FIELD AND LABORATORY EVALUATION OF TYPES

OF INTERCEPTING MEDIA USED FOR STRIP

FOUNDATION ON EXPANSIVE SOILS

Ahmed M. El Sharief *and Ibrahim M.A. Saeed**

ABSTRACT
Strip foundations are widely used on all grades of expansive soils in Sudan. Due to deficient performance of
conventional strip foundations, improvement techniques have been practiced. Intercepting layers of earthen
materials are often used between the foundation and the expansive soil in an attempt to reduce heave. This
paper reviews the current practice of using strip foundations on expansive soils in Sudan. It also presents
the results of laboratory and field tests in which several types of earthen materials like sand, boulders, plain
concrete with and without voids, cohesive non-expansive soils CNS are evaluated. Proper geotechnical
investigation was made prior to the tests. The field experiment constituted building brick walls 1.9 meters
high on treated and untreated expansive soils. The study area was then flooded and dried for two cycles for
a total period of 18 months. Measurements were taken by a precise level. The relative performance of all
treatment options was given. The field and laboratory tests showed that strip foundations, whether treated or
untreated, are not efficient when adopted in highly expansive soils and subjected to excessive wetting.
However the CNS and hardcore gave good performance and may be adopted if water is well controlled in
the built area.
‫ﺇﺳﺘﺨﻼﺹ‬
‫ ﻨﻅﺭﺍ ﻟﻘﺼﻭﺭ ﻫﺫﻩ ﺍﻻﺴﺎﺴﺎﺕ ﺍﻟﺸﺭﻴﻁﻴﺔ ﺍﻟﻌﺎﺩﻴﺔ ﻓﻲ ﻤﻘﺎﻭﻤﺔ‬.‫ﺍﻻﺴﺎﺴﺎﺕ ﺍﻟﺸﺭﻴﻁﻴﺔ ﺸﺎﺌﻌﺔ ﺍﻻﺴﺘﺨﺩﺍﻡ ﻋﻠﻰ ﻜل ﺃﻨﻭﺍﻉ ﺍﻟﺘﺭﺒﺔ ﺍﻟﻤﺘﻤﺩﺩﺓ ﺒﺎﻟﺴﻭﺩﺍﻥ‬
‫ ﺍﻷﻭﺴﺎﻁ ﺍﻻﻋﺘﺭﺍﻀﻴﺔ ﻤﻥ ﺍﻟﺘﺭﺒﺔ ﻏﻴﺭ ﺍﻟﻤﺘﻤﺩﺩﺓ ﻏﺎﻟﺒﺎ ﻤﺎ ﺘﺴﺘﺨﺩﻡ ﺒﻴﻥ ﺍﻻﺴﺎﺴﺎﺕ ﻭﺍﻟﺘﺭﺒﺔ‬.‫ﺘﻤﺩﺩ ﺍﻟﺘﺭﺒﺔ ﻓﻘﺩ ﺘﻡ ﺍﺨﺘﺒﺎﺭ ﻋﺩﺓ ﻁﺭﻕ ﺘﻘﻨﻴﺔ ﻟﻤﻌﺎﻟﺠﺘﻬﺎ‬
.‫ﻫﺫﻩ ﺍﻟﻭﺭﻗﺔ ﺘﻬﺩﻑ ﺍﻟﻰ ﻁﺭﺡ ﺒﻌﺽ ﺍﻟﺘﺠﺎﺭﺏ ﺍﻟﺤﺎﻟﻴﺔ ﻻﺴﺘﺨﺩﺍﻡ ﺍﻻﺴﺎﺴﺎﺕ ﺍﻟﺸﺭﻴﻁﻴﺔ ﻋﻠﻰ ﺍﻟﺘﺭﺒﺔ ﺍﻟﻤﺘﻤﺩﺩﺓ ﺒﺎﻟﺴﻭﺩﺍﻥ‬.‫ﺍﻟﻤﺘﻤﺩﺩﺓ ﻟﺘﻘﻠﻴل ﺘﻤﺩﺩﻫﺎ‬
،‫ ﺍﻟﺤﺠﺭ‬،‫ﻜﺫﻟﻙ ﺘﻭﻀﺢ ﻫﺫﻩ ﺍﻟﻭﺭﻗﺔ ﻨﺘﺎﺌﺞ ﺍﻟﺘﺠﺎﺭﺏ ﺍﻟﻤﻌﻤﻠﻴﺔ ﻭﺍﻟﺤﻘﻠﻴﺔ ﺍﻟﺘﻲ ﺃﺠﺭﻴﺕ ﻟﺘﻘﻴﻴﻡ ﺴﻠﻭﻙ ﺃﻨﻭﺍﻉ ﻋﺩﻴﺩﺓ ﻤﻥ ﻤﻭﺍﺩ ﺍﻟﺘﺭﺒﺔ ﻤﺜل ﺍﻟﺭﻤل‬
‫ ﺘﻡ ﺇﺠﺭﺍﺀ ﺒﺤﻭﺙ ﺤﻘﻠﻴﺔ ﺩﻗﻴﻘﺔ ﻟﻤﻭﻗﻊ ﺍﻟﺒﺤﺙ ﻗﺒل ﻫﺫﻩ‬.(CNS) ‫ ﺍﻟﺘﺭﺒﺔ ﺍﻟﻤﺘﻤﺎﺴﻜﺔ ﻏﻴﺭ ﺍﻟﻤﺘﻤﺩﺩﺓ‬،‫ﺍﻟﺨﺭﺴﺎﻨﺔ ﺍﻟﺒﻴﻀﺎﺀ ﺒﺜﻘﻭﺏ ﻭﺒﺩﻭﻥ ﺜﻘﻭﺏ‬
‫ ﺒﻌﺩ ﺫﻟﻙ ﺘﻡ ﻏﻤﺭ ﻤﺴﺎﺤﺔ‬.‫ ﻤﺘﺭ ﻋﻠﻰ ﺘﺭﺒﺔ ﻤﺘﻤﺩﺩﺓ ﻤﻌﺎﻟﺠﺔ ﻭﻏﻴﺭ ﻤﻌﺎﻟﺠﺔ‬1.9 ‫ ﺘﻀﻤﻨﺕ ﺍﻟﺘﺠﺎﺭﺏ ﺍﻟﺤﻘﻠﻴﺔ ﺒﻨﺎﺀ ﺤﻭﺍﺌﻁ ﻁﻭﺏ ﻻﺭﺘﻔﺎﻉ‬.‫ﺍﻟﺘﺠﺎﺭﺏ‬
‫ﻤﻥ ﺜﻡ ﺘﻡ ﺘﺤﺩﻴﺩ ﺴﻠﻭﻙ ﻜل ﺍﻟﻤﻌﺎﻟﺠﺎﺕ‬. ‫ ﺘﻡ ﺍﺨﺫ ﺍﻟﻘﻴﺎﺴﺎﺕ ﺒﻭﺍﺴﻁﺔ ﻤﻴﺯﺍﻥ ﺩﻗﻴﻕ‬.‫ ﺸﻬﺭﺍ‬18 ‫ﺍﻟﺒﺤﺙ ﺒﺎﻟﻤﺎﺀ ﻭﺘﺠﻔﻴﻔﻬﺎ ﻟﺩﻭﺭﺘﻴﻥ ﺍﺴﺘﻐﺭﻗﺕ ﻤﺩﺓ‬
‫ ﻏﻴﺭ ﻓﻌﺎﻟﺔ ﺇﺫﺍ ﻤﺎ ﺍﺴﺘﺨﺩﻤﺕ ﻋﻠﻰ ﺘﺭﺒﺔ‬،‫ﻤﻌﺎﻟﺠﺔ ﺃﻭ ﻏﻴﺭ ﻤﻌﺎﻟﺠﺔ‬،‫ ﺃﻭﻀﺤﺕ ﺍﻟﺒﺤﻭﺙ ﺍﻟﻤﻌﻤﻠﻴﺔ ﻭﺍﻟﺤﻘﻠﻴﺔ ﺍﻥ ﺍﻻﺴﺎﺴﺎﺕ ﺍﻟﺸﺭﻴﻁﻴﺔ‬.‫ﺍﻟﻤﺴﺘﺨﺩﻤﺔ‬
‫ ﻋﻠﻰ ﻜل ﺤﺎل ﻓﺎﻥ ﺍﻟﺘﺭﺒﺔ ﺍﻟﻤﺘﻤﺎﺴﻜﺔ ﻏﻴﺭ ﺍﻟﻤﺘﻤﺩﺩﺓ ﻭﺍﻟﺤﺠﺭ ﺃﻋﻁﺕ ﻨﺘﺎﺌﺞ ﺠﻴﺩﺓ ﻴﻤﻜﻥ ﺍﻻﻋﺘﻤﺎﺩ ﻋﻠﻴﻬﺎ ﺇﺫﺍ‬.‫ﻋﺎﻟﻴﺔ ﺍﻟﺘﻤﺩﺩ ﻭﺘﻌﺭﻀﺕ ﻟﺒﻠل ﻤﻔﺭﻁ‬
.‫ﻤﺎ ﺘﻡ ﺍﻟﺘﺤﻜﻡ ﻓﻰ ﺍﻟﻤﻴﺎﻩ ﻓﻲ ﻤﻭﻗﻊ ﺍﻟﻤﺒﻨﻰ‬

