Application of artificial neural network (ANN) for estimating reliable service life of reinforced concrete (RC) structure bookkeeping factors responsible for deterioration mechanism

摘要

Degradation of RC structures due to corrosion induced mechanism in the reinforcing steel is a serious durability problem worldwide. It occurs essentially when the reinforcement within the concrete is subjected to marine or aggressive environment. The aim of the present work is to predict the reliable service life of the RC structures by taking into consideration of various prominent models of corrosion and comparing the output with the predicted output of ANN model. Parametric studies have been conducted on four different models to study the effect of various parameters such as corrosion rate, cover thickness, bar diameter, and perimeter of bar which actively participates in the time dependent degradation of RC structures. The outcomes of the parametric inspection of the four chosen degradation models are shown in the present study. The acceptability of the prediction models in forecasting the service life of RC structures are shown through circumstantial illustrative analysis and the best suited model sorted out. However, with the application of soft computing such as ANN, a prediction has been made to determine the service life of RC structures, and the predicted outputs validated with the intended outputs thereby yielding good outcomes for envisaging service life of RC structure.