Prediction of Joint Shear Strength of RC Beam-Column Joint Subjected to Seismic loading using Artificial Neural Network.

Abstract

BC (Beam-column) joints are critical locations in RC frames subjected to severe earthquake attack. Failure of Joint which is of shear type is not an appreciable structural behaviour. Present study proposes an artificial neural network model for joint shearstrength of reinforced exterior BC connections. ANN is a component of artificial intelligence which mimics the human brain characteristics and learns from previous experiences and has recently gained popularity in the area of civil engineering. A dataset of specimen dimensions, material properties, Area of Reinforcement used in Beam and column and failure mode are established from past experimental investigations on BC joints subjected to seismic type loading and are used for ANN modelling. ANN model is developed with eleven input parameters to predict the Joint shear strength of Exterior BC Joints. The Proposed model is compared with the equation given in design codes and empirical formula. Proposed ANN model has predicted the shear strength more accurately. Thus the proposed ANN model can be used for Prediction of Joint shear strength of Reinforced concrete exterior BC joints subjected to seismic loading.