IGC 2009, Guntur, INDIA

Expansive SoilsProblems and Remedies

EXPANSIVE SOILSPROBLEMS AND REMEDIES

B.R. Phanikumar
Professor, Department of Civil Engineering, VIT University, Vellore632 014, India.
E-mail: [email protected]
ABSTRACT: Expansive soils are one of the examples of tropical soils. These soils pose problems to civil engineers in
general and to geotechnical engineers in particular. They cause damage to structures founded in them because of their potential to
react to changes in moisture regime. They undergo severe volume changes corresponding to changes in moisture content. Because
of their potential to undergo volumetric changes, civil engineering structures such as foundations, retaining walls, pavements,
airports, side walks, canal beds and linings are damaged. Various innovative practices have been devised to counteract the
volume change problem and to safeguard the structures. This paper discusses the problems and remedies in detail.
1. INTRODUCTION
1.1 Expansive Soils and Problems
Expansive soils pose problems to civil engineers in general
and to geotechnical engineers in particular (Chen 1988). They
cause damage to structures founded in them because of their
potential to react to changes in moisture regime. They undergo
severe volume changes corresponding to changes in moisture
content. They swell or increase in their volume when they
imbibe water and shrink or reduce in their volume on
evaporation of water (Chen 1988). Because of their alternate
swelling and shrinkage, they result in detrimental cracking of
lightly loaded civil engineering structures such as foundations,
retaining walls, pavements, airports, side walks, canal beds
and linings (Chen 1988).
2. EXISTING FOUNDATION PRACTICES
2.1 Mechanical, Physical and Chemical Alterations
This involves excavation of expansive soil and replacement
with non-expansive material, where the depth of active zone
(depth from ground surface wherein seasonal moisture
fluctuations occur) is small and where a suitable replacement
material is available. Sand cushion method and Cohesive
Non-Swelling (CNS) layer method are very popular.
2.1.1 Sand Cushion Method
In this method, the entire depth of the expansive clay stratum
if it is thin, or a part thereof, if it is deep enough, is removed
and replaced by a sand cushion compacted to the desired
density and thickness. It was reported that swelling pressure
varies inversely as the thickness of the sand layer and
directly as its density. Hence, sand cushions are formed in
their loosest state to avoid the possibility of excessively
increasing the swelling pressure without, however, violating
the criterion of bearing capacity. The basic philosophy of this

method is that, in monsoon, the saturated sand occupies less

volume, accommodating some of the heave of underlying
expansive soil, and in summer, partially saturated sand bulks
and occupies the extra space left by the shrinkage of the soil.
2.1.2 Cohesive Non-Swelling (CNS) Layer Method
In this method, about top 1 m to 1.2 m of the expansive soil
is removed and replaced by a cohesive non-swelling soil
layer. According to Katti (1978), with saturation of
expansive soil cohesive forces are developed up to a depth of
about 1.01.2 m and counteract heave. CNS layer creates an
environment similar to that around 1 m deep in an expansive
soil with equivalent cohesion to counteract heave.
In physical alteration, granular material is mixed with expansive
clay to minimize heave. However, the permeability of the
resulting blend would be more than that of the expansive soil
resulting in a faster ingress of water into the soil. Chemical
alteration involves addition of chemicals to expansive clay to
reduce heave by altering the nature of expansive clay minerals
(Chen 1988). Lime treatment of expansive soils is the most
widely used technique and the most effective technique of
chemical alteration to minimize volume changes and to increase
the shear strength of foundation expansive soils. The usefulness
of the treatment depends on the reactivity of the soil to lime
treatment and the extent of dispersion of lime into the soil.
In pavements, a technique called Lime-Slurry Pressure
Injection (LSPI) is also used for minimizing swelling of
soils. In this technique, lime-slurry is injected into drill-holes
under a pressure of 15 kg/cm2 (Chen 1988). In this technique,
lime slurry penetrates into the fissures in the soil mass to a
sufficient depth (usually 8 to 10 feet), and the lime-filled seams
help control the soil water content, reduce volumetric changes
and increase soil strength. However, for LSPI to be an effective
technique, the expansive clay soil must contain an extensive
network of fissures. Otherwise, lime cannot penetrate into the
relatively impermeable soil to an appreciable distance from
the injection hole to form a continuous lime seam moisture

