Stainless Tube Welding: The Complete Practical Guide

The weld looked perfect from the outside. Six weeks later the tube run was leaking at the root, and the borescope showed a black, sugared oxide layer that had been hiding inside the joint the whole time. If you have ever chased a stainless tube weld that passed visual inspection and failed in service, this guide is for you.

Why Stainless Tube Welding Is Different From Welding Carbon Steel

The joint in that opening story looked fine because everything visible was fine. The cap was uniform, the toes wet in cleanly. What failed was invisible: chemistry and atmosphere inside a joint nobody could see, plus more heat than the metal wanted. That is why stainless tube welding confuses good carbon steel welders. The skills transfer; the priorities do not.

Here is the short answer. Stainless tube welding is a contamination-control and heat-control problem far more than a bead-appearance problem. On carbon steel, a decent-looking bead usually means a decent weld. If the appearance is good, the metallurgy is probably acceptable too. On stainless, appearance and integrity come apart constantly. A beautiful cap can hide a sugared, oxide-crusted root. A clean-looking root can sit beside a heat-affected zone that has lost its corrosion resistance. Four material properties explain why.

The Four Material Properties That Change Everything

Low thermal conductivity. Austenitic stainless conducts heat at roughly 16 W/m·K, about one third the rate of carbon steel at roughly 45–50 W/m·K, as documented in the IMOA guide to fabricating austenitics. Heat does not spread away from the arc; it piles up. The pool stays fluid longer, the heat-affected zone runs hotter, and the backside of a thin wall can be glowing while the front still looks normal. Stainless rewards fast travel and punishes the instinct to soak it in a bit.

High thermal expansion. Austenitic grades expand at about 17–18 µm/m·°C against roughly 11–12 µm/m·°C for carbon steel, around 50% more movement for the same temperature rise. Combine that with trapped heat and you get shrinkage stress, bowing, and pulled fits. A tube run that measured perfectly at fit-up can close its root gap or spring out of alignment halfway around the root pass.

The chromium oxide passive layer. Stainless is stainless because of a chromium oxide film only a few nanometres thick. That film stops rust, and it only re-forms where oxygen is available. Weld it badly and you do not just create a weak spot, you create a spot with no passive layer in a warm, wet, chloride-rich environment that is exactly where pitting starts.

The backside is a weld surface too. In a tube, the root is not hidden backing, it is the wetted surface. Product flows over it, bacteria can settle on it, corrosive fluid touches it first. Weld it in air and you grow heat tint: straw, then blue, then rough black oxide, and at the extreme a crystalline crust welders call sugaring. Oxygen above roughly 100 ppm inside the tube during welding is enough to start that progression, which is why purging is not optional on stainless and effectively unnecessary on carbon steel.

The Three Pillars: Cleanliness, Atmosphere, Heat

Almost every stainless tube defect traces back to one of three things, and the rest of this guide is built around them.

• Cleanliness — no oil, grease, paint, marker, shop dust, or free iron near the joint, and dedicated stainless tooling to keep it that way. Chlorinated solvents are a separate hazard: they can trigger stress corrosion cracking, so use acetone or an approved alternative.
• Atmosphere — shielding gas on the front, a controlled purge on the back, measured with an oxygen analyser rather than guessed from a regulator gauge. Nearly every purge question reduces to oxygen level and turbulence.
• Heat input — amps, travel speed and interpass temperature managed so the metal does not linger in the sensitising range and does not distort beyond recovery.

What a Sound Weld Actually Looks Like

From the outside: a smooth, evenly rippled cap with no undercut, no arc strikes, no spatter, and heat tint that is light straw at worst. From the inside: full penetration, a root flat to slightly convex, no suck-back or excessive drop-through, and a silvery or straw-coloured surface with no grey or black oxide. If a borescope shows the same colour inside that you see outside a well-run joint, you did the job properly.

Judge colour as a proxy for atmosphere, and penetration as a proxy for heat and fit-up. Nothing else in a photograph tells you much.

What This Guide Covers

The next sections work through the chain in order: grade selection, process choice, cutting and fit-up, purging, heat and technique, post-weld cleaning and inspection, codes and paperwork, safety around hexavalent chromium and purge gas, then the defect table showing what went wrong and what rework costs. Read it as a sequence, because failures compound in that order. Bad fit-up forces extra heat, extra heat burns the purge, and a burned purge is a leak six weeks later. Start with the three pillars and the rest follows.

Stainless Steel Grades for Tube Work: 304, 316, Duplex and What Actually Matters

Most stainless tube failures I have seen traced back to a grade choice made early, often by someone reading a line off a drawing rather than thinking about the fluid, the temperature and the chloride level. You do not need a metallurgy degree to get this right. You need to know what each family is good at, where it gives up, and how the filler metal interacts with the base metal under the arc.

The Austenitic Workhorses: 304, 316, 321 and 347

Type 304 is the default tube grade worldwide: roughly 18–20% chromium and 8–12% nickel, with the chromium forming the passive oxide layer that makes stainless stainless. Type 316 keeps the same base and adds about 2–3% molybdenum, which noticeably improves resistance to chlorides and to pitting in chemical and marine-adjacent service. If a line carries chlorinated water, cleaning chemicals or coastal air, 316 is the minimum, not an upgrade.

Types 321 and 347 are the stabilized cousins. 321 is stabilized with titanium and 347 with niobium, both tying up carbon as stable carbides so it cannot cause trouble at temperature. They are common in steam, exhaust and heat-exchanger tubing where service sits in the sensitization band.

Why the “L” Grades Exist: Sensitization and Weld Decay

Standard 304 and 316 allow up to 0.08% carbon; 304L and 316L cap it at 0.03%. That matters because between roughly 425°C and 870°C (800–1600°F), chromium carbides precipitate along grain boundaries and pull chromium out of the metal beside them. Those depleted zones lose their passive layer and become the path of least resistance for corrosion. In a welded tube that band sits right next to the weld, which is why the failure pattern is called weld decay. The International Molybdenum Association’s fabrication guidance covers the mechanism and the practical limits that avoid it.

This is where heat input stops being an abstract number. A hot, slow bead with a wide HAZ parks the metal in that window longer, and so does letting interpass temperature climb. Low-carbon grades buy margin, not a licence to cook the joint.

