Oxy-Acetylene Gas Welding Stainless Steel: What Works

A fabricator digs an old torch set out of the back of the shop, grabs a length of 304 sheet, and asks the question every welding forum has argued about for twenty years: can you oxy-acetylene weld stainless steel, or is it a one-way ticket to a rusted, cracking joint? The honest answer is that you can fuse the metal — but whether the joint survives the next rainy season depends on chemistry most torch operators never think about.

Oxy-Acetylene Gas Welding Stainless Steel: The Short, Honest Answer

Every shop has one of those torch sets. The hoses are coiled on a shelf behind the chop saw, the cylinders have been chained to the wall since the last owner retired, and nobody has touched the regulators in years. Then somebody needs a bracket, a patch panel, or a repair on stainless sheet, and the old question comes back: can you actually weld stainless with this thing, or not?

You can. An oxy-acetylene torch will melt 304, fuse it to another piece of 304, and produce a joint that holds. What you pay for that convenience is a compromised heat-affected zone, a dirty surface, and a corrosion profile no inspector would sign off on. It is a real process that works, and it is also a niche, largely non-code one. Both things are true at once, which is why the forum arguments never end.

Technically this is oxyfuel gas welding (OFW), the term used in AWS A3.0 terminology. A flame melts the base metal and the filler together with no electric arc involved. The AWS C4.7/C4.7M Recommended Practices for Oxyacetylene Welding of Steel covers the process in detail, and it is worth knowing the document exists even if your job never touches it.

What Separates It From Brazing and From Arc Welding

People blur these together, so let’s draw the lines.

• Brazing melts the filler but not the base metal. The filler wets the surface and bonds without fusion, at far lower temperatures. Different animal entirely.
• Oxy-acetylene welding melts both. The puddle is the base metal itself, with filler added to build it up or bridge a gap.
• Arc processes (TIG, MIG, stick) use electrical energy and, for TIG and MIG, a shielding gas that keeps oxygen off the pool. Torch welding has none. The flame is the atmosphere, and that is the root of most of its stainless limitations.

If you want a plain-language rundown of small torch work in general, our guide to air-acetylene welding and brazing basics covers the equipment side, though air-acetylene runs much cooler and cannot fuse stainless.

Why Stainless Behaves Nothing Like Mild Steel Under a Flame

A neutral oxy-acetylene flame burns somewhere around 5,800 to 6,300 °F (roughly 3,200 to 3,480 °C), hotter than any other common fuel-gas flame. Austenitic stainless like 304 melts at about 2,550 to 2,650 °F. That looks like a comfortable gap on paper, and in practice it feels enormous right up until it doesn’t: stainless gives you very little warning before it drops through.

Three properties make it harder than mild steel:

1. Low thermal conductivity. Heat stays where you put it instead of spreading, so the puddle runs hot while surrounding metal lags behind.
2. High thermal expansion. Cooler metal around the weld moves as it heats, and it does not always move back. Distortion follows.
3. A chromium oxide skin and a sensitization window. The passive film protects the steel, but holding the metal in a certain temperature band strips chromium from the grain boundaries and leaves the joint unable to heal itself.

That third point is the one nobody mentions at the counter when they hand you a flux-coated rod. It gets a full treatment later, because it is the single biggest reason torch-welded stainless fails a year after it looked fine.

Who This Guide Is For, and What It Will Cover

This is written for hobbyists, repair techs, restorers, farm and ranch shops, and anyone weighing whether to buy a TIG setup or just make the torch work. It is not written for code welders, and it won’t pretend the process is something it isn’t.

Over the next sections we will walk through flame chemistry and heat flow, sensitization and carbon pickup, a green-light/yellow-light/red-light decision framework, equipment, filler and flux selection, a step-by-step procedure, troubleshooting, pickling and passivation, and an honest comparison against TIG, MIG, and stick.

By the time you finish, you should be able to look at a stainless job, decide in under a minute whether the torch is acceptable, and if it is, run the weld without wrecking the corrosion resistance any more than necessary. That is a modest promise, and it is the honest one.

What Actually Happens at the Torch: Flame Chemistry and Heat Flow in Stainless

A torch flame looks simple: gas in, fire out, metal turns liquid. On stainless, that flame does two jobs at once. It is a chemical tool that can add or remove carbon at the weld pool, and a heat-delivery system working against a metal that will not spread heat evenly. Get both right and the joint wets out cleanly; get either wrong and you create something that looks welded but corrodes from the inside.

If you are used to an air-acetylene torch for plumbing or HVAC brazing, the jump to an oxy-acetylene stainless weld is bigger than it looks. The principles overlap, but heat intensity and flame chemistry are different animals, as covered in this guide to air acetylene welding torch basics, brazing, and safety.

The Three Parts of an Oxy-Acetylene Flame

A properly adjusted oxy-acetylene flame has three visible zones. At the tip sits the inner cone, a sharp blue-white point where acetylene and oxygen first combine and burn hottest. Around it is the feather, a lighter blue envelope whose length shows how fuel-rich the mixture is. Beyond that lies the outer envelope, where combustion finishes using oxygen from the surrounding air.

Welders classify the flame by how those zones look:

• Neutral flame: oxygen and acetylene are balanced. The inner cone is crisp and rounded, with almost no feather. This is the default for most steel welding.
• Oxidizing flame: excess oxygen. The inner cone shortens and sharpens, and the flame may hiss. It burns hotter and leaves a strongly oxidizing pool surface.
• Carburizing (reducing) flame: excess acetylene. A distinct feather extends beyond the inner cone, sometimes with a smoky edge, introducing free carbon into the flame.

For stainless, most experienced guidance calls for a neutral or very slightly oxidizing flame. The slight oxygen excess helps burn off surface carbon and keeps the molten pool active rather than sluggish. Push it too far and the flame forms heavy chromium oxide, overheats the metal, and makes the pool hard to control. Flux can dissolve oxide, but it cannot rescue a flame that is visibly short and screaming.

