The activities started with a complete review of state of the art on bridge innovations (phase 1 of the NR2C project), which included:
- main needs and problems on civil engineering structures;
- general view of new materials applications (high performance steel, ultra high performance fibre concrete, fibre reinforced polymers, other materials);
- owners and designers' expectations (this section was mainly based on interviews of specialists, designers, architects, owners, etc);
- evolution of civil engineering structures, with a particularly interesting analysis of the evolutions made in the past which highlight that rare disruptions are linked to new materials;
- vision on new bridges for the future.
Then following activities focused on defining new material properties and modelling rules, providing to engineers and design consultants the necessary information for the design and calculation of bridges using ultra-high performance fibre-reinforced concrete or fibre-reinforced polymers.
NR2C has focused on short span bridges and in particular small industrial light structures easy to assemble on site, and likely to be used alone for smallest spans or themselves supported by structural elements, when used transversally, for most important spans.
The main idea of the design was to combine new materials, using the best performances of each material, including possible use of wood as a renewable building material.
Consequently, three main solutions families have been investigated (phase 2):
- Element of sandwich slab, combining UHPFC and FRP, filled with lightweight concrete. The sandwich construction consists of three layers: a glass fiber-reinforced polymer composite (GFRP) sheet with Tupstands for the tensile skin, lightweight concrete (LC) for the core and a thin layer of ultra-high performance fiber-reinforced concrete (UHPFRC) as a compression skin. Mechanical tests on long and short-span hybrid beams were performed using two types of LC with different fracture mechanics properties and two types of FRP-LC interface: unbonded (only mechanical interlocking of LC between T-upstands) and bonded with an epoxy adhesive. The experiments showed that ultimate load is determined by the shear strength of the LC core. A fracture mechanics-based shear strength prediction method was therefore developed. The experimental results and modeling highlighted the importance of considering not only static strength, but also fracture mechanics properties such as the LC characteristic length. Moreover, a design concept was developed demonstrating the feasibility of the suggested hybrid bridge deck and allowing the definition of the required LC material properties according to the transverse bridge.
- Solutions based on UHPFRC. The objective of the study was to develop future-oriented solutions for the construction of bridges using new materials to their best advantage in terms of durability and sustainable development. This study has explored solutions for 10 m and 25 m span bridge decks consisting of prestressed UHPFRC girders or mixed UHPFRC/composite and UHPFRC/steel girders. It showed that the most advantageous solutions in terms of materials savings are fully prestressed UHPFRC decks. For 10-metre span structures, these solutions lead to financially viable structures under current economic conditions. Mixed UHPFRC/glassfibre solutions are penalised by the poor ratio of the Young's modulus to the connection strength of the two materials which means that they have to be overdimensioned in order to respect the deflection requirements of the decks. For 10-metre span bridges, notched web girders and steel footings are interesting alternatives to fully prestressed solutions. Similarly, for 25-metre span bridges, the combination of UHPFRC and a composite can become advantageous in the case of partially prestressed UHPFRC prefabricated lattice beams. As a follow-up to this study, reflection can be pursued on the design of assemblies that will optimise the general design and building of the bridges. This reflection could include: the design of UHPFRC slab component assemblies in order to guarantee local stress transfer and water tightness of the slab; the design of UHPFRC and composite assemblies which could develop the adhesive bonding of systems that ensure the dispersion of concentrated stresses.
- Element of structure combining wood beams, slab UHPFRC, and FRP at the bottom of wood. The detailed design and large-scale validation of a prototype 10 m-span composite UHPFRC - carbon fibres - timber bridge helped drawing conclusions concerning the feasibility of the sandwich concept developed, the critical issues to be considered in practical realisation, and also the validity of the preliminary design hypotheses and detailing assumptions. These conclusions are:
- Material properties and design rules of UHPFRC and composites tend to get satisfactory mature understanding for design. On the contrary, the behaviour of the lightweight material (wood in the present sandwich concept) used for shear transmission, even though critical in the design, seems to be less controlled. Possible defects may significantly change the safety margin. Timber properties should thus deserve increased research and control efforts.
- Connection of UHPFRC pre-cast elements one against the other should deserve special attention, especially for thin elements. Stress concentrations and premature cracking may be caused by geometrical uncertainties and imperfect joint (or gluing) execution.
- Adhesive connection seems promising and ensures composite behaviour in the serviceability domain, which provides easily understandable linear behaviour and helps sound calculations. However, failure of this connection, which may be controlled by hardly detectable execution flaws, may result in very brittle structural failure. This implies an additional connecting system only for safety purpose, which may result in significant additional expense.
- Due to the orthotropic behaviour of wood, parallel reinforcement using (also orthotropic) composites does not prevent brittleness versus cracks propagating in-between the fibres. A kind of shear reinforcement (transversely to wood fibres) might be of interest for ensuring the desired non-brittleness.
- Simplified execution and material savings were searched in the tested multi-beam solution. However the beams redundancy did not result in high enough ductility, in the sense that the bearing capacity at first failure was hardly recovered when further loading the remaining beams. Possible advantage of cross-beams towards this end might be questioned.