Buildings - Special Topics in Structures
Full Session with Abstracts
339370-2 - Failure-dependent shear strength model for RC beam column joints using Machine Learning Techniques
Saturday, April 21
8:00 AM - 9:30 AM
Beam-column joints are one of the critical elements that govern the overall performance of a building under the seismic loading. This paper presents the application of machine learning techniques to predict the mode of failure and shear strength of beam-column joints. To evaluate the efficiency of various machine learning techniques, extensive experimental data comprising of 536 experimental tests was used. It has been seen from the comparison that Lasso Regression has the better efficiency in classification and prediction. An easy to use equations are suggested in this paper for identifying the failure mode and to estimate the shear strength capacity. The suggested equation can be used by the engineers to determine the mode of failure and associated shear strength of beam-column joints.