Oscar Niemeyer: The Cathedral is a Ring

Oscar Niemeyer: Sixteen Ribs and Two Rings

How the Cathedral of Brasília Was Actually Built

The Cathedral of Brasília, completed structure.


The Cathedral of Brasília is sixteen curved concrete ribs holding nothing up. There is no roof for them to carry. The glass between them is hung off their sides, not spanned across the gaps. The nave is underground. What stands above grade is a ring structure wearing sixteen faces, and everything interesting about the building is in how that ring was made.

Niemeyer's first scheme had twenty-one columns at about 130 feet. The built version is sixteen at roughly 100 feet, on a smaller base. That reduction was made with the engineer, Joaquim Cardozo, and it is the first thing worth knowing: the figure everyone photographs is a negotiated figure.

Oscar Niemeyer and the sixteen ribs under construction, c. 1959.

Two Precedents

Before the cathedral, two buildings set it up.

Church of São Francisco de Assis, Pampulha, 1943.

The Church of São Francisco de Assis at Pampulha (1943) is where the method is invented. Niemeyer drops slab-on-columns and builds parabolic reinforced-concrete vaults that are wall, roof, and room at once. One member, three jobs. The form came from hangars — utilitarian long-span shells — and he took it into a chapel. The Archbishop of Belo Horizonte refused to consecrate it for fourteen years. Researchers later had to reverse-engineer the structure from site measurement, because the original drawings were never found.

National Congress, Brasília, 1958–60.

The National Congress (1958–60) is where the engineer becomes visible. Two opposed shells on a thin continuous platform, so the cups appear merely set down on the esplanade. The inverted bowl was the hard one, solved partly by switching steel mid-job to a higher-strength bar newly available in Brazil, so the shell could stay thin enough to keep the silhouette.

Photo: Marcel Gautherot / Acervo Instituto Moreira Salles.

Photo: Marcel Gautherot / Acervo Instituto Moreira Salles.

Two Rings, Both Hidden From Outside

At plaza level, a tension ring about 197 feet in diameter ties the sixteen rib bases together. It is not a beam so much as a circular grid — four rings joined by beams, closed top and bottom by slabs. Its job is to take the outward thrust of the ribs so the foundations receive only vertical load. It rests on neoprene pads atop the wall-piers that form the perimeter of the sunken nave, and the pads let the ring move with temperature instead of cracking. From the plaza it disappears behind the pool edge. From the pews it is plainly visible — the band wrapping the room where the ribs land. Below are open caissons, roughly 28 inches in diameter with belled bases, averaging about 92 feet deep — sixteen groups, one under each rib.

Near the top, about thirty feet below the crown, a compression ring roughly 22 feet across passes through the ribs — invisible from anywhere. It is why the ribs can pinch together, separate again, and stand.

No rib is stable alone. The "two hands raised in prayer" that every guidebook offers is a description of a tension-and-compression system that happens to be beautiful.


The ribs from the plaza. Neither ring is visible from here.


The Ribs Are Partly Hollow

This is the fact the English-language literature almost never mentions, and it is the one that makes the building possible.

Each rib leaves the tension ring solid and slender. As it climbs, the section grows into a triangle — the V you see in the construction photos — and inside that enlarged section Cardozo placed caixão perdido, lost-box void formers left permanently in the concrete. The silhouette stays as large as Niemeyer drew it while the self-weight stays within what the ring and the caissons can carry. The ribs close toward each other at the compression ring, and above it they are solid again to the crown.

Formwork as the Real Drawing

Carlos Magalhães ran execution for Novacap, and his written account is the backbone of what we know. His crews built a shed on site and drew a column inside it at full scale. From that master template they cut roughly twenty transverse sections — a physical translation of a curve that existed on paper as geometry and had to exist on the ground as lumber.

Magalhães called the resulting wooden forms works of art. He was not being sentimental. Beneath them the shoring was tubular steel arranged in a fan around each rib.

Archival construction photo, c. 1959, digitally remastered.

Archival construction photo, c. 1959, digitally remastered.

The sequence: shoring first, from the base of each rib; then reinforcement; then the lost-box void formers; then the remaining steel. Rebar was placed after the forms closed. Concrete went in by crane in lifts of roughly thirteen feet, staged deliberately so every rib received the same volume at the same elevation as its fifteen siblings — an out-of-sequence pour would have loaded the unfinished ring asymmetrically. Forms were wetted continuously against the cerrado heat. Roughly twenty-eight days between a pour and its strike.

