2

Soil Densification Using Vibro-Stone Columns

Supplemented with Wick Drains

Thevachandran Shenthan
Ph.D. Candidate, Department of Civil, Structural and Environmental Engineering, University at Buffalo
Research Supervisor: S. Thevanayagam, Associate Professor and MCEER Director of Education

Summary
Saturated loose sand and non-plastic silty sand deposits are often vulnerable to liquefaction during earthquakes. Sand
deposits densified by vibro-stone column (SC) are more resistant to liquefaction, and have performed well during
earthquakes. Silty sand deposits appear to perform well when improved by SC supplemented with wick drains. Wick
drains aid dissipation of excess pore pressure induced during SC installation in low-permeable silty soils enhancing
densification. This paper presents a numerical model to simulate, and to analyze soil densification during SC
installation with and without wick drains. Design charts for SC are developed based on this work. These numerical
results are compared with field test data. Design guidelines are presented based on these design charts.

Introduction
SC installation process
involves a sequence of
processes starting with
insertion of a vibratory
probe with rotating
eccentric mass (FHWA
2001) (Figure 1) into the
ground. Once the design Stone
Columns
depth is reached, the Plan View Plan View

probe is withdrawn in 2b 2b
2b

2aWick 2a
lifts, backfilling the hole Drains
with gravel. During each
lift the probe is then
Elevation Elevation
reinserted expanding the
SC diameter. This a) Stone Columns b) SC in Sands c) SC in Silty Sands
process is repeated
several times until a Figure 1. Vibro-stone columns (with or without wick drains)
limiting condition,
measured in terms of amperage drawn by the probe, is achieved. SC have been mainly designed for
densification of relatively granular soils containing less than 15% non-plastic silt passing sieve #200
(74 m) and less than 2% of clayey particles (<2 m) (FHWA 2001). Soils containing excessive fines

Soil Densification Using Vibro-Stone Columns Supplemented with Wick Drain 9

 have been difficult to densify using SC. Current
45
practice for design of SC for liquefaction
40 mitigation relies on past experience, and field
P o s t - im p r o v e m e n t ( N 1 ) 6 0 c s

35 pilot tests. Based on pre- and post-

30 improvement SPT and CPT data from case
histories, Baez (1995) developed an empirical
25
relationship between pre- and post-
20 improvement equivalent clean sand SPT blow
15 Ar=20% counts, (N1)60cs, for a set of area replacement
10
Ar=15% ratios Ar of 5, 10, 15, and 20% (Figure 2) for
Ar=10%
fine to medium sand with fines less than 15%
5 Ar=5%
and little or no clay content. Ar is defined as the
0 ratio of stone column area to the tributary area
0 2 4 6 8 10 12 14 16 18 20 per stone column.
Pre-improvement (N 1 ) 60cs
Recent case histories show that vibro-stone
Figure 2. Pre- and post-improvement (N1)60cs
column technique may be effectively used to
densify silty sands containing fines content
exceeding 15% by using pre-installed supplementary wick drains. The supplementary drains help to
relieve excess pore pressures developed during stone column installation processes (Andrews 1998,
Luehring et al. 2001) and help soil densification.

For both cases, sands and silty soils, at present, there are no detailed analytical procedures available
to determine the densification achievable or to analyze the effects of various field soil and SC
construction parameters on the degree of improvement possible to achieve for a wide range of soils
with non-plastic silt contents.

Recent work conducted by Thevanayagam and co-workers (2002, 2003) and Shenthan et al. (2004 a
and b) focused on development of a simple analytical methodology to simulate soil response during
SC installation, quantify soil densification in saturated sands and silty soils, and assess the effect of
various construction/design choices and soil parameters on the degree of improvement achievable.
Analytical results were compared with field test data. Based on this work, guidelines were developed
for design of stone columns to mitigate liquefaction in silty soils. This paper presents a brief
summary of these findings.

Simulation of Vibro-Stone Column

Densification of saturated loose sands and silty soils during SC installation is essentially a process
involving controlled liquefaction, and consolidation of the soil leading to concurrent densification.
Shenthan et al. (2004a and b) developed an analytical technique to simulate pore pressure
developments in the soil due to vibratory energy imparted during installation and subsequent
consolidation of the soil and densification based on the simplified methodology highlighted briefly
in following sub-sections. Details are reported in Shenthan et al. (2004 a and b) and Shenthan
(2005).

