Why Your Welder Grinds The Zinc Off Our Square Tube: Fabrication Questions Buyers Should Settle Before Ordering

A customer in Mombasa sent me a photo in August of a weld that looked like a sponge. His fabricator had run a MIG bead straight onto our galvanized tube, got porosity everywhere, and told him the steel was bad. It was not bad. Nobody had ground the zinc off first, and the welder was breathing the fume without a fan anywhere near him. That conversation cost him four days of schedule, so I have started writing the fabrication details down before the order instead of explaining them after.

We run our own roll-forming line here, and I spend part of most weeks in the workshop where we cut, coat and pack, so I hear both sides. The buyer hears a price per ton. The welder at the far end of that shipment hears a machine setting that no longer works. Between the two sits a short list of things worth settling on paper.

The zinc has to go before the arc goes in

Galvanized tube welds perfectly well once you deal with the coating. Zinc boils well below steel's melting point, so if you strike an arc on bare-looking zinc you get gas in the puddle, and that gas is what turned my customer's weld into a sponge. The fix is boring and cheap: grind the coating back roughly fifty to seventy-five millimetres each side of the joint, weld, then touch up the bare strip with cold galvanizing spray or zinc-rich paint. On a frame with twenty joints that is maybe an hour of extra grinding on the whole job.

The fume is the other half of it. Zinc fume gives welders a shivering flu-like illness that shops here call metal fume fever, and it is entirely preventable with a fan or extraction pulling the plume away from his face. If your fabricator complains that our tube "smokes badly", what he is really telling you is that his shop has no extraction. Worth knowing before you hand him two hundred frames.

Corner radius is where miter joints argue

Cold formed square tube does not have knife-sharp corners. Under EN 10219 the outside corner radius can run up to about three times the wall, so a 2mm-wall tube carries an outside corner somewhere near R6, and heavier walls pull that radius out further. On a miter joint, the radius sits on the outside of the bend, which means a plain 45-degree cut leaves a small crescent gap at the corner that somebody fills with weld. Most fabricators cope with this in their sleep. What catches people out is mixing tubes from a hot finished route, where the corner is rounder still, into the same jig as cold formed stock. Two tubes with the same 60x60 callout can sit half a millimetre apart in the corner and the jig will feel it.

If your design has visible miters, a slot that the tube must click into, or a jig that was built around a sample, tell us at the order stage. We can hold pieces back from the same production run so your corner geometry stays consistent from the first container to the last.

The seam inside the tube

Roll-formed square tube is made flat and closed into a square, and the weld seam ends up on one face. On the outside you can barely find it. On the inside there is a bead, usually under a millimetre tall on the walls we run, but it is there, and it runs the whole length. For ninety per cent of jobs nobody ever notices it. The ten per cent where it matters is anything that telescopes: an adjustable leg sliding into a bigger tube, a sleeve joint, a gate frame sized so a smaller square slips inside. If the inner tube's face rides against the bead, the slide starts grabbing halfway down.

Tell us if your parts telescope. The bead position is predictable once you know the line, so we can quote stock rolled so the seam lands where your design can live with it, or talk through how much clearance your sliding joint actually needs.

Walls wander, and thin walls punish a lazy machine setting

A 2.0mm wall tube does not measure 2.000 all the way around every piece. Coil gauge wanders inside the tolerance, and on thin walls that wander is enough to matter at the torch. A setting that burns clean through 2.1mm will blow a hole through a piece that runs 1.85mm, and the welder blames the tube. He is not wrong, exactly, he just set the machine for the top of the range. On walls below 2mm we tell fabricators to set the current for the thin end and travel a touch faster, and to expect the burn-through rate to drop to almost nothing.

If your fabrication is thin-wall and appearance matters, ask us for the actual measured wall range of the batch rather than the nominal. It is one line in our packing list and it saves the argument on your floor.

Straightness and ends, before it becomes your problem

Our line holds straightness to around two millimetres per metre, so a six metre length can drift up to a centimetre or so off a perfect line. For most frames that vanishes inside the tolerance of the jig. Where it does not vanish is long guardrails and anything where two tubes must line up in mid-air, because a centimetre over six metres reads as visibly wavy once the tubes are tacked side by side. Ends matter too. Mill-cut ends are square to about a degree, which is fine for weld prep but visible if your design leaves a bare end facing the street. Deburred, square-cut ends are worth asking for when the tube is a finished architectural face.

What to put in writing

Four lines settle nearly all of it: whether the parts get welded and whether the shop has extraction, whether anything telescopes, what the jig was built around, and whether ends are structural or architectural. Send those with the enquiry and we will flag anything on our side before the steel moves, not after. The sizes and walls we run weekly sit on our galvanized square steel tube page, and if you send drawings we will tell you honestly where the tube behaves and where it needs a workaround.

My Mombasa customer's second container went through his shop without a single porous weld, because the purchase order now says "grind 60mm, extraction on" in plain words. Steel rarely fails in the hands of someone who was told what it is.