1. INTRODUCTION montmorillonite. The northern clay plains are

Expansive clay soils cover about one million generally weathered sediments derived from
square kilometers in central, eastern and the volcanic Ethiopian Highlands (2). The
southern Sudan [1]. The area covered by these southern clay soils are thought to be residual.
soils contains nearly the whole agricultural Alluvial soils occur along the White Nile, Blue
region, the majority of the population and Nile and other main watercourses.
represents the potential reserve for the future The data available at Building and Road
agricultural development. The capital of Sudan Research Institute showed the swell potential
and many major cities lie within the expansive of the expansive soils in Sudan to be medium
plains region where the climate is arid to semi- to very high. Plasticity index higher than 100
arid and the soils are mostly in unsaturated have been measured. The expansive soil layer
desiccated condition. usually extends from the ground surface to
The clay plains in central Sudan consist of depths that may exceed 20 metres. Usually
alkaline soils containing the clay mineral brownish or grayish clayey soils measuring
liquid limit greater than 50 and plasticity index

*
Associate Professor, Building and Road Research Institute, University of Khartoum
**
Head of Engineering Administration, Gezira Board, Barakat Wad Medani 49
December 2004 BRRJ Vol. 6
greater than 25 are considered as problematic
soils.

In the past people used to settle on relatively

good land near rivers or on hilly areas. Later,
due to the construction of many development
projects on the expansive clay plains, i.e.,
industrial, commercial and residential
buildings, the expansive soil problem emerged
as a great hazard. Many important structures,
e.g., factories, university buildings and
residential complexes built on expansive soils
suffer from severe distresses and damages.
The loss caused by these soils annually may
exceed millions of dollars.
Fig. (1) Areas (hatched) covered by expansive
Building foundation as normal strip or pad soils in Sudan (After Osman &
footing is very common and widely practiced Charlie)
by homeowners and local contractors. These
types of foundations are used on all types of 2. CONSTRUCTION PRACTICE OF
expansive soils mainly for the followings: FOUNDATION ON EXPANSIVE SOILS
i- Homeowners, local builders and contractors The construction practice in Sudan is to build
lack the proper understanding of the residential buildings on expansive soil from
expansive soil problem and they even lack dry grass, fired bricks or masonry. Grass huts
the knowledge of identification and are the homes of low-income people. Strip
classification of soils. foundation and pad footings are used to
ii-Limited budgets are usually available for support light and relatively heavy structures
construction and the strip foundation option irrespective of the soil type. Short piles have
is thought to be economical. been recently adopted in areas with high to
iii-Other technically viable options such as very high potential swelling mainly in
short piles and stiffened slabs on grade need Khartoum and some important governmental
specialized equipments and contractors and buildings outside Khartoum. Stiffened rafts are
relatively expensive materials such as not in use.
reinforcement steel and cement. The strip foundation practice usually involves
the following:
To minimize the deleterious effects of
expansive soils on structures when the swell i) Excavating 0.5 to 1.0 metre depth and
potential is high local, builders have practiced placing the foundation material, usually
improvement techniques. These involve using brick or masonry, directly on the
intercepting layers of earthen materials expansive soil. The foundation width is
between the natural soils and the building greater than the wall thickness.
foundation material and/or strengthening the ii) The use of improvement techniques that
structure. Most of these techniques were not involve locating an intercepting layer
based on any proper scientific approach. This between the original soil and foundation
