A comparative study of strength assessment methods for RC columns

Abstract

Realistic strength assessment of reinforced concrete structural elements, especially columns in bridges and tall buildings is a critical need not only at design time, but also when an accurate evaluation of the strength is needed for decisions such as retrofit or replacement of an existing structure. Assessment of the flexural strength of a column under a specific axial load level is usually done by constructing the axial force-bending moment interaction response curve of the section. This assessment can be done using the code procedure. However, the code does not consider the confinement effect, and is based on the “stress block” assumption for a pre-assumed failure strain for concrete. It has been shown by various experimental and analytical studies that the performance of a reinforced concrete section is affected by different factors such loading history and material behavior. A realistic performance assessment should consider not only proper models for the monotonic and cyclic response of the material, but also analytical methods and procedures that can capture the effects of loading pattern and provide realistic predictions of the section capacity. Accuracy of the analytical methods in strength assessment of reinforced concrete sections was explored in a comparative study. These methods were compared and validated against the existing experimental data. The factors considered in these analytical procedures, included the effect of confinement, and the method employed in assessment of the axial-force-bending moment interaction response of a column section. The experimental data were collected from tests conducted on circular and rectangular columns under a constant axial load. It has been shown that the axial force-bending moment interaction curve, constructed based on the moment-curvature response of a section using a more detailed analytical method such as fiber-model, considering the confining effect of the lateral reinforcement, represents the most realistic and optimal response of a cross section.