A thermal break is a conduction fix. It sets a strip of low-conductivity material — usually polyamide — into the metal so heat can't run straight across it. More resistance in the path, less conduction through it. That's the whole idea, and on its own terms it works: interrupt the continuous metal line and the number improves.
The code reaches for the same move in the wall. In a steel-framed external wall, where lightweight cladding and the internal lining fix to the same steel member, NCC 2022 calls for a thermal break of at least R0.2 on the frame — a prescriptive strip of resistance dropped in to interrupt the stud bridge. Same instinct as the window: add R, break the metal path.
Here's the catch that the whole of this piece turns on. A resistance value is a conduction idea. It assumes the layers either side of the break sit more or less still, warming in steps from the inside out. The number is only true while the assembly stays still.
The wall the code now wants doesn't stay still
The same rulebook that asks for the break also, increasingly, asks for a drained, ventilated cavity behind the cladding. Under the NCC's weatherproofing provisions (Volume Two 10.8.1 / F8D3, broadened again in NCC 2025), external walls in the climate zones it applies to get a cavity of at least ~12 mm, open top and bottom, with ventilation and drainage openings of at least 1,000 mm² per metre of wall. Its job is moisture: let any water that gets behind the cladding drain away, and let outside air move through to dry the cavity out. (It's the same cavity we pulled apart in performance isn't fixed until it's in the wall.)
Read that last part again. The cavity works by moving outside air. It's a convection device, deliberately — it has to be, or it can't drain and dry. So now the wall holds two ideas at once: a break that assumes still, stepped-up conduction, and a cavity engineered to flush outside air through itself.
What moving air does to a resistance number
Thermal-calculation practice — ISO 6946, the method underneath the numbers — treats a ventilated cavity bluntly, because it has to. Vent a vertical air layer with up to 500 mm² of openings per metre and it still counts as unventilated. Between 500 and 1,500 mm²/m it's "slightly ventilated." Past 1,500 mm²/m it's "well ventilated" — and at that point the calculation throws away the cavity's own resistance and the resistance of everything outboard of it, and places an outside-air surface condition at the back of the cavity. The NCC's minimum 1,000 mm²/m already sits in the ventilated band; a cavity open top and bottom, as the code intends, routinely reaches the well-ventilated end.
In plain terms: you can't bank the cavity as insulation, and anything sitting out in it — or outboard of it — is, thermally, outside. Moving air has flushed the warmth away before it can build a gradient across it.
A break is only worth what sits inboard of it. Put it in a cavity that's being flushed with outside air, and there's nothing warm left for it to protect.
So where is the break?
This is where the window gets caught — and it's a position question, not a product one. A window is installed into the wall, and the frame can sit at different depths in the opening. Set it so the internal aluminium is sealed and insulated in line with the wall insulation and the stud — inboard of the cavity — and the break sits in the still, warming, insulated zone it was designed for. It does its job.
Push the same frame outboard, so the internal aluminium hangs out in the ventilated cavity, and the break goes with it, into the moving air. Now the "warm" side of the break isn't in the warm wall at all; it's in a cavity the code is deliberately flushing with outside air. The conduction fix is sitting in a convection flush. The air moves past it faster than the break can hold a temperature across it — and the internal aluminium, the surface that faces the room, is dragged toward outside conditions.
That's the sting in one picture. In the stranded position, a thermally broken frame behaves much like a frame with no break at all. You specified — and paid for — the polyamide, the better extrusion, the premium. You then installed it where the cavity sweeps the benefit straight back out, and the inside surface that condenses and grows mould lands on the same line a plain aluminium frame would have given you. Same cold line. You just paid more to reach it.
The R0.2 break is the same reflex
The steel-frame R0.2 break is worth having — it genuinely interrupts the stud-to-cladding bridge, and it's the right instinct. But it's a single conductive resistance value, and a resistance value only ever describes a still assembly. On its own it can't tell you the surface temperature on the coldest line once the real wall is built — the cavity it now sits behind, the ventilation that cavity carries, the junctions where the frame meets it. The number says R0.2; the wall says "it depends where, and on what's moving." The honest answer comes from modelling the section as built, not from the strip in isolation.
