Electrical cabinet manufacturing: stamping, welding, coating

Electrical cabinet manufacturing fails at the door, not at the panel. The frame can be square to 0.5 mm, the powder can be flawless, and the cabinet still leaks because the gasket land is bowed 1.5 mm across an 800 mm span, or because the gasket was fitted before the oven and took a permanent set at 190 °C.

A cabinet is a sealing problem dressed as a sheet metal part. The route is cut, form, weld, blast, pretreat, coat, assemble, and every step either preserves the flat gasket land or destroys it. This is the route for a steel cabinet built for rail and infrastructure duty, the dimensions that decide whether it passes an IP test, and the point where the enclosure stops being the fabricator's responsibility. The part range is under railway coach sheet metal parts and enclosures, and the cluster is the railway and infrastructure hub.

What has to be true before a cabinet leaves the plant?

Five things, and only two of them are dimensions. The frame must be square, the gasket land must be flat and continuous, the coating must cover every edge, the earth path must be at 0.1 Ω or less, and the hardware must be aligned so the door closes with a single hand.

The squareness figure comes from the gasket. A typical door gasket is a 10 mm × 10 mm EPDM section compressed 30% at closure, which is 3 mm. A door bowed 1.5 mm across its width uses half that allowance at mid-span before it uses any at the corners. Once the compression at mid-span reaches zero, the cabinet is open to dust at IP5X and to driven water at IPX4, whatever the label says.

The flatness target is therefore set by the gasket, not by appearance: 1 mm over 1,000 mm on the door, and 1 mm over 1,000 mm on the frame face that the gasket lands on. Everything else on the cabinet is looser — panel positions at ±1 mm, gland plate holes at ±0.5 mm, mounting brackets at ±1.5 mm.

Who declares the IP rating: the enclosure or the board?

The empty enclosure carries a rating tested to IEC 62208, which sets general requirements for empty enclosures for low-voltage switchgear and controlgear assemblies. The assembled board carries a rating declared under IEC 61439-1 and -2, which cover the assembly as a system.

That split matters commercially. We can supply an IP54 empty enclosure with a tested gasket land and door hardware. We cannot declare the finished board IP54, because the panel builder adds cable glands, ventilation, a cable entry cut by the installer, and a door that gets opened on site and closed against a cable. If the purchase order says "IP54 cabinet" without naming who tests what, the argument arrives at the first site inspection.

Two more conditions belong in the contract, because they are routinely assumed rather than agreed. First, that the cable entry pattern is frozen before the gland plate is cut; a pattern revised after coating costs a new plate and a patch of touched-up paint. Second, that the gasket is fitted by the fabricator after cure, not by the panel builder from a roll — a gasket cut 3 mm short at a corner is a leak the enclosure supplier gets blamed for.

Where electrical cabinet manufacturing splits between panel and frame

Panels are cut and folded; the frame is welded. The split is decided by which faces carry a seal and which faces only carry stiffness.

The folded parts

Body sides, roof, door, gland plate and inner mounting plate are cut on the fibre laser and folded on the press brake. Box bends on the 200 T brake with a 4,450 mm bed handle a 1,800 mm door in one pass, which is why a cabinet door does not need a welded seam down its length. A welded door seam is a distortion source and a coating failure point at the same time.

Door thickness is the decision that most affects the seal. At 1.5 mm, an 800 mm wide door deflects visibly under a hand push, and the deflection is enough to open the mid-span gasket. At 2.0 mm the same door holds its shape and gains 5.5 kg on a 1,800 mm door. Above 2.5 mm the door needs a heavier hinge and the cabinet gains mass for no sealing benefit.

The welded frame

The base frame and the door frame are welded. The base frame is 2.5–3.0 mm folded channel or square tube, welded at the corners, because it carries the cabinet's mass and the lifting load. The door frame is welded into the body so the gasket land is one continuous surface with no rivet heads and no folded lap.

Folded laps are where cabinets leak. A lap joint on a gasket land leaves a step of one material thickness, 2 mm on a 2 mm body, and the gasket cannot bridge it. Either the lap is welded and ground, or the gasket land is designed as a single folded flange.

The door mass row changes the hardware. A 22.6 kg door on two hinges loads each hinge with about 111 N static, but the load that matters is the slam: a door closed by a driver in a hurry sees a dynamic factor of two to three at the hinge pin, and a hinge with 0.2 mm of pin clearance sags 1.5 mm at the far edge over a 1,800 mm door. That sag is the same 1.5 mm that opens the gasket.

Weld sequence: keeping the gasket land flat

The frame is welded in a fixture, because a cabinet welded on a bench is square until the fixture is not there. Three rules keep the land flat.

Stitch, do not run. A 20–30 mm stitch at 100 mm pitch holds a 2 mm body to a 3 mm frame with far less heat input than a continuous bead, and the gasket seals against the land, not against the seam. Where a continuous weld is specified for a washdown cabinet, the land is then ground and checked, and the cost goes up accordingly.

Weld from the centre outwards. Running from one corner to the other accumulates shrink at the far corner, and the diagonal moves 1–2 mm. Welding outward in both directions from the centreline splits the movement.

