See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/311708570

The use of sand columns to improve soft soil

Conference Paper · March 2002

CITATIONS READS

5 3,927

2 authors:

Alaa Hussein Al-Zuhairi Mosa J. Al-Mosawe

University of Baghdad 23 PUBLICATIONS 107 CITATIONS
62 PUBLICATIONS 70 CITATIONS
SEE PROFILE
SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Piles foundation View project

Behavior and Analysis of Piled Raft Foundation View project

All content following this page was uploaded by Alaa Hussein Al-Zuhairi on 18 December 2016.

The user has requested enhancement of the downloaded file.

 The use of sand columns to improve soft soil
Prof. Dr. Al-Mosawe M. J. Al-Zuhairi, A. H.
Director: Consulting Eng. Bureau Ph. D. Student
College of Engineering College of Engineering
University of Baghdad University of Baghdad
Abstract
This investigation describes the efficiency of using sand columns in improving soft clayey soils'
Variables such as reinforcement ratio (number and cross section of sand column)and relative density of column
material and their effect on the new soil system were studied. The effectoftheloading areaatdlengthof
column penetration is studied also.
To achieve the above aims, a detailed laboratory investigation was carried out on sand column
reinforced soil using a series of triaxial tests on the composite specimens and performing laboratory plate
loading tests on models of single sand column in soft clay.
The test results hightiglit the effect ofthevariousparametersthataffecttheperformanceofthesand
column and provide fundamental points govern the behaviour. Simple and useful charl is also suggested as a

guide for estimating the reinforcement requirements for a given soil improvement demand.

Keywords
Sand column, soft soil, soil improvement.

Introduction
Among the several methods applied in ground improvement the compacted granular column has been
increasingly usid during the past three decades as a technique to reinforce soft and very soft compressible clays,
silts for loose silty sands . This is because it represents the cheaper way among the other methods. The material
used is cheep and the technique is simple. The concept of using granular piles was first applied in France in
1830 by French rrilitary engineers to improve a native soil.
According to the grain size of column material, granular columns may be divided into two categories;
stone and sand columns. Both of the two types can improve the soft soils by increasing load bearing capacity,
accelerating consolidation process and reducing the final settlement ofthe soft soil.
Theoretically, sand columns improve foundation bearing capacity because they are stiffer than the
material which they replace (Al-Kafaji, 1996). Both column and surounding soil have amutualeffectto
increase the bearing capacity of the improved ground. The compacted sand pile takes its shear strength from
lateral support provided by the surounding soil like the action of chamber pressure in triaxial cell.
Greenwood, (1910) was one of the first who came up with mechanisms and explained the load transfer
phenomenon in granular column, (Rao, 1997). He assumed that thecolumnisloadedtriaxially,usedBell's
^(f to
Sf S) theory foi passive pressure in the estimation of the radial stress and quoted the following formula
estimate the ultimate load capacity of a single column.
ocu : Kp. *(2Co +oro).. . ... .(1)

Where,
o"u: Ultimate veftical stress carried by the granular pile'
Ko" : Rankine passive pressure coefficient for the column material.
o,o: Initial in situ radial stress'
Based on the laboratory model experimental results obtained by Hughes and Withers, (1974),they
concluded that the ultimate strength of the column is governed by the limited maximum lateral resistance of the
soil around the column during bulging, in other words, the sand column can be thought of as being confined in a
triaxial stress system where the cell pressure is limited. The authors presented their method of analysis that was
developed from the results of plasticity theory fiom which the limiting stress is given by Gibson and Anderson
in 1961 as:

orl- = oro
I (-e.
+cft+'.[e*l.,l-]
\l .(2)

ratio
Where, o,o, E, [ & c.. arc respectively, the total in situ lateral stress, the elastic modulus, Poisson's
and the undrained cohesion.
Al-Khafaji, (1996) examined the use of the sand compaction pile forreinforcingsoftclayground
through using the centrifuge modeling technique. The major parl of the study was concentrated on the settlement
reduciion obtained improvement and the following empirical equation.
 (.-c \
S^=2.I*tnl l-2.1 ..........(3)
" [E,,J
E"&E,: are the modulus of elasticity of column material and soil respectively.
In which
Laboratory work
Three types of soils were used in this study; kaolin clay for the soft media, natural sand and gravel to
be useci in the construction of columns. Two molds were designed and manufactured for this purpose to produce
100 mm annular soft clay samples to be tested in triaxial apparatus and to prepare soft clay beds required in
model testing program. See Fig.(l) and Fig.(2).
Kaolin slurry was prepared by well mixing of kaolin powder passing no. 40 sieve with a certain amount
of water that give the slury twice the liquid limit. The sluny was then left in airtight container for (5-6) days to
ensure the fulI saturation of kaolin particles. The slurry was poured in one of the molds shown in Fig.(l) and
Fig.(2) and left under a consolidation stress of 100 kPa for about (130-140) hours. The sand or gravel column
was installed in the middle of soft samples driving a thin-walled brass tube and draw it carefully to form a hole
of the required diameter. The hole was then, backhlled with column material (sand or gravel) in five successive
layers each layer was compacted with a light steel rod to get the required relative density.

