Stainless Duct for Cleanrooms: Forming, Welding, Passivation

2026-11-20

A pharmaceutical plant in Penang found rust spots on a stainless supply duct six weeks after installation. The duct was specified as 304, the mill certificate said 304, and the fabricator had every reason to believe the material was right. The rust was not in the steel. It was free iron embedded in the surface by a carbon-steel wire brush used during assembly. In a cleanroom, that is enough to fail a qualification inspection.

Cleanroom duct work is mostly about controlling contamination during fabrication. The forming, welding and cleaning all matter, and most of the mistakes happen before the duct ever leaves the shop.

Material and finish, decided with the user

304L is the standard choice for general cleanroom supply and return air. 316L goes into areas with saline cleaning agents, process exhaust, or higher corrosion risk. The "L" matters: low carbon reduces sensitisation around a weld, which is where intergranular corrosion starts.

Grade / finishTypical surface roughnessWhere it belongsCost index
304 2BRa 0.8–1.6 µmNon-product rooms, service areas100
304L 2BRa 0.8 µmGeneral cleanroom HVAC107
304L BARa 0.3–0.5 µmExposed duct in Grade B/C areas122
316L 2BRa 0.8 µmSaline or corrosive environments138
316L electropolishedRa below 0.25 µmProcess gas, high-purity zones180

Do not over-specify finish. A 316L electropolished duct in a corridor service void is money burned. Ask the user which rooms are classified, what the cleaning agents are, and whether the duct is visible or hidden. Those three answers settle the grade.

Thickness follows the cleaning, not the pressure

Cleanroom supply systems run at modest static pressure, often 400 to 700 Pa. The gauge is driven by handling and by how often the duct is wiped down. For ducts under 600 mm, 0.8 mm is normal; between 600 and 1,000 mm, 1.0 mm; above that, 1.2 mm with external stiffening. Panels thinner than 0.8 mm drum when a cleaner presses on them and eventually develop oil cans that collect dust.

Keep carbon steel away from the job

This is the rule that trips up most shops. Stainless that has been in contact with carbon steel picks up iron particles, and those particles rust. The rust is not the stainless corroding; it is the embedded iron corroding, and it looks terrible in a white room.

Dedicate a set of rolls, a shear blade and a brush to stainless work. If that is not possible, clean every tool with solvent before the stainless job and cover the bed of the machine with a plastic film. Do not use the same grinding wheel on carbon steel and stainless, and do not store the finished sections near a carbon-steel rack where swarf can blow onto them.

Keep the protective film on the sheet until the part is formed and welded. It stops most of the contamination before it starts, and it makes the finished duct much easier to clean at handover. On the two-station line, film also reduces scratching as parts slide along the outfeed table.

Forming stainless without work-hardening it

Stainless work-hardens under deformation. Bend it twice in the same place and the second bend fights back, often cracking at the outside radius. That changes three settings on the line.

First, roll gaps need to be about 10 percent tighter than the sheet thickness suggests, because stainless springs back more than galvanised. Second, bead depth should be 5 to 6 mm rather than the 8 mm used on mild steel; a deeper bead on 0.8 mm stainless will crack at the ends. Third, take less per pass on the gorelocker and let the tool do the work gradually.

Cut edges need deburring before forming. A stainless burr is harder than a mild steel burr and it will score the next roll it touches, leaving a scratch that then traps contamination. A quick pass with a deburring tool on every blank costs seconds and saves hours of polishing later.

Seams for a clean interior

Use continuously welded longitudinal seams on classified areas, with the weld ground and polished flush. Pittsburgh locks are acceptable on non-classified service ducts if they are sealed, but any crevice is a place for particles to collect. On the main supply run I weld everything and keep the interior smooth enough that a wipe test comes back clean.

Welding, pickling and passivation

TIG welding with a precisely controlled heat input is the standard. Use a small bead, keep the heat-affected zone narrow, and back-purge with argon on 304L and 316L so the inside of the weld does not oxidise. An oxidised weld is rough, discoloured, and impossible to clean back to the original finish.

After welding, the heat-affected zone needs pickling and passivation. Pickling removes the oxide layer and any free iron; passivation restores the chromium-rich surface that gives stainless its corrosion resistance. ASTM A967 is the usual specification, and the fabricator should be able to produce a certificate or an in-house procedure with the chemical concentration, temperature and contact time.