Ferrite in Austenitic Weld Metal: A Little Goes a Long Way

Here is the counterintuitive part: fully austenitic weld metal is prone to solidification cracking. A small amount of delta ferrite disrupts the low-melting films that cause centreline cracks, so austenitic stainless weld metal typically targets about 4–12% delta ferrite, often quoted as roughly 3–10 ferrite number. Too little and you risk cracking; too much can hurt toughness and, in some aggressive media, corrosion resistance.

You predict it with constitution diagrams. Schaeffler uses chromium and nickel equivalents, DeLong refines that with a nitrogen term for austenitic fillers, and WRC-1992 is the modern choice. The British Stainless Steel Association’s explanation of the Schaeffler and DeLong diagrams is a good plain-language walkthrough, and Hobart’s technical guide on ferrite shows how filler selection shifts the number. Fully austenitic consumables such as 310 or the super-austenitics need tighter heat control precisely because they have no ferrite buffer.

PREN: One Number for Ranking Corrosion Resistance

Pitting resistance equivalent number, or PREN, is commonly calculated as Cr + 3.3Mo + 16N. It is not a guarantee, but it ranks alloys sensibly against chloride pitting: 304 lands near 18–20, 316 near 24–26, and duplex 2205 in the mid-30s. The high-performance austenitics reference from the German stainless steel association sets out the formula and its limits, including the fact that PREN says nothing about stress corrosion cracking or high-temperature behaviour.

Where Duplex and Super-Duplex Fit

Duplex 2205 is roughly half austenite, half ferrite. That structure gives it about twice the yield strength of 316 plus much better resistance to chloride stress corrosion cracking, which makes it popular for process water, desalination and pulp lines. Super-duplex grades such as 2507 push PREN above 40 for severe chloride service. The trade-offs are real: duplex is less forgiving of heat input, needs nitrogen-bearing shielding and purge gas, and should not run at elevated service temperatures where the ferrite can embrittle. Filler matching is not optional either, so you will see ER2209 and ER2594 rather than a 316L rod. If your work drifts into the super-austenitic end of the family, the procedures in this guide to Alloy 20 welding are a useful comparison for how much stricter the rules become.

Questions That Actually Decide the Grade

Before the drawing gets stamped, three questions do most of the work: what is in the fluid (chlorides above all), what temperature will the tube see in service and during cleaning, and how much strength does the line need? A chloride-bearing fluid plus a warm wall points away from 304 every time. A hot wall in the sensitization range points toward low-carbon or stabilized grades. A line that needs thinner walls and higher allowable stress points toward duplex.

Common Tube Grades at a Glance

Grade	What defines it	Typical filler	Typical service
304 / 304L	18–20% Cr, 8–12% Ni; L caps carbon at 0.03%	ER308L	Food, water, general process lines
316 / 316L	Adds ~2–3% Mo for chloride resistance	ER316L	Chemical, pharmaceutical, coastal, chlorinated water
321	Titanium-stabilized against sensitization	ER347 (or ER321 where specified)	Steam, exhaust, elevated temperature
347	Niobium-stabilized	ER347	High-temperature and cyclic service
2205 duplex	~50/50 austenite-ferrite, PREN mid-30s	ER2209	Chloride process water, desalination
2507 super-duplex	PREN above 40, high strength	ER2594	Severe chloride, offshore, seawater

Get the grade right and the rest becomes a discipline problem: keeping it clean, keeping the backside under gas, and keeping the heat down. Choose wrong and no amount of technique will save the run.

Choosing the Process: TIG, Orbital, MIG, Laser and Where Each Fits

Pick the wrong process and no amount of skill saves the joint. Pick the right one and the welder’s attention goes where it belongs: the purge and the heat. On stainless tube the decision usually comes down to four things — wall thickness, position, the finish required inside the tube, and how many identical joints you have to make this week.

Manual GTAW: Still the Default

Manual TIG (GTAW) remains the workhorse for root passes and thin-wall work. Heat input is controlled precisely with a pedal or fingertip control, and no flux ever touches the joint. That second point matters more than most people realise: the inside of a stainless tube weld is a real weld surface, so any process that leaves residue behind creates a corrosion site you may never see.

From roughly 0.9 mm to 6 mm, TIG covers root passes, full-penetration single-sided joints and most field work where the tube cannot be rotated. It is slow, though, deposition rates are low, and good tube welders are expensive. For a handful of joints on a pilot line that cost is irrelevant. For 400 identical ferrule joints in a pharmaceutical skid it is the wrong tool entirely.

Orbital TIG: Closed Head and Open Head

Closed-head orbital welding seals the arc in an inert chamber around the joint, and it is the standard for hygienic, pharmaceutical and semiconductor tubing. The head clamps on, the chamber floods with argon, and the torch rotates at a programmed speed. Because the atmosphere is controlled locally, the operator is not fighting draughts or the shop’s general air quality. Repeatability is the real prize: every joint gets the same current, travel speed and purge, which is exactly what a validation document demands.

Open-head orbital works the same way mechanically but without the sealed chamber, so the tube bore still needs a proper purge. It suits larger diameters, awkward locations and heavier wall where a closed head will not fit. Both variants need a qualified procedure, and both punish anyone who forgets that the machine only does what the program says.

GMAW, FCAW and SAW: Where They Earn Their Place

MIG (GMAW) on stainless is fast and forgiving on thicker wall, but the transfer mode decides the outcome. Short-circuit transfer on thin stainless tends toward cold lap and spatter, so pulsed or spray transfer is generally preferred where GMAW is used at all. Pulsed MIG gives decent control on 3–6 mm wall and suits structural or non-hygienic tube where a little spatter can be cleaned off. Spray transfer needs higher current and flat position, which limits it on tube.

Flux-cored and submerged arc welding take over on heavy wall, typically 8 mm and above, where deposition rate drives the economics. The catch is slag and flux. Both must be removed completely, and trapped flux residue inside a tube causes corrosion that surfaces months later as a pinhole under a weld. If no grinder or brush can reach the inside of the joint, do not use a process that puts flux there.

