Professional Documents
Culture Documents
2007
Japanese Geotechnical Society
ABSTRACT
Stone columns, one of the most commonly used soil improvement techniques, have been utilized worldwide to in-
crease bearing capacity and reduce total and dierential settlements of structures constructed on soft clay. Stone
columns also act as vertical drains, thus speeding up the process of consolidation. However, the settlement of stabilised
bed is not reduced in many situations for want of adequate lateral restraint. Encasing the stone column with a geogrid
enhances the bearing capacity and reduces the settlement drastically without compromising its eect as a drain, unlike
a pile. The behavior of the encased stone column stabilized bed is experimentally investigated and analysed numerical-
ly. In the numerical analysis, material behaviour is simulated using Soft Soil, Mohr Coulomb and Geogrid models for
clay, stone material and encasement respectively and is validated with experimental results. The parametric study car-
ried out on varying the L W D ratio (Llength of the column; Ddiameter of the column) of column, stiness of ge-
ogrid and angle of internal friction of stone material gives a better understanding of the physical performance of the
encased stone column stabilized clay bed.
Key words: coupled FE analysis, geogrid encasement, settlement reduction ratio, soft clay, stone column (IGC:
E2/E13/E14/K3/K14/M9)
873
874 MALARVIZHI AND ILAMPARUTHI
and Nabil (1996) carried out triaxial tests on two types of Table 1. Properties of clay selected for this study
granular material with and without the geogrid sleeves. Specic gravity 2.68
They reported that the granular material packed within Liquid limit 55z
the cylindrical sleeve, increased the stiness of the system Plastic limit 18z
considerably. Adayat and Hanna (2005) used encapsulat- Plasticity index 37z
ed stone to improve collapsible soils. A few studies have Clay 65z
Silt 27z
reported the application of geotextile encased stone Unied Soil Classication CH
columns (Short et al., 2004; Alexiew et al., 2005; Raithel Compression Index 0.75
et al., 2005), yet the proper understanding of the behav-
iour of the columns is limited.
The scope of this study is to understand the behavior of Table 2. Properties of the stone chips
encased stone columns in soft clay and to bring out the
parameters which play dominant role in load sharing and Conning pressure, s3 Initial tangent
(kN Wm2 ) modulus, Ei (kN W m2 )
settlement reduction in the encased stone column stabi-
lized bed. With this as objective, laboratory tests were 50 2857
carried out on scaled down models approximately 1 W 20 100 4200
200 13300
size of prototype. The encased stone column stabilized Angle of shearing resistance 469
models were also analyzed using PLAXIS FEM code. An
attempt was made to include non-linear behavior of soil
and column material. The results of numerical analyses Table 3. Specications of the nets used
were compared with the experimental results of this study
as well as with the earlier theories and experiments pub- Net Wt.
Aperture size Es (kN W
m)
lished. An extensive parametric study was carried out to Id. m2
gm W
understand the inuence of L W D ratio of column, the net1 260 Diamond 1 mm 1 mm 15
stiness of geogrid and the angle of internal friction of net2 475 Square 4 mm 40
stone material on settlement reduction of the stabilized net3 730 Diamond 8 mm 6 mm 60
clay bed.
moisture content. Loss of water, if any due to evapora- Meyerhof and Sastry (1978) established that the failure
tion, was compensated before forming the bed. Care was zone extends over a radial distance of about 1.5 times the
taken to avoid entrapment of air while preparing the bed. diameter from the periphery of column. Load tests were
However, each layer of clay was tapped gently using a conducted on clay beds stabilized with stone column and
wooden plank to remove entrapped air, if present. encased stone column independently by loading the plate
in 8 to 15 equal load increments. The cumulative equal in-
Formation of Stone Column crements of load thus applied were maintained constant
The centre of the cylindrical tank was properly for a period of one hour. During this one-hour period,
marked, a casing of required diameter was placed, and settlements were recorded at an interval of 10 minutes us-
then clay bed was formed in layers around the casing. The ing two sensitive dial gauges resting at diametrically op-
stones were carefully charged into the casing and com- posite ends of the plate.
pacted using 12 mm diameter rod to achieve a density of
16 kN W m3. At the end of compaction of each layer, the
casing tube is withdrawn to a certain level and stone is MODELS USED FOR NUMERICAL STUDY
charged and compacted. This procedure is repeated until Numerical analyses of the model tests were accom-
the full length of column is formed. plished by the PLAXIS nite element software.
