PVC Window Profile Design: Chambers, Steel Reinforcement and Drainage

What a window profile actually is

A PVC window profile looks like a simple extruded shape and is in fact one of the most highly engineered profiles made. A single sash profile carries six separate functions at once: structural stiffness, thermal separation, drainage, gasket retention, hardware attachment and a visible surface that must stay white for twenty-five years.

Every one of those functions occupies a chamber, a groove or a wall in the section, and they compete for the same space. That is why window profile design is a discipline of its own rather than a variation on general profile design.

Chambers: the thermal architecture

The number of chambers is the headline specification of a window profile, and it is quoted in every brochure: three-chamber, five-chamber, six-chamber. What the number actually controls is convection inside the profile.

Table 1 — Chamber count versus frame U-value. The gains flatten above six chambers without a foam fill.

The reason more chambers help is that each chamber traps air, and still air is a good insulator — but only if the chamber is small enough that convection does not start circulating. A chamber wider than about 25 mm develops a convection cell and the insulation value falls. So the design trend is not just "more chambers" but "smaller chambers", which is why a modern six-chamber profile is not much wider than an old three-chamber one.

Above six chambers the returns diminish, and the next step is filling the largest chamber with foam — usually polystyrene or polyurethane. A foam-filled chamber takes the frame U-value to 0.9 – 1.1 W/m²K, at 15 – 25 % more profile cost and a slower extrusion line because the foam fill is a second process.

Steel reinforcement: when it is needed and how much

PVC has a modulus of about 2.8 GPa. Aluminium is 70 GPa and steel is 210 GPa. A window is a structural element that must resist wind load and its own weight over a 25-year life, and unreinforced PVC cannot do it above a certain size.

Table 2 — Reinforcement guidance. The actual requirement depends on wind load, span and the profile's own section properties.

Three rules that are frequently broken:

1. The steel must be continuous through the corners. A window with steel in the straight runs but nothing in the corner joint has a hinge at every corner. Corner connectors are screwed into the steel, and the steel must be present where the screw goes.
2. The steel must fit the chamber. A steel insert 3 mm narrower than the chamber allows the PVC to deflect before the steel takes load, which defeats the purpose. Clearance should be 0.5 – 1.0 mm per side, not 3 mm.
3. Galvanised, not painted. Condensation forms inside window profiles. Painted steel rusts from a scratch; galvanised steel does not.

Drainage and pressure equalisation

Water gets into every window. The design question is not whether it enters but whether it leaves, and whether it leaves without getting inside the building.

Table 3 — Drainage features. The pressure-equalisation slot is the one most often omitted on cheap windows.

The pressure equalisation slot is what makes the drainage work. Water leaves a chamber because gravity pulls it down and because the air behind it can escape. If the chamber is sealed, the water sits in a trapped pocket and eventually finds its way inward through a joint. A small slot — 3 – 5 mm — at the top of the same chamber lets air in and the water out.

In exposed locations, wind pressure on the outside can exceed the head of water inside the drain and blow rain back in. That is what baffles and flap valves are for, and it is why a window rated for a sheltered site and one rated for a 20-storey building are not the same product even when the profile is identical.

The rest of the section

Table 4 — The non-thermal features of a window section. Each one is a separate tooling decision.

The gasket groove deserves a note because it is the most tolerance-sensitive feature in the whole profile. A gasket is a compression seal, and it only seals within a narrow range of compression — typically 25 – 35 % of its free height. If the groove is 0.3 mm too deep, the gasket is under-compressed and leaks; 0.3 mm too shallow and it is over-compressed, takes a permanent set, and leaks a year later. That is why gasket grooves are usually specified to ±0.1 mm while the rest of the profile is at ±0.3 mm.

Tolerances and what drives them

Table 5 — Window profile tolerances. The gasket groove is the tightest feature in the section.

Straightness matters more than it does on most profiles because the profile is welded into a frame. A profile with 2 mm/m of bow produces a frame that is visibly out of square after welding, and the welder cannot correct it. The usual cause is asymmetric cooling in the calibration tank, and the usual fix is balancing the spray rather than changing the die.

Design rules

1. Minimum 2.8 mm on visible walls for Class A. Do not accept 2.5 mm on an exterior application.
2. Keep chamber widths under 25 mm to avoid convection cells.
3. Put the steel in the chamber that carries the bending load — usually the outer chamber of the sash, not the centre.
4. Design the drain before the die. A drain that has to be punched afterwards is a hole in a wall that was not designed for it.
5. Slope every horizontal chamber towards a drain. Water that cannot leave will find a joint.
6. Radius every internal corner at R1.0 minimum. Sharp corners are where PVC stress-cracks after years of thermal cycling.
7. Allow for thermal expansion in the frame. A 1.5 m white PVC profile moves 4.5 mm over a 60 K swing; a dark laminated profile moves more.
8. Match the gasket groove to a real gasket, not to a nominal dimension. Send the gasket drawing to the die maker.

The standards that apply

Table 6 — The standards a window profile is actually tested against. Heat reversion is the one that reveals residual stress.

Heat reversion (EN 478) is worth understanding because it is a direct measure of how well the profile was extruded. A 200 mm sample is placed in an oven at 150 °C for 30 minutes and the length change is measured. Well-extruded profile reverts by under 2 %; poorly cooled profile with high frozen-in stress reverts by 4 – 6 %. The test is cheap, fast and it catches a problem that no dimensional check will find.

Frequently asked questions