New Removable Bridge over the Bay of Cadiz Manuel ESCAMILLA Civil Eng./Assoc. Professor ACL/University of Cadiz Cadiz, Spain
Marcos MARTIN Managing Director Civil Works Department Sevilla, Spain
Julio CAYETANO Technical Assintance Chief Ginprosa Madrid, Spain
mescamilla@acl-estructuras.com
mmgomez@fomento.es
j.cayetano@ginprosa.es
Marta SACALUGA Civil Engineer Ginprosa Madrid, Spain
Gonzalo OSBORNE Civil Engineer ACL Cadiz, Spain
Veronica VEGA Industrial Engineer Carlos Fernandez Casado Madrid, Spain
m.sacaluga@ginprosa.es
gosborne@acl-estructuras.com
vvega@cfcsl.com
Summary The construction of the new Bridge over the Bay of Cadiz is necessary to break the secular isolation of the city. This brand new infrastructure must also be respectful and compatible with the local shipyard industry, which has many high-tech facilities inside the bay, employing nowadays more than 6000 workers. Regarding the possibility of the construction of ships or offshore structures with extraordinary dimensions, the new bridge will have a removable span, with 140 m horizontal clearance and no vertical clearance limitations. This original structure, which is in a very advanced stage of construction, will have a single 150 m long steel span with variable depth (between 3 and 8 m) and a 33,2 m wide orthotropic plate. The structural detail design and construction of the bridge is presenting many challenges in relation to its dynamic behaviour, fatigue control, durability and quality control of the manufacturing processes. Keywords: Removable Bridge, orthotropic plate, lifting, Cadiz, welding.
1.
Introduction
Cadiz, the oldest occidental city, has a 3000 year history strictly conditioned by its natural environment. Being completely surrounded by the sea, having a strategic location, near the Strait of Gibraltar and with the natural seaport configured by its bay, Cadiz has developed an incessant maritime activity, constrained however, by the restricted connections with the nearest land. The project of the New Cadiz Link has to deal, consciously, with very sensitive environmental, social and economic boundary conditions, which deeply influenced and even determined some of the main features of the final selected design. The site through which the east end of the main bridge is being developed is an industrial area, administered by The Cadiz Bay Port Authority, and strongly related with the shipyard and offshore industry. Some of the most important European companies in this sector have developed their projects in this place for decades. This gives continuity to the old naval and sailing heritage that has characterized this city. The expected improvement of the roadway and tramway accessibility and its subsequent economic and social benefits wouldn’t of benefit if the new bridge imposed some restrictions to the navigation of big vessels, and subsequently to the shipyard industrial facilities. This aspect was taken into account in order to determine the location, dimensions and proportions given to both substructure and superstructure of the new bridge: - The bridge won’t interfere with the operations at the Cadiz Bay Port, neither in its main facilities (situated in the city, outside the new link area) nor in the grain and mineral dock, sited just below the east half of the cable-stayed deck. The eastern tower is being constructed inside that dock, but away from the operational area, and the deck is so high at this point that the port cranes can pass below it. - The main span (the central span of the cable-stayed bridge), will be located above the main vessel shipping canal of the Bay, and with its 540 m length will respect the width of the existing waterway. The vertical clearance below the deck at this point is variable, with a
-
maximum of 69 m above the average sea level (66,35 m above high water level), which will suffice for the comfortable transit and manoeuvring of the existing world fleets of merchant and naval vessels [1]. In addition, and regarding the possibility of construction or operation of vessels or offshore structures with higher dimensions, the new link will have a 150 m long removable span, with no vertical clearance restrictions. This paper describes the main features of this removable span, of which completion is almost reached.
