In an ongoing construction project for a power plant, deep soil mixing (DSM) in a block configuration was proposed to reduce settlement and enhance the bearing capacity of weak subsoil. Given its critical importance for seismic performance under near-fault ground motions, this study aims to evaluate the dynamic response of DSM-improved soils subjected to various seismic excitations, including Ricker waves with varying amplitudes and frequencies, as well as unidirectional and bidirectional near-fault earthquakes, both with and without velocity pulses. The investigation focuses on the influence of key parameters such as incident angle, burial depth, DSM block width, thickness and interaction between DSM-soil-DSM. Additionally, considering the uncertainties in the geotechnical and structural parameters affecting seismic behavior, mathematical and statistical techniques are employed to develop metamodels that provide insight into DSM performance and enable sensitivity analysis of seismic responses.
Two-dimensional elastic finite element models were developed using GID software (version 14.0.2) based on project dimensions, followed by dynamic analyses performed in OpenSees (version 3.5.0). Soil and improved soil properties derived from laboratory tests were assigned to the finite element models. Lateral and bottom boundaries were modeled as semi-infinite to minimize boundary effects. Statistical analyses and metamodeling were conducted using Design-Expert software (version 12) to account for uncertainties in seismic response parameters.
The results of this study demonstrated that DSM can significantly influence horizontal and vertical accelerations at the ground surface. Increasing the DSM thickness led to a reduction in horizontal accelerations but caused an increase in vertical accelerations due to rocking motion. Therefore, increasing the DSM thickness beyond one-quarter of the soil's shear wavelength is not recommended. The presence of DSM or increasing its thickness had no considerable effect on reducing vertical accelerations. Although increasing the DSM width did not reduce horizontal accelerations, it proved an effective measure for controlling rocking motion. A width increase of more than 1.5 times the initial value is not recommended due to decreased efficiency and increased implementation costs. Given the significant influence of block-type DSM on the surrounding soil, evaluating DSM–soil and DSM–soil–DSM interactions is essential. Furthermore, applying bidirectional seismic loading led to a considerable increase in vertical accelerations.
In this study, the seismic performance of the block DSM method has been comprehensively investigated. The effects of various parameters under different loading conditions have been analysed. The results of this research can provide valuable insights to engineers for a better understanding of the behavior of this soil improvement technique.
This research was conducted based on real data from an ongoing power plant project and systematically investigated the key parameters influencing the seismic behavior of improved soils. The findings provide practical guidance for optimizing design and enhancing the understanding of the seismic performance of DSM systems.