Excess acetylene is worse. A carburizing flame deposits free carbon and can drive it into the weld pool. In austenitic stainless, that carbon is exactly what you do not want near a heat-affected zone, because it encourages chromium carbide formation and leaves the steel less able to maintain its corrosion-resistant surface film. The flame may make the puddle feel easier to move, but it trades long-term joint integrity for short-term convenience.

Why Stainless Puddles Feel Stiff and Sticky

Molten stainless does not flow like mild steel. Part of that is chemistry: the chromium oxide film melts at a much higher temperature than the base metal, so it sits on the pool like a skin until flux dissolves it. Part of it is heat flow. Austenitic stainless has a thermal conductivity of roughly 16 W/m-K, about one-third that of carbon steel. Heat stays close to the torch instead of spreading, producing a narrow, bright puddle ringed by relatively cold metal.

The practical effect: stainless rewards a tight, focused flame and quick, controlled movement. A big soft flame heats a wide area, pulls in more distortion, and gives oxide more time to build.

Heat Flow, Expansion, and the Distortion Problem

Austenitic stainless expands about 50 percent more than carbon steel for the same temperature rise. Combine that with low conductivity and you get a difficult pair: the torch creates a hot, expanding patch while the surrounding metal stays comparatively cool and rigid. The hot zone wants to grow, the cold zone resists, and something has to give. Thin sheet buckles or oil-cans, a butt joint closes up as you weld, and thick sections build shrinkage stress that can crack a weld after it cools.

On thin sheet, preheat is rarely needed and often counterproductive. It widens the heat-affected zone, increases time spent in the sensitizing temperature range, and makes buckling harder to control. Clamping, tacking, and back-stepping do more good than a rosebud tip.

Thicker stainless calls for patience rather than pressure. Cranking the regulators and running a huge flame feels productive, but it overheats the surface before the root reaches temperature, widens the HAZ, and raises the odds of burn-through or cracking. Better: a slightly larger tip at sensible pressures, gentle preheat only where the part is genuinely massive, and enough time for heat to soak in. The flame should sound soft and controlled, not like a jet engine.

Treat the flame as a precision heat source, not a blunt instrument. For a broader look at accepted oxy-fuel technique, AWS publishes Recommended Practices for Oxyacetylene Welding of Steel, a useful baseline for torch handling even though it is not a stainless-specific code.

Get the flame chemistry right and the heat input under control, and stainless welds predictably. Get either wrong and the torch produces a joint that looks acceptable until moisture, chlorides, or time expose what happened beneath the surface.

The Metallurgy Problem Nobody Warns You About: Sensitization and Carbon Pickup

Two stainless parts can leave the bench with beads that look almost identical. One lasts for years outdoors. The other develops brown weep lines along the weld within a few months. The difference is often not the operator’s hands. It is what happened to the metal just outside the puddle, in a zone you cannot see and probably cannot test in a home shop.

That zone is where oxy-acetylene welding of stainless gets complicated. The flame can produce a sound-looking joint and still leave the heat-affected area vulnerable to intergranular corrosion. Understanding why is the difference between a repair that holds and one that quietly fails.

What Sensitization Actually Means

Stainless steel does not resist rust because it contains no iron. It resists rust because chromium reacts with oxygen to form a thin, self-repairing passive film. For that film to remain stable, the alloy generally needs roughly 12 percent chromium available in the metal beneath it.

Heat changes the arrangement of that chromium. When austenitic stainless is held between about 800 and 1,500 degrees Fahrenheit (425 to 815 degrees Celsius), carbon atoms become mobile. They migrate toward grain boundaries and combine with chromium to form chromium carbides, primarily the M23C6 type. Those carbides sit at the boundaries like a chain of tiny precipitates. The chromium they consume comes from the immediately adjacent metal.

The result is chromium depletion. The matrix next to each grain boundary can fall well below the roughly 12 percent needed to rebuild a stable passive film. The alloy has not lost its corrosion resistance everywhere, but it has lost it in a microscopic network that runs through the weld heat-affected zone. This is the chromium-depletion theory, described by Bain, Aborn and Rutherford in 1933 and still the accepted explanation today. Rolled Alloys gives a clear practical summary of the mechanism.

The temperature window is the key. Below it, carbon diffusion is too slow to matter much. Above it, carbides can dissolve again. Sitting inside it, especially for minutes rather than seconds, is what creates the damage. The NRC Regulatory Guide 1.44 uses that same 800 to 1,500 degrees Fahrenheit range when discussing sensitization of austenitic stainless steels.

Why a Torch Is the Worst Thermal Profile

Arc welding is fast and concentrated. The heat-affected zone reaches high temperature and then cools quickly, so the metal spends comparatively little time in the sensitizing band. A gas torch is the opposite: the flame is broad, the heat is diffuse, and the weld is made at a slower pace. The surrounding metal can sit in the critical range during the entire pass and then cool gradually after the torch leaves.

Slow cooling is not a minor detail. Carbide precipitation is time-dependent. A long soak gives carbon more opportunity to migrate and gives the carbides more time to grow. That is why a torch weld on stainless can be more metallurgically aggressive than its mild appearance suggests. A neutral flame does not solve this. The thermal cycle itself is the problem.

Thick sections make it worse. More metal around the joint acts as a heat sink, encouraging the operator to linger, add more heat, and hold the zone in the dangerous range even longer. Thin sheet is more forgiving, but only relatively. The difference between “less bad” and “safe” is not the same thing.

The Carbon Pickup Question

Flame chemistry adds another variable. An acetylene-rich, carburizing flame can deposit free carbon on the work and introduce carbon into the weld pool. Since carbon is the element that forms the damaging carbides, adding more of it is exactly the wrong direction. A heavily carburizing flame may also produce soot and a dirty weld that is harder to inspect.

