Prediction of material property of RC beams-experimental and machine learning approach
摘要
For precise assessment of the modulus of elasticity of reinforced concrete (RC) beams, structural analysis relies on the cross-sectional area, moment of interia, and percentage (%) of tension reinforcement (pt) and compression reinforcement (pc). The behavior of steel and concrete materials in the elastic zone is properly quantified by most international codal provisions. Indian codal standards, like those of all other international codes, express the concrete’s modulus of elasticity in terms of its characteristic strength (fck). Contemporary exercise of design requires an additional set of scrutiny in addition to design. The initial set of analysis will result in the conventional estimation of the depth and corresponding reinforcing area. By considering the amount of steel in tension as well as compression zone, the actual portion of concrete and the accompanying moment of inertia may be obtained by applying the premeditated portion of steel in the initial set. At this stage, material and geometrical characteristics are essential. Finding the modulus of elasticity of a reinforced concrete beam with pt and pc between 0.193% and 0.535% is the current goal of this study. The beam cross-section selected in experimental program was 250 mm × 500 mm and models were cast with M35 grade of concrete. The machine learning approach is used to support the experiment’s findings. An attempt has been made to propose an empirical relation of effective modulus of reinforced concrete beams ERCC that can be predicted for a given set of concrete grade (fck) and reinforcement (pt and pc).