is an attempt to study the effect of several material. This layer can be of plain
improvement techniques in reducing structural concrete, layers of sand, or rubble
damage and upheaval. It is part of an on-going masonry “hardcore”. The hardcore
research project on low-cost foundations on constitutes large boulders placed on the
expansive soils at the Building and Road
Research Institute (BRRI). bottom of the excavation, sometimes to
the ground level, and then wall
construction is resumed. The practice in
some areas is to construct the hardcore
foundation before commencement of the
rainy season and leave it exposed to rain
December 2004 BRRJ Vol. 6 50
 during the rainy season. Construction is 3.1 The Laboratory Experiment
then resumed after the rainy season. The laboratory equipment constitutes a rigid steel
iii) Intercepting layers of cohesive non- cylinder 150 mm diameter and about 300 mm
expansive soils (CNS) have been recently high (Fig. 2). The expansive soil was statically
introduced when the swell potential is compacted inside the cylinder in layers. The
high and the option of using piles is not compaction moisture content was fixed at a value
practicable. The idea is borrowed from smaller than the plastic limit of the soil(19%).
India and is advocated to be technically The intercepting layer was placed on the
viable (3). expansive soil and a perforated steel plate was
placed on it. A cylinder 100mm diameter was
Local contractors usually advocate for the placed on the plate and extended outside the
success of their adopted treatment concepts. mould. Surcharge weights were placed on top of
Practice has shown that many buildings the cylinder. The plate and loads simulate a
founded as strip foundation on expansive soils foundation load which was adjusted to be 30
suffer some distresses. The factors that may KN/m2. Provision was made for measurement of
contribute to the distress of buildings founded vertical heave (Fig. 2). The intercepting materials
as strip foundations are: or treatment techniques used were: 2.5 cm
1) Locations of water source, e.g., baths, thickness CNS; 5.0 cm thickness CNS; 7.5 cm
kitchens or gardens and drainage quality thickness CNS; 5.0 cm thickness sand; large
around the buildings; most of problems gravels or hardcore; plate with 20% voids area;
are related to baths been located inside the and expansive soil stabilized with 6% hydrated
built area or improper drainage out the lime (compaction tests were carried out using
building. 2%, 4%, and 6% hydrated-lime with expansive
2) Foundation depth; strips are usually placed soil, and the 6% obtained the higher maximum
at shallow depth within the active zone dry density). The material used to represent the
that extends to 2.5 metre in central Sudan CNS is used as mortar in Wadmedani town in
(4). central Sudan and is classified as clayey silty
3) The type of building whether load bearing sand with 45% passing 0.075mm sieve. The
wall or concrete framed and the plasticity index of the fines is 13%. When the set-
strengthening techniques of the structure, up of the experiment was completed, the cylinder
e.g., introduction of ground tie beams and was submerged in a water tank and heave
tie beams at the window lintel levels. Load measurements were then taken. The test was
bearing structures with high walls are the terminated when the vertical movement was
most vulnerable to damages. about to seize. The tests results are summarized
in Table (2).
3. THE EXPERIMENTAL WORK Table (2) Laboratory Experiments Results
The laboratory and field programme are Intercepting Max. Test Average
described in this section. The field-testing was Layer Heave Period m.c %
% (days) (after test)
carried in Barakat, the headquarter of the
No- treatment 4.66 8 37.67
Gezira Scheme, about 10 kilometers south of Plate with voids 20% 4.36 11 36.83
Wadmedani town. The soil used in the 2.5 cm CNS 3.17 10 35.05
laboratory experiments was brought from the 5.0 cm CNS 2.88 25 37.25
Gravels ( 2”dia.) 2.43 10 40.03
same site. The properties of this soil are Soil +6% lime 2.24 29 32.92
summarized in Table (1). 5.0 cm Sand 1.08 10 38.38
7.5 cm CNS 1.04 16 29.75