Concrete was batched on site from regional materials — cements out of Minas Gerais, Goiás and Mato Grosso, sands from Areias and Corumbá — under lab control. Steel was Torstahl CA-50; the densest zones carry seventy to ninety one-inch bars, butt-welded where there was no room to lap, with samples sent for testing. Cover was about three-quarters of an inch: normal then, thin by anything we would accept now.

The structure was declared finished on 21 April 1960 — the day Brasília was inaugurated. Then the funding collapsed with Kubitschek's term, the project was handed to the Church to restart it, Niemeyer went into exile after the 1964 coup, and the building was consecrated in 1968 with no roof on it.

The Glass Is a Sail

Sixteen infill panels, each a triangle roughly 33 feet at the base and 98 feet tall, about 1,500 square feet apiece — something like 23,700 square feet of surface in total, weighing on the order of 88 tons.

Glazing hung from the flanks of the ribs.

The detail that matters: none of it spans between the ribs. The support is three-dimensional steel trusses — space frames — fixed along the flanks of each concrete pier with steel bars, carrying glazing on both the upper and lower faces. The panels hang off the sides of the ribs. This is why sixteen columns could replace twenty-one without the openings growing past what could be bridged, and it is why the glass reads as an interval rather than a wall.

It also closes the structural loop. Cardozo calculated wind loading on the glazing and the columns specifically, which was advanced practice for 1959. The ribs carry no roof, so wind is the governing lateral case — and the thing catching that wind is 23,700 square feet of glass hung on trusses bolted to their flanks. The tension ring at the floor exists to take what the sail collects.

What you see today is not what opened in 1970. The cathedral opened with colorless glazing; Marianne Peretti's painted fiberglass went in during the renovation begun in 1987, when the piers were also painted white. The outer glazing has since been replaced twice, in 2000 and again around 2012.

The Room He Described

Niemeyer published his intentions in 1958, and they are unusually specific about the room rather than the silhouette.

The nave sits about ten feet below the surrounding ground. Ringed around that lowered floor are the chapels and the links to the sacristy and service rooms — the sinking is not only a level change, it is what generates a perimeter of secondary spaces that never show above grade. The baptistery stays outside the temple, as baptisteries did at Ravenna and Florence.

The descending entry ramp.


The tension ring, wrapping the nave where the ribs land.

The ramp is why the floor drops. In his words, the ramped entry deliberately takes the faithful through a space of shadow before they reach the nave, heightening by contrast the light he was after. He discarded the narthex, the atrium, and the heavy visible door — the whole traditional apparatus of arrival — and replaced all three with a narrow dark tunnel floored in black granite. His sketches work the sequence as compression and release: narrow to wide, dark to light.

The circle he justified twice over. He wanted a compact solution presenting the same purity from every exterior angle — and, he adds, the circular form also gives the structure a geometric, rational, constructive arrangement. That is the architect stating plainly that his free form is buildable because it is radially symmetrical.

Two Rings and an Architect

Take away the glass and the Peretti panels and the building is a simple structural idea. A tension ring at the base holds the feet of the ribs from spreading. A compression ring near the top holds their heads from falling in. Between the two, sixteen ribs lean on each other, and none of them could stand alone.

Outside, the rings stay out of the picture. The lower one sits at grade behind the pool edge, so the ribs appear to rise from the water with nothing holding them. The upper one passes through the ribs where they touch. Inside, the lower ring is the band circling the room above your head. Only from the pews can you see what holds the building together.

Niemeyer drew the curve. Cardozo made it stand by cutting the column count, hollowing the sections, and sizing the rings. Magalhães's crews built it from a full-size template, one rib at a time and sixteen times over. The circle made that possible: one geometry, one set of forms, repeated.

Most of Cardozo's calculations are lost. What survives is the building, and the building shows the structure clearly: two rings, sixteen ribs, and glass hung in the gaps.

Structural and execution detail drawn primarily from João Teatini Clímaco's project-history and 2023 structural essays, summarizing Carlos Magalhães's written testimony (2001) and Diogo Fagundes Pessoa's 2002 UnB thesis; glazing figures via Brito et al. Pampulha structure from Pereira, Buzar and Bezzerra, Paranoá n.15 (2015); Congress from Silva and Sánchez. Construction photographs by Marcel Gautherot and archival sources, two digitally remastered. Dimensions vary between popular and engineering sources — the widely repeated 230-foot figure describes the crown's overall spread, not the 197-foot structural tension ring.


Comments