to steel, not zinc.

Tack short and cool at the ends and midpoint: 5–8 mm on thin galvanized sheet, low current, cooled before the next. If a tack is still red, you are moving too fast. On very thin panels, set the gap with a wedge or a scrap of 0.8 mm sheet, then pull it.

Sequencing and Distortion Control

Low heat input is the headline advantage: less warpage, far lower burn-through risk. But low heat is not zero heat. Work in 50–75 mm backstepped segments, alternating sides and moving away from center, and balance heat so the far side does not shrink toward the weld. Let the panel cool between passes; datasheets cap interpass temperature because bronze stays hot short. Run stringers, never weaves, which widen the HAZ, enlarge the pool, and slow travel. The mindset matches thin aluminum, where travel speed and sequence outweigh raw amperage — see our guide to aluminum MIG welding.

Puddle Size, Travel Speed, and Crater Handling

Keep the pool about the width of the wire plus a little. At 45–65 A on 0.8 mm wire it looks like a bright fluid puddle riding a dull red-brown track, not a molten lake. Tint spreading more than a few millimetres means too hot or too slow. Thin sheet runs at 4–8 mm per second with a steady crackle; popping means gas coverage is failing or the pool is boiling zinc. Craters crack, so do not snap out: pause, let the pool wet out, break the arc with a short lift, or use crater-fill. To restart, brush the crater, start slightly ahead, and wash back in.

Worked Example: Thin-Gauge Galvanized Lap

Picture 0.8 mm galvanized lapping onto 1.0 mm steel, 200 mm seam.

1. Prep. Strip zinc 10 mm back, degrease, clamp with a 0.5–1.0 mm shim gap.
2. Tack. Four 5 mm tacks, low current, cooling between; remove the shim.
3. Run. 0.8 mm ERCuSi-A, 100% argon at 12–14 L/min, 50–60 A, 45° work angle, pushing; two or three 60–70 mm backstepped stringers.
4. Finish. Brush while warm, check the ends for porosity, dress the crown if appearance matters.

A few minutes of work and the panel stays flat.

Where Technique Changes: Copper and Vertical Work

On copper base metal the heat problem inverts. Copper pulls heat away so fast that thicker sections need preheat, typically 150–300 °C, and more amperage just to establish a pool — then travel faster, because the surrounding mass sinks heat and the pool can go mushy before it wets. Angle and push stay the same. Vertical-up is possible but demanding: cut current 10–15% and keep a narrow, weave-free stringer with a slight upward push. Vertical-down is not recommended for structural laps. Our all-position welding rod guide covers the puddle-control basics.

Get angle, direction, and bead shape right and the rest becomes forgiving. Get them wrong and no parameter tweak will save the joint.

Materials: Galvanized Steel, Copper, Cast Iron, and Dissimilar Joints

Silicon bronze is not a universal filler. It is a specialist that happens to solve a few common industrial problems extremely well. Know which base metals it likes, which it tolerates, and which it should never touch, and you avoid a lot of ruined parts.

Galvanized and Coated Steel: The Headline Application

This is where ERCuSi-A earns its keep. Silicon bronze is specifically suited to welding plain or galvanized steel sheet and other coated steels because it does not destroy the coating adjacent to the bead. That single sentence explains why automotive body shops, HVAC fabricators, and ductwork manufacturers keep spools of it on hand.

Fusion weld galvanized steel with ER70S-6 and the arc takes the steel past its melting point. Zinc around the puddle burns off in a visible white cloud, leaving a bare strip of 10 to 25 mm on each side. That strip is now unprotected steel. Rust starts there, and on a lap joint it creeps invisibly along the seam until it blisters the paint.

With MIG brazing the steel never melts. The arc only needs roughly 900 to 980 °C at the interface, well below steel’s melting range, so zinc a few millimetres from the bead survives. You still lose zinc directly under the arc, but the damage footprint is far tighter. The bronze deposit is itself highly corrosion resistant, and the surviving zinc provides cathodic protection that limits rust creep along cut edges.