3.2 The Field Experiment

Table (1) Properties of the soil used for
laboratory experiments The field experiment was preceded by a soil
investigation for the test area. A borehole was
Depth m LL P.I %passing OMC MDD drilled to 10 metres depth and disturbed and
% % 0.075 mm % gm/cm3
undisturbed soil samples were taken.
1.0 –1.5 75 44 90.0 23.6 1.365 Laboratory tests were carried out for
classification and measurement of soil

December 2004 BRRJ Vol. 6 51

parameters of concern. A typical profile for the corners, for heave measurements. The
the test area is given in Table (3). study area was then flooded for 261 days,
dried for 99 days (first cycle), flooded again
for 127 days, and finally dried for 50days
(second cycle). Heave measurements were
taken by a precise level for a total period of 18
months (two cycles). The following options
were evaluated: no treatment; 25 cm CNS; 50
cm CNS; 75 cm CNS, 50 cm sand; 10 cm
plain concrete; 20 cm reinforced concrete with
20% voids; 50 cm hard core stones; and 50 cm
expansive soil stabilized with 6% of hydrated
lime. The results from field measurements are
summarized in Figure (4).

4. DISCUSSION OF THE RESULTS

A major objective of this research work is to
assess the use of cohesive non-expansive soils
as viable intercepting media between the
expansive soil and the building foundation.
The other objective is to compare the
Fig. (2) Set-up of the laboratory experiment
performance of several types of materials that
Table (3) Profile Description Of Barakat Site can possibly be used as intercepting media.
These mainly include sand, rigid plate “plain
% concrete”, and plain concrete with voids,
Clay
Depth
(m)
Soil
description
LL
%
P.I
%
passing
075mm
content hardcore, expansive soil treated with hydrated
% lime and no treatment. The soil used in the
sieve
0.5 Dark brown 68 36 88.6 34.7 laboratory experiment (Table 1) is classified as
1.0 sandy silty 80 46 93.8 60.1 clay of high plasticity CH and as highly
clay
1.5 74 44 94.8 34.7 expansive soil according to the classification
2.0 71 40 96 32.8 method proposed by Chen (1). Table 2
2.5 56 28 95.8 29.6 summarizes the results obtained from the
3.0 Light brown 45 18 97.4 21.2 laboratory-testing program. The measured
3.5 clayey silt 38 11 91.6 15.1 maximum vertical heave is given for each
4.0 with cal. 38 16 82.8 29.6
treatment option. The test was terminated
4.5 concretions 45 23 77.6 26.7
when the vertical heave almost stopped. As
5.0 Brown sandy 60 35 85.4 31.2
5.5 silty clay 57 34 85.6 33.1
expected, the no-treatment option gave the
6.0 60 33 87.4 26.3
highest values of measured heave in the
6.5 69 40 92.8 46.9
shortest time (4.66% and 8 days). The 7.5 cm
7.0 66 35 92.4 40.2 thick CNS layer measured the lowest heave
7.5 68 37 91 42.4 value (about 1%). The 5.0 cm sand came next
8.0 64 40 88.2 40.5 (1.08%). The plate with 20% voids area gave
8.5 68 41 89.8 41.1 the highest heave (4.36%) compared to the
9.0 62 35 84.4 33.4 other treatment options. The gravels that were
9.5 Light brown 45 25 85.4 36.0 aimed at simulating the hardcore in the field
10.0 clayey silt 48 25 78.6 29.6 gave a percentage heave of 2.43. It is
Typical brick walls were built in the test area. interesting to note that the lime treated
The experimental set-up is shown in Figure 3. expansive soil was good in resisting vertical
