Cut off from the toe of Italy by the 3km wide Messina Strait, Sicily is visibly and symbolically beyond the reach of Rome, and its isolation has stifled economic development.
This could all be set to change, however. In November the Italian government will vote on whether to give the go ahead for construction of a 5km suspension bridge linking the island with the province of Calabria on the mainland.
Private sector consortium Stretto di Messina has been pushing to realise the project for 20 years; it was first proposed over three decades ago. The European Union has designated a fixed link between Italy and Sicily as part of its trans-European transport network, designed to foster European political, social and economic integration. And parliament approved the use of private finance for construction of the Lira 7,100bn ($3.3bn) project 'in principle' in 1998. It is not so much a question of whether the Messina Bridge will be built as when, predicts Stretto di Messina senior engineer Enzo Vulla.
Spanning between the two land masses presents a major engineering challenge. 'At 500m offshore you are in over 100m of water, ' notes leader of the Messina Bridge design team and partner in UK consultant Brown Beech, William Brown. A multi-span structure would be out of the question due to the extraordinary technical demands and cost of building foundations on the seabed.
And a single span suspension bridge on the same scale as Japan's Akashi Kaikyo - currently the world record holder with a main span of 1,991m - would still leave the Messina crossing with its feet in unfeasibly deep water.
Messina's towers will therefore be founded on dry land. This means a main span of 3.3km, more than 1,300m beyond the current record - and a dramatic rethink of deck design. 'Messina is about controlling the deck's aerodynamic behaviour by shape rather than brute strength, ' comments Brown.
Self-destruction of the Tacoma Narrows bridge in 1940 as a result of wind induced oscillation, or 'flutter', brought about a revolution in the design of suspension bridge decks. The first Severn crossing, designed in the 1950s by Freeman Fox, used what has become the universal standard for suspension bridge decks, the aerofoil shaped steel box girder. Its aerodynamic section 'flies' through crosswinds, reducing the turbulence that contributes to oscillation.
Much of the box girder's performance, however, relies on its inherent stiffness and resistance to bending and twisting moments.
Despite impressive performance characteristics, the aerodynamic girder has been pushed to its limits. As spans get longer, increased stiffness and thus added depth is called for - the Humber Bridge, with a main span of 1,400m, required 4.5m deck depth compared to 3m depth on the 1,000m Severn Bridge.
And as deck mass increases so too does the mass of a bridge's cables and towers.
'Of course, the aim is to keep the dead weight of the bridge as low as possible, ' Brown points out.
To avoid turning Messina into a colossus, with correspondingly colossal cost, Brown has deployed an aerodynamic phenomenon he discovered while playing in a wind tunnel three decades ago.
'People talk about flutter, but I had never seen and experimented with it, ' he recalls. Having set up a flutter model in the wind tunnel he then cut out longitudinal slots from the deck, effectively leaving a series of parallel lanes. On initiating air flow once again, 'the surprise was that I couldn't get the model to flutter however fast I ran the fans'.
Brown calculates that the proportion of void to solid surface should be roughly a third. 'As you increase the width of the gap, improvements in stability tail off to the point where no extra benefit can be gained. There's no formula to working out the precise ratio, ' he says. 'It's a case of suck it and see.'
Nor is there any scientific explanation of why slotting a bridge deck should so dramatically improve performance. But Brown offers a theory: Wind passing underneath the deck is able to escape upward through the voids, reducing the net lifting force. Meanwhile, the parallel strips present not one leading edge but several. Lift on the leading edge of a conventional deck produces a powerful rotating force around its centre line.
While surfaces on the windward side of the centre line in a 'multi strip' configuration will still produce lift, it will be partially counterbalanced by lift on surfaces to the leeward side. Overall, reasons Brown, rotational and thus torsional force is reduced.
Messina is to provide both a road and a rail crossing. Two decks, 13.2m wide, carrying three lanes of traffic apiece will flank a 9.4m wide central deck supporting twin rail tracks. The decks are separated by an 8m wide grillage - for the purpose of aerodynamics, effectively empty space. A further 4.2m of grillage cantilevered out from the edges of the deck gives it a total width of 60.4m.
The grillages are designed to act as emergency lanes, giving Messina a total theoretical carrying capacity for road vehicles of 12 lanes. It has a maximum traffic capacity of 4,500 vehicles/h in each direction and more than 200 trains per day. Yet deck depth is just 2.25m.
'That's roughly half the depth of decks used on conventional single bridges, ' notes Brown. Total weight per metre of deck is only 23.4t. 'We've created stability and at the right weight.'
Stiffness has been gained from the use of curved box sections which are inherently stronger than the more usual angular profiles, not to mention more 'slippery'. Compared to the Severn bridge's drag co-efficient of 0.6 or more, Messina measures 0.4.
Still further stiffening is provided by 4.5m deep cross beams at 3m centres. It is calculated the deck will remain stable in wind speeds in excess of 270km/h, though maximum lateral deflection of 9.9m can be accommodated. Meanwhile, the structure is designed to cope with a 0.65G seismic event.
The area is one of very high seismicity - comparable to San Fransisco, according to Professor Michele Jamialkowski who has acted as geotechnical consultant to the commission. In particular the profile of the shore on the Calabrian side is a cause of concern. It slopes away relatively steeply and has a number of discontinuities caused by previous seismic shocks.
Design of the massive anchor blocks will differ on each side of the bridge because of the varying ground conditions, but the scale of the temporary work will be huge in either case. Very deep excavations will be needed and because of the poor ground, foundations will be high complex.
'Efficiencies gained in performance of the deck have contributed to making other parts of the structure lean and economical. Brown has elected to keep costs down by designing relatively short, 370m high steel towers, at least in relation to the length of the main span. The steel structures are 370m tall.
This gives Messina a shallow sag to span ratio of 1:11. Suspension is to be provided by two pairs of 1.24m diameter galvanised steel cables - heavier and pricier than a saggier, lower-stressed configuration would require, Brown concedes. But he counters: 'Increasing sag would also increase the tendency to flutter. It would enable us to use lighter cable but would result in greater deflection. And we would have to have bigger towers, or course. That would make the seismic risk far greater.' AM