Nervi Bridges: 6. Messina Straits Bridge

I've got one final bridge design by Pier Luigi Nervi to cover in this post and conclude this series.

At the end of the 1960s, Nervi was amongst engineers asked to propose a design for a bridge across the Messina Straits, between Italy and Sicily. The depth of water in the Straits meant that the bridge had to cross about 3km in a single span (although other designers, such as Leonhardt, were still proposing designs with deep-water piers at the same time - see the book by Richard Scott linked below).

Nervi's contemporary and compatriot Sergio Musmeci proposed a peculiar suspension bridge where the suspension cables are hung not directly from towers, but from cable stays which are in turn suspended from super-towers beyond the ends of the main bridge. Musmeci's idea included lateral cables either side of the deck to provide it with transverse stability.

Although this was an odd design, the proposal by Nervi was even stranger.

Nervi sought to achieve lateral stability by inclining the main suspension cables away from the deck, so that the deck hangers are no longer vertical, and the towers supporting the main cables are separated by a considerable distance.

The towers are hyperparaboloid concrete shells capped with enormous steel assemblies. They're restrained by stays to resist the incredible horizontal forces they would have to carry.

It's hard to imagine how Nervi thought this bridge might be erected. The main cables would have to be aerially spun with a conventional vertical sag, and then additional horizontal cables slung between them in order to pull them into the correct inclined alignment, at which point the deck segments could be fixed in place. I can't imagine how it could possibly work on this scale and the design is therefore, at best, fanciful rather than pragmatic.

The deck itself appears to be a trapezoidal concrete box which would be incredibly heavy and attract enormous wind loading.

On the whole, Pier Luigi Nervi was not a great bridge designer. Few of his ideas were taken up by others, largely because his attempts to maximise material efficiency were always at the cost of construction complexity, leading to greater cost overall. They also weren't always successful aesthetically.

In February I made a series of posts on the bridges of Eduardo Torroja, another concrete shell innovator, and whose bridges, like those of Nervi, are far less impressive than his better known designs. Was there something about the dedication required to become a master in one field that inhibited their intuition in a second field? None of the other great shell designers (Candela, Dieste, Isler, Hossdorf) seems to have left behind any great bridges either. If anyone has a counterexample, I'd be interested to hear it!

Further information:

• Messinacity: Il Ponte di Pier Luigi Nervi 1969 (Italian)
• In the wake of Tacoma: Suspension bridges and the quest for aerodynamic stability, Scott, 2001
• Pier Luigi Nervi: bridge designer [PDF], Manuel Cresciani, IASS Symposium 2007
• Pier Luigi Nervi and the Art of Building [PDF], Fausto Giovannardi, 2009

Nervi Bridges: 5. Opera Overpass, Ankara

Considering its highly unusual structural form and very site-specific arrangement, you might think that Pier Luigi Nervi's Ponte del Risorgimento, in Verona, was a one-off.

Surprisingly, you would be wrong, as Nervi elected to essentially repeat the design concept for a highway interchange in Ankara, Turkey, designed and built in the period from 1968 to 1973. The structure carries a highway across the Ataturk Boulevard, with associated slip roads.

It has the same basic form as the Verona bridge, a continuous concrete girder where the width of the top flange is held constant while the width of the bottom flange is varied in response to the degree of compression it carries. So the bottom flange is wider over the support positions, and the external girder web curves constantly to accommodate this.

The bridge deck is completely divorced from its supports, which are essentially twin T-shaped piers distorted to an extreme. The pier stem and crosshead morph into near-triangular masses, allowing the "crosshead" to be kept very slender at its end point. The skin of the piers is far more organically curved than in most of Nervi's column designs, but it all looks quite incongruous, a curious blend of the squat and the skinny.

The tightly curved interchange slip roads are quite different in form to the main spans. These sit on a series of bizarre supports which combine the attributes of a cone and a cruciform section, clearly intended to minimise use of concrete while maximising stability. Nervi had used cruciform section columns in several other structures, including the Corso di Francia viaduct, but I think the Ankara version is unique.

Further information:

• Structurae
• Archnet (I have used their images of the bridge, which are (C) Aga Khan Trust for Culture, with permission)
• Google maps / Bing maps
• Photograph of bridge pier

Nervi Bridges: 4. Ponte del Risorgimento, Verona

In 1961, the town of Verona held a design competition for a new bridge across the River Adige, partly to celebrate the centenary of Italian unification. Apparently, the results were uninspiring, and set to one side. Two years later, Pier Luigi Nervi was invited to prepare a design, which went on to be built, opening in 1968. The bridge was built by Edilbeton, who offered a lower tender than Nervi's own firm, Nervi e Bartoli.


The river is 131m wide, and Nervi proposed a three-span bridge with a central span of 62m and two symmetrical side-spans of 34.5m each. This partly echoed the existing upstream bridge, Ponte Catena, which had three roughly equal arch spans. It's a cellular reinforced concrete box girder, with its cross-section varied in response to the bending moment and shear force envelopes.

