A cabinet is a box that has to stay a box

A four-drawer filing cabinet is 1,320 mm high, 470 wide and 620 deep, and almost nothing about it is a strength problem. The body is 0.7 mm cold rolled sheet, the drawer fronts 0.8, the shelves 1.0, and the whole thing weighs less than the paper it will hold. What it has to do is stay flat, stay square and stay quiet for fifteen years, and those are three different specifications that pull in different directions.

Buyers ask for gauge first, because gauge looks like an answer. It is the last thing that decides whether a cabinet feels solid.

Oil canning is the first failure

Take a 300 by 470 mm side panel at 0.7 mm and push the middle with 5 kg. A simply supported plate formula puts the deflection near 8 mm, which is what the hand feels: the panel does not dent, it breathes in and out. At 1.0 mm the same push moves it about 2.7 mm, because plate stiffness goes with the cube of thickness. That is the honest argument for a heavier gauge, and also the reason it is the wrong fix: 43 per cent more steel for a panel that still gives.

What the panel needs is a fold: the shallow swages pressed into every cabinet side and door, which are not decoration.

What a 4 mm rib is worth

A rib 4 mm deep, formed from the same 0.7 mm sheet, puts two webs 2 mm either side of the skin, and bending stiffness goes with the square of that distance. Each web is 2.8 mm² of steel at a 2 mm lever, worth 11 mm⁴ on its own; the pair with their own section comes to about 30 mm⁴, against 0.17 mm⁴ for the 6 mm of flat strip the rib replaces. A factor near 170, from one forming station.

That ratio is why a plain panel looks cheap at any gauge and a ribbed one solid at 0.7, and why the press tool costs less than the steel saved by dropping a gauge.

Square is a tolerance, not a shape

The carcass is welded at the corners, and the diagonals are the only measurement that matters. On a 470 by 620 mm box the diagonals are 778 mm. Let one read 779 and the other 777 — a millimetre apart, which nobody would notice — and the sides are no longer parallel: the angle closes by 0.15°, and a 450 mm drawer travelling between them sees its clearance change by 1.2 mm. A ball-bearing slide has about 0.3 mm of float.

So the drawer runs tight at the back and the customer calls it stiff rather than out of square. The jig is checked on diagonals and the weld sequence written to pull the box into square; a cabinet welded in one pass from a single corner closes up on the last one.

The shelf is a beam wearing a sheet's clothes

Take a 900 mm shelf carrying 40 kg. As a flat 0.8 mm sheet 350 mm deep, the second moment is 15 mm⁴ and the deflection formula returns 1.2 m. That is not a shelf, it is a hammock, and the number says the sheet has no part in bending.

Fold 25 mm up at the front and the back and the same 0.8 mm becomes a tray with 8,300 mm⁴, and the sag drops to 2.1 mm against a 4.5 mm limit at span/200. Bending stress is 68 MPa, under a third of yield, so deflection governs, not strength. Doubling the fold to 50 mm takes the sag to 0.3 mm for a few grammes more steel. Doubling the thickness to 1.6 mm takes it to 1.0 mm and doubles the material.

Weld pitch, and the noise

Spot welds on 0.7 mm sheet are strong at any pitch a production line would use, so the pitch is a stiffness decision rather than a strength one. The panel between two welds is a strip with a free edge, and a long run of it drums at 20 to 30 Hz when the drawer closes. Forty to sixty millimetres is the usual answer on a body, closer at the corners where two panels meet at an angle, and closer again on the drawer front, the panel a hand touches. It is also the cheapest dimension to get right, because it costs cycle time and not material.

Powder in a box

Polyester powder at 60 to 80 µm, cured at 180 to 200 °C for ten to fifteen minutes after pre-treatment, is the standard finish and the standard problem. A cabinet is a closed box and powder is applied electrostatically: the faraday cage repels the charge, so the outer face takes 90 µm while an inside corner takes 30, below the film needed to survive a salt spray test. The answers are a second pass with the gun angled into the corner, a conductive primer, or a design that leaves an opening wide enough for the field.

The cure does a second thing. A 70 µm film shrinks as it crosslinks, and on a one-side-coated 0.8 mm panel that contraction bows the sheet by a few tenths of a millimetre — invisible in matt, obvious in gloss at a shallow angle. Coat both sides and the two films pull against each other and cancel.

Cold rolled, and where stainless earns its place

Body and drawers are DC01 in thicknesses from 0.12 to 3 mm, and ST12 where the gauge runs lighter, both chosen for a surface that takes paint without filler and for bending behaviour rather than for strength. Electro-galvanized sheet goes on interior dividers and the plinth where the paint is thin, SECCN2 sheet on the back panel and the drawer bottoms that never see daylight, and pre-painted coil on the flat side panels of a storage wall, where a coated blank is folded and used with no paint line at all.

Stainless appears in the canteen, the lab and the hospital store: square tube from 10 to 100 mm for legs and frames, 302 tube for trolleys and chair frames, 430 strip for the shelf and splashback where the budget will not carry nickel. The mild steel frame is Q235 plate cut, bent and welded, then coated after fabrication, because a frame painted before welding has bare steel in every weld and a rust line within a year.

The order

Write it as: grade and thickness per part, flatness to 1 mm in a metre, rib depth and position, weld pitch, coating system and film inside and out, colour and gloss, and the load per shelf with its deflection limit. Six of those seven are on the drawing. The load and its limit decide whether the shelf is 0.8 mm flat or 0.8 mm folded, and they are the two most often left to the supplier. A cabinet specified that way costs the same as one specified by gauge, and the difference shows the first time someone leans on it.