Details

The given work is devoted to a combined analytical and numerical investigation of structural performance of carbon fiber reinforced concrete beams subjected to three point bending. Ten different mixes were studied which are a plain C40/50 for reference and nine carbon fiber reinforced concrete mixes which were obtained by combining three fibre volume fractions (Vf = O.5% , 1.0% and 1.5%) with three aspect ratios (l/d  =40, 60,80). The methodology comprised of two complementary stages .In the first stage  an analytical method based on fib Model Code 2010 Chapter 5.6 was created in MATLAB to generate Force-CMOD response curves , residual flexural tensile strengths fR1- fR4, fracture energy and post cracking performance class (a/b/c/d) for each mix. In the second stage the geometry and constitutive parameters obtained analytically were transferred into DIANA FEA where two dimensional non linear finite element models of the  EN14651 notched beam were built. A total strain rotating crack model was used in which plain concrete used the Hordijk exponential softening law , while the other mixes were assigned the MC2010 bilinear post-crack σ–w law calibrated through parameters fFts  and fFtu  derived form the analytical residual strengths. Crack initiation , propagation ,fibre bridging were then visualized for the optimum mix. The combined results identified Vf=1,5% with l/d  =80 as the best performing option consistently. This mix delivered fR1 =1.38MPa , fR3 =0.94MPa , GF= 0,747N/mm  and a composite weighted score 5.95 times higher than the plain reference. All the mixes in the study satisfy the MC2010 minimum requirements for structural use (fR1 ≥ 1.5MPa is approached only by the strongest mix and fR3/ fR1 ≥ 0.5 is satisfied by every mix and none of them reached strain hardening behaviour . This thesis demonstrates that for non deformed straight carbon fibers aspect ratio is many times more efficient lever than volume fraction for improving post crack residual strength and that the MC2010 framework is sufficient to provide reliable input set for nonlinear FEA.