Welding Dissimilar Metals: The Complete Guide to Fillers, Challenges, and Techniques

Table of Contents

Introduction: Why Welding Dissimilar Metals Is the Ultimate Test of a Welder

You strike an arc on a stainless steel bracket welded to a mild steel frame. The bead looks perfect — smooth ripples, good tie-in, no visible undercut. Six months later it cracks along the fusion line, right where you never thought to look. Nothing about your torch technique was wrong. The joint failed because of metallurgy.

That is the humbling reality of welding dissimilar metals. It is the point where good hands meet hard science, and where the difference between a joint that lasts decades and one that fails in months usually comes down to a few decisions you make before you ever light up.

What Counts as “Dissimilar”?

Dissimilar metals are not just “two different metals.” The term covers any pairing that differs in chemical composition, mechanical properties, or crystal structure — and sometimes all three at once. Stainless steel to carbon steel. Aluminum to steel. Copper to nickel. Even two grades within the same family, like 304 to 316 stainless, behave differently enough to demand attention.

Each mismatch introduces its own complications: different melting points, different thermal expansion rates, different electrical and thermal conductivity. As the Australian Stainless Steel Development Association explains, those differences change how the weld pool freezes, how heat travels, and how the final joint behaves under load.

Why Modern Fabrication Leaves You No Choice

Engineers rarely pair dissimilar metals because they want to. They do it because it is economically and technically smart. Combining materials delivers tailor-made properties while lowering cost, since expensive corrosion-resistant alloys are used only where they are actually needed. A vessel can have a corrosion-resistant interior and a cheap, strong carbon steel exterior. Nobody pays for nickel in a place that never sees acid.

That logic shows up everywhere:

Petrochemical and oil & gas: pressure vessels, boilers, heat exchangers, and process piping join carbon steel to stainless and nickel alloys daily.
Power generation and nuclear: dissimilar metal welds connect ferritic and austenitic components in high-temperature, high-pressure systems, where AWS G2.1M/G2.1 gives formal guidance.
Automotive lightweighting: multi-material structures increasingly join aluminum to steel to cut weight and improve fuel economy.
Shipbuilding and construction: stainless balustrade posts anchor to structural steel, and doubler plates tie supports to stainless steel vessels.
Food and pharmaceutical processing: sanitary stainless systems tie into utility carbon steel with careful transitions.

Quick Note: A dissimilar joint is not automatically a weak joint. It is a joint with a different set of rules — and those rules can be learned.

What This Guide Covers

Over the next sections we work through the whole picture: the science of melting points and thermal expansion, dilution and intermetallic compounds, filler metal selection, specific metal combinations, process choice between fusion and solid-state joining, distortion and carbon migration, corrosion and galvanic concerns, and the safety and code requirements you cannot skip.

It is written for working welders, fabricators, engineers, and inspectors who need practical answers — not a textbook. Whether you are laying a root pass on a stainless-to-carbon nozzle or planning a production cell, the fundamentals are the same. If you are still building your baseline arc control, start with our guide to welding practice, because clean technique makes every metallurgical fix easier to achieve.

You will also see how process details matter. Polarity choices on a TIG setup reshape heat input and dilution, which is why we link back to TIG welding polarity when relevant. Voltage, travel speed, and bead placement all feed the same outcome. Automation raises the stakes further — high-deposition processes like robotic submerged arc welding reward good procedure design and punish bad fit-up. None of this is guesswork.

Here is the promise: with the right filler metal, the right process, and disciplined heat control, dissimilar joints can be every bit as reliable as same-metal welds. Aerospace and nuclear programs have proven it for decades. The rest of this guide shows you how to get there.

The Science Behind the Difficulty: Melting Point, Thermal Expansion, and Conductivity

Every failed dissimilar joint traces back to the same root cause: the two base metals simply do not behave the same way when heat arrives. Before a single drop of filler touches the puddle, their physical properties are already working against each other. Three mismatches matter most — melting point, thermal expansion, and thermal conductivity — and a welder who reads them can predict distortion, cracking, and cold lap before striking an arc.

Melting Point: One Metal Liquefies While the Other Barely Warms

The starkest gap appears when you pair aluminium with steel. Aluminium melts around 660 °C, while steel stays solid until roughly 1,370–1,540 °C. By the time the steel is hot enough to fuse, the aluminium has already been molten and fluid for hundreds of degrees, its protective oxide skin reforming almost instantly as it cools. That is why aluminium-to-steel fusion welds are so unforgiving, and why processes such as DC welding aluminium demand such precise heat control. Faced with this gap, many shops switch to bimetallic transition joints or solid-state methods instead of fighting a straight fusion weld. There is no single amperage that satisfies both halves of the joint, which is why so many operations reach for a process that concentrates energy at the arc.

Even between steels the gap survives in miniature: low-carbon steel melts near 1,500 °C, austenitic stainless a little lower, and cast iron far lower still. The softer the mismatch, the more forgiving the weld — but the physics never disappears.

