Carbon fiber reinforced polymer (CFRP) inserted in different configurations of the tensile zone retrofitting with microconcrete containing steel fibers to the strengthening of beams

It is researched, in this study, the strengthening technique known as Near Surface Mounted (NSM), which consists of the insertion of laminates of Carbon Fiber Reinforced Polymer (CFRP) into notches in the covering concrete structures. In the strengthening in beams, the tensile zone is found damaged for several reasons (cracking and corrosion, for instance), which demands, in the practice of engineering, its preliminary retrofitting. It should be considered that the good performance of the material used in this retrofitting is fundamental for a higher efficiency of the strengthening. Therefore, it is proposed a methodology that consists of the reconstitution of the tensile zone of the beams with a cementbased composite of high performance (CCAD), which acts as a substrate for the application of CFRP and as an element for the transfer of efforts to the part strengthened. The retrofitting of this tensile zone was performed only in the shear span, as well as throughout of the zone with a view to evaluating the influence of this aspect on the performance of the beams. The CCAD, produced from Portland cement, steel fibers and microfibers, was evaluated using the Rilem (2002), showed to be able to delay the cracking. Tests performed in the beams with the tensile zone retrofitting by CCAD and strengthening using the technique NSM showed the efficiency of the proposed methodology.


Introduction
The strengthening of structural elements of reinforced concrete with carbon fiber reinforced polymer (CFRP) has become ever more widely known due to characteristics such as the high tensile strength and corrosion resistance, low weight, and ease and speed in application.However, in many countries such as Brazil, there is still no specific standardization on the issue.Professionals use the foreign standards, recommendations of catalogs from the manufacturers of products of CFRP and the results of existing researches.
The issue becomes even more delicate, because in most case studies of engineering, the parts of reinforced concrete need to be restored even before their strengthenings.
Retrofitting the reinforced concrete beams has been considered by many researchers.In recent years much research has been performed about bending and shear retrofitting of reinforced concrete beams with different materials such as steel plates, FRP and high performance fibre reinforced cement-based composite (Ferrari, Hanai, & Souza, 2013;Moatasem, Fayyadh, & Abdul, 2014;Hamdy, Afefy, & Hussein, 2015).
In this regard, according to Ferrari, Hanai, and Souza (2013), the preliminary retrofitting of the tensile zone of reinforced concrete beams (as indicated in Figure 1) with a cement-based composite of high performance (CCAD), based on steel macro and microfibers, prevents the quick spread of critical crack at the edge of the strengthening and delays the early detachment of the blanket of CFRP.Such procedure is extremely interesting for the increase in stiffness of the part, in its load capacity and for a greater use of the resistant properties of the strengthenings.
Also according to the author, the presence of a material of higher resistance to cracking in the tensile zone of the beams, promotes a better distribution of the cracks with smaller openings along the length of the strengthening.
Based on such already existing considerations, the present study complements the study of the proposal initially made by Ferrari et al. (2013) and applied here to the strengthening through the Near Surface Mounted (NSM) technique, performed at the tensile zone and retrofitting the beams by the CCAD.
Thus, in this study the tensile zone of the beams is retrofitting with a CCAD that functions as a transition layer between the CFRP and the beam in a way to allow a better performance to the strengthened beam.
In this sense, it is proposed in this study a methodology that consists of the retrofitting of the tensile zone of beams with a cement-based composite of high performance (CCAD) that serves as layer for the application of the CFRP and also as an element for the transfer efforts to the part that will be strengthened.The retrofitting of the tensile zone was performed to the part that will be strengthened.The retrofitting of this tensile zone was performed only in the shear span of the beam, as well as throughout the entire extension of the tensile zone with a view to evaluating the influence of this aspect on the performance of the beams.To evaluate the effect of the total or partial retrofitting of the tensile zone, a beam was retrofitting in the entire tensile zone and other beam had only the shear span.

