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

net/publication/319418128

Determination of Bearing Capacity of Stone Column with Application of

Neuro-fuzzy System

Article  in  KSCE Journal of Civil Engineering · August 2017

DOI: 10.1007/s12205-017-1497-6

CITATIONS READS

0 56

2 authors:

Manita Das Ashim Kanti Dey

National Institute of Technology, Silchar National Institute of Technology, Silchar
5 PUBLICATIONS   0 CITATIONS    13 PUBLICATIONS   56 CITATIONS   

SEE PROFILE SEE PROFILE

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

PhD project View project

All content following this page was uploaded by Manita Das on 19 August 2018.

The user has requested enhancement of the downloaded file.

KSCE Journal of Civil Engineering (0000) 00(0):1-7 Geotechnical Engineering
Copyright ⓒ2017 Korean Society of Civil Engineers
DOI 10.1007/s12205-017-1497-6 pISSN 1226-7988, eISSN 1976-3808
www.springer.com/12205
TECHNICAL NOTE

Determination of Bearing Capacity of Stone Column with Application

of Neuro-fuzzy System
Manita Das* and Ashim Kanti Dey**
Received August 31, 2016/Revised May 13, 2017/Accepted June 26, 2017/Published Online August 31, 2017

··································································································································································································································

Abstract
The neuro-fuzzy controller applies the neural network learning techniques to tune the membership functions and keeps the
semantics of the fuzzy logic controller intact. Hence benefits of both the neural network and fuzzy logic controller are taken into
consideration. In this study, to predict the bearing capacity of a stone column, application of Adaptive Neuro-fuzzy Inference System
(ANFIS) is presented. To train and test the data sets, 105 data pairs are collected from the previous technical literature. These data sets
include the data of stone and sand columns. The spacing of the columns varies from 1.5 to 10 times the diameter. The undrained
cohesion varies from 7 to 400 kPa. Both experimental and analytical data are included in the collection. To test the trained ANFIS
models, data are collected from physical experiments on plate load test and numerical analysis with PLAXIS-2D. For the
comparative study, ANFIS models combined with plate load test results and analytical results, three ANFIS models are developed. A
comparative study on the accuracy of prediction by these three models is discussed.
Keywords: bearing capacity, stone columns, Neuro-fuzzy, adaptive neuro-fuzzy inference system
··································································································································································································································

1. Introduction observational errors, errors due to installation technique, weather

variations and in-situ conditions, etc. may occur. Therefore, these
Soft soils are dominating most of the coastal plain areas all over uncertainties are to be considered to obtain an accurate behavior
the world. Due to the low strength and high compressibility of the of the stone columns. Hence, some researchers prefer analytical
soft soil, the structure constructed over the soft soil may fail. studies to laboratory studies. Some computing techniques like
In case, the strength of the soft soil is increased, the bearing neural network and fuzzy control systems are being widely used
capacity will increase and extent of settlement will decrease. One in geotechnical engineering. Some researchers (Chika et al., 2014;
way to improve the strength of the soft soil is to install stone Chikb et al., 2014) used ANN to study the behavior of the stone
columns, which increase the bearing capacity of the soft soil, columns particularly the settlement. In the present work, Adaptive
reduce the settlement and also accelerate the rate of consolidation. Neuro-Fuzzy Inference System (ANFIS) is used to observe the
Hughes and Withers (1974) and Balaam and Booker (1981) used behavior of the stone columns. Three ANFIS models such as (a)
the unit cell concept to predict the bearing capacity of the single ANFIS-E with only experimental data, (b) ANFIS-A, with either
stone column. Many researchers (Madhav and Vitkar, 1978; analytical or numerical data, and (c) ANFIS-EA, with both the
Etezad et al., 2015; Afshar and Ghazavi, 2012) already developed experimental and analytical or numerical results, are developed.
some theories to find the bearing capacity of the stone columns. All these data are used for training. For the testing and validation
Bureau of Indian Standards has published one code on stone of these models, 10 experiments on a plate load test and 12
column i.e. IS 15284-(part-1), 2003. In this IS code, all the model studies with PLAXIS-2D are conducted. The diameter,
properties of a stone column like bearing capacity, settlement and spacing, length of the stone columns and friction angle of the
consolidation are available. Normally, improvement of strength stone materials are varied. Out of the three ANFIS models, the
of the soft soil reinforced with the stone columns mainly depends ANFIS-EA model shows the best result.
on the strength of the surrounding soft soil. Reduction of the
strength of surrounding soil reduces the extent of improvement due 2. Bearing Capacity of Stone Column and Plate
to a low confinement (Shivashankar et al., 2011). Some researchers Load Tests
(Hu, 1995; Ambily and Gandhi, 2004; Shahu and Reddy, 2011)
conducted experimental studies on the behavior of the stone 2.1 Bearing Capacity of a Stone Column
columns. In an experimental study, many uncertainties due to In the present study, the bearing capacity of a stone columns