Thermal Expansion: Two Metals That Grow at Different Rates

Expansion mismatch shows up as movement — or as stress when that movement is restrained. Mild steel expands about 12 μm/m/°C, while austenitic stainless steel expands roughly 17 μm/m/°C over the 0–300 °C range. The published coefficients run about 10 ×10⁻⁶/°C for ferritic AISI 430, 13 ×10⁻⁶/°C for carbon steel, and 17 ×10⁻⁶/°C for austenitic AISI 304. Weld a 304 plate to a carbon steel plate and the stainless side wants to grow roughly 40 percent more for the same temperature rise. Clamp that movement in a jig and the extra expansion converts straight into residual stress along the fusion line — the exact place where these joints tend to fail first.

Thermal Conductivity: One Side Pulls Heat Away Faster Than the Other

Conductivity decides where heat travels and how quickly each side cools. Carbon steel moves heat at roughly 49 W/m·K, while austenitic stainless manages only about 15 W/m·K at 200 °C. Energy poured into the stainless side piles up near the arc, forming a wide, slow-cooling pool; the carbon steel side wicks heat away and chills fast. That asymmetry changes penetration, bead profile, and solidification rate across the same joint. Studies of dissimilar welds confirm that this uneven heat flow governs where residual stress concentrates (Metals, 2024). It also means the same travel speed and current that weld beautifully on carbon steel can leave the stainless side undercut — one reason operators choose their current type deliberately, as our guide to AC or DC for TIG welding steel explains.

How These Gaps Drive Solidification, Cracking, and Distortion

Combine the three and you get the solidification behaviour behind most failures. The faster-cooling steel side solidifies first and contracts hard, while the stainless side is still hot and expanding. The weld metal is trapped between them. If its own expansion coefficient falls outside the range of the two parents, it becomes the weak link — which is precisely why filler selection is decisive and why an interlayer or buttering pass with an intermediate-expansion alloy buffers the mismatch. Ideally the weld metal expands at a rate between the two parent metals, so the stress it carries stays balanced.

Quick Note: You cannot change a metal’s expansion coefficient, but you can control how much it moves. Preheat, slow cooling, balanced weld sequencing, and sound fixturing turn a stress problem into a manageable one — skills worth drilling in deliberate welding practice. Heat direction matters too; understanding something as simple as TIG polarity keeps the heat where you want it.

None of these physical mismatches is a defect on its own — they are simply the rules of the game. But together they create the conditions that let the next problem flourish: dilution and the brittle intermetallic compounds that form the moment two unlike metals try to mix in the molten pool. That is exactly where we go next.

Dilution and Intermetallic Compounds: The Core Metallurgical Challenge

When you strike an arc on dissimilar metals, the weld pool is never just filler metal. It is a mixture — molten parent metal from both sides swirling together with whatever you feed into the arc. That mixing is called dilution, and it is the single biggest reason dissimilar joints behave so differently from matched ones. Handle it well and you get a sound, ductile bead. Ignore it and you build a brittle, crack-prone joint that only reveals its weakness months later.

The Lever Rule: Your Weld Is a Recipe

Think of the final weld composition as a recipe governed by a simple balance — often described with the lever rule. The chemistry of the finished bead depends on the ratio of each parent metal to the filler. Dilute 30% mild steel into a stainless-to-carbon joint and your weld metal becomes roughly a third carbon steel by composition. Change the groove angle, joint prep, or travel speed and that ratio shifts, along with strength, corrosion resistance, and cracking risk. Because dilution swings widely on the same job, it has to be measured and managed, not guessed.

Arc concentration matters too. Polarity choices that deepen or spread the melt change how much parent metal enters the pool, which is why it pays to understand how TIG welding polarity and DC welding aluminium reshape the fusion zone. The ASSDA welding dissimilar metals FAQ reflects this reality.

Intermetallic Compounds: The Brittle-Phase Problem

When dissimilar metals melt together and cool, atoms from each side combine into intermetallic compounds (IMCs) — ordered crystal structures with their own distinct chemistry and their own unfortunate properties. They are not a defect you can wish away. IMC formation actually indicates that a metallurgical bond has formed, so the goal is control rather than elimination, as the AWS G2.1 guide to joining dissimilar metals makes clear.

The most common brittle phases by system:

Aluminum–steel: Fe2Al5, FeAl3, FeAl
Aluminum–magnesium: Al3Mg2 and Al12Mg17
Aluminum–copper: CuAl2
Titanium–aluminum: TiAl3 and Ti2Al20

These phases are hard and unforgiving. In Al–Mg welds, Al3Mg2 and Al12Mg17 reach hardness values of 152–221 HV, versus only 25–60 HV for pure aluminum or magnesium. That gap is why a bead that looks flawless outside can crack clean along the fusion line, far from the visible weld.