Flame combustion products interact with the molten pool as well. Unlike TIG, oxy-fuel welding does not surround the puddle with an inert gas blanket. The flame itself performs the shielding, and its chemistry matters. A neutral or very slightly oxidizing flame is generally preferred for stainless because it avoids adding carbon and helps keep the pool clean and fluid. That said, carbon pickup from the flame is usually secondary to the thermal cycle. The dominant risk is still prolonged time in the sensitizing range.

Shop contamination belongs in the same conversation. Oil, grease, paint, marker, and carbon steel dust all introduce carbon and other unwanted elements. A stainless weld that was “protected” by a perfect flame setting can still be compromised by a dirty surface.

L Grades and Stabilized Grades: Helpful, Not a Pass

Metallurgists have two main answers to sensitization, and both help without making the torch risk disappear.

The first is low-carbon chemistry. Type 304L limits carbon to 0.03 percent maximum, compared with 0.08 percent for standard 304. Less carbon means less carbide formation, so the alloy can tolerate a longer exposure before chromium depletion becomes significant. The same logic applies to 316L versus 316.

The second is stabilization. Type 321 adds titanium, and type 347 adds niobium, sometimes with tantalum. These elements have a stronger affinity for carbon than chromium does, so they form stable carbides of their own and leave chromium in the matrix where it belongs. In theory, that protects the passive film.

In practice, neither grade is a license to weld stainless with a torch and forget the thermal history. Low-carbon steels can still sensitize under enough heat and time. Stabilized grades can lose some of their protection if the heat is high enough to dissolve the stabilizing precipitates, and their heat-affected zones can still suffer from other high-temperature effects. The NRC guidance is a useful reality check: all welding produces some carbide precipitation, but significant sensitization does not normally result from typical welding procedures with appropriate chemistry. A torch is not a typical controlled welding procedure.

If you regularly work with more aggressive service conditions, it also helps to understand how higher-alloy materials change the rules. Our guide to Alloy 20 welding covers a corrosion-resistant family where chemistry and filler selection become even more critical.

Weld Decay: How the Failure Shows Up

The field failure mode has an old name: weld decay. Corrosion attacks the chromium-depleted grain boundaries in the heat-affected zone, often a few millimeters from the fusion line. It may begin as a faint rust stain, then progress to pitting, crevice attack, or a leak along the weld. In chloride-bearing service, the attack can be surprisingly fast.

The cruel part is that the weld itself may look fine. A pretty bead, a clean toe line, and a passed visual inspection tell you almost nothing about chromium depletion at the grain scale.

Sensitization is the reason oxy-acetylene welding of stainless is a niche process rather than a general-purpose one. The flame can join the metal. The thermal cycle is what decides whether the joint has a future.

When Oxy-Acetylene Stainless Welding Is Defensible — and When to Walk Away

A torch, a length of stainless, and a repair that has to happen today: the temptation is to ask for a verdict. A better approach is a filter. Before you strike the flame, answer three questions.

1. What happens if this joint fails? A rattling tractor guard and a dairy line sit at opposite ends of the risk scale.
2. What will the part touch? Chlorides, food, process chemicals, and sustained heat all change the answer.
3. Can you finish the joint properly? If you cannot clean, pickle, and passivate the weld, the corrosion clock starts early.

Those questions sort most jobs into green, yellow, and red before heat settings enter the conversation.

The Green Light: Low-Consequence, Thin, Dry Work

Oxy-acetylene stainless welding makes the most sense when failure is inconvenient rather than dangerous. Non-structural brackets, guards, ductwork, decorative pieces that will be painted, and dry exhaust patches are the usual candidates. Thin sheet, roughly 0.030 inch and under, gives the process its best chance because fast travel limits time in the sensitizing range.

Farm and ranch repairs belong here too. If the alternative is no repair at all, a torch weld on a non-critical gate latch or auger guard is reasonable. Once the part carries structural load, lifts people, or holds fluid under pressure, it moves out of this category. A joint that will be cleaned and passivated afterward also earns a green light.

The Yellow Light: Proceed With Eyes Open

Some jobs are not automatically wrong, but they demand grade selection, good fit-up, and honest inspection. Dry exhaust is the classic example. It runs hot, vibrates, and often uses thin-wall tubing with little margin for burn-through. Grade 409 ferritic stainless is common there: magnetic, lower cost, and designed for exhaust, but with limited toughness and weldability that differs from austenitic sheet. Cracking and oxidation remain real risks.

Cosmetic and lightly loaded parts, removable panels, and non-pressure covers also sit in the yellow zone. Use stainless filler and flux, keep the bead small, and plan on cleanup. If the part flexes repeatedly or sees road salt, move it to red.

The Red Light: Walk Away

Do not torch-weld sanitary or food-service tubing, pharmaceutical or dairy piping, marine and pool components, pressure vessels or pressure piping, or anything governed by ASME or AWS structural codes. The same applies to joints exposed to chlorides, high-temperature service, and any weld inspected to a corrosion or radiographic standard.

Codes matter here. AWS D1.6 and similar structural stainless documents are built around arc processes, not a gas flame. Even AWS’s dedicated oxy-acetylene guidance, C47.7, is written for steel rather than stainless. Marine work raises the stakes because chloride-bearing environments are the classic trigger for stress corrosion cracking in sensitized austenitic stainless, as Rolled Alloys explains. A torch-smoothed marine repair can look fine and fail from the inside out.

Grade Selection Changes the Math

The austenitic grades 304, 304L, 316, 316L, 321, and 347 account for most stainless fabricated in shops. The L grades hold carbon at 0.03 percent maximum instead of 0.08 percent, leaving less carbon to form chromium carbides during heating. Stabilized grades 321 and 347 tie carbon up with titanium or niobium, keeping it in stable carbides instead of stripping chromium from grain boundaries.