The walls were 1.9 meters high and founded at movement (2.24%) and took the longest time
0.60m depth. Various treatment techniques to reach the maximum heave (about a month).
were used, however one treatment is applied to The CNS was effective in resisting vertical
each wall. Trenches were excavated parallel to heave. However thickness has an important
the walls on both sides to allow for even role. The 2.5cm CNS measured 3.17% while
wetting of the foundation soils. Horizontal the 5.0cm CNS measured 2.88%. The 7.5cm
rods were placed at four locations, mainly near CNS considerably reduced the developed
December 2004 BRRJ Vol. 6 52
 of the figure may be based on two basic
trends: one based on the shape of the curves
and the other based on the relative magnitudes
of movements for the different treatments with
time. In terms of shape and magnitude three
regions may be depicted in the first cycle:
a primary swell region (0 to about 74 days), a
secondary swell region (74 to 261 days) and a
shrinkage region (261 to 360 days). Also , two
regions may be depicted in the second cycle, a
swell region (360 to 487 days) and a shrinkage
region (487 to 537 days).
However, based on shape of curves, the
patterns of different treatments are shown
consistent (no intersection between them) in
the primary swell region of the first cycle (0 to
74 days) while inconsistent in the other
regions of the two cycles.
The maximum magnitudes of field heave are
summarized in Table 4. During the primary
swell region of the first cycle, the 75cm CNS
gave the best results (2.59 cm) followed by the
hardcore (2.72 cm) and the 50cm lime treated
expansive soil. The 50cm CNS and concrete
without and with voids followed. The 50cm
sand, 25 cm CNS, and finally the non-treated
expansive soil gave the worst result. During
the secondary swell region, the CNS of
different thicknesses gave the worst results
and the hardcore gave the best result followed
by the 10cm plain concrete. During the
shrinkage region, the 25cm and 50cm CNS
gave the maximum shrinkage values followed
by the hardcore and the 75cm CNS. The
arrangement of treatments, according to heave
amount, in the swell region of the second
cycle seems to be as same as that of the
secondary swell region of the first cycle.
Fig. (3) View of part of the field experiment During the shrinkage region of the second
cycle, the concrete without and with voids
gave the maximum values of shrinkage and
the lime treated expansive soil gave the lowest
heave. This emphasizes the importance of value followed by the 75 cm CNS.
layer thickness for the CNS option.
Figure 4 shows the results from the field
experiment. The results cover a test period of
18 months (two cycles). The average vertical
heave was plotted against time. The discussion

December 2004 BRRJ Vol. 6 53

 7 shrinkage
secondary swell swell Shrinkage
primary swell region region region region reg.
6

5
HEAVE IN CMS

3 Drying Drying
wetting period
wetting period period
2 period
First Cycle Second Cycle
1

0
5 8 23 38 56 74 84 121 154 183 196 217 232 261 299 327 360 393 432 453 487 520 537
DAYS

A1 A2 A3 B1 B2 B3 C1 C2 C3

A1: 10 cm plain conc.; A2: 20 cm r.c. with voids; A3: hardcore; B1: 25cm CNS; B2: 50cm CNS; B3: 75cm
CNS; C1: no-treatment; C2: soil+6% hydrated lime; C3: 50cm sand

Fig. (4) Summary of field maximum heave measurements

.