907

Expansive SoilsProblems and Remedies

Addition of calcium chloride (CaCl2) to expansive soil has

also proved efficacious in altering the swelling properties of
the soil. Calcium chloride is a hygroscopic material and hence,
is pre-eminently suited for stabilization of expansive soils,
because it absorbs water from the atmosphere and prevents
shrinkage cracks occurring in expansive soils during summer.
Addition of calcium chloride to expansive soils reduced
Plasticity Index (PI), free Swell Index (FSI%), swell potential
and (H/H, %) and swelling pressure, ps, significantly. Another
additive that has been found to be quite promising in reducing
the swelling characteristics and improving the engineering
behavior of expansive soils is fly ash. The efficacy of fly ash
as an additive to expansive soils in ameliorating their properties
was reported.

granular pile was varied as 50%, 60% and 70%. The height
of the expansive clay bed and the granular pile-anchor were
the same in all the tests. A square mild steel plate of size 100
mm 100 mm was used as the surface footing in the heave
tests (set up not shown here). The rate and amount of heave
of the unreinforced clay bed were compared with those of the
clay bed reinforced with granular pile-anchor. The
unreinforced expansive clay bed attained a final heave of 9%
in 9 days. However, the heave of the expansive clay bed
reinforced with granular pile-anchor attained a reduced
amount of heave of 1.15% in a short period of 3 days (see
Fig. 1). As surface area of the GPA increased, heave of the
clay bed decreased (see Fig. 2).
10
9
8
7

% Heave

barrier Lime-soil columns were also tried to stabilize

expansive clays in-situ. It was reported that diffusion of lime
into the ambient soil is effective up to a radial distance of
about 3 times the diameter of the lime-soil column.

2.2 Special Foundation Techniques

The problem of heave or uplift of foundations caused by
expansive soils is one of tension developed in the soil due to
swelling. Hence, tension-resistant foundations are required
for counteracting the heave problem. Various special foundation
techniques such as drilled piers, belled piers (Chen 1988) and
under-reamed piles are all tension-resistant foundations.

A total of 81 laboratory model tests were conducted for

studying the heave behaviour. The relative density of the

S'

Dry unit weight of soil = 14 kN/m

wi of soil = 14%
Length of pile = 300mm
Diameter of pile = 40mm
Relative density = 0.6

3
2
1
0
0

1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000 12000

Time (minutes)

Fig. 1: Rate of Heave

2.4
Length = 300mm
Length = 400mm
Length = 500mm

2.0

1.6
S'

Granular Pile-Anchor (GPA) technique has been a recent

innovation over the conventional granular pile, modified into
an anchor (Phanikumar 1997). This paper discusses GPA
system in detail as it is the latest most successful tensionresistant foundation system. Laboratory testing performed to
study the efficacy of Granular Pile-Anchor Foundation
(GPAF) system in reducing the amount of heave and improving
the engineering behavior of expansive clay beds gave promising
results. The effect of length, diameter and relative density of
GPAs was studied. In a GPA, the foundation is anchored at
the bottom of the granular pile to a mild steel plate through a
mild steel rod to render the granular pile tension-resistant by
the effect of anchor, which is able to counteract the uplift
force exerted on the foundations. In a granular pile-anchor,
the resistance to uplift is developed mainly due to, (i) the
weight of the granular pile acting in the downward direction
and (ii) uplift resistance due to friction mobilized along the
cylindrical pile-soil interface, and precludes the possibility of
heave of foundations. The upward force acting on the
foundation is because of swelling of the expansive soil on
imbibition of water. The resisting force acting in the
downward direction and counteracting the uplift force is due
to friction mobilized along the pile-soil interface, which
depends on the shear strength parameters of the interface.

