Its masterpiece is the lifting system, located between the two independent lifting spans, which features pairs of reinforced concrete hollow shafts supporting an ingenious butterfly–shaped, cross-leverage system, optimising both structural and mechanical designs.

Rouen’s historical harbour city on the River Seine is half way between Paris and Le Havre. For more than 30 years, urban planners have been suggesting that a new north-south highway connection would be required, with a river crossing at a location where maritime traffic would have to be maintained to allow large cruise-ships, as well as historical sailing ships, to reach the city centre.

The project is located half way between the historical city, dominated by the cathedral which was made famous by impressionist Claude Monet, and the harbour; hence a design competition was organised for the bridge concept. The winning team was led by Arcadis, with Michel Virlogeux as consultant, architect Aymeric Zublena and M&E consultant Eurodim. developed a new concept, with lifting towers located in between the two independent spans.

Maritime traffic from Le Havre towards Paris stops in Rouen, from which point only inland traffic is allowed to navigate upstream. As a matter of fact, all five existing bridges over the river in Rouen allow inland traffic only, while all three bridges downstream at Normandy, Tancarville and Brotonne, allow maritime traffic.

But it was decided that this new crossing, located downstream and referred to as ‘the 6th crossing’, had to allow seagoing ships up to 40000 DWT, to sail through it. Although this traffic is halted just 1.5km upstream by the Guillaume le Conquérant Bridge, it was considered mandatory that cruise ships, as well as ‘Armada’ large sail ships, should be able to navigate as close as possible to the historical city centre.

A high rise bridge was not acceptable because of its visual impact at this location, while a tunnel option was considered too expensive, hence a lift bridge was adopted.

At this location, the River Seine is 180m wide, but the basic clearance requirements were a minimum opening of 86m, and a minimum height of 55m when lifted, 7m when lowered. Typically, lift bridges require two high towers which carry the lifting equipment and guide the spans. To provide a vertical clearance of 55m, these towers had to be about 80m high, so they still have a significant impact upon the surrounding landscape.

Lift bridges are considered the most appropriate option among movable bridges when a long span is required. Bascule or swing bridges are more limited in this respect because of the amount of space needed to accommodate their moving decks. The longest span lifting bridge in France is the Recouvrance Bridge in Brest which has a lifting span of 87.5m.

But this pales into insignificance against some of the world’s largest lift bridges such as those in New Jersey where many waterways are designed to accommodate maritime traffic. The longest lifting span is 170m, belonging to the Arthur Kill Bridge, although this is no longer in operation. Another significant structure is the Gil Hodges Memorial Bridge, which has a main lifting span of 164m.

However these bridge decks are steel trusses which are relatively heavy and industrial-looking. The design of the Rouen bridge was a significant departure from this, in order to meet the owner’s expectations of elegance and fitness with the surrounding urban and industrial environment.

The design brief specified that the bridge should have two independent spans carrying three carriageways and a 2.5m-wide walkway; making each deck 18m wide. Furthermore, in order to make maintenance easier, it had to be designed so that each span was lifted independently.

The main challenge facing the team was the design of the towers, which required an extensive development process. The initial idea was to place the towers on each side of the spans, either independently or connected to form an arch, but these solutions appeared rather clumsy. Furthermore, the bridge alignment is at a skew angle with the river, and the towers must be perpendicular to the bridge alignment. Such wide towers would have led to a significantly longer central span and would have interfered with the water flow.

This led the team to develop a new concept by placing the towers in between the spans, with lifting equipment overhanging on both sides, symmetrically. Two lifting principles were considered: either cantilever beams connected to a vertical element located inside the towers, or cables attached to overhanging structures located at the top of the towers. In both case, the lifting motion was created by a combination of counterweights and winches.

Choosing between these two concepts was clearly a key issue, so the team fully analysed the advantages and disadvantages of both. The second option was eventually selected because of its simplicity, reliability and lower cost.