Laser and Hybrid Laser: Speed for Repeat Parts

Laser welding, and hybrid laser-arc variants, run fast with narrow heat-affected zones and low distortion, which makes them attractive for thin-wall tube assemblies produced in volume. The barriers are capital cost, joint fit-up tolerance and a genuine safety program. A gap a TIG welder would close with filler will simply fall through a laser beam. For how these systems behave in practice, our guide to alpha laser welding covers the practical side.

Process Comparison at a Glance

Process	Typical wall	Deposition rate	Skill level	Best application	Main limitation
Manual GTAW	0.9–6 mm	Low	High	Root passes, thin wall, field work	Slow, operator-dependent
Closed-head orbital	1–12 mm	Medium	Medium (programming)	Hygienic, pharma, semiconductor	Fits only certain geometries
Open-head orbital	3–25 mm	Medium	Medium	Large diameter, awkward access	Needs external purge discipline
Pulsed GMAW	3–10 mm	High	Medium	Structural tube, fill passes	Spatter, wider HAZ
FCAW / SAW	8 mm+	Very high	Medium	Heavy-wall pipe and tube	Slag and flux removal
Laser / hybrid	0.5–6 mm	Very high	High (setup)	High-volume thin-wall parts	Cost, tight fit-up

Filler Metal and Autogenous Welding

Filler selection follows the base metal, not the process: ER308L for 304L, ER316L for 316L, ER347 for 347, and ER2209 for duplex 2205. An L-grade filler on L-grade tube keeps carbon low and reduces sensitisation risk; a 308 filler on a 316 tube throws away the molybdenum you paid for. Ferrite control and the rest of that metallurgy are covered in the grade section and in this Hobart guide to ferrite in austenitic weld metal.

Autogenous, or no-filler, welds are viable only on thin wall with tightly controlled fit-up, because there is no filler to compensate for gaps or shrinkage. In practice that means thin-wall instrumentation tubing, some laser and orbital work, and joints that were machined rather than eyeballed. On anything with a gap, autogenous welding produces a concave, underfilled root.

The Trade-Offs Nobody Mentions in the Brochure

• Travel speed: faster processes cut heat input per millimetre, but demand better fit-up and steadier torch or head alignment.
• Cleaning: TIG leaves no slag but does leave heat tint. FCAW leaves slag and spatter. Laser leaves almost nothing, provided the joint was clean to begin with.
• Spatter: on stainless it is not cosmetic. Every embedded particle is a corrosion nucleus, and grinding it out thins the wall.
• Access: a closed head needs clearance around the joint. If the tube sits 50 mm from a wall, that plan dies on site.

Where the work is thin, positional and inspected from the inside, TIG and orbital TIG dominate for good reason. Where it is aerospace-grade thin wall with code-driven documentation, the discipline overlaps heavily with aerospace TIG welding practice. Whatever you choose, the process is only the vehicle. The purge and the heat input still decide whether the weld survives.

Cutting, Fitting and Cleaning: The Prep Work That Decides the Weld

Most stainless tube welds are won or lost before the arc starts. No torch hand rescues a joint with a 4 mm gap, a bevel smeared with cutting oil, or a root face chewed up by a carbon steel disc. Prep sets the tolerances the weld has to live inside, so treat it as part of the procedure rather than a chore before it.

Cutting: The Method Sets the Contamination Baseline

Any cut is acceptable if it leaves a square, undamaged end and no foreign metal behind. The methods are not equal, though:

• Bandsaw or cold saw with a dedicated stainless blade: the shop default. Clean, square, minimal heat, no embedded grit.
• Abrasive chop saw: fast and cheap, but the wheel sheds abrasive and can carry carbon steel grit from previous jobs. That free iron gets hammered into the cut face and shows up months later as rust blooms. Keep a stainless-only wheel and still dress the face afterwards.
• Plasma cutting: fine for heavier wall and field tack-cutting, but it leaves a nitrided, oxidised edge that needs grinding back to bright metal.
• Lathe or machining: the best finish available, and the usual choice for hygienic and high-purity tube where the bore must stay burr-free.

Whichever you use, the cut face is a weld surface. Treat it like one.

Squaring, Beveling and Fit-Up Geometry

Square the end to within a degree or so, then bevel. A common tube bevel is a 37.5-degree land per side, giving a 75-degree included angle, with a root face of roughly 1.5–2 mm on heavier wall. Thin hygienic tube is often welded square-butt with no bevel at all, so check the WPS before you touch the grinder.

Then measure. Root gap should be tight and consistent around the circumference. Too much gap forces the welder to add filler just to bridge the joint, which raises heat input, increases shrink and pulls the bore out of alignment. High-low (mismatch) matters too: an internal step creates a ledge that traps product, promotes incomplete fusion and cannot be welded over. If the gap opens up on one side, the fit is wrong — re-cut rather than compensate with wire.

Deburr inside and out. The internal lip left by a saw or a tube cutter is a crack starter and, in sanitary service, a bacteria trap. A half-round file or internal deburring tool, used dry and clean, takes seconds.

Cleaning: Solvent First, Stainless-Only Tools Always

Wipe the bevel and the first 25 mm of outside and inside surface with acetone or an approved solvent, using clean lint-free cloth, and let it flash off. Grease, marker ink, tape adhesive and cutting fluid all turn into porosity or carbon pickup under the arc. If you also work aluminium, the same prep logic is covered in Aluminum Cleaner for Welding: How to Prep Aluminum Right.

Keep grinding discs, wire brushes, files and flapper wheels reserved for stainless only. A brush that has touched carbon steel transfers free iron in one stroke, and that iron becomes a corrosion site. Industry fabrication guidance is blunt about it: embedded iron contamination is one of the most common causes of rust staining on new stainless work.

Storage, Handling and Protecting the Bevel

Store tube off the ground, away from carbon steel racking that sheds rust, and cap or tape the ends. Bevels pick up grease from gloves, dirt from truck beds and moisture from overnight damp. A plastic cap or a wrap of clean tape costs pennies; re-cutting a contaminated end costs an hour. ASTM A380 covers cleaning practice and contamination control in detail and belongs on the shelf.