Model the air regime, not just the metal
The fix is the move we make everywhere: stop trusting the catalogue value and model the real thing — the frame where it actually sits in the opening, with the real cavity and the real insulation line, resolved in two dimensions to ISO 15099 and cross-checked against THERM (more than 1,000 times over, in a public register). Then you can see the one thing the resistance value can't give you: whether the break is in the warm zone or stranded in the flush, and what the coldest inside surface actually does as a result — an fRsi condensation screen and, taken to the junction, the AIRAH Mould Index via TIRS. A break is cheap insurance when it's in the right plane, and dead money when it isn't. The only way to know which you've bought is to model it.
Did your thermal break land in the warm zone — or the cavity?
Send us one thermally broken window with its installation detail. We'll model the frame where it actually sits in the wall — the break, the cavity, the insulation line — in 2D, and show you whether the break is doing its job or stranded in moving air, and what the coldest inside surface does either way. The first one costs you nothing.
Start with one window →Common questions
Does a thermal break still work if the window sits out in the ventilated cavity?
Often not. A thermal break is a conduction fix — it adds resistance to interrupt the solid metal path, and it relies on the layers each side of it staying relatively still. A drained, ventilated cavity is the opposite: it moves outside air through itself by design, to drain and dry. If the frame is installed so the internal aluminium and the break hang out in that cavity rather than sealed and insulated in line with the wall insulation, the break sits in moving outside air — the warmth is flushed away before a gradient can build, and the frame behaves much like one with no break at all.
Why does a ventilated cavity reduce a wall's thermal resistance?
Because moving air carries heat away. ISO 6946 classes a vertical air layer by its ventilation openings: up to 500 mm² per metre it is treated as unventilated and keeps its resistance; between 500 and 1,500 mm²/m it is slightly ventilated; above 1,500 mm²/m it is well ventilated, and the calculation discards the air layer's resistance and that of every layer outboard of it, applying an outside-air surface condition at the back of the cavity. The NCC's minimum 1,000 mm²/m already puts a drained-ventilated cavity in the ventilated range, so its insulating value can't be banked.
Is the NCC R0.2 steel-frame thermal break enough on its own?
The R0.2 break required under NCC 2022, where lightweight cladding and internal lining fix to the same steel member, is a real and worthwhile improvement — it interrupts the stud-to-cladding bridge. But R0.2 is a single conductive resistance value, and a resistance value only describes a still assembly. It doesn't by itself tell you the surface temperature on the coldest line once the real wall — cavity, ventilation, junctions and all — is built. For that you model the section in 2D.
How do I know if my thermally broken windows were installed in the right plane?
Model the frame where it actually sits in the wall, in 2D, with the real cavity and insulation line. The question is whether the break and the internal aluminium sit inboard — sealed and insulated in line with the wall insulation and the stud — or whether they project out into the ventilated cavity. The first keeps the break in the warm zone it was designed for; the second strands it in moving outside air. The model shows which, and what the coldest inside surface does as a result.
Illustrative mechanisms. The install-plane effect described — a thermal break losing its benefit when the internal aluminium is positioned outboard, in a drained-ventilated cavity, rather than sealed and insulated in line with the wall insulation — is a mechanism by which a thermally broken frame can perform like a non-broken one; whether and to what degree it occurs depends on the specific system, cavity and installation. A thermal break in the correct plane is a genuine performance improvement over plain aluminium, and the NCC R0.2 steel-frame break is a required, worthwhile measure. ISO 6946 air-layer categories and the NCC cavity and thermal-break provisions are summarised for explanation; apply the current clauses to your project. Project-specific thermal simulation to ISO 15099 / NFRC 100, with condensation assessed via surface temperature and the Mould Index (TIRS); modelled results are indicative to the systems and details provided, and final NCC compliance is determined by your assessor or certifier. AFRC Lab (Advanced Fenestration Research & Certification) is an independent company; it is not the AFRC and is not affiliated with WERS. TIRS™ is a trademark of AFRC Lab Pty Ltd (patent pending).