Other coated steels behave similarly: Galvalume, zinc-aluminium coatings, and most paint-primed sheets tolerate the low heat input far better than a fusion process. For automotive work, follow the OEM procedure. I-CAR’s guidance on MIG brazing and regional body repair standards make clear that permission to braze depends on the specific steel grade and model year.

Mild Steel Sheet and Light-Gauge Material

Plain low-carbon sheet from about 0.6 mm to 3 mm is the sweet spot. Thin panels that would warp or blow through under a fusion arc braze beautifully because the heat stays low and the puddle stays small. Silicon bronze is also used for brazing malleable iron and light-gauge steel, which covers a surprising amount of repair and restoration work.

Above roughly 3 mm the economics shift. You need more preheat and more passes, and a conventional steel MIG with ER70S-6 is usually faster and cheaper unless coating preservation or distortion control decides the job.

Copper, Copper-Silicon, and Copper-Zinc Base Metals

Here silicon bronze is on home ground. It joins copper, copper-silicon, and copper-zinc base metals to themselves and to steel. Copper pulls heat away fast, so expect to preheat, often to 150 to 300 °C on thicker sections, and expect a sluggish puddle until the part warms. Brasses are more delicate: zinc boils at a low temperature and can cause porosity, so keep the arc moving and the current modest. Leaded brasses are best avoided entirely.

Cast Iron and Dissimilar Joints

Silicon bronze is a classic choice for cast iron repair because the low heat input reduces the thermal shock that cracks grey iron. Preheat helps on heavy sections and peening between passes relieves stress. The result is a sound, machinable joint rather than a hardened, crack-prone fusion zone.

Dissimilar combinations are where the filler really shines. Steel to brass, steel to copper, and bronze to steel all work because the filler wets both surfaces without requiring either to melt. For a deeper look at how filler selection changes with a different soft, conductive metal, see our guide to aluminum MIG welding.

Stainless and High-Strength Steel: Proceed With Caution

Silicon bronze can braze stainless, but the joint is brazed, not welded. On austenitic stainless in corrosive service the copper-steel interface becomes a galvanic weak point. On advanced high-strength steels in vehicle structures, heat input alters the metallurgy, so never braze without the OEM procedure in hand.

Galvanic Corrosion in Wet Service

Bronze joined to steel in a wet, salty environment creates a galvanic couple. The steel becomes the anode and corrodes preferentially. In dry interior service this is negligible. On marine or road-salt-exposed assemblies, seal the joint, keep it painted, and avoid trapping moisture.

Build-Up and Surfacing

Silicon bronze is widely used for surfacing parts exposed to corrosive environments. Pump housings, shaft journals, and valve bodies can be built up with a work-hardening bronze layer that resists corrosion and wear. Deposit thin passes, keep interpass temperature in check, and remember the deposit work-hardens under service load rather than in the as-deposited state.

Base Metal Combination	Recommended Approach	Key Caution
Galvanized steel sheet to itself	MIG braze, 100% argon, low current	Clean coating, ventilate fume, follow OEM spec
Mild steel, 0.6–3 mm	MIG braze, lap or fillet	Control distortion; stringer beads only
Copper / brass	Preheat, modest current, fast travel	Zinc vaporization porosity; avoid leaded grades
Cast iron repair	Braze with preheat, peen passes	Thermal shock cracking on heavy sections
Steel to brass or copper	MIG braze, wet both surfaces	Galvanic couple in wet service
Stainless / AHSS	Only per a qualified OEM procedure	Not a structural substitute; heat alters base metal

Troubleshooting: Porosity, Cracking, Spatter, and Bird-Nesting

Almost every silicon bronze failure that lands on a bench gets blamed on the operator. Most trace back to three things instead: too much heat, a contaminated feed path, or a torch angle that traps zinc vapor where the puddle is trying to form. Bronze is unforgiving in a specific way. It can produce a weld that looks acceptable and bonds poorly, so diagnosis has to go beyond appearance.

