Acta Polytechnica CTU Proceedings doi:10.14311/APP.2019.22.0123 Acta Polytechnica CTU Proceedings 22:123–127, 2019 © Czech Technical University in Prague, 2019 available online at http://ojs.cvut.cz/ojs/index.php/app MECHANICAL PROPERTIES IMPROVEMENT OF FIBER REINFORCED CONCRETE Jan Trejbala, ∗, Václav Nežerkaa, Radim Hlůžeka, Zdeněk Prošeka, b a Czech Technical University in Prague, Faculty of Civil Engineering, Thákurova 7, 166 29 Prague, Czech Republic b Czech Technical University in Prague, University Centre for Energy Efficient Buildings, Třinecká 1024, 273 43 Buštěhrad, Czech Republic ∗ corresponding author: jan.trejbal@fsv.cvut.cz Abstract. Fiber reinforced concrete mechanical properties are limited due to low adhesion be- tween polymer fibers and cement matrix. To ensure a strong interaction between the two materials, polypropylene fibers (d = 0.305 mm) were modified by an oxygen plasma treatment. The interface interaction was moreover activated using finely ground concrete recyclate, whose individual grains (1–64 µm) ensure an adhesion improvement in interfacial zones. The adhesion enhancement was verified by pull-out tests, when reference and modified fibers were pulled-out from cement matrix specimens. Such obtained results were used as a crucial parameter to numerical simulations of bending tests of specimens (550 × 150 × 150 mm) with properties following fiber reinforced concrete. It was shown that samples reinforced with modified fibers and contained activating recyclate reached on higher residual bending strength then those with reference fibers. Keywords: Fiber reinforced concrete, macro fibers, polymer fibers, interfacial shear stress. 1. Introduction Fiber reinforced concrete (FRC) has became popu- lar at production of prefabricated concrete materials, shotcretes, and industrial high-loaded floors. Such material is composed from polymeric macro-fibers (amount ca. up to 1 % vol. of whole mixture), cement, and aggregate [1, 2]. Technical standards EN 14845-1 and EN 14845-2 describe FRC as structural concrete reinforced with fibers having static effect and fulfilling requirements of EN 14889-2. During three-point bending test of notched specimens 550 × 150 × 150 mm, such FRC has to exhibit, besides, residual strength at least 1.5 MPa at crack mouth opening displacement (CMOD) of 0.5 mm (corresponding deflection 0.47 mm). It is clear that such behavior differentiates the FRCs from strain hardening or engineered composites, where strain-hardening response is required after reaching the elastic limit [1]. D. J. Kim et al. explained that strength limit of a fibrous composite material (including FRC) is a function of fibers volume, fibers length to diameter ratio, and the interfacial interaction between fiber surfaces and matrix. In the field of FRCs, it means that increasing fibers amount weakens the cement matrix mechanical properties in the stage of elastics response during loading. It is therefore clear that fibers amount should be as small as possible. On the other side, once matrix limit of proportionality is reached and the matrix is damaged by the crack, fibers transfer the acting stress across the crack (crack bridging) and thus ensure macroscopic integrity of whole material. Amount of stress transfered via fibers depends especially on their number and on adhesion between fiber surfaces and the cement matrix [3]. However, such adhesion is mostly too poor, espe- cially between polymeric fibers and the cement matrix due to fibers smooth and chemically inert surfaces (re- lated to cement matrix) [4, 5]. Mechanical potential of fibers – tensile strength – is therefore unused. To avoid issues connected with the poor adhesion be- tween the two materials, some researches have applied additional treatment of fibers in order to decrease their surface free energy and to increase their mor- phology, both ensuring improvement of bond to the matrix. For these purposes, several types of treatment may be employed, e.g. chemical (use of high alkali solutions) and physical (mechanical roughening) [6–8]. Plasma modification has shown to be a promising technology, combining both the chemical (etching) and