SlideShare a Scribd company logo
Geomechanics and Engineering, Vol. 4, No. 1 (2012) 67-78 67
Settlement analysis of viscoelastic foundation under
vertical line load using a fractional Kelvin-Voigt model
Hong-Hu Zhu*1
, Lin-Chao Liu2
, Hua-Fu Pei3
and Bin Shi1
1
School of Earth Sciences and Engineering, Nanjing University, Nanjing, China
2
School of Civil Engineering, Xinyang Normal University, Xinyang, China
3
Department of Civil and Structural Engineering, The Hong Kong Polytechnic University, Hong Kong, China
(Received January 23, 2011, Revised February 19, 2012, Accepted March 05, 2012)
Abstract. Soil foundations exhibit significant creeping deformation, which may result in excessive
settlement and failure of superstructures. Based on the theory of viscoelasticity and fractional calculus, a
fractional Kelvin-Voigt model is proposed to account for the time-dependent behavior of soil foundation
under vertical line load. Analytical solution of settlements in the foundation was derived using Laplace
transforms. The influence of the model parameters on the time-dependent settlement is studied through a
parametric study. Results indicate that the settlement-time relationship can be accurately captured by
varying values of the fractional order of differential operator and the coefficient of viscosity. In comparison
with the classical Kelvin-Voigt model, the fractional model can provide a more accurate prediction of
long-term settlements of soil foundation. The determination of influential distance also affects the calculation
of settlements.
Keywords: soil foundation; fractional viscoelastic model; the Flamant-Boussinesq solution; settlement;
Laplace transform.
1. Introduction
Research results show that the soil foundation exhibits significant creeping behaviour and the
ground deformation under loading from superstructures is highly time-dependent (Bjerrum 1967).
For the estimation of the increase in stresses at various points and associated displacements caused
in soil foundations due to external loading, there have been a number of viscoelastic, elastic-visco-
plastic (EVP) or elastoplastic-viscoplastic models (e.g., Christie 1964, Kaliakin and Dafalias 1990,
Yin and Graham 1994, Justo and Durand 2000). Among them, viscoelastic foundation models can
provide satisfactory simulation of the time-dependent behaviour when the stress level in soil is low
and yielding of soil is thus not reached. The ideal assumption of the theory of viscoelasticity,
namely that the foundation is homogeneous, viscoelastic, and isotropic, is not quite true in most
cases. However, it provides a close estimation of long-term settlement and failure possibility of
superstructures and, using proper safety factors, a safe design can be developed.
Because integer-order differential operators are used in the viscoelastic constitutive equations, the
*Corresponding author, Associate Professor, E-mail: zhh@nju.edu.cn
Technical Note
68 Hong-Hu Zhu, Lin-Chao Liu, Hua-Fu Pei and Bin Shi
kernel functions (creep modulus and relaxation modulus) of viscoelastic models are normally a
combination of exponential functions. Thus these models have difficulties in predicting settlement of
soil foundation accurately and their applicable range is significantly limited. Fractional constitutive
models of viscoelastic materials were first introduced by Gemant (1936). In the constitutive equations,
integer-order differential operators were replaced by fractional-order differential operators. These models
have shown powerful capability in capturing static and dynamic stress-strain-time relationships of
viscoelastic materials and have been applied in various fields in the past few decades (Bagley and
Torvik 1983 and 1986, Welch et al. 1999, Zhang and Shimizu 1998, Li et al. 2001, Rossikhin and
Shitikova 2010). However, due to the complexity involved in obtaining the analytical solution, these
fractional models have not been applied in the field of geotechnical engineering until recent years
(Atanackovic and Stankovic 2004, Dikmen 2005, Liu et al. 2006, Paola et al. 2009, Zhu et al. 2011).
In this paper, a fractional Kelvin-Voigt viscoelastic constitutive model was proposed to study time-
dependent settlement of soil foundations. The analytical solution of a half-space under a vertical line
load is derived using Laplace transforms. The influences of the model parameters and the predetermined
influential distance on the calculated settlements at the foundation surface are further explored using
a parametric study.
2. Fractional Kelvin-Voigt model
The generalized viscoelastic constitutive equation using fractional derivatives can be described by
(Padovan 1987)
(1)
where m and n are positive integers; and are Riemann-Liouville fractional differential
operators defined by (Miller and Ross 1993)
(2)
where Γ(u) is the Gamma function, .
If m = n = 1 and b1 = 0, Eq. (1) can be expressed by
(3)
It is obvious that if α = 1, Eq. (3) represents the constitutive equation of the classical Kelvin-Voigt
viscoelastic model (KVM). Thus this new model can be called a generalized fractional Kelvin-Voigt
model (FKVM), in which the integer-order differential operator is now replaced by a fractional-
order one.
In the theory of viscoelasticity, the stress-strain relationships are
(4)
σ t( ) +
m 1=
M
∑ bm
d
βm
dt
βm
---------σ t( ) = E0ε t( ) +
n 1=
N
∑ En
d
αn
dt
αn
--------ε t( ) 0 βm 1,0 αn 1< < < <( )
d
βm
/dt
βm
d
αn
/dt
αn
d
α
f t( )
dt
α
-------------- =
1
Γ 1 α–( )
--------------------
d
dt
----
0
t
∫
f τ( )
t τ–( )
α
----------------dτ 0 α 1< <( )
Γ u( ) =
0
∞
∫ t
u 1–
e
u–
dt
σ t( ) = E0ε t( ) + E1
d
α
ε t( )
dt
α
---------------
P′Sij t( ) = Q′eij t( )
Settlement analysis of viscoelastic foundation under vertical line load using a fractional Kelvin-Voigt model 69
(5)
where Sij(t) and σij(t) are deviatoric and hydrostatic stress tensors; eij(t) and εij(t) are deviatoric and
hydrostatic strain tensors; P', Q', P'' and Q'' are linear differential operators consisting of viscoelastic
parameters.
The volumetric and shear strains of foundation soils under low stress levels are assumed to be
time-dependent. The FKVM shown in Fig. 1 is thus utilized for the description of the viscoelastic
behaviour of the bulk and shear moduli,, which describe the volume and shape changes of soil mass
under external loading. Thus
P' = 1 (6)
(7)
P'' = 1 (8)
(9)
where G, K and η are the shear modulus, the bulk modulus, and the coefficient of viscosity, respectively.
