Short answer: To stop sugaring and heat tint on stainless, back purge the inside of the joint with argon so the root never sees oxygen. Dam the purge volume down, feed gas low and vent high, and wait for an oxygen analyzer to read your target before striking an arc: commonly under 50 ppm for sanitary tube and under 100 ppm for general process piping. Then cut the flow back while welding.
Back purging means filling the inside of a stainless pipe, tube or vessel with inert gas, usually argon, before and during welding so the root side of the weld never sees oxygen. Skip it and the inside of the joint oxidizes into a crusty gray-black growth that welders call sugar. Do it badly and you get heat tint instead: a band of straw, blue or purple oxide that looks harmless and is not.
This article covers why both happen, how to set up a purge that holds, what oxygen levels to aim for, and how to clean up tint you could not avoid.
What sugaring and heat tint actually are
Austenitic stainless such as 304L and 316L resists corrosion because of a thin, chromium-rich passive film on its surface. Welding wrecks that film locally. If the hot metal sees air while it is above roughly 800 to 1,000°F, chromium and iron react with oxygen and grow a much thicker oxide layer. The metal directly under that oxide gives up some of its chromium to make it. That chromium-depleted skin corrodes faster than the base metal around it.
Sugaring
Sugaring, also called sugar or granulation, is the extreme case. The root side of the weld is open to air, and the bead comes out rough and granular, nearly black, like burnt sugar on the bottom of a pan. It is heavily oxidized, often porous, and a reject on essentially any pressure, sanitary or code job. You cannot brush it off. It gets ground out and rewelded.
Heat tint
Heat tint is the milder version. It is a colored oxide on and beside the weld, running from pale straw through gold, brown, blue and purple to dull gray as the oxide thickens. The color tells you roughly how much oxygen was there while the metal was hot. Pale straw usually passes. Blue and darker usually does not on food, pharma or chemical service, because crevice corrosion, pitting and microbial attack like to start in that oxidized band.
Why the inside of the weld matters more than the outside
On stainless pipe or a tank, the product touches the inside. That is the surface that has to stand up to chlorides, hot caustic, acid washes and sanitizers. You can see the outside cap and clean it in a few minutes. The root side you usually cannot reach once the joint is closed, so whatever you leave inside the pipe is what goes into service.
Torch gas protects the top of the puddle and nothing else. On thin-wall tube, say 0.065 inch, the arc penetrates all the way through and the back of the weld runs about as hot as the front. With no purge gas behind it, that back side oxidizes the moment it forms. On heavier pipe with an open root, the root bead hangs out into the bore and sees the same thing.
Purge oxygen levels and the color they produce
Air is about 20.9% oxygen, or roughly 209,000 parts per million. A good purge takes that down by four orders of magnitude. The usual reference for what that looks like is the AWS D18.1 and D18.2 color chart, which shows 316L tube welded with progressively more oxygen in the purge. The values below are approximate and follow that chart. Treat them as a guide, and use whatever acceptance limit your job specification actually sets.
| Oxygen in purge | Approximate root color | Where that level is commonly targeted |
|---|---|---|
| About 10 ppm | No visible tint | Semiconductor and ultra-high-purity gas lines |
| About 25 ppm | Very faint straw | Pharma and biotech tubing |
| About 50 ppm | Light straw | A common ceiling for dairy, food and beverage tube |
| About 100 ppm | Straw | General stainless process piping, where the spec allows |
| About 200 to 500 ppm | Gold to brown, blue starting | Usually needs cleaning, often rejected on sanitary work |
| About 1,000 to 5,000 ppm | Blue to purple | Reject on most corrosion-critical service |
| Above about 12,500 ppm | Gray, moving toward sugar | Reject |
Two things stand out. First, the jump from acceptable to bad happens at levels a welder cannot judge by eye or by how long the gas has been running. You need an oxygen analyzer at the vent that reads in ppm. Second, chart color depends on heat input too. A hot, slow root at 50 ppm can tint darker than a fast one at 100 ppm, so the purge reading and the welding technique both have to be right.