Fig. 1: Longitudinal view of the cable-stayed deck at its main span The 3092 m long and 33,20 m average wide deck of The New Bay of Cadiz Bridge, has been planned as a conjunction of 4 different deck typologies: the west approach bridge has 9 spans of variable length (from 55 to 150 m), and it will be composed by a continuous composite deck structure to be built by incremental launching over the sea, and a removable 150 m long steel span. The central structure is an 1180 m long cable-stayed bridge, with a span distribution of 120 m + 200 m + 540 m + 200 m + 120 m. Finally, the east approach bridge has a 1183 m long continuous prestressed concrete deck, which has 75 to 32 m long spans from the 14th pier (of a total of 36) to the second abutment.
2.
Planning and designing the New Bridge
2.1 Alternative proposals In order to make feasible future developments in the shipyard and offshore industries inside the Bay, the Spanish Government decided to provide a new large ships navigation channel, with no vertical clearance limitations, below the western approach viaduct. Two main options were then discussed: the first one was to build a “conventional� 185 m long movable span, with 245 m total length and 140 m clear width. The other one, which is the final selected option, was a simple-supported 150 mlong removable deck. Pros and cons for each alternative were analyzed, focusing in life-time costs, construction difficulties and operability and maintenance requirements. The main advantage of the movable span was the reduced manoeuvre time needed to open and close it, which would allow the traffic to be re-established within a few hours after having interrupted it. However, the opening needs for this particular facility are supposed to be circumstantial rather than usual, so the operation time became a less important factor than the others mentioned. As a result, the economy, lightness, feasibility and reduced maintenance needs of the removable span, tipped the balance in its favour. The need for such a solution turned into an undoubted necessity when, in 2008, having begun the foundation works on site, it was decided to include a two tramway lane in the deck, making it necessary to design it wider and suitable to resist higher dead, live and accidental (seismic) loads, preventing at the same time an increase of its self-weight.
2.2 Design requirements and criteria. The developed design had to fulfil the basis of design defined in European and Spanish technical normative, and relative to both roadway and railway actions. The analysis of the dynamic and vibration effects and the prevention against fatigue failure became crucial in the design process, either for the global structural behaviour of the whole span or for the local and detailed design of the different elements (orthotropic deck, longitudinal and transversal stiffeners, etc.) and welded joints. The tramline was included eccentrically into the deck, occupying the left side. This had an important influence in the modal and deformational behaviour of the deck, because of the asymmetrical distribution of permanent and live loads. Accordingly with the normative requirements, the dynamic effects of the railway loads were taken into account, limiting the vertical acceleration, the maximum deformations (global and local, in all directions) or the torsional twist of the deck. Fig. 2: Longitudinal view of the New Removable Bridge Finally, the transportation and lifting manoeuvre required specific assessments and local adaptations to concentrated reactions. 2.3 Structural design process. As discussed, the main challenge faced was to design an economic and easy-to-build solution, able to house 4 lane roadway and 2 lane tramway traffic, and without delaying the scheduled construction plan, already started. Carlos Fernandez Casado (CFCSL), GINPROSA and ACL, former companies of the joint venture in charge of the project and work surveillance at site, decided then to establish a design process based on the “Two Man Rule�: two independent project teams would work at the same time to define the solution. The leading team was formed in CFCSL main offices in Madrid, and the second one was placed at the site facilities. The work developed was coordinated but at the same time independent, using different structural models and software packages to process them, and channelizing the communication exchange. This system proved its efficiency as an internal and dynamic mutual supervision, which led to a final design that passed a further owner’s review without any objections, preventing an expensive delay in the scheme planned. Since the very beginning of the design process, the contractor (a Dragados and FPS joint venture) played a key role in the decisions adopted; this circumstance led to a final design ready to be built up. A complete finite element model was Fig. 3: 3D view of the complete finite element model developed, in order to properly assess the dynamic and vibration behaviour of the structure, and to control the interaction between flexural and torsional effects. It was very useful also for the further detailed design, the verifications done to check the intermediate construction phases and the final transportation and lifting process. 2.4 General description of the deck. The removable deck is an S 355 steel structure, 150 m-long, 33,2 m-wide and with a longitudinal variable depth (shifting linearly from 3 m at the edges to 8 m at the centre). The deck cross section