Dry unit weight of soil = 14 kN/m3

wi of soil = 14%

1.2
Granular pile
Relative density =
0.6

0.8

0.4

0.0
0

10

12

14

16

18

Lg/Dg

Fig. 2: Effect of l/d Ratio on % Heave

for a Given Length
The pull-out capacity of the granular pile-anchors increased
with increasing length of the pile-anchor and relative density
of the granular material. For a pile of length of 250 mm, the
load for 30 mm upward movement was about 250 N,
whereas for a pile of length of 350 mm, it was equal to about

908

Expansive SoilsProblems and Remedies

510 N, indicating an increase of about 100%. The % increase

in the pull-out load when the relative density was increased
from 50 to 70% was about 30% for an upward movement of
30 mm (see Figs. 3 and 4). It was found that the load-carrying
capacity of the expansive clay bed also increased by the
installation of granular pile-anchors. The loading intensity
for a settlement of 25 mm in the case of the reinforced clay
bed increased by 2 times over that for the same settlement
in the case of the unreinforced clay bed (see Fig. 5).
60

Granular pile
Diameter = 30 mm
Relative density = 0.6

50

used in this investigation was collected from a depth of 1.5 m

below the ground level in Amalapuram, Andhra Pradesh,
India. The soil was a highly swelling soil with Free Swell
Index (FSI) of 180%. It was a CH soil. Table 1 shows the
index properties of the expansive soil. The maximum dry
unit weight was 14 kn/m3 and the optimum moisture content
was 27%. The granular material used for the installation of
granular piles was a blend of 20% stone chips and 80% coarse
sand. The particle sizes of stone chips and coarse sand were
the same as the granular material used in laboratory testing.
It may be mentioned here that all the granular piles in the test
program were compacted at a relative density of 70%.
2

40

10

11

0
30

Granular pile alone

Composite ground (soil and pile)
Soil alone

6
20

Length = 0.35m

12
Settlement (mm)

Load x 10 (N)

Stress (kN/m )

Length = 0.30m

10

Length=0.25m

0
0

10

20

30

40

50

60

Deformation (mm)

18
24
30
36

Fig. 3: Influence of Length of Granular Pile-Anchor

on the Pullout Behavior

42
48

80

Fig. 5: Stress-Settlement Curves

Granular pile
Diameter = 30 mm
Length = 0.30 m

70

Table 1: lgp/dgp Ratios of Granular Pile-Anchors

Diameter of
Length of the pile, mm
the pile, mm
500
750
1000

Load x 10 (N)

60
50
40

100

7.5

Relative density=0.7

150

3.33

6.67

Relative density=0.6

200

2.5

3.75

30
20
10

10

Relative density=0.5

Granular pile-anchor system was also tested in field

conditions by performing in situ tests (Rao et al. 2004). It
was found that heave reduction and the rate of heave were
similar to those obtained in the laboratory model tests.

In all the tests the expansive clay beds were compacted at a

placement water content of 15% and dry unit weight of
14 kN/m3. The thickness of expansive clay beds was fixed at
1000 mm. the length of the granular pile-anchor (lgp) was
varied as 500, 750 and 1000 mm. for each length, the
diameter of granular pile-anchors (dgp) was varied as 100,
150 and 200 mm. Hence, the range of the length to diameter
ratio (lgp/dgp) of granular pile-anchors varied from 2.5 to 10.
Table 1 shows the l/d ratios of the GPAs tested. Heave of the
clay bed at different depths and radial distances from the
center of the granular pile-anchor was also continuously
monitored as the clay bed was inundated.