Critical Thickness: Why Too Little and Too Much Both Fail

An IMC layer can be tolerated if it stays thin, uniform, and continuous. Embrittlement of aluminum–steel welds can be minimized when the intermetallic layer stays below about 10 μm. For aluminum–steel friction welding, a reaction layer as thin as 1–2 μm is recommended for strong bond strength. The trap is that more bonding is not automatically better. Drive heat and time too far and the layer thickens into a brittle, crack-initiating band; pull back too far and you never achieve real metallurgical continuity.

Quick Note: That 1–2 μm window is narrow for a reason. Bonding must carry load, yet stay shallow enough that brittle phases never get room to grow.

Heat Input Is the Master Variable

IMC growth is governed by temperature and time — in other words, heat input. Lower the peak temperature and shorten the time above the reaction threshold, and the layer stays thin. Underwater friction stir welding proved the point in published metals research: it reduced IMC thickness from about 7 μm to about 1 μm while lowering peak temperature from 1015 K to 883 K.

In arc welding, manage the same physics with lower amperage, faster travel, controlled interpass temperature, and fit-up that avoids excessive root penetration. Wire-feed setups reward that discipline; keeping consumables and settings predictable — as covered in our contact tip welding guide and gasless MIG guide — keeps dilution repeatable. Automation helps most: a robot reproduces identical heat input pass after pass, which is why robotic submerged arc welding is trusted on heavy dissimilar sections. Holding these variables steady is pure craft, earned through welding practice.

The takeaway is simple: dilution sets your weld’s chemistry, and heat input sets how thick the brittle intermetallic layer grows. Neither is random. Both are yours to control — and that control separates a joint that lasts decades from one that cracks along a line you never thought to inspect.

Filler Metal Selection: The Ingredient That Makes or Breaks the Joint

If metallurgy decides what’s possible in a dissimilar joint, the filler metal decides what you actually get. It’s the one variable you control completely — the designer’s strongest tool for steering weld chemistry away from brittle intermetallics and toward something ductile and predictable. Pick wrong, and a flawless-looking bead can crack along the fusion line months later. Pick right, and the joint quietly outlasts the machine around it.

That’s why it pays to know exactly what you’re feeding into the arc. Our MIG wire composition explainer breaks down how a consumable’s alloying elements shape the finished weld — the same logic drives every choice below. It’s also a skill worth drilling, as our guide to building real welding skills makes clear.

The 309 rule of thumb

Standards such as AS/NZS 1554.6 give you a starting point worth memorizing: carbon steel joined to 304(L) takes a 309L filler, and carbon steel to 316(L) takes 309LMo. Those “over-alloyed” austenitic fillers arrive with extra chromium and nickel so that, even after dilution by the carbon-steel side, the weld still solidifies as tough austenite with a little ferrite instead of hard martensite.

The major filler families

Austenitic stainless (309L, 309LMo): the workhorse for carbon-to-stainless work. Cheap, forgiving, and widely stocked. They tolerate moderate dilution but lose corrosion resistance and ferrite balance if the carbon-steel melt-in gets excessive.
Nickel-based (ERNiCr-3, ENiCrFe-3): defined by AWS 5.14 and AWS 5.11, these are the standard consumables for joints running above roughly 400 °C. Nickel weld metal dissolves a wide range of diluting elements without cracking — but it is sensitive to lead, sulfur, and phosphorus, so clean base metal is non-negotiable.
Nickel-copper (Monel-type): excellent for nickel-to-steel and some dissimilar service, but it tolerates only about 10% iron or 5% chromium before cracking risk climbs — and even less if the part is thermally stress relieved.
Cupronickel and aluminum bronze: used for copper alloys, bronzes, and some steel-to-copper joints, offering good wear and seawater resistance where nickel fillers are overkill. They demand clean, tight fit-up and steady, low-dilution technique.

Quick Note: Carbon and low-alloy steel fillers are not normally recommended for dissimilar welding. Diluted alloying elements raise hardenability in the weld pool and promote post-weld cracking — the opposite of what you want at a restrained joint.

Don’t underestimate the delivery side, either. A worn contact tip or inconsistent wire feed can wreck the deposition rate and dilution you carefully planned, so keep that part of the equation tight.

Why dilution tolerance is everything

Every filler has a dilution budget. Austenitic stainless fillers accept a fair amount of iron before ferrite content and corrosion resistance drift. Nickel-based fillers accept far more — that’s their superpower — but their tight limits on lead, sulfur, and phosphorus mean oily, painted, or sulfur-bearing steel must be cleaned back to bright metal first.

Buttering: building a bridge before the real weld

Buttering means depositing a compatible layer on one parent metal before the final joint is made. A high-nickel buttering layer (such as N08800 or N08330) gives a thermal expansion coefficient intermediate between carbon steel and stainless steel, buffering the mismatch that would otherwise tear the fusion line. It also blocks carbon migration from the carbon-steel side into the stainless weld. Butter the carbon-steel face, stress relieve it if the code requires, then complete the joint — a technique that turns an impossible mismatch into a routine weld.

