Indian Geotechnical Conference

IGC 2022
15th – 17thDecember, 2022, Kochi
Kochi Chapter

Pile Capacity Estimation Considering Variability in Soil

Adhesion Factor

C Supritha Reddy1, Goushya Begum2 and B. Hanumantha Rao3

1 Undergraduate
Student, School of Infrastructure, Indian Institute of Technology Bhubanes-
war, Argul, Bhubaneswar, Odisha-752050, E-mail: [email protected]
2 Research Scholar, School of Infrastructure, IIT Bhubaneswar, Argul, Bhubaneswar, Odisha-

752050, E-mail: [email protected]

3Assistant Professor, Department of Civil Engineering, IIT Bhubaneswar, Argul, Bhubaneswar,

Odisha-752050, E-mail: [email protected]

Abstract. Estimation of the ultimate bearing capacity of the pile is considered the main
challenge in geotechnical engineering from a technical and economical view point. To
avoid pile failure under design axial superstructure load, end bearing and skin friction
capacity of the pile should be carefully estimated. Calculation of skin friction capacity
of the pile for soft soils is crucial when compared to other soils. As soft soils are con-
sidered to possess less shear strength, they undergo large deformation leading to varia-
tion in skin friction capacities. This paper focuses on the estimation of skin friction
capacity of pile for a cohesive soil stratum, considering the variation in adhesion factor
along the depth of the strata, using PLAXIS software. As observed by many authors,
the adhesion factor for soil pile is dependent on different parameters. Therefore, by the
numerical modelling, the adhesion factor values calculated based on different empirical
equations is validated.

Keywords: Adhesion factor, Skin friction, undrained shear strength, soft soils

1 Introduction

Any high-rise building constructed on weak-bearing soils needs a robust, deep foun-
dation system to prevent significant settlements over time. According to
Wrana.B (2015), a pile is considered to be a structural element that utilizes its end point
and skin friction to transfer the axial load of the superstructure to the soil. However, the
soil-pile interaction has a significant impact on a pile's skin friction capability. Large
pile settlements may occur if the soil strata are composed of soft soils.
Skin friction of pile in clayey soil is calculated taking into consideration, predomi-
nantly, of soil-pile interface parameter i.e. adhesion factor, which in clay is indeed de-
pended on the soil parameters like plasticity index, pile geometry, undrained shear
strength, overburden stress etc (Cherubini, C., & Vessia, G. (2007)). Arifin et al. (2022)
have analyzed the large bored pile through numerical approach (PLAXIS 2D) and con-
firmed that the adhesion factor of 0.97 and 0.94 for very soft and soft clayey layer soils,
while 0.56 and 0.48 for stiff and very stiff clayey layer soils. Kulhaway (1989) and

TH-03-040 1
Coduto (1989), in their studies, have asserted that the obtained adhesion factors for
various undrained shear strength of soil are in good agreement with adhesion distribu-
tion. As per Banerjee (2022), it was found that required piled group-raft area ratio for
minimizing the differential settlement of a raft in a layered soil should be within a range
of 0.4 to 0.6.
A squared pile raft foundation was modeled using PLAXIS 3D for different pile
lengths, spacing between piles, and a number of piles to investigate the pile behavior in
terms of pile load capacity and load settlement curve under uniform vertical load. At a
specific load, the settlement of the pile increases as the pile spacing increases.
Generally, the ultimate bearing capacity of a pile is determined by pile load tests.
The obtained results from pile load tests can further be employed to ascertain the quality
of numerical models. As per Ezzat et al (2019), three different numerical models i.e.
Modified Mohr-Coulomb (MMC), Mohr-Coulomb (MC) and Soft soil models can be
used to simulate the field load tests. As we know that MC model is elastic perfectly
plastic model, the other two models have a non- linear relation between axial strain and
deviatoric stress. The primary difference to note in these models is that stiffness in MC
model is constant whereas in SS and MMC model stiffness is stress dependent. It was
observed that very good agreement was found between field settlements and the calcu-
lated settlements using MMC model.
The objective of the present study is to calculate the skin friction capacity of a pile
under axial loading by numerical approach (PLAXIS-3D) and compare it with the pile
load test of the field. The numerical results are used to assess the accuracy of the em-
pirical equations of adhesion factor distribution suggested by different authors.

2 Methodology

The skin friction of a pile for cohesive and granular soils can be determined by the
analytical equation given in IS 2911 (Part 1). As the adhesion factor is dependent on
various parameters, it is estimated using empirical equations suggested by them. Con-
sequently, the skin friction of the pile is calculated by the analytical equation suggested
by the IS 2911 code provision. In the present study, a bored pile of 1200 mm diameter
and 25 m length is modeled using PLAXIS-3D software with an application of axial
point load. The soil–pile interaction factor (adhesion factor) suggested by different au-
thors is given as input data to find out the variation in skin friction capacity obtained
from the load settlement curve, as interpreted in IS 2911 (Part 4). The load settlement
curve obtained for different adhesion factor values is compared with the field load set-
tlement curve to assess the accuracy of the adhesion factor determined as per the em-
pirical equations.

