ICE - Institution of Civil Engineers - Virtual Library

This paper describes the first known application of fibre-reinforced polymer (FRP) composites for structural re-decking of a railway bridge for full railway and derailment load. Calder Viaduct crosses the River Calder near Sellafield in Cumbria. The original bridge comprised three spans of half-through-type steel girders with timber decking and ballasted rail, constructed in 1922. Due to the coastal location of the viaduct, significant corrosion of the steelwork and deterioration of the timber decking had occurred. This type of timber railway decking on a metallic supporting superstructure is common throughout the UK. The structure was chosen by Network Rail for a pilot project for replacement of the timber decking with FRP composite materials. The chosen FRP decking solution comprised a standard pultruded glass-fibre-reinforced polymer (GFRP) deck with bonded and bolted GFRP top plate for ballast retention and wear resistance. Other types of FRP deck, such as bonded FRP pultruded sections and bespoke vacuum-cured FRP decks, were also considered. The choice of an FRP decking system was driven by a number of factors; ease of installation, adaptability, minimum depth/weight, compliance testing, durability/design life and cost. The installation method, design challenges such as achieving compliance for derailment loading, and procurement issues such as developing a generic specification to enable competition, are discussed. Conclusions are given on the factors that require consideration by the designer to deliver a competitive FRP solution in accordance with best practice, and a brief comparison with equivalent performance steel decking is included.

This paper discusses the development of advanced polymer composite material applications in civil and structural engineering over the past three decades. It illustrates the rapid growth in its utilisation in comparison with other construction materials, and the reasons for this. The material has extraordinary mechanical and important in-service properties, which when combined with other materials are utilised to improve stiffness, strength and durability and lower the whole-life cost and environmental impact. The paper demonstrates the types of structure that have been developed from the material, and the most advantageous way to employ composites in civil engineering. There is, however, generally a 20-year period from the fundamental innovation stage to the final use of the product by industry: this seems a very slow process. The paper concludes by summarising key successes of advanced polymer composites in construction.

Network Rail owns and operates Britain's railway infrastructure, and has been a leading client in the adoption of fibre-reinforced polymers (FRP), initially for bridge strengthening and then for other civil engineering related applications. It now owns over 20 FRP-strengthened bridges, three fully FRP bridge, and several fibre-reinforced polymer platform decks, and has recently installed the first known application of an FRP secondary decking system for carrying railway live loading. Other applications of FRP are under development. This paper describes the background to the adoption of FRP for railway civil engineering, and the safeguards put in place to ensure satisfactory performance. A number of completed schemes are used to illustrate typical applications, and those planned in the immediate future are described. The paper also looks briefly at FRP related research supported by Network Rail.

This paper reports on the repair and reinstatement of fractured historic cast iron elements of the High Level Bridge in Newcastle upon Tyne, UK. As a part of a full structural refurbishment of this grade 1 listed structure, three cast-iron girders were extracted from the historic bridge and subjected to fatigue testing. This destructive testing was carried out under the condition imposed by English Heritage that the cross-girders could be reinstated with the least visual impact. To satisfy this requirement, the damaged elements were repaired using a refurbishment procedure employing strengthening with externally bonded fibre-reinforced polymer (FRP) plates and ‘cold' metal stitching with alloy inserts. The paper discusses the numerical analysis, structural modelling and the quality control measures adopted during construction for the repair of the girders. The fatigue testing will be presented in a separate paper.

The development of a lightweight bridge deck for use in railway applications provided an ideal opportunity to combine outputs from three-dimensional analysis techniques, empirical data and simple numerical methods. Add in a modern anisotropic material and the scene is set for an exciting project.

Two finite-element (FE) models for evaluating the lateral buckling loads of simply supported single- and two-span continuous unstiffened and carbon-fibre-reinforced plastic (CFRP) strip stiffened pultruded glass-reinforced plastic (GRP) wide-flange (WF) beams are presented. Eigenvalue analyses of the FE models are used to calculate the lateral buckling loads of unstiffened and CFRP stiffened beams. The effects of varying the stiffener geometry (thickness, width, etc.) and stiffener arrangement (CFRP strips bonded to bottom flange, top flange, etc.) on the critical buckling loads are quantified. It is shown that stiffeners increase the buckling loads of single-span beams more than two-span continuous beams and that only bonding CFRP strips to the top flange is a near-optimal stiffener arrangement.

Failure tests on pultruded glass-reinforced-plastic single-bolt tension joints are described. Four joint layouts, defined in terms of ratios of end distance (E) and plate width (W) to bolt diameter (D), were used. The objective was to rank the deleterious effects of four factors on joint failure loads: (a) bolt/hole clearance; (b) angle between tension and pultrusion directions; (c) elevated temperature; and (d) water immersion period. Three levels were selected for each factor: (a) 0·2, 1·2 and 2·2 mm; (b) 0°, 45° and 90°; (c) ambient temperature (∼20°C), 60°C and 80°C; (d) 0, 6·5 and 13 weeks. The joint test matrix was defined by a Taguchi three-level balanced array for four factors. Altogether 108 joints (comprising three replicates of nine joint groups, four factors and three factor levels) were fabricated and tested. Four methods were used to rank the effects of the factors. Their rank order (highest to lowest) was identified as: elevated temperature, water immersion period, bolt/hole clearance, and angle between tension and pultrusion directions. Environmental factors were shown to reduce failure loads by up to 93%.

This paper presents the effect of unevenly prestressed carbon-fibre-reinforced polymer (CFRP) sheets used for strengthening a prestressed concrete beam (L = 3·6 m). An integrated anchor system, consisting of steel brackets and anchor plates, is used to prestress the sheets. To predict the behaviour of the CFRP sheet during a prestressing operation, a three-dimensional finite-element analysis model is built. The model includes the reinforced concrete element, structural solid element and elastic shell element to represent the concrete, anchorage and CFRP sheet respectively. The model shows good agreement with test data, including a maximum error of less than 5%. Stress concentrations in the CFRP sheet are predicted near the tip of the anchor plates. Uneven prestressing effects are observed across the CFRP sheet because of the anchor geometry and uneven bonding of the sheet in the laboratory. These effects cause uneven rupture of the sheet when the strengthened beam is flexurally loaded to failure. To avoid such a failure mode, care should be exercised when the strengthening system is prepared with adequate anchorage.

The increased use of fibre-reinforced polymers (FRPs) in reinforced concrete construction is largely driven by the requirement for improved durability, especially for those applications subjected to the most severe environmental conditions. However, the adoption of FRP as embedded reinforcement in new structures has been much slower than in repair and strengthening applications. This is owing to the limited availability of curved FRP reinforcing elements. In addition, the mechanical performance of the bent portions of FRP bars is significantly reduced under a multiaxial combination of stresses, and the tensile strength can be as low as 40% of the uniaxial tensile strength. This paper presents and discusses potential issues relating to the use of curved FRP bars as embedded reinforcement in concrete and uses as an example one of the case studies that was examined during the European funded project, CurvedNFR. A 6 m long concrete plank reinforced with thermosetting FRP bars as longitudinal reinforcement and thermoplastic FRP strips as shear reinforcement was manufactured and tested. The use of FRPs allowed the reduction in the required concrete cover without compromising durability. This study shows that current design recommendations for FRP concrete structures are effective in predicting deflections and crack widths at service load. It is also shown that in FRP concrete, serviceability limit state can control the design.