Welding Stainless Steel to Carbon Steel: Filler Selection, Procedures, and Corrosion Control
Table of Contents
A maintenance crew needs to tie a new stainless steel pipe to an existing carbon steel line. Without the right filler metal and precautions, a seemingly simple joint can become brittle, prone to cracking, or start corroding faster than either metal alone.
1) Introduction: Why SS to Carbon Steel Joints Are Common
Almost every processing plant, fabrication shop, and maintenance department eventually faces the same task: joining an austenitic stainless steel component to a carbon or low-alloy steel one. Technically this is called dissimilar metal welding, and in practice it means taking two materials with different chemistry, different thermal behavior, and different microstructures and fusing them into a single load-bearing, leak-tight joint. The stainless side is usually a 304, 304L, 316, or 316L grade; the carbon side is typically plain carbon steel or a low-alloy structural or pressure-steel grade such as A36, A106 Grade B, or A516 Plate.
According to the British Stainless Steel Association, welding stainless steels to other steels is a routine, everyday activity across the process and construction industries. It is not an exotic or experimental procedure. It is, however, a procedure where the ordinary habits used for steel-to-steel welding will get you into trouble.
Where These Joints Show Up
If you have ever walked a plant floor, you have already seen dozens of these transitions. Typical examples include:
- Piping tie-ins — a new stainless line connected to an existing carbon steel header, often in a shutdown window with tight schedule pressure.
- Vessels and tanks — stainless liners, nozzles, or internal cladding joined to carbon steel shells and heads.
- Structural transitions — stainless handrails, balustrades, and architectural trim welded to mild steel supports or base plates.
- Equipment repairs — replacing a stainless spool, flange, or wear plate on machinery with a carbon steel frame.
- Heat exchangers, stacks, and exhaust work — where corrosion resistance is only needed on the wet or hot-gas side.
The reason is simple economics and logic: corrosion resistance is only required where the product touches the metal, while strength and cost favor carbon steel everywhere else. Why pay for a full stainless assembly when only the wetted surface needs to survive the environment?
The Principal Question This Article Answers
The core question is straightforward: how do you weld stainless steel to carbon steel so the joint is crack-free, correctly diluted, mechanically sound, and not a corrosion liability? The short answer, which the rest of this article unpacks in detail, is that you design the weld as a stainless weld. The BSSA guidance is explicit on this point: in a stainless-to-carbon joint, the weld and the corrosion design are treated as stainless, not as carbon steel. The carbon steel side contributes strength and cost savings; it does not get to dictate filler chemistry or corrosion behavior.
Quick Note: The most common and most damaging mistake is assuming the joint is “mostly steel” and welding it with a carbon steel filler. That produces a brittle, hydrogen-crack-prone fusion zone. Dissimilar joints are designed stainless-first, from filler selection onward.
Scope of This Guide
This is a pragmatic guide for routine fabrication and maintenance work, not a research paper on interface metallurgy. It focuses on the decisions a welder, welder-qualifier, or supervisor actually has to make on the shop floor or in the field: which filler to pull from the cabinet, how much preheat the carbon side really needs, how to keep distortion under control, how to prepare and clean the joint, and how to keep the finished transition from becoming a galvanic corrosion cell in wet service.
The supporting questions are the ones that generate the most confusion in real jobs: why a 309L wire is used for a 304-to-A36 joint instead of matching either base metal; whether preheat is required at all; why the stainless side cannot simply be pickled after the weld; what changes when duplex stainless such as 2205 is involved; and how restraint and heat input affect a joint where the two sides expand at different rates. The Australian Stainless Steel Development Association and AWS guidance such as the AWS Guide to Welding Cast Iron and dissimilar-metal welding references provide the standards backbone for the procedures described here.
The common combinations covered are the ones you will actually meet: 304 or 316 stainless welded to plain carbon and low-alloy structural grades, plus the less frequent cases involving coated steel, high-restraint thick sections, and duplex grades. Electrode and consumable choice on the carbon side of a repair, including alloy steel welding electrodes, is treated where relevant, and field repair processes such as stick welding are discussed alongside TIG and MIG approaches.
By the end of the article you should be able to walk up to a stainless-to-carbon transition, identify what filler belongs in it, decide whether preheat and hydrogen control apply, plan the weld sequence to control distortion, and specify the corrosion protection the joint will need in service. That is the whole job: correct filler, controlled dilution, managed heat, and a joint that does not corrode itself apart after installation.
2) Metallurgy of Dissimilar Joints: Dilution, Martensite, and Ferrite Control
When welding stainless steel to carbon steel, the arc doesn’t just join two metals. It creates a mixed weld pool with its own chemistry. The key to a sound joint is controlling that chemistry from the moment it solidifies.
Dilution: The Pool Changes Composition
Dilution is the percentage of the weld pool that comes from the base metals rather than the filler. For a root pass running along the carbon steel side, it’s common to see 30 to 40 percent of the pool melted from the base metal, and sometimes more with a wide, washed-out bead. This influx of iron and carbon pulls chromium and nickel levels down. If enough of these are lost and the filler can’t compensate, the pool can solidify as austenite but transform to martensite on cooling. That martensite is hard, low in ductility, and highly sensitive to hydrogen, creating a real risk for cold cracking in restrained joints.