physical (roughening caused by a ion bombard- ment) treatment, as proven by Li et al. from the early 1990s [9, 10] and many other researches later, e.g. [11– 14]. It is also worth noting that such treatment has been extended through many industrial fields over the past few years, especially for the surface treatment (roughening, activating, cleaning) of polymeric materi- als [15]. Therefore, there is no obstacle to apply such technology during surface treatment of the fibers. Although a benefit of fiber surface treatment was proven from the perspective of “surface science” many times, this was not achieved from the practical point of view, including the field of FRCs. To connect theoretical findings with praxis of civil engineering, we studied an influence of plasma modified fibers on 123 http://dx.doi.org/10.14311/APP.2019.22.0123 http://ojs.cvut.cz/ojs/index.php/app J. Trejbal, V. Nežerka, R. Hlůžek, Z. Prošek Acta Polytechnica CTU Proceedings mechanical properties of FRC samples using numerical simulations, following EN 14845-1. 2. Interaction between fiber and matrix Post-cracking response of FRC is influenced by be- havior of one fiber that is pulled out from the matrix. This phenomenon was described by Ch. Li et al. and C. Redon et al. [16, 17]. The behavior is divided into two stages, the first describes chemical interaction between the two materials (Pdeb), while the second mechanical interaction activated by fiber movement out of the matrix (Ppull), concretely: Pdeb = √ π2Efd 3 f 2 (τ0u + Gd) (1) and Ppull = πdfτ0  1 + β ( u − ( 2τ0L2e Ef df + √ 8GdL2e Ef df )) df   × [ Le − u + ( 2τ0L2e Efdf + √ 8GdL2e Efdf )] , (2) and τ0, as a frictional stress after sudden drop fol- lowing the peak pull-out load, by: τ0 = Ppull πdfLe (3) where Ef is Young’s modulus of elasticity of fibers; df, fibers diameter; τ0, interfacial shears tress be- tween fiber surface and matrix; u, fiber free-end dis- placement; Gd, interfacial bond strength; β, a shear retention factor, parameter describing slip soften- ing / hardening behavior; Le, fiber embedded length. 3. Fibers and their treatment 3.1. Polymeric fibers Polymeric macro-fibers were used for all experiments described below. Their geometrical and mechanical properties were as follows: material, polypropylene (PP); morphology, smooth; length, 60 mm; diameter, 305 µm; density, 900 kg/m3; Young’s modulus of elas- ticity, 6.1 GPa; tensile strength, 440 MPa; elongation, 8 %. Mechanical properties have been determined experimentally, as reported in [18]. 3.2. Fiber Treatment Two fiber types were used: reference (further marked as R) and plasma treated (P30 and P120 according to treatment duration). Plasma treatment was executed using Tesla VT214 device. Treatment parameters were: plasma, cold; gas, oxygen; power of RF source, 100 W; gas pressure, 20 Pa; treatment duration, 30 and 120 seconds. Name Cement Recyclate W/C [wt. %] [wt. %] [-] Ref 100 0 0.40 Rec 70 30 0.41 Table 1. Weight proportions of cement matrices. 4. Pull-out tests Two matrices were used to carry out pull-out tests. The reference matrix (Ref) was made from Portland cement CEM I 42.5R and the modified matrix (Rec) contained 30 wt. % concrete recyclate as a substitu- tion for cement at the form of finely ground powder. The recyclate was used to fill interfacial zones be- tween fiber surfaces and surrounding matrix and thus to ensure the adhesion between the two materials with individual grains. Grain size differed between 1–64 µm, as measured by Blain’s method. Matrices compositions are summarized in Table 1. Specimens made from such matrices had dimensions equal to 25 × 20 × 30 mm, contained a single fiber in their cen- terline (fiber embedded length was equal to samples height – 30 mm). Results of this experiment were used as basic input values for numerical modeling of FRC bending tests. The whole experiment was carried out using loading frame Veb Tiw Rauenstein FP100. The specimen was anchored by its matrix body to a static part of the frame, while the single fiber, protruded from the specimen body, was caught by a moving frame part. The experiment was displacement controlled at the constant rate of 3 