747

Experimental Observations and Parametric Study

of Piled Raft Foundation Using Plaxis
N. Venkateshwaran

Faculty,
National Institute of Technology,
Tiruchirapalli, 620015, India
Email: [email protected]

Dr.S.P. Jeyapriya
Assistant Professor,
Government College of Technology,
Coimbatore 641013, India
Email: [email protected]

AbstractThe foundations for tall buildings are usually
subjected to a combination of vertical, lateral and
overturning forces. Piled raft foundation is an
economical and efficient foundation for tall buildings in
which the load from the superstructure is transferred to
the soil by load sharing mechanism between raft and pile
and thereby reducing the settlement. In the present
study, laboratory model tests and numerical analysis
were carried out to study the behaviour of piled raft
foundation subjected to vertical load. The test was
conducted on plain raft and piled raft model with
different configurations viz. 1x1, 2x2 and 3x3 in
homogeneous loose sand. The experimental results were
analysed and calibrated using PLAXIS 2D. The
percentage settlement reduction observed experimentally
and numerically were compared and reported. It was
observed that the settlement has reduced considerably
for piled raft foundation and it was about 71% reduction
for 3x3 configuration. Also, the load carrying capacity
has increased significantly from 234N to 578N when 3x3
pile configuration was used which accounts about 146%
increase .

Keywords-component; piled raft modellng;
experimental program; PLAXIS.

I. INTRODUCTION
One of the most important aspect of Civil
Engineering projects is the foundation system.
Designing a foundation system carefully and properly,
will not only lead to a safe and efficient structure, but
also an overall economy of the project. Until recently,
there were some individual systems such as shallow
foundations like rafts and deep foundations like piles.
By combining these two systems, foundation
engineers will provide necessary values for the design,
obtain the required safety and also come out with more
economical solutions. A piled raft is a composite
foundation in which the piles are used as settlement
reducers and they share with the raft, the load from the
superstructure [5],[8]. In the design of piled rafts, the
load is shared between the piles and raft. The piles in
piled raft are used up to a load level that can be of the
same or even greater magnitude as that of a
comparable single pile.
Therefore, the piled raft foundation allows
reduction of settlements in a very economic way as
compared to the traditional foundation concepts. The
capacity of piles is not to be downgraded for negative
skin friction [6],[7]. The piles are then interconnected
with a rigid system of raft to ensure uniform
settlement. In this research work, a small scale
laboratory model test was conducted on plain raft and
piled raft in homogeneous loose sand condition. The
properties of river sand, which is used as the
foundation medium was analysed and reported. The
laboratory model test was carried out with a
rectangular tank of size 450mm x 300mm x 450mm
and modelled plain raft of size 70mm x 70mm and
thickness 6mm [2]. In piled raft, the raft is of same
size and piles of 8mm diameter and length 180mm
with a spacing of 3.5 times the diameter of pile was
used. The configuration of piles used in this study was
1x1, 2x2 and 3x3. Results showed that the piled raft
foundation in sand considerably reduces the
settlement compared to that of plain raft foundation.
Methods of Analysis
Numerous methods were available to
establish the interaction of piled raft combination. This
include simplified calculation methods, approximate
computer-based methods and more rigorous computer
based methods. The simplified calculation methods are
empirical methods with numerous simplifications,
mostly concerning the soil profile and the load
condition. In approximate computer-based methods,
the raft is modelled as a strip and the soil and the piles
are modelled with springs of varying stiffness. The
more rigorous computer-based methods consist of
different numerical methods and this research work is
mainly focused on the numerical analysis using
PLAXIS 2D software.

Plane strain FEM-model for piled rafts
The main problem when modelling a piled
raft with a plane strain model is the transition from
three to two dimensions. The wall element is defined
per meter in the plane strain analysis using PLAXIS,
the raft and piles were modelled as series of beam
elements with the appropriate geometrical parameters
and geometrical boundaries.

