Skip to content
Ungewitter

All lessonsLesson 15

The flyer’s range of thrust

A flying buttress carries any thrust between two limits; it works if they meet the vault’s.

Lesson 6 met the flying buttress as Ungewitter introduces it: first of all a strut, wedged between the clerestory wall and the outer buttress. It is also an arch, with a weight and a line of thrust of its own, and like every arch of Lesson 13 it admits a whole family of lines. The book already knows this. The modern reading turns it into the flyer’s defining property: a flying buttress does not push with one force but with any force in a range.

A range, not a force

From the Lehrbuch

„Die Hauptforderungen an einen guten Strebebogen kann man dahin zusammenfassen, dass er nicht zu schwer ist, für gewöhnlich nur einen mässigen Schub ausübt, in besonderen Fällen aber einen bedeutenden Gegenschub leisten kann.“The main requirements of a good flying buttress can be summed up thus: it should not be too heavy, should ordinarily exert only a moderate thrust, but should be able to provide a considerable counter-thrust in special cases.Ungewitter 1890, p. 160
„Die Bogenform muss so beschaffen sein, dass sie alle für die vorkommenden Belastungen möglichen Drucklinien sicher beherbergen kann.“The form of the arch must be such that it can safely contain all the lines of pressure possible under the loads that occur.Ungewitter 1890, p. 160, Fig. 402

Maria Nikolinakou, Andrew Tallon and John Ochsendorf (2005) surveyed the flyers of early Gothic churches and analysed each by its lines of thrust. The steepest line that fits inside a flyer gives the least horizontal thrust it can exert: its passive state, the flyer merely resting between its supports. The flattest line gives the greatest thrust it can resist before it would snap upward: its active state. Between the two, the flyer carries whatever the wall asks of it. Ungewitter’s “moderate thrust, but a considerable counter-thrust in special cases” is this range in words. The 1890 method finds it from lines through the kern points of the end joints; the modern one lets the lines use the whole flyer.

Meeting the vault

The vault has a range of its own, from its least to its greatest thrust. The two are in balance only if they share a value: the flyer is adequate when its range overlaps the vault’s. In the figure the vault is a 10 m nave vault whose range is estimated from Table I’s load; the flyer is a quarter circle, Ungewitter’s normal form (Fig. 404). Change its depth.

H min 14 kNH max 20 kNwallbuttress
0.5 m

The flyer can carry 14–20 kN; the vault pushes with 27–33 kN. The ranges miss. flyer: 14–20 kN; vault: 27–33 kN; weight: 56 kN; 1890 kern points: 19–21 kN.

flyer
14–20 kN
vault
27–33 kN
weight
56 kN
1890 kern points
19–21 kN
Figure 1A quarter-circle flyer from the clerestory wall at 15 m to the outer buttress at 10 m, 0.6 m broad, sandstone. Tint: every admissible line of thrust; the arrows are its least and greatest thrust at the head. The vault’s range is an estimate for a 10 m vault with 100 kN on the wall.

A thin flyer cannot hold the vault: even its flattest line resists less than the vault’s least thrust, and it would be pushed up into a mechanism. Thicker, its range widens upward, mostly at its top end, and meets the vault’s. At first sight a heavier flyer is only stronger. But its least thrust rises too, and that is the thrust it always exerts, in calm weather, against a wall that may not need it. Ungewitter warns of exactly this: “the same large thrust that the arch exerts on the wall above also acts at its lower end, which is why an excessively heavy flying buttress also demands a particularly strong buttress” (Part II, p. 386).

The form decides

Depth is not the only lever. The second figure keeps the flyer 0.6 m deep and changes its form between the same two points.

H min 15 kNH max 28 kNwallbuttress
Flyer form

This flyer can carry 15–28 kN: a range 13 kN wide. flyer: 15–28 kN; vault: 27–33 kN; weight: 68 kN; 1890 kern points: 23–26 kN.

flyer
15–28 kN
vault
27–33 kN
weight
68 kN
1890 kern points
23–26 kN
Figure 2Three flyers of the same depth, 0.6 m, between the same two points: a quarter circle, a flatter segment of radius 8 m and a straight strut.

The flatter the flyer, the higher both ends of its range. The straight strut is the extreme. Any thrust along its axis finds a line inside it, so under Heyman’s assumptions its greatest thrust is unbounded; only crushing would limit it. But its least thrust is high too, because its weight must hang between its two ends on a funicular curve with almost no room to sag. Here it exceeds the vault’s greatest thrust: the strut would push the clerestory wall inward even with the vault in place. The curved forms keep the least thrust low and the reserve large, which is why the book takes the quarter circle or the half pointed arch as its normal form (Fig. 404).

Over two aisles

A five-aisled basilica must carry the nave’s thrust across two aisles. One flyer can leap both, with a radius larger than their total width (Paris, Ulm), but it is long and heavy, and its least thrust grows with it. The book’s usual way divides it into two spans (Part II, pp. 389–390): the upper flyer lands on an intermediate pier raised on the row between the aisles, the lower one carries the thrust on to the outer buttress, and the intermediate pier, loaded almost only from above, can stay slender. Where the lower flyer’s head is set a little higher, as in the choir of St-Ouen (Fig. 896a), its range meets the upper one’s head on: two shorter, lighter flyers in series instead of one long one. Open Cologne in the Studio and switch the choice of Doubled aisles in the elevation group to compare the three systems.

The whole section

In a basilica the flyer is one link in a chain: vault, clerestory wall, flyer, outer buttress, ground. The page 1890 against modern statics evaluates a Studio design link by link, by Ungewitter’s kern rule and by limit analysis: the geometric factor at the joints, the outer buttress as a block and cracked, and the flyer’s range against the vault’s.

Challenge

The link opens the basilica of Lesson 6 on the modern page. Does its flyer’s range meet the vault’s, and how large is the geometric factor of its outer buttress once cracked? Then widen the central vessel in the Studio and check again: which gives out first, the flyer or the cracked outer buttress?

Check a basilica by modern statics →