Elastoplastic behaviour of stone-column is modeled by
Formation of Encased Stone Column Mohr-Coulomb yield criterion employing a non-associ-
For the encased stone column, geogrid W net was stitch- ated ow rule. The non-associated ow rule in this case
ed into a tube of required diameter as explained earlier has a signicant meaning in the sense that the dilation of
and inserted along with the casing. The stones were com- the stone-column on shearing can be represented by ad-
pacted in the same way as it was done in the case of the justing the dilatancy angle.
stone column. The non-linear behaviour of clay is represented by the
The bed thus prepared was loaded with a seating pres- modied critical state model, the Soft soil model. The
sure of 2.5 kN W m2 to the entire area of the bed for 24 Soft soil model is extensively described in Vermeer and
hours to obtain uniform bed, which also ensured proper Brinkgreve (1998), which is similar to the modied Cam
contact between clay and encased stone column. The test clay model, without softening behaviour. The Soft Soil
carried out after 24 hours of preparation of the bed also model is meant for primary compression of near nor-
ensured gain in the strength of disturbed clay. mally-consolidated clay-type soils. Some features of the
Soft-Soil model are:
Experimental Setup for the Load Test Stress dependent stiness (logarithmic compression be-
Tests were conducted on a single stone column of di- haviour).
ameter 30 mm for various L W D ratios (Llength of the Distinction between primary loading and unloading-
column; Ddiameter of the column) on a standard load- reloading.
ing frame as a stress-controlled test. The loading arrange- Memory for pre-consolidation stress.
ment is shown in Fig. 1. The size of the tank was 300 mm Failure behaviour according to the Mohr-Coulomb
in diameter and 300 mm in height, (H). The equivalent di- criterion.
ameter of the tributary area for a spacing of 2D in a Soft Soil model requires the following material constants:
square pattern is 2.3D (1.13 times spacing). So the modied compression index, l*, modied swelling index,
stabilized clay bed is loaded through a steel rigid circular k*, cohesion, c?, friction angle, q, dilatancy angle, c,
plate of diameter 2.3D. The lateral dimension of the tank poison's ratio and k0, coecient of lateral stress in nor-
was chosen to be such that the minimum free distance be- mal consolidation. The dierence of this model from the
tween the periphery of the column and the side of the Modied Cam clay model is that it is a function of volu-
tank does not interface with the failure wedge. metric strain instead of void ratio. The parameter l* is
the modied compression index, which determines the
compressibility of the material in primary loading given
by Eq. (1). It is assumed that there is a logarithmic rela-
tion between the volumetric strain, en, and the mean eec-
tive stress, p?, which can be formulated as:
ene0nl* ln pp
?
0 (1)
l* and k* are obtained from the one-dimensional com- both the experimental and the numerical results of Lee
pression test. A relationship exists with the international- and Pande (Fig. 2). Therefore, these two models were
ly recognized parameters for one-dimensional compres- adopted for further analysis.
sion, Cc given by Eq. (2). The properties of the clay and stone are presented in
Cc Table 4. These properties of stone material and clay were
l * (2) determined through triaxial tests conducted on specimens
2.3(1e) in the laboratory as per ASTM D4767. When the stone
The factor 2.3 is obtained from the ratio between the material alone was tested at low conning pressures and
logarithm of base 10 and the natural logarithm. The ratio the volume change was measured through drained triaxial
l* Wk* (l W k) ranges, in general, between 3 and 7. The tests on stone specimens of 75 mm diameter, the dilatan-
elastic behaviour is described by Hooke's law and it re- cy was observed to be 269. However, the dilatancy of the
lates linear stress dependency on the tangent bulk modu- stone material was found to be only 49when it was pack-
lus as in Eq. (3). ed within the geogrid encasing. The initial tangent modu-
lus obtained from triaxial test on the encased stone
Eur p?
