Abstract:

There are many test methods in the international testing standards for the evaluation of FRP reinforcement-concrete bond. However, differences in test conditions of these methods affect FRP reinforcement-concrete bond. Therefore, even in the tests with identical test parameters, different bond strength values are attained due to experimental differences. In this study, the factors affecting FRP reinforcement-concrete bond were determined individually for four different test methods (hinged beam, beam-end, spliced beam and pullout tests) by using the gauss process regression (GPR) method. Accordingly, eight parameters affecting the bond strength, namely the diameter, fiber type, surface texture, embedment length, location in concrete and clear cover of reinforcement, the concrete compressive strength and the presence and degree of confining by means of transverse reinforcement were adopted as test parameters. In this way, the effects of each variable on bond strength were investigated for each test method and the methods were compared in terms of each parameter. The analyses on the four methods for identical test parameters depicted that there were differences in bond strength values as great as four times from one test parameter to another. The average bond strength values from the hinged beam tests were about 31, 3 and 146 % higher than the respective values from the pullout tests, beam-end tests and spliced beam tests, respectively. The degree of influence of each parameter on bond strength was established to depend on the test method and the same parameter was found to even have opposite effects on bond strength in different test methods.

Keywords: