[EN] At present, hybrid structures in flexural elements are being gradually incorporated in the field of civil and building structures. The low weight and high durability of these structures make use viable. Hybrid structures are usually composed of polymeric materials reinforced with glass or carbon fibers (GFRP or CFRP) in structural shapes of any type of section, in I, in box, trough-shaped, as elements working in tension, and the concrete in the compressed head of the resistant section. Due to this, the union of these materials and their excellent qualities allow for optimizing their mechanical performance in these positions. The failure modes of these structures have different opinions (points of view) among researchers. Therefore, this thesis focuses first of all on the possibility of designing a beam with ductile failure mode that moves the break to the compressed head by fiber reinforced concrete(Appendix I). It is found in this study that it is not possible to analyze this option because after "shielding" all possible failure modes of the structure the break occurs by the grazing effort between GFRP and concrete , without checking the influence of HRF. For this reason, the study is redirected and focused on the bond between GFRP and concrete by analyzing the influence of a series of surface treatments applied to the GFRP profile and the use of mechanical fasteners. To do this without losing sight of applications in civil engineering and construction, and after analyzing the scientific information available, given that there are no standard tests of this type of hybrid beams trials, the need to provide a simple testing methodology arises to enable the study of this phenomenon. That is why the adaptation of the pull-out test, usually for bars, is presented for rectandular section profiles. The complexity of this adaptation is solved by providing a number of modalities to the approach depending on the degree of confinement of the concrete over GFRP profile and the possibility of applying torque to the prestressed bolts, finally generating patterns 2C, 1CA, 1CE and 1CL. These patterns in turn are represented in the forms of connecting profiles with "in situ" concrete: flange profile embedded in the compression layer, or the top of the flange profile in contact with the compression layer. In the modality 2C, the concrete is passively confining all faces of the GFRP profile; in the form 1CA, the concrete confines the two main faces of elements but one of them does not have surface treatment; in the form 1CE, one side of the profile is confined with surface treatment and the other one is not in contact with the concrete; and finally, in 1CL one side is in contact with the concrete and the other one is liberated from this to access profile and give torque to conduct a post-tensioning of the mechanical fastener, thus achieving an active confinement of concrete over GFRP profile. Once clarified trial methodology, it is proceed to the characterization of the different proposed bond variables like a series of surface treatments such as sandblasting, and textured resins and a combination of both. Also fixed and moveable mechanical fastening elements are used providing them with a torque, sometimes only for fixing the profile GFRP and in other cases to achieve an active confinement of the concrete on the profile. After the results, a proposal for characterizing the bond parameters is made and a variety of behavioral models based on bond stress-slip curves are analyzed, bringing innovations to the scientific community. TESIS