has been designed as a 3-cells closed box, configured by a 10 m-wide central box (which longitudinal depth varies) and 2 lateral 11,6 m-wide cantilever boxes with a constant longitudinal depth and a transversal depth varying from 0,7 m to 3 m. An orthotropic upper plate has been designed to hold the 4 cm pavement layer and the tramline equipment; its thickness varies from 18 mm up to 30 mm. There are transversal stiffeners each 3,75 m, constituted by tee-shaped welded crossbeams in all boxes, connected to diagonal tubular sections by gusset plates, in order to prevent distortion and restrict transversal vibration within admissible levels. The longitudinal stiffening of plates is materialized by trapezoidal cross section stiffeners in all cases (orthotropic plate, webs, bottom flange and lateral flanges), although there is a need to reinforce the longitudinal stiffening of the bottom flange at the midpoint of the span, because of the deviation forces due to the variable depth; at this point, trapezoidal and simple plate stiffeners are used. With 3800 tonnes of total self-weight, 41 % Fig. 4: Internal view of the central box of the corresponds to the orthotropic plate, 17 % to the deck lateral bottom flanges, 16 % to the central bottom flange, 12 % to the webs (all including the corresponding longitudinal stiffeners) and 14 % to the crossbeams and diagonal tubular stiffeners. The connection with the substructure will be provided at both ends of the girder by a double corbel system, constituted by two box cross section steel corbels supported by elastomeric bearings placed over concrete corbels in the piers. This design allows the lifting up and lateral movement of the deck, so there is no need for retractable devices whose reliability and maintenance requirements would have been difficult and expensive. The corrosion protection of the steel has been entrusted to high durability protective paint systems (according to ISO-12944-1 [2]), suitable Fig. 5: 3D drawing of the deck end at piers to the high atmospheric corrosion level C4 (moderate salinity coastal area, according to ISO-12944-2 [3]). The exterior protective system consists of a 3 layer, 325 microns dry film thickness coating, with a top 125 microns DFT highperformance siloxane-epoxy. This product was selected because of its abrasion, chemical and saltfog resistance, flexibility and versatility, together with its colour and brightness stability (see [4]). 2.5 Detailed design Once the principal design process had concluded, a master drawings collection was edited. At the same time, the contractor and the steel structure manufacturer planned a fabrication procedure, and determined the workforce and technical resources at disposal. Later on, the detailed design began, with the design team working together with the contractor technical staff, whose technical queries were analysed and in most cases, taken into account. All of this process was directed by the owner’s technical personnel, throughout numerous meetings. As a result, more than 800 shop drawings were made-up.
Fig. 6: Detail of welding connections on the central box girder
Fig. 7: Finite Element model of the gusset plate connection within the transverse stiffening
An exhaustive assessment of the welded joints was carried out, focussing on their feasibility, fatigue resistance and durability. This has a particular importance in the orthotropic plate, because of its sensitivity, density of the stiffening system and responsibility of the welds. The previous experiences designed by Carlos Fernandez Casado (see [5]), the technical state-of-the-art (see [6]), and the recommendations given by the European Normative [7] were attended to achieve the final design. It was decided to weld the longitudinal stiffeners to the upper plate making a complete penetration joint, due to the high level qualification, experience and quality control abilities of the steel manufacturer (Dragados Offshore), which has been involved in many outstanding structural projects (the Oresund Bridge, as an example, see [8]). The welded connection between the tubular diagonal elements and the corresponding gusset plate was also studied; a skewed cut was first designed at the ends of the tubes, but in order to facilitate the cutting and welding process, a transverse cut was modeled and thus selected. The location, size (to a proper weld around) and corrosion protection of the cut-outs, the airtight sealing of inaccessible areas to the surface preparation and protection, the design of man holes for safety and maintenance requirements and the transportation and maneuvering operations were details deeply discussed.