An extensive test program was conducted in situ for establishing

the efficacy of granular pile-anchors in expansive clay beds.
Amount of heave and rate of heave were observed both in the
case of un-reinforced expansive clay beds and clay beds
reinforced with granular pile-anchors. The expansive soil

The un-reinforced expansive clay bed attained a final amount

of heave of 154 mm or final % heave of 15.4%. As there was
no technique provided in the clay bed for arresting heave, the
clay bed heaved to its fullest extent. However, in the case of
clay beds reinforced with GPAs in situ, heave was reduced

0
0

10

20

30
40
Deformation (mm)

50

60

Fig. 4: Effect of Relative Density on the

Pull-Out Behavior
3. FULL SCALE TESTING ON GPA SYSTEM

909

Expansive SoilsProblems and Remedies

considerably. The respective amounts of reduced heave of

clay beds reinforced with GPAs of length 1000 mm and
diameters 100,150 and 200 mm were 48, 27 and 15 mm. Hence,
the final % of heave was 4.8%, 2.7% and 1.5% in the above
cases (Phanikumar et al. 2007). The amount of heave was
reduced quite effectively by reinforcing the clay beds with
GPAs due to the effect of anchorage and shear resistance
mobilized along the pile-soil interface (see Table 2). Regarding
the rate of development of heave, the un-reinforced expansive
clay bed required about 210 days or 7 months of continuous
wetting to attain the final amount of heave (154 mm), whereas
the clay bed reinforced with granular pile-anchors required a
much less time. For example, the clay bed reinforced with
granular pile-anchors 1000 mm long and 100, 150 and 200
mm in diameter required a time period of 120 days or 4 months.
The expansive clay beds reinforced with gpas adjusted quickly
to moisture changes. In all the cases it was found that the
amount of heave decreased with depth from the top, reaching
zero amount of heave at the bottom of the pile-anchor or clay
bed. It may be mentioned here that the length of the pile-anchor
and the thickness of the clay bed are equal in this case.
Table 2: Reduced % heave (S) of Clay Beds by GPAs
(Heave of un-reinforced expansive clay bed = 154 mm)
Diameter
of GPA
(mm)
100
150
200

Length of GPA, mm
500

750

1000

75
63
46

64
47
38

48
27
15

However, in the case where the length of the pile-anchor was

less than thickness of the clay bed, some amount of heave was
recorded at the bottom of the pile-anchor. This was because
the expansive clay beneath the base of the pile-anchor
swelled. Heave increased with increasing radial distance from
the center of the granular pile-anchor in curvilinear fashion.
Further, a test program was conducted in the field on
granular pile-anchors embedded in expansive clay beds. The
reinforced clay beds were saturated and the amount and rate
of heave were observed and compared with those of the
unreinforced clay beds. After complete saturation, pullout
load was applied on the granular pile-anchor and the upward
movement noted. In total, nine field pullout tests were
conducted on single granular pile-anchors varying the length
and diameter of granular pile-anchor. One pullout test was
conducted to study the group behavior also (Rao et al. 2007).
Figure 6 shows the pullout load-upward movement curves
for single granular pile-anchors of a uniform diameter of 200
mm with lengths varying as 500, 750 and 1000 mm. The

curves indicate the influence of length of granular pile-anchor.