That helps, but it does not make a torch safe. A slow flame still holds the heat-affected zone in the sensitizing range far longer than a fast arc process, and no grade eliminates that exposure. Grade 316 adds molybdenum for chloride resistance, yet it will still crack in a sensitized marine joint. Duplex stainless is a harder no: it depends on a balanced ferrite-austenite ratio and controlled heat input, making torch welding especially risky. Ferritic 409 and 430 are different: magnetic and lower cost, but with limited toughness and weld behavior that does not follow austenitic rules.

Green, Yellow, and Red Applications at a Glance

Application	Light	Why
Decorative thin sheet, painted or coated	Green	Low consequence; coating hides tint
Non-structural farm repair	Green	Failure is inconvenient, not dangerous
Dry exhaust patch	Yellow	Heat, vibration, thin wall
Light bracket or guard	Yellow	Needs fusion and cleanup
Sanitary, food, or pharma tubing	Red	Contamination and inspection risk
Marine, pool, or chloride service	Red	SCC risk in sensitized steel
Pressure vessel or code piping	Red	Qualification and toughness
High-temperature service or duplex	Red	Sensitization and phase balance

The practical rule: torch-weld stainless only when the part is thin, dry, non-structural, and cleanable. When in doubt, step up to TIG. If your torch skills are rusty, review torch basics, brazing, and safety before experimenting on stainless.

Equipment, Gases, Filler Metal, and Flux: Setting Up for Stainless

Torch-welding stainless punishes sloppy setup. Mild steel will forgive a rich flame, a mild-steel filler rod, or a brush that has been in the shop since 1998. Stainless will not. Every piece of gear you touch either protects the alloy’s chemistry or quietly ruins it. Here is the setup, in order of importance.

Torch and Tip Selection for Thin Stainless

Thin stainless wants a small, soft flame with a lot of control. Think of it as the opposite of the “crank it up” instinct that works on quarter-inch mild steel plate. For sheet in the 0.020 to 0.060 inch range, a small torch body with a No. 0 to No. 2 tip is plenty, and many welders run even smaller for 0.030 inch and under. Pick the smallest tip that will still bring the puddle up in reasonable time. If the puddle forms instantly and then runs, the tip is too big.

Smaller tips also let you keep the cone close to the joint without dumping heat into a wide area, which matters enormously for stainless because the heat-affected zone is where the metallurgical damage happens.

Regulator Settings and the 15 psig Rule

This is the one pressure number you must never treat as a suggestion. Acetylene becomes unstable and can decompose violently above roughly 15 pounds per square inch gauge. OSHA 29 CFR 1910.253(a)(1) prohibits generating, piping, or utilizing acetylene at pressures above 15 psig (103 kPa gauge). That is a hard ceiling written into federal regulation, and it exists because people have died ignoring it.

In practice, oxy-acetylene welding almost never needs anywhere near that. Typical settings look like this:

Material Thickness	Tip Size	Acetylene (psig)	Oxygen (psig)
0.020–0.030 in	No. 0	2–3	3–4
0.030–0.045 in	No. 1	3–4	4–6
0.045–0.075 in	No. 2	4–5	5–8
0.075–0.125 in	No. 3	5–6	6–10

Treat that chart as a starting point, not gospel. Hose length, tip condition, and ambient temperature all shift the numbers, and a slightly clogged tip will force you to raise pressure to get the flame you want, which is the wrong fix. Clean the tip instead. Feel genuinely matters more than the gauge: the flame should be soft enough to hold on thin sheet without the puddle running away from you.

OSHA 1910.253 also requires approved apparatus, flashback protection, and that acetylene cylinders be stored and used upright with valves closed when not in use. Flashback arrestors belong on both the regulator outlet and the torch inlet, not one or the other. Hoses should be the correct grade (red for fuel, green for oxygen), inspected for cracks, and never swapped. If you are new to gas rigs generally, the fundamentals in this guide to air acetylene torch basics transfer directly to oxy-acetylene work.

Filler Metal: ER308L, ER316L, and Matching the Base Metal

AWS A5.9/A5.9M classifies bare stainless steel welding electrodes and rods, including ER308L, ER316L, ER321, and ER347. The principle is simple: match the filler to the base metal, or undermatch slightly. ER308L is the common match for 304 and 304L. ER316L suits 316/316L and gives you molybdenum for better chloride resistance. ER347 is the choice when the base metal is stabilized with niobium (columbium).

The “L” matters. Low-carbon fillers (0.03 percent max) leave less free carbon available to form chromium carbides in the heat-affected zone, the reaction covered in the metallurgy section. If you have a choice on a torch weld, take the L-grade every time.

Do not use mild steel filler. It sounds obvious, but it happens constantly in farm shops where ER70S-2 is the only rod in the drawer. A mild-steel-filled stainless joint looks acceptable through a hood and corrodes from the inside out within a season. The filler dilutes the chromium and nickel that make the alloy stainless, and no amount of pickling brings that back.

Flux: Mandatory, Not Optional

Stainless forms a tenacious chromium oxide film that melts at a far higher temperature than the base metal itself. Without something to dissolve it, the puddle stays stiff and refuses to wet out, no matter how much heat you add. Oxy-fuel welding of stainless therefore requires a fluoride- or borate-based flux.

These fluxes are typically borax and boric acid blends, sometimes with fluorides added for the more stubborn oxides. Mix them with water to a thin paste, apply a light coat to both the joint faces and the filler rod, and let it dry. Too much flux creates slag pockets and fumes; too little and the puddle fights you. A thin, even film is what you want.

This is not brazing flux. Brazing flux is designed for lower temperatures and different oxides, and it will not do the job on a fusion-welded stainless joint. Buy flux specifically labeled for stainless steel oxy-fuel welding. For contrast on how flux chemistry changes with the alloy being joined, this breakdown of how aluminum flux welding works is a useful comparison — different metal, same principle that the flux has to match the oxide.

The Small Stuff That Decides Whether the Joint Lasts

Contamination control is half of stainless welding, and torch work gives you more chances to fail than TIG does because there is no shielding gas to blow debris away.