The Wrong Filler Creates a Brittle Structure
There is only one rule to follow here: never fuse austenitic stainless steel to carbon steel using carbon steel filler or weld metal. As the Australasian Stainless Steel Development Association sets out in its technical guidance, this configuration can form martensite, giving low ductility and a high risk of cracking (ASSDA Technical FAQ 9). The British Stainless Steel Association echoes this, noting the aim should always be to keep the deposited weld metal fully austenitic and higher in alloy than either parent (BSSA, welding stainless steels to other steels).
| Scenario | Pool composition | Resulting structure | Risk |
|---|---|---|---|
| Carbon steel filler, high dilution | Low Cr, elevated C | Martensite / hard mixed structure | Hydrogen cracking, low toughness |
| Matched stainless filler, low dilution | Close to parent 304 | Austenite, sometimes with hot-crack sensitivity | Solidification cracking |
| Over-alloyed filler (309 type) with moderate dilution | Cr and Ni restored above parent level | Austenite plus a little ferrite | Sound, ductile, crack-resistant |
Over-Alloying and Controlled Ferrite
The fix is deliberate over-alloying. A 309L filler carries roughly 24 percent chromium and 13 percent nickel, well above 304 stainless. Even with around 30 percent carbon steel dilution, this restores enough alloy content to solidify as austenite instead of martensite.
However, austenite alone is vulnerable to solidification cracking. As impurities like sulphur and phosphorus concentrate at grain boundaries, a small amount of delta ferrite is needed to break up that continuous network and allow for shrinkage strain. For this reason, standards such as AS/NZS 1554.6 and AWS filler metal specifications typically call for 3 to 10 ferrite number (FN) in the finished weld. Too little ferrite leaves the weld crack-prone, while excessive ferrite can promote brittle intermetallic phases, such as sigma, during long-term high-temperature service.
Quick Note: Ferrite is measured with a ferrite number gauge, not by eye. Always specify an FN range on the welding procedure and verify it, as a weld can appear sound while falling outside of that range.
Schaeffler and WRC-1992: Predicting the Result
Rather than guessing, these diagrams allow you to predict the weld structure before striking an arc. The Schaeffler diagram plots chromium equivalent against nickel equivalent to identify austenite, martensite, ferrite, or mixed fields. The newer WRC-1992 diagram is more accurate for Fe-Cr-Ni weld metals and can predict ferrite number directly, making it ideal for procedure specification. Both are endorsed as practical selection tools in ASSDA technical FAQs, with their weldability reasoning also covered by the CWB. Always plot for the most dilute condition, usually the root pass, as it presents the highest risk.
The Carbon Steel HAZ Can Harden Too
Cracking isn’t limited to the weld pool. On the carbon steel side, the heat-affected zone (HAZ) just beside the fusion line can also transform. Under rapid cooling and restraint, this zone can form hard martensite, especially with hydrogen present. The carbon equivalent, most often calculated with the IIW formula (CE = C + Mn/6 + (Cr + Mo + V)/5 + (Ni + Cu)/15), is used to gauge this hardenability. While stainless steel’s lower thermalconductivity than carbon steel, so heat doesn’t race away from the joint as fast as you might expect, and the stainless side tends to slow cooling in the adjacent region. However, restraint still works against you. An overstitched, fully boxed joint cools under stress, which is exactly when a hardened HAZ is most likely to crack.
For steels with a higher carbon equivalent, preheat, controlled heat input, and dry, low-hydrogen consumables become necessary, as covered in our guide to Alloy Steel Welding Electrodes. Many fabricators also set a hardness limit, commonly around 250 HV in the carbon steel HAZ, and verify it with a portable hardness tester after welding.
The metallurgical takeaway is simple. Keep the weld pool out of the martensite field by using a higher-alloyed filler, maintain just enough ferrite in the deposited metal to control solidification cracking, use the Schaeffler or WRC-1992 diagrams to prove your dilution assumption, and never treat the carbon steel HAZ as inert.
3) Filler Metal Selection: 309L, 309LMo, and When to Consider Nickel Alloys
Choosing the filler metal is the single most important decision when you weld stainless steel to carbon steel. Get it right and the joint behaves like a well-designed transition. Get it wrong and you can end up with a brittle, crack-sensitive weld or a corrosion problem that shows up months later. The good news is that the rules are well established, and once you understand why they exist, they are easy to remember.
The Golden Rule: Never Weld Stainless to Carbon Steel with Carbon Steel Filler
The most common field mistake is welding the joint with a plain carbon steel or low-alloy filler such as ER70S-6 or E7018 because it is already in the truck. Resist the temptation, even on a “temporary” repair. As the Australian Stainless Steel Development Association explains, carbon steel filler diluted with stainless steel produces a weld metal chemistry that can form hard, brittle martensite in the mixed zone, with a real risk of hydrogen cracking in service or during fabrication Technical FAQ 9. The joint also loses the corrosion resistance you were counting on. For these joints, the filler metal is almost always selected on the stainless side of the equation.
Quick Note: Treat the joint as a stainless steel weld that happens to be tied to a carbon steel component. The filler, gas, cleaning, and post-weld finishing should all follow stainless practice.
The Standard Pairing: 309L and 309LMo
Standards bodies worldwide converge on essentially the same answer for austenitic stainless steel welded to carbon or low-alloy steel:
- Carbon steel to 304 or 304L: use 309L (AWS ER309L / E309L-15/-16) filler.
- Carbon steel to 316 or 316L: use 309LMo (AWS ER309LMo / E309LMo-15/-16), which adds molybdenum to match the 316 side.