Predicting the result with Schaeffler and WRC diagrams

You don’t have to guess at ferrite content. The Schaeffler and WRC-1992 constitutional diagrams plot chromium and nickel equivalents, letting you estimate the weld’s final microstructure from the filler and the expected dilution. Aim for a few percent ferrite to resist solidification cracking without tipping into excessive hardness. For consistent results, dial in your process with TIG polarity and heat-input control, and treat filler selection as an engineering decision — not a coin flip. The diagrams won’t do the welding for you, but they stop you guessing.

Common Dissimilar Metal Combinations and How to Weld Each One

Metallurgy stops being abstract the moment you pick a filler for a specific pair. Here are the combinations that land on real shop floors, with the filler, process, and the trap waiting in each.

Carbon and Low-Alloy Steel to Austenitic Stainless

This is the pairing welders get wrong most often. Carbon steel must not be welded directly to austenitic stainless without a high-alloy filler, because the solidified weld metal forms low-ductility martensite that is likely to crack as it contracts. Use over-alloyed consumables such as 309L (ER309L) for SMAW, GTAW, and GMAW, or nickel-based fillers like ERNiCr-3 where service is severe. Keep polarity clean too — DCEN holds a stable arc on the stainless side, as our TIG polarity guide explains.

Heat also drives carbon diffusion from the ferritic base steel toward the austenitic weld. The result is a decarburized zone roughly 50–100 μm wide on the ferritic side, paired with a harder carbide-rich band on the stainless side. The Australian Stainless Steel Development Association notes that this migration accelerates above roughly 425°C (800°F), which is why these joints are typically limited to lower service temperatures unless a nickel barrier layer is used. For related heat-affected trouble in hardenable grades, see our guide to welding hardened steel.

Steel to Galvanized Steel

Galvanized steel is weldable, but the zinc coating is the enemy. Zinc from galvanized coatings must be fully removed from the heated zone, or it causes liquid embrittlement and cracking along the zinc penetration line. Grind back the coating at least 25–50 mm on both sides, weld in a ventilated area, and treat the fumes as a serious exposure risk — see OSHA’s welding guidance. Restore protection afterward with a zinc-rich coating.

Aluminum to Steel

Arc welding processes are generally viewed as nonrecommended for aluminum-steel because of melting and solidification problems — the two metals refuse to form a sound fusion bond and brittle intermetallics (especially Fe₂Al₅) form at the interface. Friction-based solid-state processes such as friction stir welding and friction welding are preferred. If you work aluminum on its own, our guide to DC aluminum welding covers polarity and setup.

Aluminum to Copper

Both conduct heat aggressively, so you need fast travel and generous heat input. Use aluminum-silicon (ER4043) or aluminum-copper fillers, expect a brittle Al-CuAl₂ intermetallic at the interface, and keep dilution low. Keep the arc moving fast, too — the copper wicks heat away while the aluminum edge can collapse before the puddle forms. Brazing is often the safer commercial choice.

Copper to Steel

Copper and iron mix completely in the liquid state but have limited solid solubility, giving wide freezing ranges and a high risk of hot cracking. Preheat the copper side heavily (200–500°C depending on thickness) because it pulls heat away fast, use nickel-based or copper-nickel fillers such as ERCuNi, and favor processes that limit dilution.

Titanium to Steel and Nickel

Titanium is incompatible with iron; solid solubility is limited and brittle intermetallic compounds form in the solid state. Direct fusion is off the table. The standard fix is an insert or transition piece — a vanadium or copper-nickel barrier, or buttering with a compatible filler — plus inert shielding so strict it belongs in a glovebox, not a garage.

Magnesium, Nickel Alloys, and the Rest

Magnesium-to-aluminum joints form the brittle Al₁₂Mg₁₇ phase and are almost always joined mechanically or by solid-state welding. Nickel alloys are the universal bridge: they tolerate dilution from both ferritic and austenitic sides and appear in heat exchangers, transition joints, and power-plant piping. Monel and Inconel fillers earn their price here. Read the service environment before choosing a bridge metal, because thermal cycling and moisture decide whether the joint survives.

Pairing	Preferred Filler	Key Precaution
Carbon steel + austenitic stainless	309L, ERNiCr-3	No direct fusion; limit service temperature
Carbon steel + galvanized steel	ER70S-6	Strip zinc 25–50 mm; ventilate
Aluminum + steel	None reliable	Solid-state joining only
Copper + steel	ERCuNi, nickel-based	Preheat copper; control dilution
Titanium + steel	Insert or buttering	Full inert shielding

Quick Note: When in doubt, over-alloy the filler. A richer consumable absorbs more base-metal dilution before the deposit turns brittle — cheap insurance for any mixed joint. For formal guidance, AWS G2.1M/G2.1 covers dissimilar joins in detail.