The soffit is curved to enhance the girder depth at the bridge piers, where the hogging bending moment (and also shear force) is greatest. The bottom flange thickness also varies, increasing at the piers to carry the resulting increased flange compression force. These are both concepts common to bridges of this type.


However, Nervi also varied the width of the bottom flange in response to the moments, so that is wider at the piers, where greater compression capacity is required, and narrower at midspan. The top flange is kept at a constant width, so Nervi created a bridge where the slope of the side walls varies continuously.

At the piers, the cross-section is of a trapezium with the bottom flange wider than the top, while at midspan, the reverse is true.


There is a hydraulic as well as structural logic - the side wall of the deck juts out above the piers to form something like a cutwater, helping to divide flood flow and creating a downward thrust under flood. The shaping at midspan may reduce hydraulic drag under flood conditions.

To the modern engineer, the varying girder depth and flange details are sufficient for an economic design - the curved side walls, shaped as hyperbolic paraboloids, require expensive varying formwork, and make detailing of the internal reinforcement more difficult.

Personally, I'm not sure that the visual effect justifies the effort. It's an intriguing design, but not so startlingly attractive as to make the cleverness seem the result of necessity rather than whim.

Further information:

• Google maps / Bing maps
• Wikipedia (Italian)
• Structurae
• Pier Luigi Nervi, bridge designer [PDF], Manuel Cresciani, IASS Symposium 2007
• Pier Luigi Nervi: Architecture as Challenge, Silvana Editoriale, 2010

Nervi Bridges: 3. California

Following the Corso di Francia viaduct, Nervi designed a 160m span roof support structure for the Burgo Paper Mill in Mantova in 1961 which was clearly bridgelike, if not actually a bridge. The picture (courtesy of Michele Molinari on Flickr) shows one half of it, and it has the same form as a suspension bridge (the half not shown in the picture is symmetrical).

In 1962, Nervi was asked to prepare bridge designs for the Kaiser Steel Company in San Francisco. He developed both steel and concrete designs for an elevated highway, pictured below:


This concept shows Nervi's usual concern with the shaping of the bridge pier legs both to achieve structural efficiency and also improved appearance. The portal legs are more attractive than the T-shaped supports for the Corso di Francia Viaduct, but it's certainly not an especially innovative design.

The following year, California's Department of Public Works invited Nervi to submit a proposal for a bridge at San Mateo Creek, some 550m long and up to 76m tall.

It's worth comparing this to Nervi's 50m tall River Tenza Viaduct design of eight years before, with its peculiar reinforced concrete trestles. Perhaps the experience at Corso di Francia had allowed Nervi to determine that elevated viaducts didn't require oversized frames to provide their stability, and that a more conventional bridge pier arrangement could be made to work.

However, the San Mateo Creek still has hangovers from the early days of concrete design, such as the use of hinge joints at the base of the pier legs and between the two crossbeam elements. Hinges were normally adopted in reinforced concrete either to make the design calculations easier, or to facilitate differential movement e.g. caused by mining-induced ground settlement. It's hard to see why they were adopted here, in a seismically active state, since Nervi had dispensed with them for the Corso di Francia Viaduct earlier.

The bridge deck illustrated also seems to be the V-shaped beams Nervi had used at Corso di Francia, although the picture is unclear.

A recent book on Nervi's works, Architecture as Challenge, lists Californian proposals by Nervi at San Mateo Creek, Fort Sutter Viaduct and Embarcadero Viaduct, and pictures a further design with giant X-trusses which might be a double-decker viaduct.

Further information:

• Structurae (Burgo Paper Mill)
• Great Buildings Online (Burgo Paper Mill, including drawings)
• Pier Luigi Nervi, bridge designer [PDF], Manuel Cresciani, IASS Symposium 2007

Nervi Bridges: 2. Corso di Francia Viaduct, Rome

In 1956, Pier Luigi Nervi became involved in a number of structures for the 1960 Rome Olympics. Most notably, this included the two indoor sports arenas, Palazzo dello Sport and Palazzetto dello Sport, and the outdoor stadium at Flaminio. It also included the Via Olimpica viaduct, now called the Corso di Francia Viaduct.

The Corso di Francia runs right through the middle of the Olympic Village. Apparently, it divided the housing for male athletes, to the east, from that for female athletes, to the west, although any notion of propriety was ruined by the use of the viaduct by Italian men to spy into athlete's bedrooms (I don't know how true this story is).

The road had to be elevated to allow free passage of athletes and spectators throughout the site, and the use of T-shaped piers by Nervi may have been an attempt to maximise space at ground level and reduce visual obstruction (see picture left - all colour photos are courtesy of Nicolas Janberg at Structurae).

While the structure is more conventional than the two early bridge designs I featured last time, it still has its share of idiosyncrasy.

The columns have an unusual varying section, from rectangular at the top to a cruciform section at the base. This enhances stability while also adding to the visual interest, and Nervi frequently provided varying columns in his designs. The Palazzetto dello Sport has some, as does his roughly contemporary UNESCO building in Paris.