Electropolishing is a different process and a different price. It removes a thin layer of metal and produces a much smoother surface, typically below Ra 0.25 µm. Use it only where the user has asked for it, usually on process gas lines rather than general HVAC duct.

One practical warning: pickling acid is dangerous stuff. It needs a dedicated area, plastic containment, neutralising facilities and trained operators. If a shop cannot do that safely, send the sections to a subcontractor. Doing a bad passivation job in-house is worse than paying for a good one outside.

Cleaning and inspection before handover

Four checks, done in this order.

Visual and wipe test. Wipe a 300 mm length with a lint-free cloth and isopropyl alcohol. Any visible soil or colour on the cloth means the duct goes back for cleaning.

Ferroxyl test. A reagent that turns blue in the presence of free iron. Use it on the welds and anywhere a carbon-steel tool may have touched. A positive result means the area must be re-passivated.

Borescope on every main run. Look at the inside of the welds and the corners. A camera on a 20 mm entry point shows more in ten seconds than a man can see by looking down a 6 m duct.

Leakage test. Cleanroom supply ducts are often tested to a tighter class than normal HVAC, so confirm the specification before testing. Seal failures at the flange joint are usually a gasket problem, not a welding problem.

Package each section in two layers of clean polythene with desiccant inside the outer layer, and seal the openings. A duct that arrives clean and open in a warehouse will be dirty by the time it is installed, and the client will hold the fabricator responsible.

Cleaning agents and what they do to stainless

The cleaning regime specified by the user is a design input, not an afterthought. Cleanrooms use a surprising range of chemicals, and not all of them are kind to stainless or to the duct seals.

Quaternary ammonium compounds and hydrogen peroxide are common and generally safe on 304L and 316L if the surfaces are rinsed. Chlorinated cleaners are the problem: even at low concentration, repeated exposure causes pitting, especially around welds and in crevices. If a site uses bleach-based products, specify 316L and pay attention to the gasket material at every joint.

Alcohol-based wipes are fine for daily cleaning and leave little residue. Alkaline detergents need a proper rinse, because a dried alkaline film on a bright-finish duct shows as a white haze that customers mistake for poor material. Tell the cleaning contractor which products are approved and ask the facility manager to keep the list current.

A cleaning sequence that works

Start at the supply end and work towards the return, so dirty water never runs over a surface that has already been cleaned. Use two cloths: one to apply the solution and one to dry. A single wet cloth just moves contamination along the duct.

Change cloths often. A cloth picked up from the floor or used on a carbon-steel beam will undo all the contamination control from the shop. In a classified room, colour-code the cloths by area and keep them in sealed bags until use.

Why sections get rejected at handover

Three reasons, almost every time. Free iron on the surface from a tool or a brush, an oxidised weld that was never pickled, and dust left inside the duct because the openings were not sealed during transport. All three are preventable in the shop, and all three are expensive to fix once the duct is on a scaffold in a live facility.

My last check before despatch is a white glove test inside a newly opened section. If the glove comes out grey, the duct goes back for cleaning. It is a crude test and it has never let me down.

FAQ

Can I run stainless on the same rolls as galvanised sheet?

Not without a thorough clean, and even then it is a risk. The smallest carbon-steel particle embedded in a roll can transfer iron to the stainless surface. Dedicated rolls are the only reliable answer for classified work.

Do I need to passivate 430 stainless?

430 is often used without a full passivation step in less critical areas, but a proper clean and a light passivation still improve its resistance. In a cleanroom, treat all stainless the same and keep the procedure simple.

How do I know if the surface finish I ordered is right?

Compare it with a reference coupon from the mill, and measure Ra with a portable roughness meter if the specification states a number. Visual descriptions like "bright" or "brushed" are not measurable and lead to disputes.

Why do my welds discolour even with argon?

Usually the purge flow is too low, the purge ends too early, or the gas is not reaching the back of the joint. Purge until the weld has cooled below the oxidation range. A few extra seconds of gas costs nothing compared with reworking a weld.

Cleanroom work rewards patience. The forming is not difficult, the welding is routine, and the contamination control is what separates a duct that qualifies from one that gets pulled out. If you are choosing a line for this market, look at the beading and forming options with shallow-depth settings in mind rather than the maximum figures in the brochure.

Read0
share
Write your comment right