Determination of the model parameters can be obtained from laboratory and field testing.
3. Analytical solutions of settlements using fractional Kelvin-Voigt model
3.1 Flamant-Boussinesq elastic solution
Fig. 2 shows the case where a line force of p0 per unit length is applied on the surface of a half-
space. The elastic solution for this plane-strain problem was developed by Flamant (1892) by
modifying the three-dimensional solution of Boussinesq (Timoshenko and Goodier 1970). The
Flamant-Boussinesq solution using an elastic model (EM) is widely used in foundation engineering
for the calculation of stresses and estimation of displacements.
In the polar coordinate system, assume that at a point P( ) on the ground surface, ,
i.e., the influential distance of ground surface settlement is assumed to be r0. The displacements at
any point in the foundation can be obtained as
P″σij t( ) = Q″εij t( )
Q′ = G + η
d
α
dt
α
-------
Q″ = K + η d
α
dt
α
-------
r0, π/2 uθ( )θ=
π
2
---,r=r0
= 0
Fig. 1 Three-parameter fractional Kelvin-Voigt model
70 Hong-Hu Zhu, Lin-Chao Liu, Hua-Fu Pei and Bin Shi
(10)
(11)
where E and μ are the modulus of elasticity and Poisson’s ratio, respectively; ur and uθ denote the
displacements in the radial and circumferential directions, respectively.
Taking E = 144 MPa, μ = 0.2, p0 = 1 MPa and r0 = 15 m, the distributions of displacements within
the soil foundation using the elastic model (EM) are depicted in Fig. 3. It is seen that the surface
settlement at the loading point approaches infinity and the settlements within the foundation reduce
sharply with increasing depth. For locations far away from the loading point, heave of the foundation
occurs.
In the case of , Eqs. (10) and (11) for radial and circumferential displacement components
at the ground surface can be given as the following simple form
(12)
(13)
ur =
2p0 1 μ
2
–( )
πE
--------------------------cosθ ln
r0
r
---- +
1 2μ–( ) 1 μ+( )p0
πE
-----------------------------------------θ sinθ −
1 μ+( )p0
πE
---------------------cosθ
uθ =
2p0 1 μ
2
–( )
πE
--------------------------– sinθ ln
r0
r
---- +
1 2μ–( ) 1 μ+( )p0
πE
-----------------------------------------θ cosθ
θ = π/2
ur( )θ=
π
2
---
=
1 2μ–( ) 1 μ+( )p0
2E
----------------------------------------- = −
3p0
4 3K G+( )
------------------------
uθ( )
θ=
π
2
---
=
2p0 1 μ
2
–( )
πE
--------------------------ln
r0
r
---- = −
p0ln
r0
r
----
2π
--------------
3
3K G+( )
--------------------- 1
G
----+
Fig. 2 Schematic illustration of a foundation subjected to vertical line load at the surface
Settlement analysis of viscoelastic foundation under vertical line load using a fractional Kelvin-Voigt model 71
Fig. 3 Displacement distributions calculated from the Flamant-Boussinesq elastic solution
72 Hong-Hu Zhu, Lin-Chao Liu, Hua-Fu Pei and Bin Shi
Eqs. (12) and (13) indicate that there is a logarithmic relationship between the distance to the
loading point and the vertical displacement, and that the horizontal displacement at the ground
surface is a constant that is not affected by r.
3.2 Viscoelastic solution using classical Kelvin-Voigt model
Considering a line load applied on the foundation shown in Fig. 2, the loading condition is
specified as
(14)
where H(t) is the unit-step function defined as .
To obtain the viscoelastic solution, the volumetric and shear strains of the foundation soil satisfies
the KVM. Thus Eqs. (7) and (9) turns into
(15)
(16)
Taking the Laplace transforms of Eqs. (6), (8), (15) and (16), and combining them with Eqs. (12)
and (13), the displacements at the ground surface can be expressed by
(17)
(18)
Taking the inverse Laplace transforms of Eqs. (17) and (18), we get
(19)
(20)
where , . Eqs. (19) and (20) show that the time-dependent displacements
are exponential creep functions. It is obvious that if , the ultimate displacements equal the
Flamant-Boussinesq solution using elastic model (EM).
3.3 Viscoelastic solution using fractional Kelvin-Voigt model
Taking the Laplace transforms of Eqs. (6) to (9) and combined with Eqs. (12) and (13), the
p = p0H t( )
H t( ) =
1 t 0≥
0 t 0<⎩
⎨
⎧
Q′ = G + η
d
dt
----
Q″ = K + η
d
dt
----
ur( )θ=
π
2
---
= −
3p0
4
--------
1
3K G 4ηs+ +( )s
---------------------------------------
uθ( )θ=
π
2
---
= −
p0ln
r0
r
----
2π
--------------
3
3K G 4ηs+ +( )s
---------------------------------------
1
G ηs+( )s
-----------------------+
ur( )θ=
π
2
---
= −
3p0
4 3K G+( )
------------------------ 1 e
t
τ1
----–
–
⎝ ⎠
⎜ ⎟
⎛ ⎞
uθ( )
θ=
π
2
---
= −
p0ln
r0
r
----
2π
-------------- 3
3K G+( )
--------------------- 1 e
t
τ1
----–
–
⎝ ⎠
⎜ ⎟
⎛ ⎞ 1
G
---- 1 e
t
τ2
----–
–
⎝ ⎠
⎜ ⎟
⎛ ⎞
+
τ1 = 4η/ 3K G+( ) τ2 = η/G
t ∞→
Settlement analysis of viscoelastic foundation under vertical line load using a fractional Kelvin-Voigt model 73
displacements at the ground surface can be expressed by
(21)
(22)
In order to take the inverse Laplace transform, note the following characteristics of the fractional
derivative in the Laplace transform for zero initial condition
(23)
where Eα is the Mittag-Leffler function defined as .
In view of Eq. (23), taking the inverse Laplace transforms of Eqs. (21) and (22), we get
(24)
(25)
where , .
It is obvious that when , , .
Thus the ultimate displacements equal the elastic solution.