Tacks: Small, Fully Fused and Consumed

Tacks should be small, fully fused and either consumed by the weld or ground out completely. A cracked, cold-lapped or oxidised tack becomes a root defect the moment the arc reaches it, and grinding it out now is far cheaper than cutting the joint later. Use two tacks on small bore and three or four on larger tube, placed symmetrically so shrink stays balanced. Purge before tacking, not after, or you bake a sugared patch into the bore at every tack.

Field Conditions and a Fit-Up Checklist

Wind scatters shielding gas, rain puts water in the joint, and dirty pipe racks put grit exactly where you do not want it. Build a shelter, keep the joint capped until the last minute, and never weld a wet bevel.

1. Cut square; bevel per WPS at 75 degrees included where specified.
2. Root face 1.5–2 mm on heavier wall; square butt on thin hygienic tube.
3. Deburr inside and out; no lip in the bore.
4. Gap consistent all round; mismatch within the project limit.
5. Solvent-wiped, dry, no grease or marker.
6. Stainless-only discs, brushes and files.
7. Tacks small, fused, purged, and scheduled for consumption.
8. Ends capped and bevel protected until welding starts.

Get these eight right and the root pass is a matter of technique. Get them wrong and no downstream skill will save the joint.

Purging and Shielding Gas: The Make-or-Break Variable

If one variable decides whether a stainless tube weld survives, it is the atmosphere inside the tube. The torch shield protects the cap; the purge protects the root. On a tube, the root is not a hidden backing surface. It is exposed to product, cleaning chemicals and stagnant water. A perfect cap over a sugared root is a corrosion failure waiting for a service cycle.

Why the Backside Is a Weld Surface

When the root becomes molten, it is just as reactive as the visible side. Oxygen and nitrogen in the tube react with chromium, forming oxides and depleting the passive layer that makes stainless stainless. The result ranges from light heat tint to the rough black crust welders call sugaring. Purging displaces that air so the root solidifies clean and retains its corrosion resistance.

Oxygen Targets and Heat Tint

Heat tint is an oxide film, and its colour gives a rough indication of oxygen exposure. A widely applied industry maximum at the start of welding is about 0.1% oxygen (1,000 ppm). Many critical and hygienic specifications require below 100 ppm, and some demand tighter still. ESAB’s gas selection guidance and practical purge tips from Huntingdon Fusion both stress that measurement, not guesswork, should drive the decision.

Oxygen at the root	Typical internal appearance	Where it fits
Above 0.1% (1,000 ppm)	Straw, blue, grey or black sugar	Reject for most critical tube work
0.05–0.1% (500–1,000 ppm)	Light straw	General utility, only where the specification allows
0.01–0.05% (100–500 ppm)	Pale gold	Good practice for process tubing
Below 100 ppm (0.01%)	Near-colourless	Hygienic, high-purity and critical service
Below 20 ppm	Bright silver	Semiconductor and ultra-high-purity work

Colour is a useful warning, not a perfect gauge. Heat input, travel speed and grade also affect tint, which is why fabrication guidance for austenitic stainless steels treats atmosphere and heat as a combined problem.

Dams, Tape and Purge Chambers

Gas has to be held where it is needed. Inflatable or foam purge dams seal the tube on both sides of the joint; water-soluble dams disappear after the weld and are useful where retrieval is difficult. Tape works for simple open-root joints but is easy to disturb. For fittings and short sections, a sealed purge chamber uses less gas and gives more reliable coverage. Always leave a vent so displaced air can escape, and avoid dead pockets where gas cannot circulate.

Pre-Purge Math and Flow Control

Purging is volume displacement, not a quick squirt. The chamber must be flushed with several times its volume. A 3 m length of 50 mm inside-diameter tube holds about 5.9 litres; five volume changes mean roughly 29.5 litres. At 15 litres per minute, that is about two minutes before welding begins. Use a high flow to purge, then reduce to a gentle flow during welding. Excessive flow creates turbulence that can pull shop air into the joint, producing intermittent discoloration that looks like a gas problem but is really an flow problem.

Choosing the Gas

Argon is the default for austenitic stainless because it is inert and dense enough to displace air effectively. Nitrogen suits some applications, particularly duplex grades where it helps maintain phase balance, but it is not a universal substitute and should follow the procedure. Argon-hydrogen forming gas, commonly 95/5, reduces residual oxides and can help achieve lower oxygen levels, but hydrogen is flammable and may not suit every grade or workplace. Treat it with proper ventilation, leak checks and a hot-work permit. The same principle of matching gas to material applies elsewhere, including aluminum MIG welding, where shielding behaves quite differently.

Measurement, Open Roots and No-Analyzer Days

An oxygen analyser sampling at the vent is the only real confirmation that the purge is ready. A regulator gauge shows delivery pressure, not oxygen content. Open-root joints need a longer purge zone and well-secured dams; closed chambers are easier to control but still need a vent and enough flush time. Without an analyser, purge by calculated volume, add a safety margin, weld a test coupon and inspect the root colour before committing to production. ASTM A380 is a useful reference for subsequent cleaning, but it cannot rescue a contaminated root.

Before striking an arc, confirm the dams are sealed, the vent is open, the analyser has stabilised below the specified limit, and the weld procedure states the purge flow for both pre-purge and welding. AWS D18.2 provides reference discoloration levels for the inside of austenitic stainless tube welds, making it a practical acceptance tool when the specification is silent. Get the atmosphere right and the rest of the weld has a chance. Skip it, and every later inspection is just a search for the damage already done.

Heat Input, Technique and Distortion Control

Stainless steel conducts heat about a third as well as carbon steel and expands roughly 50% more when it gets hot. Those two facts explain almost everything that goes wrong on a tube joint. Heat piles up in a narrow band beside the weld, that band tries to expand, the cold metal around it refuses to move, and something has to give. Usually it is the fit-up you spent twenty minutes getting right. Controlling heat input is how you stop that.

The Heat Input Formula, and What It Actually Tells You

Heat input is the energy delivered per millimetre of weld, in kilojoules per millimetre:

Heat input (kJ/mm) = (volts × amps × 60) ÷ (travel speed in mm/min × 1000)

Worked example: 90 A at 12 V, travelling 80 mm/min. That gives (12 × 90 × 60) ÷ (80 × 1000) = 64,800 ÷ 80,000 = 0.81 kJ/mm. Now go to 130 A at 14 V but move at 150 mm/min: (14 × 130 × 60) ÷ 150,000 = 0.73 kJ/mm. More current, less heat in the plate, because speed went up. That is the most useful lesson in this section: amps alone do not define heat input, and a welder who raises the amps and speeds up often produces a cleaner joint than one who lingers at low amps.