• Dedicated stainless wire brush. Marked, stored separately, never used on anything else. A brush that has touched carbon steel transfers free iron straight into the surface you just cleaned.
• Dedicated grinding discs. A disc previously used on mild steel is contaminated. Keep a stainless-only set, and consider a separate grinder if you do a lot of both.
• Aluminum oxide or zirconia abrasives. The safer choices for stainless. Avoid abrasives with iron-bearing compounds, and never use a wire wheel that has been on carbon steel.
• Clean gloves and clean bench. Sounds fussy. It is not. Grease and shop dust are porosity sources.
• Fresh solvent. Acetone or a dedicated stainless cleaner, applied with clean rags. Contaminated solvent just spreads oil around.

If your shop handles a lot of aluminum too, the cross-contamination rules there are equally strict; the preparation habits in this aluminum welding prep guide are a good model for how seriously to treat dedicated tooling.

One more setup note: keep a scrap piece of the same stainless grade on the bench for test welds. Run a bead on scrap before you commit to the workpiece. It tells you in thirty seconds whether your flame, flux, and filler are cooperating, far cheaper than discovering it halfway across a finished panel.

Step-by-Step: Welding Stainless Steel With an Oxy-Acetylene Torch

This procedure assumes thin 304, 316L, or similar austenitic sheet, the correct stainless filler and flux, and a clean workspace. The sequence matters more than speed. Rushing the prep or overheating one pass can undo an otherwise sound joint.

Joint Preparation and Fit-Up

Start with clean metal. Degrease both faces with acetone or isopropyl alcohol and wipe them with a clean, lint-free cloth. Avoid chlorinated brake cleaners because heat can create dangerous fumes. Then remove surface oxide with a stainless-only wire brush or a zirconia flap disc. Do not use a brush or grinding wheel that has touched carbon steel.

For thin sheet, fit a square butt joint with almost no gap. Stainless holds heat at the weld, and a generous gap invites burn-through. For stock over roughly 1/8 inch, bevel to a 60–75 degree included angle and leave a small root gap. Clamp the parts to a copper backing bar or heat sink where possible.

Mix the oxy-fuel stainless flux into a thin paste and apply it to both joint faces and the filler rod. The flux is not optional. It dissolves oxides and helps the puddle wet out, and skipping it is one of the fastest ways to produce a rough, contaminated weld.

Tacking and Distortion Control

Tack both ends first, then add a tack in the middle. Use a back-step sequence for the tacks and the finished weld: run each short segment toward the previous one rather than continuously in one direction. Back-stepping and skip-welding spread heat through the part and reduce residual stress, which matters on thin stainless sheet that likes to oil-can.

Lighting the Torch and Setting the Flame

Open the acetylene valve slightly, light the torch with a striker, then add oxygen until the acetylene feather disappears and a sharp inner cone forms. That is a neutral flame. For stainless, keep the cone crisp and avoid a heavy acetylene feather, which can add carbon to the pool. Some welders crack the acetylene just enough for a faint feather to soften the puddle, but it should be subtle. If you are new to torch handling, this guide to air-acetylene torch basics covers lighting, adjusting, and shutting down safely. The broader procedure follows conventional oxy-fuel practice, with AWS C47:2020 as a useful reference.

Torch Angle, Travel, and Filler Technique

Hold the torch at roughly 60–70 degrees leading the weld, with the filler at a shallow 15–20 degree angle. Forehand welding, with the torch pushing the puddle ahead, gives good visibility and suits thin sheet. Backhand welding, with the torch trailing and the filler following, can produce deeper penetration and a narrower heat-affected zone, but it is harder to control on thin stainless.

Keep the filler rod inside the outer, reducing envelope of the flame. Do not withdraw the rod into open air between dips, where the tip oxidizes and picks up contamination. Dip small amounts at the leading edge of the puddle, then move and dip again. Stirring the puddle or dragging the rod through the arc only makes a mess.

Heat-Input Discipline and Back Purging

Run short passes, roughly one to two inches at a time, and let the part cool between them. Common industry guidance caps interpass temperature for austenitic stainless at about 300–350°F (150–175°C) to limit sensitization and distortion. Check with a temperature stick or infrared gun rather than guessing. Never quench hot stainless in water.

When welding thin tube and the root must stay oxide-free, purge the inside with argon or use soluble purge dams. Air backing will leave a black, sugared root that becomes a corrosion initiation site.

The Procedure, Step by Step

1. Degrease both faces and the filler rod with acetone or isopropyl alcohol.
2. Brush the joint with a stainless-only brush, bevel thicker stock, and fit with minimal gap.
3. Apply stainless flux thinly to the faces and rod.
4. Tack the ends and center using a back-step sequence.
5. Light the acetylene first, then add oxygen to a crisp neutral flame.
6. Set a 60–70 degree leading torch angle and a shallow filler angle.
7. Establish a small puddle and let it wet the joint faces before adding filler.
8. Dip the rod inside the flame envelope, adding small amounts as the puddle advances.
9. Weld in short forehand passes, back-stepping and skip-welding to manage heat.
10. Let each pass cool, check interpass temperature, then clean and passivate after welding.

Practice the sequence on scrap of the same grade and thickness before touching a finished part. A clean puddle, a crisp flame, and disciplined cooling do more for joint quality than any single torch trick.

Defects, Diagnosis, and Troubleshooting: What Goes Wrong and Why

Torch-welded stainless rarely fails on the bench. It looks fine the day it ships, then blooms rust weeks later or cracks at the crater under vibration. Because the feedback loop runs in weeks, not seconds, troubleshooting is hard. Here are the failures and the fixes.

Porosity: Almost Never the Gas Settings

The instinct is to blame the regulators, but stainless porosity usually traces to contamination. Oil, grease, cutting fluid, fingerprints, and shop dust sit on the surface until the pool swallows them, and so does iron embedded by a grinding disc or brush that has touched carbon steel.