If , . Thus Eqs. (24) and (25) turn into
(26)
(27)
Eqs. (26) and (27) indicate that the displacements at the ground surface are not time-dependent
any more and the displacement values are half of those from the elastic solution. Therefore, the
FKVM collapses to be a ideal elastic model as .
ur( )
θ=
π
2
---
= −
3p0
4
-------- 1
3K G 4ηs
α
+ +( )s
-----------------------------------------
uθ( )θ=
π
2
---
= −
p0ln
r0
r
----
2π
--------------
3
3K G 4ηs
α
+ +( )s
-----------------------------------------
1
G ηs
α
+( )s
-------------------------+
L
1– 1
G ηs
α
+( )s
------------------------- =
1
G
---- 1 Eα–
t
τ
---
⎝ ⎠
⎛ ⎞
α
–
⎩ ⎭
⎨ ⎬
⎧ ⎫
Eα t( ) =
0
∞
∑ t
n
Γ 1 α n+( )
-------------------------; τ =
η
G
----
ur( )
θ=
π
2
---
= −
3p0
4 3K G+( )
------------------------ 1 Eα–
t
τ 1
-----
⎝ ⎠
⎛ ⎞
α
–
⎩ ⎭
⎨ ⎬
⎧ ⎫
uθ( )
θ=
π
2
---
= −
p0ln
r0
r
----
2π
--------------
3 3Eα
t
τ 1
------
⎝ ⎠
⎛ ⎞
α
––
3K G+( )
----------------------------------------
1 Eα–
t
τ 2
------
⎝ ⎠
⎛ ⎞
α
–
G
-------------------------------------+
⎩ ⎭
⎪ ⎪
⎨ ⎬
⎪ ⎪
⎧ ⎫
τ1 =
4η
3K G+
------------------ τ2 =
η
G
----
0 α< 1≤ limt ∞→ ur( )
θ=
π
2
---
=
3p0
4 3K G+( )
--------------------------– limt ∞→ uθ( )
θ=
π
2
---
=
p0ln
r0
r
----
2π
---------------–
3
3K G+( )
----------------------
1
G
----+
α 0→ Eα t( )
0
∞
∑ t
n
Γ 1( )
----------- =
0
∞
∑ t
n
=
1
1 t–
---------- 1– t 1< <( )→
ur( )
θ=π
2
---
= −
3p0
8 3K G+( )
------------------------
uθ( )
θ=
π
2
---
=
p0ln
r0
r
----
4π
--------------–
3
3K G+
-----------------
1
G
----+
α 0→
74 Hong-Hu Zhu, Lin-Chao Liu, Hua-Fu Pei and Bin Shi
Table 1 Parameters of the FKVM
Item Value
Young’s modulus E (MPa) 144
Poisson’s ratio μ 0.2
Shear modulus G (MPa) 60
Bulk modulus K (MPa) 80
Coefficient of viscosity η (MPa·d) 1000
Fractional differential order α 0.5
r0 (m) 15
If α = 1, . Obviously, Eqs. (23) and (24) currently become the
viscoelastic solution, which means that the FKVM turns into a classical KVM.
4. Numerical example and analysis
4.1 Comparison of elastic and viscoelastic models
How to determine the fractional model parameters is a major concern in practical applications. To
carry out a preliminary investigation of these parameters’ influence on the calculation results, a
parametric study is presented as following. Here, we analyze the quasistatic problem of a viscoelastic
foundation subjected to line loading p0 = 1 MPa using the FKVM. The basic parameters used in this
analysis are listed in Table 1.
Fig. 4 shows the surface settlement results using the fractional model at t = 10 d and 50 d,
together with the results from the classical KVM solution and the Flamant-Boussinesq elastic
solution. The present results show that the discrepancy between the results from FKVM and KVM
increases as t increases. The calculated settlements using KVM develops rapidly with the elapsed
time while those using FKVM will take much more time.
4.2 Parametric study of the FKVM
To study the influence of the coefficient of viscosity in the FKVM, η is set to 500, 1000 and
1500 MPa·d. Fig. 5 depicts the settlement-time curves at r = 1 m over time using the FKVM, the
KVM and the EM. It is shown that the coefficient of viscosity mainly affects the overall creeping
rate of the foundation soil. For the same time point, the settlements increase with the decrease of η.
Note that the ultimate settlement in the viscoelastic solution is equal to that in the elastic solution, η
may determine the magnitude of creeping by controlling the time duration to achieve the ultimate
settlement. If η decreases, the creeping effect will function more slowly and the ultimate settlement
will be obtained for a much longer time. Results from the classical Kelvin-Voigt model (α = 1),
which is shown in Fig. 6 by a dashed line, shows larger settlements and less time to get to the
ultimate settlement.
Eα t( ) =
0
∞
∑ t
n
Γ 1 αn+( )
------------------------- =
0
∞
∑ t
n
n!
----- = e
t
Settlement analysis of viscoelastic foundation under vertical line load using a fractional Kelvin-Voigt model 75
To investigate the influence of the fractional order of differential operator in the FKVM, α is set
to 0.1, 0.3, 0.5, 0.7 and 0.9 while η is fixed to 1000 MPa·d. Fig. 6 illustrates the settlement at r = 1
m over time. It is shown that the fractional order α has a significant influence on the long-term
performance of the foundation surface settlements. At t1/2 = 10 d, the settlements are half of the
ultimate settlement (5.75 mm) for any value of α. t1/2 is found to have an approximately linear
relationship with η and 1/E. After that time point, with the increase of α, the foundation settlement
and corresponding settlement rate increase. Therefore, α may determine the functioning law of
creeping effect by separating the time-dependent deformation of soil foundation into two stages. The
first stage may associate with the consolidation process of the foundation soil and the second stage
is mainly due to secondary compression. Therefore, a small α value is fit for sand foundation,
which has a small permeability coefficient. While for soft clay foundation, the α value should be
Fig. 4 Comparsion of settlements calculated using elastic, Kelvin-Voigt and fractional Kelvin-Voigt models
76 Hong-Hu Zhu, Lin-Chao Liu, Hua-Fu Pei and Bin Shi
increased. In the KVM where α = 1 (the dashed line in Fig. 6), a minimum settlement rate at stage
I and a maximum settlement rate at stage II is obtained.
