Plate bonding : strengthening of existing concrete structures with epoxy bonded plates of steel or fibre reinforced plastics

This thesis deals with several aspects of the plate bonding problem, i.e. when steel or fibre reinforced plastics (FRP) plates are bonded to a concrete structure with the use of an epoxy adhesive. Both theoretical work and laboratory tests are presented as well as a full scale test on a strengthened concrete bridge. Earlier work in the area of plate bonding as well as the use of epoxy adhesives in civil engineering are discussed. This report presents several original and new features, both in theory and practice. A linear elastic model is introduced for calculation of shear and peeling stresses at the end of an epoxy bonded plate situated underneath a concrete beam subjected to bending. For the non linear behaviour, when the concrete starts to fracture, a non linear fracture mechanics (NLFM) model is introduced. The report is divided into two main parts, the main text and the appendices. The appendices are mainly used to give detailed derivations and test results. One exception is an illustrative example of how to strengthen a T-beam. is given in Appendix H. The main text begins with an introduction and is followed by a presentation of earlier work in the area of plate bonding. The theory for plate bonding is presented, beginning with the simple truss model which is useful for design of plated structures for both bending and shear. The theory of elasticity applied to bars loaded in pure shear is investigated. Moreover, linear elastic formulas are derived for calculation of the shear and peeling stresses in the adhesive layer. The original differential formulas are quite easy to put together but very complicated to solve; therefore, a simplification is needed It has thus been ass~ that the steel plate has negligible stiffness in comparison to the concrete beam being strengthened. This is a fully acceptable simplification and leads to solutions that give good results compared with FE-analysis. When the concrete starts to fracture, linear elastic theory can not be used. Here a new concept for NLFM is introduced. This concept is based on criteria from Hillerborg, 1977, and also ideas from Gustavsson, 1987. For stiff adhesives is it easy to calculate the failure load by means of the fracture energy, Gf. However for softer bond zones numerical calculations are needed. The laboratory tests performed show that the derived truss model is acceptable for design purposes and that good agreement with test results are attained for both bending and shear. Also the theory of Volkersen, 1938, corresponds very well with the test results. When using FRP plates, however, strain levels must be taken into considerations in the design stage. The tests shows that CFRP (Carbon Fibre Reinforced Plastics) material is preferable to (GFRP) Glass Fibre Reinforced Plastics) material. Finally a full scale test on a strengthened bridge is presented. The results from this test show that the strengthening method, being both useful and practical, is acceptable for strengthening of existing concrete structures.Godkänd; 1994; 20070426 (ysko