2.1 Analytical Equations for estimating Adhesion Factor Values

Skin friction for cohesive soils can be estimated as per IS 2911 (Part 1), given as:
𝑄𝑠 = ∑𝑛𝑖=0(𝛼𝑖 ∗ 𝑐𝑖 ∗ 𝐴𝑠 ) (1)
where, Qs = ultimate skin friction; αi = adhesion factor of the ith layer of soil strata; ci =
undrained shear strength of ith layer strata; As = surface area of the pile.

TH-03-040 2
The values of the adhesion factor, in Eq. (1), are determined based on the empirical
approaches proposed by different researchers. Please note that each empirical approach
accounts for different soil properties. Thus, there is a chance to understand and signify
which soil parameter has major impact on the skin friction capacity of a pile installed
in cohesive soils. The various empirical approaches are given herein.
As per the API (1984), the adhesion factor is a function of undrained shear strength
alone.
α = 1 – (cu -25)/90 - when 25 kPa < cu<70kPa (2)
α = 1 - when cu<= 25 kPa
α = 0.5 - when cu> 70 kPa
The adhesion factor equation proposed by Sladen (1992) is dependent on two soil pa-
rameters such as effective stress and undrained shear strength.
α = 0.5 * (σv’/cu)0.45 (3)
According to Karlsud (2005), the adhesion factor varies with the plasticity index of α
clayey soils. The correlation of cu/σv’ and Plasticity Index (Ip) with α is introduced in
the form of trend lines described in the below graph:

Fig. 1 Variation of α with cu/σv’ and Plasticity Index (Ip)

The adhesion factor estimated by Kulhaway (1989) is dependent on effective stress and
undrained shear strength parameters of soil:
α = 0.21 +0.26 *(cu/Pa) (4)
for cu/Pa <= 3 and α <=1
Variation of α with cu/Pa and slenderness ratio (L/D) as proposed by Kolk and Van der
Velde (1996) is given as:
α =0.5*(D/L)0.2*(cu/ σv’) 0.3 (5)

TH-03-040 3
2.2 Finite Element Modeling
The pile was modeled using Plaxis-3D in layered cohesive soil whose parameters are
given in Table 1. PLAXIS 3D is a three-dimensional program for deformation, stabil-
ity, and flow analyses for different types of geotechnical applications. The program
uses a comfortable graphical user interface to quickly create a geometry model and a
finite element mesh. It provides different models to simulate the soil behavior which
are linear elastic, Mohr-Coulomb, strain hardening, and soft soil creep models. The soil
behavior is assumed to follow the Mohr-Coulomb failure criterion in the model. In
PLAXIS, to simulate the Mohr-Coulomb model five parameters of soil are required.
Young’s Modulus (E) and Poisson’s ratio (ν) are inputted as stiffness parameters and
cohesion (c), Friction angle (φ), and Dilatancy Angle (ψ) are used as strength parame-
ters. The value of these parameters are listed in Table 1.
As regards to boundary conditions, the bottom of soil body is fixed in all three direc-
tions. The side faces of the soil body are fixed in X and Y directions but free to move
along the Z-direction.
The analyzed pile is cylindrical and isolated element subjected to axial loading, allow-
ing for three-dimensional embedded pile simulation. Soil properties vary considerably
from layer to layer. The embedded pile model can be used to model the pile and param-
eters used in PLAXIS 3D are given in Table 2. The embedded pile is connected to the
adjacent soils by special interfaces named skin interfaces and foot interfaces. A point
load is applied over the pile with increments and the maximum load applied is 3800
kN.
Medium mesh is selected as the optimum mesh generation element considering exces-
sive time consumption of fine, and very fine meshes and the deformed soil body mesh
is shown in Figure 3.

Interface Modelling

The interface between soil and pile is modeled by input parameter called interface re-
duction factor (Rinter) [11]. If Rinter is 1, the soil is fully bonded to the surface of pile.
Interface properties indirectly depend on the soil strength. Undrained shear strength of
interface can be determined by Eq. 6 and adhesion of soil-pile interface is given by Eq.
7.
su,i = Rinter su,soil (6)
sa,i = α su,soil (7)
where su,i is undrained shear strength of interface, su,soil is undrained shear strength of
soil, sa,I is the adhesion of soil and α is the adhesion factor of clay.
Eq. 6 can be compared to Eq. 7 when the adhesion of soil is equal to adhesion factor
multiplied by the undrained shear strength of soil. So, the Rinter resembles the adhesion
factor of soil and su,i resembles the adhesion of interface (mobilized shear strength). So,
Rinter which is the adhesion factor of clay is given as input in the PLAXIS and varied
for the clayey soil located at a depth of 13 m. Adhesion factor of clayey soil of thickness
5 m suggested by various authors are given in the Table 4. So, the Rinter parameter is

TH-03-040 4
varied and the differences in the plot of load displacement are observed for each value
of adhesion factor (Rinter).