The key feature of the FKVM illustrated in Figs. 5 and 6 is the capability to represent generalised
viscoelastic behaviour over many decades of time with only three parameters. It is also indicated
from the above parametric study that the classical viscoelastic model may overestimate the
development of long-term settlement (Stage II) and fail to describe the attenuation creep of soil
mass accurately. Using the fractional viscoelastic models, we can obtain the best fitting effect of the
actual soil deformation characteristics by selecting appropriate values of the fractional differential
order and the coefficient of viscosity. In geotechnical design, this model can provide more accurate
settlement prediction and thus save construction costs of foundations or embankments.
Fig. 6 Settlement-time relationship at r = 1 m under different α values
Fig. 5 Settlement-time relationship at r = 1 m under different η values
Settlement analysis of viscoelastic foundation under vertical line load using a fractional Kelvin-Voigt model 77
4.3 Determination of the influential distance
The determination of influential distance due to loading is another major engineering concern. Fig.
7 shows the settlement results when the value of r0 is set to 10, 15 and 20 m while α is fixed to be
0.5. It is shown that with the increase of r0, the time-dependent and ultimate settlement (t = 1000 d)
will increase but the settlement profile is not affected significantly. If the influential distance of
settlement is overestimated (like r0 = 20 m), a larger settlements will be predicted.
5. Conclusions
Based on the theory of viscoelasticity and fractional calculus, a fractional Kelvin-Voigt viscoelastic
model is developed to account for the time-dependent behaviour of soil foundations under
concentrated line load. Analytical solution of displacements in the foundation was derived using
Laplace transforms. The fractional theory shows its potential in modeling the long-term foundation
deformation. The presented analytical procedure can be extended to solve other geotechnical problems.
From a parametric study, the fractional order of derivative differential operator and the coefficient of
viscosity exert a significant influence on the time-dependent settlements of foundation. Compared
with the KVM, the FKVM can accurately capture the actual deformation of soil foundations under
loading.
However, the empirical relationship between the model parameters and soil properties is an
important issue to be solved, which requires the support of abundant measurement data from
laboratory and field experiments. Careful attention should be paid to the predetermined influential
distance, as well.
Fig. 7 Settlement distributions at the foundation surface under different predetermined r0 values
78 Hong-Hu Zhu, Lin-Chao Liu, Hua-Fu Pei and Bin Shi
Acknowledgements
The financial supports provided by the National Basic Research Program of China (973 Program)
(Grant No. 2011CB710605) and National Key Technology R&D Program of China (2012BAK10B05)
are gratefully acknowledged.
References
Atanackovic, T. M. and Stankovic B. (2004), “Stability of an elastic rod on a fractional derivative type of
foundation”, J. Sound Vib., 277(1-2), 149-161.
Bagley, R. L. and Torvik, P. J. (1983), “Fractional calculus-a different approach to the analysis of viscoelastically
damped structures”, Am. Inst. Aeronaut. Astronaut. J., 21(5), 741-748.
Bagley, R. L. and Torvik, P. J. (1986), “On the fractional calculus model of viscoelastic behavior”, J. Rheol.,
30(1), 133-155.
Bjerrum, L. (1967), “Engineering geology of Norwegian normally-consolidated marine clays as related to
settlement of buildings”, Géotechnique, 17(2), 81-118.
Christie, I. F. (1964), “A re-appraisal of Merchant's contribution to the theory of consolidation”, Géotechnique,
14(4), 309-320.
Dikmen, Ü. (2005), “Modeling of seismic wave attenuation in soil structures using fractional derivative scheme”,
J. Balkan Geophys. Soc., 8(4), 175-188.
Gemant, A. (1936), “A method of analyzing experimental results obtained from elasto-viscous bodies”, J. Appl.
Phys., 7, 311-317.
Justo, J. L. and Durand, P. (2000), “Settlement-time behaviour of granular embankments”, Int. J. Numer. Anal.
Meth. Geomech., 24(3), 281-303.
Kaliakin, V. N. and Dafalias, Y. F. (1990), “Theoretical aspects of the elastoplastic–viscoplastic bounding surface
model for cohesive soils”, Soil Found., 30(3), 11–24.
Li, G. G., Zhu, Z. Y. and Cheng, C. J. (2001), “Dynamical stability of viscoelastic column with fractional
derivative constitutive relation”, Appl. Math. Mech. -Engl. Ed., 22(3), 294-303.
Liu, L. C., Yan, Q. F. and Sun, H. Z. (2006), “Study on model of rheological property of soft clay”, Rock Soil
Mech., 27(S1), 214-217. (in Chinese)
Miller, K. S. and Ross, B. (1993), An Introduction to the Fractional Calculus and Fractional Differential
Equations, Wiley, New York .
Padovan, J. (1987), “Computational algorithms for FE formulations involving fractional operators”, Comput.
Mech., 2(4), 271-287.
Paola, M. Di., Marino, F. and Zingales M. (2009), “A generalized model of elastic foundation based on long-
range interactions: Integral and fractional model”, Int. J. Solids Struct., 46(17), 3124-3137.
Rossikhin, Y. A. and Shitikova M. V. (2010), “Application of fractional calculus for dynamic problems of solid
mechanics: novel trends and recent results”, Appl. Mech. Rev., 63(1), 010801, 1-52.
Timoshenko, S. P. and Goodier, J. N. (1970), Theory of Elasticity, McGraw-Hill, New York.
Welch, S. W. J., Rorrer, R. A. L. and Ronald, G. D. Jr. (1999), “Application of time-based fractional calculus
methods to viscoelastic creep and stress relaxation of materials”, Mech. Time-Depend. Mater., 3(3), 279-303.
Yin, J. H. and Graham, J. (1994), “Equivalent times and one-dimensional elastic viscoplastic modelling of time-
dependent stress-strain behaviour of clays”, Can. Geotech. J., 31(1), 42-52.
Zhang, W. and Shimizu, N. (1998), “Numerical algorithm for dynamic problems involving fractional operators”,
J. Soc. Mech. Eng. Int. J. Ser. C., 41(3), 364-370.
Zhu, H. H., Liu, L. C. and Ye, X. W. (2011), “Response of a loaded rectangular plate on fractional derivative
viscoelastic foundation”, J. Basic Sci. Eng., 19(2), 271-278. (in Chinese)
CC

More Related Content

DOC
Paper sanjay tiwari
PDF
Analysis of piled rafts
PDF
Compósitos de Polimeros - Artigo
PDF
10 simple mathematical approach for granular fill
PDF
Gas foil bearing analysis and the effect of bump foil thickness on its perfor...
