PREDICTING THE SEISMIC RESPONSE OF REINFORCED CONCRETE STRUCTURES USING ARTIFICIAL NEURAL NETWORKS

  • Deepthy S Nair, Beena Mol M,
Keywords: Artificial Neural Network; Seismic Response, Reinforced Concrete, Performance-Based Design, Earthquake Engineering.

Abstract

Aftershocks from previous earthquakes have demonstrated that they can exacerbate the vulnerability of structures damaged by mainshocks. However, the primary challenge is to incorporate these threats into the design and retrofit processes. The article describes the use of Artificial Neural Networks [ANNs] to predict the seismic behavior of Reinforced Concrete [RC] structures that have been revealed to seismic events. An ANN system is developed, trained, and validated using existing evaluation details extracted from pertinent documentation on the RC structural elements. The Finite Element Method [FEM] is used in studies to determine the magnitude of vibration-induced structural damage. FEM is effective when evaluating a small number of well-defined structural elements but is ineffective when evaluating larger assets. As a result of these constraints, the model used Artificial Neural Networks to develop a new model for estimating earthquake-induced damage. It was discovered that the lower floors are most exposed, whereas the rooftop is least affected. The method is to define both the entire structure and floor movement using a wide variety of structural and floor motion characteristics. Modeling earthquake technology is a computationally complex domain in which ANNs could be used to simulate the architectural response under stationary or adaptive loads. PBD is a novel concept in structured framework earthquake engineering in which structural efficiency is evaluated across a range of risk scales, necessitating significant computational resources. A modern adaptation technique is suggested for this analysis in order to determine the non-linear structural performance when seismic acts of increased intensity are considered. The projected structural contribution of ANNs could be incorporated into the PBD model while conducting thorough analyses to avoid incurring unnecessary computational costs. The ANN's efficiency was evaluated using a variety of scenarios, and it was determined that the ANN is capable of accurately predicting damages.

Published
2021-10-29
How to Cite
Beena Mol M, , D. S. N. (2021). PREDICTING THE SEISMIC RESPONSE OF REINFORCED CONCRETE STRUCTURES USING ARTIFICIAL NEURAL NETWORKS. Design Engineering, 8239-8249. Retrieved from http://www.thedesignengineering.com/index.php/DE/article/view/5867
Section
Articles