10 Highway Project
Dissipated Energy
Considering the energy source in the vibratory probe as a point-source and assuming that the energy
propagates spherically outward, the energy loss per unit time per unit volume of soil at distance r can
be approximated by,

w = W0 Exp 2 ( r r0 ) .Exp ( ru )av
2 r 2 (1)

where r0 is the radius of the probe, W0 is the energy imparted by the probe into the soil per unit time
(= 0P0), P0 is the power rating of the vibratory probe, 0 is the probe efficiency, (ru)av is the average
excess pore pressure ratio within the soil surrounding the probe up to an effective radial distance re,
which is assumed to be the same as the center to center spacing between stone columns in this
study, is the coefficient of attenuation (Richart et al. 1970, Dowding 1996), and = a constant.

In the soil around the vibratory probe, as excess pore pressure develops due to vibration, the soil
becomes weak. Since the amplitude of vibration of the probe is limited (FHWA 2001), the energy
imparted to the surrounding soil would decrease, resulting in a reduced efficiency. When the pore
pressures dissipate, and the soil is sufficiently densified, the energy transfer rate would increase. In
this study, this phenomenon has been taken into account, considering the energy transfer rate to
decay with increasing excess pore pressure, represented by the last exponential term in the above
equation.

Pore pressure generation and dissipation

Based on experimental data and theoretical considerations, excess pore water pressure generated due
to cyclic loading has been related to frictional energy loss in the soil by ru = 0.5Log10[100(Ec/EL)]
(Thevanayagam et al. 2003), where ru = pore pressure ratio u/0, 0= initial effective confining
pressure, Ec = cumulative energy loss per unit volume of soil, EL = energy per unit volume required
to cause liquefaction, and (Ec/EL)>0.05.

The governing equation for excess pore pressure dissipation in soil around the vibratory probe is:

u k 2 u 1 u 1 2 u k v 2 u u g
= h 2 + + 2 + +
t w mv r r r r 2 w mv z 2 t
(2)

where kh and kv are the hydraulic conductivities of the soil in the horizontal and vertical directions,
respectively; mv is the volume compressibility of the soil; u is the excess pore water pressure at
coordinates (r, , z); t is time; w is the unit weight of water; r, z, and are radial, vertical, and
angular coordinates, respectively; ug is the excess pore pressure generated due to vibration of the soil
during SC installation (or due to seismic excitation during an earthquake following installation of
stone columns).

Design Charts
Numerical simulations were conducted for silty soils with three different pre-improvement (N1)60cs
(normalized clean sand equivalent SPT blow counts) of 7, 11 and 16. For each (N1)60cs, three different
area replacement ratios (Ar = 5.6, 10, and 22.5%) were considered. For each case, simulations were
done for stone column without wick drains, and with wick drains, respectively. The effect of fines

Soil Densification Using Vibro-Stone Columns Supplemented with Wick Drain 11

content is reflected by varying the hydraulic conductivity (k) from 10-4 to 10-8 m/s. It was assumed
that the vibro-stone columns are installed at a triangular pattern with wick drains pre-installed at
midpoints between stone column locations. The power rating of the vibratory probe was assumed to
be 120 kW.

Figure 3 shows the simulation results for pre- and post- improvement (N1)60cs for soils at a depth
with effective vertical stress of 100 kPa and a range of k values. Results show that, at low area
replacement ratio, the effect of supplementary wick drains is negligibly small and soil densification is
primarily affected by stone columns. No significant additional densification is achieved by stone
columns with wick drains compared to stone columns without wick drains in soils with hydraulic
conductivities less than about 10-6 m/s. At high area replacement ratio of about 22.5%, wick drains
significantly contribute to the drainage of excess pore pressures induced during stone column
installation and soil densification. The combined system is effective for soils containing non-plastic
silt and hydraulic conductivity as low as 10-8 m/s. However, the degree of improvement decreases
with increasing silt content and decreasing hydraulic conductivity.

40 40 40
Pre - (N 1 ) 60 cs = 7 Pre - (N 1 ) 60 cs = 11 Pre - (N 1 ) 60 cs = 16
Post (N1)60 cs

Post (N1)60 cs
A r = 5.6% A r = 5.6%
Post (N1)60 cs

30 30 A r = 5.6% 30

20 S C + W icks 20 20
S C Alone
10 10 SC + W icks 10 SC + W icks
SC A lone SC A lone
0 0 0
1E-8 1E-7 1E-6 1E-5 1E-4 1E-8 1E-7 1E-6 1E-5 1E-4 1E-8 1E-7 1E-6 1E-5 1E-4
(a) k (m /s) k (m /s) k (m /s)