This solution was then refined, with the main change being that the cables were crossed, which avoided applying bending to the towers.

Each tower consists of two hollow concrete shafts, resting upon an elliptical caisson. In order to minimise interference with the river flow, this caisson is parallel to the river main direction, while the tower shafts are at right angle to the bridge alignment.

At the top, an elegant steel structure supports the pulleys on three parallel frames, designed in such a way that members carry normal forces only; its shape led to it being named the ‘butterfly’. Each span may be lifted independently through the action of winches located in the large caissons supporting the towers.

For architectural reasons, the conventional lifting design using lateral truss beams was abandoned in favour of a box beam. To minimise the weight of the spans – an important consideration for lift bridges - the spans are entirely made of steel, and the highway overlay is a 12mm-thick epoxy layer.

The cross-section of the deck evolved from a rectangular box with large overhangs stabilised by inclined struts, to a streamlined box with inclined webs. This evolution was dictated by wind engineering.

Each span is lifted by 16 cables, eight at each end, four each side of the roadway. Of those four cables, two are connected to a ‘dead’ counterweight, and the other two to a counterweight connected to a winch. Each end is therefore lifted by two winches, and a total of eight winches is installed, four in each tower. Winches are synchronised with each other.

This design is extremely reliable, because of its simplicity and because all the pulleys are independent, which avoids risks linked to even the slightest differential movements. The cables are coupled so that if a cable ruptures, its load is transferred to the adjacent one. Another benefit is that cable fatigue is minimised: load variations either do not exist for the cables connected to dead counterweights, or are relatively low for the cables connected to active counterweights.

When the decks are in the upper position, locking is achieved through the electrical motors of the winches - to keep things simple there is no mechanical locking. As a consequence, the cables apply a permanent uplift force to the spans, which significantly reduces the load these spans apply to their supports. This is one of the reasons why the deck had to be streamlined, in order to minimise the overturning moment resulting from wind action. By using the motors as brakes, it is possible to lock the decks at any height.

In order to control horizontal movements during lifting operations, spans are guided at each extremity by rollers that sit in a groove alongside the tower shafts. Preliminary design showed that wind would have significant effects and that a thorough analysis was necessary, including wind tunnel testing. Furthermore, these tests were necessary to take into account interaction between the two adjacent spans, in their various positions.

Two specific aspects had to be addressed. Firstly, when the spans are resting on their permanent bearings, wind forces create overturning and uplift forces which could potentially result in uplift at the bearing locations. This is aggravated by the fact that permanent loads are significantly reduced by cables uplift forces. Therefore overturning and uplift forces had to be carefully checked.

Wind-induced vibrations also had to be analysed by wind tunnel testing, and this led to the addition of tuned mass dampers on the decks, at mid-span and at both ends.

Large cruise ships of up to 40000DWT will be allowed to pass under the bridge. Although these ships travel very slowly, at approximately 2m/s, the piers have been protected from impact by four large gravity dolphins up and downstream of each pier. These are large, 20m-diameter concrete cylindrical boxes filled with gravel, which rest directly upon the river bed.

The construction contract was awarded in 2004 to a team consisting of Quille (Bouygues Group), Eiffage, Eiffel and Victor Buyck. Construction of the main crossing has now been completed, with the first test lifting carried out in April of this year. Approach viaducts are due for completion by spring 2008.

Caissons and gabions were built in situ in two steps: they were cast above water level, on the permanent piles, for caissons, or on temporary piles for gabions. Both were then lowered into place by means of a system of jacks and cables by subcontractor VSL. The tower shafts were cast in situ using jump-forms.

Steel sections for the deck were fabricated in two sites: at Eeklo in Belgium by Victor Buyck and at Lauterbourg in France by Eiffel. The decks and tower-top ‘butterflies’ were towed to Rouen on barges and lifted into place by very large cranes on pontoons.

Michel Moussard is head of the civil engineering & bridges department at Arcadis

Printer friendly version

Email this article to a friend