No abstract

The uniform hazard spectrum (UHS) and conditional mean spectrum (CMS) are commonly used as target spectra in selecting and scaling of records to be used in nonlinear response history analysis (NLRHA). When using CMS with tall buildings, CMS ground motions conditioned at multiple natural periods of the buildings should be considered. The application of CMS ground motions in NLRHA to estimate seismic demands for design of tall buildings located on soft-soil layers in Bangkok is investigated in this study. The seismic demands computed using multiple sets of CMS ground motions are compared with those computed using a single set of UHS spectral matching ground motions. Four existing tall buildings subjected to earthquake excitations in Bangkok were considered. The NLRHA was conducted using multiple sets of CMS ground motions, where periods of interest T  were considered at the periods closest to the periods of the first-three translational modes of the building in the direction of excitation. It was found that CMS ground motions conditioned at the higher-mode periods result in larger force demands than CMS ground motions conditioned at the fundamental period for some locations along the height of the building. The envelope of demands obtained by using multiple sets of CMS ground motions conditioned at different periods should be used in design but requires significant computational effort. Using UHS spectral matching ground motions can provide results close to such an envelope and reduce the computational effort significantly.

Abstract. ASCE 7 allows structural engineers to use Response Spectrum Analysis (RSA) procedure to compute the seismic design forces of the structures. However, seismic shear demands of reinforced concrete (RC) walls determined by RSA have been found to be inadequate by many researchers. This paper aims to investigate the seismic shear demands of RC core walls from low-rise to high-rise buildings. RC split core walls in five buildings varying from 5 to 25 stories subjected to earthquake ground motions in Bangkok and Chiang Mai of Thailand were first designed by RSA procedure in ASCE 7-10. Then nonlinear response history analysis (NLRHA) was conducted to compute more accurate seismic demands of the structures. The results demonstrated that shear demands of core walls from NLRHA were significantly larger than those from RSA procedure. The shear amplifications of core walls in cantilever-wall direction were larger than those in coupledwall direction. The two building locations having different spectrum shapes led to different shear amplifications. Hence, an empirical formula cannot be applied to every location. In Bangkok, it is found that Rejec et al. (2012)'s equation could well estimate shear forces in cantilever direction of core walls but it significantly overestimated shear forces in coupled direction of core walls. In Chiang Mai, Luu et al. (2014)'s equation provided good estimation of shear forces in both directions of core walls. Beside these two equations, the shear magnification factor equation in EC8 is found acceptable to be adopted to multiply with shear force from RSA procedure before using it as design shear force of RC core wall in both Bangkok and Chiang Mai.

An over-strength factor in seismic design plays an important role in computing actual forces in a structural member designed to remain elastic. However, sources contributing to this over-strength have not yet been systematically quantified for tall buildings. This paper aims to investigate the contribution from different sources of the over-strength factor in a reinforced concrete (RC) tall building. The effect of how floor slabs are modeled in nonlinear structural models to compute lateral load capacity of the building is also investigated. 39-story RC building subjected to earthquake ground motions in Bangkok was first designed according to the current building codes. Then, pushover analysis was conducted to compute lateral load capacity of the building with three different specified strengths: design strength (with  factor), nominal strength (without  factor), and actual strength (with material over-strength). It was found that modeling floor slabs by elastic shell elements in nonlinear structural model should not be used in computing the ultimate lateral load capacity of the building because the contribution from slab-column framing action is unrealistically large at large roof displacement. When floor slabs are modelled by inelastic effective beam width approach, slab-column framing action contributes about 60% of the ultimate lateral load capacity of the building. The building has an overall lateral over-strength factor of 3.36 to 3.71. The over-strength factor arising from design process is 2.12 to 2.42 in which the contributions from strength reduction factor, material over-strength, and other sources involving the design requirements are about 1.10, 1.17, and 1.77, respectively. The over-strength factor arising from redundancy due to the redistribution of internal forces is about 1.55 and the contribution from steel strain hardening to the over-strength factor is relatively small.

No abstract