Sliding Door Hardware from China: Tracks, Rollers and the Weight Rating Trap

A hotel project I was involved in got fitted with 84 sliding bathroom doors, all from one supplier in Guangdong, all with a "80kg rated" track. Within a year, eleven of them were dragging. Nothing had failed dramatically; the rollers had simply flattened. The rated load was the load at which the track itself did not bend, which is a completely different question from the load at which the rollers still turn freely.

Sliding door hardware looks like a commodity and is not. The category lives on four parts, and each has a number that is easy to overstate.

The track is mostly a wall-thickness question

Aluminium tracks are commonly sold as 6063 T5, and the alloy is rarely the problem. Wall thickness is. A nominal 2.0mm profile often measures 1.55–1.65mm with a cheap caliper because the extruder runs light to save metal. On a 2.4m span under a 60kg door, that difference shows up as visible sag in the middle, which then makes the door scrape the floor or jump the roller.

Bring a caliper, cut a section of sample, measure at three points along the extrusion, and compare with what the product page says. Anything below 90% of nominal is worth a conversation. Also ask whether the extrusion die belongs to the factory or the extruder; if you are paying for a custom profile, that die should be yours or at least written as unavailable to competitors.

Rollers: bearing type decides the life

The good version is a proper ball bearing with a nylon or polyacetal tyre, sealed or shielded. The cheap version is a plain bush or, worse, a nylon wheel running on a steel pin with no bearing at all. The bush version feels fine in the showroom and starts squealing at month four.

Three checks I do on rollers. Spin one and listen; a sealed ball bearing has a smooth, slightly dense feel rather than a dry rattle. Load a single hanger with the rated door weight and pull it by hand along a metre of track; it should move with a light consistent effort, not with a stiction-then-lurch feel. And check the tyre material by pressing a thumbnail in: soft nylon deforms and recovers, cheap recycled plastic marks. If your door is a bathroom door in a humid climate, a roller with an exposed steel pin and no shield will rust and seize, so specify stainless or a shielded version and expect to pay for it.

Weight ratings are not what you think

Ask the supplier, in writing, what the rated weight means: the static failure load of the track, the safe working load per hanger pair, or the load at which the door still operates smoothly over 20,000 cycles. Most published numbers are the first. What you need is the third, and in practice I plan the safe working load at around half of the published figure, a bit less for tall doors with a wide centre of gravity.

If you have a wide door, say 1.2m or more, the load on the top hangers is not just vertical. The door acts as a lever and the track takes a horizontal pull every time it is opened, so double top guides rather than single ones make a real difference. This is one of those details that never appears on a quote and decides whether you have callbacks.

Soft-close: where the cheap versions give themselves away

Soft-close mechanisms are usually a damper plus a spring. The cheap ones use a damper with a low cycle rating and a spring that loses preload. Test it in the way it will be used: open and close 50 times in a row and see whether the closing speed stays constant and whether the latch still engages fully at the end. On a good unit, the 50th close feels like the first. On a poor one, the damper has already gone soft and the door finishes with a small bang.

Also look at whether the mechanism is field-adjustable. On commercial projects you want the ability to correct a door that starts drifting, without ordering new hardware.

Finish, screws and the parts nobody quotes

Stainless covers are usually 304 for interior work; 316 appears on coastal projects and costs noticeably more. Salt spray hours matter more than the grade label, and a supplier quoting 480 hours should be able to name the test house. Screws are the quiet failure point: a track supplied with zinc-plated screws into a damp wall will show rust bleeding down the tiles in a year. Specify A2 stainless fixings and check the box separately from the hardware, because it is common for the sample to come with good screws and the bulk to come with the cheapest available.

Finally, ask for spare rollers and a spare damper per 20 doors in the purchase order. It is a small cost and it turns a warranty claim into a 10-minute repair. Suppliers who resist this are usually the ones whose spares are not interchangeable across batches.