Steel members strengthened with carbon fibre reinforced polymers - A study of local and global structural behaviour

Doctoral thesis, 2009

Investigations have been conducted by several research groups to study the effect of using carbon fibre reinforced polymers (CFRP) as reinforcement on steel beams and a brief summary of this work is given in this thesis. Laboratory studies of reinforced steel beams demonstrate that the bending stiffness and the ultimate load-carrying capacity can be significantly increased. Furthermore, field applications of this reinforcement method have been used on bridge elements made of steel or wrought iron, with the aim of either increasing the load-carrying capacity of the beams or restoring the origin capacity of beams that have deteriorated. In addition, field applications have also been made to increase the fatigue life of bridge members. The results of these field applications reveal that the main aims of the reinforcement application could be realised and it was also revealed that this reinforcement method could be cost efficient in comparison with traditional reinforcement techniques. Investigations have been conducted by the author to investigate the global behaviour of double-symmetrical steel beams that were reinforced with CFRP laminate on the tension flange. This took the form of a parametric study in which laminates and adhesives with different geometric and material properties were used. The study was carried out using laboratory tests and analytical and FEM analyses. There was good agreement between the results produced by the different methods in terms of global behaviour. It was revealed from this study that the maximum increase in the load-carrying capacity of the beam may be limited by the resistance of the steel sections to supporting compressive stresses. However, studies of the shear stress distribution in the adhesive layer at the end of the bond line revealed that it is difficult to define the interfacial stress distribution from laboratory tests using strain gauges applied to the laminates. The shear and peeling stress distribution in the adhesive layer was found to be similar in reinforced steel beams with high sections and steel plates reinforced with laminates on both sides and loaded in tension. A laboratory investigation was therefore conducted on double-sided reinforced steel plates where the shear and peeling strain distribution in the adhesive layer was defined with an optical measurement method. The laboratory investigation revealed that, in the case of non-linear adhesives with high ductility, a substantial redistribution of the stresses in the adhesive layer took place. The redistribution of the stresses resulted in a stress concentration close to the adhesive-steel interface, followed by the successive degradation of the adhesive joint until debonding failure occurred. Due to the redistribution of the stresses in the adhesive layer, predicting the stress distribution could be very complex. The laboratory tests also showed that using a linear-elastic model for predicting the joint capacity may result in a very conservative design, particularly for non-linear adhesives with high ductility. This thesis also presents a suggestion for a new design method, which considers the force in the laminate as the only governing parameter. This method inherently takes the redistribution of stresses in the adhesive layer into consideration.