PDF
IRJET- Design of Girder using Thin Beam Walled Theory with Influence of S...
PDF
A Rapid Location Independent Full Tensor Gravity Algorithm
Paper sanjay tiwari
Analysis of piled rafts
Compósitos de Polimeros - Artigo
10 simple mathematical approach for granular fill
Gas foil bearing analysis and the effect of bump foil thickness on its perfor...
IRJET- Design of Girder using Thin Beam Walled Theory with Influence of S...
A Rapid Location Independent Full Tensor Gravity Algorithm

What's hot (18)

DOC
Papa 1
PDF
Fractrue
PDF
Research 2
PDF
GaripovTT_Ecmor13
PDF
Numerical simulation of laterally loaded pile
PDF
Jresv6n5p857 a2b
PDF
Problem for the gravitational field
PDF
Rock mass properties
PDF
Pseudo static passive response of retaining wall supporting c-φ backfill
PDF
2013 siryabe - holes effects in plane periodic multilayered viscoelastic media
PDF
DOC
Fixity depth of offshore piles in elastoplastic soft clay under dynamic load
PDF
Structur e very helpfull
PDF
Fundamentals of geophysical hydrodynamics
PDF
Some Properties of M-projective Curvature Tensor on Generalized Sasakian-Spac...
PDF
Pc x-2019 adebayo-56
PDF
GJI_Saul_Lumley13
PDF
Optimization of 3 d geometrical soil model for multiple footing resting on sand
Papa 1
Fractrue
Research 2
GaripovTT_Ecmor13
Numerical simulation of laterally loaded pile
Jresv6n5p857 a2b
Problem for the gravitational field
Rock mass properties
Pseudo static passive response of retaining wall supporting c-φ backfill
2013 siryabe - holes effects in plane periodic multilayered viscoelastic media
Fixity depth of offshore piles in elastoplastic soft clay under dynamic load
Structur e very helpfull
Fundamentals of geophysical hydrodynamics
Some Properties of M-projective Curvature Tensor on Generalized Sasakian-Spac...
Pc x-2019 adebayo-56
GJI_Saul_Lumley13
Optimization of 3 d geometrical soil model for multiple footing resting on sand
Ad

Similar to 1 viscoelastic foundation on soil under vertical load (6)

PDF
COMPUTATION OF HARMONIC GREEN’S-FUNCTIONS OF A HOMOGENEOUS SOIL USING AN AXIS...
PDF
30120140504021 2
PDF
Soil structure interaction effect on dynamic behavior of 3 d building frames ...
PDF
Effect of Soil Flexibility on Analysis and Design of Building
PPT
7 vertical stresses below applied loads (1)
PPT
sm2-lec-3_soil_stresses advanced geotechnical.ppt
COMPUTATION OF HARMONIC GREEN’S-FUNCTIONS OF A HOMOGENEOUS SOIL USING AN AXIS...
30120140504021 2
Soil structure interaction effect on dynamic behavior of 3 d building frames ...
Effect of Soil Flexibility on Analysis and Design of Building
7 vertical stresses below applied loads (1)
sm2-lec-3_soil_stresses advanced geotechnical.ppt
Ad

More from Jeorge Esrom Chambi (8)

PDF
Suarez burgoa2018book
PDF
Tabla de-dosificaciones-y-equivalencias1
PDF
08 isoyetas 100 años
PDF
06 isoyetas 25 años
PDF
07 isoyetas 50 años
PDF
09 estaciones
PDF
Informe rev03
Suarez burgoa2018book
Tabla de-dosificaciones-y-equivalencias1
08 isoyetas 100 años
06 isoyetas 25 años
07 isoyetas 50 años
09 estaciones
Informe rev03

Recently uploaded (20)

PDF
86236642-Electric-Loco-Shed.pdf jfkduklg
PDF
COURSE DESCRIPTOR OF SURVEYING R24 SYLLABUS
PPTX
Artificial Intelligence
PDF
Level 2 – IBM Data and AI Fundamentals (1)_v1.1.PDF
PPTX
Nature of X-rays, X- Ray Equipment, Fluoroscopy
PDF
The CXO Playbook 2025 – Future-Ready Strategies for C-Suite Leaders Cerebrai...
PDF
Unit I ESSENTIAL OF DIGITAL MARKETING.pdf
PDF
Soil Improvement Techniques Note - Rabbi
PDF
Integrating Fractal Dimension and Time Series Analysis for Optimized Hyperspe...
PPTX
UNIT 4 Total Quality Management .pptx
PPT
A5_DistSysCh1.ppt_INTRODUCTION TO DISTRIBUTED SYSTEMS
PDF
Visual Aids for Exploratory Data Analysis.pdf
PPT
Occupational Health and Safety Management System
PPT
INTRODUCTION -Data Warehousing and Mining-M.Tech- VTU.ppt
PDF
737-MAX_SRG.pdf student reference guides
PDF
null (2) bgfbg bfgb bfgb fbfg bfbgf b.pdf
PDF
Automation-in-Manufacturing-Chapter-Introduction.pdf
PPTX
introduction to high performance computing
PPT
Introduction, IoT Design Methodology, Case Study on IoT System for Weather Mo...
PDF
Mitigating Risks through Effective Management for Enhancing Organizational Pe...
86236642-Electric-Loco-Shed.pdf jfkduklg
COURSE DESCRIPTOR OF SURVEYING R24 SYLLABUS
Artificial Intelligence
Level 2 – IBM Data and AI Fundamentals (1)_v1.1.PDF
Nature of X-rays, X- Ray Equipment, Fluoroscopy
The CXO Playbook 2025 – Future-Ready Strategies for C-Suite Leaders Cerebrai...
Unit I ESSENTIAL OF DIGITAL MARKETING.pdf
Soil Improvement Techniques Note - Rabbi
Integrating Fractal Dimension and Time Series Analysis for Optimized Hyperspe...