*Research Scholar, Civil Engineering Dept., National Institute of Technology Silchar, Silchar Assam-788010, India (E-mail: [email protected])
**Professor, Civil Engineering Dept., National Institute of Technology Silchar, Assam-788010, India (Corresponding Author, E-mail: [email protected])

−1−
 Manita Das and Ashim Kanti Dey

Fig. 2. Plan of Seven Stone Columns

Fig. 1. Experimental Setup

was determined through model tests in a steel tank of size 1 m ×

1 m × 1 m (inside). Fig. 1 shows the test setup used for this study.
The inside portion of the tank was wrapped with a polyethene
sheet to overcome the friction between the soil and the tank. The
clayey soil collected from the nearby paddy fields was pulverized,
freed from foreign matter and grass, thoroughly mixed with Fig. 3. Variation of Friction Angle with a Density of Stone Column
water keeping the water content near to the liquid limit, dumped Material
in the tank layer by layer. The thickness of the soil layer was kept
as 0.9 m. The soil was kept idle for seven days to attain a uniform
consistency. The average undrained cohesion of three samples It was intended to maintain a constant cu value for all the tests.
collected from three depths after the load test showed a value Soon after completion of the tests, block samples were collected
around 25 kPa. For the construction of a stone column, a borehole from the tank and were tested for undrained cohesion. It was
was dug using an auger of diameter 50 mm. The stability of the observed that the value varied in the range 22-21.5 kPa.
hole was checked with the insertion of a hollow tube of 50 mm Similarly, it was tried to maintain a constant f value for the stone
diameter and length 400 mm. In most of the cases the tube column. In reality, it was observed that total quantity of stone
entered freely, otherwise, the auger was inserted again to clear aggregates for filling the hole was not same. Hence direct shear
the hole. Once the stability of the hole was ensured, stone chips tests were conducted for different densities of the stone
of size 10-12 mm were poured into the hole in three layers. The aggregates. Fig. 3 shows a variation of f values against density.
weight of each layer was pre-calculated considering 20% voids. This figure was used during the analysis of results.
Each layer was compacted with a circular steel tamper of diameter The double tangent method as prescribed in IS 15284-I (2003)
25 mm and weight 2 kg. The compaction effort was kept light to was used to obtain the ultimate bearing capacity of the stone
avoid lateral bulging as suggested by Ghazavi, M. and Afshar columns. The method consists of drawing two tangents on the
(2013). Total seven stone columns were constructed in a triangular load-settlement curve, one at the initial portion and the other at
pattern to obtain the behavior of the central column (IS-15284 the rear portion of the curve, the intersection point gives the
PART-1, 2003). The plan view of seven stone columns is presented ultimate bearing capacity. The tangents are drawn as shown in
in Fig. 2. One footing of diameter 10 cm was placed over the
central stone column. The vertical load was applied through a load
cell of capacity 5 ton using a hydraulic jack and settlements were
measured through two LVDTs of sensitivity 0.01 mm placed
diagonally opposite ends. Two dial gauges were also placed along
the other diagonal. One data acquisition system was connected to
the load cell and the LVDTs.
The bearing capacity (qu) mainly depends upon the spacing
between the stone columns (s), the diameter of the stone column
(d), friction angle of stone column materials (f) and undrained
cohesion of clay (cu). Table 1 shows the bearing capacity of stone
columns by varying these parameters. Fig. 4. Load Versus Settlement Graph of Group Stone Column

−2− KSCE Journal of Civil Engineering

 Determination of Bearing Capacity of Stone Column with Application of Neuro-fuzzy System

