Ground Response Analysis: Comparison of 1D, 2D and 3D Approach

Abstract

Generally, one-dimensional (1D) wave propagation or ground response analysis (GRA) is preferred to evaluate the effect of local site conditions subjected to an earthquake ground motion [1]. For a site with complex and irregular stratigraphy, two-dimensional (2D) and three-dimensional (3D) ground response is preferred over 1D wave propagation for more realistic evaluation of ground response under seismic load. In present study, 1D, 2D (plane strain) and 3D (solid) finite element (FE) models are developed using Abaqus [2] considering two different soil profiles (with multilayer linear viscoelastic materials) of different dynamic characteristics. The kinematic constraints are used along the lateral boundaries of FE models, whereas the base is considered to be fixed in vertical direction. The maximum size of used elements is selected according to the recommendation of ASCE/SEI 4-98 [3] and ASCE/SEI 4–16 [4], for wavelength corresponding to 10 Hz. A recorded ground motion is applied at base of FE models (1D/ 2D/ 3D) and 1D wave propagation model developed in SHAKE2000. The simulated ground motions are compared in terms of transfer functions. On the other hand, a recorded ground motion is de-convoluted through the considered soil profiles using SHAKE2000 [5] separately and then de-convoluted ground motion is applied at the base of FE models. Again, the simulated ground motions (in terms of response spectra and acceleration time history) of different FE models are compared with the recorded one. The results are found to be in excellent agreement for all the considered cases.