Design of a test specimen to assess the effective bond length of carbon fiber-reinforced polymer strips bonded to fatigued steel bridge girders

Single-lap and double-lap specimens have been widely used to determine the shear strength of epoxy adhesives for many applications, including mechanical joints and retrofit of wing skins. Although it has been known that the stress state in the adhesive is not uniform in shear, but rather a combined stress state of peeling stress and shear stress, these specimens are useful to determine the bond strength if the application of the adhesive is similar in shape and material properties to the test setup. However, when the application does not have a shape similar to the single-lap or double-lap specimens, the test results of the lap specimens may not be applicable to the practical application. This is often the case for the use of epoxy adhesives in structural engineering applications. In this paper, one example of an application in which a single-lap or double-lap specimen is not appropriate for determination of the bond strength of the epoxy adhesive will be presented. The application of the epoxy adhesive discussed herein involves rehabilitation with bonded carbon fiber-reinforced polymer (CFRP) strips of the tension flanges of fatigued steel I-griders used in bridges. This application provides a cost-effective means of repairing these bridge girders, so long as the effective bond strength of the CFRP and adhesive are sufficient. In this work, the stress distribution in the adhesive layer is analyzed and compared between a prototype repaired bridge girder and various specimen models to determine an appropriate specimen and test setup for assessing the effective bond length of the adhesive. This study points out the strengths and weaknesses of standard single-lap and double-lap specimens and proposes that the new test setup is a suitable alternative for a wide range of applications in which the adherend is subjected to tension plus flexure.