Choosing the Right Process: Fusion Welding vs. Solid-State Joining

Not every dissimilar joint should be welded with an arc. The process you pick decides how much heat enters the part, how much the two base metals dilute into each other, and how thick the fragile intermetallic layer grows. Get that choice right and the metallurgy mostly behaves. Get it wrong and no amount of steady hands will save the joint. The goal is to match the process to the metals, the thickness, and the service environment rather than defaulting to whatever machine sits closest to the bench.

The Fusion Toolbox and Its Heat-Input Trade-Offs

Fusion welding melts both metals and the filler into one pool. That is exactly why it is so sensitive to dilution. Each process brings a different heat and dilution profile:

GMAW/MIG — versatile and productive, but higher heat input widens the HAZ. On aluminium and stainless work, polarity matters, so review TIG polarity: DCEN, DCEP, and AC and the related DCEN vs DCEP logic before dialing in a procedure.
GTAW/TIG — the most controllable arc, ideal for thin sections and precise dilution control, but slow.
FCAW and SMAW — rugged for repair and field work; high heat input and slag management add variables.
SAW — enormous deposition for thick plate, best left to automated setups like robotic submerged arc welding.
Laser welding — high power density with very high cooling rates and short diffusion times, which limits intermetallic growth and makes it a strong pick for thin dissimilar sheets.

The enemy in fusion is excessive heat. Lower energy input means a thinner intermetallic compound (IMC) layer, and adaptive GMAW control—systems that regulate heat input, HAZ size, and IMC composition in real time—has measurably improved feasibility for dissimilar joints. Cold metal transfer (CMT) takes this further by mechanically retracting the wire to reduce current and spatter, giving precise, low-heat deposition. Even modest drops in current can pay off, because IMC thickness tends to grow with time at temperature.

Quick Note: If a joint must be fusion welded, choose the process that delivers the least heat consistent with full penetration. On dissimilar work, “just enough” beats “plenty.”

Solid-State Joining: Avoiding the Melt Entirely

Solid-state processes bond metals below their melting points, so dilution and IMC growth are largely side-stepped. Friction stir welding (FSW) was invented at The Welding Institute (TWI), Cambridge, in 1991 and adopted by NASA Marshall in 1995. A rotating tool stirs the plasticised metal into a forged joint—no filler, no shielding gas, no external heat—and in aluminium alloys it delivers higher joint efficiencies and fewer defects than fusion welding. Because the exit hole left by conventional friction stir spot welding concentrates stress and traps corrosion, refill FSSW was developed to close that hole completely. That lower thermal load also means less distortion and fewer residual stresses, which matters on thin, heat-sensitive assemblies.

Friction and diffusion welding — solid-state options for bars, tubes, and precision interfaces; diffusion bonding suits reactive or exotic pairs.
Ultrasonic, explosive, and magnetic pulse welding — purpose-built for sheet, dissimilar metal plate, and tubular overlap joints where melting is unacceptable.
Brazing and braze welding — lowest-temperature alternatives; the filler melts but the base metals do not, so dilution stays minimal. Great for thin sections and repair.

When Arc Welding Is Simply the Wrong Answer

Reach for solid-state or low-temperature joining when sections are thin, when IMC formation is acute (titanium to steel, aluminium to copper), when a leak-tight or hermetic seal is needed, or when the service environment punishes galvanic corrosion. Stick with fusion when you need high deposition on thick plate, field portability, or a qualified procedure that already exists. Gasless options like flux-cored wire can travel where shielding gas cannot—see gasless MIG and flux-cored welding for those cases.

Whichever route you take, the process only rewards you if your fundamentals are sound. Build those on consistent welding practice first, then let the metallurgy guide the machine.

Controlling Distortion, Residual Stress, and Carbon Migration

Two metals that expand and contract at different rates will fight your every pass. Distortion isn’t just cosmetic—it loads the joint with residual stress that can sit quietly for months before a crack finds the fusion line. Getting this right is mostly discipline with heat, and it starts long before the first arc.

Preheat: More Often Optional Than You Think

Ambient temperature is perfectly fine for most steel-to-stainless work—preheat isn’t automatic. It becomes necessary when the carbon steel side is over 0.2% carbon, thicker than 30 mm, or under high restraint. In those cases, about 150 °C is usually adequate: enough to slow the cooling rate and let hydrogen escape without overheating the austenitic side, a balance outlined in AWS G2.1. Heat the carbon steel and let conduction carry warmth across the joint rather than torching both plates.

Quick Note: Preheat the carbon steel, not the stainless. A torch dragged across a 304 plate mainly adds distortion and contamination risk without improving the joint.

Heat Input and Interpass Temperature

Keep heat input low and, more importantly, consistent. Stringer beads beat wide weaves because they limit the volume of hot metal per pass. Cap interpass temperature too—150 °C is a sensible ceiling for austenitic stainless, since the damage comes from accumulated heat exposure, not a single pass. A temperature crayon or contact thermometer takes the guesswork out. Process choice helps here as well: DCEN concentrates heat and keeps the puddle tight, as we explain in our guide to TIG welding polarity. These habits are the ones real welding practice builds over time.