Composites analyzed
Six different cement-based composites were analyzed.They are formed from the variation in the volume of steel microfibers and fibers as indicated in the Table 1.The steel fiber specified by the letter 'A' has 25 mm in length and terminal hooks.On the other hand, the steel microfiber specified by the letter 'C' is not yet produced for commercial purposes and has 13 mm in length only, terminal hooks and 0.75 mm in diameter, as are the fiber type A. For each group, five composites with the same characteristics were molded.The composites were stored at + 20°C and humidity of 95% until preparation for testing.
In order to produce the cementitious matrix of microconcrete the following materials have been used: Portland cement of high early strength and density of 3.15 g cm -3 , medium sand with a finemess modulus of 2.60 and an apparent specific gravity of 2.70 g cm -3 , and a maximum aggregate size of 10 mm and a particle density of 2.78 g cm -3 .A superplasticizer that acts as a dispersant of the cementitious material has also been used and all composites had 0.5% of superplasticizer by weight of cement.The cement consumption for the microconcrete was 443 kg m -3 .In order to evaluate the tensile strength of the CCAD, the three point bending test of prismatic specimens with central notch, as recommended by the RILEM TC 162-TDF (Rilem, 2002), has been applied.In the Figure 2, it is observed the general configuration of the test that was conducted under control of the crack mouth opening displacements (CMOD -crack mouth opening displacement) using, for this purpose, an electrical extensometer of the clip gauge type.

Results of the three-point bending tests: loads, strengths and P-CMOD curves
In order to determine the flexural toughness of the composites, the recommendations prescribed by the work group of the RILEM TC 162-TDF (Rilem, 2002) have been followed.The Table 2 shows the average values of loads and strengths calculated based on these recommendations.These values were extracted directly from the mean curve that is representative of the group.(2002).
In the Figure 3, the behavior of each composite is compared based on the P-CMOD.The curve used to each composite is the curve of intermediate behavior, which can be representative of the other curves of the group.
All composites presented an initial behavior characterized by a linear stretch, where the cement matrix is presented in full.In this behavior phase it is highlighted the performance for the composites CPM1A2C, CPM1.5A1C and CPM1.5A1.5C,characterized by higher stiffness, higher values of loads applied until the matrix cracking (F M ) and higher strengths (f fct.L ).As special focus, it is emphasized the values obtained for the composite CPM1.5A1.5C.The behavior of this composite is also emphasized in relation to other composites, after the matrix cracking, when it is noticed the existence of a stretch with horizontal plateau (Figure 2) due to the contribution given by the fibers and microfibers.

Characteristics
Eight beams were molded with cross section of 170 x 350 mm, length of 3600 mm and free span of 3200 mm.The lower longitudinal reinforcement was composed of two CA-50 steel bars with 12.5 mm in diameter and the upper reinforcement of two CA-50 bars with 6.3 mm in diameter.The stirrups were formed by steel bars with 6.3 mm in diameter with a uniform spacing of 120 mm.The beams were divided into four groups and their characteristics are described in the Table 3.The beams of the group A, without strengthening, are references for other beams that were strengthened.The beams of this group were sized with low rate of longitudinal reinforcement so that the ultimate limit state could be characterized by the excessive strain of the reinforcement without breaking the compressive zone.
The beams of the group B were strengthened through the insertion of two sheets of CFRP into notches performed in the covering concrete of the reinforcement.On the other hand, beams of the group C and D were subjected to a removal process of concrete from the tensile zone and the retrofitting with the composite CPM1.5A1C.In the Figure 4, it is showed the region of the tensile zoned that was retrofitting and strengthened on the beams.
The only difference between the group C and D is related to the extension of the region of concrete removal of the tensile zone.In the group C, this removal was performed in the entire extension of the span of the beams, while in the group D, it was restricted only to the region of the shear span of the beams.
Another aspect to be mentioned is in relation to the depth of 80 mm used for the concrete removal of the tensile zone.This measure was used based on the recommendations of Ferrari et al. (2013), i.e., for allowing partial exposure of the reinforcement of the stirrups and, therefore, a greater connection between the concrete of the beam and the cement-based composite.
Reconstitution of the tensile zone of the beams and application of the strengthening Fourteen days after concreting, markings for the regions to be removed from the tensile zone of each beams were performed.This removal was done mechanically with hammer drill and concluded with club hammer, flat chisel and pointed chisel.In order to remold the tensile zone from the beams of the groups C and D, it was applied the composite CPM1.5A1C.
It is important to mention that despite the best responses obtained with the composite CPM1.5A1.5C, the CPM1.5A1C was the composite applied in the reconstitution of the tensile zone of the beams, solely because of the low supply of steel microfibers, which is not commercially available and was produced on request for the present research.
The exposed surface of the concrete of the tensile zone received a cleaning with air jet, followed by water to remove all dust accumulated.It was sought to keep the surface of the concrete dry, however, saturated.
The composite was applied using two wooden formworks in plastic-coated plywood positioned on the two sides of the beams.The composite was manually inserted inside these formworks (Figure 5).In order to characterize the cement-based composite, six cylindrical specimens (100 x 200 mm) were molded.
The strengthening was performed on the beams eight days after the remolding of the tensile zone.The notches were made with the help of a cutting disc, seeking to keep them with an opening of 5 width and 18 mm depth.It was used a twocomponent epoxy resin, of brand name Sikadur 30, for fixing the sheets in the notches.