40 40 40
Pre - (N 1 ) 60 cs = 7
Post (N1)60 cs

Post (N1)60 cs
Post (N1)60 cs

30 A r = 10% 30 30
SC + W icks SC + W icks
20 20 SC A lone 20 SC A lone

10 S C + W icks 10 Pre - (N 1 ) 60 cs = 11 10 Pre - (N 1 ) 60 cs = 16

S C Alone A r = 10% A r = 10%
0 0 0
1E-8 1E-7 1E-6 1E-5 1E-4 1E-8 1E-7 1E-6 1E-5 1E-4 1E-8 1E-7 1E-6 1E-5 1E-4
(b) k (m /s) k (m /s) k (m /s)

40 40 40
Post (N1)60 cs

Post (N1)60 cs
Post (N1)60 cs

30 30 30
S C + W icks SC + W icks SC + W icks
S C Alone SC A lone
20 20 20 SC A lone

10 Pre - (N 1 ) 60 cs = 7 10 Pre - (N 1 ) 60 cs = 11 10 Pre - (N 1 ) 60 cs = 16

A r = 22.5% A r = 22.5% A r = 22.5%
0 0 0
1E -8 1E -7 1E -6 1E -5 1E -4 1E-8 1E-7 1E-6 1E-5 1E-4 1E-8 1E-7 1E-6 1E-5 1E-4
(c) k (m /s) k (m /s) k (m /s)

SC + Wicks = vibro-stone column with wicks; SC = vibro-stone column without wicks

Figure 3. Vibro-stone columns design charts

Field Test
Field tests were conducted, in collaboration with Hayward Baker, Inc. and Advanced Geosolutions,
Inc. at a site in Marina Del Rey, CA, to assess the applicability of the above design charts for field
conditions (Shenthan 2005). Figs.4a-b show a typical soil profile and pre- and post-improvement
CPT profiles at the test location at the site. Liquefaction risk analysis suggested inter-bedded silty

12 Highway Project
soil layers from 2.7 m up to about 6.0 m depth were liquefiable for a design earthquake of M = 7.5
and amax = 0.35 g. The silt layer to be improved at the test location has a silt content ranging from 20
to 40%. The hydraulic conductivity was estimated to be in the range of 10-6 to 10-7 m/s. Ground
improvement recommendations called for SC with a diameter of 0.9 m at a spacing of 2.4 m (Ar
=11%) supplemented with pre-installed wick drains (100 mm x 5 mm) at mid point between stone
columns. Hayward Baker Inc. installed the stone columns. Pore pressures, ground vibrations, and
energy delivered to the soil during stone column installation within a selected test section were
monitored using seismic piezocones (Fugro Geosciences, Inc., CA), retrievable seismic
accelerometers (www.nees.ucla.edu), and a current sensor. Three pre-improvement CPT tests and
two post-improvement CPT tests were completed at the test location. A detailed analysis of these
data is presented elsewhere (Shenthan 2005). Approximate average water table at the time of CPT
tests and stone column installation is shown on the soil profile using dashed-line. CPT test results
were interpreted following the procedures outlined by Youd et al. (2001). Normalized CPT
resistance results were converted to (N1)60cs using correlations recommended by Robertson and
Wride (1998).

Average pre-improvement (N1)60cs for the sand to silty sand to sandy silt layer (from 3.0 to 4.3m
depth, Figure 4b) was estimated to be about 11, and the average post-improvement (N1)60cs for this
layer was estimated to be about 22. This data point is superimposed on the design chart
corresponding to Ar=10%, which is the closest to the actual field Ar value. The field results are in
good agreement with simulation results. Additional field data are needed to further validate the range
of applicability of the simulation results.

(a ) N o rm . T ip R e s is ta n c e , (b ) A v e ra g e
qc1N S o il P r o file
0 50 100 150 200
0 0 .0

S a n d to S ilty S a n d
S ilty S a n d to
5 1 .5 S a n d y S ilt
40
C la ye y S ilt to C la y
S ilty S a n d to Field Test
S a n d y S ilt
Pre-(N1)60cs = 11
30
Post (N1 ) 60 cs

C la y e y S ilt to C la y
10 3 .0 Ar =11%
Depth (m)
Depth (ft)

S a n d to S ilty S a n d to
S a n d y S ilt SC + Wicks
20 SC Alone
15 4 .6

P re C P T - A v e 10 Cha rt for
C la ye y S ilt to C la y Pre - (N1 )60 cs = 11
P ost C P T - A ve
20 6 .1 A r = 10%
0
1E-8 1E-7 1E-6 1E-5 1E-4
25 7 .6 k (m /s)

(a) pre- and post-improvement CPT profile (b) Soil profile (c) Comparison with design chart