The Tacking Sequence

Sequence your tacks so contraction balances across the whole joint: ends first, then the centre, then quarter points, then eighth points. That order locks the assembly evenly and pushes movement toward the free edges instead of letting it pile up at one end. Pair it with balanced, backstep passes and you give residual stress somewhere to go other than the fusion line.

Restraint is the other half of the story. A heavily restrained joint—thick sections, gussets, or a fixture that won’t let the metal move—converts shrinkage into stress instead of distortion, and that stress concentrates at the fusion line. Tacking order, preheat, and sequence all matter more, not less, as restraint rises.

Post-Weld Heat Treatment: A Double-Edged Sword

Here’s the dilemma that catches people out: post-weld heat treatment that relieves stress in carbon steel can actively harm the stainless side. Holding at 550 °C or 700 °C increases carbide precipitation, stripping chromium from grain boundaries—sensitization that can begin in as little as seven minutes at 700 °C in steel with only about 0.056% carbon, as documented by the Australian Stainless Steel Development Association. Selecting a lower-carbon grade like 304L reduces that risk, but on duplex and high-alloy grades, the same temperatures promote sigma phase formation, which depends on the Cr/Ni ratio and time at temperature. Sometimes the right call is choosing a filler and procedure that simply don’t require PWHT.

Carbon Migration, Sensitization, and Sigma Phase

At elevated temperature, carbon diffuses from the higher-carbon steel toward the stainless side, leaving a soft, decarburized band and a hard, brittle carburized zone beside the fusion line. A nickel buttering layer works as a diffusion barrier: in one CK45 carbon steel to AISI 304 joint, buttering prevented both the carburized and decarburized zones from forming. Thicker butter layers help, but even a thin, well-fused deposit changes the diffusion path. Above roughly 425 °C service temperature, carbon migration accelerates enough that high-nickel fillers such as ENiCrFe-3 become the practical choice.

Finally, keep an eye on your toes. The low heat input that protects these joints can tempt you into a lazy, slow puddle—and that’s how you create an overlap defect, a cold lap that hides incomplete fusion beneath a tidy-looking bead. Fast, controlled passes reduce both distortion and overlap at once.

Corrosion and Galvanic Concerns in Dissimilar Joints

You can nail the bead, pass the bend test, and still lose the joint to chemistry. When two dissimilar metals are electrically connected in an electrolyte — rain, condensation, salt spray, even damp concrete — they form a battery. The more active metal becomes the anode and corrodes; the more noble metal becomes the cathode and is protected. That is galvanic corrosion, and it is why dissimilar joints fail long after the weld looks fine.

The Galvanic Series: Predicting Who Corrodes

The galvanic series ranks metals by electrical potential in a given environment. The further apart two metals sit on that list, the stronger the corrosion driving force. Carbon steel sits well below stainless steel, so when the two are coupled, the steel is sacrificed. Crawford and colleagues showed this in a study of steel fasteners in concrete: coupling carbon steel to stainless steel raised the carbon steel corrosion rate by roughly 1% through the concrete path, versus about 2.4% for a passive carbon steel couple. Small lab numbers, but a persistent, decades-long drain on the anodic side.

Quick Note: The weaker, less passivated metal in a couple always pays the price. Design assuming the anode will corrode and give it extra material or a coating, not the cathode.

The Area Effect: Why Weld Beads Are Corrosion Magnets

Here is the part most fabricators underestimate. Galvanic corrosion severity depends on the ratio of cathodic to anodic area. When the corroding half of a couple is small relative to the cathodic half, corrosion accelerates because all the cathodic current concentrates on a tiny anode. A weld bead is usually small compared with the parent plate — so a small, active weld in a large, noble stainless plate is a textbook setup for rapid localized attack along the fusion line.

Carbon Steel Welded to Stainless: The Classic Problem Site

This combination turns up everywhere — bracket to frame, handrail to post, tank nozzle to shell. A common but imperfect fix is to paint the carbon steel plus the weld and carry the coating 10–50 mm onto the stainless steel. That overlap blocks the electrolyte path across the couple. It is imperfect because coatings scratch, and any pinhole becomes a concentrated anode.

Surface quality matters here, too: rough, undercut, or spattered beads trap moisture. Reading a bead is a corrosion-control skill as much as a cosmetic one — see our guide to good and bad MIG welds for what to look for. A clean, smooth profile helps coatings perform better, while good fit-up and consistent technique, built through deliberate welding practice, reduce crevices that initiate attack.

Microscopic Galvanic Cells and Hydrogen Embrittlement

Galvanic effects are not only macro-scale. In highly alloyed welds, segregation during solidification creates tiny composition differences that act as microscopic galvanic cells, so corrosion can begin inside the weld itself. Martensitic fusion zones add a second hazard. Research on resistance spot welding of carbon steel to stainless steel found a martensitic fusion zone with hardness around 480 HV and a 42 ± 5% reduction in maximum elongation after hydrogen charging — a clear hydrogen embrittlement risk on top of the galvanic one.

