Evaluation of the adequacy of the response spectrum analysis for the seismic analysis of moment-resisting and concentrically-braced buildings according to the seismic design codes

According to the seismic design codes, the response spectrum analysis (RSA) method can be used in the seismic analysis of tall buildings since it can consider the effect of higher modes. On the other hand, the nonlinear time history analysis is the most accurate method of the evaluation of seismic responses of structures. Consequently, it is needed to study the accuracy of the RSA method and to compare the seismic responses computed by the RSA to those from the nonlinear time history analysis. To this end, six 3D structures with 4-, 10- and 20-story heights are investigated in this paper. The lateral load resisting systems of the structures include special steel moment-resisting frames (MRFs) and concentrically braced frames (CBFs). To conduct the nonlinear time history analyses, four sets of ground motion records including three groups of near-fault records with different characteristics and one set of far-fault records are used. The near-fault ground motion sets include forward directivity effect, fling step and no pulse characteristics. All sets comprise seven seismic motion records. The results indicate that the seismic responses, obtained by the RSA, are mostly underestimated and non-conservative in comparison with those from the nonlinear time history analysis. In general, the more the height of the structure, the larger the error in the seismic responses derived from the RSA. Also, the largest error in the RSA relative to the rigorous time history analysis occurs in the case of the near-fault ground motions with fling-step effect.