Weld Cleaning & Prep:
Removing Heat Tint
That rainbow around the weld isn't a stain. It's chromium that used to be in your steel, and the metal it came out of is now measurably worse at resisting corrosion. Polish the color off and you can leave the problem behind — shiny, invisible, and shipping. Here's what's actually happening, and how to fix it properly.
Weld cleaning is two jobs that share a name, and shops routinely do the second one well and the first one not at all.
Before the arc: get the joint clean, or the contamination you left there ends up in the weld as porosity, inclusions, or lack of fusion. After the arc: remove the spatter, the slag, and the heat tint — and the tint is where the real technical content lives, because removing the color and removing the problem are two different operations.
That distinction is the whole article. Everything else is procedure.
Prep: The Order Matters
Most shops know they should clean before welding. Fewer know that doing the steps in the wrong order makes the joint dirtier than not cleaning it at all.
Degrease First. Abrade Second. Always.
Hit an oily surface with an abrasive and you don't remove the oil — you drive it into the surface, work it into the grain structure, and contaminate the abrasive so it carries oil to the next joint. Solvent comes off with solvent. Once the joint is dry and clean, then the abrasive has something useful to do.
Grinding a greasy joint is the single most common prep mistake in fabrication, and it looks exactly like doing the job right.
What has to go, and why
| Contaminant | What It Does to the Weld | How It Comes Off |
|---|---|---|
| Oil, grease, cutting fluid | Hydrogen source. Porosity, and on hardenable steels, cracking. | Solvent or alkaline degrease. First, before any abrasive touches it. |
| Moisture / condensation | Hydrogen source. Porosity. Worst on aluminum and hardenable steels. | Dry the joint. Watch parts brought in from cold — they'll condense in a warm shop. |
| Mill scale | Iron oxide with a melting point above the base metal. It doesn't melt out — it gets trapped as an inclusion, and it blocks fusion. | Abrasive. Flap disc, fiber disc, or non-woven depending on how heavy it is. |
| Rust | Oxide plus trapped moisture. Porosity and inclusions. | Abrasive down to bright metal. Pitted rust goes deeper than it looks. |
| Paint, primer, coating | Burns off into the arc. Porosity, contamination, and fumes you don't want to breathe. | Abrasive or chemical stripper. Check the coating's SDS — some produce genuinely hazardous decomposition products. |
| Marker, crayon, layout ink | Often overlooked. Many contain chlorides or sulfur — both bad news in a weld. | Solvent. Use welder-safe layout markers on stainless in the first place. |
| Aluminum oxide | Melts far hotter than the aluminum under it. Doesn't melt out — traps as inclusions and blocks fusion. | Dedicated stainless wire brush, immediately before welding. It reforms in minutes — clean it and weld it, don't clean it and go to lunch. |
| Previous shop's iron | On stainless: embedded free iron rusts in service, weeks after delivery. | Dedicated stainless-only abrasives. This is a housekeeping discipline, not a product spec. |
↔ Scroll table horizontally on mobile
Chlorinated Solvents and Stainless Don't Mix
Chlorinated degreasers leave chloride residue, and chlorides are the classic trigger for stress corrosion cracking in austenitic stainless — particularly under heat and stress, which is precisely what a weld provides. Use a solvent rated for stainless, confirm it against the SDS, and make sure it fully evaporates before the arc strikes.
Aluminum's Clock
Aluminum oxide reforms on a freshly cleaned surface almost immediately — meaningfully within minutes. That changes the workflow: on aluminum, cleaning isn't a prep stage you do to a batch in the morning, it's the step immediately before the arc. Clean the joint, weld the joint. Use a stainless brush dedicated to aluminum and nothing else, and keep it clean.
Reading the Heat Tint Spectrum
The colors around a stainless weld aren't decoration and they aren't random. They're a readout — and they run in a fixed order, every time, because they're produced by a fixed physical process.
Colours run in this order, always. Colour is a proxy for oxide thickness, not a thermometer — it reflects time, temperature, and oxygen availability together. Darker means thicker oxide, more chromium pulled out of the metal beneath, and more material you have to remove to fix it. Colours shown are indicative; judge against the standard your work is held to.