Passivation: What You Cannot Do Here

Acid pickling and passivation restore the protective chromium oxide layer on stainless welds, and they are standard practice on all-stainless work. But they cannot be applied to a carbon-steel-to-stainless fabrication because the acids aggressively attack the carbon steel. You lose your best stainless restoration tool exactly when the joint is most vulnerable — which is why mechanical cleaning, coating strategy, and joint design carry the load instead.

Corrosion control in dissimilar joints is a design decision, not an afterthought. As AWS and industry guidance on dissimilar welding stress, material selection, isolation, and drainage often matter more than the weld itself. Get the electrochemistry right before the arc ever starts. For the broader picture, review our shop planning guide to build corrosion-aware practices into daily work.

Safety, Codes, and Procedure Qualification You Cannot Skip

A dissimilar joint is not a place to improvise. Even when the materials, filler, and technique look right on paper, the code framework exists because these joints fail in ways a visual inspection can’t catch. A crack that starts at a fusion line or a carbide-rich band may not show itself until the part is in service, under load, at temperature. Treat the paperwork and the PPE as part of the weld, not red tape around it. Skip either and you inherit a risk you can’t inspect away.

The Standards That Govern Dissimilar Welding

The backbone is ASME Boiler and Pressure Vessel Code Section IX, which governs welding procedure and performance qualification. Here’s the point many shops miss: a dissimilar joint still requires a qualified WPS and a supporting PQR. You cannot simply borrow the carbon steel procedure that’s been on the wall for fifteen years — the combination is a new essential variable. A change in filler classification, base metal grouping, or preheat often invalidates the old qualification entirely. That means new test coupons, new mechanical testing, and a welder re-qualified on the actual joint configuration. Documentation isn’t bureaucracy; it’s the trail that proves the joint was built to a repeatable process.

Beyond ASME, structural work leans on AWS D1.1 for steel and AWS D1.6 for stainless. For nickel and austenitic stainless joining, the AWS G2.1 and G2.3 guides are essential references. In Australia and New Zealand, AS/NZS 1554.6 covers structural steel welding. The Australian Stainless Steel Development Association reinforces that dissimilar welds demand tighter controls than matching-metal work, from joint prep through interpass temperature. Each of these documents exists because a matching-metal assumption has failed in the field.

Quick Note: Preheat is a code issue, not just a comfort one. Higher-carbon or heavily restrained plain carbon steel frequently needs preheat to prevent post-weld cracking, and your WPS must state the minimum.

The Fumes You Cannot See Coming

Dissimilar work stacks hazards. Zinc-coated steel produces fume linked to metal fume fever, and zinc must be removed before welding stainless to avoid embrittlement. Stainless and nickel alloys release hexavalent chromium and nickel fume — respirable carcinogens that demand local exhaust ventilation, not just an open door. Nickel alloys are especially sensitive to sulfur and phosphorus, which form low-melting-point eutectics that crack and later fail at high temperature. Since you can’t see any of this, your control is engineering: fume extraction at the arc, supplied-air or P100 respiratory protection, and proper eye and skin coverage. Beryllium-bearing copper alloys and cobalt alloys add further risk and deserve a written exposure assessment. Our guide to protective eyewear for welding breaks down lens shades for high-current work.

Safety also lives in the building itself. Layout, ventilation, and extraction are facility decisions made long before anyone strikes an arc — see our guide on planning a safe welding shop. OSHA regulates welding, cutting, and brazing safety under 29 CFR 1910.252, and NIOSH fume guidance underpins the exposure limits. Get the airflow wrong and every subsequent weld is a hazard regardless of how good the bead looks. Dedicated extraction and a written exposure plan turn guesswork into control. Whether you’re running practice plates or production, the same rules apply.

Note: Confirm your procedure is qualified before the first arc, document the filler and preheat, and ventilate for the worst fume in the mix — usually the stainless or nickel side.

Practical Tips, Common Mistakes, and Final Takeaways

If you have followed this guide from the top, one theme keeps surfacing: welding dissimilar metals rewards preparation and metallurgical thinking, not raw speed. A fast, beautiful bead means nothing if the filler, the heat input, or the dilution was wrong. The welder who understands what happens inside the puddle will always outlast the one who simply chases a clean cap. Here is how to turn all that theory into a repeatable shop habit.