Table 1. Experimental Test Results

cu (kPa) s/d ratio φ (degree) L (mm) qu (kPa)
22 1.5 40 400 210
21.5 2 40 400 190
21.3 2.5 40 400 165
21.5 3 40 400 148
21.2 3.5 37.27 400 120
21.5 1.5 37.27 500 145
21.5 2 37.27 500 134
21.5 2.5 35 500 128
21.5 3 35 500 118
21.5 3.5 35 500 105

Fig. 4 and the ultimate bearing capacity for this test was obtained
as 118 kPa.
Fig. 5. Plot of Deformed Mesh (extreme total displacement)
2.2 Plaxis2D Analysis
PLAXIS 2D is a finite element package which is used for two-
dimensional analysis of the deformation and stability in geotechnical Table 2. Properties of Soils Used for Plaxis-2d Model
engineering. In this study, Plaxis 2D was used to find out the Materials Stones Soft soil
bearing capacity of the stone columns of diameter 50 mm, 60 Unsaturated γ in kN/m3 19 15
mm and 70 mm and length of 400 mm, 450 mm, 500 mm, 600 Saturated γ in kN/m3 20 16
mm and, 650 mm. In the present study, the strength characteristic Poissons ratio, ν 0.3 0.35
of the soil was studied using Mohr-Coulomb failure criteria. By Elastic modulus, E in kPa 1200 30000
Permeability in horizontal
considering the drained condition, a medium size finite element 1.24 × 10−4 2.9 × 10−9
direction in m/s
mesh was generated. In the case of a group stone column, the
Permeability in vertical
triangular pattern was considered and the spacing was varied as 1.24 × 10−4 1.2 × 10−9
direction in m/s
1.5, 2, 2.5, 3 and 3.5 times the diameter of the stone column. The Cohesion, c in kPa 0.276 Varied from 21.5 to 25
material characteristics were listed in Table 2. For both the stone Friction angle, φ Varied from 35 to 40 0.043
column materials and clays, the elasticity modulus, permeability,
and density values were obtained from the laboratory tests and Table 3. Plaxis2D Results
Poisson's ratios were taken from literature (Bowles, 1988). cu (kPa) s/d ratio φ (degree) L (mm) qu (kPa)
The medium mesh was generated in this Mohr columb model 25 1.5 40 400 232
to obtain a time-effective model. Fixed supports were chosen at 25 2 40 400 210
the bottom portion of the mesh and roller supports for the vertical 25 2.5 40 500 190
portion of the boundaries. At the interface between the soft soil 25 3 40 500 175
and the column, no slip elements were considered, because, 25 3.5 37.27 500 150
under an applied load, both the soft soil and rigid raft will undergo 22.5 1.5 37.27 600 182
vertical displacement together and differential settlement will be 21.5 2 37.27 600 165
zero. Ambily and Gandhi (2007) constructed the plaxis-2d model in 21.5 2.5 35 600 142
the same manner as discussed. 23 3 35 650 125
In this plaxis-2d model, the boundaries for the finite element 21.5 3.5 35 650 103
mesh are shown in Fig. 5. In this case, the boundaries were 22 2 36 400 195
considered to be far away from the zone, where the failure 23 2 37 500 202
mechanism would take place and the stresses would not get
disturbed during loading. To simulate this scenario both the
horizontal and vertical movements were kept restrained i.e. ux combination of both the fuzzy logic and the neural network was
and uy are chosen as zero (horizontal and vertical fixity) except used in the present study. This combination is termed as Adaptive
the top. The results of the Plaxis analysis are shown in Table 3. Neuro-Fuzzy Inference System (ANFIS). It takes the reasoning
capability of the fuzzy system and learning capability of the
3. ANFIS Model for Prediction of Bearing Capac- neural network. Among all the algorithms, the algorithm proposed
ity of Stone Column by Jang (1993) was used here for constructing the model of the
stone column.
For predicting the bearing capacity of the stone columns a Fuzzy logic is actually the mapping procedure from a known

