Flying Buttress

The structural problem flyers solved

Gothic builders wanted height. They wanted the nave to soar, the windows to fill with stained glass, and the wall mass to disappear. But masonry vaults — the curved stone ceilings spanning the nave — push outward as well as down. A barrel or rib vault generates lateral thrust at its springing line equal to 30–50% of the vertical load. For a wall to resist that thrust without buttressing, it has to be thick at the base and stay continuous to the foundation. The result is a Romanesque cathedral: heavy walls, small windows, low ceilings.

The flying buttress moves the resistance to thrust outside the building. A vertical pier is built at some distance from the wall — far enough that the aisle and clerestory windows fit between them. From the pier, an arched strut (the flyer) reaches inward and upward to meet the wall exactly at the height where the vault springs. The flyer catches the lateral thrust and transfers it down through the pier to the foundation. The wall in between can be hollowed out almost completely.

          ╱─── high vault thrust ───╲
         │                            │
         │ ⌒  ← FLYER (slanted strut) │
         │  ╲                         │
       PIER  ╲────► CATCHES THRUST    │
       (with  ╲     AT VAULT SPRINGING│
       PINNACLE)╲                     │
         ║       ╲ ◄── clerestory     │
         ║        ╲    windows fill   │
         ║         ╲   wall area      │
         ║          ◯                 │
         ║       AISLE WALL           │
         ║         ◯                 │
         ║                            │
       GROUND ════════════════════════│

The thrust line idea

An unreinforced stone arch carries load in pure compression along an internal "thrust line" — the locus of the resultant force at every cross-section. For the arch to stand, the thrust line must remain inside the masonry (it can't be pulled by tension because there's no tensile capacity). For the arch to stand stably, the thrust line must remain inside the middle third of every cross-section — otherwise the windward face goes into tension and the joint opens, even if the average stress stays compressive.

The flying buttress's job is to keep the thrust line inside the pier all the way down. The forces it sees are:

• Horizontal thrust H from the vault (≈ 250–400 kN per flyer for a 13th-century French cathedral).
• Self-weight of the flyer arch (≈ 50–100 kN).
• Self-weight of the pier (≈ 500–2,000 kN depending on pier height).
• Pinnacle weight at the top (≈ 50–200 kN).
• Wind load on the upper roof (variable).

The pinnacle is critical. Without it, the resultant force at the top of the pier — vault thrust H plus flyer self-weight, both directed roughly horizontally inward and downward — would angle outward as it travels down through the pier. By the foundation level, the resultant could pass outside the middle third of the pier base, opening a tension crack on the building-side face. The pinnacle adds vertical weight that bends the resultant back toward the centre of the pier. It's structural ballast disguised as decoration.

Buttress types compared

Type	Geometry	Era	Best for	Limitation
Pier (wall) buttress	Thick projection from wall, full height	Romanesque, early Gothic	Modest heights, thick walls	Blocks aisle windows, dark interior
Flying buttress (single)	One arched strut from pier to wall	High Gothic 12th–13th c.	Mid-rise vaults (20–30 m)	Single stack of thrust limits height
Flying buttress (double-tier)	Two stacked arched struts on same pier	Late Gothic 14th c.	Tall vaults (30–40 m), wind load	More stone, more weight on pier
Hidden buttress	Pier inside aisle roof, no exposed flyer	English Gothic, Italian	Low broad churches	Aisle wall absorbs lateral thrust internally
Lateral (transverse) buttress	Spans across an internal aisle	Cistercian, abbey churches	Two-aisle plans	Visually intrusive
Buttress with pinnacle vs without	Decorative spire vs flat top	All Gothic	Tall thin piers	No pinnacle: resultant exits middle third

Famous examples and their numbers

The flyers of Notre Dame de Paris (vault height 33 m, completed circa 1250) span roughly 15 m from the upper nave wall to their piers. Each flyer carries an estimated 350 kN of lateral thrust. The piers are about 1.5 × 4 m in plan and rise 35 m to their pinnacles. Several flyers were replaced over the centuries; the 19th-century restoration by Viollet-le-Duc rebuilt some original elements that had been removed in the 18th century, including pinnacle ornamentation lost during the French Revolution.

Beauvais Cathedral in northern France pushed Gothic proportions to their limit — the choir vault rose to 48 m, the tallest ever attempted. The first vault collapsed in 1284, twelve years after completion, partly because the flyer geometry was too shallow for the thrust generated. The rebuild added a tier of flyers and additional pier mass, but the cathedral was never finished — they ran out of money and engineering confidence.

Cologne Cathedral (vault height 43 m, completed 1880 to 14th-century plans) uses double-tier flyers with both upper and lower arched struts on each pier. The lower flyer catches aisle vault thrust; the upper catches nave vault thrust. Each pier carries roughly 2.5 MN of vertical load including pinnacle weight.

The Sainte-Chapelle in Paris (vault height 20 m, completed 1248) is unusual — its walls are almost entirely glass and yet it has no flying buttresses. The trick: iron tension chains hidden inside the wall above the windows take the lateral thrust, with conventional pier buttresses externally. It's an early hybrid system.

Medieval design rules

Master masons used proportional rules — no force calculations were done. Common rules of thumb:

• Pier base width ≈ ¼ to ⅓ of vault span.
• Flyer springing height ≈ ⅔ of pier height.
• Flyer rise to span ratio ≈ 1:3 (low arch).
• Pinnacle height ≈ ¼ to ⅓ of pier height.

These rules came from Vitruvius (1st century BC) via the Carolingian masons, refined through the medieval lodge tradition. They weren't always right — Beauvais being the famous counterexample — but they were good enough for buildings that have stood 700–800 years.

Modern analysis using limit-state methods (Heyman 1966, Como 2013) confirms that the medieval geometries are remarkably efficient. The flyers operate well within their compressive capacity; the failure mode is almost always foundation settlement or tensile cracking from differential movement, not crushing of the stone.

Common failure modes

• Pier crushing where the flyer meets it. The flyer concentrates its thrust on a small bearing area at the pier face. If the contact stress exceeds the stone's compressive strength (≈ 30–60 MPa for limestone), local crushing follows. Notre Dame's flyers landed on integral capitals carved from a harder limestone variety to mitigate this.
• Foundation settlement. Differential settlement of pier vs main wall foundations rotates the flyer geometry, opening a hinge crack at the keystone. This is why Gothic cathedrals are routinely surveyed and grouted; small settlements over centuries add up.
• Wind on the upper roof. Hurricane-force wind on the steep wood roof above the vault adds horizontal load that the flyers must resist. Late Gothic double-tier flyers exist partly because of repeated wind damage to single-tier predecessors.
• Loss of the pinnacle. French Revolution iconoclasm removed pinnacles from many cathedrals (decorative purpose was assumed). At Reims and Strasbourg, post-revolutionary surveys found the unloaded piers showing tension cracks — proving the structural function the medieval masons understood intuitively.
• Fire damage to the wood roof. The 2019 Notre Dame fire destroyed the timber roof and spire above the stone vault. The vault held — proof that the buttress system was independent of the roof — but two bays of the high vault collapsed where roof debris fell through, putting unplanned vertical load on the rib intersections.

Modern legacy

The flying buttress was abandoned for new construction by the 17th century when domes and barrel vaults replaced rib vaults. But the underlying idea — externalising lateral force resistance through inclined struts — survives in modern engineering. Cable-stayed bridges, exposed steel braced frames, and the Sydney Opera House shells all use the same logic: catch the thrust, redirect it to the ground via the shortest stiff path. The Gothic builders worked it out 800 years before the math caught up.