Enhancing Seismic Resilience in Multi-Storey Structures on Uneven Terrain: A Comparative Evaluation of Response Spectrum and Nonlinear Time History Analyses with AI-Driven Optimization
- Version
- Download 60
- File Size 588.14 KB
- File Count 1
- Create Date 17 September 2025
- Last Updated 17 September 2025
Enhancing Seismic Resilience in Multi-Storey Structures on Uneven Terrain: A Comparative Evaluation of Response Spectrum and Nonlinear Time History Analyses with AI-Driven Optimization
T. Sravya1, P. Praveen2, V. Anusha3, P Mohan4, Matta Venu5
1T. Sravya, PG Student, Department of Civil Engg. & Vikas College of College Engineering & Technology.
2,P. Praveen, Assistant Professor, Department of Civil Engg. & Vikas College of College Engineering & Technology
3,V. Anusha, Assistant Professor, Department of Civil Engg. & Vikas College of College Engineering & Technology
4,P. Mohan, Assistant Professor, Department of Civil Engg. & Vikas College of College Engineering & Technology
5, Matta Venu, PG Student, Department of Civil Engg. & Vikas College of College Engineering & Technology.
Abstract - Combined with the growing requirements of urban development within hilly areas, the number of constructions of multi-storey buildings on sloping slopes has significantly risen, and the topographic aspect of the requirements is extremely irregular, as well as the interaction of soil and structure, which makes the seismic design very challenging. This paper suggests a detailed seismic analysis of the multi-storey reinforced concrete (RC) structures standing on the sloping ground, using the Response Spectrum Method (RSM) as well as the Nonlinear Time History Analysis (NTHA) to analyze the dynamic behavior of such structures under seismic loading conditions. The ultimate goal is to compare the efficiency of the approaches to capture the structural response, i.e. base shear, inter-storey drift and displacement, and introduce the use of artificial intelligence (AI)-based optimization to enhance the efficiency of planning and seismic resistance.
The results of the finite element analysis of a set of multi-storey RC building models of different configurations (e.g., step-back and set-back) on slopes between 10 and 30 are compared. RSM applied in the analysis is based on the linear dynamic analysis and used to evaluate the modal behavior and design forces, as per applicable seismic codes, whereas the NTHA, which accounts nonlinear material and geometrical property is used to model the response of the building to actual earthquake ground motions. The simulations are realistic because site requirements such as soil stiffness, slope and seismic zoning are considered during the analysis. The most important response parameters, such as maximum lateral displacement, inter-storey drift ratios, base shear distribution, and others are compared to determine the conservatism and accuracy of RSM compared to the more computationally intensive NTHA.
The structural performance is further improved by using an AI-based optimization framework that is trained to streamline key design variables, including column sizes, shear wall location and foundation stiffness, as well as other variables, using machine learning algorithms, such as genetic algorithms and neural networks. The results indicate that NTHA would provide a more realistic view of nonlinear behavior particularly when the buildings are sounding on steep slope, but RSM would grossly overstate design forces in certain options. The AI-optimized designs show better seismic performance up to 15 percent material consumption and still with the safety standards. This paper highlights the significance of incorporating new analytical tools and AI-based solutions to develop seismically robust structures on inclined sites and present meaningful information to the engineers and policymakers working in hilly areas with a high risk of earthquakes.
Key Words: Seismic Analysis, Sloping Ground, Response Spectrum, Time History, AI Optimization