Developing machine learning models for estimating the shear capacity of barbell squat shear walls under seismic forces

Reinforced concrete shear walls are crucial in providing lateral resistance and stability to structures during seismic events. Despite extensive research on shear walls, a significant gap exists in accurately estimating the shear capacity of barbell squat shear walls, especially when subjected to seismic forces. This gap presents a critical challenge in ensuring the safety and resilience of structures in earthquake-prone regions. Accordingly, the study aims to develop advanced machine-learning models to predict the shear capacity of barbell squat shear walls under seismic loads. By utilizing a comprehensive dataset of experimental results, the study seeks to create high-accuracy predictive models. The significance of this research lies in its potential to enhance the understanding and prediction of shear wall behavior for earthquake-resistant design, contributing to the construction of safer and more resilient buildings. Furthermore, the machine learning models developed in this study can be a valuable tool for engineers and researchers for preliminary assessment of shear capacity within the range of the compiled experimental database, while recognizing that response-related variables such as εcc limit direct design-stage use unless omitted or estimated independently. Additionally, feature importance analysis is conducted to identify the most influential variables affecting shear capacity. This approach improves model transparency and offers more profound insights into the key factors affecting the performance of barbell squat shear walls under seismic conditions. The outcomes of this research are expected to bridge the current knowledge gap and pave the way for more informed decision-making in seismic design practices.