How to set up a back purge that holds
Most purge problems come from setup, not from the welder’s hands. This is the sequence we run on stainless pipe and tube:
- Clean the joint inside and out. Acetone on a lint-free rag about an inch back from the bevel. Marker, cutting fluid and tape residue all burn into the root. Use stainless-only brushes and flap discs that have never touched carbon steel.
- Fit it up tight. Square, deburred ends. On thin tube the gap is zero. On pipe, use the root opening your WPS calls for, commonly 1/16 to 1/8 inch, and keep hi-lo to a minimum.
- Dam the volume down. Put purge dams a few inches either side of the joint so you are purging a small chamber instead of 40 feet of line. Water-soluble paper dams work on spools that get flushed. Inflatable bladder dams on a tether work on bigger pipe and existing lines.
- Tape the gap. Wrap the open root with aluminum foil tape or purge tape and peel it back a couple of inches at a time as you weld.
- Gas in low, vent high. Argon is about 1.4 times as dense as air, so feed it in at the bottom and vent at the top. Make the vent at least as large as the inlet.
- Give it time, then measure. Plan on at least five to ten volume changes, then confirm with the analyzer before striking an arc.
- Turn the flow down to weld. High purge flow pressurizes the pipe and pushes the root concave, or blows the puddle out on the final tie-in. Drop the flow once you hit your number, and make sure the gas still has a way out when the last inch of tape comes off.
- Keep the purge on through the hot pass. On thin wall, the second pass reheats the root enough to tint it again. Hold the purge until the root side is back below about 800°F.
A quick volume example
Take 4 inch schedule 10 stainless pipe. The bore is about 4.26 inches, which works out to roughly 0.1 cubic feet per foot of length. With dams 12 inches either side of the joint, you are purging about 0.2 cubic feet. Ten volume changes is 2 cubic feet, so at 20 CFH you are looking at around six minutes before the meter starts reading low. Try the same joint with no dams on a 30 foot run and it is closer to an hour and a half, and a lot of argon.
Choosing the purge gas
Welding-grade argon is the default. Ultra-high-purity argon, 99.999% or better, goes on semiconductor and some BPE work, often run through stainless or specially rated purge lines so the hose itself does not let moisture and oxygen in. Nitrogen is cheaper and some specs allow it on austenitic grades. Nitrogen-hydrogen forming gas can give a bright root on 300-series stainless, but keep hydrogen away from ferritic, martensitic and duplex grades, where it can cause cracking. Duplex procedures sometimes add nitrogen to the purge on purpose to protect the ferrite-austenite balance at the root.
Protecting the outside of the weld
Heat tint on the cap side is easier to fix, but it is better not to make it. A gas lens and a bigger cup, #10 or #12 on pipe, spread the argon into a calm column instead of a jet. Keep post-flow long enough to cover the tungsten and the crater as they cool, usually 8 to 15 seconds depending on amperage. Keep interpass temperature down too. Many stainless procedures cap it around 350°F, and running hot is the fastest way to turn a straw bead blue.
On long seams in plate, or wherever the heat input is high, a trailing shield behind the torch keeps argon over the cooling bead. A properly shielded stainless TIG cap should come off straw to light gold. If yours are coming off blue on every pass, look at amperage, travel speed and gas coverage before blaming the material.
When you cannot purge
Some joints cannot hold a purge: a repair on a vessel with no access, or a weld deep inside equipment. Backing fluxes painted on the root side can stand in on some carbon and stainless joints, but the residue has to come off afterward and many sanitary specs will not accept them. Grooved backing bars with gas fed through them work on plate. If the joint matters and none of those options fit, the honest answer may be to cut a different joint location that can be purged.
Cleaning heat tint after welding
Where you can reach the tint, it can be removed. The goal is to take off both the colored oxide and the chromium-depleted layer under it, then let the passive film reform.