UNIT 4 Total Quality Management .pptx
A5_DistSysCh1.ppt_INTRODUCTION TO DISTRIBUTED SYSTEMS
Visual Aids for Exploratory Data Analysis.pdf
Occupational Health and Safety Management System
INTRODUCTION -Data Warehousing and Mining-M.Tech- VTU.ppt
737-MAX_SRG.pdf student reference guides
null (2) bgfbg bfgb bfgb fbfg bfbgf b.pdf
Automation-in-Manufacturing-Chapter-Introduction.pdf
introduction to high performance computing
Introduction, IoT Design Methodology, Case Study on IoT System for Weather Mo...
Mitigating Risks through Effective Management for Enhancing Organizational Pe...

1 viscoelastic foundation on soil under vertical load

  • 1. Geomechanics and Engineering, Vol. 4, No. 1 (2012) 67-78 67 Settlement analysis of viscoelastic foundation under vertical line load using a fractional Kelvin-Voigt model Hong-Hu Zhu*1 , Lin-Chao Liu2 , Hua-Fu Pei3 and Bin Shi1 1 School of Earth Sciences and Engineering, Nanjing University, Nanjing, China 2 School of Civil Engineering, Xinyang Normal University, Xinyang, China 3 Department of Civil and Structural Engineering, The Hong Kong Polytechnic University, Hong Kong, China (Received January 23, 2011, Revised February 19, 2012, Accepted March 05, 2012) Abstract. Soil foundations exhibit significant creeping deformation, which may result in excessive settlement and failure of superstructures. Based on the theory of viscoelasticity and fractional calculus, a fractional Kelvin-Voigt model is proposed to account for the time-dependent behavior of soil foundation under vertical line load. Analytical solution of settlements in the foundation was derived using Laplace transforms. The influence of the model parameters on the time-dependent settlement is studied through a parametric study. Results indicate that the settlement-time relationship can be accurately captured by varying values of the fractional order of differential operator and the coefficient of viscosity. In comparison with the classical Kelvin-Voigt model, the fractional model can provide a more accurate prediction of long-term settlements of soil foundation. The determination of influential distance also affects the calculation of settlements. Keywords: soil foundation; fractional viscoelastic model; the Flamant-Boussinesq solution; settlement; Laplace transform. 1. Introduction Research results show that the soil foundation exhibits significant creeping behaviour and the ground deformation under loading from superstructures is highly time-dependent (Bjerrum 1967). For the estimation of the increase in stresses at various points and associated displacements caused in soil foundations due to external loading, there have been a number of viscoelastic, elastic-visco- plastic (EVP) or elastoplastic-viscoplastic models (e.g., Christie 1964, Kaliakin and Dafalias 1990, Yin and Graham 1994, Justo and Durand 2000). Among them, viscoelastic foundation models can provide satisfactory simulation of the time-dependent behaviour when the stress level in soil is low and yielding of soil is thus not reached. The ideal assumption of the theory of viscoelasticity, namely that the foundation is homogeneous, viscoelastic, and isotropic, is not quite true in most cases. However, it provides a close estimation of long-term settlement and failure possibility of superstructures and, using proper safety factors, a safe design can be developed. Because integer-order differential operators are used in the viscoelastic constitutive equations, the *Corresponding author, Associate Professor, E-mail: zhh@nju.edu.cn Technical Note
  • 2. 68 Hong-Hu Zhu, Lin-Chao Liu, Hua-Fu Pei and Bin Shi kernel functions (creep modulus and relaxation modulus) of viscoelastic models are normally a combination of exponential functions. Thus these models have difficulties in predicting settlement of soil foundation accurately and their applicable range is significantly limited. Fractional constitutive models of viscoelastic materials were first introduced by Gemant (1936). In the constitutive equations, integer-order differential operators were replaced by fractional-order differential operators. These models have shown powerful capability in capturing static and dynamic stress-strain-time relationships of viscoelastic materials and have been applied in various fields in the past few decades (Bagley and Torvik 1983 and 1986, Welch et al. 1999, Zhang and Shimizu 1998, Li et al. 2001, Rossikhin and Shitikova 2010). However, due to the complexity involved in obtaining the analytical solution, these fractional models have not been applied in the field of geotechnical engineering until recent years (Atanackovic and Stankovic 2004, Dikmen 2005, Liu et al. 2006, Paola et al. 2009, Zhu et al. 2011). In this paper, a fractional Kelvin-Voigt viscoelastic constitutive model was proposed to study time- dependent settlement of soil foundations. The analytical solution of a half-space under a vertical line load is derived using Laplace transforms. The influences of the model parameters and the predetermined influential distance on the calculated settlements at the foundation surface are further explored using a parametric study. 2. Fractional Kelvin-Voigt model The generalized viscoelastic constitutive equation using fractional derivatives can be described by (Padovan 1987) (1) where m and n are positive integers; and are Riemann-Liouville fractional differential operators defined by (Miller and Ross 1993) (2) where Γ(u) is the Gamma function, . If m = n = 1 and b1 = 0, Eq. (1) can be expressed by (3) It is obvious that if α = 1, Eq. (3) represents the constitutive equation of the classical Kelvin-Voigt viscoelastic model (KVM). Thus this new model can be called a generalized fractional Kelvin-Voigt model (FKVM), in which the integer-order differential operator is now replaced by a fractional- order one. In the theory of viscoelasticity, the stress-strain relationships are (4) σ t( ) + m 1= M ∑ bm d βm dt βm ---------σ t( ) = E0ε t( ) + n 1= N ∑ En d αn dt αn --------ε t( ) 0 βm 1,0 αn 1< < < <( ) d βm /dt βm d αn /dt αn d α f t( ) dt α -------------- = 1 Γ 1 α–( ) -------------------- d dt ---- 0 t ∫ f τ( ) t τ–( ) α ----------------dτ 0 α 1< <( ) Γ u( ) = 0 ∞ ∫ t u 1– e u– dt σ t( ) = E0ε t( ) + E1 d α ε t( ) dt α --------------- P′Sij t( ) = Q′eij t( )
