Backpropagation Neural Network in Flexural Members: Prediction of Capacity Reduction of Beams with Cold Joints using the Angle of Inclination and Discontinuity Location
Dante Silva, Aurus Jodeo Tiam, Kevin Lawrence M. de Jesus, Raphaela Lois Ejera, Bernard S. Villaverde, Reiana Dennise Sarmiento, Ray Armand Gappi, Broderick V. Flores
Abstract
In practically any organization, delays are a persistent concern, and the construction sector is no exception. Construction joints, whether intentional or unintentional, are unavoidable in the construction industry particularly for concrete structures. In this study, the two major variables of cold joints were used as a factor in analyzing the beam's flexural strength using Artificial Neural Network, Garson's Algorithm, and ANSYS Software. The major purpose of this study is to generate a Backpropagation Neural Network-based model for forecasting beam flexural strength while considering two variables: the cold joint surface position and the angle of inclination. The study aimed to understand the performance and behavior of an ordinary reinforced concrete beam having cold joint with different conditions based on these major variables. After conducting data gathering, the study determines the Capacity Reduction of Beams considering the Angle of Inclination and Location of Cold Joint using ANSYS Software which were interpreted and were utilized for the formulation of the model which helped project load capacity of beams with cold joint. Based on the sensitivity analysis, it was conclusive that, in order of importance, the angle of inclination is the variable that is considered to have an impact on beam capacity.