| dc.description.abstract | The marine bridge differs from a regular bridge by using buoyancy units, often so-called pontoons, to obtain vertical equilibrium instead of rigid foundations. The term may include floating bridges, such as typical pontoon bridges, as well as submerged floating tunnels (SFTs). The marine bridges under development today are usually composed by subsystems proven in other fields, combined in a new way. These concepts use different mechanisms to resist vertical and horizontal offset, and the most common buoyancy elements were originally developed for offshore platforms. In Norway today, the driving force behind the development of marine bridges is the Coastal Highway Route E39 Project, where the deep and wide fjords demand new technological solutions.
There are several load components that must be accounted for when designing and analysing a marine bridge. According to the guidelines for bridge engineering in handbook N400 (2015), the characteristic loads are divided into permanent, variable, and accidental loads. Variable loads include loads induced by enviromental phenomena such as wind, wave, current, and tide. To ensure that equilibrium is obtained and that the static deformation is acceptable, static analyses are performed, while dynamic analyses are required when dynamic amplification is significant. This implies that the equation of motion is solved with damping and inertia terms included. If the eigenfrequencies and eigenmodes of a structure are of interest, modal analyses should be conducted. The results can subsequently be assessed in relation to the relevant excitation loads.
The Master Thesis investigates an innovative marine bridge concept for the Bjørnafjord, suggested by the Norwegian Public Roads Administration. The S-shape may enable the bridge girder to flex to thermal loads and to carry transverse loads without use of mooring lines.A number of marine bridge concepts for the Bjørnafjord have been discarded because cost estimations indicate that they are too expensive. The proposed concept is one of these concepts, which despite use of new technological solutions Bjørnafjorden is part of the Coastal Highway Route E39 Project, and is selected as a basis for the concept design because of the of its extreme width of more than 5 km. However, the relevancy of the Master Thesis is not necessarily confined to extreme crossings and can prove helpful in development of future bridge designs for other sections. The concept incorporates \SI{1000}{m} long submerged twin-tubes in combination with an S-shaped pontoon bridge which extends over the remaining 5125 m. A rectangular transitional structure is used to connect the two components.
The motivation behind using an S-shaped girder is that the curved design can enable the bridge girder to flex to thermal loads and to carry transversal loads through arch action. The latter consequence may make horizontal mooring to sea bed along the pontoon bridge unnecessary. The submerged tunnels are incorporated to provide the required navigation channel in the northern end of the bridge, while the rectangular transitional structure is included to accommodate the presumingly increased wave and tidal loads acting where the tunnels enter the water. 34 so-called semi-pontoons are employed, which due to their geometry give reduced wave loads and tidal loads compared to a regular single-column pontoon. To avoid buckling and to increase the effective stiffness of the bridge structure, a longitudinal pretension force is applied. Ideally, the pretension level should be tuned to avoid critical excitation frequencies, while ensuring that the stress levels in the bridge are in an acceptable range.
The Finite Element software ANSYS Mechanical APDL is used to model the bridge and to execute static and modal analyses. In the static analyses, effects of permanent loads, tidal variations and thermal loads are examined, and two different methods to obtain pretension are tested. The results indicate that a vertical displacement of the transitional region is inefficient, while an axial prescribed displacement works well. It is evident that the latter method can be employed to counteract the thermal expansion within the whole range that is tested for. Due to a segmentation fault, the modal analyses must be run on an alternative, simplified model. Although the results from these analyses have limited transferability to the original model, the results are included to enable comparison if modal analyses are performed for the other model at a later stage and to demonstrate how the eigenperiods of a similar structure develop with pretension. The results from the modal analyses with pretension agrees with the theory; the additional geometrical stiffness increases the eigenfrequencies, moving them further away from the critical region where resonance would be excited by the specified sea states. | en |