For austenitic stainless tube, controlled procedures typically land between roughly 0.5 and 2.0 kJ/mm, with the low end preferred on thin wall. Above that window you buy distortion, carbide precipitation and a wider heat tint band. Below it you risk lack of fusion, which is worse because you cannot see it. The International Molybdenum Association’s fabrication guidance is a solid reference for how heat input and interpass temperature interact with corrosion resistance.

Starting Parameters by Wall Thickness

These are bench starting points for manual TIG on 304/316 tube with argon shielding and a purged root. They are not a procedure.

Wall thickness	Current (A)	Travel (mm/min)	Filler	Target heat input
1.6 mm	55–75	70–100	1.6 mm ER308L/316L	0.4–0.8 kJ/mm
2.0 mm	70–90	70–100	1.6–2.4 mm	0.5–0.9 kJ/mm
3.2 mm	90–120	80–120	2.4 mm	0.8–1.4 kJ/mm
6.0 mm+ (multi-pass)	110–140	90–140	2.4 mm	1.0–2.0 kJ/mm

Short Note: Every number in that table must be qualified by a WPS and proven by a procedure qualification test. Treat it as a direction to travel, not a destination.

Bead Shape, Sequence and Restraint

Use stringer beads or a very narrow weave, kept to about three times the electrode diameter. Anything wider is a slow, hot puddle spreading heat sideways instead of forward. On multi-pass work, let each bead cool slightly before the next.

• Backstep welding: start each short segment slightly ahead of the previous one and weld back toward it. Heat spreads over a longer distance and the start crater lands in already-deposited metal.
• Block sequence: break a long circumferential joint into segments and weld opposite sides in rotation rather than running continuously around.
• Symmetric welding: on two-sided or mirrored joints, alternate sides so shrinkage pulls against itself.
• Chill blocks: clamped copper or aluminium backing pulls heat out fast. It is the same principle that makes aluminum MIG welding manageable on thin sections, and it works just as well on stainless tube.
• Fixtures: rigid fixtures restrain movement, but over-restraint plus high heat input equals residual stress and cracking. Let the fixture guide alignment, not strangle it.

Interpass Temperature

On austenitic stainless, keep interpass temperature below about 150 °C. Above that you accelerate carbide precipitation in the 425–870 °C range and lose dimensional control. Duplex is tighter and grade-specific, so follow the qualified procedure rather than a rule of thumb.

Measure it properly with a contact thermometer or temperature-indicating sticks on the tube a short distance from the weld. Not the back of your glove, and not a hand held near the joint. Gloves are useless as instruments and your hand adapts to heat within seconds.

Root Pass Technique

The root is where tube welding is won or lost. Establish the keyhole, then hold it steady. It should stay a consistent size and shape, not open and close like a pulse. Add filler as the keyhole starts to widen, and time the dip so it bridges rather than blocks. Torch angle around 70–80° from the surface, pushing slightly, arc kept tight.

Fit a gas lens. It gives better shielding coverage and lets you run a larger ceramic cup, which matters where access is restricted and the torch is fighting a fitting or a wall. Root purge flow should be established before the arc starts and held steady during the pass; a surge mid-root can stir the atmosphere and pull oxygen in. Practical purge setup is covered in ESAB’s guidance on purge gas selection.

Walking the cup is worth learning where you have room: it stabilises travel speed, produces consistent tie-in and reduces the stop-start heat spikes of freehand welding. It is a poor fit on tight fittings and small-diameter tube where the torch cannot oscillate.

Arc Strikes, Tungsten and Knowing When to Stop

An arc strike outside the joint is localised melting followed by rapid quench: a hard, corrosion-susceptible spot that is difficult to remove and frequently a rejectable defect. Strike in the groove or on a dedicated run-off tab, never on the pipe surface. Tungsten contamination means stop, break out, regrind and re-establish the puddle from clean metal.

And the hardest discipline on the job: when a weld starts going wrong, stop. Cutting it out costs an hour. Chasing it with more heat, more filler and a wider weave costs a repair weld on sensitised, distorted metal that will fail inspection or fail in service. More heat has never fixed a stainless weld.

Post-Weld Cleaning, Passivation and Inspection

You finish the weld, the colour looks fine under shop lights, and the job ships. A month later a rust bloom shows up along the joint line in a humid plant room. Nothing was wrong with the filler or the amps. The oxide film was left on the surface, and nobody dealt with it.

On stainless, the weld is not finished when the arc stops. The finish is part of the weld quality.

What Heat Tint Really Is

Heat tint is a thickened chromium oxide formed when steel is heated in the presence of oxygen. Straw, brown, blue, purple and black are interference colours set by oxide thickness, not dirt or dye. Underneath, the metal is chromium-depleted: chromium has migrated outward to build the oxide, leaving the layer below short of the element that provides corrosion resistance. A tinted surface is a locally weakened surface. In chloride service it becomes a pitting initiation site, and pits do not stay small in tubing.

Cleaning Methods Compared: Mechanical, Pickling, Passivation, Electropolishing

These four are used interchangeably in conversation. They do different jobs.

Method	What it does	Removes tint?	Where it fits
Mechanical cleaning	Dedicated stainless brushes, grinding, blasting	Yes, back to bright metal	Touch-up, spatter, accessible surfaces
Pickling	Nitric-hydrofluoric acid dissolves oxide and a thin skin of metal	Yes	Weld seams, HAZ, internal runs
Passivation	Nitric or citric acid removes free iron	No	Post-fabrication cleanup
Electropolishing	Electrolytic removal of a controlled metal layer	Yes, and levels the surface	Hygienic, pharmaceutical

Pickling is the heavy hitter. Nitric-hydrofluoric acid dissolves the oxide and takes a thin skin of base metal with it, which is why pickled welds look uniformly grey and corrosion resistance returns. It is also why pickling is hazardous: HF attacks glass and skin, and spent acid is regulated waste.