These pairings are the ones recommended by the Stainless Steel Pipe and Tube Fabricators’ manual guidance summarized in ASSDA’s Technical FAQ 9 and reflected in structural welding standards such as AS/NZS 1554.6 for dissimilar joints. The naming is not accidental: the “L” denotes low carbon, which reduces carbide precipitation sensitivity in the stainless side of the weld, and the “Mo” grade keeps molybdenum-bearing corrosion resistance from being lost when 316 base metal is diluted into the pool.
Why Over-Alloyed Fillers? The Dilution Problem
Filler 309L contains roughly 23–25% chromium and 12–14% nickel, far more than 304 stainless itself. That surplus is deliberate. When you weld, the arc melts the filler together with both base metals, and the carbon steel side dilutes the weld pool with iron and, critically, carbon. If you started with a filler matching 304 chemistry exactly, that dilution would drag the chromium and nickel content below the level needed to keep the weld fully austenitic. The result: mixed chemistry that can transform to hard martensite under rapid cooling, especially in the first few passes near the carbon steel fusion boundary.
By starting over-alloyed, the 309L weld metal still sits in the austenitic region of the Schaeffler or WRC constitution diagrams even after 30–40% dilution from the carbon steel side. The diagrams are the same tools covered in our earlier discussion of alloy steel welding electrodes and microstructure prediction. Keep the weld metal at roughly 5–10% ferrite for most applications: enough ferrite to suppress hot cracking and tolerate dilution, not so much that the weld becomes brittle or suffers sigma-phase embrittlement at elevated service temperature.
When to Move to Nickel-Base Fillers
Over-alloyed austenitic filler solves most routine dissimilar joints. But there are situations where nickel-based fillers are the better engineering answer:
- High restraint or thick sections: Fillers like ERNiCr-3 (AWS A5.14, commonly known as Inconel 82) or ERNiCrMo-3 (Alloy 625) tolerate even higher dilution while staying tough, and their thermal expansion sits between carbon steel and stainless, reducing residual stress at the transition.
- Thermal cycling service: The British Stainless Steel Association notes that nickel-based fillers are preferred where the joint sees repeated heating, because the closer thermal expansion match lowers the risk of fatigue and distortion-driven cracking BSSA guidance on welding stainless to other steels.
- Corrosive or high-temperature environments: Alloy 625 filler provides resistance to a wide range of corrosive media that a 309 weld metal may not survive, making it a standard choice for refinery and chemical plant dissimilar welds.
The trade-off is cost: nickel alloys typically cost several times more per pound than 309L, so they should be specified where the service justifies it, not used by default.
Matching Consumables to Process and AWS Classifications
Whatever filler you select, buy consumables classified to the relevant AWS specifications: AWS A5.9/A5.9M covers bare stainless welding rods and wires (ER309L, ER309LMo), AWS A5.22/A5.22M covers coated stainless electrodes and strip cladding electrodes (E309L, E309LMo), and AWS A5.14 covers nickel-alloy bare wires such as ERNiCr-3 and ERNiCrMo-3. For stick work, E309L-16 electrodes are the field-friendly equivalent; for GTAW and GMAW, ER309L and ER309LMo wire are the workhorses. If you are stepping up from a different welding discipline, our guides on Alloy 20 welding procedures and all-position welding rods cover the same selection logic for other corrosion-resistant alloys.
Finally, remember that a filler choice alone does not guarantee performance. Dilution control, shielding gas purity, and joint preparation — covered in the next sections — all influence whether the specified filler actually delivers its designed microstructure and corrosion resistance in the finished weld.
Welding Processes, Preparation, and Preheat Requirements
When joining austenitic stainless steel to carbon or low-alloy steel, the process itself is rarely the issue. According to the Australasian Stainless Steel Development Association, TIG, MIG, and stick can all produce reliable dissimilar joints. What determines success is preparation, cleaning, and careful heat management.
Process Selection: TIG, MIG, and Stick
GTAW (TIG) is the go-to for root passes. Its precise heat input allows for tight control over dilution, limiting how much carbon steel melts into the weld pool. The trade-off is speed, so it’s common to run a TIG root followed by faster fill passes.
GMAW (MIG) is efficient for fill and cap passes, especially on thicker sections. Either short-circuit or spray transfer can be used with 100% CO₂ or an Ar/CO₂ mix. Regardless of gas, consumables must be clean and dry. Damp wire introduces hydrogen to the joint, which is most problematic on the carbon steel side. For a refresher on gas selection and cylinder handling, see our guide to Airgas welding supplies.
SMAW (Stick) remains a solid option for field repairs and out-of-position work. A 309L electrode (AWS classification E309L-XX) is the standard for dissimilar joints. Much like the selection process for alloy steel welding electrodes, a plain carbon rod should never be substituted for this over-alloyed filler metal.
Joint Preparation: Two Metals, Two Priorities
Prep must be tailored to each side. On the carbon steel, grind or bevel to remove mill scale, rust, and any surface contamination within roughly 25–50 mm of the joint. Mill scale can trap moisture and create porosity once diluted into the weld pool.
On the stainless steel, only dedicated stainless tools and abrasives should be used. Never use a carbon steel wire brush or a grinder that has previously been used on carbon steel. Embedded iron particles will rust over time and stain the finished weld. Bevel geometry should also account for dilution, with a slightly wider root face on the carbon steel side to help control penetration.
Remove Zinc Coatings Before Welding
If the carbon steel is galvanized or zinc-coated, the coating must be stripped from the weld zone before striking an arc. Zinc melts and boils at a much lower temperature than steel, which can cause zinc embrittlement, porosity, and a sharp drop in corrosion resistance. Both the British Stainless Steel Association and ASSDA flag this as a common mistake. Strip the coating by grinding, abrasive blasting, or chemical stripping well back from the joint. Adequate fume extraction is also mandatory, as zinc oxide fume poses a genuine health hazard.