3.
Construction and Quality Control
There are two independent quality control teams working on site at the same time: the first one is the internal Q.C. department of the steel manufacturer, and the second belongs to the work surveillance joint venture. Both complementary and cross-check inspections and testing are being developed. As discussed before, the complete penetration joint which connects the longitudinal stiffeners to the orthotropic upper plate was a key detail, so in order to prevent weld defects, some test specimens were first carried out, calibrating the influence of the surface preparation, welding angular orientation and gap width. As a result, a specific welding procedure was designed, a previous abrasive surface blasting was required, and an intensive non-destructive testing campaign was developed since the very beginning of the fabrication (100 % of the welds was firstly inspected). The satisfactory results obtained led to a reduced inspection ratio when the 15 % of this type of weld was finished. The steel deck will finally require about 50 km of welds and a total 5 m3 amount of welding consumables. 34 % of the total welded length corresponds to the longitudinal stiffeners weld to the orthotropic plate and 40 % of the mentioned length is being made by fillet welding. The figure 8 shows the main welds being done, and the following table 1 describes the type, position and welding techniques being applied. 70 % of the welds are being done by totally automatic devices (50 % of submerged arc weldings and 20 % of gas metal arc weldings), and the resting 30 % are semiautomatic flux cored arc weldings.
Fig. 8: 3D Cross Section view, naming the different welded joints
Table 1: Description of welded joints, according to the figure above CODE ELEMENTS WELDED Longitudinal stiffener to 1 deck plate Longitudinal stiffener to 2 bottom and lateral plates Web of crossbeam to 3 stiffener and deck plate Web of crossbeam to its 4 flange Deck/Bottom/Lateral 5 plates 6 Web to deck plate Diagonal brace to gusset 7 plate/crossbeam 8 Crossbeam to web 9 Lateral plate to web 10 Web to bottom plate Exterior web to 11 crossbeam
TYPE
POSITION
WELDING PROCEDURE
Complete penetration
Flat
Automatic (GMAW-MAG)
Fillet welding
Flat
Automatic (SAW)
Fillet welding
Flat / Vertical
Semiautomatic (GS-FCAW)
Fillet welding
Flat
Automatic (GMAW-MAG)
Complete penetration
Flat
Automatic (SAW)
Partial penetration
Horizontal
Semiautomatic (GS-FCAW)
Fillet welding
Vertical/Over head
Semiautomatic (GS-FCAW)
Fillet welding Complete penetration Complete penetration
Vertical/Over head Over head Horizontal
Complete penetration
Horizontal
Semiautomatic (GS-FCAW) Semiautomatic (GS-FCAW) Automatic (GMAW-MAG) Automatic (GMAW-MAG) Semiautomatic (GS-FCAW)
The deck has been divided into 11 modules (3 of them correspond to the central box girder and 4 to each of the lateral ones) to be welded indoors at the factory facilities. Once the welding works is finished, each module is transported to the painting facilities, from where it is moved to the final linking area, placed outdoors. The welding works there are being developed with special provisions in order to prevent defects, like preheating devices in the welding affected areas. The internal transportation of the central box plate girder (50 m long x 10 m wide x 8 m height, 700 t weight), from the welding facilities to the painting area, required a specific structural study. A detailed finite Fig. 9: Internal view of a finished element model was developed to assess the effect of the cantilever box module. temporary supports on the bottom plate, and local additional stiffening was required. The substructure of the removable bridge is under construction, and 2 m-diameter, 30 m-long piles have already been completed to the deep foundation. Within the next days, the watertight caissons needed to build the pier footings will be placed on site; the main body of the piers will be concreted using climbing formwork, and at the top, corbels will be made to house the structural bearings. At this time, the shipping and lifting project is under development. High capacity pontoon and strand-jack units (placed at the pier top) will be used.