Table 1. Undrained parameters of Paradip soil used in PLAXIS-3D software

Parameter Sand Clay Sand
Range of Depth (m) 0 -13 13 -18 18 -40
Material model Mohr coulomb Mohr coulomb Mohr coulomb
Material behavior Undrained (A) Undrained (B) Undrained (A)
Su(KN/m^2) 0 22 0
Young's modulus 8000 8000 12000
(KN/m^2)
Poisson's ratio 0.33 0.2 0.33
Angle of internal fric- 33 0 34
tion
Strength reduction fac- 1 Variable - 1
tor 0.27,0.52,1

Table 2. Parameters of the pile in Paradip soil

Pile data Value

Pile Length 25 m
Pile Diameter 1.2 m
E (MPa) 29580
Model of pile Embedded beam
Material behavior Linear elastic
Poisson's Ratio 0.2

13 m

5m

22 m

Fig 3.1. Deformed mesh of soil body Fig 3.2. Pile centered in the soil

TH-03-040 5
2.3 Pile Load Test Data

The pile load test conducted at Paradip port which is a natural deep-water port on the
East coast of India in Odisha was used for the study purpose. The static load test (SLT)
was performed on a cylindrical concrete pile having a length of 25 m and diameter of
1200 mm. The soil parameters based on the geotechnical investigations are given in
Table 3.

Table 3. Bore log details of Paradip

Depth of soil Group of SPT Cohesion Plasticity Index

(m) soil (kPa) (Ip)
0.5 SM -- -- --
1.5 SM 6 -- --
3 SC 19 -- 19
4.5 SP-SM 22 -- --
6 SP-SM 25 -- --
7.5 SP 33 -- --
9 SP 40 -- --
10.5 SP 33 -- --
12 SP 18 -- --
13.5 CI 2 -- 21
15 CI -- 22 22

3 Results

3.1 Analytical results of adhesion factor using various methods

The adhesion factor of clay located at 13 m depth estimated by different empirical

approaches are given in Table 4. The adhesion factors suggested by different authors
are in range of 0.27 to 1.16. As various parameters are taken into consideration while
estimating adhesion factor, differences in suggested values of α is also high. The adhe-
sion factor for other layers are ignored as the remaining layers are granular soils (non-
cohesive soils) for which the skin friction capacity of pile is calculated by the effective
stress approach as specified by IS 2911 Section 1.
On the other hand, the skin friction of the soil sample is interpreted from load dis-
placement curve of the plate load test as per the IS 2911 code (Fig. 3). Furthermore,
skin friction is calculated as per adhesion factor (Table 4) and the values of it are pre-
sented in Fig. 4. The load capacity of pile predicted using the empirical equations is
overestimated when compared to the skin friction obtained from pile load test.

TH-03-040 6
 Table 4. Adhesion factor estimated by various methods for Paradip soil

Depth Adhesion factor, α

API Sladen Karlsud Kulhaway Kolk and
(1984) (1992) (2005) and vander
Jackson Velde
(1989) (1996)
0m-2m -- -- -- -- --
2 m - 13 m -- -- -- -- --
13 m - 18 m 1 1.16 1 0.27 0.52
18 m - 27 m -- -- -- -- --

Load (KN)

0 1000 2000 3000 4000 5000

0
-5
Displacement (mm)

-10
-15
-20
SKIN FRICTION -
-25 2416.94 KN
-30
-35
-40
-45
-50

Fig. 3. Interpretation of skin friction as per IS method for Paradip Soil

8486.2
8789

8855

8789
SKIN FRICTION (KN)

2416.94

API Sladen Karlsud Kulhaway Pile load test

Fig. 4. Skin friction as per the adhesion factor suggested by different authors

TH-03-040 7
3.2 Results of Numerical Analysis
In addition to estimation of skin friction capacity, the load displacement curve is
predicted using PLAXIS-3D for static axial load applied on the top of pile. The varia-
tion in the load displacement curve obtained for Rinter (adhesion factor) suggested by
different authors and field load displacement curve is plotted in the Fig. 5.
Load (KN)
0 1000 2000 3000 4000 5000
0

-10
Displacment (mm)