The most common technician error reported by I-CAR is setting machine output too high and making the weld too hot. Check that first.

The Diagnostic Table

Symptom	Likely cause	Fix
Porosity, scattered or clustered	Zinc vapor trapped by excess heat or wrong torch angle; gas loss; dirty liner	Lower current, push the torch, raise gas flow, replace liner
Cracks at centerline or across bead	Overheated hot-short filler; restrained joint cooling fast	Smaller puddle, cap interpass, adjust bead sequence
Cold lap, poor wetting	Heat too low, oil or coating on surface, travel too fast	Clean to bare metal, add slight energy, slow down
Heavy spatter, harsh arc	Wrong gas mix, dirty wire, poor ground	Argon-dominant gas, clean drive path, fix work clamp
Bird-nesting at the feeder	Drive tension too high, wrong roll profile, kinked liner	Formed rolls, minimum tension, fresh liner
Burnback, tip welded shut	Wire sticking, feed hesitation, oversized tip bore	Correct tip bore, check spool drag, trim wire
Wandering bead width	Travel speed drift, inconsistent standoff	Brace the torch, count the rhythm, watch puddle width

Porosity: Zinc Vapor Versus Shielding Loss

Zinc boils at roughly 907 °C, well below arc temperature, so on galvanized sheet the coating turns to vapor the moment the puddle forms. Run too hot, or angle the torch so vapor has to pass through the molten pool, and it gets trapped — porosity plus inadequate fusion at the same time. The countermeasure is mechanical: lower current, a pushing angle that lets vapor escape ahead of the puddle, and gas coverage across the full bead width.

Shielding-related porosity looks similar but sits differently, usually along bead edges or at restart points. Check the nozzle for clogs, the diffuser for spatter, and the shop for a fan pulling gas away before touching the dials.

Cracking: The Hot-Short Trap

Silicon bronze is hot short, meaning it loses ductility in a narrow band just below solidification. Overheat it and the bead cracks along its centerline or pulls apart at the crater as it cools. Restraint makes it worse: a tightly clamped lap joint with no room to move puts the shrinking bead in tension.

The countermeasure is boring and effective. Keep the pool small, cap interpass temperature, and let each bead cool before stacking the next. If a crack shows up in service rather than during welding, suspect insufficient bonded area rather than filler strength.

Cold Lap and Poor Wetting

Bronze does not fuse steel; it wets and tins the surface, so the joint depends on cleanliness and capillary action. Oil, paint, mill scale, and silicone residue block wetting, and a bead laid over them sits on top like a loaf instead of tying in.

A crowned bead with edges that lift under a chisel means a dirty surface or travel that is too cold and too fast. Clean to bright metal, then add just enough energy to see the puddle wet out and flatten at the toes.

Bird-Nesting, Shaving, and Feed Faults

Silicon bronze wire is soft enough to deform in the drive rolls, so it behaves much like aluminum in the feed path. Pinching or abrading it causes shaving, hesitation, and porosity as damaged wire drags contamination into the arc. Use formed or half-round rolls, set tension to the minimum that feeds without slipping, and keep the gun lead straight. The setup principles in our aluminum MIG welding guide apply almost word for word here.

Spatter and arc harshness usually point at gas rather than technique. An argon-dominant mix with a small CO2 addition stabilizes the arc, but too much CO2 raises heat input and oxidizes the alloy. Burnback and tip failure are feed problems in disguise: hesitation lets the wire stub into the puddle and the tip welds shut.

Reading the Cross-Section and Running a Tear-Out Test

Cut a finished lap joint across the bead and look at the interface. A good braze weld shows a thin, continuous tinning line where bronze met steel, with no voids or lift-off. Gaps, black oxide lines, or a bead on an unbonded shelf mean the joint never wet out, however smooth the cap looks.

For a field check, weld a scrap lap coupon and tear it apart in a vise. A proper brazed lap joint tears out along the bead on the top coupon, with parent metal failing while bronze stays attached. If the bead peels off cleanly, bonding was poor. I-CAR’s MIG brazing guidance describes the same tear-out method used in collision repair training.

