3. Key Issues and Assumptions
3.1 Key Issues
The scheme options have been assessed with respect to the key issues and criteria which govern the overall design of the bridge. The major design objectives are to provide an elegant, unique and instantly recognisable structure which is durable and straightforward to maintain. With the procurement programme being a key concern for delivery of the project, constructability is also very important as this will lead to a reduced construction period as well as reduced costs.
Aerodynamic stability is an issue whose importance for any long span structure has been well established. Whilst a programme of wind tunnel testing is required to fully investigate these phenomena, the use of correlations established from previous tests as well as computational modelling has been used to provide preliminary guidance on performance.
An unusual structural feature of the Forth Replacement Crossing is the provision of double main spans which are inherently less stiff than a traditional system where the pair of towers which flank the single main span is anchored back to ground by the stay cables in each side span. In this case the central tower has no back stays and may deflect significantly under asymmetric loading of the bridge resulting in relatively large deck deflections as well. This bridge will include the world’s longest multiple cable stayed spans and developing an understanding of the flexibility associated with this structural system as well as appropriate mitigation measures has been influential in the structural design. This is described in more detail in Appendix D.
The bridge crosses a navigable waterway and maintaining safe navigation clearance at all times governs the vertical alignment of the bridge. Furthermore the potential consequences to the bridge due to errant ships impacting the towers and piers are critical to the design of the lower sections of the towers and the foundations. The possibility of subsequent explosions and pool fires if the ships contain hazardous flammable materials is also being studied.
3.2 Consideration of longer main spans
The general arrangement of two main spans each of 650 m, centred around Beamer rock, means that the location of the south tower is in relatively deep water. Bed level is approximately -22 mOD, and bedrock is tentatively assumed at approximately -55 mOD although the tower is beyond the extent of previous borehole investigations so this level is uncertain. More accurate levels will be determined as part of the ongoing marine ground investigation. The depth to rock is expected to lead to a substantial piled foundation at the south tower.
Studies have been carried out to investigate whether it is feasible to locate the south tower further south and increase the main span lengths. The potential benefits are a saving in the foundation costs which could arise from two improvements to the foundation conditions:
- Bedrock level being not as deep (to be assessed by marine ground investigation)
- Reduction in design ship impact force due to greater distance from the typical vessel transit paths (to be assessed by ship impact investigation)
There would also be further benefits arising from a reduction in the number of approach span piers in the water. However, the superstructure costs would be increased. An assessment of the overall merit is required once the investigations that determine the potential benefits are completed. It is not obvious whether longer spans will prove to be of benefit and therefore the assessment of scheme options at this stage has been carried out on the basis of 650 m spans. However the general arrangement of potential longer span bridge configurations is illustrated in Drawings FRC/C/076/S/002 and FRC/C/076/D/013 contained in Appendix B.
3.3 Stay Cables
Two different types of stay cable system are suitable for large cable-stayed bridges: parallel wire cables or multi-strand cables. Alternative cable types of locked coil strand or spiral strand are not appropriate due to their poor fatigue performance, low stiffness and lesser ultimate tensile strength (in typically manufactured cable sizes).
Parallel wire cables have a very compact cross section and are factory manufactured to the specific lengths required. Galvanised wires are arranged into the required pattern, and a polyethylene sheath is extruded onto the outer surface. The cables are wound onto reel and transported to the bridge site, where substantial lifting equipment is required to handle them. Very large jacks are needed to stress the cables. Cable length adjustment can be made with either shim plates, or a large nut on the threaded portion of the cable socket, depending on the system adopted.
Multi-strand cables are assembled on site. After the cable sheath is placed between the tower and deck anchorages, individual strands (each consisting of 7 galvanised wires within a polyethylene sheath) are fed through and secured using wedges at each end. The diameter of the cables is larger than for parallel wire cables of the same capacity, as each strand has its own corrosion protection sheath, and spare space is required within the outer sheath to allow strand installation. Stressing of individual strands can take place using small stressing equipment to adjust the lengths, and care must be exercised to ensure an even force distribution between all strands. Any de-stressing must be done using a large stressing jack to adjust a nut on the anchor so as to avoid disturbing the wedges holding the individual strands.
Cable replacement for parallel wire cables involves removing the entire cable, and replacing it with another one. Large lifting and stressing equipment is required. For multi-strand cables it is possible to withdraw, inspect and replace individual strands by reversing the assembly method. Although still a major operation, it can be performed using relatively small equipment and without major disruption to operation of the bridge. In practice it may be that once the cables have reached the end of their design life removal of the entire cables may be required which would involve similar procedures as for the parallel wire cables. Nevertheless, the ability to inspect individual "witness" strands at periodic intervals is a definite advantage.
The compact nature of the parallel wire cables enables equipment to be clamped onto the cables at any location along its length. If either cross–ties to link stay cables together, or external damping devices are required to limit unforeseen vibrations, or if street lighting equipment is to be suspended from the stay cables this can be an advantage. For multi-strand cables, provision for these types of equipment must be planned in advance of installing the stay cables and a special fixing point formed in the outer sheath.
At this stage, multi-strand cables appear favourable, due to long term inspection and replacement considerations. As the cable diameters, and therefore wind loading, are larger for this system, designing the structure accordingly does not preclude the use of parallel wire cables if they prove more advantageous. For example it could be considered to tender the project allowing either stay type in order to obtain the most competitive price.
A maximum cable size of 127 strands has been assumed, as although some cable manufacturers include larger cable sizes in their literature, experience and suitable equipment for fatigue testing and installation is extremely limited. There are a number of manufacturers that have a stay system with this size as their limit. If larger sizes are demanded there may be a restriction to competition.
