The Rion-Antirion Bridge
Located 250 km west of Athens (Greece), the Rion-Antirion Bridge will, as of 2004, provide a new crossing over the Gulf of Corinth. With a reference span of 560 meters, the Rion-Antirion Bridge ranks as one of the ten largest bridges: built on four piers with each foundation reaching a diameter of 90 m, this cable-stayed bridge features the longest suspended deck in the world (2,252 meters). The bridge has been designed to withstand the collision of a 180,000-ton oil tanker traveling at 16 knots, winds of up to 250 km/hour and earthquakes of magnitude 7 on the Richter scale.
Physical centrifuge modelling for large civil engineering structures
Each large civil engineering structure displays the particularity of constituting a prototype: engineers are not always given prior examples to serve as references in predicting the behaviour of their structures. Moreover, major structures being built throughout the world tend to be increasingly bold and often rely on concepts based on latest innovations. The tools available to the engineer for designing and sizing an exceptional structure lie in the domains of numerical and physical modelling. Numerical modelling provides the transcription, in mathematical terms, of an idealized approach to conceiving the structure, whereas physical modelling consists of conducting actual experiments on a reduced-scale model of all or part of the structure. A physical model therefore enables not only validating theoretical models, but visualizing and quantifying structural behaviour as well, especially in extreme situations of non-linear behaviour or failure loading (earthquakes, shocks, violent storms, etc.).
These physical models still require quite sizable installations (wind tunnels, vibrating tables, geotechnical centrifuges, etc.). In order to study foundations and soil-structure interactions, the rules of similitude (established for the most part by Edouard Phillips in 1869) lead to testing reduced-scale models under macrogravity conditions in a centrifuge. The centrifugal acceleration must be n times greater than gravity on earth if the model is built to a 1/n scale.
The challenge of this project is to build a structure that connects Peloponnesus to the continent under extremely adverse environmental conditions: a depth of 65 m midway across the strait, foundation soils of mediocre quality (Quaternary alluvial deposits exhibiting poor characteristics on beds with thicknesses exceeding 500 m), high seismic risk (an earthquake of magnitude 7 capable of having its epicenter on a fault located approximately 8 km away, thereby causing maximum accelerations of 0.5 G at ground level). The structure must also be able to adapt to significant tectonic movements (2 m offsets between two consecutive supports, both vertically and horizontally).
The foundation-laying principle represents a breakthrough innovation for construction in a seismic zone. Each tower is supported by a 90 m diameter caisson, which in turn lies on the seafloor whose ground has been previously reinforced by means of rigid vertical inclusions. Numbering between 150 and 200 underneath each support, these inclusions are tubular (2 m in diameter, 25 to 30 m in length) and driven in a grid approximately 7 m by 7m. A 3.60 m layer of ballast is inserted between the upper surface of the marine clay and the lower surface of the basement slab enclosing the tops of the inclusions, thus ensuring no contact between foundation and inclusions at all times.
This entire layout (gravel plus rigid inclusions) serves both to control the foundation failure mode in the case of an exceptional event and to limit the forces transmitted to the superstructure. The gravel acts as a fuse and the ground reinforcement enables avoiding the formation of deep rupture surfaces, which would induce sizable rotational movements on a 230 m high tower.
This innovative foundation-laying process was adopted by the GEFYRA consortium, which was commissioned to build the bridge and within which the VINCI DUMEZ-GTM Group was named to coordinate the project.
The tests were conducted on the centrifuge at the LCPC Nantes Laboratory upon the request of Mr. Alain Pecker, Director of the firm GEODYNAMIQUE & STRUCTURE (contracted to design the bridge foundation system). The clay deposits on site were used for these experiments. Mixed with water in proportions to obtain a more fluid material, the clay was first cleaned of all coarse elements (shells, rocks, seaweed, etc.) and consolidated into several layers in 89-cm diameter containers. The inclusions were then driven into the clay (some inclusions had been fitted with strain gauges to measure the force absorbed); next, the ballast and circular foundation were set into place.
Placement of the inclusions at 1 G in the (natural clay) soil mass
The preparation stage is concluded by installing the loading devices (servo-jack and electrical motor) and setting up the instrumentation (monitoring cameras and a series of sensors to measure: applied forces, vertical and horizontal displacements of the foundation, settlement, pore pressure, etc.).
Submerged foundation and experimental device
The container was then placed into the centrifuge and first submitted to a consolidation phase at 100 g for several hours until the requisite consolidation ratio, as determined according to the "Asaoka" method, had been reached. A series of static penetrometer tests were performed in order to verify the mechanical characteristics of the clay, without having to stop the centrifuge. The loading tests could then get underway and the foundation was submitted to various types of loads of increasing magnitude: horizontal loads associated with reverse moments and alternating static and cyclical loading sequences. The final loading was deliberately pushed up to the point of causing foundation system failure. Once the centrifuge had been stopped, the soil block was cut into vertical cross-sections to determine both the internal deformations and rupture kinematics..
Visualization of failures
after cyclic horizontal loading These experiments enabled verifying the effective resistance of the foundation system to loading conditions, in addition to deriving both the maximum allowable loading diagram and the rupture mechanisms. The experimental data necessary for validating the theoretical and numerical approaches were also obtained. The tests performed on reduced-scale models helped to optimize the number and position of reinforcement inclusions.
It may be added that the group of international experts assigned to verify the design (including Professor Peck) was on hand to witness a test. Moreover, the response of the bridge to wind effects was studied on reduced-scale models in the wind tunnel operated at the Nantes Testing Center of the CSTB (building industry research organization) (aerodynamique.cstb.fr) and the stay cables were tested on the fatigue testing benchat LCPC’s Nantes facility.
For further information:
GARNIER J., PECKER A. (1999). Use of centrifuge tests for the validation of innovative concepts in foundation engineering. 2nd Int. Conf. on Earthquake Geotechnical Engineering, Lisbon, June, 7 p.
PECKER A. (1998). Apport des essais en centrifugeuse au dimensionnement d’une fondation de pont. Colloquium entitled: "Mécanique et géotechnique. Jubilé scientifique de Pierre Habib". Luong (Ed.) ISBN2 7302 0521 3, pp. 143-153.
PECKER A. (2000). Pont de Rion-Antirion : fiabilité et conception parasismique des fondations. La sécurité des grands ouvrages, Presses ENPC, pp. 21-51.
RUSSEL H. (2000). Greek triumph - A bridge is born. Bridge design & engineering, Fourth quarter, pp. 29-36.
TEYSSANDIER J.-P., COMBAULT J., PECKER A. (2000). Rion-Antirion, le pont qui défie les séismes. La Recherche 334, September, pp. 42-46.
ARCHIMEDE-ARTE 2002 Conception du pont Rion-Antirion (5 march). Video.