BOND BEHAVIOUR OF NSM CFRP-CONCRETE SYSTEMS: ADHESIVE AND CFRP CROSS-SECTION INFLUENCES J. Cruz 1 , A. Borojevic 1 , J. Sena-Cruz 1 , E. Pereira 1 , P. Fernandes 1 , P. Silva 1 and A. Kwiecien 2 1 ISISE, School of Engineering, University of Minho, Portugal. Email: jsena@civil.uminho.pt 2 Department of Civil Engineering, The Cracow University of Technology, Poland. ABSTRACT Near Surface Mounted (NSM) strengthening technique has been used in a sustainable way for retrofitting existing structures. This technique, which utilizes CFRP laminates inserted in the concrete cover, has been used due to the several advantages when compared with the technique based on the application of these reinforcing materials on the concrete surface (EBR technique). Although several studies have been developed on this topic in the recent past, open issues still deserve research, such as the influence of the adhesive type on the performance of the NSM-CFRP system. The present work details an experimental program carried out in order to assess the effect of using three adhesives with distinct mechanical properties on the bond behavior of the NSM-CFRP system, through direct pullout tests (DPT). Thus, the following variables were considered in the present study: (i) the type of adhesive; (ii) the cross-section of the laminate; and, (iii) the bond length. The experimental pullout force-slip responses were obtained and digital image correlation (DIC) was used for obtaining additional information about the bond mechanisms developed. In general, two of the three adhesives, with similar mechanical characteristics, provided essentially similar bond behavior, with high level of effectiveness, whereas the third adhesive, which had a much lower elastic modulus than the other two, provided the lowest effectiveness in terms of the investigated parameters. KEYWORDS NSM, CFRP, Direct Pullout Test, bond, adhesive. INTRODUCTION Fibre reinforced polymers (FRP) have been extensively investigated for repairing and/or strengthening existing structures. These materials can be introduced in the concrete cover of the element to be strengthened through the near surface mounted (NSM) technique. An epoxy adhesive is commonly used to fix the CFRP laminate to concrete. This bonding agent plays a critical role on the composite performance of the system. Extensive research has been developed to assess the bond behaviour of this strengthening system using carbon fibre reinforced polymers (NSM CFRP system). According to Coelho et al. (2015) the performance of the NSM CFRP system depends mainly on the: (i) geometry of the groove and of the FRP; (ii) mechanical properties of the concrete; (iii) mechanical properties of the adhesive; (iv) FRP cross-section and its external surface; and, (v) surface roughness in the groove. Digital Image Correlation (DIC) is a method that allows to evaluate the displacement fields at the surface of a structural element, as well as to compute the deformation fields during the test. Essentially the method is based on comparing two consecutive images of the element surface, before and after its deformation, through the application of an appropriate correlation technique (Chu et al. 1985). More information about this technique can be found elsewhere (e.g., Pereira et al. 2012; Carloni and Subramaniam 2013). In the literature few investigations can be found dedicated to analyzing the influence of adhesive and cross-section geometry of the laminate on the bond behavior of NSM CFRP system, e.g. Macedo et al. (2008). In this research the effect of three adhesives for fixing CFRP laminates to concrete substrate according to the NSM technique is analyzed, by means of direct pullout tests. The main motivation for testing two stiff adhesives and one of much lower stiffness lies on the reported advantage of using highly deformable (flexible) adhesives in external bonding (EB) of CFRP laminates to RC beams as strengthening (Derkowski et al. 2013). In this research the type geometry of the CFRP cross-section as well as the bond length were also analyzed and as study variables. The experimental program is detailed and the main results are presented and analyzed in the subsequent sections.