Kur (3) column was used for the stone material within the encase-
3(1 2nur) k*
ment to incorporate the initial connement eect oered
The sux `ur' stands for unloading reloading. Neither by the geogrid encasement.
the elastic bulk modulus, Kur, nor the elastic Young's The geogrid was modeled as linear elastic continuum
modulus, Eur, is used as an input parameter. Instead, nur element whose axial stiness was taken as the initial tan-
and k* are used as input constants for the part of the gent modulus, obtained from the tension test. The initial
model that computes the elastic strains. The poisson's tensile modulus, (EA) of the geogrids (ratio of the axial
ratio used is the well known pure elastic constant rather
than a pseudo-elasticity constant as used in the Mohr
Table 4. Parameters used for material modeling
Coulomb model. It is also possible to specify undrained
behaviour in an eective stress analysis using eective Stone Encased Stone
Parameter Clay
model parameters. The presence of pore pressures in a column column
soil body, usually caused by water, contributes to the Mohr Mohr
total stress level. According to Terzaghi's principle, total Model Soft soil
Coulomb Coulomb
stresses s can be divided into eective stresses s? and pore E [kPa] Not reqd. 2500 9000
water pressures sw, given by sxxs?xxsw; syys?yysw. n 0.2 0.3 0.3
g [kN Wm 3] 11 16 16
However, water is supposed not to sustain any shear
F?[9] 24 48 48
stress, and therefore the eective shear stresses are equal c? [kPa] 6 0.1 0.1
to the total shear stresses (sxys?xy). Cohesion and angle c [9] 0 26 4
of internal friction are the eective stress parameters k (m Wday) 5.94e4 1 1
obtained from the drained triaxial test. The permeability, l* 0.136 Not reqd. Not reqd.
k* 0.054 '' ''
k is calculated from the cv obtained from the consolida-
eo 1.42 '' ''
tion test.
Application of these two material models were veried
with the published results of Lee and Pande (1998)
wherein Cam clay and Mohr Coulomb models were used
to analyze the stone column stabilized clay bed. The
result of numerical study compares reasonably well with
Fig. 2. Validation of FEM with Lee and Pande (1998) Fig. 3. Typical FE model of clay bed and column
ENCASED AND CONVENTIONAL STONE COLUMN 877
Column type m2 )
Expt. (kN W m2 )
FEM (kN W
Clay 18 16
D5)
sc (L W 22 23
D7.5)
sc (L W 27 28
D10)
sc (L W 40 38
Fig. 7(b). Load vs. Settlement response of oating and end bearing
columns with encasement (net3)
Column D 5
LW D7.5
LW D10
LW
sc 21 27 40
net1 22.5 29 45
net2 30 32 63
net3 35.5 39 81
m2 at
Fig. 11. Load settlement behavior: incremented load of 10 kN W
an interval of 10 days
Fig. 9. Variation of hoop force in encasements
Stone Encased
Parameter Clay
column stone column
Mohr- Mohr-
Model Soft-
Coulomb Coulomb
Soil
E [kPa] Not reqd. 24000 48000
n 0.2 0.35 0.3
g[kN Wm 3] 12 16 16
f?[9] 30 3248 3248
c? [kPa] 6 0.3 0.3
c [9] 0 2W 3f 1W 3f
k (m Wday) 5.94e4 1 1
l* 0.136 Not reqd. Not reqd.
k* 0.054 '' ''
eo 1.42 '' ''
Fig. 10. Stress concentration ratio in column
ENCASED AND CONVENTIONAL STONE COLUMN 881
Fig. 19. Stiness of geogrid on SRR for dierent pressures Fig. 21. Horizontal displacement vs. Stiness of geogrid for dierent
q? values
3. The load-settlement curves obtained from nite ele- Foundation Engineering, 114.
ment analyses compare well with experimental 4) ASTM: D476704Consolidated Undrained Triaxial Compression
Test for Cohesive Soils.
curves suggesting that material models viz. Soft
5) ASTM: D663701Determining Tensile Properties of Geogrids by
soil, Mohr Coulomb and geogrid models used in the the Single or Multi-Rib Tensile Method.
