Seismic design of reinforced concrete buildings : an inelastic response spectrum approach

The only rigorous technique for evaluating the response of an inelastic, multidegree of freedom (MDoF) structure to an arbitrarily defined ground motion is a step-by-step integration of the equations of motion. This procedure, while theoretically feasible, is extremely costly and impractical for normal design. The Inelastic Response Spectrum Approach (IRSA), consisting of the superposition of modal responses obtained from inelastic earthquake spectra, is not a rigorous analytical technique because it relies on the principle of superposition which does not apply past the elastic limit. However, the simplicity of the IRSA and its ability to explicitely account for all the important variables have made it a promising procedure in the search for an improved version of the presently applicable earthquake design approaches.

The primary purpose of this investigation was to test the applicability of the IRSA as a tool for the design of moment-resisting, reinforced concrete frames under seismic exitation. The applicability of this method can be measured by the accuracy (or lack of it) of the predicted inelastic demands when compared to those obtained from a rigorous, step-by-step, inelastic dynamic analysis. In order to test the applicability of the method, two frames were designed for various combinations of ground motions and design ductilities and then analyzed using DRAIN-2D, a computer program with inelastic dynamic analysis capabilities.

Another objective of this dissertation was to develop design procedures to insure that the inelastic action in the frames designed according to the IRSA would be restricted to the beams.

From the comparisons between predicted inelastic demands from the IRSA and those obtained from step-by-step inelastic dymanic analyses, it was found that the story drift ductility (defined as the ratio between the maximum story drift and the yield value) is in good agreement with the design ductility, u, used to develop the inelastic response spectrum. It was also shown that the plastic hinge rotations in the beams can be calculated with sufficient accuracy from the structure's yield drifts (which depend on the elastic properties) and the design ductility, µ. Although not enough cases were considered to permit a definite conclusion, the results obtained indicate that, as one may expect, the accurracy of the inelastic demands predicted using the IRSA decreases as the design ductility increases.

In order to insure a strong column-weak beam behavior (i.e., to restrict the formation of plastic hinges to the beams) a special column design procedure was developed (the term, design procedure, is used here to refer to the determination of the critical column actions). This column design procedure was found to be efficient in preventing any significant yielding in the columns of the reinforced concrete moment resisting frames considered in this investigation.