2-Arm vs 4-Arm Spider Fittings: Load & Aesthetics

On a point-supported glass facade, the spider fitting is the visible node where four glass corners meet at a single stainless-steel hub. But not every node on a facade is a four-way intersection. Along a vertical mullion, at a canopy edge, or at a T-junction, a 4-arm spider would carry arms that end in mid-air. That is why fabricators offer 2-arm, 3-arm and 4-arm spiders as separate products — and why the choice between them is an engineering decision, not an aesthetic afterthought.

This guide compares 2-arm and 4-arm spider fittings on the parameters that actually matter: how load reaches the glass, what happens at the drilled hole, where each arm count belongs on a facade, and what to ask a supplier before ordering. For the underlying dimensions and load ratings, see our guide to spider fitting load capacity.

What the arm count actually defines

A spider fitting is a cast or machined stainless-steel body with radiating arms, each ending in a bolted rosette that clamps a hole in a glass pane. The arm count matches the geometry of the node:

• 1-arm — a single rosette on a standoff, used at free edges, canopy corners, or where glass meets a wall.
• 2-arm (180°) — two opposite arms along a straight line, used on vertical mullions, horizontal transoms, and linear glass runs.
• 3-arm (90°/180°) — three arms for T-junctions and facade corners.
• 4-arm (cross) — four arms at 90° intervals, used where four glass panes meet at a single node on a regular grid.

Standard hole-to-hole centres are 200 mm, 210 mm and 220 mm. A typical 2-arm casting weighs about 1.5 kg; a 4-arm casting of the same series weighs about 2.2 kg. Most manufacturers publish an allowable axial load around 2800 N and an allowable radial load around 1400 N per spider body, but these values apply to the casting — the glass at the hole is often the controlling element, not the steel.

Load distribution: where the two-arm and four-arm fittings differ

The critical difference is how wind load reaches the glass holes. On a 4-arm node, four glass panes share the node and each rosette reacts against its own corner. The upper rosettes carry the dead load of the glass below; the lower rosettes are left free to move with wind deflection, which is the reason spider facades can flex without prying the bolts out of the holes.

On a 2-arm node, the fitting runs along a mullion and connects only two panes at a time. The same wind pressure over a larger tributary area funnels into two holes instead of four. Finite-element work on point-fixed glass consistently shows that stress around drilled holes is the controlling design limit, not stress across the broad surface of the pane. Researchers report maximum principal stresses at holes reaching 90–144 MPa under test loading, against a large-surface design value of roughly 84 MPa for 12 mm toughened glass. The fewer the holes sharing the load, the higher the local stress at each one.

This is why the glass specification often drives the arm-count decision. Point-supported facades require toughened or toughened-laminated glass, typically 10–15 mm thick, drilled before tempering, with holes at least 6 mm from the edge and a bolt diameter of roughly 2–3× the glass thickness. Standards call for a minimum safety factor of 2.5 on wind loads and 4.0 on dead loads for overhead glazing. Where a 2-arm node must carry a large panel, the engineer may specify thicker glass, SGP interlayer, or add a glass fin versus spider system behind the pane to share the load.

Aesthetic and visual trade-offs

The 4-arm spider is the iconic point-fixing detail: a symmetrical cross at every grid intersection, visible through the glass as the high-tech language of the 1990s atriums and airport terminals that defined the system. It reads as deliberate, engineered, and repetitive.

The 2-arm spider is deliberately quieter. Mounted on a slim vertical fin or a steel tube, it reads as a series of small linear clamps rather than a grid of crosses. Architects choose it for canopies, entrance walkways, and edges where a four-way hub would look heavy or out of place. Because the arms sit along one axis, the visible metalwork is also easier to align flush with the facade plane.

Button-head versus countersunk bolts

Within either arm count, the rosette bolt head changes the visual and structural picture. A button-head bolt leaves a small raised disc on the exterior face. A countersunk bolt sits flush into a conical recess machined in the glass, giving a cleaner face but concentrating stress more sharply around the cone. Test data shows the clamping effect of a button-head fitting reduces stress near the hole, especially on thinner panels, while a countersunk fitting offers little clamping relief. The flush look the architect usually wants therefore comes with a stricter glass-stress check — another reason to involve the engineer before finalising the specification.

When to specify which

Material grade follows the environment: 304 for interior and dry facades, 316 for coastal, poolside, and high-humidity exteriors, with passivation after fabrication. A supplier that cannot produce a mill test certificate traced to a heat number, a third-party load test report, and a dimensional drawing should be asked to explain why — or replaced. The same diligence applies to the rest of the supply chain; our overview of the curtain wall hardware supply chain covers documentation and sourcing for these components.

Common installation mistakes to avoid

Fabricators point to a short list of recurring errors that turn a correct specification into a field failure: using non-stainless bolts that stain the surrounding glass; leaving too small a contact area between the rosette and the glass; using undersized arms that bend under design load; omitting structural silicone between adjacent panes; and cutting or re-drilling spiders on site to fit a misaligned node. Each of these is invisible in a catalogue photo and expensive to fix after glazing.

Bottom line

Choose a 4-arm spider where four glass panes genuinely meet on a structural grid and the architectural language calls for the visible cross. Choose a 2-arm spider where the node runs along a mullion, edge, or canopy and the load must be carried by two holes rather than four. In both cases, the glass at the hole — not the steel casting — is the component that limits the design, so the arm-count decision should be locked with the structural engineer before drawings go to fabrication.