NUMERICAL STUDY ON THE INFLUENCE OF SUBTERRANEAN LEVELS ON THE SEISMIC RESPONSE CHARACTERISTICS OF BUILDING FRAMES

Abstract

The seismic interaction of the structures with compliant soil has been the subject of various experimental and numerical studies assuming the elastic or inelastic behavior of both structure and foundation soil. These coupled interactions are complex and significantly important in understanding and interpreting the real performance. The seismic performance of building structures is often influenced by the interaction between various integrated components such as superstructure, foundation, subterranean levels, and the supporting and surrounding soil deposit with the effects of dynamic filtering, which is widely referred to in the literature as soil-structure interaction (SSI). The seismic SSI analysis can be helpful in realistically predict the deformation characteristics, seismic response force demand, and the behaviour of the entire system by incorporating the soil strata as a single integral compatible structural component. Many studies have been carried out in recent decades with a strong emphasis on seismic SSI. Most investigations to far have focused on buildings with shallow and deep foundations, including piles. Because of fast urbanization or infrastructural growth, as well as a lack of space in cities, the majority of buildings in metropolitan areas are often planned with one or more subterranean levels for parking and other amenities. Very limited experimental and numerical studies have been presented on the building frame with subterranean levels. However, numerous issues remain concerning the effects of the subterranean levels on the design and seismic response parameters of multi-storey buildings. In the present study, several novel objectives are identified and investigated. (1) to evaluate the influence of subterranean levels on foundation input motion (FIM); (2) to appraise the applicability of theoretical solutions to predict the FIM in comparison to the nonlinear numerical model incorporating the kinematic SSI effect; (3) assess the influence of different modeling approaches and subterranean levels on the seismic response history analysis of the superstructure; and (4) investigate the effect of subterranean levels and their embedment depth on the seismic response of building frame considering SSI. As a result, to effectively address these issues, realistic numerical models are developed and used in a practical manner. These models encompass a wide range of soil types and soil-structure system (SSS) under the influence of various earthquake ground motions. vi A 15-storey medium-rise reinforced concrete moment resisting frame (RC-MRF) building with subterranean levels resting on a raft foundation was adopted as a reference for various investigations in the present study. Three deep homogenous subsoil conditions underlain by bedrock were adopted: soil types C, D, and E, denoting dense, medium, and soft soil deposits, respectively. Three alternative base conditions, five different models, and seven earthquake input motions have been considered. The comprehensive numerical studies was conducted utilizing the finite element method (FEM) by ABAQUS software to address the above-mentioned issues. Several indicators for both soil and structural response are established in terms of response spectral acceleration, transfer functions with respect to free-field motion (FFM), and seismic response demands such as lateral displacements and drifts. According to the investigation conducted in the present study, the major findings are given below. From the numerical analyses, there is in general a reduction of the motion at the subterranean level of a building as the subterranean level depth increases, whereas in contrast more intense than the FFM revealed especially at lower periods and for profounder depths of embedment, predominantly in soil type E (soft soil). Similarly, the effect of soil properties on foundation input motion (FIM) exhibits the same characteristics as the stiffness of the soil properties decreases from medium-dense to soft soil. These results demonstrate clearly how the embedded stiff subterranean level existence in different subsoil conditions causes the high frequencies to be filtered or the reduction of FIM with respect to FFM. The applicability range of theoretical transfer function models was also compared with numerical model results. The models properly anticipate a constant transfer function value for higher periods, which is consistent with the numerical analyses, whereas the divergence is more significant, especially within a small time range. Variations between analytical and numerical transfer function models could also be due to the underlying assumptions applied to build theoretical models. It can be seen that the numerical models can estimate more consistently taking into account the effects of embedment depths and soil nonlinearity. Nevertheless, the analytical models fail to account for certain crucial features such as subterranean levels flexibility, nonlinear behavior of soil deposit, and frequency-dependent amplification of foundation level motion relative to the FFM. vii From the major findings of the numerical investigations in the present work, the interaction between the superstructure, subterranean levels, foundation, and subsoil conditions plays a substantial role in altering the seismic response behavior of the building. It is observed that the superstructure seismic response demands in terms of storey level relative lateral displacement and inter-storey drift ratio (DR) values are intensified when incorporating the subterranean levels and SSI, as well as due to the variation of soil density. The influence is more considerable predominantly in the case of flexible-base models resting in medium and soft soil. On the other hand, generally, when the embedment depth of the subterranean levels increases the storey level lateral displacements and DR values are decreased. Based on the undertaken investigations, it can be concluded that the nonlinear seismic design of an RC-MRF building structure excluding the substructure level and SSI is not adequate to assure structural safety. Also, various modeling strategies often used for the analysis of the seismic response of building structures considerably alter the seismic response characteristic and demand of the superstructure. Although building codes used simplified procedures where the base of the superstructure at the ground surface is assumed to be fixed and the analysis is carried out under the influence of FFM, may often be correct enough. According to the investigation in this study, however, the incorporation of the influence of subterranean levels and seismic SSI is required to predict accurately the seismic response characteristics and demands of the superstructure with great rigor, especially the building structure resting on medium and soft soil deposits with shallow embedment depths.