All the granular pile-anchors were tested up to failure. The
curves presented in the figure show that the failure pullout
load increased with increasing length of the granular pileanchor. The curves also indicate that, at all stages of loading,
the upward load required to be applied on the granular pileanchor to cause a given upward movement increased with
increasing length of granular pile-anchor. However, it may
be observed that, up to an applied uplift load of 4 kN, there
was no significant upward movement in the granular pileanchors. This is because of the weight of the granular pileanchor (about 0.5 kN) and the shear resistance mobilized in
the downward direction along the cylindrical pile-soil
interface (about 3 kN) because of the anchorage. Figure 7
shows the pullout behavior of granular pile-anchors of 1000
mm length but of diameters varying as 100, 150 and 200
mm. The curves in the figure reflect the effect of diameter of
Granular Pile-Anchor (GPA). All the curves in the figure
show that the failure pullout load increased with the
increasing diameter and weight of the granular pile-anchor.
The applied upward load was observed to increase with
increasing diameter of granular pile-anchor at all stages of
the test as the resistance to uplift increased with increasing
surface area of the pile-soil interface (Rao et al. 2007).
Figure 8 shows the variation of failure pullout load (kN) with
the l/d ratio of Granular Pile-Anchors (GPA). The curves
present the data for different lengths of the granular pileanchors. For a given l/d ratio the failure pullout load
increased with increasing length of the granular pile-anchor.
This is attributed to the increase in the pullout resistance with
increasing surface area of the granular pile-anchor. Similarly,
for a given length of the Granular Pile-Anchor (GPA), the
failure pullout load increased with increasing diameter or
decreasing l/d ratio. Increasing diameter increases the surface
area and pile-anchor weight and consequently uplift resistance
and results in increased failure pullout load (Rao et al. 2007).
Figure 9 shows, by comparison, the pullout behavior of the
laboratory scale granular pile-anchor and field scale granular
pile-anchor. In both the cases, the granular pile material was
compacted to a relative density of 70%. Similarly the l/d ratio
of granular pile-anchor in both the cases was equal to 10. While
the diameter and length of laboratory scale granular pileanchor were 30 and 300 mm, those of the field scale granular
pile-anchor were 100 and 1000 mm, respectively. The
thickness of the expansive clay bed was equal to the length
of the Granular Pile-Anchor (GPA) in both the cases. The
curves in the figure show that the uplift load (kN) required to
be applied for any given upward movement (mm) of the
granular pile-anchor was significantly higher in the field
scale Granular Pile-Anchor (GPA) than that in the case of
laboratory scale Granular Pile-Anchor (GPA).

910

Expansive SoilsProblems and Remedies

The large variation in the behavior of the Granular PileAnchors (GPAs) in the laboratory study and the field study
was due to the scale effect (Rao et al. 2007).

Pullout Load (kN)

l = 1000 mm
12

l = 750 mm

l = 500 mm

0
0

10

20

30

40

50

Upward Movement (mm)

Fig. 6: Pullout Behavior of Granular Pile-Anchors

of Diameter 200 mm

Pullout laod (kN)

16
12
8

Diameter = 100 mm
= 150 mm

= 200 mm

0
0

10

20

30

40

50

Figure 10 shows, by comparison, the variation of failure

pullout load with l/d ratio of granular pile-anchors in
laboratory scale study and field scale study. The data pertain
to three field scale Granular Pile-Anchors (GPA) of length
1000 mm (diameter = 100, 150 and 200 mm) and three
laboratory scale Granular Pile-Anchors (GPA) of length 500
mm (diameter = 30, 40 and 50 mm). The expansive clay beds
in both the cases were compacted at placement water content
of 15% and dry unit weight of 14 kN/m3. All the granular
piles were compacted to a relative density of 70%.
Irrespective of the l/d ratio the failure pullout load for field
scale granular pile-anchors was much higher than that for
laboratory scale granular pile-anchors. This was because of
the increased surface area, frictional resistance and the
weight of granular material in the case of field scale granular
pile-anchor and the consequent increase in uplift resistance.
Moreover, the lateral swelling pressure of expansive clay bed
in situ, which confines the Granular Pile-Anchor (GPA) is
also significantly high compared to that in the laboratory
scale expansive clay bed (Rao et al. 2007).

Upward movement (mm)

Failure pullout load (kN)
for field scale GPA

Fig. 7: Pullout Behavior of Granular Pile-Anchors

of 1000 mm Length

Failure pullout load (kN)

16

16

0.92

12

0.88

0.84

Field scale GPA

12

Failure pullout load (kN)

for laboratory scale GPA

16

Laboratory scale GPA

0.8
2

10

14

18

l = 500 mm
l = 750 mm

l/d

l = 1000 mm

Fig. 10: Variation of Failure Pullout Load with l/d in Field

and Laboratory Scale Granular Pile-Anchors

4
2

10

12

l/d

Fig. 8: Variation of Failure Pullout Load with l/d Ratio

of Granular Pile-Anchor

40

Pullout behavior
(field scale)
Pullout behavior
(laboratory scale)