Vol. 00, No. 0 / 000 0000 −3−

 Manita Das and Ashim Kanti Dey

input to an unknown output. This mapping provides the decision friction angle of stone column material. The one and only output
capability of the fuzzy logic. Generally a fuzzy system has five is the bearing capacity of the stone column.
parts such as (a) fuzzification of the input variables; (b)application of
the fuzzy operator (AND or OR) in the antecedent; (c) implication 3.2 Training and Testing of ANFIS Models
from the antecedent to the consequent; (d) aggregation of the As discussed above, three ANFIS models were introduced
consequents across the rules; and (e) defuzzification. with the data collected from the previously published literature
In practice, by using laboratory plate load test, the bearing (Poorooshasb and Meyerhof, 1997; Murugesan and Rajagopal,
capacity of stone column is determined. But it is always not 2006; Black et al., 2007; Malarvizhi and Ilampurthi, 2007;
possible to conduct laboratory experiment because laboratory Murugesan and Rajagopal, 2008; Golait et al., 2009; Hassen et
experiments are time-consuming and costly. Moreover there is a
practical limitation to the number of tests conducted in the
laboratory or the field to decide effective results. Sometimes Table 4. Details of Collected Data
because of a small number of experiments, it becomes very cu (s/d) φ L qu
difficult to determine statistically meaningful or reliable conclusions. kPa ratio degree mm kPa
Minimum 5 1.5 20 10.5 20
Again for analytical or numerical results, it becomes difficult to
Maximum 32 4 48 600 905
get effective results due to various errors i.e. assumption of
Mean 14.77 2.43 36.95 298.76 260.02
various properties, calculation error etc. Therefore, to eliminate
Standard deviation 8.71 0.73 11.54 191.96 208.49
all the uncertainties, Adaptive Neuro-Fuzzy Inference System
COV 58.96 30.26 31.24 64.25 80.18
(ANFIS) was used to observe the behavior of the stone columns
in the present work. The ANFIS models were conducted with
three combinations of data, namely, (a) ANFIS-E, where only al., 2010; Shivashankar et al., 2011; Deb et al., 2011; Shivashankar
experimental data were used as inputs, (b) ANFIS-A, where et al., 2011; Choobbasti et al., 2011; Zhang et al., 2013; Zhang et
either analytical or numerical results were used as inputs, (c) al., 2013; Andreou and Papadopoulos, 2014; Afshar and Ghazavi,
ANFIS-EA, where both the experimental and analytical or 2014; Mohanty and Samanta, 2015) for the training purpose.
numerical results were used as inputs. The ANFIS models were Data collected from 10 plate load tests and 12 Plaxis 2D analyses
constructed in the toolbox available in Matlab R2013a. were used for testing purpose. The ranges of all these data are
The structure of the rules of ANFIS model is shown in Fig. 6. shown in Table 4.
After completing the training for each model, the trained data
3.1 Model Architectures were recalled again to check the trained ANFIS models. A
In this study, to construct the ANFIS models, the data were comparison between the predicted value and the measured value
collected from available literature which was used for training is shown in Fig. 7. From the figure it is seen that the maximum
purpose. For the testing and validation of these models, 10 plate points of the model ANFIS-EA are nearer to the diagonal line,
load tests and 12 PLAXIS-2D analyses were conducted. The i.e. the accuracy is maximum in case of ANFIS-EA than any
diameter, spacing, and length of the stone columns and friction other models. The covariance (COV) of the measured to predicted
angle of the stone chips were varied. It is reported that when the values are 11.5%, 7.41% and 5.54% for the models ANFIS-E,
lateral deformation decreases, the failure stress of the stone ANFIS-A, and ANFIS-EA respectively. The mean values of the
column increases with the increase in diameter (Indraratna et al.,
2015). In the case of ANFIS-EA model, the results show more
accuracy than any other models. Based on the consideration of
the variables those have an effect on the bearing capacity of a
stone column, the number of input nodes is evaluated. In this
study, the variables used as input parameters are undrained
cohesion of soft soil, the diameter of the stone column, the length
of the stone column, the spacing between the stone column, and

Fig. 6. The Structure of the ANFIS (After Jang, 1993; Na et al., Fig. 7. Comparison of the Three ANFIS Models using Training
2009) Data

−4− KSCE Journal of Civil Engineering

 Determination of Bearing Capacity of Stone Column with Application of Neuro-fuzzy System

Table 5. Statistical Parameters for ANFIS Model for Training Data

Statistical parameters ANFIS-E ANFIS-A ANFIS-EA
RMSE 7.55 8.26 2.06
R2 0.971 0.96 0.986
MAPE(%) 21.11 25.12 8.2
MPR 1.013 1.076 1.0047

Table 6. Statistical Parameters for ANFIS Model for Testing Data

Statistical parameters ANFIS-E ANFIS-A ANFIS-EA
RMSE 7.4 7.9 1.45
R2 0.964 0.94 0.979
MAPE(%) 19.91 26.44 9.2
MPR 1.025 1.09 1.007