Testing programme
Eleven CU-triaxial compression tests were performed to cover the studied parameters (reinforcement
ratio a, and relative density of column material D,). Three values of reinforcement ratio were taken 1176,117 and
1/4. Also, three values of D. were considered 20oA,60yoand85o/o for each selectedreinforcementratio. The
consolidation pressures were 100, 200 and 300 kPa.
The model testing procedure was carried out according to ASTM D1194-72 Standard on eleven models
to study the following variables:
a. The ratio of loading areato the cross-sectional area of the sand column (Ap/A.). Four values were chosen.
b. The length of the column to its diameter ratio (L/D). Four values were taken 7,6,4 and2.
c. The relative density of the sand column (D,) 20o , 60%o and 85oh.
Results and discussion
Results of triaxial tests
Figure (3) shows the results of CU-triaxial compression test on 100mm composite sample of soft kaolin
clay reinforced by 37.5mm sand column with D,:85% (as an example for all tests on composite samples).
To show the effect of reinforcement ratio (a,) on the performanceof sandcolumninsoftclay,the
factor (F6) was introduced and obtained from:
*u^.(o1. )
r., "" "(4)
-u*.(o1u)
Where, o1,, o1u,:Totol axial shess of reinforced and unreinforced sample'
Table (1) shows the effect ofboth a, and D, on the values ofF6.
/ \
\
Also, the effect of both variables on the normalized undrained shear strength I " ["t,1
I can be seen in Fig.(4)'

In order to show the improvement gained using sand column a factor F, is introduced and defined as:
(s, )reinr.soir
-.' =
/(\
""""""\J/
G, Lnr.*r.,
Where, So:is the undrained shear strength.
Similar to Fig.(4), Figure (5) is formed which the relationship between F, , a, and D, cdn be obtained as

follows:

Fr =1+Cl *as +C2xa"2 (6)

Where,

Cl = 2.Sle - 21.23* Dr * 18.61*D12

CZ = -13.02 +153.74* Dr 87'96*D12

The range of application of Fig. (5) can be extended by substituting another values ofa, and D' and
calculating the conesponding f. from Equation (6). Fig. (6) presents the results as a design chaft from which an
urr"rr*.rrlt of the improvemint method can be obtained. This figure is useful to some extent to estimate the
required reinforcement ratio or volume of sand replacement per cubic meter of untreated land for a given
improvement requirement and specihed relative density of sand used in the column.
Results of model tests
The ratio of ultimate bearing capacities of treated to untreated soil (qu,/q,,) was decreased with the
increasing in the ratio of loading area to the area of sand column used (Ao/A,). This result may be supported by
the following equations with aid of Fig. (7).
P.
P = P, + ....(7)
o*Ap -os *A, +o. *A. """(8)
where, A": loaded area of clay, Ac = Ap - At
.'.o*Ap -os *A, +o. -(oo *or) (e)

o.,-1 =o'*A'*AP*A' ,..(10)

o'c oc Ap Ap

But, n = 5 i, concentrationratio
oc
o
Thus.--l+-i-l--
AsG -,) (11)
oc Ap

At failure, - is approximately equal to (q*/q,). Hence, (q,/qJ is decreased with increasing of (Ao/A')
oc
keeping (n) unchanged.
It was found that with increasing of (L/D) ratio, the (q*/qJ ratio was increased. For example, when
(t.iD) changed from 2 to 4 the ultimate bearing capacity of reinforced soil increased 2 folds. This may be related
to the fbllowing:
1. When the column is floating in the soft clay (i.e. LID <7) the stresses at the tip of the column can not be
reflected above, but dissipated in the soft clay.
2. The granular column takes its capacity from bulging which is usually happened in long columns. Shofi
columns rarely exhibit this phenomenon and may be considered as a short pile.

