Sustainable strengthening of concrete deep beams with openings using ECC and Bamboo: An equation and data-driven approach through abaqus modeling and GEP

Abstract

• Successfully demonstrated effective use of ECC and bamboo reinforcement in strengthening reinforced deep concrete beams with large openings to address brittle shear failures and contribute to a sustainable construction approach. • ECC configurations increased shear capacity by 13 % to 237 kN at 8.4 mm deflection, and bamboo configurations increased capacity by 10 % to 230.5 kN at 10.2 mm deflection. • Beam depth and compressive strength influenced load capacity positively with correlation coefficients of 0.60 (ECC) and 0.68 (bamboo). • Developed 400 FEM simulations in Abaqus using nonlinear material properties and critical parameters such as compressive strength, beam depth, and hole diameter to determine the best configurations. • Machine Learning Model Genetic Expression Programming (GEP) exhibited high predictive power, with R² > 0.85 through training, testing, and validation. • This study will contribute to sustainable infrastructure development by integrating innovative materials and data-driven approaches for improving the resilience of RC deep beams. This study investigates the innovative strengthening of concrete deep beams with large openings using engineered cementitious composites (ECC) and bamboo reinforcement, addressing challenges associated with traditional reinforced concrete (RC) deep beams prone to brittle shear failures. Initially, the model was validated with experimental data, followed by an analysis of various ECC and bamboo configurations to select the most economical strengthening approach for each material. A comprehensive dataset of 400 Abaqus models was developed, incorporating key parameters such as compressive strength (f’ c ), beam depth (D), span length (L), and reinforcement details. Pearson correlation analysis showed that beam depth and compressive strength positively influenced maximum load capacity, while hole diameter had a moderate negative effect. Gene expression programming (GEP) was used to develop predictive models for shear capacity, providing an interpretable framework for optimizing beam design. Although error analysis indicated areas for further calibration, the GEP models exhibited high predictive accuracy with R² > 0.85 in both training and validation datasets. This research offers a data-driven approach to sustainable beam strengthening, emphasizing the optimization of beam depth, concrete strength, and opening size to enhance structural resilience.

Topics & Concepts