-20

-30

-40

-50

-60

-70
Kulhaway (Rinter - 0.27)
Kolk and Van der velde (Rinter =0.52)
API and Karlsud (Rinter =1)
Pile Load test

Fig. 5. The load displacement curve obtained from numerical modeling and pile
load test

3000

2416.94
2500
Skin Friction(kN)

2000
1537.43 1537.43
1500 1241.97 1255.04

1000

500

0
Kulhaway Kolk and API Karlsud Pile load
Va der test
velde

Fig 6. The skin friction capacity from load displacement curves obtained from
numerical modelling as per IS 2911 code
As the Rinter is increased to 1, the load displacement curve is also approached towards
the field load displacement curve. The skin friction for the load displacement curve

TH-03-040 8
calculated as per IS code 2911 is shown in Fig. 6. The variation in skin friction capac-
ities is very low as adhesion factor of clay is changed and also the skin friction of pile
obtained by numerical model by varying the adhesion factor is underestimated wile
compared to that obtained from plate load test. The skin friction calculated as per ad-
hesion factor values suggested by API and Karlsud are close to that of Field load test.
So, the adhesion factor estimated by API and Karlsud as 1 can be used for the clayey
soil in Paradip situated at a depth of 13.5 m. The settlement of pile in the soil body were
comparatively high with those of field settlements as shown in Fig 7.

Fig 7. The contour of settlement of pile in soil strata

4 Conclusions

Based on the numerical studies and analysis of experimental data, arrived at the fol-
lowing conclusions given below:
 The variation in skin friction capacity is noticed to be very low with the change
in the adhesion factor values.
 Settlements obtained by the numerical analysis are high when compared with that
of field data.
 It is found that the load settlement curve obtained when adhesion factor of soft
clay of unity is close to that of the field load settlement curve.
 Adhesion factors estimated by empirical methods are excessively simplified and
cannot validate the reason for the change in adhesion factor for all kinds of soils.
 It is necessary to check the wide range of experimental data to know the reason
behind the change in adhesion factor for different soil types and its effect on the
calculation of skin friction.
 The skin friction predicted by the numerical approach is lesser when compared
with that of experimental test. This may account for the constant stiffness of soil
assumed in model and stiffness does not as the soil undergoes settlement.

5 References
1. American Petroleum Institute, API Recommended Practice for Planning, Designing and Con-
structing Fixed Off-shore Platforms, API, Washington, DC, 1984\

TH-03-040 9
 2. Arifin, Y. F. (2022). Uji Plagiasi (Turnitin) Adhesion Factor of Large Diameter Bored Piles
in Soft to Stiff Clay in the Sei Alalak Bridge Replacement Project‒South Kalimantan, Indo-
nesia.
3. Banerjee, R., Bandyopadhyay, S., Sengupta, A., & Reddy, G. R. (2022). Settlement behaviour
of a pile raft subjected to vertical loadings in multilayered soil. Geomechanics and Geoengi-
neering, 17(1), 282-296.
4. Cherubini, C., & Vessia, G. (2007). Reliability approach for the side resistance of piles by
means of the total stress analysis (α Method). Canadian geotechnical journal, 44(11), 1378-
1390.
5. Ezzat, M., Zaghloul, Y., Sorour, T., Hefny, A., & Eid, M. (2019). Numerical simulation of
axially loaded to failure large diameter bored pile. International Journal of Geotechnical and
Geological Engineering, 13(5), 283-297.
6. IS 2911 (Part 1/Sec 1) : 2010 Indian Standard design and construction of pile foundations —
code of practice part 1 concrete piles
7. IS 2911 (Part 4) : 2013 Design and construction of pile foundations — code of practice part
4 load test on piles
8. KARLSRUD K., CLAUSEN C.J.F., AAS P.M., Bearing Capacity of Driven Piles in Clay, the
NGI Approach, Proc. Int. Symp. on Frontiers in Offshore Geotechnics, 1. Perth 2005, 775–
782
9. Kulhawy, F.H., and Jackson, C.S. 1989. Some observations on un-drained side resistance of
drilled shafts. Foundation Engineering:current principles and practices CongressEvanstonIlli-
nois,2:1011–1025.
10. Park, D., Park, D., & Lee, J. (2016). Analyzing load response and load sharing behavior of
piled rafts installed with driven piles in sands. Computers and Geotechnics, 78, 62-71.
11. PLAXIS 3D - version 9.02, Reference manual, Delft University of Technology, Delft, Neth-
erlands,2018.
12. Sladen, J. A. (1992). The adhesion factor: applications and limitations. Canadian Geotech-
nical Journal, 29(2), 322-326.
13. Wrana, B. (2015). Pile load capacity–calculation methods. Studia Geotechnica et Mechan-
ica, 37(4), 83-93

TH-03-040 10