Behavior of confined concrete columns under axial load and flexure

Sixteen large size reinforced concrete specimens were tested under combined axial force and flexure. The main purpose of this research program was to investigate the influence of several variables on the behavior of concrete columns confined with rectilinear ties. Variables considered in this program included: (1) Distribution of longitudinal and lateral steel (steel configuration), (2) Level of axial load, (3) Spacing of ties, and (4) Amount of lateral steel.; All the specimens were 12 in. (305 mm) square and 9 ft (2.74 m) long except one which was 13 in. (330 mm) square. The specimens were tested under monotonic flexure to large inelastic deformations while simutaneously subjected to constant axial loads. The axial load varied from 0.46f{dollar}spprimesb{lcub}rm c{rcub}{dollar}A{dollar}sb{lcub}rm g{rcub}{dollar} to 0.78f{dollar}spprimesb{lcub}rm c{rcub}{dollar}A{dollar}sb{lcub}rm g{rcub}{dollar}. The volumetric ratio of the tie steel varied between 0.8% and 1.6%. The tie spacing ranged from {dollar}2{lcub}1over8{rcub}{dollar} in. to {dollar}6{lcub}13over16{rcub}{dollar} in. (54 to 173 mm).; Test results showed that better distribution of laterally supported longitudinal bars result in better performance of columns. Small tie spacing and large lateral steel content increased flexural strength and ductility of confined concrete. High axial force significantly reduces ductility of concrete columns. An amount of lateral steel larger than that specified in the ACI Code is required to ensure ductile behavior of columns under combined high axial load and flexure. Unsupported longitudinal bars and crossties with 90{dollar}spcirc{dollar} hooks confine concrete effectively at small deformations. However, at large deformations, buckling of unsupported bars and opening of 90{dollar}spcirc{dollar} hooks result in a rapid deterioration of the column behavior.; Four confined concrete models available in the literature were used to predict the experimental results. Only one analytical model, proposed by Sheikh and Uzumeri, provided consistent results. This model was modified to include the effects of strain gradient and the level of axial load. Predictions from the modified model yielded excellent comparisons with the experimental results.