Passivation is a different operation. Nitric or citric acid dissolves free iron smeared on by tooling, with no meaningful metal removal. Citric has become popular because it is safer to handle and easier to dispose of, and it removes free iron well. But passivation does not remove heat tint. Passivate a tinted weld and you have a clean tinted weld.

Electropolishing, covered by ASTM B912, reverses the process electrochemically: the workpiece is the anode, and a controlled layer of metal is dissolved, leaving a bright, smoothed surface with lower roughness. For acceptance, ASTM A380 covers cleaning, descaling and passivation of stainless parts, including weld discoloration, while ASTM A967 specifies passivation treatments and the tests that verify them.

Practical Pickling and Waste Reality

Pastes suit field work and vertical seams; baths suit small shop parts. The rules are the same:

• Degrease first, or the acid works unevenly.
• Control dwell time: too short leaves tint, too long etches.
• Rinse with clean water, then neutralise and collect the rinse water.

Note: Never pickle near carbon steel fabrications, and keep HF-bearing product away from galvanised surfaces and bare concrete.

Inspection: Matching the Method to the Defect

No single test finds everything. Each has a blind spot, and on tube the geometry is the biggest obstacle.

Visual, Borescope and Colour Acceptance

Visual inspection remains the most informative check, because oxidation colour is a direct proxy for shielding quality. On tube the root is invisible, so a borescope is not optional on critical work. AWS D18.2 grades internal discoloration of austenitic tube from light straw to heavy black oxide; most hygienic work accepts only the lightest one or two levels. Write the acceptance level into the WPS and inspection plan.

Penetrant and Radiography

Dye penetrant finds surface-breaking cracks, pinholes and lack of fusion, but nothing below the skin. Clean first or tint and slag give false indications. RT suits volumetric defects like porosity and tungsten inclusions, though small-bore thin-wall tube makes source-to-film geometry awkward and drops sensitivity. A clean radiograph says nothing about heat tint or sensitisation.

Hydrostatic, Leak and Ferrite Checks

Hydro proves containment, not quality: a joint can hold pressure today and crack in six months. For gas service, helium leak testing finds what hydro misses. Austenitic weld metal normally carries a few percent ferrite to resist solidification cracking, and too much or too little causes problems, especially in duplex work. Handheld ferrite scopes give a quick reading, but the instrument must be calibrated and the reading taken on a prepared, representative surface; root ferrite can differ from the cap.

Documentation and Traceability

A defensible package for a stainless tube run includes:

• Heat numbers for tube and fittings, tied to mill certificates
• Filler classification plus lot number
• Purge records: gas type, oxygen reading, verifier
• Cleaning method, product and rinse confirmation
• Inspection results by location, method and inspector

If a failure happens two years later, these records prove what was done. Without them, every weld on the run is suspect. Our guide to Alloy 20 welding procedures and filler metals follows the same traceability logic on a more specialised alloy, and it carries across to 316L tube work.

Where This Leaves You

Remove the tint properly, know which process does what, and inspect with methods that cover the defects you actually worry about. Then write it down.

Codes, Procedures and Qualification: What the Paperwork Actually Requires

Ask a tube welder what they need before striking an arc and you’ll hear gas, filler and fit-up. Ask a quality manager and you’ll get a folder. Both are right. The procedure records what was proven; the qualification records are the evidence an auditor asks for first.

The WPS and PQR Relationship Under ASME Section IX

ASME BPVC Section IX governs procedure and performance qualification. It doesn’t tell you how to build a pipe; construction codes like ASME B31.3 do, and reference Section IX for qualification.

• PQR (Procedure Qualification Record) — the evidence. A coupon welded to a written specification is tested in tension, bend and, where required, Charpy; the record captures actual amps, volts, travel speed, gas flow, filler lot and results. What happened, not a plan.
• WPS (Welding Procedure Specification) — the plan built from that PQR: ranges for amperage, voltage, travel speed, gas type and flow, filler classification, preheat, interpass limits and cleaning.

Write a WPS first and qualify it later from memory and you have paper with a signature on it.

Procedure Qualification vs. Performance Qualification

Keep them apart: PQR proves the method (once per procedure, owned by the company); WPQ proves the person (coupon tested, record listing process, position, thickness range and expiry).

A welder qualified on 3 mm wall in 6G isn’t cleared for 10 mm or another process, and someone who has never run a purged root can pass a bend test and still ruin a hygienic tube.

Which Essential Variables Change for Stainless

From a carbon steel WPS, these are what break on stainless tube:

• Filler classification — ER308L is not ER70S-2; F and A numbers change.
• Shielding and purge gas — composition, flow, purged root or not.
• Position and thickness — 1G to 6G; ranges come from the coupon.
• PWHT — little for austenitic grades, more rules for duplex.
• Joint design — open root versus consumable insert.

Purge appears twice, as a gas variable and a joint-detail variable. That’s the stainless trap.

The Construction and Product Standards Around It

Section IX is the qualification engine; these govern the finished tube.

Standard	Covers	Where you’ll meet it
ASME B31.3	Process piping design, fabrication, examination	Chemical, pharmaceutical
ASME BPE	Bioprocessing: finish, weld acceptance, documentation	Biopharma tubing
AWS D18.1	Austenitic stainless tube and pipe, sanitary	Food, dairy, beverage
AWS D18.2	Weld interior discoloration levels	Purge acceptance
EN ISO 15614-1	Procedure qualification	European projects
ISO 9606-1	Welder testing, steel	European certification
EN 10217-7	Welded stainless pressure tube	Procurement

ASME B31.3 sets examination percentages, acceptance criteria and Section IX qualification. ASME BPE goes further — surface finish (Ra), weld acceptance, drainability, documentation — and its discoloration limit is often tighter than a piping code. AWS D18.1 covers sanitary austenitic work; AWS D18.2 turns tint into a number.

EN ISO 15614-1 covers procedure qualification and ISO 9606-1 welder testing; EN 10217-7 is the pressure-tube product standard, worth checking against the certificate. For fabrication behaviour, the IMOA austenitic fabrication guidance is a solid reference.

