Seismic Vulnerability of the Highway System

Task E2-1: Ground Remediation for Silty Soils Using Stone Columns

Subject Area: Geotechnical and Foundation Engineering 
Research Year 2

Principal Investigator(s) and Institution(s)

Sabanayagam Thevanayagam, University at Buffalo


This task will develop an improved remediation technique and design method to mitigate liquefaction hazards in silty soils using stone columns. It involves laboratory studies on the behavior of silty soils, numerical studies on pore pressure generation, dissipation, and densification of silty soils during stone-columns installation, and performance evaluation using scale-models. The end product from this research will be a report summarizing the improved design method with typical design charts and guidelines for installation of stone columns.


Many of the past earthquake-induced ground failures and large deformations observed in the built-environment have occurred mostly in sites containing non-plastic silty soils. The need to develop new methods to assess liquefaction potential of these soils and to cost-effectively remediate such sites to mitigate earthquake induced damages has been widely recognized. Experience indicates that silty soils behave differently than clean sands. Silty soils have low permeable and are relatively more compressible. Their liquefaction (power pressure) behavior is also different. There is consensus among engineers and contractors that the traditional remediation techniques (densification, reinforcement, or solidification (grouting, admixtures) used for sandy soils need to be modified to suite the need for silty soils. The approach to the solution is two-fold, as described below.

The first issue that needs to be addressed and understood is the matrix behavior of silty soils and what (constituents) controls the behavior of silty soils. Recent studies by the investigators (Thevanayagam 1998a-b) have shed some important light on this issue. It is now recognized that a silty soil can be judiciously characterized in terms of the behavior of the host silts and/or host sands. At fines contents less than typically 20 to 30%, the intergranular void ratio controls the behavior. Beyond that the silt primarily controls the behavior with minor reinforcement effect by the sand grains. Further work is still needed to quantify the matrix effects. However, with this framework, it is now possible to analytically study and develop appropriate techniques for remediation of silty soils.

Among suitable techniques vibratory stone columns has appeal. It combines densification, reinforcement, and provision for drainage of excess pore water pressures. But how to account for these effects in silty soils is the question. The central issue is the effects of low permeability and compressibility characteristics of the silt matrix on pore pressure generation, dissipation, and densification of the soil during stone-column installation, and dissipation during earthquakes. They need to be understood and incorporated in the analysis in order to appropriately design and install stone-columns, with confidence that they will mitigate liquefaction damage in silty soils.

As a result, this task will be comprised of the following subtasks:

Subtask 1: A laboratory study of silts under undrained and drained loading conditions will be performed aimed at characterizing liquefaction (pore pressure) behavior, permeability, and compressibility characteristics of these soils at conditions relevant during installation and earthquakes. This will increase our understanding of liquefaction behavior and post-liquefaction pore pressure dissipation phenomenon in natural silty soils. To this effect natural silty soils recovered from Oregon, the San Fernando Dam site, Memphis area (mid American site), and other (eastern) silty sites will be used. They will be subjected to undrained monotonic and cyclic loading followed by drained dissipation conditions. Their liquefaction, pore pressure generation, permeability, and compressibility characteristics will be determined at density (silt void ratio) levels relevant for detrimental field conditions. Effects of sand content will also be studied. These data will provide input variables for numerical studies in subtask 2.

Subtask 2: Numerical and semi-theoretical studies of pore pressure generation, dissipation, and densification of silty soils during stone-column installation and earthquakes. This subtask will quantify the effects of stone-column design variables (installation procedure, diameter, stone drainage characteristics, etc.) and silt characteristics on the level and extent of pore pressure generation and dissipation and densification achievable in the surrounding silty soils during installation, and assess the effectiveness of dissipation during earthquakes. The use of wick drains to enhance dissipation of the installation pore pressures will also be studied. Semi-theoretical numerical analyses will be carried out addressing the effects of the above design variables and material characteristics on the effectiveness of stone columns.

Subtask 3: During the course of this research, new field opportunities have developed. It is now possible to evaluate the actual performance of stone column, along with wick drains, installed in silty soil sites in the field against the results from subtasks 1 and 2. The US Bureau of Reclamation and Hayward Baker, Inc. (Contractor) are planning to install stone columns along with wick drains at a silty soil site in Washington. Additional field data also have become available. This subtask will utilize these field pore pressure generation/dissipation and densification data around the stone columns for comparison with the analytical/numerical predictions suing the numerical model developed in subtask 2. This will be used to refine and calibrate the model where necessary. Field soils from the above specific sites will also be collected for testing to obtain soil parameters required in silty soils.

Subtask 4: Guidelines for design and installation of stone columns will be developed. The results from the above three tasks will provide the necessary insight and information needed to develop a design method and guidelines for installation of stone columns in silty soils.

During Research Year 1, work on the laboratory characterization of silty soils neared completion, and numerical simulation of pore pressure generation and dissipation during stone column installation and seismic excitation was initiated.


  • Summary report on the liquefaction behavior, permeability, compressibility, and pore pressure generation/dissipation characteristics of silts
  • Final report presenting the improved design method with typical design charts and guidelines for installation of stone columns. The relevant data and interpretations will be presented in an Appendix to the report.


return to Task Statement List