Reliability analysis of strength models for CFRP-confined concrete cylinders
Afaq Ahmad, Qaiser uz Zaman Khan, Ali Raza
Abstract
The present work describes a reliability analysis of strength models for predicting the confinement effect of concrete cylinders with Carbon Fibre Reinforced Polymers (CFRPs) through the Artificial Neural Network (ANN). For the reliability analysis, not only current available equations are used but present work also proposes the new empirical equation (Eq-Q) based on the previously conducted experimental work. For which, the database of 708 samples is developed with the detailed information of critical parameters, i.e.; diameter of cylinder (d), height of cylinder (h), unconfined concrete strength ( f c o ' ), thickness (n t ), elastic modulus of CFRP (E f ), and the ultimate strength of confined concrete ( f c c ' ). The aim of this investigation is achieved in two phases: In first phase, the comparison of current available equations with the proposed ANN model and proposed Eq-Q. In second phase, for the validation purposes of proposed ANN model and Eq-Q, new experimental scheme is developed for the concrete cylinders confined with CFRP and its nonlinear finite element analysis (NLFEA). The NLFEA model after calibration with experimental results is used for further parametric studies . The Comparative studies show that the proposed ANN model, proposed Eq-Q and NLFEA results are in close agreement with the experimental results and give better results as compared to their counterparts.