3.4 Ship Impact
The bridge crosses a navigable waterway with approximately 5,500 significant vessel transits per year in the Forth Deep Water Navigation Channel travelling to and from Grangemouth and other upstream ports. Vessels up to 39,000 DWT pass under the bridge but the number of passes of such large vessels is very low. Over half of the vessel traffic is less than 6,000 DWT and only 1% of the traffic is larger than 20,000 DWT. The Rosyth Navigation Channel also passes below the northern main span of the bridge but the volume of shipping using this channel is an order of magnitude lower than the Forth Deep Water Navigation Channel and the subsequent risk from ship impacts is also very low.
The importance of ship impact loads for the design of the foundations was recognised during the Setting Forth studies which recommended a design ship impact load of 130 MN based on a 33,000 DWT ice strengthened tanker travelling at 12 knots. A force of this magnitude would govern the design of the foundations and would require significantly more piles than are needed to resist the ordinary in-service loads of self-weight, traffic and wind.
Considering the very low volumes of large ships it is possible that a statistical analysis could conclude that the probability of a large vessel striking one of the towers or piers at full speed is extremely low and therefore acceptable such that the design ship impact scenario could involve a smaller, more typical, vessel and/or travelling at a lower speed. The American design standard AASHTO provides a detailed and prescriptive methodology for carrying out such a statistical analysis which would result in a design impact load of approximately one third that recommended by the Setting Forth studies.
However, some of the target criteria, correlations and formulae used by the AASHTO method are superseded by guidance in the Eurocode and recent research. On the other hand, the Eurocode does not provide a prescriptive methodology for the statistical analysis. A project specific statistical methodology is currently being developed which would be compliant with Eurocode but may include some of the statistical components of the AASHTO methodology where they are believed to be relevant.
An important component of the statistical analysis is the probability that a ship will lose control in the vicinity of the bridge. Loss of control can be contributed to both human error and mechanical failure and the probability of these incidents occurring can be significantly reduced by piloted and tug-assisted vessels. Discussions have been held with Forth Ports which indicate that high rates of pilotage and tug-assistance are expected for the larger vessels which should be included in the statistical analysis.
From the preliminary results of the statistical study it is believed that the design ship impact force will be lower than that estimated by the Setting Forth studies and can be accommodated by moderate strengthening of the foundations compared to those required to resist in-service loads and the assessment of scheme options has been carried out on that basis. Furthermore, slender tower elements near the waterline have been considered unacceptable due to excessive vulnerability to large ships travelling in ballast at high states of tide.
3.5 Other Issues
3.5.1 Surfacing Thickness
The road surfacing system adopted will depend on the structural nature of the deck. Generally thinner surfacing is used for steel structures compared to concrete or composite structures because of the significant weight saving and hence reduction in structural quantities. In the past very thin surfacing systems have been adopted in the UK with a 38 mm mastic asphalt system being used on a number of steel bridges but this has in some cases resulted in poor ride quality and difficulties in maintaining the system. If an orthotropic steel box girder is adopted, a surfacing thickness of approximately 70 mm will be suitable on top of the stiffened steel deck plate which is consistent with European practice for steel bridges. This thickness will result in reasonable ride quality and allow the upper wearing course to be replaced without disturbance to the lower base course. 70 mm of surfacing also allows a 2 mm reduction in the deck plate thickness compared to thinner surfacing due to composite action in reducing fatigue stresses. An assessment will be made of the most suitable material to use considering either Gussasphalt, mastic asphalt or epoxy asphalt systems on top of the waterproofing layer.
For a concrete deck slab, as would be adopted for a composite deck solution or a truss solution, the weight penalty associated with thicker surfacing is proportionally less and a standard 125 mm surfacing layer has been assumed in this assessment (hot rolled asphalt or stone masic asphalt weairing course with appropriate base layer). This may result in a slightly better ride quality and more standard maintenance and replacement procedures.
3.5.2 Vehicle Restraint Systems and Parapets
Along the edges of each carriageway, vehicle restraint systems will be provided in accordance with the relevant standards. A zone immediately behind each barrier will be kept free of any structural components, so that in case of an accident which leads to deformation of the barrier, the risk of a vehicle striking the structure is extremely small. Nevertheless, vehicle impacts on the structure will be considered in the design. The barriers systems adopted will have been proven to comply with the relevant standards and appropriate limits of deformation.
The barriers will provide segregation between the traffic and the pedestrians / cyclists. At the outer edges of the walkways, pedestrian parapets will provided. These could be combined with the wind screens to make use of the same supporting structure.
Due to the critical function of the bridge as a key link in Scotland’s transport network, it is important that it remains operational at all times for traffic use. The exact criteria for the maximum wind speeds across the carriageway will be defined as part of a study of other major bridge crossings, and research into the effects of gust wind on road and light rail vehicles.. The criteria will need to be met under all wind conditions when traffic can still use other parts of the network such as the approach roads leading to the bridge. Wind screens will be provided on the bridge to achieve this.
The three corridor cross section will require windshields at each edge of the deck, but as the section is very wide, it may be necessary either for these to be very high, or to have additional wind screens surrounding the multi-modal corridor in order to provide suitable wind shielding across the full width of the deck. For the double level cross section, windshields will be required on the edges of both the top deck and probably also the bottom deck.
Additional windshields may be required along short lengths close to the towers, where sudden changes in cross wind can occur due to the shielding nature of the tower structure.
The windshields along the edges of the deck will be designed to be difficult to climb over.
Possible layout of anti-climb windshields