Short Note: With bronze, a bad weld and an ugly weld are different things. A crowned, slightly rippled bead that passes tear-out is serviceable. A smooth, pretty bead that peels off is scrap. Judge the interface, not the cosmetics.

Strength, Testing, and Safety: What the Numbers Do and Don’t Tell You

Silicon bronze is often sold on its numbers: roughly 50,000 psi tensile strength and high elongation. Those figures are real, but they describe a test specimen pulled along its length, not the lap joint you just ran across a rocker panel. Understanding the gap between the two is what separates a sound repair from a good-looking failure.

What 50,000 psi Means in a Real Joint

That tensile result comes from a standardized coupon loaded uniformly along its axis. In a lap joint, load enters through the overlap, spreads across the bonded interface, and concentrates at the toes. The filler’s strength matters far less than how much bonded area you created and how that area is loaded. A narrow, crowned bead with poor wetting can test strong on paper and peel apart in service.

High elongation is a genuine advantage. The bronze stretches before it tears, which absorbs impact and forgives minor fit-up error. It also means the joint can bend rather than hold a precise shape, so don’t treat it as a rigidity solution. Butt joints deserve special caution: this is a braze-welding process, and a full-penetration butt joint in steel is rarely the right design. Where a butt configuration is unavoidable, bevel and gap the joint so the bronze can wet both faces rather than simply bridge the top.

Peel, Shear, and Fatigue: Where Brazed Joints Differ

Brazed joints are strongest in shear and weakest in peel. A lap joint pulled straight apart concentrates stress on a thin line at the bead toe, and that is exactly where a bronze joint will start to open. Design overlaps so service loads run along the interface, not perpendicular to it. Fatigue behavior is also different from a fused steel weld: the softer bronze and the sharp toe transition can initiate cracks under repeated loading, even when the static strength looks adequate. Never substitute bronze for a structural weld that an engineer designed as fusion welding.

When You Need a Procedure and Destructive Testing

For cosmetic or non-structural work, careful technique plus a tear-out test on a scrap coupon may be enough. For structural, pressure, or OEM-certified repairs, you need a written procedure, documented settings, and destructive or non-destructive verification. That is the same discipline used in procedure qualification for specialty alloys. In collision repair, I-CAR guidance on MIG brazing exists precisely because OEM procedures govern whether the process is permitted at a given joint. Generic internet advice does not override the vehicle maker’s document.

The Safety Numbers That Matter

Bronze MIG runs cooler than steel fusion welding, but the fume is not harmless. Copper fume and zinc oxide are the two exposures to control. These are documented occupational limits, not theoretical concerns.

Substance	OSHA PEL (8-hour TWA)	Notes
Copper fume	0.1 mg/m³	Copper dusts and mists: 1 mg/m³
Zinc oxide fume	5 mg/m³	ACGIH recommends 2 mg/m³ respirable TLV

The copper fume limit is low enough that ordinary shop air can exceed it quickly in a small bay. The values sit in OSHA’s air contaminants table and are summarized in OSHA’s copper fume data sheet. Zinc oxide has its own health fact sheet worth reading before the next galvanized job.

Metal Fume Fever and Galvanized Steel

Metal fume fever is the classic galvanized-steel outcome: fever, chills, muscle aches, nausea, and a dry throat that show up several hours after exposure and usually clear within 24 to 48 hours. NIOSH notes an estimate that 30 to 40 percent of all welders have experienced it at some point. Because symptoms arrive hours later, people often blame a cold instead of the fume.

Ventilation, Respirators, and the Beryllium Exception

Low-heat brazing generates less fume than fusion welding, but it does not eliminate the need for ventilation. Use source-capture extraction at the arc, keep your head out of the plume, and reserve respirators for situations where engineering controls cannot keep exposure below the limits. A half-face respirator with particulate filters is the usual minimum; check the cartridge or filter against the specific metals involved.

