A framed door is a frame that happens to hold glass. A frameless glass door is the opposite: the glass is the structure, and everything else is a fitting bolted to it. That single inversion changes the entire failure analysis, and it is the reason maintenance teams who are competent with aluminium systems often misdiagnose a glass one.
The distinction matters more in the Gulf than most places, because frameless glazing has become the default for shopfronts, office partitions, hotel lobbies and villa bathrooms across the region. A lot of it is now ten to fifteen years old and entering the part of its life where the hardware starts talking.
The glass carries the load
In a framed assembly, the aluminium or steel section resists bending, carries the hinge loads and transfers everything to the structure. The glass is infill. Crack it and you have an expensive but contained problem.
In a frameless assembly there is no section. The pane resists its own bending, carries the weight of itself plus the fittings, and transfers all of that through a handful of clamped or bolted connections. The glass is simultaneously the cladding, the beam and the column.
This is perfectly sound engineering. It just means the tolerance for local damage is completely different. A chip on the edge of a framed pane is cosmetic. A chip on the edge of a structural pane is a stress riser on a member that is holding the door up.
Toughened glass cannot be repaired, and the reason is in how it is made
Thermally toughened glass is produced by heating float glass to roughly 620 °C and then quenching it rapidly. The surfaces cool and set first; the core cools afterwards and pulls inward as it contracts. The result is a locked-in stress state — the surfaces held in compression, the core in tension — and that balance is what gives the pane its strength and its granular break pattern.
The consequence is absolute: the stress state cannot be altered after the fact. You cannot cut a toughened pane, drill it, grind the edge or enlarge a hole. Every cut-out, every bolt hole, every notch for a patch fitting has to exist before the glass enters the furnace.
For anyone maintaining these doors, that has a hard practical edge. There is no such thing as “adjusting” a toughened pane to fit. If the opening has moved, or a replacement hole is in the wrong place by five millimetres, the pane is scrap. And once a toughened pane fails, it fails completely — the whole sheet dices at once, with no partial state to limp along in.
The nickel sulfide clock
Toughened glass carries a specific latent defect that framed systems largely escape, because it is a property of the toughening process rather than the installation.
Nickel sulfide inclusions are microscopic particles that can form in the glass melt. In the finished pane they sit in the tensile core, and over time they undergo a slow phase change that increases their volume. If an inclusion is large enough and sits in the wrong place, that expansion eventually exceeds what the core can hold, and the pane breaks — with no impact, no load and no warning. It is a genuinely spontaneous failure.
The industry answer is the heat soak test, defined in EN 14179. Toughened panes are held at 290 ± 10 °C for roughly two hours, which accelerates the phase change so that panes carrying a critical inclusion fail in the oven rather than in a lobby. Glass that passes is designated ESG-H, as distinct from plain toughened glass under EN 12150.
Two points matter for anyone looking after an existing installation. First, heat soaking reduces the residual risk sharply but does not eliminate it. Second, there is a timeline: Saint-Gobain’s technical guidance reports that around 90% of nickel sulfide breakages occur within six to eight years of production, and that breakages seldom occur more than ten years after installation. Failures beyond twenty years have been recorded, but they are exceptions.
So a frameless installation that has passed its first decade without a spontaneous break has very probably cleared the worst of that particular risk. What it has not cleared is everything bolted to it.
Patch fittings concentrate the entire problem
A patch fitting is a clamp. It grips the glass across a small area — typically a few square centimetres of bearing surface either side — and takes the full weight of the leaf plus every dynamic load the door experiences.
That concentration is the whole design, and it is also the whole vulnerability. Three things go wrong there, in order of frequency:
The gasket layer compresses. A patch fitting never contacts glass directly; there is a soft interlayer, usually a polymer or fibre pad. That pad takes a permanent set over years of clamping load and thermal cycling, and when it does, the clamping force drops. The door develops a few millimetres of play that was not there before.