Table (4) Summary of maximum heave lying layers. Therefore only the results from the
measurements of field test (cm) primary region express the interaction between
the expansive soil and the treatment layers.
First Cycle Second Cycle
Wetting Period Drying Rewetting Redrying Excessive long-term wetting caused all the
Treatment period Period Period options to heave, substantially.
Primary Secondary Shrinkage Swell Shrinkage
Swell Swell Region Region Region The field experiment showed that the strip
Region Region foundation option is not viable when placed on
75cm CNS 2.59 2.72 3.31 2.42 0.33
50cm 2.72 1.43 3.47 1.38 0.85 expansive soil of high potential and subjected to
Hardcore
Soil+6% 2.89 2.55 2.46 2.14 0.15
continuous long-term wetting. However some
lime(50cm) treatment techniques may be effective in case of
10cm plain 2.93 1.64 2.57 1.52 1.05
conc
short term limited wetting of the foundation soil.
50cm CNS 3.12 2.82 3.89 3.07 0.36 A comparison between the laboratory and field
20cm 3.22 2.03 3.06 1.93 1.03
r. conc
results (primary region of first cycle) shows
with 20% about consistent trend. The thick CNS and
voids
50cm sand 3.36 2.68 2.81 2.2 0.44
hardcore can be considered as the best options
25cm CNS 3.42 2.82 3.69 3.32 0.48 followed by the lime treated expansive soil.
No- 3.47 2.67 2.86 2.77 0.56
treatment However both experiments emphasize the
importance of layer thickness when CNS is
The authors believe that the real interaction used. The hardcore that is widely used in
between the expansive soil and foundation took different areas in Sudan gave sound results. The
place during the primary swell of the first cycle. sand layer gave acceptable results in the
The movement after that , i.e., during the laboratory but poor results in the field. Sands
secondary swell, is caused by the heave of under- usually allow easy penetration of water and
therefore their use is doubtful.
December 2004 BRRJ Vol. 6 1
 4. CONCLUSION [3] Elsharief A. M. (1987). “Field and
Laboratory Experiments on Expansive
This paper reviewed the construction practices Soil Heave and the Behavior of Short Piles
of foundations on expansive soils in Sudan. It in Expansive Soils.” M.Sc. Thesis, BRRI,
also presented the results of laboratory and University of Khartoum.
field experiments aimed at evaluating several
improvement techniques for strip foundation [4] Katti, R. K. (1979). “Search for Solutions to
Problems in Black Cotton Soils.” First
on expansive soils with high potential for Annual Lecture, Indian Geotechnical
swelling. The improvement techniques Society Journal, Vol. 9, No. 1, pp. 1-80
included the use of layers of several types of
earthen materials as intercepting media
between the soil and foundation material. The
objective was to reduce vertical heave. The
media used were: cohesive non-expansive soil
CNS, sand, concrete with and without voids,
hardcore and expansive soil stabilized with
lime. The effect of CNS thickness was also
evaluated. The following conclusions are
drawn from the experiments:

i. The strip foundation option may not be

technically viable on soils of high potential for
swelling when the foundation is subjected to
excessive wetting.

ii. The 750mm CNS and hardcore are effective

intercepting media when limited localized
wetting is experienced.

iii. Wherever the CNS option is considered

adequate thickness should be used. The
thickness should not be less than 750mm.

iv. The use of plain concrete with or without voids

and sand should be avoided.

v. Excessive wetting leads to high values of

shrinkage of expansive soils (first cycle of
Fig.(4))

5. REFERENCES
[1] Chen F. H. (1975). “Foundations on
Expansive Soils.” Elsevier-Amsterdam ,
the -the Netherlands.

[2] Osman M. A. & Charlie W. A. (1983).

“Expansive Soils in Sudan.” BRRI current
Paper no. CP. 3/83, Building and Road
Research Institute, University of
Khartoum, Khartoum, Sudan

December 2004 BRRJ Vol. 6 2