- Mechanical cleaning. Stainless wire brushes, non-woven abrasive pads or light grinding. Fast on caps, but it smears rather than removes on heavy tint, and any carbon steel contamination will show up later as rust spots.
- Pickling paste. Usually a nitric and hydrofluoric acid mix. It removes oxide and the depleted layer. It is hazardous, needs proper PPE and neutralizing, and has to be rinsed completely.
- Electrochemical weld cleaning. A wand with a mild electrolyte, usually phosphoric-based, that strips tint as you pass over it. Clean and quick on visible welds.
- Passivation. Nitric or citric acid treatment to ASTM A967, often after cleaning to ASTM A380. Passivation restores the passive film but does not remove heavy heat tint on its own.
Inside a closed tube there is usually no way to reach the root, apart from circulating chemical passivation through the whole system. That is the real reason purge quality matters. Prevention is the only method that works everywhere.
Common purge mistakes we see on repair work
A lot of cracked or pitted stainless that comes in for repair traces back to a weld that was never purged properly. The usual causes:
- A hose stuck in the open end of a pipe with no dam. Argon rolls out, air rolls back in.
- Too much flow, so the root goes concave or a pinhole blows out on the tie-in.
- No vent, or a vent smaller than the inlet.
- Striking the arc on a timer instead of an oxygen reading.
- Leaks at branch connections, thermowell taps or loose hose fittings pulling air into the purge.
- Carbon steel tools, clamps or grinding discs on stainless, which embeds iron and rusts later.
- Pulling the purge right after the root on thin wall, so the hot pass re-tints it.
Most of these cost nothing to fix except time and attention.
Purging in the shop and in the field
In the shop, a spool on a rotator with dams at both ends is about as easy as purging gets. The welder stays in one position, the purge chamber is small, and the analyzer sits right at the vent. That is a big part of why we prefer to build stainless spools in the shop and keep field welds to closures and tie-ins.
Field tie-ins on existing lines are harder. You may have to dam inside a pipe you cannot see into, feed gas through a drain or instrument tap, and deal with product residue upstream. It is still doable with inflatable dams on a retrieval line and some patience. That kind of tie-in is what our mobile TIG welding service is set up for.
If your job is food, dairy or pharma tube, read our article on 3-A and ASME BPE sanitary tube welding, which covers the acceptance side in more detail. For general work, see our stainless steel TIG welding and pipe and tube TIG welding services. Tell us the alloy, sizes, the acceptance limit your spec sets and where the welds are, and we will plan the purge with you.
Frequently asked questions
Sugaring happens when the root side of a stainless weld is exposed to air while it is molten and hot. Chromium and iron oxidize heavily and form a rough, black, granular surface. It is caused by no back purge, a purge that leaks, or pulling the purge too early. Sugared welds have poor corrosion resistance and need to be ground out and rewelded.
It depends on the spec. Common targets are under about 100 ppm for general stainless process piping, under about 50 ppm for food, dairy and beverage tube, around 25 ppm or lower for pharma, and near 10 ppm for semiconductor high-purity lines. Air is about 209,000 ppm oxygen, so use a ppm-reading analyzer at the vent rather than guessing.
Sometimes. Nitrogen is allowed by some specifications on 300-series austenitic stainless and costs less than argon. It is not suitable for every alloy or code, and nitrogen-hydrogen forming gas should be kept off ferritic, martensitic and duplex grades because hydrogen can cause cracking. Check the WPS and the project spec before switching gases.
Not heavy heat tint. Passivation with nitric or citric acid under ASTM A967 restores the passive film, but it does not reliably remove colored oxide or the chromium-depleted layer under it. Pickling paste, electrochemical cleaning or mechanical cleaning removes the tint first. Passivation is usually the final step after the tint is gone.
Long enough for at least five to ten volume changes and a confirmed low oxygen reading. With purge dams 12 inches either side of a joint on 4 inch pipe, that is often around five to ten minutes at 20 CFH. Without dams on a long run, it can take an hour or more. Time alone is not proof; measure at the vent.