II. PARAMETRIC STUDY
Parametric study of piled raft refers to
studying the behaviour of the structure by changing
their dimensions [1], [4]. The details of model are
Case 1: Plain raft 70mm 70mm square and of 6mm
thick.
Case 2: Piled raft of 70mm 70mm with raft
thickness of 6mm and pile length of 180mm. The pile
configuration is 1x1 and is named as PR 1x1.
Case 3: Piled raft of 70mm 70mm with raft
thickness of 6mm and pile length of 180mm. The pile
configuration is 2x2 and is named as PR 2x2.
Case 4: Piled raft of 70mm 70mm with raft
thickness of 6mm and pile length of 180mm. The pile
configuration is 3x3 and is named as PR 3x3.
III. LABORATORY INVESTIGATION
Laboratory tests were conducted on the
materials used in the experimental study

748
A. Properties of Sand
All the laboratory tests were performed as per
IS specifications. The sand was tested to determine the
following preliminary properties and the results are
tabulated in Table 1.
1. Gradation of sand
2. Specific Gravity
3. Unit weight
4. Maximum and minimum dry densities
5. Angle of internal friction
6. Maximum and minimum void ratio.
Unit Weight
Dry unit weight of sand was determined by
vibrating table method conforming to IS 2720-Part 14.
To determine the unit weight, several trials have been
carried out for varying heights of fall. It was found
that as the height of fall of sand increases, the density
of sand also increases. To verify this, a cylindrical
calibrating container conforming to IS 2720-Part 28
was used to pour the sand with the help of a hopper.
The cylinder was filled with sand for different heights
of fall of 0cm, 10cm, 20cm, 25cm, 30cm, 35cm,
40cm, 50cm, 60cm, 70cm, 80cm, 90cm and 100cm.
The height of fall of 25cm, 30cm and 35cm was fixed
for filling tank to achieve loose sand condition
corresponding to 20.55%, 26.63% and 32.97% relative
density respectively.
Void ratio and Relative density
To determine the void ratio and relative
density of sand for varying heights of fall,
corresponding value of unit weights obtained from
previous study was used. The value of relative density
varied from 0% to 100% for a void ratio of 0.72 to
0.47 respectively. The results of unit weight, void ratio
and relative density corresponding to various heights
of fall are shown in Table 2. The relationship of void
ratio vs height of fall,, relative density vs height of
fall, unit weight vs height of fall are illustrated in
figures 1 to 3 respectively.
Modulus of Elasticity (E)
Shear strength parameters of soil was
determined from Direct Shear test. The Modulus of
Elasticity was indirectly determined from shear
modulus and the value was found to be 10.24 MN/m
2

for the relative density of 26.63%.





























TABLE 1 PROPERTIES OF SAND

Sl.
No.
Description Values
1 Specific gravity 2.65
2 Coarse sand 4.6
3 Medium sand 71.7
4 Fine sand 23.7
5 Uniformity coefficient 2.55
6 Coefficient of curvature 0.76
7 Classification of sand
Poorly graded
sand (SP)
8 Maximum dry unit weight 17.63 kN/m
3

9 Minimum dry unit weight 15.10 kN/m
3

10 Maximum void ratio 0.72
11 Minimum void ratio 0.47
12 Unit weight 15.71 kN/m
3

13 Relative density 26.63%
14 Angle of internal friction 30
15 Modulus of Elasticity 10.24 MN/m
2

TABLE 2 HEIGHT OF FALL METHOD

Sl.
No.
Height
of fall
(cm)
Unit
weight
(kN/m
3
)
Relative
density
(%)
Void
ratio
Angle
of
intern
al
frictio
n ()
1 0 15.10 0 0.72 24
2 10 15.18 3.67 0.71 24
3 20 15.42 14.46 0.69 27
4 25 15.57 20.55 0.67 28
5 30 15.70 26.63 0.66 30
6 35 15.85 32.97 0.66 31
7 40 16.00 39.20 0.62 32
8 50 16.40 55.23 0.59 34
9 60 16.78 69.77 0.55 37
10 70 17.10 81.52 0.52 38
11 80 17.28 87.91 0.50 39
12 90 17.34 90.00 0.49 40
13 100 17.38 91.41 0.49 40



Fig 1. Relationship of void ratio and height of fall



Fig 2. Relationship of relative density and height of fall

749

Fig 3. Relationship of Unit weight and height of fall

IV. EXPERIMENTAL PROGRAM AND
NUMERICAL ANALYSIS
Mild steel rods were used as model piles
having diameter 8 mm and length 180 mm, forming
configurations 1x1, 2x2, 3x3 and the piled raft model
used in the study is shown in figures 4 to 6. The piles
are rigidly fixed to the raft by welding. The property
of mild steel is shown in Table 3.