Quick Note: If the stainless side has a protective plastic film, remove it and clean any residue with a dedicated welding surface cleaner. Burned-on film is a source of carbon contamination that can pit the stainless steel later.
Preheat: When It Helps and When It Hurts
Preheat isn’t required for every dissimilar joint. Based on ASSDA and BSSA guidance, the following is a practical approach:
- Carbon steel below ~0.20% carbon: Preheat is usually not required, even when joined to austenitic stainless.
- Carbon steel above ~0.20% carbon, sections over ~30 mm, or heavily restrained joints: A modest preheat of around 150°C is typically specified.
- Austenitic stainless steel: No preheat is needed for metallurgical reasons. Keep interpass temperatures below ~150°C to avoid sensitization and loss of corrosion resistance.
The purpose of preheat is to slow the cooling rate of the carbon steel heat-affected zone to prevent hydrogen-induced cold cracking. Rapid cooling can form hard martensite, and with hydrogen and restraint present, cracks can nucleate. For heavy restraint, buttering the carbon steel edge with 309L filler before completing the joint can be used to limit this risk. Lastly, store low-hydrogen consumables in holding ovens and keep flux-cored wires sealed, as dry consumables are just as vital as controlled heat input.
5) Thermal Expansion, Distortion, and Restraint Management
Two metals, two rates of movement. Tie a stainless steel nozzle into a carbon steel header and each side of the joint expands and contracts on its own schedule, while the weld metal sits somewhere in between. That mismatch is often a bigger day-to-day problem than the metallurgy itself. It shows up as a plate that won’t stay flat, a pipe that pulls out of alignment, or a root pass that cracks because the joint was boxed in while it cooled.
Why Stainless and Carbon Steel Move Differently
The numbers are worth memorizing. According to the Austenitic Stainless Steel Fabricators Association technical FAQ, mild steel expands at roughly 12 µm/m/°C, while austenitic stainless steel expands at about 17 µm/m/°C over the 0–300°C range. Stainless moves nearly 40% further for the same temperature change. On a long seam or a large gusset, that difference accumulates into visible bowing and twisting.
Thermal conductivity adds a second layer of trouble. Carbon steel pulls heat away from the arc faster than austenitic stainless, so the stainless side holds heat locally while the carbon side sheds it. The two sides therefore reach peak temperature, cool, and shrink at different moments. The ASSDA guidance on welding dissimilar metals and the BSSA article on joining stainless to other steels reach the same conclusion: plan for asymmetry rather than fighting it after the fact.
Where the Stress Actually Ends Up
Distortion is not the only consequence. A restrained dissimilar joint locks residual stress into the carbon steel heat-affected zone, which is exactly where hydrogen cracking likes to start. Restraint raises the effective hardness and the required preheat on the carbon side, so a joint free to move needs fewer precautions than one boxed in by heavy fixtures or a fully built-up structure. If your connection details involve welded alloy steel components, our Alloy Steel Welding Electrodes: Selection, Use & Safety guide covers preheat and hydrogen control in more depth.
Practical Distortion Control
- Fit-up first. Hold root gap and land within tolerance, keep angular misalignment tight, and never pull parts into alignment with tacks. A tack that strains the joint to close a gap stores elastic energy that releases later as distortion.
- Tack symmetrically. Match tacks on opposite sides at regular intervals, placing them where they counteract rather than add to the anticipated pull.
- Sequence away from the problem. Back-step, skip, or stagger beads so no single continuous pass runs the full length, and alternate sides of a butt joint rather than finishing one side first.
- Balance weld volume. Keep deposited metal roughly symmetric about the joint centerline. Where geometry makes that impossible, compensate with sequencing or a buttering layer.
- Control heat input. Smaller beads, faster travel, and a strict interpass temperature keep total heat low. Many fabricators cap austenitic stainless interpass around 150°C (300°F) unless the procedure specification says otherwise, because excess heat also degrades corrosion resistance.
Quick Note: Buttering outline — deposit a sound layer of 309L (or 309LMo) on the carbon steel edge first, inspect it, then complete the joint between that layer and the stainless. The first pass now dilutes into carbon steel only, later passes dilute into an already stainless-compatible layer, and final weld chemistry is easier to predict. Buttering also gives thick sections a graded transition that spreads contraction stress.
Buttering Thick Sections on the Carbon Steel Side
For thick carbon steel butted against stainless, industry practice is to butter the carbon steel side before making the joint. The buttering layer does three jobs: it controls dilution at the fusion boundary, it lets the welder temper-bead or peen lightly in a controlled sequence, and it allows the joint to contract in stages rather than all at once. Do it while the carbon steel can still be preheated, keep the layer thin and well fused, and don’t grind it back so aggressively that you reintroduce a dilution-heavy first pass.