Hydrogen is the other culprit. An oxy-acetylene flame is hydrogen-rich, and unlike TIG there is no inert envelope shielding the pool. Entrapped hydrogen from the flame and moisture causes porosity, and stainless freezes too fast for it to escape. The fix: solvent-clean until a white rag stays white, use a stainless-only brush, keep flux fresh, and run slightly oxidizing.

Incomplete Fusion and Burn-Through: Two Ends of Heat Control

Incomplete fusion looks like a bead perched on the joint rather than sinking into it. Stainless carries a tenacious chromium oxide film that melts far hotter than the base metal; if the flux has not dissolved it, the puddle rides over cold metal. Heavy oxide, low heat, fast travel, and a dragged rod all give the same result.

Burn-through is the opposite failure, and thin sheet punishes impatience. Heat concentrates at the torch instead of spreading, so a 0.030-inch panel can go from “not wetting” to “hole” in under a second. Smaller tip, tighter fit-up, more tacks, and faster travel solve it. If you catch yourself adding pressure to go faster, stop: that is technique, not gas. Torch control basics are worth reviewing in Air Acetylene Welding: Torch Basics, Brazing & Safety.

Cracking: Hot Cracks, Crater Cracks, and Restraint

Austenitic stainless is ductile but susceptible to hot cracking when sulfur and phosphorus are high or restraint is heavy. Torch welding stacks the deck: slow travel, a wide mushy pool, and slow solidification give low-melting grain-boundary films time to open. Crater and centerline cracks trace to a concave crater and rapid solidification at the end of a pass. Cure: taper the flame back, dab filler as you withdraw so the crater fills convex, and backstep long seams rather than running them straight through clamps.

Heat Tint and Sugaring: The Corrosion Seed

That straw-to-purple rainbow is not decoration. Heat tint is chromium oxide that signals chromium depletion, and it should be removed before service because the depleted skin cannot rebuild its passive film properly.

Sugaring is worse: the rough black oxide that forms on a stainless root when oxygen is not excluded, porous and slaggy, a textbook corrosion initiation site. Back purging or a root-side flux is the only real prevention, and once sugared the root cannot be cleaned back to sound metal without gouging. The metallurgy behind chromium depletion is covered well by Rolled Alloys’ explanation of sensitization.

Distortion, Oil-Canning, and Undercut

Low thermal conductivity plus high thermal expansion means the metal next to the weld moves a lot while the weld stays hot. On sheet that shows as oil-canning: a dished panel that pops between two shapes. Undercut comes from excess heat and a thin filler feed, carving a groove at the toe that becomes a stress riser. Backstep, copper chill bars, and cooling between short passes beat any torch change.

Defect	Likely Cause	Correction
Porosity	Contamination, embedded iron, flame hydrogen	Clean, stainless-only brush, fresh flux, oxidizing flame
Incomplete fusion	Oxide, low heat, fast travel	Remove oxide, more heat, slower travel
Burn-through	Too much heat, poor fit-up	Smaller tip, tighter fit-up, faster travel
Hot/crater cracks	High S or P, restraint, concave crater	Fill crater, backstep, less restraint, low-sulfur filler
Sugaring	Oxygen at the root	Back purge or root flux
Undercut	Excess heat, thin filler feed	Less heat, steadier rod, correct angle

The Over-Brazing Trap

The most seductive failure is running mild steel or bronze filler into a stainless joint. It flows beautifully at low heat, wets out like a dream, and photographs well, yet nothing fused. The joint relies on a dissimilar-metal bond, creates a galvanic couple in any moisture, and fails months later at the interface. If a “stainless weld” was suspiciously easy and the filler was not ER308L, ER316L, or ER347, cut it out.

The discipline that separates a decent operator from a lucky one is diagnosis before the grinder comes out. Decide which failure mode you are looking at and treat the cause, not the appearance. The recommended practice for oxy-acetylene welding in AWS C4.7/C4.7M assumes clean metal, correct flux, and controlled heat; anything less, and this defect list becomes a menu.

Post-Weld Cleanup: Pickling, Passivation, and Restoring Corrosion Resistance

A torch-welded stainless joint is not finished when the flame goes out. It is finished when the chromium oxide film is back. That straw-gold to black band running along the bead is heat tint, a thickened oxide that forms where the metal got hottest and where chromium has been pulled out of solution. Leave it there and you have built a corrosion initiation site into the part. The weld may look fine in the shop and show rust within weeks outdoors.

So cleanup is corrosion control, not cosmetics. Heat tint removal restores the chromium-rich passive film that corrosion resistance depends on, and on an oxy-acetylene weld — with its wide, slow-cooled heat-affected zone — that film is exactly what you damaged.

Step One: Mechanical Removal Without Contaminating Anything

Start mechanically, with tools that have never touched carbon steel. A stainless wire brush kept in a labeled drawer, grinding discs reserved for stainless only, and aluminum oxide or zirconia abrasives will do it. Carbon steel contamination is the quiet killer here: one pass with a shop brush that stripped mill scale off mild steel can embed free iron into the surface, and that iron rusts first and takes the surrounding stainless with it.

Before you reach for acid, check whether iron is present. A ferroxyl test or a copper sulfate test will show free iron on the surface. If it tests positive, the part needs pickling, not just passivation. Light surface tint on thin sheet often comes off with a fine zirconia disc alone — but do not grind thin gauge into a dish trying to chase a shadow of color. For solvent and prep work generally, this guide to a proper welding prep cleaner covers the degreasing mindset that applies to any alloy.

Pickling vs. Passivation: Two Different Jobs

People use these words interchangeably. They are not the same process, and after a torch weld you usually need both, in order.

Process	What it does	What it removes
Pickling	Chemically dissolves the contaminated surface layer and heat tint	Oxide, a thin skin of base metal, embedded iron
Passivation	Rebuilds the chromium oxide film on clean metal	Free iron and surface debris only — no tint

Pickling gels are typically based on nitric acid and hydrofluoric acid. Applied by brush or spray, left to dwell, then rinsed thoroughly, they strip the tint and the chromium-depleted skin beneath it. Passivation then follows on clean metal, usually nitric or citric acid, and lets the passive film reform. ASTM A380 covers cleaning, descaling, and passivation of stainless parts, equipment, and systems; ASTM A967 sets out the chemical passivation treatments themselves. Citric acid passivation is a lower-hazard option many shops use for non-critical work, though it is gentler and not a substitute for pickling where heavy tint remains.

