A curtain wall isolator profile carries the glass dead load of a transom across the thermal break, sustained, for the 25-to-50-year life of the envelope. F1 Composite's own isolator laminates do this today with multi-layer multiaxial fabric that balances the 0°/90° modulus most pultrusion runs at 4-to-5-to-1. That is a solved problem. The harder one sits one layer out: replacing the aluminum mullion and transom itself — not the thermal break inside it, the primary structural member — with pultruded FRP, at the stiffness a modern unitized facade actually demands.
Nobody disputes the case for trying. Aluminum conducts heat at roughly 500 times the rate of pultruded FRP, and every mullion is a thermal bridge whether or not it carries an isolator. F1's own construction-grade profiles already meet EN 14024 and AAMA 507 performance criteria for curtain wall framing, in halogen-free flame-retardant resin rated Class B-s1,d0 to EN 13501-1. What is missing is not the thermal case. It is the stiffness case, at the sizes and load combinations the biggest unitized bays actually put on a primary member — and that is a genuinely different engineering problem than the isolator solved.
The mechanics expert's view: why modulus runs out before strength does
Mullion and transom sizing on a curtain wall is almost never governed by ultimate strength. It is governed by deflection, because glass and gaskets fail from movement long before the frame fails from stress. The industry convention limits frame deflection to L/175 of clear span up to about 4 m, and L/240 plus 6.4 mm beyond that — tight enough that on a tall unitized bay, stiffness (EI, modulus times moment of inertia), not strength, decides the section.
For a fixed architectural depth — and depth is exactly what a specifier is trying to shrink when they move off aluminum — the only lever left to cut deflection is E. EN 13706's standard grades top out at E23, about 23 GPa longitudinal. F1's own E40 sunshade-grade laminate, developed for facade shading fins and louvers, pushes full-section modulus to roughly 40 GPa by stacking multi-layer fabric rather than relying on straight unidirectional roving. That closes real ground. It does not close all of it. On the deepest unitized bays under design wind pressure, converted to a line load by w = p x b across the panel's tributary width, even E40 pultrusion needs a deeper section or a supplementary reinforcement bar to hold L/175 — and depth is the one thing the architect asked the specifier not to spend.
The load case compounds from there. A wind-blade spar cap is designed overwhelmingly against one bending axis; that is why F1's WE-G80 glass laminate and WE-C100 carbon laminate for spar caps are unidirectional-dominant, and why WE-C100 reaches 147 GPa tensile modulus and 1920 MPa tensile strength at 62 percent fiber volume — pure 0° fiber, doing one job extremely well. A curtain wall mullion does not get that luxury. It carries primary bending from wind pressure, torsion from eccentric loads at operable-vent hardware and asymmetric glazing pockets, and — on transoms, exactly as the isolator profile already does — sustained transverse load from glass dead weight at the setting blocks. Stack a straight unidirectional-carbon laminate into that section and the torsional and transverse numbers come back weak, because pure 0° fiber does very little for a load path that runs 90 degrees to it.
The curtain wall engineer's view: a mullion inherits a system, not a spec sheet
A spar cap ships to one customer, in one blade, under one qualification program. A curtain wall mullion has to slot into an existing ecosystem of pressure plates, gaskets, setting blocks, and — on four-side structural silicone glazed (SSG) facades — a bonded glass connection that loads the frame differently than a captured, gasketed one does. Every one of those interfaces raises a question the isolator profile did not have to answer, because the isolator sits buried inside the section, isolated from the weather and from direct connection hardware.
Move FRP into the primary structural role and it now interfaces directly with the aluminum pressure-plate and gasket system most curtain wall hardware is built around — which raises real bolting, bonding, and galvanic-isolation questions that a captive isolator profile never faced. Fire performance scrutiny escalates too: a thermal break buried inside a section gets one level of code attention; a primary structural member carrying the building's glazing gets another, which is exactly why F1's curtain-wall-grade resin is already formulated to Class B-s1,d0 per EN 13501-1 with ASTM E84 Class A available on request, and why any new higher-modulus grade has to clear the same bar, not a lighter one.
Durability follows the same logic. A curtain wall's design life runs the same 25-to-50-year order of magnitude as a wind blade's, under a comparable regime of cyclic wind and thermal load — which is precisely why the tension-tension fatigue discipline F1 already runs for WE-G80 (P95/95-percent-confidence S-N design lines to ISO 13003, not the 50-percent mean curve) is the right qualification path to carry over, not a new one to invent. A structural building component that has not been fatigue-qualified to that standard is not a credible primary member, whatever its static modulus claims.