For Tube and Sanitary Work, There's a Standard
If you're welding stainless tube to a sanitary or high-purity spec, discoloration isn't judged by eye against a blog graphic. AWS D18.2 is the guide to weld discoloration levels on the inside of austenitic stainless tube, and it exists precisely because "how blue is too blue" needed an answer that two parties could agree on. If your work is governed by it, that document is the authority — not this one. Get the current edition.
The Colour Isn't the Problem. The Metal Under It Is.
Here's what's physically happening, because it explains everything else.
Stainless resists corrosion because chromium in the alloy reacts with oxygen to form a thin, tight, self-healing chromium oxide film at the surface. That film is what "stainless" means. It's a few atoms thick and it's invisible.
Heat the metal in the presence of oxygen and that film grows. As it thickens, it starts interfering with light — the same thin-film interference that makes an oil slick rainbow on a puddle. Different thickness, different colour. That's why the sequence is fixed: straw, gold, bronze, brown, purple, blue, black. You're not watching the metal change colour. You're watching an oxide layer get thicker.
The chromium in that thickened oxide came from somewhere. It came out of the metal immediately beneath it — leaving a chromium-depleted layer that is measurably less corrosion resistant than the alloy around it.
So the tint band is two things stacked: a visible oxide on top, and an invisible impoverished layer underneath. And that leads directly to the failure mode nobody talks about:
You Can Remove the Colour and Leave the Problem
Buff the tint lightly and the colour disappears. The part looks finished. But if you only took the oxide and not the depleted metal under it, you've removed the evidence and kept the defect — a bright, clean-looking weld zone with reduced corrosion resistance, heading to a customer who will find out in service.
This is the weld-cleaning equivalent of painting over rust. The tint was doing you a favour by being visible.
Which raises the obvious question: how much do you take off? Honest answer — it depends on the service environment and the spec, and it's not something to eyeball. A decorative handrail and a pharmaceutical vessel have wildly different requirements from the same tint band. Where the part faces a corrosive environment, the governing specification will tell you what's required, and it's usually written in terms of removing the heat-affected material, not just the colour. On non-critical work, cosmetic removal is genuinely fine. Know which one you're doing.
Carbon Steel: Same Colours, Different Meaning
Carbon steel tempers to similar colours, but there's no chromium and no passive layer, so there's no depletion issue. The tint is cosmetic and an indicator of heat input — which on a hardenable steel might mean you've affected the temper, a separate concern. Clean it for appearance and for coating adhesion, but the corrosion argument in this section is a stainless problem.
Three Ways to Remove It
Mechanical, chemical, electrochemical. Each one is right for some work and wrong for other work. Here's the honest comparison, including where abrasives aren't the answer.
| Method | How It Works | Strengths | Limits & Cautions |
|---|---|---|---|
| Mechanical Abrasive |
Non-woven discs, convolute wheels, flap discs physically cut the oxide and the metal under it away. | No chemicals, no permits, no neutralisation. Removes oxide and depleted layer in one operation. Controls the final finish. Works on any geometry you can reach. | Changes surface finish — you have to blend it back. Risk of contamination from shared abrasives. Operator-dependent. Requires dust control. |
| Chemical Pickling |
Acid paste, gel, or bath dissolves the oxide and the chromium-depleted layer. | Reaches everything the paste touches, including complex internal geometry. No mechanical skill. Removes depleted metal reliably. | Typically uses hydrofluoric acid. Genuinely dangerous — see below. Leaves a matte finish. Neutralisation, rinse water, and disposal are regulated. Often a specialist operation. |
| Electrochemical Weld cleaner |
Electrolyte plus current at a brush or carbon head lifts the oxide electrolytically. | Fast on accessible welds, tidy, far less hazardous than HF pickling, minimal finish change. Popular for stainless fabrication shops. | Line of sight only — no internal or blind geometry. Equipment cost. How much depleted metal it removes depends on the process and electrolyte, so confirm against your spec rather than assuming. |
↔ Scroll table horizontally on mobile
Hydrofluoric Acid Is Not Ordinary Shop Chemistry
Most stainless pickling products contain hydrofluoric acid, and HF is in a different category from other shop acids. It penetrates skin and causes systemic fluoride poisoning — burns can be severe with little immediate pain, and symptoms can be delayed for hours while the damage progresses. Exposures involving relatively small skin areas have been fatal.
