Technical editorial

Steel doesn't punish insulation — it punishes insulation it can short-circuit

We modelled six South African wall assemblies with a 2-D finite-element thermal engine. The results settle two arguments: what steel framing really does to insulation, and where the vapour control layer belongs in each climatic zone.

Technopol SA · June 2026 · 4 min read

South Africa's two dominant wall types — cavity brick and light steel frame (LSF) — were solved per EN ISO 10211, uninsulated and insulated, across the three design climates that bracket the SANS 204 zones. The headline numbers are stark: a bare plastered cavity-brick wall achieves a total resistance of just R 0.68 m²K/W, and an empty steel-frame wall R 0.52 — far below the SANS 10400-XA minimum of 2.2 for non-masonry walls. The insulated versions reach R 1.99 (masonry + 50 mm external EPS) and R 3.46 (LSF, fully insulated).

Bar chart of total thermal resistance for six wall assemblies — Whole-wall thermal resistance, solved by 2-D FEM. Red: uninsulated baselines. Blue: insulated LSF. Green: masonry with external EPS.

The steel-frame trap

Here is the result every specifier should know. Fill a 90 mm steel-stud bay with mineral wool and the studs quietly take back 38 % of the wall's resistance — upgrade to a denser wool and they take 44 %, because the penalty is set by the contrast between steel and what surrounds it. Add a continuous 50 mm EPS layer outside the frame — a layer the stud cannot cross — and the penalty halves to 18 % while the wall climbs to R 3.46. The thermal images make it visible: in the bare wall the cold reaches the lining; in the EPS-wrapped wall the isotherms crowd into the insulation and the structure stays warm.

Solved temperature field of an uninsulated steel frame wall — Solved isotherm fields (Cape Town winter, interior left). Left: uninsulated LSF — inner surface at f_Rsi 0.71, mould-active at 60 % indoor humidity. Right: insulated LSF with continuous EPS — f_Rsi 0.93, the stud line barely visible.

Solved temperature field of a fully insulated steel frame wall with continuous EPS — Solved isotherm fields (Cape Town winter, interior left). Left: uninsulated LSF — inner surface at f_Rsi 0.71, mould-active at 60 % indoor humidity. Right: insulated LSF with continuous EPS — f_Rsi 0.93, the stud line barely visible.

Bare walls don't just leak heat — they grow mould. At 60 % indoor humidity in a Johannesburg winter, the inner surface of an uninsulated steel-frame wall runs at 87 % surface humidity: above the 80 % mould threshold, every cold night.

The vapour layer must follow the climate

The condensation screens (EN ISO 13788) found two genuine failure modes, and both are specification errors, not product failures. An interior polyethylene VCL — correct and necessary in the heating zones 1, 2 and 4 — causes condensation in humid Durban, where air-conditioning reverses the vapour drive (105 % saturation on the film). The mirror error also fails: the zone-5 wall with its exterior vapour retarder condenses in winter zones (up to 119 %). One wall does not serve all zones. Masonry with external EPS, notably, passes every climate tested without a dedicated VCL — external insulation keeps the structure warm, so the dew point never enters it.

A1 wall assembly build-up for cold and wet zones — The cold/wet-zone LSF assembly: VCL on the warm side, continuous EPS outside the frame, lipped C-stud 90×40×12×0.8 at 600 mm centres.

The full 17-page study includes the solved fields for all six assemblies, the Glaser condensation matrix across three climates, zone-by-zone specifications and complete worked calculations.

Download the full study (PDF)

All temperature fields, R-values and condensation screens computed with the THERMA v2.2 two-dimensional finite-element engine (validated against EN ISO 10211 reference cases) using typical SA material properties. Substitute certified product declarations for compliance work.