TABLE 3 PROPERTIES OF MILD STEEL SPECIMEN

Sl.No. Properties Values
1 Yield stress (N/mm
2
) 416.74
2 Ultimate stress (N/mm
2
) 436.59
3 Modulus of Elasticity (N/mm
2
) 2.34 x10
5

Fig.4 Piled raft model of configuration 1x1
Fig.5 Piled raft model of configuration 2x2




Fig.6 Piled raft model of configuration 3x3

A. Preparation of foundation medium
The sand was placed in the tank with
predetermined relative density of 26.63%, by
maintaining a dropping height of 30cm upto top level
or by taking the weight of sand equal to 21.60 kg that
is poured to the tank at every 100cm height until the
tank is filled upto the surface that corresponds to a
relative density of 26.63%. After each test, the sand
box should be emptied to a depth below the zone of
influence (which was considered as L below the tip of
pile, where L is the pile length). During the process of
sand raining, the piled raft model was placed at the
center of the tank and under the loading ring, a bubble
balance was used to ensure the level of the raft, then
the sand raining was continued upto the top slightly
lower than the raft. The final layer of the sand is
levelled by a sharp edge ruler.

B. Loading Setup
The load frame consists of proving ring of
2kN capacity with 0.0025kN accuracy for load
determination and two numbers of 25mm dial gauge
with 0.01mm sensitivity for settlement prediction. The
entire loading setup is shown in figure 7.
Fig.7 Loading setup of piled raft

750
C .Application of vertical load and calibration in
PLAXIS
The vertical load was applied at a constant
loading rate of 1mm/min throughout the entire testing
program. The settlement of the system was measured
using dial gauges for each load increment until the
settlement is less than 0.02mm/min. The load vs
settlement graph was plotted for each configuration of
piled raft system. The obtained curves are calibrated in
PLAXIS 2D and parametric study for these
configurations were analysed. The ultimate load is
determined by considering 10 % width of the footing.
Figures 8 to 10 show the parametric study using
PLAXIS 2D.


Fig.8 PR 1x1 in PLAXIS 2D

Fig.9 PR 2x2 in PLAXIS 2D


Fig.10 PR 3x3 in PLAXIS 2D





V RESULTS AND DISCUSSIONS
From the experimental study and FEM
analysis, it was observed that for plain raft and piled
raft (PR 1x1, PR 2x2 and PR 3x3), the load settlement
variation between the experimental study and PLAXIS
2D is within 15% upto the ultimate load and also the
FEM analysis underestimates the settlement and
overestimates the ultimate load when compared to that
of experimental study. Figures 11 to 14 show the
comparison between experimental study and PLAXIS
2D. In all these graphs, it was found that the numerical
analysis showed slightly higher value of load than the
experimental work. Figure 15 shows the comparison
between load vs settlement for plain raft and piled raft
with different configurations namely PR 1x1, PR 2x2
and PR 3x3. As the number of piles is increased, the
load carrying capacity of piled raft system is increased
significantly which can be observed from Table 4 and
Figure 16. Figure 17 shows the settlement reduction
with pile raft combination as reported in Table 5 for
varying area ratio (ratio between the cross sectional
area of pile and area of raft).



Fig.11 Comparison of Plain raft in experimental and PLAXIS study


Fig.12 Comparison of PR 1x1 in experimental and PLAXIS study

751


Fig.13 Comparison of PR 2x2 in experimental and PLAXIS study



Fig.14 Comparison of PR 3x3 in experimental and PLAXIS study



Fig.15 Load vs Settlement of plain raft and piled raft
(From PLAXIS study)

TABLE 4 PERCENTAGE INCREASE IN LOAD CARRYING
CAPACITY BY CONSIDERING 10% WIDTH
Sl.No.
Footing
model
Area
ratio

Ultimate
load
(N)
% increase
1 Plain raft 234 ---
2
Piled raft
1x1
0.010 390 67
3
Piled raft
2x2
0.041 505 116
4
Piled raft
3x3
0.092 578 147