This is not a "wear gloves" caution. It requires specific PPE, specific training, calcium gluconate on hand, an emergency plan, and disposal that satisfies your local regulations. If your shop isn't already set up for it, that's an argument for a different method — not an argument for being careful. Read the SDS and mean it.
Which One, Honestly
For most fabrication — accessible welds, no hygienic spec, finish matters — mechanical is the practical answer, and it's the only one of the three that also fixes the surface finish while it's there. For complex internal geometry or where a spec demands verified removal of the depleted layer, pickling or passivation by people equipped for it. For high-volume accessible stainless weld cleaning where finish change is unwelcome, electrochemical earns its equipment cost fast.
We make abrasives. We'd still tell you to pickle a vessel interior.
The Mechanical Method, Step by Step
Run only the steps your part needs, in this order.
-
Knock down spatter and excess
Flap disc · ceramic · 60–80 gritIf there's spatter, slag, or bead standing proud, it goes first. Ceramic on stainless — it cuts cool and self-fractures instead of glazing. Keep the tool moving; a stationary disc concentrates heat exactly where the alloy can't conduct it away, and you'll add tint while removing tint.
-
Take the tint and the metal under it
Non-woven · Medium grade · aluminium oxideThis is the operation that matters. Medium non-woven cuts the oxide and the depleted layer without gouging the base metal or undercutting the weld toe. Work the full width of the tint band, not just the darkest part — the pale straw at the edges is still oxide, and it's still depletion. On light tint, Medium may be more than you need; test on scrap.
-
Blend to uniform
Non-woven · Very Fine · straight-line passesStep to Very Fine and blend the cleaned zone into the surrounding panel. Match the grain direction of the parent surface or the repair announces itself under any light at any angle. Straight-line passes, consistent pressure.
-
Deburr and break edges if required
Convolute wheel · density 8 · aluminium oxide · MediumIf the weldment has machined or cut edges needing a break, bring it to the bench. A convolute wheel conforms to the part and cuts the burr rather than folding it over — which matters more on stainless than on anything else you weld.
-
Clean, inspect, and passivate if the spec says so
Per applicable specificationRemove abrasive debris and inspect under good light from multiple angles — tint hides at low angles. The passive layer rebuilds on its own in air, but embedded contamination prevents it. If your customer's print or industry calls for passivation, that's a specified chemical process with its own standards. Follow the spec you're held to.
The Contamination Rule, Again
Never run an abrasive on stainless that has touched carbon steel. The open web of a non-woven product is very good at trapping iron particles and carrying them to the next part, where they embed and rust — often weeks after delivery, at your customer. Dedicate stainless abrasives, store them separately, label them. No product on the market prevents this for you.
Dust, Fume, and RPM
Stainless grinding dust contains chromium and nickel, both of which carry occupational exposure requirements — use extraction and work to your shop's respiratory programme. Welding fume is its own hazard with its own controls. And check the maximum safe speed on any non-woven product against your tool's rated speed before mounting: the web is not a bonded wheel and does not tolerate overspeed. Guards on, eye and face protection, every time.
The Cheapest Tint Removal Is Not Making Tint
Every minute spent cleaning tint is a minute paying for a problem the welding process created. Tint forms when hot metal meets oxygen — so the levers are heat and oxygen, and both are controllable.
- Purge properly. On tube and pipe, back purging displaces oxygen from the root side. Lower oxygen in the purge means less tint. This is the single biggest lever on internal discoloration, and it's why sanitary work specifies purge quality rather than hoping.
- Cover the topside. Adequate shielding gas coverage and flow — and a trailing shield where the geometry allows — keep oxygen off metal that's still hot enough to react.
- Manage heat input. Lower heat input and faster travel mean less time in the temperature range where oxide grows. Stringer beads over wide weaves, and let interpass temperatures come down.
- Don't stop and admire it. Metal cools through the tint-forming range whether you're watching or not. Post-flow matters — pulling the torch away early exposes hot metal to air at exactly the wrong moment.