Your Dissimilar Welding Checklist

Before you strike an arc, run this list. It takes two minutes and saves far more:

Confirm the parent metals. Positive identification beats a guess — a magnet tells you almost nothing about the difference between 304 and 316 stainless.
Match the filler to the more highly alloyed parent. For steel to stainless, 309L or a nickel alloy is the single most effective way to prevent cracking, as the Australian Stainless Steel Development Association (ASSDA) explains.
Estimate dilution. Know how much parent metal will mix into your weld before you choose current and joint design.
Keep heat input low. In aluminum-to-steel work, reducing heat input is the primary lever for controlling intermetallic compound (IMC) layer thickness.
Plan coatings and galvanic protection early rather than grinding and reworking later.
Qualify your procedure. A written procedure specification is your insurance policy against a failure nobody sees coming.

Tight, clean fit-up matters even more when thin walls and close joints are involved, because there is no filler volume to hide a poor gap. That discipline carries straight over to weld tube fittings, where root control decides whether a joint seals or weeps.

Common Mistakes That Crack Joints

Using carbon steel filler on a stainless joint. It dilutes chromium, invites cracking, and often corrodes first.
Ignoring dilution when setting current. Run too hot and you pull more parent metal into the puddle than your filler can tolerate.
Welding over galvanized coatings. Zinc must be fully removed — the fumes are toxic and porosity follows. Review OSHA’s welding and cutting guidance on ventilation and metal-fume exposure.
Applying post-weld heat treatment blindly. A PWHT cycle suited to carbon steel can sensitize or embrittle the stainless side.
Over-welding. Extra passes add heat, grow the IMC layer, and buy you nothing.
Skipping procedure qualification. Hoping a joint holds is not a plan.

Quick Note: Every one of these mistakes shares a root cause — treating a dissimilar joint like an ordinary one. Slow down and weld the metallurgy, not just the gap.

Smart Habits for Joints That Last

A thin, controlled IMC layer — on the order of 1–2 μm for aluminum-steel — is associated with strong joints, while thick layers reduce tensile strength and ductility, as materials research shows. Heat control is therefore everything. Keep operating stress on stainless-to-carbon joints to a minimum so thermal fatigue doesn’t find the fusion line years down the road. For joints expected to run above roughly 400 °C, high-nickel alloys are usually selected for their carbon tolerance, creep strength, and oxidation resistance.

Buttering a joint face as a buffer, calculating dilution honestly, and choosing solid-state processes for difficult pairs like aluminum to steel all stack the odds in your favor. That same planning-first mindset shows up in metal rolling and welding working together, where process order decides final quality.

Short Note: When in doubt on a difficult pair, practice on scrap first. The habits you build in structured welding practice are exactly what keep a dissimilar joint from surprising you.

Final Takeaways

Dissimilar metal welding is a thinking welder’s game. The chemistry, the filler choice, and the heat input matter more than how pretty the bead looks on day one. Choose the higher-alloyed parent’s compatible filler, control dilution and heat, buffer where you can, and respect that a thin IMC layer is a healthy one.

Start with good joint preparation, qualify your procedure, and let metallurgy guide your decisions instead of raw speed. Keep records of what worked so the next job starts from experience rather than hope. Do that, and the crack that never came will be the proof your preparation worked.

Frequently Asked Questions

Can you weld any two dissimilar metals together?

No. Many combinations can be fusion welded with the right filler and process, but others, such as aluminum to steel or titanium to steel, form brittle intermetallic compounds that make direct fusion welding unreliable. These pairs usually require solid-state processes, interlayers, or brazing instead.

What filler metal do you use to weld stainless steel to carbon steel?

The standard choice is a higher-alloy filler such as 309L for carbon steel to 304L, or 309LMo for carbon steel to 316L. For demanding or elevated-temperature service, nickel-based fillers like ERNiCr-3 are often preferred.

Why does stainless steel to carbon steel crack along the weld?

One common cause is using a filler that dilutes to a martensitic microstructure, which has low ductility and cracks as it contracts. Carbon migration and thermal expansion mismatch between the two steels also contribute to cracking and premature failure.

What is dilution in dissimilar metal welding?

Dilution is the proportion of melted base metal that mixes into the weld pool. It determines the final weld chemistry, so controlling it through current, travel speed, joint design, and filler selection is essential to avoid brittle or cracked joints.

Can you weld galvanized steel to stainless steel?

Yes, but the zinc coating must be completely removed from the heated zone first. Left in place, zinc can cause liquid embrittlement and cracking, and its fume is a serious health hazard.

Why is aluminum so hard to weld to steel?

Aluminum melts at around 660 °C while steel melts above 1,370 °C, and the two form brittle iron-aluminum intermetallic compounds. Fusion arc welding is generally not recommended for this pair; friction stir welding and friction welding are preferred.

What is buttering and when should you use it?

Buttering means depositing a compatible layer, often a high-nickel alloy, onto one parent metal before making the final joint. It buffers thermal expansion mismatch, blocks carbon migration, and improves weldability on difficult dissimilar combinations.

How do you prevent galvanic corrosion in dissimilar metal joints?

Keep the anodic and cathodic areas favorable, use coatings or barrier layers, and extend paint from the less noble metal over the joint and slightly onto the more noble metal. In some cases, electrical isolation or careful material selection is the best protection.