mm/min, finished after reaching to 4.5 mm of fiber-free end displacement (only R and P30 were further used). Results from pull-out tests – dependence between fiber free-end displacement and force resisting to that – are summarized in Figure 1. It is clear from these results that the maximal force recorded during pull- out reference fiber (R) from the reference matrix (Ref) slightly overcame 10 N, while in the case of 30 seconds plasma modified fibers (P30) and matrix containing concrete recyclate (Rec), the force reached on more than 14 N. Despite of the assumptions, fibers exposed to plasma for 120 seconds (P120) exhibited adhesion to the matrix worse than these modified for 30 seconds. This could be caused by their diameter reduction due to too long treatment. Therefore, such fibers were not used for numerical simulations described below. According to equation (4), shear stresses were calcu- lated from thus obtained results. It was calculated that τ0,R = 3.24 · 105 Pa andτ0,P30 = 4.56 · 105 Pa in the case of reference and 30 seconds modified fibers, respectively. 5. Numerical simulations Numerical simulations followed procedure of three- point bending test described in technical standard 124 vol. 22/2019 Mechanical Properties Improvement of Fiber Reinforced Concrete Fo rc e [N ] 0 2 4 6 8 10 12 14 16 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Fiber Free-end Displacement [mm] R P30 P120 Figure 1. Pull-out behavior of reference and modified fibers. Figure 2. Geometry of FRC specimen (top) and mesh of computational model (bottom). EN 14845-2. As already mentioned in Introduction, notched FRC specimen 550 × 150 × 150 mm, contain- ing 0.5 vol.% of fibers, has to exhibit at least 1.5 MPa of bending strength in post-cracking phase during midspan deflection of 0.47 mm. To avoid lengthy experimental testing, numerical simulation was em- ployed. SHCC material model [19, 20], suitable also for FRCs, was applied. Mash of the 2D model, count- ing 3080 of predominantly triangle linear elements, was created in Salome software [21]. Non-linear nu- merical analysis was conducted using OOFEM soft- ware [22], proceeded in 800 steps. Numerical experi- ment was controlled by displacement. Solution was searched by Newton-Raphson’s method. The stiffness matrix was compiled in 2D plain-stress preposition. Geometry of the specimen and mesh of finite elements are shown in Figure 2. Two calculations were done; the first contained ref- erence, while the second one 30 seconds plasma modi- fied fibers. Mechanical properties of matrix was set to correspond common concrete. All parameters set to numerical model were as follows (reference / modified fibers): E Young’s modulus of elasticity of matrix, 20 / 20 GPa; ν Poisson’s ratio of matrix, 0.2 / 0.2; Gf fracture energy of matrix, 5.0 / 5.0 N/m; ft tensile strength of matrix, 2.5 / 2.5 MPa; softType a parame- ter describing post-peak behavior, 3 / 3: Hordijk’s soft- ening; shearType a parameter describing shear stiffness of cracked material, 1 / 1: constant shear retention; shearStrengthType a parameter limiting the magnitude of resulting shear stress acting on crack plane; 1 / 1: the threshold is set to the value of the tensile strength; Vf fiber volume ratio, 0.005 / 0.005; Df fiber diameter, 0.305 / 0.305 mm; Df fiber length, 60.0 / 60.0 mm; Ef Young’s modulus of elasticity of fibers, 6.1 / 6.1 GPa; Gf shear modulus of fibers, 1.0 / 1.0 GPa; τ0 frictional shear stress between the fiber and the matrix during debonding, 0.324 / 0.456 MPa; f snubbing coefficient; 0.5 / 0.5; kf fiber cross-section shape correction fac- tor, 0.9 / 0.9; FSStype a class describing type of fiber bond shear strength, 0 / 0: constant shear strength; fiberType class of reinforcing fibers, 2 / 2: short ran- domly oriented fibers; nCracks maximal number of cracks, 2 / 2; M exponent related to fiber unloading, 1 / 1; fibreActivationOpening, 10−6 / 10−6; dw0 lower bond allowing to smoothen the traction-separation law for fibers, 10−7 / 10−7; dw1 upper bond allowing to smoothen the traction-separation law for fibers, 10−7 / 10−7. 