Figure 4. Field test data

Soil Densification Using Vibro-Stone Columns Supplemented with Wick Drain 13

 C h o o se ch arts set fo r d ep o sit p re-(N 1 ) 6 0 cs
S tart w ith S C alo n e
Based on the above results, a flow chart shown
in Figure 5 was developed as a guideline for C o n sid er trial
A r o f 5 .6 %
liquefaction mitigation design in silty soils using
stone columns supplemented with pre-installed E n ter ch art w ith k (m /s)
wick drains. F in d p o st-(N 1 ) 6 0 cs

Concluding Remarks Y es
(N 1 ) 6 0 cs d es (N 1 ) 6 0 cs req

Design guidelines for liquefaction mitigation of No

non-plastic silty soils using vibro-stone columns In crease A r
combined with wick drains have been
presented. Field data was used to assess the Y es
A r is p ractical?
applicability of these design guidelines. Results
indicate that degree of improvement achievable No
by SC at a site depends on hydraulic U se S C w ith w ick d rain s
conductivity and silt content. A significant
increase in post-improvement (N1)60cs can be C o n sid er an o th er tech n iq u e

achieved by vibro-stone columns combined

P rin t fin al d esig n p aram eters
with wick drains for soils with hydraulic W icks, A r , S C d iam eter, sp acin g
conductivity as low as 10-7 to 10-8 m/s,
provided that the area replacement ratio Figure 5. SC design procedure
exceeds about 20%. The recommended
guidelines presented herein are expected to Note: (N1)60cs,req = minimum (N1)60cs profile required to mitigate
liquefaction at the site, obtained using a liquefaction potential
advance the use of vibro-stone columns analysis; pre-(N1)60cs = average pre-improvement (N1)60cs for the
combined with wick drains to mitigate liquefiable layer at the design site; post-(N1)60cs = (N1)60cs,des.
liquefaction potential in non-plastic silty soils, average post-improvement (N1)60cs for the liquefiable layer at the
design site obtained from the chart (Figure 3).
and reduce the reliance on expensive field trials
as a design tool.

Acknowledgements
This research was carried out under the supervision of Dr. S. Thevanayagam, and primarily
supported by FHWA, under award number DTFH61-98-C-00094 to the Multidisciplinary Center
for Earthquake Engineering Research.

References
Andrews DCA (1998): Liquefaction of silty soils: susceptibility, deformation, and remediation. Ph.D.
Dissertation, Dept. of Civil Eng., USC, CA.
Baez JI (1995): A design model for the reduction of soil liquefaction by vibro-stone columns. Ph.D.
Dissertation, USC, Los Angeles, CA.
Dowding CH (1996): Construction vibrations, Prentice Hall, NJ.
FWHA (2001): Stone columns. Ground Improvement Technical Summaries, II, Publication No. FHWA-
SA-98-086R, (7-) 1-84.

14 Highway Project
Luehring R, N Snorteland, L Mejia, and M. Stevens (2001): Liquefaction mitigation of a silty dam
foundation using vibro-stone columns and drainage wicks: a case history at salmon lake dam. Proc.
21st USSD annual meeting and lecture, Denver, CO.
Richart FE Jr, Hall JR, Woods RD (1970): Vibrations of soils and foundations, Prentice-Hall, Inc.,
Englewood Cliffs, NJ.
Robertson PK, Wride CE (1998): Evaluating cyclic liquefaction potential using the cone penetration
test. Canadian Geotechnical Journal, 35: 442-459.
Shenthan T, Nashed R, Thevanayagam S, Martin,G R (2004a): Liquefn. mitigation in silty soils using
composite stone columns and dynamic compaction. J. Earthq. Eng. & Eng. vibrations, 3(1).
Shenthan T, Thevanayagam S, Martin GR (2004b): Densification of saturated silty soils using
composite SC for liquefaction mitigation. Proc., 13th WCEE, Vancouver, BC, Canada.
Shenthan T (2005): Liquefaction mitigation in silty soils using composite stone column. Ph.D.
Dissertation, University at Buffalo, Buffalo, NY.
Thevanayagam S, Shenthan T, Mohan S, Liang J (2002): Undrained fragility of sands, silty sands and
silt. ASCE, J. Geotech. & Geoenv. Eng., 128 (10), 849-859.
Thevanayagam S, Shenthan T, Kanagalingam T (2003): Role of intergranular contacts on
mechanisms causing liquefaction and slope failures in silty sands. USGS Report, Award#
01HQGR0032 and 99HQGR0021, US Dept of Interior.
Youd et al. (2001): Liquefaction resistance of soils: summary report from the 1996 NCEER and
1998 NCEER/NSF workshops on evaluation of liquefaction resistance of soils. J. Geot. & Geoenv.
Eng., ASCE, 127(10): 817-33.

Soil Densification Using Vibro-Stone Columns Supplemented with Wick Drain 15