Weld Maps, Travellers and Purge Records in Practice

A weld map is an isometric with every joint numbered (W-001, W-002) so a weld, welder, filler lot and inspection result trace to a plant location. A traveller moves with the spool: joint numbers, WPS number, welder ID and stamp, filler lot, purge gas and oxygen reading, sign-offs at each stage.

Purge logs matter because purge quality is invisible afterwards. “Purged” is worthless; “0.004% O₂, 8 minutes pre-purge, 12 L/min during weld, analyser calibrated 14/03” is evidence. Same discipline as TIG technique and controlled procedures, different acceptance criteria.

When the Client Specification Is Stricter Than the Code

The stricter requirement governs. If B31.3 allows a tint your client’s spec forbids, weld to the spec — the code is a floor, not a ceiling.

• Read the spec before qualifying; it may demand tighter oxygen limits, filler or inspection.
• Get written clarification on conflicts — verbal reassurance is not a deviation approval.
• Qualify to the stricter requirement and document any deviation.

Documents to See Before You Strike an Arc

Before the torch is lit, you should be able to lay hands on:

1. The approved WPS for the joint, material and thickness, plus the PQR behind it.
2. Your current qualification matching process and position.
3. The weld map with your joint number, and the traveller.
4. Filler certificate and lot, tube certificate and heat numbers.
5. The purge procedure and analyser calibration record.
6. Acceptance criteria: D18.2 level or client tint limit.
7. Safety paperwork: JSA, hot work permit, confined space entry.

If any is missing, stop and ask. Refusing to weld without a WPS isn’t being difficult — it protects the client, the company and you, because the first question after a failure is “what procedure were you working to?” For how pathways differ by process, see our aerospace welding standards and careers guide.

Safety: Hexavalent Chromium, Fume and Confined Spaces

Stainless steel is often described as a clean material to weld. The bead is brighter, the spatter is lower, and there is no flux cloud to sweep away. That “clean” label refers to the finish, not the air. Welding stainless creates a complex mixture of metal fume, ozone, nitrogen oxides and intense ultraviolet radiation, and the purge gas protecting the root can displace oxygen without warning. On tube work, those hazards are often concentrated in pits, racks and vessel interiors where ventilation is poor.

The Fume Is the Main Exposure Risk

Stainless consumables contain chromium and nickel. Under the arc, a fraction of the chromium becomes hexavalent chromium, or Cr(VI), a carcinogen with an unusually tight exposure limit. OSHA’s permissible exposure limit is 5 µg/m³ as an 8-hour time-weighted average, while OSHA’s hexavalent chromium guidance also requires employers to use feasible engineering controls before relying on respirators. NIOSH goes further, recommending 0.2 µg/m³ as an 8-hour TWA, which is 25 times stricter than the OSHA PEL.

Manganese is another concern, particularly with consumables that contain it. NIOSH’s welding safety guidance recommends a manganese REL of 1 mg/m³ as a TWA with a 3 mg/m³ short-term limit, whereas OSHA’s manganese PEL is 5 mg/m³ as a ceiling. NIOSH also recommends keeping total welding fume exposure at or below 5 mg/m³ as an 8-hour TWA. Those numbers are difficult to meet in a small fabrication bay with natural draft and an open roller door.

TIG Is Not Automatically Low-Risk

TIG welding produces ozone and nitrogen oxides, and its UV output is intense enough to burn nearby skin and eyes even when the welder is fully covered. The arc can affect helpers, fitters and anyone working behind a screen that does not block UV. OSHA’s welding fact sheet covers fume, radiation and ventilation together because these hazards arrive as a package. If your shop is adding laser or hybrid tube welding, the hazard profile changes again; this Alpha laser welding safety guide is a useful starting point for the additional optical and fume controls.

Purge Gas Can Kill Without Warning

Argon and nitrogen are simple asphyxiants. They have no smell and no colour, so a purge discharge leaking into a pit, tank or small weld bay can lower oxygen without anyone noticing. A welder who leans into a vessel to inspect a root may be breathing an invisible atmosphere that will cause collapse within seconds. Treat every purge outlet as a confined-space hazard. Vent discharge to a safe outdoor location, test the atmosphere before entry, use a permit and an attendant, and keep retrieval equipment at the opening. Never enter a space to look for a suspected leak.

Controls in Order of Effectiveness

Start with engineering controls, not respiratory protection. Local exhaust ventilation at the arc, fume extraction torches, fixed fume arms and ventilated weld booths remove contamination before it reaches the breathing zone. In confined spaces, mechanical ventilation and continuous air monitoring are essential.

• Engineering: capture fume at the source, ventilate enclosed areas, isolate purge gas discharges.
• Administrative: exposure monitoring, hot-work and confined-space permits, training, restricted access.
• PPE: supplied-air respirators or powered air-purifying respirators for high-exposure tasks; UV-rated screens for helpers.

A nuisance dust mask is not adequate for Cr(VI) fume, ozone or nitrogen oxides. It seals poorly, captures particles inefficiently and does nothing for gases. Air monitoring tells you whether controls are working; a visual check of the haze does not.

Pre-Job Safety Checklist

• Is local exhaust positioned at the arc, not across the room?
• Has the confined space been tested for oxygen and purge-gas accumulation?
• Is the purge discharge routed outdoors, away from intakes and pits?
• Are UV screens protecting nearby workers?
• Is the correct respirator selected, fit-tested and available?
• Has exposure monitoring been done for Cr(VI), manganese and total fume?

The best stainless welders are not the ones who tolerate smoke best. They are the ones who arrange the job so there is nothing to tolerate.

Troubleshooting, Common Mistakes and the Cost of Rework

Every defect on a stainless tube weld traces back to one of three things: contamination, atmosphere, or heat. The trick is reading the symptom correctly, because the fix for a rust bloom is not the fix for a centreline crack, and treating one as the other is how a small problem becomes a cut-out.

