Accidental design eccentricity: Is it important for the inelastic response of buildings to strong earthquakes?

Abstract

To deal with earthquake-induced torsion in buildings due to some uncertain factors, difficult to account for directly in design, modern codes have introduced the so-called accidental design eccentricity (ADE). This provision has been based primarily on elastic investigations with special classes of multi-story buildings or with simplified, one-story inelastic models. In the present paper, the effectiveness of this provision is investigated using inelastic models, both of the typical one-story, 3-DOF type, and the more sophisticated MDOF, frame idealizations of the plastic hinge type. One, three and five story, realistic, frame buildings with different natural eccentricities were designed for different ADEs, including those specified by the EC8 and IBC codes. The evaluation is made using mean peak ductility factors of the edge frames as measures of their inelastic response, obtained from dynamic analyses for ten pairs of semi-artificial earthquake motions. The simplified models indicate that the accidental design eccentricity is very effective in reducing ductility demands, especially for very stiff systems. However, this is not confirmed by the more accurate and detailed plastic hinge building models, which show that designs accounting for accidental eccentricity do not exhibit any substantial reduction or better distribution of ductility demands, compared to designs in which accidental eccentricity has been entirely ignored. These findings suggest that the ADE provisions in codes, especially the more complicated ones as in the IBC, should be re-examined, by weighting their importance against the additional computational work they impose on designers. In the cases examined herein this importance can be characterized as marginal. Obviously additional studies are required, to include more building types and earthquake motions, in order to arrive at firm conclusions and recommendations for code modifications.