A good weekend of mostly last-weekend tourism in Sicily with a small but dedicated Academy crew. Us mild-mannered construction historians, especially of the modern ilk, don’t have much to offer on those kind of trips–I usually just gawk at the ruins (Greek variety, this time), ask dumb questions, and am ferociously happy to be so out of my usual element.
But Sicily does have one Nervi claim to fame. Between the more famous hangars at Orvieto and Orbetello (which I’ve blogged about here), he designed and built a pair of hangars at Marsala. These were simpler–instead of the diamond-shaped lamella roof systems of the other sets, these were relatively boring single-span arches, with buttresses on the sides. They weren’t published as widely, and architecturally they aren’t much to look at…
Until you zoom way in and notice that the arches are, in fact, made up of precast truss elements. And here’s where the Marsala hangars are actually kind of interesting and important.
There’s a kind of mysterious image of a hangar roof prototype being built in the Magliana yard that was published in The Works of Pier Luigi Nervi, which doesn’t jibe with either the Orvieto or the Orbetello hangars. You can see, for instance, that the arches are parallel to the span–not diagonal. But they’re definitely made of precast elements–vierendeel-like components in the foreground and more traditional truss shapes in the back. Interestingly, behind the arch is a prototype for the system that was eventually used for the Orbetello/Torre del Lago series. On that one you can see that the components are set on diagonals, and that they’re vierendeel-like elements. (You can also see the Square Colosseum at EUR to the right of the picture–a nice touch).
So, there are two things going on here. First, Nervi was experimenting with assembling long spans out of precast components–a response to the self-critique of the Orvieto series that focused on the inefficient formwork required for a monolithic pour of such a large roof. Second, he’s clearly wondering about the complexity of the lamella pattern–would it be easier to just span the hangars in a single direction? Or is the structural form of the original hangar design still valid even if the diagonals have to be made up of small units?
At Orbetello, Nervi proved that the geometric complexities could be handled relatively simply. But the Marsala structures came before those, and it seems possible that he wanted to master the technique of prefabrication first, and then deploy it in a more complex geometry. (This is pure speculation–the dates of the Marsala hangars aren’t totally clear, for instance, and are usually given as 1935, the same year as Orvieto. Just saying).
But if you look closely (OK, I’ll look closely for you), you can see that the arch prototype yielded some decisions–for instance, the type of truss work. The skin of the hangars is badly deteriorated, but that at least telegraphs the structure behind–and you can see that Nervi opted for the more conservative truss arrangement instead of the moment-resisting vierendeel. That ended up being his arrangement of choice on the subsequent hangars as well.
So while these aren’t as dramatic as the famous ones, they do represent an important turning point in Nervi’s work. They show that he proceeded fairly cautiously with precasting, and with thinking of long span structures as component based rather than monolithic. The Marsala hangars also have the distinction of being the only ones left–the Allied invasion of Sicily happened rapidly enough that the retreating Axis forces didn’t have time to demolish these, as they did with the ones on the mainland. From their appearance, however, time and neglect may well accomplish what German dynamite didn’t. There was no chance of getting in, and we had lunch in the city planned carefully enough that we didn’t have time to bail someone out for jumping a military fence. But these might well be a good excuse to go back to Sicily, which I wouldn’t complain about at all…