The Economics Nobody Runs
Weld cleaning is almost always accounted for as consumables and shop time, which makes it invisible as a process cost. Run it the other way: if better purge and gas coverage cut your tint band in half, you've cut the cleaning operation roughly in half across every weld on every job. Compare that against the cost of the gas. On production work the arithmetic is usually not close.
Six Ways Weld Cleaning Goes Wrong
- Grinding before degreasing. You don't remove oil with an abrasive — you drive it in and contaminate the disc. Solvent first, always.
- Buffing the colour off and calling it done. The colour was the indicator, not the defect. On corrosion-critical work, the depleted metal underneath has to go too.
- Cleaning only the dark part of the band. Pale straw at the edges is thinner oxide, but it's still oxide and still depletion. Work the whole affected zone.
- Using a carbon steel wire brush on stainless. Guaranteed free iron contamination. The part will rust and it won't be the alloy's fault.
- Cleaning aluminium and then walking away. The oxide is back before you get to the torch. Clean it and weld it.
- Adding tint while removing tint. Leaning on a disc and holding it in one place cooks the surface. Stainless can't conduct the heat away — that's the whole problem with stainless — so you generate exactly what you're there to remove.
Answers, Short Version
What causes the rainbow colours around a stainless weld?
Heat plus oxygen thickens the chromium oxide film on the surface. As the film gets thicker it produces optical interference — the same effect that makes an oil slick rainbow. Different thickness, different colour, which is why the sequence always runs straw, gold, bronze, brown, purple, blue, black in that order.
Does heat tint actually matter, or is it just cosmetic?
Depends entirely on service environment. The chromium that formed the visible oxide came out of the metal directly beneath it, leaving a chromium-depleted layer with reduced corrosion resistance. On a decorative part in a dry environment, that's cosmetic. On anything in a corrosive service environment, it's functional — and the governing spec will say so.
Can I just polish the colour off?
You can, and that's exactly the trap. Removing the oxide without removing the depleted metal under it gives you a bright weld zone that still has reduced corrosion resistance — you've removed the evidence and kept the defect. Fine on non-critical work. A real problem on anything that has to resist corrosion.
What grade non-woven removes heat tint?
Start at Medium and step to Very Fine to blend. Coarse works but takes more surface than the job needs. Aluminium oxide is the mineral for stainless. Work the full width of the tint band and keep the tool moving — a stationary disc adds heat where the alloy can't shed it.
Pickling vs. electrochemical vs. abrasive — which is best?
None of them universally. Abrasive for accessible welds where finish matters and you want no chemicals — it's the only method that fixes the surface finish at the same time. Pickling for complex internal geometry or where a spec demands verified removal, but hydrofluoric acid is genuinely dangerous and usually specialist work. Electrochemical for high-volume accessible stainless where you don't want to change the finish.
How do I prevent heat tint in the first place?
Reduce oxygen and reduce heat. Back purge tube and pipe properly, get adequate shielding gas coverage on top, consider a trailing shield, keep heat input down and travel speed up, and don't cut the post-flow short. On production work the gas usually costs less than the cleaning operation it eliminates.
Do I need to clean carbon steel weld discoloration?
For appearance and for coating adhesion, yes. But the corrosion argument doesn't apply — carbon steel has no chromium and no passive layer, so there's no depletion happening. The colours are an indicator of heat input, which on a hardenable steel might tell you something about temper, but that's a different concern.
Why did my stainless weld rust after I cleaned it?
Two usual suspects. Free iron contamination — an abrasive, brush, table, or tool that touched carbon steel embedded particles that rusted on their own. Or a chromium-depleted heat tint zone that was polished over rather than removed. Both look perfect leaving the shop.
Built for the Work.
Priced for the Worker.
Pro-Graad is an independent brand making non-woven discs, flap discs, and convolute deburring wheels for fabricators who measure performance in parts finished, not in logos on the box. Full grade ranges, honest specs, no premium tax. Available direct at pro-graad.com.
Shop the Full Range Convolute Wheel Selection Guide
info@pro-graad.com · pro-graad.com
This guide is general technical reference, not a substitute for your print, your customer's specification, a product SDS, a welding procedure specification, or your shop's safety programme. Where a specification or standard applies, it governs.