Symptom, Cause and Corrective Action

Symptom	Likely cause	Corrective action
Sugaring (dark, granular oxide inside the root)	Oxygen in the purge, purge lost mid-pass, insufficient pre-purge	Heavy sugaring typically means cut-out and re-weld, since the oxide cannot be reliably removed from the inside of an installed joint
Heat tint beyond acceptance	Excess heat input, poor gas coverage, wrong cup size	Remove by pickling or mechanical means, then re-inspect against the specified colour standard
Porosity	Damp filler, dirty joint faces, gas flow turbulence, drafts	Clean, dry, re-establish shielding, cut out if clustered or exceeding acceptance
Incomplete penetration	Root gap too tight, land too thick, travel too fast, low current	Re-fit; never chase penetration by dumping in heat
Lack of fusion	Insufficient energy, contaminated faces, wrong torch angle	Cut out at the affected length; grind-out risks thinning the wall
Centreline solidification cracks	Low ferrite content, high heat input, high sulphur or phosphorus, poor bead shape	Change filler classification for more ferrite, reduce heat input, remelt and correct bead profile — see the IMOA guide to fabricating austenitic stainless steels
Crater cracks	Arc terminated without filling the crater or without slope-out	Use proper slope-out and taper; grind and re-weld the crater
Arc strikes	Striking outside the joint, poor ground placement	Grind out and inspect; arc strikes are corrosion initiation sites
Distortion	Uneven sequence, excessive heat, weak fixturing	Re-sequence, chill blocks, mechanical correction only where permitted
Weld decay / knife-line attack	Sensitization from prolonged time in the 425–870°C band	Replace section with low-carbon or stabilized grade; review heat input
Rust blooms within days to weeks	Free iron contamination from shared tooling or carbon steel contact	Re-clean, pickle and passivate per ASTM A380

The Ten Mistakes That Cause Most of It

• Sharing carbon steel tooling. Wire brushes, grinding discs and files transfer free iron in seconds.
• Skipping the purge check. A regulator gauge tells you flow, not oxygen.
• Welding too hot and too slow. Stainless has nowhere to dump the heat, so it stays in the joint.
• Over-weaving. Wide, wandering beads widen the hot zone and drag in the heat-affected area.
• Bad fit-up. Gaps and mismatch get “fixed” with current, which is where penetration problems are born.
• Contaminated filler. Filler left out, dropped, or handled with dirty gloves introduces carbon and moisture.
• Wrong filler match. Matching the base metal by name instead of by service environment — relevant for exotic alloys too, as covered in Alloy 20 Welding: Procedures, Filler Metals, and Best Practices.
• No post-weld cleaning. Tint left in place is a corrosion cell waiting for moisture.
• Welding without a qualified WPS. You cannot repeat a result you never defined.
• Ignoring interpass temperature. Stacked passes push sensitization without anyone noticing until the metallurgy has changed.

What Rework Actually Costs

A repair on a hygienic tube joint often costs several times the original weld once cutting, re-purging, re-inspection and documentation are counted. Add schedule knock-on and it climbs further. Then consider a hydro test failure after final finishing: stripping insulation, cutting out welds over again and repeating the entire inspection cycle, with the client watching. A corrosion failure in service is a different order of magnitude — product loss, downtime, and a very uncomfortable investigation into your purge records.

Pre-Weld Checklist

1. Grade and filler verified against the WPS and heat numbers.
2. Dedicated stainless tools, clean gloves, acetone-wiped faces.
3. Fit-up: gap, land, mismatch, tack condition all checked.
4. Purge verified with an analyser, not a gauge.
5. Parameter sheet at the bench, interpass measured not guessed.
6. Post-weld cleaning method agreed before the arc starts.

What to Do Differently Next Time

That leaking run at the start of this guide was not a welding skill failure. It was a purge check nobody insisted on. Stainless tube welding rewards the boring disciplines: clean metal, verified atmosphere, controlled heat, and a colour standard you actually compare against. Everything else is improvisation, and improvisation on stainless shows up six weeks later — from the inside.

Frequently Asked Questions

Can you TIG weld stainless steel tube without purging?

You can strike an arc, but you cannot produce a sound, corrosion-resistant root. Without purge, the backside oxidises into heat tint and eventually sugaring, which destroys the passive layer and is very difficult to remove from an installed tube. Purging is mandatory for any joint where the interior will be in service or must meet a hygienic specification.

What oxygen level should I purge to for stainless tube welding?

A widely used industry maximum is about 0.1% oxygen (1000 ppm), but many critical specifications call for below 100 ppm, and some hygienic or semiconductor work goes lower. The right target comes from the applicable code or client specification, and it must be measured with an oxygen analyser, not estimated.

Do I need to remove heat tint after welding stainless steel?

Yes, wherever corrosion resistance or cleanliness matters. Heat tint is a thickened, chromium-depleted oxide, so the metal beneath it is less resistant to pitting. Pickling dissolves the tint, passivation removes free iron, and mechanical methods work only if they do not smear or contaminate the surface.

What filler metal should I use for 316L stainless tube?

Use ER316L for matching 316L tube, and ER316L or ER309L when joining 316L to carbon steel depending on the procedure. Never substitute carbon steel filler, and keep stainless filler segregated and clearly labelled to avoid mix-ups on site.

Why does my stainless weld crack down the middle?

Centreline solidification cracking usually points to low delta ferrite in the weld metal, excessive heat input, a deep narrow bead shape, or high sulphur and phosphorus in the base metal. Using the correct filler, controlling heat input and avoiding a concave, deeply penetrating bead profile all help.

Is stainless steel harder to weld than carbon steel?

It is different rather than simply harder. Stainless has lower thermal conductivity and higher expansion, so heat concentrates and distortion is greater, and the backside oxide problem does not exist with carbon steel. Once you control purge, fit-up and heat input, stainless is very weldable.

Can I use a carbon steel wire brush on stainless?

No. Carbon steel brushes, grinding discs and files embed free iron that rusts and can start pitting. Keep dedicated stainless-only tools, store them separately, and never use a disc that has touched carbon steel.

What is the difference between pickling and passivation?

Pickling uses an acid mixture, typically nitric-hydrofluoric, to dissolve heat tint and a thin layer of base metal. Passivation uses nitric or citric acid to remove free iron contamination without meaningful metal removal, so it cleans but does not remove heavy tint.

Do stainless tube welds need to be radiographed?

Only if the applicable code, client specification or service demand requires it. Radiography is good at finding volumetric defects like porosity and incomplete penetration, but it is poor at tight lack-of-fusion and does not assess heat tint, which is why borescope inspection and colour criteria matter for tube interiors.