Nonlinear time-history analysis conducted as part of a performance-based seismic design approach often require that the ground motion records are selected and then scaled to a specified level of seismic intensity. In such analyses, besides an adequate structural model, a set of acceleration time-series is needed as the most realistic representation of the seismic action. In this paper, the effects of scaling procedure on seismic demands of steel frames are investigated. To this, two special steel moment-resisting frames with considerable higher mode effects, and two sets of ground motions, including near-fault and far-fault motions are considered. Moreover, three scaling procedures are introduced for performing nonlinear dynamic time-history analysis of structures. Among different demands, lateral roof displacement and interstory drift are selected as seismic demands. The height-wise distribution of demands shows that the inelastic seismic demands of the near-fault pulse-like ground motions differ considerably from those of far-fault ones. These results show that the story drifts are mostly larger for far-fault motions in the upper story levels in comparison to near-fault records and in the lower floors, the reverse is true. Thus, the scaling procedures directly affect the results of seismic demands and choosing different methods would result in varying responses. Moreover, a low-cost and fairly effective procedure is proposed to estimate the target displacement demands of buildings from response-spectrum analyses, considering near-fault effects. The precision of this method is verified by nonlinear time-history analysis results, as the benchmark solution, and acceptable improvements have been achieved.
Nonlinear time-history earthquake analysis for steel frames
