SUMMARYThe original structural design of this case study consisted of five basement floors and a 34-story hotel tower in Kaohsiung, Taiwan. The construction started in 1993, and the erection of the entire steel frame and the pouring of concrete slabs up to the 26th floor were completed before 1996. However, construction of the original hotel was subsequently suspended for 10 years. Recently, this building has been retrofitted for residential purposes. Buckling restrained braces (BRBs) and eccentrically braced frames were incorporated into the seismic design of the new residential tower. This paper presents the seismic resisting structural system, seismic design criteria, full-scale test results of one BRB member and the as-built welded moment connections. Test results confirm that the two side web-plate stiffening details can effectively improve the rotational capacity of welded moment connection. The paper also discusses the analytical models for simulating the experimental responses of the BRB members and the welded moment connections. Nonlinear response history analyses (NLRHA) indicate that the inelastic deformational demands of the original and the redesigned structures induced by the maximum considered earthquakes are less than those found in the seismic building codes or laboratory tests. This paper also proposes a ground motion scaling method considering multi-mode effects for NLRHA of the example building. It is shown that the proposed scaling method worked well in reducing the scatter in estimated peak seismic demands.

As nonlinear response history analysis (NLRHA) becomes a frequently used procedure for the seismic demand evaluation of multistory buildings, it becomes increasingly important to develop a ground motion scaling method which properly includes the dominating modes in the seismic demand estimates. This paper proposes a multimode ground motion scaling (MMS) method, which applies the square root of the sum of the squares (SRSS) or complete quadratic combination (CQC) rule in computing peak seismic demands. The aim is to minimize the weighted sum of the square differences between the spectral responses of a given ground motion and the design response spectrum for the first few modes. Using four case studies, this paper compares the effectiveness of the MMS method with the other common scaling procedures. It is illustrated that the MMS method is effective in reducing the scatter in the peak seismic demands computed from both the response spectral analysis (RSA) and the NLRHA. Recommendations for selecting the ground motion records for the application of MMS method are also provided.