FEM analyses are eective in idealising the behav- 6) Balaam, N. P. and Poulos, H. G. (1983): The Behavior of founda-
iour of soft clay, stone column and encasement re- tions supported by clay stabilized by stone columns, Proc. Specialty
spectively. Sessions, VII European Conference on SMFE, 2.
7) Barron, R. A. (1948): Consolidation of ne-grained soils by drain
4. Numerical studies conrmed the bulging mechan-
wells, Transactions of the American Society of Civil Engineers,
ism of stone column. The bulging of stone column 113, paper 2346, 718724.
is eective up to the depth of 4 times the diameter of 8) Bauer, G. E. and Nabil, A. J. (1996): Laboratory and analytical in-
the column, which is in conformity with classical vestigation of sleeve reinforced stone columns, Geosynthetics: Ap-
theories of Hughes and Withers (1974) and Green- plication, Design and Construction, 463466.
9) Biot, M. A. (1941): General theory of three-dimensional consolida-
wood (1970) and the present experimental study.
tion, J. Appl. Phys., 12, 1941a, 155165.
5. The mobilised hoop force in the geogrid material in- 10) Bouassida, M. and Hadhri, T. (1995): Extreme load of soils rein-
creases with increase in surcharge pressure. Initial- forced by columns: The case of an isolated column, Soils and Foun-
ly, the hoop force is mobilised over the top 1D dations, 35(1), 2135.
depth, and as the pressure increases, the hoop force 11) Burland, J. B. (1965): The yielding and dilation of clay, G eotech-
nique, 15, 211214.
is mobilised over the length of column apart from
12) Datye, K. R. and Nagaraju, S. S. (1981): Design approach and eld
increase in its magnitude. Further, the hoop stress is control for stone columns, Proc. 10th International Conference on
always maximum at 1D depth of the column for the Soil Mechanics and Foundation Engineering, 637640.
parameters analysed in this study. 13) Deshpande, P. M. and Vyas, A. V. (1996): Interactive encased
6. Encasing the stone column increases the stress con- stone column foundation, 6th International Conference and Exhi-
bition on Piling and Deep Foundation, DFI'96, ISSMFE, 119.
centration on the column, thereby reducing the load
14) Greenwood, D. A. (1970): Mechanical improvement of soils below
on clay, consequently reducing the settlement. The ground surface, Conference on Ground Engineering, Institution of
parametric study shows that the settlement reduc- Civil Engineering, 1122.
tion ratio in the encased stone column bed is about 15) Han, J. and Ye, S. L. (2001): Simplied method for consolidation
50z of stone column bed for identical conditions. rate of stone column reinforced foundations, J. Geotechnical and
Geoenvironmental Engg, 127(7), 597603.
7. The stress concentration factor increased with the
16) Hughes, J. M. O. and Withers, N. J. (1974): Reinforcing of soft co-
stiness of encasement and is always higher than in hesive soils with stone columns, Ground Engineering, 7(3), 4242
conventional stone column irrespective of applied and 4749.
pressure. 17) Katti, R. K., Katti, A. R. and Naik, S. (1993): Monograph to Anal-
8. The parametric study shows that the increase in ysis of Stone Columns with and without Geo-synthetic Encasing,
CBRI publication, New Delhi.
stiness of encasement reduces the settlement, but
18) Lee, J. S. and Pande, G. N. (1998): Analysis of stone column rein-
when the stiness is increased beyond 2000 kN W m2 W forced foundations, Int. J. Numerical and Analytical Meth in Ge-
m, the contribution to settlement reduction ratio omech., 22, 10011020.
becomes insignicant for the conditions analysed in 19) Madhav, M. R., Miura, N. and Alamgir, M. (1994): Analysis of
this study. granular column reinforced ground, 5th International Conference
on Geotextile, Geomembranes and Related Products.
