683
Evaluation of the efficiency of different ground improvement techniques
Évaluation de l'efficacité des différentes techniques d'amélioration des sols
Bryson S.
Stantec Consulting Ltd., Fredericton, NB, Canada
El Naggar H.
University of New Brunswick, Fredericton, NB, Canada
ABSTRACT: There are two primary reasons why developments continue in areas with poor subgrade soil conditions. The first is the
lack of space and increased pressure to develop within a particular region. The second is for economic reasons such as road
construction where it is not feasible, and may not be physically possible, to modify routes to avoid crossing areas of soft soils. This
paper investigates the performance of various ground improvement methods used in construction of embankments on soft soils using
three-dimensional numerical modeling.
Multiple forms of ground improvements were analyzed including deep soil mixing, light-weight fill, and stone columns. The
efficiency of each type of ground improvement was evaluated based on the reduction of predicted settlement compared to a baseline
model where improvements were not used. It is suggested that the economic feasibility of ground improvements be highly dependent
on the geographic location of the site; however, the gain in performance may be worth the extra material costs in some cases.
RÉSUMÉ : Il y a deux raisons principales pour lesquelles on continue
à développer
dans les zones
où les conditions du sol de
fondation sont mauvaises.
La première est le manque d'espace, ainsi qu’une pression accrue à se développer dans une région donnée.
La seconde étant, pour des raisons économiques telles que la construction de routes où il n'est pas probable, et peut-être pas
physiquement possible, de modifier les routes pour éviter de traverser les zones de sols mous. Cet article examine la performance des
différentes méthodes d'amélioration des sols utilisées dans la construction de remblais sur sols mous en utilisant la modélisation
tridimensionnelle numérique.
De multiples formes d'amélioration du sol ont été analysées, y compris un sol de mélange profond, légèrement rempli et en colonnes
de pierre. L'efficacité de chaque type d'amélioration des sols a été évaluée en fonction de la réduction de tassement prédit par rapport à
un modèle de référence où des améliorations n'ont pas été utilisées. Il est suggéré que la faisabilité économique des améliorations du
sol est fortement tributaire à la situation géographique du site, cependant dans certains cas le gain en performance vaut la peine
d’avoir des coûts supplémentaires de matériels.
KEYWORDS: Ground improvement, deep soil mixing, light-weight fill, stone columns, embankments, soft soils
1 INTRODUCTION
Lack of space, increased pressure to develop within a particular
region or any other economic or political motivations are all
valid reasons for developments to continue in areas with poor
subgrade soil conditions. There are several methods of
improving the properties of soft soils to reduce the post-
construction settlement or to improve the stability and the
overall performance of embankments and dams. Improvement
techniques used in construction of embankments and dams on
soft clay such as stone columns, deep soil mixing,
vibrocompaction, etc., are becoming increasingly popular in
North America. The San Pablo Dam, Sunset North Basin Dam,
the Clemson Upper and Lower Dams, and the I-95/Route1
project are all examples of case studies in the United States
where ground improvements have been implemented under
foundations of dams and embankments.
Methods of construction of embankments on soft soils have
been well documented; however, to the authors’ knowledge,
there is a lack of literature related to the comparison and process
of selecting ground improvements. The main purpose of this
study was to investigate the performance of various ground
improvement methods used in construction of embankments on
soft soils using three-dimensional numerical modeling to
identify which method is most efficient at reducing settlements
for the considered case.
2 NUMERICAL MODEL
The parametric analyses completed for this study were done
using an explicit three dimensional finite difference numerical
model. Materials modeled using the finite difference method are
represented by polyhedral elements to which variables are
assigned at discrete locations and the zones will behave
independently. The explicit time marching scheme calculates
the velocity and stresses for every element during each time step
based on the initial values. The new stresses and velocities are
then applied to the elements to be used in the following time
step. The model uses the velocity and time step to determine the
displacements for each element. The finite difference methods
allow grid points within the model to move and deform as
incremental displacements are applied during each time step
which makes it better suited to be used for analysis of nonlinear
large strain problems, such as the deformation of soft soil
during embankment construction (Itasca 2009).
2.1 Model details
For the purpose of the parametric study a baseline model was
developed using assumed stratigraphy, embankment
dimensions, and material properties. The assumed stratigraphy
was a 6 m thick layer of loose silt, overlying 2 m of normally
consolidated, soft clay, over sandstone bedrock. The design
embankment used in the model was 6 m high, with a crest width
of 6 m and 3H:1V side slopes (see Figure 1). It was assumed
that the ground improvements would extend as far as the toe of
the embankment side slopes to the bottom of the soft clay.