  • 3. Settlement analysis of viscoelastic foundation under vertical line load using a fractional Kelvin-Voigt model 69 (5) where Sij(t) and σij(t) are deviatoric and hydrostatic stress tensors; eij(t) and εij(t) are deviatoric and hydrostatic strain tensors; P', Q', P'' and Q'' are linear differential operators consisting of viscoelastic parameters. The volumetric and shear strains of foundation soils under low stress levels are assumed to be time-dependent. The FKVM shown in Fig. 1 is thus utilized for the description of the viscoelastic behaviour of the bulk and shear moduli,, which describe the volume and shape changes of soil mass under external loading. Thus P' = 1 (6) (7) P'' = 1 (8) (9) where G, K and η are the shear modulus, the bulk modulus, and the coefficient of viscosity, respectively. Determination of the model parameters can be obtained from laboratory and field testing. 3. Analytical solutions of settlements using fractional Kelvin-Voigt model 3.1 Flamant-Boussinesq elastic solution Fig. 2 shows the case where a line force of p0 per unit length is applied on the surface of a half- space. The elastic solution for this plane-strain problem was developed by Flamant (1892) by modifying the three-dimensional solution of Boussinesq (Timoshenko and Goodier 1970). The Flamant-Boussinesq solution using an elastic model (EM) is widely used in foundation engineering for the calculation of stresses and estimation of displacements. In the polar coordinate system, assume that at a point P( ) on the ground surface, , i.e., the influential distance of ground surface settlement is assumed to be r0. The displacements at any point in the foundation can be obtained as P″σij t( ) = Q″εij t( ) Q′ = G + η d α dt α ------- Q″ = K + η d α dt α ------- r0, π/2 uθ( )θ= π 2 ---,r=r0 = 0 Fig. 1 Three-parameter fractional Kelvin-Voigt model
  • 4. 70 Hong-Hu Zhu, Lin-Chao Liu, Hua-Fu Pei and Bin Shi (10) (11) where E and μ are the modulus of elasticity and Poisson’s ratio, respectively; ur and uθ denote the displacements in the radial and circumferential directions, respectively. Taking E = 144 MPa, μ = 0.2, p0 = 1 MPa and r0 = 15 m, the distributions of displacements within the soil foundation using the elastic model (EM) are depicted in Fig. 3. It is seen that the surface settlement at the loading point approaches infinity and the settlements within the foundation reduce sharply with increasing depth. For locations far away from the loading point, heave of the foundation occurs. In the case of , Eqs. (10) and (11) for radial and circumferential displacement components at the ground surface can be given as the following simple form (12) (13) ur = 2p0 1 μ 2 –( ) πE --------------------------cosθ ln r0 r ---- + 1 2μ–( ) 1 μ+( )p0 πE -----------------------------------------θ sinθ − 1 μ+( )p0 πE ---------------------cosθ uθ = 2p0 1 μ 2 –( ) πE --------------------------– sinθ ln r0 r ---- + 1 2μ–( ) 1 μ+( )p0 πE -----------------------------------------θ cosθ θ = π/2 ur( )θ= π 2 --- = 1 2μ–( ) 1 μ+( )p0 2E ----------------------------------------- = − 3p0 4 3K G+( ) ------------------------ uθ( ) θ= π 2 --- = 2p0 1 μ 2 –( ) πE --------------------------ln r0 r ---- = − p0ln r0 r ---- 2π -------------- 3 3K G+( ) --------------------- 1 G ----+ Fig. 2 Schematic illustration of a foundation subjected to vertical line load at the surface
  • 5. Settlement analysis of viscoelastic foundation under vertical line load using a fractional Kelvin-Voigt model 71 Fig. 3 Displacement distributions calculated from the Flamant-Boussinesq elastic solution
  • 6. 72 Hong-Hu Zhu, Lin-Chao Liu, Hua-Fu Pei and Bin Shi Eqs. (12) and (13) indicate that there is a logarithmic relationship between the distance to the loading point and the vertical displacement, and that the horizontal displacement at the ground surface is a constant that is not affected by r. 3.2 Viscoelastic solution using classical Kelvin-Voigt model Considering a line load applied on the foundation shown in Fig. 2, the loading condition is specified as (14) where H(t) is the unit-step function defined as . To obtain the viscoelastic solution, the volumetric and shear strains of the foundation soil satisfies the KVM. Thus Eqs. (7) and (9) turns into (15) (16) Taking the Laplace transforms of Eqs. (6), (8), (15) and (16), and combining them with Eqs. (12) and (13), the displacements at the ground surface can be expressed by (17) (18) Taking the inverse Laplace transforms of Eqs. (17) and (18), we get (19) (20) where , . Eqs. (19) and (20) show that the time-dependent displacements are exponential creep functions. It is obvious that if , the ultimate displacements equal the Flamant-Boussinesq solution using elastic model (EM). 3.3 Viscoelastic solution using fractional Kelvin-Voigt model Taking the Laplace transforms of Eqs. (6) to (9) and combined with Eqs. (12) and (13), the p = p0H t( ) H t( ) = 1 t 0≥ 0 t 0<⎩ ⎨ ⎧ Q′ = G + η d dt ---- Q″ = K + η d dt ---- ur( )θ= π 2 --- = − 3p0 4 -------- 1 3K G 4ηs+ +( )s --------------------------------------- uθ( )θ= π 2 --- = − p0ln r0 r ---- 2π -------------- 3 3K G 4ηs+ +( )s --------------------------------------- 1 G ηs+( )s -----------------------+ ur( )θ= π 2 --- = − 3p0 4 3K G+( ) ------------------------ 1 e t τ1 ----– – ⎝ ⎠ ⎜ ⎟ ⎛ ⎞ uθ( ) θ= π 2 --- = − p0ln r0 r ---- 2π -------------- 3 3K G+( ) --------------------- 1 e t τ1 ----– – ⎝ ⎠ ⎜ ⎟ ⎛ ⎞ 1 G ---- 1 e t τ2 ----– – ⎝ ⎠ ⎜ ⎟ ⎛ ⎞ + τ1 = 4η/ 3K G+( ) τ2 = η/G t ∞→