Heat Input Balances Two Risks at Once
Heat input is the dial that governs both problems. Too much and you get wide heat-affected zones, heavy distortion, excessive dilution, and sensitization on the stainless. Too little and the carbon side may lack fluidity for good fusion while hydrogen cracking risk climbs under restraint. Aim for the middle of the procedure’s qualified range, watch interpass temperature with a thermometer rather than by eye, and note that Aluminum MIG Welding: The Complete Practical Guide applies the same back-step and balanced-bead habits to a far more thermally sensitive material.
| Situation | Move to make |
|---|---|
| Long butt seam, thin stainless | Back-step or skip sequence, low heat input, frequent tacks |
| Heavy, heavily restrained joint | Butter the carbon side, preheat if required, low-hydrogen dry consumables |
| Pipe tie-in out of round | Tack to a continuous line first, weld in quarters, check alignment between segments |
| Angular distortion after welding | Correct mechanically while warm; avoid forceful cold straightening across the weld |
In practice: treat distortion as a fit-up and sequencing problem, treat restraint as a cracking problem, and use buttering plus disciplined heat input to solve both on thick dissimilar joints.
6) Corrosion: Galvanic Effects, Bimetallic Corrosion, and Cleanup
A dissimilar joint can be structurally perfect and still fail in service, and the failure usually starts on the surface. When stainless steel and carbon steel sit together in a wet line, a drain, a coastal plant, or an outdoor yard, they form a galvanic cell. Understanding that cell, and what you can realistically do about it, is the difference between a transition that lasts twenty years and one that pits through in two.
How a Galvanic Cell Forms at a CS-SS Transition
Galvanic corrosion needs three things: two metals with different electrochemical potentials, metallic contact between them, and an electrolyte such as water, condensate, salt spray, or soil moisture. Satisfy all three and the pair behaves like a battery. Current flows through the electrolyte from the more anodic metal to the more cathodic one, and the anode dissolves. As a dissimilar metal welding guide from HXSCO explains, dissimilar joints can corrode faster than either metal alone when no protection is applied.
In a stainless-to-carbon steel assembly, carbon steel is almost always the anode. The practical twist is the area ratio: a large stainless surface acting as the cathode driving corrosion on a small exposed carbon steel area is the worst case, because the corrosion current concentrates on a tiny anode. That is exactly what happens at a weld where the stainless is left bare and the carbon steel coating was burned back a few inches during welding.
Quick Note: The galvanic risk is rarely at the weld bead itself. It starts in the heat-affected zone and the exposed base metal beside it, where paint, mill scale, or galvanizing was destroyed by arc spatter and heat. Treat that halo of bare metal as the zone that needs protection first.
Protection Options for Wet Service
There is no single fix. The right choice depends on whether the joint is intermittently damp, continuously submerged, buried, or insulated inside a piping system.
- Barrier coatings: A compatible coating system over both metals keeps the electrolyte out of the circuit. Overlap well past the heat-affected zone on both sides.
- Stripe coating: An extra-rich coat applied directly over welds, edges, and blast-cleaned surfaces, where film thickness naturally falls short. Standard detail on pipework and tanks.
- Insulation and dielectric fittings: Non-conductive gaskets, sleeves, or unions break the electrical path between the two metals.
- Cathodic protection: Sacrificial anodes or impressed current for submerged or buried transitions, typically combined with coating.
- Design change: Where possible, keep the transition out of the wetted zone, or use a full-lined or clad spool instead of bare carbon steel.
Marine work makes the point vividly. An aluminum hull fitting bolted to steel hardware corrodes with remarkable speed for the same reason an SS/CS joint does, which is why aluminum boat welding procedures spend so much time on isolation and dissimilar-metal detailing rather than on the bead itself.
Why Pickling and Passivation Don’t Translate
On stainless-to-stainless fabrication, weld heat tint and contamination are removed with an acid pickle followed by passivation, restoring the chromium oxide layer. On stainless-to-carbon steel, that routine generally falls apart. As the Stainless Steel Industry of Australia technical FAQ on welding dissimilar metals notes, acid pickling and passivation that work well for stainless-to-stainless joints are often not possible for carbon steel to stainless assemblies because of the effect of the acid on the carbon steel.
Practically, you cannot dunk the joint. Carbon steel dissolves rapidly in the nitric-hydrofluoric baths used for pickling, leaving a roughened, carbon-rich surface that defeats the purpose. Work mechanically and locally instead:
- Remove heat tint from the stainless side with stainless-only wire brushes, discs used exclusively on stainless, or pickling paste brushed strictly onto stainless areas.
- Follow with a citric- or nitric-based passivating gel on the stainless surface only, observing the manufacturer’s time and temperature limits and the applicable AWS passivation standard.
- Neutralize and remove all residues from the carbon steel side, then dry quickly to prevent flash rusting before coating.
- Inspect for residual blue heat tint and iron contamination with a ferroxyl or copper sulfate test where service is critical.
Where zinc galvanizing survives near the joint, the same ASSDA guidance warns that a stripe coat or barrier paint may be needed to avoid galvanic dissolution of the zinc itself. Zinc is anodic to steel, so a bare steel weld next to intact galvanizing sacrifices the coating preferentially, leaving white rust and then unprotected steel. Either strip the zinc back far enough to coat sound metal, or seal the boundary with a high-build stripe coat.
Surface preparation matters as much as the weld before final coating. The same discipline of cleaning, deoxidizing, and keeping one metal’s residue off the other shows up across processes, and the preparation logic in aluminum cleaner for welding applies in spirit here: contaminants left behind become the corrosion site later.
Note: If the joint will be pickled, passivated, or left coating-free in an aggressive environment, specify the corrosion protection in the procedure, not after the fact. Post-weld corrosion control that is not written into the WPS is almost never done on site.
Field Checks Before Handover
- Is the electrolyte path broken, coated, or cathodically protected?
- Did the coating extend at least 100 mm past the heat-affected zone on the carbon steel side?
- Was the stripe coat applied over the weld crown, edges, and any pinholes?