4. Discussion, Acknowledgements
Conclusions
and
Many aspects of such a singular structure can arouse discussions and further analysis. The authors would like to highlight herein two particular aspects: Projecting large public infrastructures, there is always a controversy between those who believe that the construction tender must include the project, so the candidates (usually led by a construction and/or a contractor company) have the possibility to develop their own constructive alternative, which would be Fig. 10: Internal transportation of the fitted to their abilities and resources; this is the way to largest module (central box girder) obtain the optimum solution, economic, feasible and with a realistic construction period, they assert. On the other side, the defenders of the owner’s interests maintain that the project should be directed by the public administration, being completely independent from the contractor; this leads to the most economical, technical and functional option, they maintain. In the case here considered, both the fist general design and the further construction and detailed design was leaded by the public administration, but the participation of the contractor and the steel deck fabrication company, together with the communication exchange and technical queries presented to the project team, were a key item in the second mentioned phase. The result will be soon observable. The second item to debate is related with the construction methods to be applied in the erection of the bridge. As described before, the deck of the removable bridge is being designed to be transported by pontoon and lifted to its definitive position using strand-jack units. High-tech devices will be utilized in this procedure, and to an outsider it would seem to be an innovative scheme. But, as a matter of fact, this process is analogous to the one employed in the construction of the first plate box girder ever built: the Britannia Bridge in the Strait of Menai (United Kingdom, 1850), which with its 142 m-long main spans was a real outstanding solution, with no previous
references (see [9]). Good ideas last forever. The authors would like to sincerely thank to all people, companies and administrations involved in the project, especially to: The Spanish Ministry of Public Works (owner); Carlos Fernandez Casado S. L. (Javier Manterola, Antonio Martinez and all their team) and GINPROSA (led by Juan Tardon), companies in charge of the project; All in UTE Puentebahia (joint venture) and ACL (led by Alejandro Castillo) staff, owner’s consulting and technical advisor companies during construction; Dragados-FPS joint venture, the construction contractor; Dragados Offshore (led by Javier Fernandez and Saturio Solano), the steel deck manufacturer.
5.
References
[1]
LARSEN O., Ship Collision with Bridges. The Interaction between Vessel Traffic and Bridge Structures, IABSE Structural Engineering Documents, No. 4, Zürich, 1993, p. 10. UNE-EN ISO 12944-1: 1998, Paints and varnishes. Corrosion protection of steel structures by protective paint systems. Part 1: General introduction, AENOR, Madrid, 1999, p.10. UNE-EN ISO 12944-2: 1998, Paints and varnishes. Corrosion protection of steel structures by protective paint systems. Part 2: Classification of environments, AENOR, Madrid, 1999, p.11. NAVAJAS P. and LOPEZ A., Proteccion y Durabilidad de las Estructuras de Acero, Publicaciones APTA, Madrid, 2009, pp. 82-83. MANTEROLA J., PUENTES. Apuntes para su diseño, calculo y construccion, Tomo I, Colegio de Ingenieros de Caminos, Canales y Puertos, Madrid, 2006, pp.360-369. VIÑUELA L. and MARTINEZ J., Proyecto y Construccion de Puentes Metalicos y Mixtos, Publicaciones APTA, Madrid, 2009, pp. 425-453. EN 1993-2: 2006, Desing of steel structures. Part 2: Steel bridges, pp.70-90. FALBE-HANSEN K. and LARSSON Ö., “The Oresund Bridge: Project Development from Competition to Construction”, IABSE Conference Proceedings: Cable-Stayed Bridges, Malmö, Sweden 2-4 June, 1999. FERNANDEZ L., Tierra sobre agua. Vision Historica Universal de los Puentes, Tomo II, Colegio de Ingenieros de Caminos, Canales y Puertos, Madrid, 2004, pp. 43-44.
[2] [3]
[4] [5] [6] [7] [8]
[9]