nervi hangars


Happy new year from ArchitectureFarm…

I’ve taken advantage of a quiet week to hunker down and get to grips with two fundamentally important Nervi projects–the Air Force hangars from 1935 and 1939-1942.  These were incredibly important projects for Nervi’s technical development, but they’re confusing–a week or so with sketchup has given me a better handle on them, but there’s still plenty of figuring out to go.

The images here never happened, of course; Nervi and Bartoli won the commission to design and build the first set of hangars (in the foreground) in 1935 in Orvieto, and the second set of hangars (background) in 1939 for Orbetello and Torre del Lago.  But it seems important to see the two types next to one another, because the intervening four years saw some important changes.  While the hangars are similar in form, they’re radically different in their details and their processes.


The Orvieto hangars were built entirely out of reinforced concrete, using a geodetic pattern that, as Mario Sassone and Edoardo Piccoli have shown, paralleled developments in geodetic design by U.K. aeronautical designer Barnes Wallis.[i]  (Nervi served in the Italian airship corps in WWI, so this link makes sense, but so far there’s no smoking gun showing the influence…but watch this space).  This structure was efficient, combining the deep, hollow shape of a concrete vault with space frame principles to create a ribbed roof shell that was relatively lightweight and hyperstatic; loads on the roof had multiply redundant paths they could follow, which made the roof impossible to calculate but super-efficient and robust even if some ribs were damaged in combat.  Nervi ended up testing large-scale models of the hangars to determine their behavior, beginning an important relationship with engineer Arturo Danusso that would prove vital throughout both of their careers.

These first hangars had problems, however.  They required a huge amount of formwork, since their shells had to be poured all at once to guarantee monolithic behavior.  Nervi’s design for almost continuous buttressing along the hangars’ sides and backs also burned through a tremendous amount of material and timber forming, and the dead weight of the roof–even with the efficiencies of the vault and the ribs–induced bending stresses that required a substantial amount of steel reinforcing.  Steel was, by 1935, a politically problematic material as Italy had few natural iron sources.  The Fascist government had begun its autarchic economic policy, which stressed nationally produced materials and that was on the verge of outlawing steel even for concrete reinforcement.  Finally, when the hangars entered service they suffered from thermal expansion issues.  After several ceiling tiles came loose from the shells, each hangar had to have expansion joints cut into their roofs.


In 1939, when the Air Force again ran a competition for new hangars, Nervi and Bartoli submitted a new iteration of their previous design.  This second version (top) was much taller, with a parabolic cross section instead of a cylindrical one–a more efficient shape that right away reduced the need for steel reinforcing.  But they also proposed lightening the roof further by pre-casting the rib elements on the ground and hoisting them into place by crane.  This let them use a modular scaffold that could be taken down (“disarmed” in the far more colorful Italian) and re-used multiple times, radically reducing the amount of timber.  The lighter roof also enabled a far simpler support system–six large buttresses instead of the forty or so required in the first iteration.
ImageEach precast truss was joined to its neighbors by a poured-in-place joint that connected short rebar loops cast into each truss element.  A handful of solidly cast arches provided additional stability to the system, along with a continuous edge truss that held the roof’s sectional shape.  The walls of the hangar were structurally separate from the overhead roof, allowing for considerable deflection.
ImageThe results were structurally efficient but also visually striking–Nervi commissioned the well-known Vasari Studio to photograph the hangars during construction, before the solid roof panels were put in place and when the roof structure, set against the sky, was almost uncannily ephemeral.
Famously, none of these hangars survive–the German army dynamited them during their northward retreat in early 1944.  Nervi was either devastated by their loss (“Recall’s Nervi’s son Vittorio (three of his four sons work with their father): ‘He wanted to crawl under those hangars and die with them.’–Time, Nov. 11, 1957) or secretly relieved that the Germans, and not his daring design, had been responsible for their collapse (an unfortunately unverified anecdote, but as the Italian saying goes, “whether it’s true or not, it’s a good story.”)

ImageDespite their untimely end these hangars were, I think, the crucial moment in Nervi’s career.  Between the two iterations, he made a turn from conventional, if extraordinarily competent, reinforced concrete engineering to something far more innovative.  In particular, his use of precasting looked ahead to his mastery of precast ferrocemento sections that would define much of his work in the 1950s.  And the astonishing vision of lightweight, even lacy concrete seemingly suspended in mid-air must have impressed upon him just how poetic this material could be.

ImageRe-building these things digitally has been a healthy bit of research on its own.  Not to give too much away, but my limited sketchup abilities neatly parallel the limitations of casting and pre-casting concrete on an Italian job site in the 1930s; instead of starting off with a geodetic network, I’ve actually laid up the roofs based on Nervi’s original drawings and built a 3-d ‘scaffold’ on which I’ve placed, scaled, and rotated each family of trusses.  Hunter S. Thompson once said that he’d typed out The Great Gatsby “just to get a feel for the thing,” and this is a pretty good exercise for figuring out what’s going on in buildings as well–especially when they’re made of mass produced elements and you can option-copy a single module a bunch of times.

[i] Mario Sassone and Edoardo Piccoli, “The Grid Structures of Nervi’s First Hangars in Orvieto: Innovation in Context.”  Journal of the International Association for Shell and Spatial Structures, Vol. 54, nos. 2&3.  159-168.