• P-ISSN 0974-6846 E-ISSN 0974-5645

Indian Journal of Science and Technology


Indian Journal of Science and Technology

Year: 2016, Volume: 9, Issue: 32, Pages: 1-6

Original Article

Prediction of Flexural Performance of Confined Hybrid Fibre Reinforced High Strength Concrete Beam by Artificial Neural Networks


Objectives: This paper presents an outcome of an experimental study to analyse the flexural behavior of confined hybrid fibre high strength over-reinforced concrete with varying volume fraction of steel and polypropylene fibre and to predict the performance using Artificial neural network model. Methods: 13 over-reinforced beams were casted with the size 120mm×200mm×1200mm.Out of this one beam was kept as reference (without fibre) and remaining 12 beams are casted with hybrid fibres with different proportion such as Steel 100% / Polypropylene 0%, Steel 80% / Polypropylene 20% and Steel 60% / Polypropylene 40% . Purposefully all the beams are made as an over-reinforced section in order to ensure the ductile behavior. An Artificial Neural Network has been created to predict the output parameters of the process. Findings: Beam specimens were tested as per ASTM and the deflection were measure for the consecutive loads using dial gauges fixed over the beam. The experimental results revealed that hybrid fiber with volume fraction 2% (Steel 80% -polyolefin 20%) permutation sample develops the flexural performance substantially associated with that of reference beam and steel fiber reinforced high strength concrete beam. Improvements: Use of this fiber reinforced concrete with the confined manner improved the ductile behavior and the energy absorption capacity even if it is an over-reinforced beam. Many networks are developed by trial and error method were in Radial Basis Network is found to have good performance than other network created (Back propagation Network). Artificial neural network has the stuff called ability which approximately resembles to their skill to model any given function. Artificial neural network software is proficient of learning and simplifying from illustrations and familiarities, which makes this network becoming an influential tool to solve many complicated problems in the civil engineering projects. In accordance with that a proposed model are very effective in predicting the performances with a greater accuracy
Keywords: Artificial Neural Network, Confinement, High Strength Concrete, Steel Fibre 


Subscribe now for latest articles and news.