954
Proceedings of the 18
th
International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013
Proceedings of the 18
th
International Conference on Soil Mechanics and Geotechnical Engineering, Paris 2013
Seasonal fluctuations such as floods or heavy rainfalls can
raise the water table up to or beyond the footing level and
produce additional settlements of shallow foundations. The soil
loses its stiffness when submerged, and settles more. Substantial
additional settlement may occur when the groundwater level
changes, which can exceed the tolerable limit for settlement and
threaten the integrity of structure. Very few works have been
found in the literature investigating the influence of fluctuating
water level on shallow foundation settlements. Some
researchers suggested using a water table correction factor,
which can be used as a multiplier on the settlements predicted
for footings resting on dry sands, to get the settlements in
submerged condition. Limited laboratory model tests have been
conducted in the past, which did not cover the effect of
foundation shape or varying stress level on additional settlement
induced by water table rise.
In this paper, the authors have described a comprehensive
laboratory test program carried out to quantify the additional
settlement due to rise in water table with varying footing shape,
soil density, water table depth and stress level. This was
followed by modeling the experimental set up in geotechnical
modeling software FLAC, and the results were compared with
the experimental data.
2 WATER TABLE RISE AND CORRECTION FACTOR
Terzaghi (1943) made an intuitive suggestion that when dry
sand becomes saturated, the soil stiffness (Young’s modulus)
reduces by approximately 50%. He noted that, the effective
vertical stress on soil under the water table reduces roughly to
half; which reduces the effective confining stress by 50%. This
leads to loss of stiffness of saturated soil to half of that in the
dry condition. As a result, settlement in soil below the water
table gets doubled.
When the water table rises to some depth below the footing,
a correction factor for the new location of water table is used in
the design of shallow foundations. The settlement under dry
conditions is multiplied by this factor, to give the settlement
expected due to the water table rise. The correction factor
C
w
is
greater than or equal to 1 and increases with rise in water table.
It is defined as:
C
w
=
dry sand
in
settlement
level
footing
the
below
table
water
with
settlement
)1(
Various researchers (Terzaghi and Peck 1948, Teng 1962,
Alpan 1964, Bazaraa 1967, Peck 1974, Bowles 1977) proposed
correction factors to quantify the additional settlement due to
the water table rise below the footing. The depth below the
footing where the water table fluctuation will not have any
effect is not unanimously agreed upon. The depth of embedment
of the footing also affects the influence of water table on
settlement, as the surcharge due to embedment increases the
settlement in raised groundwater level. Throughout this paper,
the correction factor for water table, foundation width, depth of
water table below the foundation and the depth of embedment
are denoted by
C
w
,
B
,
D
w
and
D
f
, respectively, as illustrated in
Figure 1.
Shahriar et al. (2012) made a critical review of the current
state-of-the-art for predicting shallow foundation settlement due
to rise in water table in granular soil. Theoretical studies by
Vargas (1961), Brinch Hansen (1966) and Bazaraa (1967)
suggested a maximum correction factor of 1.7, when the water
table rises to the base of the foundation. Limited field
investigations suggest that submergence of granular soil doubles
the settlement when compared to dry condition, agreeing with
Terzaghi’s proposition. Numerical modeling conducted by
Shahriar et al. (2012) shows that the settlement gets doubled in
submerged sand if linear elastic model is used, but the use of
hyperbolic non-linear elastic soil model gives higher additional
settlements at high stress levels.
Ground Surface
D
f
Footing
Level
B
D
w
Water Table
Figure 1. Schematic diagram of a shallow foundation.
Very little laboratory studies have been conducted so far
and contradictory results have been found. Agarwal and Rana
(1987) conducted tests on square footings of three different
sizes. Their results support Terzaghi’s proposition that the
settlement gets doubled when the sand gets submerged. Murtaza
et al. (1995) also used three different sized square footings and
conducted the tests with loose, medium dense and dense sands.
The results showed 8 to 12 times more settlement in submerged
condition. Morgan et al. (2010) carried out settlement tests with
a square footing in two different types of soils and found that
the increase in settlement in submerged sand can be 5.3 times
the dry sand. However, these experimental programs were small
in scale and none of these considered the effect of varying
footing shape and stress level.
3 LABORATORY MODEL STUDY
A Perspex rectangular tank 800 mm x 800 mm in plan and 600
mm high was built to carry out the settlement test. Various
footing shapes were used. A circular footing of 100 mm
diameter and square and rectangular footings with
B/L
=1.0,
0.75, 0.50, 0.25 were used where the width,
B
was fixed to 100
mm in each case. A locally available granular soil was used. In
a model footing having smaller dimensions, the settlement
might get affected by change in soil stiffness in a partially
saturated area. From laboratory testing, it was observed that the
capillary rise is higher in well graded soil. Hence, the finer
particles were sieved out from the test soil to get a uniformly
graded soil with soil grains large enough to significantly reduce
the capillary height. The rate of capillary rise of the sieved soil
was then tested using soil filled Perspex tubes protruding from
water. At five minutes, the capillary height observed were 40
mm and 53 mm in loose and dense sands respectively. Five
minutes was the maximum time to get the water level static
during the settlement tests, so the capillary rise is expected to be
limited within the range of 40-53 mm. In fact, the height of
capillary rise was limited to 50 mm for most of the time during
the tests. This height is reasonable when compared to the
footing width (100 mm). In case of granular soil, the elastic
modulus of the soil is a key parameter in predicting foundation
settlement, and Vanapalli and Mohamed (2007) showed that the
elastic modulus of unsaturated soil can be significantly
influenced by matric suction. However, by limiting the capillary
rise within a shorter range, the unsaturated zones in the model
tests were kept quite small and hence, their effect on the overall
settlement was negligible. The soil properties of sieved out sand
are: effective size
D
10
=0.67 mm, co-efficient of uniformity,
C
u
=1.64, co-efficient of curvature
C
c
=0.89, specific gravity,
G
s
=2.61, maximum and minimum dry densities =1.53 t/m
3
and
1.382 t/m
3
respectively. Two different relative densities (37.6%
and 77.4%) of the sand were used. Since the model tests
represent the larger footings with higher densities in the field,
maximum relative density was limited to 77.4%.