6. Results It was found from numerical simulations that residual strength of FRC specimen reinforced with plasma modified fibers (P30) at amount of 0.5 vol.% tightly exceeded 1.5 MPa at CMOD of 0.47 mm. In the same stage, the samples containing reference fibers exhibited only ca. 1.3 MPa. Results from both simulations are imagined in Figure 3, where the green line highlights the minimal bending strength required by EN 14845- 2. It is obvious from these results that only FRC containing modified fibers fulfilled these requirements, so this can be considered as structural. These simulations also revealed that the behavior of both specimens was practically identical in phases 125 J. Trejbal, V. Nežerka, R. Hlůžek, Z. Prošek Acta Polytechnica CTU Proceedings Fiber Free-end Displacement [mm] 0 0 0.1 0.2 0.3 0.4 0.5 0.6 1.0 1.5 2.0 2.5 3.0 0.5 St re ss [M Pa ] Midspan deflection [mm] R P30 Figure 3. Normal tension as a function of midspan displacement. until the fibers have not been activated yet. After, fibers bridged the opening crack, transfered acting stress and thus ensured macroscopic integrity of speci- mens. Based on the specimens post-cracking behavior, it is clear that the adhesion between modified fibers and the matrix was increased than in case of reference fibers. This finding proves that interfacial shear stress between the two materials plays an important role from the mechanical response point of view. 7. Conclusions This work deals with searching of mechanical behavior of fiber reinforced concrete using numerical simula- tions of three-point bending tests. Reference and plasma modified polypropylene fibers (d=0.305) were used as reinforcement. Adhesion between these fibers and two types of the cement matrix was examined em- ploying pull-out test. The purpose of the simulation was to find residual bending strength of specimens, following relevant technical standards. Finding were as follows: • Adhesion between 30 seconds oxygen plasma treated fibers and the cement matrix containing concrete recyclate was higher by approx. 15 % then in the case of reference fibers and reference cement matrix. • 120 seconds lasting plasma treatment did not bring any benefits in terms of adhesion improvement. Conversely, such treated fibers showed worse ad- hesion to the matrix than those 30 seconds exposed to plasma. This was probably caused by their di- ameter reduction as a consequence of too intensive ion bombardment. • Numerical simulations revealed that residual strength of reference FRC at the midspsan deflec- tion of 0.47 mm was less than 1.5 MPa, so this material did not meet requirements of technical standard EN 14845-2. On the other side, if plasma treated fibers and the cement matrix containing 30 wt.% of concrete recyclate were used, residual strength overcame minimal 1.5 MPa. Acknowledgements This work was financially supported by Czech Technical University in Prague – SGS project SGS16/201/OHK1/3T/11 and SGS18/037/OHK1/1T/11, and by the Czech Science Foundation research project 17-06771S. Special thanks to Tereza Horová for research assistance. References [1] M. Prisco, G. Plizzari, L. Vandewalle. Fibre reinforced concrete: new design perspectives. Materials and Structures 42(9):1261–1281, 2009. [2] P. Tatnall. Fiber-reinforced concrete. In Significance of Tests and Properties of Concrete and Concrete- Making Materials, chap. 49. ASTM International, 2006. [3] D. Kim, E. Naaman, S. El-Tawil. High performance fiber reinforced cement composites with innovative slip hardending twisted steel fibers. International Journal of Concrete Structures and Materials 3(2):119–126, 2009. doi:10.4334/IJCSM.2009.3.2.119. [4] M. Halvaei, M. Jamshidi, M. Latifi. Investigation on pullout behavior of different polymeric fibers from fine aggregates concrete. Journal of Industrial Textiles 45(5):995–1008, 2006. doi:10.1177/1528083714551437. [5] W. Choi, S. Jang, H. Yun. Interface bond characterization between fiber and cementitious matrix. International Journal of Polymer Science pp. 995–1008, 2015. doi:10.1155/2015/616949. [6] V. Machovič, V. Lapčák, L. Borecká, et al. Microstructure of interfacial transition zone between pet fibres and cement paste. Acta Geodyn Geomater 10(1):121–127, 2013. doi:10.13168/AGG.2013.0012. [7] A. Tagnit-Hamou, Y. Vanhove, N. Petrov. Microstructural analysis of the bond mechanism between polyolefin fibers and cement pastes. Cement and Concrete Research 35(2):364–370, 2005. doi:10.1016/j.cemconres.2004.05.046. [8] L. Yan, R. Pendleton, C. Jenkins. Interface morphologies in polyolefin fiber reinforced concrete composites. Composites Part A: Applied Science and Manufacturing 29(5–6):643–650, 1998. doi:10.1016/S1359-835X(97)00114-0. 