Experimental Study On Durability Of Bond Behavior Between FRP And Concrete

Along with the widespread application of FRP reinforcement technology, durablity of FRP reinforced concrete structure has aroused many researchers' interest. In many actual projects, FRP is often used in outside directly or under the adverse circumstance. Durability of FRP reinforced concrete structure under these environments is decided by the durability of bond between FRP and concrete. This question must be payed more attention and solved reasonably before FRP is used in such environments.The durability of bond between FRP and concrete is studied by shear experiment. The samples are manufactured according to research contents. The experimental environments contain thaw-freeze cycles, wet-dry cycles, the acid environment, the alkali environment, the salt water environment and the water environment. Study mainly includes the following several works: destruction process and destruction characteristic of samples; the change tendency of interface bearing capacity and the reasons causing the change; the bond-slip performance of surface between FRP and concrete; the distribution of stress and strain in FRP along bonding length. Following conclusion are draw from this study: Through experiments of 6 different environments, the destruction characteristic of samples have obvious change; Interface bearing capacity assume the drop tendency basically and the drop reasons are various;The bond-slip performance of surface has no change essentially. Through experiments of thaw-freeze cycles and wet-dry cycles, the distribution of stress and strain in FRP has no change essentially.Through the analysis of samples' performance, influence rules are studied, providing data and reference for the bond durability research of FRP reinforced concrete structure and instructing the application in the actual projects.Using the elementary theory of BP neural network, the neural network parameters are designed in view of the questions of experiments, appraising the test data and forecasting the interface bearing capacity according to the actual situation. The results confirm this method is feasible and effective.