20

Failure pullout load (N)

for laboratory scale
GPA

60

Failure pullout load

(kN) for field scale
GPA

12

0
0

10

20

30

40

Upward movement (mm)

Fig. 9: Comparison of Pullout Behavior of Laboratory Scale

and Field Scale Granular Pile-Anchors (l/d = 10; Dr = 70%)

Figure 11 shows, by comparison, the pullout behavior of the

granular pile-anchor when tested single and when tested
under group effect. The granular pile-anchor under group
effect resulted in increased uplift load for a given upward
movement in comparison to that of a single granular pileanchor tested. This is because of the influence of the granular
pile-anchors in the group on the test granular pile-anchor. As
heave of the expansive clay bed was reduced significantly on
the installation of group of granular pile-anchors, which also
act as tension members, there was not much reduction in the
dry unit weight of the expansive clay bed. As a result the
pile-soil interface friction for the group would be much more
than in the case of the single granular pile-anchor. This also
resulted in a higher lateral swelling pressure (confining the
test granular pile-anchor radially) than in the case of a single
granular pile-anchor. There is an arching action between the
911

Expansive SoilsProblems and Remedies

0
0
-5
-10

Stress (kN/m )
400

600

800

1000

un-reinforced expansive clay bed

composite ground (Phanikumar et al.
2004) - laboratory scale
composite ground - in situ (150 mm dia
granular pile-anchor)
p ile-anchor alone ( Phanikumar et al.
2004) - laboratory scale
p ile-anchor alone -in situ (150 mm dia)

Settlement
(mm)
Settlement
(mm)

-15

200

-20
-25
-30
-35
-40
-45
-50

Fig. 11: Stress-Settlement Behavior of Un-Reinforced

Expansive Caly Bed and Clay Beds Reinforced with
Granular Pile-Anchors
Field-scale plate load tests were also conducted on in situ
GPAs installed in expansive clay beds to study the compressive
load response (Rao et al. 2007). Load-settlement behavior
was studied in the following cases:
(a) un-reinforced expansive clay bed
(b) composite ground with the test plate spreading on both
the granular pile-anchor and the expansive clay bed
(c) Granular Pile-Anchor (GPA) alone with the test plate
bearing only on the granular pile-anchor, and
(d) expansive clay bed reinforced with a group of granular
pile-anchors.
Figure 11 shows the stress-settlement curves for the unreinforced expansive clay bed, the composite ground (granular
pile-anchor and expansive clay bed together) and GPA alone.
The data plotted in the figure show that the clay bed reinforced
with granular pile-anchor gave an improved compressive
load response in comparison to the un-reinforced clay bed as
reflected in the stress-settlement curve for composite ground.
For example, the stress required to cause a settlement of
25 mm in the un-reinforced expansive clay bed was
200 kN/m2 whereas the stress required to cause the same

amount of settlement in the case of composite ground was

500 kN/m2. This indicates that the load-carrying capacity of
the clay bed reinforced with granular pile-anchor improved
by 150%. The compressive load response for the expansive
clay bed reinforced with group of granular pile-anchors
showed a significant improvement in load-carrying capacity
over the un-reinforced clay bed. Though the test plate was
placed on saturated clay alone in both the cases, the resistance
to compressive load offered by the clay reinforced with group
of granular pile-anchors was higher than that by un-reinforced
clay. Heave was reduced significantly in the expansive clay
bed reinforced with granular pile-anchors. This means that
the decrease in dry unit weight or the increase in void ratio of
the clay bed owing to heaving of clay was arrested by
granular pile-anchors. Hence, the reinforced clay bed could
resist the compressive load better than the un-reinforced clay
bed wherein complete heave occurred leading to significant
reduction in dry unit weight or increase in void ratio. The
stress required to cause a settlement (d) of 25 mm in unreinforced expansive clay bed was 200 kN/m2, whereas it
increased to 330 kN/m2. Figure 12 shows the effect of length
of GPAs on stress-settlement behaviour for a given diameter
of 150 mm. The figure shows the stress-settlement curves for
un-reinforced clay bed and clay bed reinforced with single
granular pile-anchors having different lengths (500 mm,
750 mm and 1000 mm) but uniform diameter of 150 mm. It
may be mentioned here that the data shown in Figure 12
pertain to load tests performed on granular pile-anchors alone
(test plate bearing only on the granular pile). Increasing
length of the GPA resulted in higher amount of stress for a
given settlement (d). For example, the stress required for
25 mm settlement was respectively 360, 505, 690 kN/m2 for
500 mm, 750 mm and 1000 mm long granular pile-anchors.
A similar trend was observed with increasing length of GPA
2