Groups
Be    The highlighted behavior of the beams of the group C in relation to the other beams, both in terms of stiffness and load capacity, is the result of the retrofitting of the tensile zone by the CCAD.Such behavior revealed by the Figure 9, indicates that the material CCAD contributed to delay the beginning of the cracking of the beam and to reduce the opening and the spread of the cracking.
In the Table 5, there are the comparisons of the values of loads supported by each beams for the maximum vertical displacement allowed (span/250 = 12.8 mm), considering the limit state of service.Highlighted is the load capacity of the beams of the group C, achieving an increase of up to 30% in relation to the reference beam and of up to 16% in relation to the strengthened beam.Thus, it is verified that the beams of the group C, both for the ultimate limit state and for the limit state of service present the best load capacity.

Strain in the laminate
In Figure 10, it is showed the strain progress in the laminate registered in the middle section of the span of each beam.The maximum strains registered in the strengthening, ranged between 8.0 (beam VD2) and 9.6‰ (beam VC2), reaching, therefore, 88% of the maximum capacity of strain of the CFRP sheets, which is of 10.9‰.

Conclusion
In the present study, it was employed the technique of flexural strengthening NSM, by which strips of sheets of CFRP were inserted into notches performed on the tensile zone of reinforced concrete beams.Some of these beams had their tensile zone previously removed and reconstituted through the application of a cement-based composite of high performance.The following conclusions were, thus obtained: It was possible to develop a cement-based composite with the addition of conventional steel microfibers and fibers in order to potentiate a greater contribution of the cement matrix for the composite strength and for the improvement of the transfer mechanism of stresses from the matrix to the fibers; The developed composite presented a behavior of pseudo strain-hardening, because, with the cracking of the matrix, the transfer of stresses was facilitated, firstly, by the steel microfibers that, in great amount in the matrix, conditioned the progress of the cracks with the increase of the level of loading; Comparatively to the beam that was only strengthened, it was verified that the complete reconstitution of the entire tensile zone of the beams provided increase of load capacity and increase of stiffness due to the reduction in the opening of the cracks and consequently to a better use of the sheets of strengthening; The presence of the material CCAD in the tensile zone that was completely reconstituted provided a better use of the resistant capacity of the flexural strengthening with sheet of CFRP; The partial reconstitution of the tensile zone did not significantly improve the beam stiffness, and not even the load capacity, when compared to the beam that was only strengthened.For the history of

Figure 2 .
Figure 2. General setup of the three point test of prismatic specimens.

Figure 4 .
Figure 4. Beams of the groups C and D.
Figure 6 zone of th Test con The Structu indicate an actu nomina the spe Figure 7. Schema of the gro viors.Aspect nce of the c shear span d of the beams.

Figure 10 .
Figure 10.Evolution of the strains of the strengthening in the middle of the span versus load.

Table 2 .
Loads and strengths calculated according to Rilem

Table 3 .
Characteristics of the reinforced concrete beams.
Evolution of the vertical displacement of the beams in the middle of the span versus load.

Table 5 .
Loads supported by the beams for the limit state of service.