چكيده به لاتين
High performance fiber reinforced cementitious composites (HPFRCCs) are a kind of cementitious composites that exhibit strain hardening behavior after the first crack, instead of a brittle behavior. Engineered cementitious composites (ECCs) are a group of HPFRCC that uses short fibers as reinforcers and have a tensile strain of about 3-6%. In current study, polypropylene (PP) and polyvinyl alcohol (PVA) short fibers and their combinations have been used as ECC reinforcers. Both fibers were used in four different volume fractions of 0.5%, 0.1%, 1.5% and 2.0%. The PP fibers were in four different shapes including hollow, circular, delta and trilubal. The PP fibers were used in lengths of 6, 9 and 12 mm, and the PVA fibers had a length of 6 mm. The bending behavior of the samples was obtained by a universal testing machine. Also, in addition to the flexural strength, the toughness of the samples was also calculated from the area of the load-deflection curve. Then, the flexural strength results were modeled in MATLAB using regression, neural network and ANFIS. The results showed that PVA and PP fibers improved flexural strength and toughness, respectively. Hollow fibers with a length of 12 mm were the best case of using single fibers and when it used in 2 vol.%, flexural strength and toughness increased up to 32% and 37% in comparision to the blank sample, respectively. Also, in combination of two fibers, the sample containing PP delta fibers with a length of 9 mm and a volume fraction of 1.5% and PVA fibers with a volume fraction of 0.5% was the best, which increased the flexural strength and toughness up to 71% and 146%, respectively. Modeling results also showed that neural network and ANFIS are better methods for modeling and provide more accurate results. In neural network, a model with a neuron numbers of 8 and 7 was the best. In anfis, a model with bell-shaped membership function was the most accurate one.