One hard stop: beryllium-bearing copper alloys require far stricter controls and fall outside ordinary silicon bronze work. If the base metal or filler contains beryllium, stop and get expert guidance before striking an arc. For everything else, the honest summary is simple. The strength numbers are a starting point, joint design and bonded area decide real performance, and the fume limits decide whether you can safely make the weld at all.

Cost, Alternatives, and the Practical Verdict

Silicon bronze is not the cheap filler. It is often the cheaper job. The wire price is one line on the estimate; heat damage, coating repair, grinding time, warpage, and comebacks are the others.

Process	Heat and coating	Speed and skill	Appearance and finishing	Cost reality
Silicon bronze MIG	Low heat, preserves much of the zinc nearby	Fast on long seams, moderate skill	Minimal spatter, soft bead dresses quickly	High filler cost, extra setup cost, low rework
ER70S-6 steel MIG	Higher heat, burns coating and can warp thin sheet	Fast, familiar, easy to learn	Spatter and hard grinding, visible seams	Cheap wire, higher finishing and rework hours
TIG brazing	Very controlled, low heat	Slow, high skill, excellent for small joints	Clean, precise bead	High labor cost, low consumable waste
Squeeze-type resistance spot welding	Localized heat, OEM-preferred on many steel assemblies	Very fast where access allows	No visible bead, minimal dressing	High equipment cost, low per-joint time
Structural adhesive	No welding heat, excellent sealing	Moderate speed, strict surface prep and fixturing	Invisible joint, paint-ready	Material and cure time, validated procedures required

Copper-alloy filler costs substantially more per pound than mild-steel MIG wire, so the economics depend on avoiding rework and corrosion rather than on filler price alone. Dedicated soft drive rolls and a Teflon, plastic-graphite, or carbon-fiber liner add real setup cost, and sharing a feeder with steel wire risks carbon contamination. Soft-wire feeding follows the same discipline as aluminum MIG welding: low drive-roll tension, a clean liner, and no sharp bends in the conduit.

Those costs are repaid quickly on thin galvanized panels. Reduced warpage and burn-through translate directly into fewer pulled dents, less hammer time, and fewer ruined coatings. Minimal spatter and a soft bead that dresses fast offset some filler cost in cosmetic work. For gas, liners, and contact tips, a reliable local supplier matters more than a few dollars saved online; our Airgas welding supplies guide covers what to keep on the shelf.

Decision Checklist for Your Job

• Does the OEM procedure allow brazing on this joint, or specify spot welding?
• Is the steel coated, thin, high-strength, or heat-sensitive?
• Is the joint loaded in shear, or in peel where bronze is weakest?
• Can you reach both sides for cleaning, fit-up, and inspection?
• Do you have a dedicated liner, soft rolls, and 100% argon?
• Can you run a scrap coupon and a tear-out check before the panel?

Squeeze-type resistance spot welding remains the OEM-preferred method on many modern steel assemblies, and MIG brazing is often specified only for particular joints. Follow the vehicle maker’s published position, not a generic forum answer, and check I-CAR’s repair guidance when the procedure is unclear.

When Silicon Bronze Is the Wrong Answer

Skip it where full fusion is required, where the joint sees high peel stress, or where a code-qualified structural weld is mandatory. It is also a poor choice when the OEM says spot welding or adhesive bonding, when the part is thick enough that preheat and deep penetration are needed, or when nobody can keep steel wire out of the shared feeder.

Five Steps to Start a Job Properly

1. Confirm the method. Read the OEM procedure and confirm brazing is permitted for that exact joint.
2. Dedicate the setup. Fit soft drive rolls, the correct liner, and a matching contact tip, then purge and clean the torch.
3. Prepare the surfaces. Remove paint and heavy zinc at the weld zone, leave the surrounding coating intact, and set a small root gap where the joint allows it.
4. Prove it on scrap. Tune low, run a test bead on the same gauge and coating, then tear or peel the coupon to confirm wetting.
5. Weld cool and inspect. Push the torch, keep the bead slightly crowned, watch the coating, and dress only after the joint cools.