Stress (kN/m )
0

100

200

300

400

500

600

700

800

900

0
un-reinforced expansive clay bed

Settlement (mm)

granular pile-anchors, which offers more resistance to uplift.

The uplift load required to be applied on the granular pileanchor to cause an upward movement of 25 mm was 13.7 kN
when tested under group effect as against an uplift load of
11.25 kN for the same amount of upward movement of
25 mm when tested single. This indicated a percentage increase
of 22.22% in the applied uplift load when the granular pileanchor was tested under group effect. The failure pullout
load of the granular pile-anchor when tested under group was
18 kN as against a failure pullout load of 12 kN for the granular
pile-anchor when tested single, indicating an improvement of
50% in the failure pullout load (Rao et al. 2007).

-10

3 - group granular pile-anchors

composite ground (150 mm dia
granular pile-anchor)

-20

-30

-40

-50

Fig. 12: Stress-Settlement Behavior of Unrienforced

Expansive Clay Bed and Clay Bed Reinforced with Group of
Granular Pile-Anchors

912

Expansive SoilsProblems and Remedies

for other diameters also. Figure 12 shows the stresssettlement behavior of un-reinforced expansive clay bed and
clay bed reinforced with group of GPAs.

Phanikumar, B.R. (1997). A Study of Swelling

Characteristics of and Granular Pile-Anchor Foundation
System in Expansive Soils, Ph.D Thesis Submitted to JN
Technological University, Hyderabad, India.

4. CONCLUSION

Rao, A.S., Phanikumar B.R., Suresh, K. and Sudhakar, V.

(2004). Compression Test on Granular Pile-Anchors
Embedded in Expansive Soils, Indian Geotechnical
Conference, Warangal, India, pp. 141144.
Rao, A.S., Phanikumar, B.R., Babu, R.D. and Suresh, K.
(2007). Pullout Behavior of Granular Pile-Anchors in
Expansive Clay Beds In-situ, Journal of Geotechnical
and Geoenvironmental Engineering, Vol. 133, Issue 5, pp.
531538.

Expansive soils, being problematic soils with their innate

potential for severe volume changes upon changes in moisture
content, need to be treated with special techniques. Sand
cushion, CNS layer, chemical and physical alteration techniques
have been successful. Granular Pile-Anchor (GPA) technique
has been a recent technique developed with some innovation
introduced into granular piles. GPA has met with considerable
success as observed in both laboratory and field scale tests.
Heave was reduced significantly and the surrounding soil
was found improved. Results of pullout tests and compressive
load tests indicated that length, diameter and relative density of
GPA govern heave and engineering behaviour of GPA system.
REFERENCES
Chen, F.H. (1988). Foundations on Expansive Soils, Elsevier
Scientific Publishing Company, Amsterdam.

Rao, A.S., Phanikumar, B.R. and Suresh, K. (2008).

Compressive Load Response of Expansive Clay Beds
Reinforced with Granular Pile-Anchors In-situ, Ground
Improvement, Vol. 161, No. 3, pp. 121129.
Phanikumar, B.R., Rao, A.S. and Suresh, K. (2008). In situ
Behavior of Expansive Clay Beds Reinforced with Granular
Pile-Anchors, Ground Improvement (in press-proof
checked).

913