Handling Pickling Chemicals Safely

Hydrofluoric acid is not a chemical to treat casually. It attacks skin and bone, and burns may not hurt immediately. Nitric acid adds its own hazards. Required: acid-resistant gloves, a face shield over safety glasses, apron, and real ventilation — outdoors or with fume extraction, not a closed shop corner. Never mix pickling waste with other chemicals, and dispose of it through a licensed handler per local rules. Store acids in labeled, secondary-contained containers away from acetylene cylinders.

Inspection Checklist

• Visual: no tint, no discoloration, uniform matte-to-bright surface, no residual gel in corners.
• Dye penetrant: for surface-breaking defects near the weld toes, especially on thin sheet where undercut hides easily.
• ASTM A262: request intergranular corrosion testing when the joint is critical — these practices detect susceptibility to intergranular attack in austenitic grades.

Do the mechanical work, pickle what needs pickling, passivate what is now clean, and inspect with your eyes and — where it matters — with a penetrant or a lab. Skipping this stage is the most common reason a torch-welded stainless repair fails months later, far from the shop that made it.

Oxy-Acetylene vs TIG, MIG, and Stick for Stainless: An Honest Comparison

The comparison usually starts with price and ends with regret. A basic torch rig is cheaper than a TIG machine, and it works when the shop has no power. That makes oxy-acetylene tempting for stainless repair. But stainless is not mild steel, and the cheapest way to fuse it is rarely the cheapest joint once corrosion, cleanup, and rework enter the bill.

The table below compares the four processes on the factors that actually decide whether the part lasts.

Factor	Oxy-acetylene	TIG	MIG	Stick
Heat input and control	High, broad, slow; heat and filler not independent	Precise; independent heat and filler control	Higher, but controllable	High and less precise
Cleanliness	Flux, heat tint, oxides	Cleanest; inert gas shield	Good; some spatter	Slag and spatter
Corrosion resistance	Poorest; sensitization risk	Best	Good	Fair
Speed	Slow	Slow	Fastest	Moderate
Equipment cost	Lowest upfront	Highest	Mid to high	Low
Portability	Best; no electricity	Poor	Fair	Good
Skill required	Deceptively high	High	Moderate	Moderate
Positional capability	Limited, mostly flat	Excellent	Good; pulsed helps	Excellent
Thickness range	Thin sheet	Thin to mid	Mid to thick	1/8 inch and up
Code acceptance	Rarely; not structural	Widely accepted	Widely accepted	Widely accepted

TIG Is the Reference Standard for Thin Stainless

TIG earns that status for two reasons: an inert argon or argon-helium shield, and independent control of heat and filler addition. You can let off the pedal at the edge of thin sheet without stopping the filler, or add wire without changing amperage. On 0.030-inch stainless, that control is the difference between a flat bead and a hole. It also keeps oxygen away from the pool, so the finished joint has the best chance of keeping its corrosion resistance. TIG also leaves the least cleanup, which matters when every heat tint must be pickled. The same precision is why TIG dominates critical thin-wall work, from aerospace TIG welding to sanitary tubing.

MIG and Stick Have Their Place

MIG is the speed choice. Depending on thickness and position, it runs short-circuiting, spray, or pulsed transfer, and on thicker stainless it will outrun TIG by a wide margin. On thin sheet, its speed can outrun the puddle and cause incomplete fusion, so the advantage shrinks as gauge drops. For structural fittings and heavier repairs, though, it is efficient.

Stick with stainless electrodes is the outdoor, field-ready option. It handles wind and dirty conditions and works well on 1/8-inch and thicker material, but it is rough on sheet. Electrode selection matters, so use a proper all-position welding rod guide rather than grabbing whatever is in the rod oven.

The Cost Math Is Not the Purchase Price

A basic oxy-acetylene rig costs far less than a TIG machine plus an argon cylinder, regulator, and torch consumables. Argon and TIG consumables are an ongoing cost that the torch largely avoids. The torch also needs no electrical power, which is a genuine advantage for remote field repairs and farm calls.

That is where the savings usually end. Torch welding stainless consumes filler and flux, produces more cleanup, and carries a higher risk of sensitization damage because heat goes in slowly and leaves slowly. That damage can force a repair a TIG joint would never have needed. Post-weld pickling and passivation add time and chemical cost. AWS publishes recommended practices for oxy-acetylene welding, but structural stainless falls under AWS D1.6, which is built around arc processes. For anything inspected, the torch is usually on the wrong side of the code line.

Practical Bottom Line

Choose TIG for thin, visible, sanitary, or corrosion-critical stainless. Choose MIG when speed and thickness matter. Choose stick for outdoor structural work on heavier plate. Keep oxy-acetylene for non-critical thin sheet, dry service, and places where no power exists. If the joint must pass inspection or survive chlorides, leave the torch in the truck.

Safety, Common Mistakes, and the Practical Bottom Line

Torch welding rewards preparation and punishes shortcuts. The safety rules below are not bureaucratic extras, and the mistakes in the second half of this section are the ones that send perfectly good stainless parts to the scrap bin. Read both before you light up.

Safety Rules That Are Not Optional

Start with the fire watch. OSHA 1910.253 requires combustibles to be moved at least 35 feet from the work or protected with approved covers. Sparks travel farther than most people expect, and a torch produces them continuously. Keep a fire extinguisher within reach and stay at the work for a while after the flame goes out.