The architect's view: every GPa is millimeters of sightline
An architect asking for a slimmer curtain wall grid is asking, whether they use this language or not, for more EI at the same depth — the deflection math in the mechanics section above is the actual constraint behind the sightline they are drawing. Aluminum's problem was never stiffness; it was thermal bridging and weight. FRP's problem, once the thermal case is solved, is exactly the opposite: get the stiffness up without growing the section, or the sightline win evaporates into a deeper mullion or a bolted-on reinforcement bar.
The rest of the architectural case is already proven at the fenestration scale and should transfer directly. F1's powder-coated profiles already match aluminum's finish system — AAMA 2604/2605, any RAL color, Qualicoat Class 2 — so a higher-modulus curtain-wall grade does not reopen the finish question. Pultrusion's die economics (a custom profile die typically runs 6 to 10 weeks) make bespoke mullion geometries viable at project volumes that would not justify a new aluminum extrusion die. And there is a design-language argument beyond economics: a visibly slender, corrosion-immune structural grid is not just an engineering deliverable, it is an expressive move for a facade, in the way exposed structure has always been for architects willing to let the frame read as part of the design rather than something to hide behind glass.
None of that argument works, though, if the higher-modulus mullion needs a deeper section than the one it is replacing. The architectural case and the mechanics case are the same case, stated in two vocabularies.
What F1 is developing: E50-class carbon/glass hybrid, braided pultrusion
The direction this points to is a laminate architecture built for the mullion's actual load case, not adapted from one that was built for something else. Two moves address it directly.
The first is carbon/glass hybridization rather than a straight swap to unidirectional carbon. Carbon fiber pultrudes to a tensile modulus range of roughly 120 to 500 GPa depending on grade, well above E-glass's 40-to-45 GPa ceiling, and F1's own WE-C100 spar-cap laminate already proves the process can hold 147 GPa at production scale. But published hybrid-composite research shows carbon/glass laminates retain a useful property pure carbon does not: once the carbon fibers reach their failure strain, the glass fraction keeps carrying load, giving the laminate a secondary, lower stiffness rather than a sudden loss of capacity. For a life-safety building structural member — as opposed to a component inside a wind blade's own separate certification regime — that residual load path after first-fiber-failure is not a footnote, it is a real design consideration.
The second is braided reinforcement architecture at the profile's outer layers, in addition to the unidirectional carbon/glass core. Braid-pultruded (or "pull-braided") composites add fiber oriented in the peripheral, helical direction rather than only 0° and 90°, and published testing on braid-pultruded rods is consistent on the trade-off: shear and torsional stiffness go up substantially, tensile and flexural modulus come down somewhat relative to an equivalent unidirectional laminate at the same fiber volume. That is not a strictly-better material — it is a different one, chosen because a mullion's torsion and edge-bearing demands at unitized panel joints and operable-vent hardware are exactly the load case a purely 0°/90° laminate under-serves.
Put together, the target is a laminate F1 is calling E50-class internally — not an EN 13706 designation, since the standard does not define a grade above E23, but a full-section modulus target in the same naming convention as the E40 sunshade grade, aimed at roughly 50 GPa. It is a carbon/glass hybrid core for longitudinal bending stiffness and post-carbon-failure residual capacity, wrapped in a braided outer architecture for torsional and transverse stiffness, qualified against the same fatigue discipline already proven on WE-G80 and WE-C100.
Where this stands today
This is a development program, not a shipped, certified product line. E50-class carbon/glass hybrid braided pultrusion for primary curtain wall structure has not completed EN 14024/AAMA 507 structural qualification or long-duration fatigue testing at building-envelope scale, and no project should spec it as a drop-in aluminum replacement yet. What is real: the fiber-volume discipline, the carbon pultrusion process, and the fatigue-qualification methodology all already exist in production at F1, proven on WE-G80 and WE-C100 for wind energy. The work ahead is adapting that proven process to a multi-axis structural load case and running it through the same rigor.
For engineers and facade consultants tracking this development, or interested in an early pilot mullion/transom qualification program, F1 Composite's KNOWHOW engineering group takes project inquiries through [the contact form](/contact) or the [KNOWHOW services page](/technology/knowhow-services).
