Seismic response and uncertainty quantification of deep soil mixing-improved Nevada sand under unidirectional and bidirectional loading
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Block-type deep soil mixing (DSM) is a widely accepted ground improvement technique for large-scale infrastructure projects, particularly for the foundations of power plants. This study examines the dynamic response of a Nevada sand layer enhanced with DSM through numerical simulations in GID V14.0.1 and OpenSees V3.5.0. The sand was modeled as an elastoplastic material based on the PDMY03 constitutive model, whereas the DSM columns were assumed to be elastic. The models were subjected to Ricker wave excitation and near-fault ground motions characterized by velocity pulses and forward directivity effects, with absorbing boundaries simulated using free-field columns and viscous dampers. A comprehensive parametric study was conducted. Results show that increasing DSM thickness reduces horizontal accelerations but amplifies vertical accelerations due to rocking motion, suggesting a preliminary design recommendation of DSM thickness equal to one-fifth of the soil shear wavelength. Increasing the DSM width has a minimal effect on horizontal accelerations but reduces vertical responses. Lattice-shaped DSM layouts effectively decrease horizontal accelerations while generating higher vertical responses than block-shaped configurations. The type and thickness of DSM significantly influence the redistribution of energy across frequency bands, which should be considered in seismic analyses of power plant structures and equipment. Bi-directional seismic loading minimally affects horizontal accelerations but substantially amplifies vertical responses, highlighting the importance of tri-directional analyses, particularly in near-fault regions. To address uncertainties, six key parameters were investigated using a Design of Experiments approach, and surrogate models were developed to predict the seismic performance of DSM-improved soils.