Handle cylinders like the pressure vessels they are. Acetylene cylinders must be stored and used valve-end up because the gas is dissolved in acetone inside a porous mass; laying one on its side can send liquid acetone into the regulator. Secure them upright, keep them away from heat, and open valves slowly. Flashback arrestors belong on both the torch and the regulator end of the lines. OSHA also prohibits using oxygen as a substitute for compressed air, a mistake that has killed people who thought they were just blowing dust off a part.

Then protect your lungs. Stainless welding fume contains chromium and nickel compounds, and hexavalent chromium is a recognized carcinogen. OSHA 29 CFR 1910.1026 sets a permissible exposure limit of 5 micrograms per cubic meter as an 8-hour time-weighted average. Use local exhaust ventilation or a fume extractor at the work, not just an open shop door. If you weld stainless regularly, have your exposure checked.

The Mistakes That Cause the Most Rework

• Using a carburizing flame: Excess acetylene adds carbon to the pool and makes a dirty, brittle weld. Set a neutral flame, then feather it only slightly.
• Welding dirty metal: Oil, grease, paint, and mill scale turn into porosity and corrosion sites. Clean until a white rag stays white.
• Reusing a carbon steel brush: Embedded iron particles rust and seed corrosion. Keep a stainless-only brush, clearly marked.
• Skipping flux: Oxy-fuel stainless flux is mandatory. Without it, the chromium oxide film will not break down properly and the puddle will not wet out.
• Quenching hot stainless in water: It can introduce residual stress and is not a substitute for controlled cooling. Let parts cool in still air.
• Over-grinding thin sheet: It removes sound metal and creates a thinner, weaker joint. Grind to blend, not to reshape.
• Assuming a pretty bead means a sound joint: A shiny cap can hide lack of fusion, sensitization, or a sugared root. Judge the process, not the cosmetics.

The Practical Bottom Line

Oxy-acetylene can join stainless for non-critical, dry-service, thin-gauge work when you use the right filler, flux, flame, and cleanup. It is a real option for a farm repair, a decorative bracket, or a part that will be painted and kept dry. It is not a substitute for a controlled arc process on anything that matters structurally.

For structural work, sanitary and food-service tubing, marine or pool environments, pressure piping, or any joint inspected to a corrosion or radiographic standard, TIG or MIG is the correct tool. The AWS C47 recommendation covers oxy-acetylene welding of steel, but that does not make every torch joint code-acceptable for stainless service. If the joint has to pass inspection, choose the process the code recognizes.

A Short Checklist Before You Strike the Torch

1. Clear combustibles 35 feet or cover them, and stage a fire extinguisher.
2. Stand acetylene cylinders upright and check flashback arrestors.
3. Set up fume extraction before the first spark.
4. Clean and degrease both faying surfaces.
5. Confirm stainless-only tools, brushes, and abrasives.
6. Set a neutral flame with a crisp inner cone, then feather slightly.
7. Use the correct stainless filler and flux for the grade.
8. Plan the post-weld pickle and passivation before you weld.

If you are comparing torch work with other processes, our guide to air-acetylene torch basics and safety covers the smaller rig many home shops already own, and the advantages of stick welding explains why SMAW still earns its place in field repair. Both are useful context before you decide which process deserves your time on stainless.

Frequently Asked Questions

Can you really oxy-acetylene weld stainless steel?

Yes, the metal will fuse, and the process is used occasionally for thin sheet, repairs, and decorative work. However, torch welding gives you less heat control and no inert shielding, so the finished joint usually has lower corrosion resistance than a TIG weld. Treat it as a niche repair method rather than a production process.

Why does torch-welded stainless rust so quickly?

The flame holds the heat-affected zone in the 800-1500 degrees F range long enough for chromium carbides to form at the grain boundaries, which strips chromium from the surrounding metal. On top of that, heat tint and any iron contamination from tooling destroy the passive film. Together those effects leave the weld zone unable to resist moisture and chlorides.

What filler rod should I use for 304 stainless with a torch?

ER308L is the standard match for 304 and 304L, and ER316L is used for 316 and 316L base metal. Both are classified under AWS A5.9. Never substitute mild steel filler, because the resulting joint will rust at the weld line and can fail prematurely.

Do I need flux to oxy-acetylene weld stainless steel?

Yes. Stainless steel forms a refractory chromium oxide film that the flame alone will not remove, so a borate- or fluoride-based oxy-fuel flux is required to keep the pool clean and let the metal wet out. The flux residue must be washed off completely after welding because it is corrosive.

Is an oxidizing or neutral flame better for stainless?

Most guidance calls for a neutral flame or a very slightly oxidizing one, which helps keep carbon out of the pool and keeps the molten metal fluid. A carburizing flame with excess acetylene can add carbon and soot, which is exactly what you do not want in stainless.

Can oxy-acetylene welding be used on sanitary or food-grade stainless?

It should not be. Sanitary tubing requires a smooth, oxide-free, fully penetrated root and a documented procedure, and torch welding cannot reliably deliver that. TIG with an argon purge is the accepted process for dairy, food, beverage, and pharmaceutical work.

How do I remove heat tint after torch welding stainless?

Remove it mechanically with a stainless-only brush or a zirconia/aluminum oxide abrasive, then chemically with a pickling gel, and finally passivate the surface to rebuild the chromium oxide layer. Citric acid passivation is a lower-hazard alternative for some applications, and ASTM A967 describes accepted passivation treatments.

Is oxy-acetylene stainless welding allowed by welding codes?

For structural and pressure work, oxyfuel gas welding is generally not accepted without a qualified procedure and testing, and many stainless structural codes are written around arc processes. If a job is inspected, engineered, or covered by AWS D1.6 or ASME, confirm the permitted process before you strike a flame.

What is the biggest safety concern when torch welding stainless?

Fume is the one people underestimate. Stainless welding fume contains hexavalent chromium and nickel compounds, and OSHA limits hexavalent chromium exposure to 5 micrograms per cubic meter as an 8-hour average. Use local exhaust ventilation or supplied air, and keep the fire watch and 35-foot combustible clearance rules in place.