Railway seat frame manufacturing: stamping, weld, coating

A coach seat frame rarely fails because the steel was thin. It fails at the weld toe where the backrest tube meets the pan, at the pedestal foot that rocks on a floor plate that is 2 mm out, or at the recline pivot where powder film grew the bore shut. Each of those is a manufacturing decision, and each one is cheaper to settle on the drawing than on a stripped seat in a maintenance shed.

Railway seat frame manufacturing is four decisions in sequence: which members are pressed and which are welded, where the fixture datum sits, what the mask schedule protects, and what inspection proves the frame before dispatch. Get the sequence right and a frame leaves the plant at 4–6 kg, square to ±1.5 mm, with a pivot that accepts its bush by hand. The parts range and documentation route for coach work sit under railway coach sheet metal parts, and the wider cluster is in the railway and infrastructure hub.

Why does a seat frame crack at the backrest weld?

Because the pan and the tube frame move against each other after welding, and the weld toe is where that movement concentrates. A 1.6 mm tube welded to a 1.6 mm pan produces a heat-affected zone roughly 3–5 mm wide on each side of the bead. That zone is harder and less ductile than the parent metal, and every passenger lean cycles it.

The fix is not a bigger fillet. It is a shorter weld in the right place. A 2.5–3 mm fillet on 1.6 mm tube is at the practical ceiling: a 4 mm leg on that section burns through the tube wall and leaves a concave bead with a crack starter at the root. Where the drawing asks for a continuous 4 mm fillet around a pan, the pan needs to be 2 mm, and the frame mass goes up about 0.9 kg.

Three other failure modes show up in service more often than parent-metal yield:

• Pedestal foot rocking. The foot is welded square to the pedestal, the coach floor plate is not flat, and the frame is shimmed at one corner. The shim takes the load, the bolt loosens, and the seat squeaks. A slotted foot with a ±3 mm adjustment removes the shim.
• Pan oil-canning. A pan pressed from 1.2 mm CRCA with no stiffening rib snaps through under a knee load, and the noise is reported as a loose seat. A 6 mm rib or a 1.0 mm bead along the front edge raises the snap-through load out of reach.
• Recline pivot seizure. Powder in the bore, or a bore that was laser cut at nominal and then coated. This one is purely a mask schedule decision and is covered below.

Where railway seat frame manufacturing splits between press and weld

The split is driven by volume and by shape, not by preference. A pan with a formed channel edge, a rib and two pierced holes is a pressing at 5,000 pieces a year and a laser-and-brake part at 400.

The pressed members

Seat pans and back pans carry the shape. They are 1.0–1.6 mm CRCA to IS 513, which specifies cold reduced sheet in drawing, deep drawing and extra deep drawing grades; the grade matters because a 1.6 mm D grade pan with a 6 mm rib will split at the rib radius where DD will not. Pressing runs on our 20–300 T mechanical presses with 250 T punching capacity, and the tooling is built in our own tool room, quoted separately from piece price.

Press tooling is worth it above roughly 500 pans a year. Below that, a laser-cut blank formed on the 200 T, 5-axis brake costs more per piece but carries no tool amortisation, and a design change after the first 50 seats costs a drawing revision rather than a die modification.

The welded members

Tube frames and pedestals are welded. Backrest perimeters are usually 20 × 20 × 1.6 mm or 25 × 25 × 2 mm square tube, notched and bent to a radius, with the pan welded or riveted inside. Pedestals are 50 × 50 × 2 mm tube or a 2 mm formed channel, with a welded foot plate and a bolted or welded top plate.

Riveting beats welding where the pan will be replaced in service or where the OEM wants the frame flat-packed. A 4.8 mm steel blind rivet in a 1.6 mm pan gives about 1.2 kN shear per rivet; six rivets per side is more than the foam and trim will ever transfer, and the joint does not distort the pan.

The table that settles the argument

Two rows deserve a decision before the drawing is frozen. The fillet row sets the pan thickness: a design note that reads "4 mm fillet all round" against a 1.6 mm pan will either be ignored or executed badly, and both outcomes reach the inspection table eventually. The mask row sets the bore tolerance: if the pivot must be a 12 H8 fit, powder is not allowed anywhere near it, and the mask costs about 40 seconds per frame.

How the weld sequence keeps a frame square

Sequence, not skill, controls the diagonal. Three rules do most of the work on a tube frame.

Alternate about the centreline. Run the left backrest joint, then the matching right joint, then return to the left. A welder who finishes one side first pulls the frame that way and stays there; the second side then fights the first, and the diagonal ends up wherever the two effects settle.

Weld the pedestal last, into a fixture that also holds the frame. The pedestal is the stiffest member in the assembly and the least forgiving. Welding it into a free frame racks the backrest, because the pedestal joint sits off the frame centreline and shrinks asymmetrically.