The beams are 16m long, precast and prestressed, a simple and economical solution for bridge viaducts that seems unusual for Nervi, who was not a great user of prestressing. More unusually, they are V-shaped in cross-section, with a very narrow bottom flange, and with the top flanges placed in contact. The overall cross-section of each viaduct therefore consists of a "corrugated" or V-wave profile, which is essentially similar to the approach Nervi had taken on a number of roof structures, including the Palazzo dello Sport.

Nervi's design for the viaduct is featured in Fritz Leonhardt's Bridges. Leonhardt notes that the bridge did not age well, with water seeping through joints and down the piers (I imagine the deck was discontinuous, as on the Midland Links viaducts in the UK, which have been notoriously plagued with leakage and associated concrete degradation).

Leonhardt also criticises the adoption of column and crosshead piers to support the deck beams. Although this is the most straightforward solution to build, it is visually more congested than the alternative of having the crosshead integral with the deck and within its depth. The latter approach, however, requires temporary propping of the deck beams and construction of a more difficult joint detail.

Further information:

• Google maps / Bing maps
• Structurae
• Pier Luigi Nervi's works for the 1960 Rome Olympics [PDF], Iori & Poretti, Actas del Cuarto Congreso Nacional de Historia de la Construcción, Cádiz, 2005

Nervi Bridges: 1. Early bridges

Pier Luigi Nervi (1891-1979) was one of the greatest Italian structural engineers of the twentieth century. Indeed he was one of the greatest designers anywhere in the world, a pioneer of large-scale reinforced concrete roof structures, highly successful as the engineer, architect and builder of his own works.

His many astonishing designs include St Mary's Cathedral in San Francisco, the Palazzo and Palazzetto del Sport in Rome (the former is pictured, right, courtesy of Jeroen Meijer on Flickr), the Papal Audience Hall in the Vatican, and exhibition buildings in Turin.

A hallmark of his approach was the use of ribbing to stiffen thin-shell construction, leading to geometrically intricate vaulting reminiscent of the Gothic architecture that Nervi undoubtedly admired.

Unlike his contemporary and compatriot Riccardo Morandi, Nervi is less well known for his bridge designs. Nonetheless, he produced enough of interest that I'm going to cover them in a short series of posts here.

Nervi designed several reinforced concrete bridges early in his career. I've found references to:

• Bridge over River Cecina, Pomarance, Pisa (1920-22, demolished 2001)
• Bridge over Pescia Creek, Pistoia (1922-23)
• Widening of Bisenzio Bridge, Prato, Florence (1923-32)
• Fosso Biedano, Rome (1934, unbuilt)
• Bridge at Arno (1945, unbuilt)
• Reno, near Sasso Marconi, Nologna (1951, unbuilt)
• River Tenza, Salerno (1955, unbuilt)

The proposed bridge at the Fosso Biedano was a reinforced concrete trestle viaduct. The A-frame legs stiffen the curved deck against centrifugal forces, while the three levels of bracing serve to reduce the buckling length of the frame columns.

I'm not aware of many concrete trestle bridges which were built: this was a design essentially harking back to the great timber trestle viaducts of the 19th century.


Two decades later, Nervi's plan for a bridge over the river Tenza was equally unconventional. According to The Works of Pier Luigi Nervi, the designer's main challenge was to reduce the amount of scaffolding required for a viaduct which was some 50m above the valley floor.

The 319m long viaduct had trestle supports at a maximum 50m interval, four with identical geometry and hence the option to reuse the falsework and formwork. These trestles support main concrete cantilever beams, which vary in width, having a thicker web above their support positions, where the shear is highest.

The cantilever beams in turn support precast span beams, lifted into position from ground level. These span only 16m, sitting on roller bearings on their half-joints.

In 1947, Nervi designed the roof structure for the Conte Trossi Wharf in Genoa, which if not a bridge, was certainly bridge-like. The relationship of the reinforced arch to the "deck" slab above is more than a little reminiscent of Robert Maillart's bridge design at Tavanasa.

In 1959, Nervi published a critique of five bridges (Cinque ponti), discussing the Salginatobel Bridge, Rodenkirchen Bridge, Golden Gate Bridge, Elbe River Bridge, and a bridge project by Paolo Soleri. Nervi used his critique to contrast two possible methodologies for structural design:

1. "designers, following the statical, constructive and economical needs, and within the margins of freedom in defining structural and formal details, which they always and anyway concede, try to express their own aesthetic sensibility."


2. "designers start from a purely formal conception, very close to what sculptors could have, and then begin to improve it, with the sub-conscious thought that the 'calculator' will manage to make it stand and the constructor will be able to build it."

Evidently, Nervi supported only the first of these choices.

Further information:

• Wikipedia (Italy) (Ponte sul fiume Cecina)
• The Works of Pier Luigi Nervi, Pier Luigi Nervi, The Architectural Press, 1957 (Biedano, Conte Trossi Wharf, Tenza)
• Pier Luigi Nervi, Ada Huxtable, Mayflower, 1960 (Conte Trossi Wharf, Tenza)
• Pier Luigi Nervi, bridge designer [PDF], Manuel Cresciani, IASS Symposium 2007