DYNAMIC ANALYSIS OF SOIL-STRUCTURE INTERACTION ON EARTHQUAKE-RESISTANT BUILDING USING SCALED BOUNDARY FINITE ELEMENT METHOD

• Main Author: ABDILLAH, ALFI
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/61000
• Institution: Institut Teknologi Bandung

It widely proven that the soil-structure interaction is suggested to be considered in the dynamic analysis of earthquake loads, especially on soft soils. As the field of soil-structure interaction on structures evolves, various methods have been developed to calculate the structural responses due to the dynamic behavior of the soil. An easier and commonly used method is the soil-spring method where the soil is modeled as point springs at the soil-structure interface that includes pairs of stiffness and radiation damping values. This method is stated in SNI 1726-2019 which refers to NIST GCR 12-917-21. The more advanced method is to model the structure, substructure, and soil with finite elements. The structure as a bounded domain is modeled using the finite element method, while the soil, which is an unbounded domain, is modeled using the scaled boundary finite element method (Song, 2018). This method (SBFEM) only discretizes the boundary region which results in a reduction of the spatial discretization by one. The analytical solution in the radial direction permits the boundary condition at infinity to be satisfied exactly. The dynamic behavior of the soil is obtained by using the infinitesimal finite-element cell method (Wolf & Song, 1994) which generates the unit-impulse acceleration response matrix in the time domain and dynamic stiffness matrix of the soil in the frequency domain. Research on soil-structure interactions in 2D earthquake-resistant buildings (special moment resisting frame with raft foundation) is carried out using two numerical methods, namely soil-spring method and combined method (FEM-SBFEM). Structures and soils are modelled linearly and subjected to time history loads. Comparisons are made between the two methods with variations of site condition (stiff and soft soil). Direct integration with the Newmark method were carried out to obtain the response of the soil-structure due to the time history loads which were scaled to the target response spectrum. Based on this study, it is found that the roof displacement and the base shear from the combined method are smaller when compared to the soil-spring method and the fixed-based method. For stiff soil (Site class D), base shear of fixed-based can be reduced by 33.76% with soil-spring method and 64.16% for combined method. For soft soil (Site class E), base shear of fixed-based can be reduced by 68.42% with soil-spring method and 76.29% for combined method. In conclusion, soil-structure interaction modeling with combined method (FEM-SBFEM), although needing more work and more advanced analysis, yields more economical results especially in soft soil. However, the weakness of the combined method in this study is that it has not been able to model heterogeneous soils for each depth and nonlinear parameters. And the results of this study are also limited to 2D linear time history analysis.