Fig. 8. Validation of the Three ANFIS Models using Testing Data

Sets

ratio of predicted and measured weights are 1.1, 1.077 and 1.007
respectively. It is noticed that the ratio of predicted to measured
value and COV are less in the case of ANFIS-EA model than
any other models. That means ANFIS-EA model gives the most
accurate prediction of bearing capacity of the stone column.
Testing of ANFIS models: Similar to training data, the graph
between predicted and measured values is shown in Fig. 8 below.
The ratio of the mean value of measured to predicted value is
1.0693, 1.0761 and 1.028 for the models ANFIS-E, ANFIS-A
and ANFIS-EA respectively. The COV values are 21.5%, 14.2%
and, 4.04% for the above models respectively. In this case also, Fig. 9. Sensitivity Analysis of Each Input Data
the result shows that the model ANFIS-EA gives a better
accuracy than any other method. From the results, the maximum
error can be noticed in the case of the models ANFIS-E followed Where,
by ANFIS-A, but ANFIS-EA predicts the bearing capacity with ai = The actual bearing capacity of a stone column,
a good accuracy. pi = Predicted value and
N = Total sample number.
4. Results and Discussion
All the statistical parameters for the prediction of the accuracy
of training and testing data such as RMSE, R2, MAPE, and MPR
4.1 Error Estimation of ANFIS Model
are presented in Table 5 and 6 respectively. From the Table 5 and
To predict the accuracy of each of the three ANFIS models, the
6, it is seen that the statistical parameters RMSE, R2, MAPE, and
statistical parameters like Root-mean-square Error (RMSE), absolute
MPR for ANFIS-EA are smaller than those for other two
fraction of variation (R2), Mean Absolute Percentage Error
models. Therefore, the model ANFIS-EA can predict the bearing
(MAPE) and Mean Prediction Ratio (MPR) were determined.
capacity of the stone columns most accurately.
The Eqs. (1)-(4) show the expressions of these parameters.
1 2 4.2 Sensitivity Analysis of Input Parameters
RMSE = ---- ∑ a – p (1)
N i i
It is important to carry out the sensitivity analysis of the input
parameters. Normally sensitivity means percentage change in
2
2 ⎛ ∑ (a – p ) ⎞ output due to 1% increase in input. In this study, by increasing 1
R = 1 – ⎜ ---------------------------
i i i

2 ⎟ (2)
⎝ ∑(p ) ⎠ i
unit increase in the input parameters, the effect on the output
parameter was observed for the sensitivity analysis. The sensitivity
1 of the input parameters is shown in Fig. 9. It is clearly seen that
a – p ) × 100
MAPE = ---- ∑ (----------------
- i i
(3)
N p i i
the ratio of spacing to the diameter of the stone column has the
strongest effect on the bearing capacity of the stone columns in
1 p ANFIS system, whereas the length of the stone column has the
MPR = ---- ∑ ⎛ ----⎞ i
(4)
N ⎝a ⎠ i i least effect on the bearing capacity of the stone column.