Conclusions
1. The undrained shear strength of reinforced sample (S") was found to be depended on both reinforcement
ratio a, and relative density of the sand in the column D,. The improvement gained was highly affected by
the number of sand columns used rather than the relative density of soil in the column.
2. The increase in bearing capacity is inversely proportional with the ratioof footingareatothesand
column cross-section al area (Ap/A,) and directly with L/D ratio and relative density of sand.
3. The reduction in settlement is affected in the same manner mentioned in point 2 above.
References
1. Al-Khafaji, Z. A., Reinforcement of Soft Clay Using Granular Columns. Thesis presented to the
University of Manchester at UK, in partial fulfillment of the requirements for the degree of Ph. D (1996).
2. Greenwood, D. A., Mechanical improvement of soils below ground surface. Proceedings Ground
Errgineering Conference, Institution of Civil Engineering, June 1970'
3. Hughes,- J. M. O. and Withers, N. J., Reinforcing of soft cohesive soils with stone columns. Ground
Engineering, Vol. 7, No. 3, May 1974,pp.42-49.
 tffi- Perforatedsteeldisc
\
ffi Porousstone

ttt
ltt
-
Irt aoo,mm
I l- 0loommPipe
I Kaolin I

I sturv
III I

ttt
Ltl
ttt
tlt
I 1_ _l a Porousstone

,,ffirr-
'
Herorateo
steel 0tsc

Consolidation Cell

Figure (1): O 100mm consolidation apparatus

0.01rrm
Dailprur

Figure (2): Soft clay bed preparation

 1.0

a
@
I nr
a
!
o
.S 0.6
()
Consolidation pressure
o tl 4
N-" r 100 kPa

i-.i n, r 200 kPa

o'-
o 300 kPa

0.0
048121620
Axial strain, (%)
A: Normalized Deviator Stress-Strain Relationship
o. 0.6
B
a.
a
a
3
x
o.+
o
Consolidation pressure
()
.! or r 100 kPa
(o r 200 kPa
o
o 300 kPa
z 0.0
048121620
Axial strain, (o%)
B:Normalized Excess PWP-Axial Strain Relationship
240

s Consolidation pressure
a t80
}Z
d
I 100 kPa

R t20 r 200 kPa

a o 300 kPa
2 a
()
E60
a I a
t
I a
a
I

0 I 00 200 300 400

Mean effective principal stress, p' (kPa)
C: Effective Stress Paths
Figure (3): CU-Triaxial compression test results

Table (l): Effect ofa, on values ofF6

Values of Fa

Reinforcement D,:20 oh D,: 60 oh D,: 85 Yo

ratio (a,) o":100 o o

o o o o'":100 o o
":100 ":200 ":300 ":200 ":300 ":200 ":300
kPa
kPa kPa kPa kPa kPa kPa kPa kPa

1116 1.03 1.01 L01 1.08 r.08 I .08 1.121 t.t2 1.14

U7 1.04 1.03 r.03 1.14 t.12 t.12 t.25 1.26 t.24

U4 1.22 t.20 120 1.63 t.31 1.52 2.15 2.10 r.89
 t.2

Pl0
o
H C) losse sand column reinf. soil

G
() o medium sand column reinf. soil
€ o.s
o dense sand column reinf. soil

d
.d

f 0.6
o
N
(t

.o 0.4
A

o.2
0.0 0.1 0.2 0.3
Reinforcement ratio, as

Figure (4): Relation between Normalized Undrained Shear Strength and Reinforcement Ratio
for Different Relative Densities of Sand in the Column

Soil reinforced bv:

rf 3.0 C Loose sand column

o
o
O Medium sand column
-(o
O Dense sand column
()
6 rn
o
L
q

Reinforcement ratio, a"

Figure (5): Relation between Improvement Factor and Reinforcement Ratio for Different
Relative Densities of Sand in the Column
 o\
tr
a
a
() Values of F
u,
\\ \%,
C)

()
lr
\\\\\Yoo ,\'oo
\!
?\\
E\\oo\\ oo\
o
\o.
\
a
\\
\ \
\\
\\
\ Qoo'

\ \, \,*
,"s?:8--i \,*

0. r0 0.20 0.30 0.40

Reinforcement ratio, as

Figure (6): Prediction of Improvement Factor F.

Loading
Plate

Sand
Column

Figure (7): Analysis of Stresses in Sand Column Model

View publication stats