126 http://dx.doi.org/10.4334/IJCSM.2009.3.2.119 http://dx.doi.org/10.1177/1528083714551437 http://dx.doi.org/10.1155/2015/616949 http://dx.doi.org/10.13168/AGG.2013.0012 http://dx.doi.org/10.1016/j.cemconres.2004.05.046 http://dx.doi.org/10.1016/S1359-835X(97)00114-0 vol. 22/2019 Mechanical Properties Improvement of Fiber Reinforced Concrete [9] V. Li, Y. Chan, H. Wu. Interface strengthening mechanisms in polymeric fiber reinforced cementitious composites. In Proceedings of conference on Brittle Matrix Composites 4. IKE and Woodhead, Warsaw, 1994. [10] V. Li, H. Stang. Interface property characterization and strengthening mechanisms in fiber reinforced cement based composites. Advanced Cement Based Materials 6(1):1–20, 1997. doi:10.1016/S1065-7355(97)90001-8. [11] B. Felekoglu, K. Tosun, B. Baradan. A comparative study on the flexural performance of plasma treated polypropylene fiber reinforced cementitious composites. Journal of Materials Processing Technology 209:5133– 5144, 2009. doi:10.1016/j.jmatprotec.2009.02.015. [12] J. Trejbal, V. Nežerka, M. Somr, et al. Deterioration of bonding capacity of plasma-treated polymerfiber reinforcement. Cement and Concrete Composites 89:205– 215, 2018. doi:10.1016/j.cemconcomp.2018.03.010. [13] A. Pamreddya, D. Skácelová, M. Haniǎinec, et al. Plasma cleaning and activation of silicon surface in dielectric coplanar surface barrier discharge. Surface and Coatings Technology 236:326–331, 2013. doi:10.1016/j.surfcoat.2013.10.008. [14] T. Homola, J. Matoušek, B. Hergelová, et al. Activation of poly(ethylene terephthalate) surfaces by atmospheric pressure plasma. Polymer Degradation and Stability 97(11):2249–2254, 2012. doi:10.1016/j.polymdegradstab.2012.08.001. [15] J. Reece Roth. Industrial Plasma Engineering. Volume 2: Applications to Nonthermal Plasma Processing. CRC Press, Boca Raton, 2001. [16] V. Li, C. Leung. Steady-state and multiple cracking of short random fiber composites. Journal of Engineering Mechanics 118(11):2246–2264, 1992. doi:10.1061/(ASCE)0733-9399(1992)118:11(2246). [17] C. Redon, V. Li, C. Wu, et al. Measuring and modifying interface properties of pva fibers in ecc matrix. Journal of Materials in Civil Engineering 13(6):399–406, 2001. doi:10.1061/(ASCE)0899-1561(2001)13:6(399). [18] R. Hlůžek. Analysis and modification of interfacial zones between cementitious matrix and fiber reinforcement. Diploma thesis. CTU in Prague, Faculty of Civil Engineering, Czech Republic, 2018. [19] K. Toshiyuki, P. Kabele, H. Fukuyama, et al. Strain hardening cement composites: Structural design and performance. In State-of-theArt Report of the RILEM Technical Committee 208-HFC, SC3. Springer, Dordrecht, 2012. [20] P. Havlásek, P. Kabele. A detailed description of the computer implementation of SHCC material model in OOFEM, CTU in Prague. [21] SALOME Platform, on-line, available from: http://www.salomeplatform.org/user- section/about/mesh. [22] B. Patzák. OOFEM – an object-oriented simulation tool for advanced modeling of materials and structures. Acta Polytechnica 6(6):59–66, 2012. 127 http://dx.doi.org/10.1016/S1065-7355(97)90001-8 http://dx.doi.org/10.1016/j.jmatprotec.2009.02.015 http://dx.doi.org/10.1016/j.cemconcomp.2018.03.010 http://dx.doi.org/10.1016/j.surfcoat.2013.10.008 http://dx.doi.org/10.1016/j.polymdegradstab.2012.08.001 http://dx.doi.org/10.1061/(ASCE)0733-9399(1992)118:11(2246) http://dx.doi.org/10.1061/(ASCE)0899-1561(2001)13:6(399) Acta Polytechnica CTU Proceedings 22:123–127, 2019 1 Introduction 2 Interaction between fiber and matrix 3 Fibers and their treatment 3.1 Polymeric fibers 3.2 Fiber Treatment 4 Pull-out tests 5 Numerical simulations 6 Results 7 Conclusions Acknowledgements References