Vol. 00, No. 0 / 000 0000 −5−

 Manita Das and Ashim Kanti Dey

5. Conclusions Chika, Z., Aljanabi, Q. A., Kasa, A., and Taha, M. R. (2014). “Tenfold
cross-validation artificial neural network modeling of the settlement
In this study, a neuro-fuzzy inference system is presented to behavior of a stone column under a highway embankment.” Arab. J.
Geosci., Vol. 7, pp. 4877-4887, DOI: 10.1007/s12517-013-1128-6.
predict the bearing capacity of stone columns. Three models i.e.
Chikb, Z. and Aljanabi, Q. A. (2014). “Intelligent prediction of settlement
ANFIS-E, ANFIS-A, and ANFIS-EA are developed. In ANFIS- ratio for soft clay with stone columns using embankment improvement
E, only experimental results are used as input parameters. In techniques.” Neural Comput & Applic, Vol. 25, No. 1, pp. 73-82,
ANFIS-A, only analytical results are used as inputs and in DOI: 10.1007/s00521-013-1449-0.
ANFIS-EA, both experimental and analytical results are used as Choobbasti, A. Z., Zahmatkesh, A., and Noorzad, R. (2011) “Performance
input parameters. For the training of these models, input data set of stone columns in soft clay: Numerical evaluation.” Geotech.
are collected from the previous literature. Geol. Eng., Vol. 29, pp. 675-684, DOI: 10.1007/s10706-011-9409-x.
To validate the trained ANFIS models, the plate load test Deb, K., Samadhiya, N. K., and Namdeo, J. B. (2011). “Laboratory
model studies on unreinforced and geogrid-reinforced sand bed over
results and PLAXIS-2D results are used. The plate load tests are
stone column-improved soft clay.” Geotextiles and Geomembranes,
conducted with different diameters, different spacings, different Vol. 29, No. 2, pp. 190-196, DOI: 10.1016/j.geotexmem.2010.06.004.
friction angles and different stone column lengths. From the Etezad, M., Hanna, A. M., and Ayadat, T. (2015). “Bearing capacity of a
results of the testing of the trained models, the accuracy of the group of stone columns in soft soil.” International Journal of
prediction of the bearing capacity of the stone column can be Geomechanics, ASCE, ISSN 1532-3641/04014043(15).
noticed. The whole ANFIS system procedure mainly depends on Golait, Y. S., Satyanarayana, V., and Raju, S. S. V. (2009). “Concept of under
the quality and quantity of the input parameters. The more reamed cemented stone columns for soft clay ground improvement.”
number of data will give the more accuracy in prediction of IGC 2009, Guntur, INDIA.
Hassen, G., Buhan, P. D., and Abdelkrim, M. (2010). “Finite element
bearing capacity of the stone column. Especially, ANFIS-EA
implementation of a homogenized constitutive law for stone column-
model will help the geotechnical engineers to design the stone reinforced foundation soils, with application to the design of structures.”
columns. The model also concludes that out of the four input Computers and Geotechnics, Vol. 37, Nos. 1-2, pp. 40-49, DOI:
parameters, namely, undrained cohesion of a soft soil, the 10.1016/j.compgeo.2009.07.002.
spacing to diameter ratio of the stone column, the length of the Hassen, G., Buhan, P. D., and Abdelkrim, M. (2010). “Finite element
stone column, and friction angle of stone column materials, the implementation of a homogenized constitutive law for stone column-
spacing to diameter ratio is the most important parameter for reinforced foundation soils, with application to the design of structures.”
predicting the bearing capacity of the stone column. Computers and Geotechnics, Vol. 37, Nos. 1-2, pp. 40-49.
Hughes, J. M. O. and Withers, N. J. (1974). “Reinforcing of soft cohesive
soils with stone columns.” Ground Eng., Vol. 7, No. 3, pp. 42-49.
Referrences Indraratna, B., Ngo, N. T., Rujikiatkamjorn, C., and Sloan, S. W. (2015).
“Coupled discrete element–finite difference method for analyzing
Afshar J. N. and Ghazavi, M. (2014). “A simple analytical method for the load-deformation behaviour of a single stone column in soft
calculation of bearing capacity of stone column.” International Journal soil.” Computers and Geotechnics, Vol. 63, pp. 267-278, DOI:
of Civil Engineering, Vol. 12, No. 1, Transaction B: Geotechnical 10.1016/j.compgeo.2014.10.002.
Engineering, January 2014. IS 15284-1 (2003). “Design and construction for ground improvement-
Afshar J. N. and Ghazavi, M. (2014). “Experimental studies on bearing Guidelines.” Part-1: Stone columns [CED 43: Soil and foundation
capacity of geosynthetic reinforced stone columns.” Arab. J. Sci. engineering], ICS 93.020.
Eng.,, Vol. 39, No. 3, pp. 1559-1571, DOI: 10.1007/s13369-013- Jang, J., S. R. (1993). “ANFIS: Adaptive-network-based Fuzzy inference
0709-8. system.” IEEE Transactions on systems, Man, and Cybernetics,
Ambily., A. P. and Gandhi, S. R. (2004). “Experimental and theoretical Vol. 23, No. 3, May/June 1993.
evaluation of stone column in soft clay.” ICGGE. Juang, C. H., Wey, J. L., and Elton, D. J. (1991). “Model for capacity of
Andreou, P. and Papadopoulos, V. (2014). “Factors affecting the settlement single piles in sand using fuzzy sets.” Journal of Geotechnical
estimation of stone column reinforced soils.” Geotech. Geol. Eng., Engineering, ASCE, Vol. 117, No. 12, December, 1991, ISSN 0733-
Vol. 32, No. 5, pp. 1175-1185, DOI: 10.1007/s10706-014-9788-x. 9410/91/0012-1920.
Balaam, N. P. and Booker, J. R. (1981). “Analysis of rigid rafts supported by Madhav, M. R. and Vitkar, P. P. (1978). “Strip footing on weak clay
granular piles.” Int. J. Numer. Anal. Methods Geomech., Vol. 5, stabilized with a granular trench or pile.” Can. Geotech. J., Vol. 15,
No. 4, pp. 379-403. No. 4, pp. 605-609.
Black, J. A., Sivakumar, V., Madhav, M. R., and Hamill, G. A. (2007). Malarvizhi, S. N. and Ilampurthi, K. (2007). “Comparative study on the
“Reinforced stone columns in weak deposits: Laboratory model behavior of encased stone column and conventional stone column.”
study.” Journal of Geotechnical and Geoenvironmental Engineering, Soils and Foundations, Japanese Geotechnical Society, Vol. 47,
Vol. 133, No. 9, September 1, 2007, ASCE, DOI: 10.1061/(ASCE) No. 5, pp. 873-885, Oct. 2007.
1090-0241(2007)133:9(1154). Mohanty, P. and Samanta, M. (2015). “Experimental and numerical
Bowles (1988). “Foundation analysis and design.” 4th edition, McGraw studies on response of the stone column in layered soil.” Int. J. of
– Hill International Editions, New Delhi. Geosynth. and Ground Eng., pp. 1-27, DOI: 10.1007/s40891-015-
Castro, J. (2014). “An analytical solution for the settlement of stone 0029-z.
columns beneath rigid footings.” ActaGeotechnica, DOI: 10.1007/ Murugesan, S. and Rajagopal, K. (2006). “Geosynthetic-encased stone
s11440-014-0358-4. columns: Numerical evaluation.” Geotextiles and Geomembranes,