That is the verdict. Silicon bronze earns its price on thin coated steel and mixed-material joints where heat control, coating survival, and fast finishing matter more than raw tensile strength. It is a specialist tool, not a universal replacement for steel MIG.

Return to that rocker panel. The shop did not need a hotter arc or a better welder. It needed a filler that melted below the steel, kept the zinc alive along the seam, and left a soft bead that sanded flat. Six months later, the difference is not in the weld appearance. It is in whether the rust has started.

Frequently Asked Questions

Can you MIG weld silicon bronze with a regular MIG welder?

Yes, in most cases. A standard GMAW power source can run ERCuSi-A provided it holds a stable arc at low current. You will need soft or half-round drive rolls, a Teflon, plastic-graphite, or carbon-fiber liner, 100% argon shielding gas, and electrode-positive polarity. A synergic or pulsed machine makes the process considerably easier to control.

Is MIG brazing with silicon bronze as strong as steel MIG welding?

It depends entirely on joint design and loading. The deposited bronze typically tests around 50,000 psi tensile with high elongation, and a well-made lap joint can be comparable in shear to a steel weld. However, because the base metal is not fused, the joint is generally less tolerant of peel and certain fatigue loadings, so it should not be substituted blindly for a specified fusion weld.

Why does my silicon bronze weld keep cracking?

Overheating is the usual cause. Silicon bronze is hot short, meaning it loses ductility at elevated temperature, so a large puddle, slow travel, or excessive interpass temperature leaves the bead vulnerable to cracking as it contracts. Keep the pool small, use stringer beads, control interpass temperature, and avoid weaving.

What shielding gas should I use for silicon bronze MIG welding?

100% argon is the standard choice at roughly 25–30 cfh. For thicker sections, a 75% argon / 25% helium mix improves heat input and wetting. Some shops add a small amount of CO2 for arc stability, but pure argon remains the default and the safest starting point for thin galvanized sheet.

Can silicon bronze be used on galvanized steel without grinding the coating off?

That is precisely why the process is popular. Because the filler melts well below the temperature at which zinc vaporizes in bulk, the coating immediately adjacent to the bead is largely preserved, and the zinc that is disturbed can partly recover. You should still clean oil and dirt from the joint, and you still need ventilation for zinc fume.

Can I use silicon bronze to join steel to copper or brass?

Yes. ERCuSi-A is specifically intended for joining copper, copper-silicon, and copper-zinc alloys to themselves and to steel. It is one of the more forgiving ways to make those dissimilar joints, though you should preheat thick copper sections and keep the interpass temperature under control.

Does MIG brazing count as welding for structural repair?

Technically it is a braze-welding process, not fusion welding, and that distinction matters. Automotive OEM repair procedures specify exactly where MIG brazing is and is not permitted, and those procedures take precedence over general advice. Never substitute brazing for a specified structural weld without documented approval.

Is silicon bronze welding dangerous?

It carries the same fume hazards as any arc process, with specific concerns for copper fume and, on galvanized steel, zinc oxide fume. OSHA limits copper fume to 0.1 mg/m³ and zinc oxide fume to 5 mg/m³ as eight-hour averages. Use local exhaust ventilation, position the torch so vapor escapes the breathing zone, and wear appropriate respiratory protection when ventilation is inadequate.

What size silicon bronze wire should I start with?

For thin galvanized sheet in the 0.8–1.2 mm range, 0.030 in (0.8 mm) or 0.035 in (0.9 mm) wire is the usual starting point at low amperage. Step up to 0.045 in for heavier sheet and 1/16 in for thick sections or build-up work, adjusting wire feed speed and voltage to match.

Can I run silicon bronze through a spool gun?

Yes, and a spool gun is often the better choice for the soft 0.030 in wire because it shortens the feed path and reduces the chance of shaving or bird-nesting. If you use a push feeder, fit a low-friction liner and keep the torch lead as straight as possible.