Fig.16 Percentage increase in load carrying capacity

TABLE 5 PERCENTAGE REDUCTION IN SETTLEMENT
Sl.No.
Footing
model
Area
ratio
Settlement
(mm)
%
Reduction
1 Plain raft 7.00 ---
2
Piled raft
1x1
0.010 3.19 54
3
Piled raft
2x2
0.041 2.40 66
4
Piled raft
3x3
0.092 2.02 71



Fig.17 Percentage settlement reduction vs Area ratio

The load carrying capacity with the association of
piles with raft has increased considerably and further
the settlement of the footing had considerably reduced.
The ultimate load increased from 234N for plain raft
to 578N for piled raft of 3x3 configuration. Similarly,
the settlement reduces from 7mm for plain raft to
2.02mm for the same combination. The reason for
increase in settlement reduction and percentage load
carrying capacity is due to the raft soil interaction and
pile soil interaction. The increase in normal stress of
subsoil is due to the above mentioned soil structure
interactions along with pile- raft interaction and pile
pile interaction.

752
VI CONCLUSIONS

Laboratory study with plain raft and piled raft
on loose sand foundation medium was conducted
and the following observations were made.
(1) The failure load has increased from 234N for
plain raft to 390N, 505N and 578N for piled raft
of 1x1, 2x2 and 3x3 configurations respectively.
This account to about 67%, 116% and 146%
increase in the carrying capacity for piled raft of
1x1, 2x2 and 3x3 respectively.
(2) The settlement of plain raft is 7mm and it has
reduced to 3.19mm, 2.40mm and 2.02mm when
piled raft of 1x1, 2x2 and 3x3 configuration is
used. This is about 54%, 66% and 71% for the
respective combination.
(3) With the increase in area ratio, the % reduction in
settlement initially increase rapidly and and
thereafter attains a marginal value and also the %
increase in load carrying capacity increases
rapidly at first and then it maintains a gradual
increase .
(4) Analysis using PLAXIS 2D compares well with
the experimental values and hence it can be used
for studies on piled raft system of other
configurations.

















































VII REFERENCES

1. MeisamRabiei, 2009, Parametric Study for
Piled Raft Foundations, Electronic Journal
of Geotechnical Engineering ,Vol.14.
2. S.P.Bajad, R. B. Sahu, 2012, An
Experimental Investigation on Interference of
Piled Rafts, Civil and Environmental
Research, Vol.2, No.2, pp 49-58.
3. Rolf Katzenbach, Christian Gutberlet, Gregor
Bachmann, 2007, Soil-Structure Interaction
aspects for ultimate limit state design of
complex foundations, First International
Symposium on Geotechnical Safety & Risk,
Shanghai, Tongji University, China, pp 585-
596.
4. M. H. Baziar, A. Ghorbani , R. Katzenbach,
2009, Small-Scale Model Test and Three-
Dimensional Analysis of Pile-Raft
Foundation on Medium-Dense Sand,
International Journal of Civil Engineerng,
Vol.7, No.3, pp 170-175.
5. Poulos.H.G., 2001,Piled Raft Foundation
Design and Applications, Geotecnique,
Vol.51, No.2, pp 95-113.
6. Young-Kyo Seo and Kyung-sik Choi, Sung-
Gyo Jeong, 2003, Design Charts of Piled
Raft Foundations on Soft Clay, Proceedings
of The Thirteenth International Offshore and
Polar Engineering Conference, Honolulu,
Hawaii, USA, May 2530, pp 753-755.
7. K. Yamashita, T. Yamada, 2009, Settlement
and load sharing of a piled raft with ground
improvement on soft ground, Proceedings of
the 17
th
International Conference on Soil
Mechanics and Geotechnical Engineering, pp
1236-1239.
8. H.G. Poulos, J.C. Small, H. Chow, 2009,
Piled Raft Foundations for Tall Buildings,
Geotechnical Engineering Journal of the
SEAGS & AGSSEA, Vol. 42, No.2, pp 78-
84.

AUTHORS PROFILE
1. N. Venkateshwaran, M.E., Geotech

Faculty,
National Institute of Technology,
Thiruchirapalli, 6201015, India
Email: [email protected]

2. Dr.S.P. Jeyapriya
Assistant Professor,
Government College of Technology,
Coimbatore 641013, India
Email: [email protected]

753