Effect of boundary condition on the cyclic response of I-shaped steel columns: Two-story subassemblage versus isolated column tests

Recent studies on isolated steel columns under combined axial load and cyclic lateral drift showed that the column response is affected by the boundary condition. To consider more realistic boundary conditions of first-story columns in a frame, two-and-a-half story beam-column subassemblages were tested to investigate their behavior under reversed cyclic loading. The subassemblages, composed of a steel column with steel beams at two floor levels, used highly ductile I-shaped sections with web slenderness of either 37 or 49. Two isolated fixed-fixed columns were tested to facilitate a direct comparison with the subassemblage testing. An axial compression force corresponding to 20% of the yield force was applied. Test results showed that the realistic boundary condition at the top end of the first-story column significantly alters the column plastic hinging, moment distribution, and out-of-plane deformation reported previously based on isolated column testing. Plastic hinging of the adjoining beams caused the inflection point of the first-story column to move upward initially, then moved downward when plastic hinging developed at the column base. Out-of-plane deformation observed from isolated columns did not occur in the subassemblage column because hinging did not occur at the top end. Plastic components of the backbone curves for both the subassemblage column and isolated column were very similar; the former delivered a larger story drift angle rotation due to a larger yield rotation, dependent on the elastic lateral stiffness of the column. A closed-form solution that calculates this stiffness for any boundary condition was developed and verified from test results.