Weld the base frame before the body panels go on. The base is the stiffest set of joints in the cabinet, and welding it into an assembled body racks the door opening.

Distortion that survives the fixture is corrected before blasting, by mechanical adjustment in a jig, not by heat. We do not flame straighten a cabinet. Heat straightening on a 2 mm panel leaves a visible wave in the skin, and the wave reappears after coating because the powder cure releases it.

Coating a cabinet: edges, threads and the earth path

The coating schedule is where a cabinet gains or loses its outdoor life. The route is shot blasting where the specification calls for a profile, seven-tank pretreatment, an electrostatic powder coat and an oven cure at 180–200 °C, with DFT and cross-hatch adhesion testing recorded per batch.

Three details decide the result. The first is edge coverage. A laser-cut edge has close to zero radius, and powder pulls away from a sharp edge during cure, leaving 20–40 µm instead of 70–90 µm. That thin edge is where rust starts on a trackside cabinet, and deburring before pretreatment is the only fix that lasts. The second is internal coating. A cabinet that is coated outside and left bare inside rusts from the inside out, because condensation forms on the cold steel. The third is masking: earth studs, gland plate faces, hinge pins and any sliding surface stay bare, and the mask land should be 3 mm wider than the washer that will sit on it.

Gaskets, locks, hinges and filters go on after cure. A gasket fitted before the oven takes a compression set at temperature, and a hinge pin fitted before coating collects overspray and binds.

The five checks that catch a cabinet before dispatch

Inspection on a cabinet is concentrated on the faces that seal, and it happens after coating because the oven moves the geometry.

Door flatness is measured on a surface table with the door supported at its hinges, not laid flat on its face, because a door measured face down reads flat when it is not. Frame diagonal is measured corner to corner inside the gasket land with a tape or a laser, and recorded before and after cure. Gasket land continuity is checked by running a finger and a straight edge along the full perimeter; a 0.4 mm step is a reject. Earth continuity is checked at 0.1 Ω or less from the stud to the frame. Hardware is checked by closing the door with one hand and watching for lift at the latch side.

The rework causes we see most often on this type of cabinet are all upstream of inspection. A gland plate cut before the cable schedule was frozen. A door handed the wrong way, which is a mirror error in the flat pattern and cannot be corrected after folding. A gasket fitted before cure. A hinge welded to the wrong face of the door frame. A powder touch-up over a weld made after coating. Four of the five are drawing-control problems rather than manufacturing problems, which is why the drawing fields at the end of this article matter more than the machine list.

The frame, booth and oven behind a cabinet

Cabinet panels start on the 12 kW fibre laser, whose 40 ft × 16 ft bed lets a body side, a door and two gland plates nest together so the hole patterns match across a batch. Folding runs on the 200 T, 5-axis CNC brake with the Delem control, with the bend programme stored against the part number. Frames are welded at the manual MIG/MAG stations or in one of the three CLOOS robotic cells where the cabinet repeats, in fixtures built in our own tool room.

Finishing is in-house and in sequence: a 16 ft × 16 ft shot blast chamber, a seven-tank pretreatment line, an 18 ft × 14 ft electrostatic powder booth and a curing oven, with DFT and cross-hatch adhesion checks recorded per batch. Indian Railways is a current customer, and the documentation route — mill certificate traceability, in-process inspection and dimensional reports — is the one described in railway coach panel manufacturing.

Two neighbouring articles cover the parts of this route in detail: box bending for electrical enclosures for the fold sequence that keeps a deep box square, and laser cutting for electrical enclosure panels for hole quality and the cut-edge condition that decides coating adhesion. Where the cabinet is galvanised before painting, the preparation rules are in degreasing electrical enclosures before coating, and the wider finishing route is under paint shop and powder coating.

What to put on the cabinet drawing

• Sheet thickness per part, body, door, base and gland plate, with the CRCA grade.
• Gasket land defined as a face, dimensioned flat, with no laps or rivets.
• Door handing and opening direction, stated as a note and shown on the view.
• Gland plate pattern and cable schedule revision, frozen before cutting.
• IP requirement and who tests what, empty enclosure to IEC 62208, board to IEC 61439.
• Finish specification: pretreatment, powder type, colour, DFT range, internal coating.
• Masking schedule: earth studs, gland faces, hinge pins, slide faces.
• Hardware list with the fit stage, so gaskets and locks are fitted after cure.
• Documentation: first article report, PPAP where required, DFT and adhesion records.

The RFQ checklist carries these fields, which is the difference between a quotation that holds and one that comes back with six questions and a revised price.

When a fabricated steel cabinet is the wrong answer

A welded steel cabinet is wrong for a small terminal box, where a folded and riveted enclosure at 1.2 mm is cheaper and the IP rating is easier to hold on a short gasket perimeter. It is wrong on a coastal or heavily chlorinated site where stainless or GRP is specified, and wrong where the OEM wants a moulded enclosure with integral mounting bosses.

We do not apply a certified IP test on the finished board, do not machine gland threads onto a plate, do not fit the electrical components, and do not declare a salt spray life in hours. What we do is build the enclosure, record the dimensions that the seal depends on, and hand over a coated, square, gasketed cabinet whose flatness figures can be checked against this table on arrival.