The bolt tension relaxes. Once the interlayer has compressed, the bolts are no longer at their design tension. The leaf starts to move within the fitting under its own weight, which wears the interlayer further. The loop is self-reinforcing.
The alignment drifts. A door that has developed play no longer closes on the same line. It begins catching the floor, the head, or its opposing leaf — and because the glass cannot be trimmed, the only remedy is to re-set the hardware.
None of those three are glass failures. All three present as a door that does not close properly, which is exactly the symptom people assume means a replacement.
The floor spring is a hydraulic device buried in the floor
The other major component is the floor spring: a hydraulic closer sunk into the slab, carrying the bottom pivot and controlling how the door swings, closes and latches.
Being hydraulic, its behaviour depends on fluid viscosity, and viscosity depends on temperature. A floor spring in a shopfront with direct sun on the threshold runs hot and closes faster; the same unit in a chilled lobby runs slower. That is why closing speed appears to drift seasonally when nothing has been touched.
The failures that actually matter are these. Seals age and the unit develops a slow leak, after which closing force fades and the door stops latching reliably. Grit works into the pivot. The cement box the spring sits in cracks or fills with water during a wash-down, which corrodes the housing from underneath.
A floor spring is a replaceable component. The difficulty is purely one of access — getting it out means lifting the door and opening the floor — which is why a failed floor spring gets quoted as if it were a failed door.
Where glass doors actually fail
Run the list and the pattern is clear:
- Interlayer pads in the patch fittings, compressed and no longer clamping
- Bolt tension that has relaxed as a consequence
- Floor spring seals, fluid and pivot wear
- Edge damage to the glass — chips at a corner or near a fitting, which is the one item on this list that genuinely compromises the structure
- Alignment, which is a symptom of the first three rather than a fault of its own
Four of those five are hardware. Only one is the glass, and even then only when the damage sits near a fitting or a loaded edge.
What this changes about repair
The diagnostic order for a frameless door is the reverse of a framed one. On a framed door you check the frame first. On a frameless door you check the fittings first, because the glass is either intact or it is in pieces on the floor — there is no middle condition to assess.
That is the practical reason a competent approach to glass door repair in Dubai starts at the patch fittings and the floor spring rather than at the pane. The pane is the expensive item and the one that cannot be modified; the hardware is the part that wears, and the part where almost every presenting fault originates.
Replacement genuinely is the answer in two cases: edge or corner damage near a loaded fitting, and a cut-out that no longer matches the hardware. Outside those, a glass door that swings badly is usually a hardware job on a pane that is in perfect condition.
A ten-minute inspection
For anyone assessing an existing installation:
- Run a fingernail around every exposed edge, particularly within 100 mm of a fitting. Any chip you can feel is a stress riser on a structural member.
- Grip the leaf and try to rock it within the patch fittings. Movement means the interlayer has compressed and the bolts have lost tension.
- Let the door close from 90 degrees and time it. Compare against its opposite number if there is a pair. A fast, uncontrolled close means the floor spring is losing fluid.
- Check whether it latches every time, not most times. Intermittent latching is a closing-force problem, not an alignment problem.
- Look at the floor around the spring box for staining, lifted tiles or standing water after cleaning.
- Check the glass stamp. EN 14179 or ESG-H means heat-soaked; EN 12150 alone means it was not, which matters for a pane still inside its first decade.
The summary
Frameless glass doors do not fail the way framed doors fail, and treating them the same way produces the wrong quotation. The glass is structural, cannot be modified after toughening, and carries a defined early-life spontaneous-breakage risk that heat soaking reduces and time retires. Everything else — the clamping interlayers, the bolt tension, the floor spring — is conventional hardware on a conventional wear curve.
Diagnose the hardware before pricing the glass. On most doors that are swinging badly, the glass was never the problem.
References
- EN 12150 vs EN 14179 — toughening and heat-soak test parameters.
- Saint-Gobain Glass technical guidance on nickel sulfide inclusions in toughened and heat-strengthened glass.