−6− KSCE Journal of Civil Engineering

 Determination of Bearing Capacity of Stone Column with Application of Neuro-fuzzy System

Vol. 24, No. 6, pp. 349-358, DOI: 10.1016/j.geotexmem.2006. and Geoenvironmental Engineering, ASCE, Vol. 137, No. 12, DOI:
05.001. 10.1061/(ASCE)GT.1943-5606.0000552.
Murugesan, S. and Rajagopal, K. (2008). “Performance of encased stone Shivashankar, R., Dheerendra Babu, M. R., Nayak, S., and Manjunath,
columns and design guidelines for construction on soft clay soils.” R. (2010). “Stone columns with vertical circumferential nails: Laboratory
Proceedings of the 4th Asian Regional Conference on Geosynthetics, model study.” Geotech Geol. Eng., (2010), Vol. 28, No. 5, pp. 695-
June 17 - 20, 2008 Shanghai, China. 706, DOI: 10.1007/s10706-010-9329-1.
Na, U. J., Park, T. W., Feng, M. Q., and Chung, L. (2009). “Neuro-fuzzy Shivashankar, R., Dheerendra Babu, M. R., Nayak, S., Rajathkumar, V.,
application for concrete strength prediction using combined non- (2011). “Experimental studies on behaviour of stone columns in
destructive tests.” Magazine of Concrete Research, 2009, Vol. 61, layered soils.” Geotech. Geol. Eng., Vol. 29, pp. 749-757, DOI:
No. 4, May, 245-256, DOI: 10.1680/macr.2007.00127. 10.1007/s10706-011-9414-0.
Poorooshasb, H. B. and Meyerhof, G. G. (1997). “Analysis of behavior Zhang, L., Zhao, M., Shi, C., and Zhao, H. (2013). “Settlement calculation
of stone columns and lime columns.” Computers and Geotechnics, of composite foundation reinforced with stone columns.” International
Vol. 20, No. 1, pp. 47-70, 1997. Journal of Geomechanics, ASCE, Vol. 13, No. 3, pp. 248-256, DOI:
Shahu, J. T. and Reddy, Y. R. (2011). “Clayey soil reinforced with stone 10.1061/(ASCE)GM.1943-5622.0000212.
column group: model tests and analyses.” Journal of Geotechnical

Vol. 00, No. 0 / 000 0000 −7−

View publication stats