Leave the pivot brackets until the frame is square. Fit and weld the recline bracket after the structural welds have cooled, in a secondary fixture that holds the pivot axis. This is the single biggest difference between a frame that assembles on the seat line and one that is reamed on site.

To size the effect, work the diagonal. For a backrest frame 500 mm wide and 450 mm high, the diagonal is 672.7 mm. If one long member loses 1 mm to transverse shrinkage, the diagonal changes by (500 ÷ 672.7) × 1 = 0.74 mm. Lose 2 mm on each long member and the diagonal shifts about 1.5 mm — which is the whole of a ±1.5 mm allowance spent on one side.

We do not flame straighten a welded seat frame. Before coating, an out-of-square frame goes back into the fixture, the offending joint is cut, and the frame is re-welded and re-measured. After the oven cycle, a frame that has moved is re-made. Straightening heat on a coated structural part destroys both the film and the record.

Coating a seat frame: what gets masked

A seat frame in a coach interior is normally powder coated, and the mask schedule is decided by function. Pivot bores, earth bonding studs and the faces that sit on the floor rail stay bare. Everything else is coated, including the inside of the pedestal, because a coach interior sees condensation and cleaning chemicals for 20 years.

The build matters. At 60–90 µm a polyester powder covers a laser-cut edge and a spot-weld mark without bridging them. At 120 µm and above, the film softens the shape of a weld that a quality inspector needs to see, chips at the frame corner where the trim clips on, and adds 0.24 mm to a bore diameter if it reaches the bore wall.

Masking has two economics. Hand masking a pivot bore takes about 40 seconds per frame; silicone plugs take 5 seconds but cost per plug and lose their shape after a few hundred cycles. On a coach programme that runs for years, plugs win.

Cure is where geometry moves. Powder cures at 180–200 °C, and that cycle relaxes the residual stress the fixture was holding. A frame that measured 1.0 mm of diagonal before coating can read 1.6 mm after. Measure both figures and record both, or the site team inherits a tolerance argument nobody can win.

The press, brake and coating line behind a seat frame

Frames start as laser-cut blanks on the 12 kW fibre laser, whose 40 ft × 16 ft bed lets a pedestal, two pivot brackets and a pan blank share one nest, so the hole pattern between them cannot drift within a batch. Formed parts run on the 200 T, 5-axis brake with the Delem control, which stores the bend programme against the part number. Pans that justify tooling run on the 20–300 T presses with dies from our own tool room.

Welding is done in fixtures we design and build in the tool room, on one of three CLOOS robotic MIG cells for repeat frames or at the manual stations where the joint access suits a person. Coating is in-house: seven-tank pretreatment, a 16 ft × 16 ft shot blast chamber where the drawing calls for a blast profile, an 18 ft × 14 ft electrostatic powder booth and a curing oven, with DFT and cross-hatch adhesion testing in process.

Indian Railways is a current customer, and material is traceable to mill certificates with inspection recorded in process and dimensional reports issued with the batch. That documentation route is described in railway coach panel manufacturing, which follows a panel from coil to fitted part, and in railway underframe bracket manufacturing for the heavier structural side of the same coach. Where a frame is welded into a seat module, the joint tolerances in welded assembly tolerances are the ones that apply.

What to put on the seat frame drawing

• Member thickness and grade per part, not one note for the whole frame. A 1.6 mm pan and a 1.2 mm back pan are different IS 513 grades.
• One datum, normally the recline pivot bore, with the mounting holes dimensioned from it.
• Diagonal as a tolerance, not a note. "Square" is not a dimension.
• Mask schedule, face by face, including earth studs and floor contact faces.
• Finish specification: pretreatment sequence, powder type, colour, DFT range.
• Weld symbols with a sequence number, and the joints that must be welded after the frame is square.
• Test and documentation: first article report, PPAP where the coach builder requires it, mill certificate traceability.
• Packing: frames travel nested and braced, and the brace points belong on the drawing so they do not land on a coated face.

The RFQ checklist turns those fields into a form that goes out with the drawing, which usually removes one round of questions from the quotation.

When a welded steel frame is the wrong route

A welded steel frame is wrong when the seat is a reclining sleeper berth with a mechanism the OEM buys complete, when the coach interior specification calls for aluminium to meet a weight target, and when the programme is 50 seats for a refurbishment that will not run again. In the first case the frame is a mechanism housing and the OEM owns the tolerance stack. In the second, aluminium welding changes the distortion rules and the frame normally moves to extrusion and bolted joints.

We do not machine pivot bores after welding, do not press aluminium pans, and do not test seats to a rail seat standard — the static and fatigue proof loads in EN 16139, for example, at 1,600 N on the seat surface and 560 N on the back for severe non-domestic use, are the OEM's declaration, not ours. We build the frame to the drawing and prove the dimensions. If the design needs a machined bore, a bolted bushing housing is the cheaper answer than a machining operation bolted onto the end of the route.