Stainless chairs and dissimilar metal joints
Stainless steel does not rust because the grade is wrong nearly as often as buyers assume. It rusts at joints: where a zinc-plated bolt meets a 304 boss, where an oak seat rail is screwed tight against a tube, or where a washer clamps a crevice that no rinse water can reach. Those details are drawn once and lived with for ten years, so we treat joint isolation as a specification, not an assembly habit.
Published by JPChairs · Chair manufacturer in Foshan, China · 2026-11-07
Where stainless frames actually corrode
Corrosion on a stainless frame almost never starts on a visible face. It starts where two parts are clamped so close that wash water cannot leave: under a washer, inside a threaded connection, along the faying surface of a bracket. A gap under 0.3 mm holds electrolyte long after the chair has dried, and the chloride concentration inside that crevice climbs several times higher than on the open surface, where 304 keeps its passive film with about 18 per cent chromium. Inside a wet crevice the chlorides break the film locally, and pin rust appears within a single season.
- Washer crevices: the top failure location, because the joint is under constant preload and the gap is guaranteed.
- Thread roots: wash water sits in the last two threads of a horizontal bolt.
- Contamination: a wire brush used earlier on carbon steel leaves particles that show freckle rust within 24 to 48 hours in humid storage.
None of these show at inspection on the day of packing, which is why isolation is designed in rather than inspected in.
Isolating stainless from carbon steel and zinc
The classic pair on a mixed frame is a zinc-plated bolt or brace against a 304 tube. In wet contact the couple is worth roughly 0.5 volts, the zinc side sacrifices itself first, and a plated flat bar bolted bare against stainless shows white rust in weeks. On coastal installs the plated part can perforate within 30 months while the stainless only stains from the oxide bleeding across it.
- Isolation washers: POM-C, 1.5 to 2.0 mm thick, at every carbon or zinc contact face.
- Sleeves: nylon 6 bushings in the bolt hole with about 0.5 mm diametral clearance, so the shank never touches the tube wall.
- Acceptable contact: dry interiors sealed from wash water, same metal family, no isolation needed.
- Tooling rule: dedicated brushes and grinding discs for stainless only; a shared consumable re-contaminates a clean frame in one pass.
Isolation hardware adds 0.30 to 0.50 USD per frame, which is the cheapest corrosion protection on the chair.
Timber and upholstery against a stainless rail
Wood against stainless looks harmless and is not. Oak and chestnut leach tannic acid, and a seat rail at 18 per cent moisture content keeps a damp line against the tube for most of the year. The stain is cosmetic on the metal, but the timber also swells and shrinks across seasons, working the screws loose over 12 to 18 months.
- Barrier tape: a 1 mm EPDM strip along the contact face, or a 2 mm POM spacer block at fixing points.
- Sealed end grain: end grain absorbs water three to four times faster than face grain, so it is sealed before assembly.
- Drainage slot: a 3 mm slot under the seat rail lets condensation leave instead of pooling on the top face of the tube.
- Moisture gate: timber above 16 per cent moisture content is held back for 48 hours of conditioning before it is fitted.
Fabric is simpler: keep the staple line off the tube and onto a timber or POM capping strip, because a staple driven through fabric into stainless pierces the passive layer at 60 points per seat.
Torque values by fastener and boss material
Stainless hardware galls and plastic bosses strip, so torque is written per material, not per bolt size. A2-70 M6 bolts go to 8 Nm and M8 to 18 Nm into a steel boss. Into a moulded POM boss the same joint stops at 2.5 to 3.0 Nm on an M5, and a nylon self-tapping screw into a plastic boss takes only 1.5 to 2.0 Nm. Stainless-to-stainless joints above M10 get anti-galling paste, because a dry M10 tightened past 30 Nm can seize and shear on the next service visit.
Because EPDM washers relax 10 to 15 per cent in the first day, we re-torque a 10 per cent sample 24 hours after assembly and record the second value on the batch card. A joint torqued once and never re-checked is the joint that rattles loose in a beach club by mid-season.
Welded joints, repair and re-passivation
Every weld damages the passive layer in the heat-affected zone, so weld dressing is part of the joint specification rather than an optional polish. We dress to 120 grit maximum, pickle for 20 to 30 minutes, neutralise and rinse. Repair areas that cannot be pickled get citric passivation, 10 minutes at 60 C. Grinding a weld with a disc that has touched carbon steel does more damage than skipping the dressing entirely, which is why the tooling rule from the isolation section applies here without exception.
Bolted joints are not automatically better: they introduce crevices and hardware. Welded joints are not automatically worse: a dressed, pickled weld has no faying surface at all. The choice is made per joint on the drawing, with the isolation method written next to it.
Bench checks before a frame is packed
Joint isolation is verified by disassembly, not by looking. One frame per batch is taken apart and every isolation washer and sleeve is checked for presence and seating. Torque is audited on 10 per cent of joints per batch with a calibrated wrench, and the sample is re-read at 24 hours to catch relaxation. For coastal projects we add a crevice coupon: a bolted stainless pair with a deliberate 0.2 mm gap goes into salt spray for 240 hours, and the pass line is no rust bleeding from the crevice.
The frames built to this joint specification are listed under stainless steel chairs for coastal venues, and the isolation details are agreed on drawing through OEM joint isolation specifications before the first sample is welded.