9. As the L W D ratio of column increases, settlement
20) Meyerhof, G. G. and Sastry, V. V. R. N. (1978): Bearing capacity
reduces and if the L W D ratio is more than about 10, of piles in layered soils, part I and II, Canadian Geotechnical Jour-
it does not contribute much to settlement reduction. nal, 15, 171189.
10. The angle of shearing resistance of column material 21) Mitchell, J. K. and Huber, T. R. (1985): Performance of a stone
also aects the SRR of encased stone column but column foundation, Journal of Geotechnical Engineering, 111,
205223.
not to the extent of stiness of encasing material.
22) Narasimha, Rao. S. N. (2000) Studies on groups of stone column in
Eciency of encased stone column is higher if the soft clays, Symposium on Ground Improvement Techniques For
column material is compacted well to achieve high Practicing Engineers, India, 8493.
angle of shearing resistance. 23) Nayak, N. V. (1983): Recent advances in ground improvements by
stone volumn, Proc. Indian Geotechnical Conference, 1, V19.
24) Poorooshasb, H. B. and Meyerhof, G. G. (1997): Consolidation
REFERENCES settlement of rafts supported by stone columns, Journal of Ge-
otechnical Engineering, 8392.
1) Aboshi, H., Ichimoto, E., Enoki, M. and Haraka, K. (1979): The 25) Priebe, H. J. (1976): An evaluation of settlement reduction in soil
CompozerA method to improve characteristics of soft clays by improved by vibroreplacement (en alem an), Bautechnik, (53),
inclusion of large diameter sand columns, Proc. International Con- 160162.
ference on Soil Reinforcement: Reinforced Earth and other Tech- 26) Raithel, M., Kirchner, A., Schade, C. and Leusink, E. (2005):
niques, 1, 211216. Foundation of constructions on very soft soils with geotextile en-
2) Adayat, T. and Hanna, A. M. (2005): Encapsulated stone columns cased columnsState of the Art, Geofrontiers 2005, GSP 131 In-
as a soil improvement technique for collapsible soil, Ground Im- novations in grouting and soil improvement, 111.
provement, 9(4), 137147. 27) Rao, B. and Ranjan, G. (1988): Closure of the paper Settlement
3) Alexiew, D., Brokemper, D. and Steve, L. (2005): Geotextile En- analysis of skirted granular piles, Journal of Geotechnical En-
cased Columns (GEC): Load capacity, geotextile selection and pre- gineering Division, ASCE, 114, 729736.
design graphs, Geofrontiers 2005, GSP 131 Contemporary Issues in 28) Schlosser, F., Jacobsen, H. U. and Juran, I. (1983): General report
ENCASED AND CONVENTIONAL STONE COLUMN 885
on soil reinforcement, Proc. 8th European Conference on SMFE, cal Journal, 41, 299312.
3, 11591180. 31) Van Impe, W. Y. and De Beer, E. (1983): Improvement of settle-
29) Short, R. D., Prashar, Y. and Metcalf, B. (2005): Repairing a rail- ment behaviour of softy layers by means of stone columns, Proc.
way spur roadbed failure using geotextile encased columns, TRB 8th International Conference on SMFE, Helsinki, 309312.
2005 Annual Meeting Report. 32) Vermeer, P. A. and Brinkgreve, R. B. J. (1998): Plaxis Finite Ele-
30) Sivakumar, V., Mckelvey, D., Graham, J. and Hughes, D. (2004): ment Code for Soil and Rock Analyses, Rotterdam, Balkema.
Triaxial tests on model sand columns in clay, Canadian Geotechni-