- Is the zinc boundary either removed or sealed?
- Does the operations manual tell the owner how to inspect and recoat this transition during maintenance?
Handle corrosion at the design and coating stage and the joint will outlive the piping around it. Leave it until after hydrotest and the transition becomes the first place the inspector points at.
7) Special Cases: Duplex Stainless to Carbon Steel and Other Combinations
Most of this article covers austenitic grades like 304 and 316, which account for the overwhelming majority of stainless-to-carbon joints. Plenty of real projects push past that: a duplex 2205 pump casing tied into a carbon steel skid, a super duplex fitting spliced onto an A36 header, or an Alloy 20 vessel nozzle welded to a mild steel shell. The core principle stays the same — always weld to the stainless side with an over-alloyed filler — but these combinations add a variable the austenitics don’t have: a carefully balanced two-phase microstructure that a single pass of wrong heat input can wreck.
Why Duplex Joints Behave Differently
Duplex stainless steels are named for their roughly 50/50 mix of austenite and ferrite. That balance produces their high strength and excellent stress corrosion cracking resistance, and it is not accidental — it is engineered through chemistry and then governed by thermal history. When you join duplex to carbon steel, the weld pool ranges from fully diluted carbon steel at one fusion boundary to nearly undiluted duplex at the other. In between, composition sweeps through a range where the local microstructure can shift dramatically depending on cooling rate and heat input.
The practical consequence is that filler selection alone is not enough. A filler that behaves beautifully at one heat input can produce an unacceptable microstructure at another.
Filler Choices: ER2209 vs. Alloy 625
Two fillers dominate dissimilar duplex-to-carbon joints:
- ER2209 — a 22Cr-9Ni-3Mo duplex filler matched to 2205 base metal. It keeps the weld metal duplex instead of converting it fully austenitic, which preserves strength and holds corrosion performance closer to the base metal. The AWS A5.9 classification governs its chemistry, and it is the default when the joint must perform as duplex.
- Alloy 625 (ERNiCrMo-3) — a nickel-based filler with substantial niobium and molybdenum. Nickel fillers tolerate far higher dilution than stainless fillers without forming hard martensite, so they are frequently specified where metallurgical compatibility matters more than matching base-metal strength. It is the same logic behind Alloy 20 welding procedures leaning on high-nickel fillers when joining to plain carbon steel.
GTAW is the preferred process for research-grade and production root passes here because it gives precise control over the two variables that decide the outcome: heat input and dilution.
The Hard Martensitic Layer at the A36 Fusion Boundary
Barnhouse and Lippold’s work at OSTI on joining duplex 2205 to A36 carbon steel with GTAW exposed the key hazard directly. In the ER2209 welds, they observed a narrow martensitic region with high hardness along the A36-side fusion boundary — exactly where the weld metal picks up enough iron and carbon from the mild steel to push that thin band into a hard, brittle structure. It is the same martensite mechanism discussed earlier for austenitic fillers, just compressed into a narrower band by the duplex filler’s higher alloy content.
Critically, the same study found that the hard martensitic layer was not observed with Alloy 625 filler. The high nickel content keeps that interfacial region austenitic across a much wider dilution range, which is why 625 is often the specified answer where hydrogen cracking or brittle interfacial failure is a concern.
Quick Note: If your code or service environment treats the carbon steel side as structurally critical, it is worth specifying a 625 butter layer and having it covered by a qualified WPS. Nickel filler is expensive; a cracked interface in service costs far more.
Heat Input, Ferrite, and Chromium Nitrides
Barnhouse and Lippold also found that corrosion resistance with ER2209 improved at higher heat input. Slow cooling gives nitrogen time to diffuse into austenite, producing more austenite and reducing chromium nitride precipitation that robs the matrix of chromium and creates corrosion initiation sites. Too little heat input shifts the balance toward excess ferrite and invites nitrides at the grain boundaries.
The catch: higher heat input means a wider heat-affected zone and more distortion on the carbon steel side. It is a balance, not a target to chase blindly. The BSSA guidance on welding stainless to other steels reinforces the point — dissimilar joints demand attention to heat input and filler chemistry together, not one or the other.
The 625 Root Pass Warning
Here is the counterintuitive finding that trips people up. In the same testing, the 625 root pass showed severe corrosion attack in certain environments, even though 625 won on the metallurgical front at the fusion boundary. Galvanic and compositional gradients at a thin root pass can concentrate attack in a way thicker, multi-pass sections avoid. The lesson is not that 625 is bad — it is that the filler solving the cracking problem can create a corrosion problem, and both must be judged against the actual service environment rather than assumed.
For context on how environment-specific these decisions are, the Australian Stainless Steel Development Association’s technical FAQ and the CWB weldability articles on differing stainless grades both stress that no universal dissimilar-metal procedure exists — service medium, temperature, and inspection access all change the answer.
Practical Selection Guide
| Situation | Preferred Filler | Why |
|---|---|---|
| 2205 to carbon steel, general service | ER2209 | Preserves duplex weld metal; adequate corrosion performance |
| 2205 to carbon steel, high restraint or cracking risk | Alloy 625 (ERNiCrMo-3) | No hard martensitic layer at fusion boundary |
| Highly corrosive or uncertain environment | 625, with service-specific review | Dilution tolerance; verify against actual medium |
| Super duplex (e.g. 2507) to carbon steel | Matching super duplex filler or 625 | Match PREN; same dilution logic applies |
The overarching rule still holds: treat the joint on the stainless side, verify the WPS against the actual base metal pair, and never assume a filler that worked on your last mild steel job belongs anywhere near duplex. If your team works regularly across metal families, building the same discipline around alloy steel welding electrode selection and preheat as you apply to stainless fillers will prevent most of the failures seen in the field.
8) Common Mistakes, Troubleshooting, and Practical Checklist
Most failures in stainless-to-carbon steel joints are not exotic metallurgical mysteries. They are ordinary oversights, repeated under time pressure, that show up weeks later as a cracked root, a rusting heat tint, or a joint that fails inspection before the system ever goes into service. This section collects the errors welders and supervisors run into most often, what they look like on the bench, and what to do about them.
The Eight Most Frequent Errors
- Welding with a carbon steel filler. This is the classic mistake. The CWB’s guidance on weldability is blunt: using a carbon steel electrode on stainless produces a dilution-sensitive, crack-sensitive martensitic weld metal. Carbon from the steel side migrates into the pool, hardness climbs, and hydrogen does the rest. The fix is always the same: an over-alloyed austenitic filler such as 309L or 309LMo. If you are unsure which stick or wire is which, a refresher on alloy steel welding electrodes saves a lot of rework.
- Carrying over the wrong shielding gas. The gas mix tuned for carbon steel MIG is not the gas for this joint. The Australasian Stainless Steel Development Association is explicit that shielding gas must not include the oxygen often found in carbon steel mixtures, since oxygen pickup roughens the bead and degrades corrosion resistance. Note also the opposite trap: adding nitrogen to boost productivity can reduce weld metal ferrite, and the same ASSDA technical FAQ warns that this loss of ferrite invites hot cracking. Pure argon for GTAW and a clean argon-CO2 or tri-mix verified as oxygen-free for GMAW is the conservative starting point. Suppliers such as Airgas welding supplies publish mix data sheets that make verification easy.
- Inadequate cleaning, especially zinc. Galvanized carbon steel must have the coating mechanically removed a good 10–25 mm back from the joint before any arc is struck. Welding over zinc produces porosity, unstable arcs, and toxic fume. Grind, do not just wipe.
- Ignoring restraint and preheat. A heavily restrained joint on a thicker or higher-carbon steel needs preheat and, sometimes, a buttering layer. Skipping it on a “small repair” invites hydrogen cracking on the carbon steel side days later.
- Over-penetration and weld pooling on thin stainless. Stainless conducts heat poorly and melts readily. Excessive current on thin sheet causes burn-through and a slumped, oxidized bead. Back off the amperage, travel faster, and let the joint geometry do the work.
- Inadequate ferrite control. Too little ferrite risks hot cracking; too much, especially in the presence of nitrogen or strain, can embrittle the weld on cryogenic or highly restrained service. Aim for the ferrite range specified by the applicable WPS.
- Neglecting galvanic protection. The joint may be sound while the adjacent carbon steel corrodes in wet service. Coating, insulating, or cathodic protection must be designed in, not added later.
- Wrong process choice for the position. Vertical-up GMAW on a root that should have been TIG-sealed leads to lack of fusion. If you need a refresher on rod behavior off-flat, see this all position welding rod reference.
Quick Note: If the bead cracks while it is still cooling, suspect filler selection or hydrogen. If it cracks weeks later in service, suspect galvanic corrosion or sensitization. The timing of the failure tells you where to look first.
Troubleshooting at a Glance
| Symptom | Likely cause | Corrective action | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Cracking in or beside the weld | Carbon steel filler, high dilution, no preheat on restrained joint | Switch to 309L/309LMo, reduce dilution with buttering, preheat per WPS | ||||||||||||||||
| Porosity, spatter, blue-grey bead | Zinc or oil contamination, oxygen-bearing gas | Strip coating, clean 25 mm either side, verify gas purity | ||||||||||||||||
| Hot cracking / wormhole fissures | Ferrite too low from nitrogen pickup or excessive dilution | Re-check gas, control travel and current, confirm ferrite by WRC chart | ||||||||||||||||
| Rust bloom on the stainless side | Free iron, heat tint, no passivation | Remove tint by pickling or abrasive method, passivate | ||||||||||||||||
9) Practical Tips, Tables, and Decision GuideEverything covered so far meets at the workbench. This section consolidates the guidance into a filler decision matrix, a buttering outline, and a set of shop-floor checks you can run before striking an arc on a stainless-to-carbon joint. Filler Metal Decision MatrixThe fastest way to avoid a bad joint is to start with the right spool or rod. The matrix below reflects standard practice for joining austenitic stainless steel to carbon or low-alloy steel, with the over-alloyed 309 family as the default.
Filler classifications come from the AWS A5.9 and AWS A5.22 family of specifications, and the designator tells you the service intent: the “309” prefix means the deposit is engineered to tolerate carbon steel in the puddle. Heat Input and Dilution ControlDilution is the variable you control most directly. Too hot, and the carbon steel side melts deeply, dragging carbon into the pool and pushing the microstructure toward hard martensite; too cold, and you risk lack of fusion, especially on the stainless side. The Stainless Steel Industry of Australia technical FAQ on dissimilar welding recommends moderate heat input with 309L precisely so dilution stays in the range the filler chemistry can absorb. Practical levers:
Buttering: The Outline for Thick SectionsWhen the carbon steel is thick, restrained, or prone to hardening, deposit a corrosion-resistant layer on the carbon steel face first, let it cool, clean it, then make the final joint into the stainless. A typical sequence:
Quick Note: Buttering adds time, but it moves the risky metallurgy out of the final weld. On heavy, restrained assemblies it is usually cheaper than a repair after a crack appears. The approach is well documented in HXSCO’s dissimilar metal welding guide, which notes the same technique for managing hardness on the carbon steel side. Distortion Control on the FloorStainless expands roughly one and a half times as much as carbon steel for the same temperature rise, so asymmetric joints pull. Tack generously in a balanced pattern, weld in short staggered sequences rather than one long bead, and alternate sides where the joint allows. Tack with the same filler you will use for the finished weld. If one side is much thicker than the other, see Alloy Steel Welding Electrodes: Selection, Use & Safety for the carbon steel preheating and electrode handling side of the job, since preheat changes distortion behavior as much as it changes hardness. Corrosion Protection ChoicesDecide the protection scheme before welding, not after. In wet or buried service, plan for coating, electrical insulation at the transition, or cathodic protection, because the galvanic couple between stainless and carbon steel persists once an electrolyte is present. Where the fabrication must also be pickled, remember that acid treatments acceptable for stainless will attack the carbon steel side, so the British Stainless Steel Association advises barrier or stripe coating over the completed transition instead of a full pickle. Inspection Checklist
Rules of Thumb
10) ConclusionA maintenance crew can tie a new stainless pipe into an old carbon steel line and walk away confident, but only if the joint was planned as a dissimilar metal weld from the start. Nearly every failure mode covered in this article, from a hairline crack in the heat-affected zone to rust bleeding from a flange weeks after start-up, traces back to treating the joint as ordinary carbon steel welding with a stainless-looking bead. The base metals are compatible. The procedures usually are not. The single most durable rule is also the simplest: select the filler as if the whole joint were stainless. The British Stainless Steel Association’s guidance on welding stainless steels to other steels puts it plainly, and the practice follows the same logic as the AWS filler classifications. You do not dilute the joint down toward carbon steel; you build it up away from it. Carbon steel filler on stainless is the classic error, because the weld metal loses its corrosion resistance and picks up a hard, hydrogen-sensitive structure on the fusion boundary. The Takeaways Worth Carrying to the Bench
Quick Note: Before the arc is struck, confirm the filler classification against a published standard, write the procedure down, and decide in advance how the finished joint will be cleaned, passivated or coated. A qualified weld procedure and a documented corrosion plan will save far more money than any repair later. Ferrite control, dilution limits and preheat decisions are easier to argue about in the abstract than to apply correctly under time pressure, which is why verified procedures and recognized references matter more than shop folklore. Lean on published guidance such as the AWS Procedure Handbook of Arc Welding when documenting a new joint, and treat…treat every stainless-to-carbon tie-in as a stainless weld that happens to be joined to carbon steel, not a carbon steel weld with stainless cladding. Get that inversion right and the rest of this article falls into place as a checklist rather than a rescue mission. Confirm the filler class against a published specification, verify dilution and ferrite expectations with a diagram instead of an assumption, decide preheat and sequencing before fit-up, and settle how the finished joint will be cleaned and protected in its actual service environment. The dissimilar joint that survives is the one someone finished designing before the first tack was struck. Frequently Asked QuestionsCan you weld stainless steel directly to carbon steel with carbon steel filler?No. Welding carbon steel directly to austenitic stainless with carbon steel filler can produce a weld metal prone to martensite formation, resulting in low ductility and high crack risk. Use an over-alloyed filler like 309L or 309LMo instead. What filler metal should I use for welding 304 stainless to mild steel?For 304(L) to carbon steel, the standard recommendation is ER309L. For 316(L) to carbon steel, use ER309LMo to account for molybdenum requirements and maintain corrosion resistance. Do I need preheat when welding stainless to carbon steel?Usually not for carbon steels below ~0.20% carbon at ambient temperature. Preheat of around 150°C may be needed if carbon content is above 0.20%, thickness exceeds ~30mm, or there is high restraint. Always dry consumables due to hydrogen cracking risk on the carbon steel side. Will a stainless to carbon steel weld corrode faster?It can in wet/electrolytic environments due to galvanic corrosion between dissimilar metals. Proper design (coatings, insulation, or cathodic protection) and avoiding large exposed carbon steel areas near the joint can mitigate this risk. Should I remove galvanizing before welding stainless to galvanized steel?Yes. Remove zinc coating around the joint before welding. Molten zinc can cause embrittlement and reduced corrosion resistance in the weld. Is pickling/passivation possible after welding SS to carbon steel?Generally no, or only in very limited cases. Acid pickling/passivation used for stainless-stainless fabrications is corrosive to carbon steel, so CS-SS assemblies typically cannot be fully pickled/passivated; plan corrosion protection differently. When should I use nickel-based filler like Inconel instead of 309L?Consider nickel-based fillers (e.g., ERNiCr-3/Inconel 625) when lower thermal expansion mismatch is critical, for certain corrosive environments, or for some duplex/complex dissimilar combinations where 309L may not provide adequate properties. Follow project specifications. How can I reduce distortion when welding stainless to carbon steel?Control heat input, use proper fit-up and sequencing, tack weld strategically, consider balanced welding, and for thick sections use buttering on the carbon steel side. The mismatch in expansion/conductivity makes distortion management important.
NH
Nicholas HarperEditorial TeamCovering welding processes, equipment selection, and shop-floor practice for WeldZen — with every claim checked against the governing code. Related Studies
Technical Dispatch
Get The Next Benchmark FirstNew field studies and calibration sheets, once a week. No spam. |