  • 7. Settlement analysis of viscoelastic foundation under vertical line load using a fractional Kelvin-Voigt model 73 displacements at the ground surface can be expressed by (21) (22) In order to take the inverse Laplace transform, note the following characteristics of the fractional derivative in the Laplace transform for zero initial condition (23) where Eα is the Mittag-Leffler function defined as . In view of Eq. (23), taking the inverse Laplace transforms of Eqs. (21) and (22), we get (24) (25) where , . It is obvious that when , , . Thus the ultimate displacements equal the elastic solution. If , . Thus Eqs. (24) and (25) turn into (26) (27) Eqs. (26) and (27) indicate that the displacements at the ground surface are not time-dependent any more and the displacement values are half of those from the elastic solution. Therefore, the FKVM collapses to be a ideal elastic model as . ur( ) θ= π 2 --- = − 3p0 4 -------- 1 3K G 4ηs α + +( )s ----------------------------------------- uθ( )θ= π 2 --- = − p0ln r0 r ---- 2π -------------- 3 3K G 4ηs α + +( )s ----------------------------------------- 1 G ηs α +( )s -------------------------+ L 1– 1 G ηs α +( )s ------------------------- = 1 G ---- 1 Eα– t τ --- ⎝ ⎠ ⎛ ⎞ α – ⎩ ⎭ ⎨ ⎬ ⎧ ⎫ Eα t( ) = 0 ∞ ∑ t n Γ 1 α n+( ) -------------------------; τ = η G ---- ur( ) θ= π 2 --- = − 3p0 4 3K G+( ) ------------------------ 1 Eα– t τ 1 ----- ⎝ ⎠ ⎛ ⎞ α – ⎩ ⎭ ⎨ ⎬ ⎧ ⎫ uθ( ) θ= π 2 --- = − p0ln r0 r ---- 2π -------------- 3 3Eα t τ 1 ------ ⎝ ⎠ ⎛ ⎞ α –– 3K G+( ) ---------------------------------------- 1 Eα– t τ 2 ------ ⎝ ⎠ ⎛ ⎞ α – G -------------------------------------+ ⎩ ⎭ ⎪ ⎪ ⎨ ⎬ ⎪ ⎪ ⎧ ⎫ τ1 = 4η 3K G+ ------------------ τ2 = η G ---- 0 α< 1≤ limt ∞→ ur( ) θ= π 2 --- = 3p0 4 3K G+( ) --------------------------– limt ∞→ uθ( ) θ= π 2 --- = p0ln r0 r ---- 2π ---------------– 3 3K G+( ) ---------------------- 1 G ----+ α 0→ Eα t( ) 0 ∞ ∑ t n Γ 1( ) ----------- = 0 ∞ ∑ t n = 1 1 t– ---------- 1– t 1< <( )→ ur( ) θ=π 2 --- = − 3p0 8 3K G+( ) ------------------------ uθ( ) θ= π 2 --- = p0ln r0 r ---- 4π --------------– 3 3K G+ ----------------- 1 G ----+ α 0→
  • 8. 74 Hong-Hu Zhu, Lin-Chao Liu, Hua-Fu Pei and Bin Shi Table 1 Parameters of the FKVM Item Value Young’s modulus E (MPa) 144 Poisson’s ratio μ 0.2 Shear modulus G (MPa) 60 Bulk modulus K (MPa) 80 Coefficient of viscosity η (MPa·d) 1000 Fractional differential order α 0.5 r0 (m) 15 If α = 1, . Obviously, Eqs. (23) and (24) currently become the viscoelastic solution, which means that the FKVM turns into a classical KVM. 4. Numerical example and analysis 4.1 Comparison of elastic and viscoelastic models How to determine the fractional model parameters is a major concern in practical applications. To carry out a preliminary investigation of these parameters’ influence on the calculation results, a parametric study is presented as following. Here, we analyze the quasistatic problem of a viscoelastic foundation subjected to line loading p0 = 1 MPa using the FKVM. The basic parameters used in this analysis are listed in Table 1. Fig. 4 shows the surface settlement results using the fractional model at t = 10 d and 50 d, together with the results from the classical KVM solution and the Flamant-Boussinesq elastic solution. The present results show that the discrepancy between the results from FKVM and KVM increases as t increases. The calculated settlements using KVM develops rapidly with the elapsed time while those using FKVM will take much more time. 4.2 Parametric study of the FKVM To study the influence of the coefficient of viscosity in the FKVM, η is set to 500, 1000 and 1500 MPa·d. Fig. 5 depicts the settlement-time curves at r = 1 m over time using the FKVM, the KVM and the EM. It is shown that the coefficient of viscosity mainly affects the overall creeping rate of the foundation soil. For the same time point, the settlements increase with the decrease of η. Note that the ultimate settlement in the viscoelastic solution is equal to that in the elastic solution, η may determine the magnitude of creeping by controlling the time duration to achieve the ultimate settlement. If η decreases, the creeping effect will function more slowly and the ultimate settlement will be obtained for a much longer time. Results from the classical Kelvin-Voigt model (α = 1), which is shown in Fig. 6 by a dashed line, shows larger settlements and less time to get to the ultimate settlement. Eα t( ) = 0 ∞ ∑ t n Γ 1 αn+( ) ------------------------- = 0 ∞ ∑ t n n! ----- = e t
  • 9. Settlement analysis of viscoelastic foundation under vertical line load using a fractional Kelvin-Voigt model 75 To investigate the influence of the fractional order of differential operator in the FKVM, α is set to 0.1, 0.3, 0.5, 0.7 and 0.9 while η is fixed to 1000 MPa·d. Fig. 6 illustrates the settlement at r = 1 m over time. It is shown that the fractional order α has a significant influence on the long-term performance of the foundation surface settlements. At t1/2 = 10 d, the settlements are half of the ultimate settlement (5.75 mm) for any value of α. t1/2 is found to have an approximately linear relationship with η and 1/E. After that time point, with the increase of α, the foundation settlement and corresponding settlement rate increase. Therefore, α may determine the functioning law of creeping effect by separating the time-dependent deformation of soil foundation into two stages. The first stage may associate with the consolidation process of the foundation soil and the second stage is mainly due to secondary compression. Therefore, a small α value is fit for sand foundation, which has a small permeability coefficient. While for soft clay foundation, the α value should be Fig. 4 Comparsion of settlements calculated using elastic, Kelvin-Voigt and fractional Kelvin-Voigt models
  • 10. 76 Hong-Hu Zhu, Lin-Chao Liu, Hua-Fu Pei and Bin Shi increased. In the KVM where α = 1 (the dashed line in Fig. 6), a minimum settlement rate at stage I and a maximum settlement rate at stage II is obtained. The key feature of the FKVM illustrated in Figs. 5 and 6 is the capability to represent generalised viscoelastic behaviour over many decades of time with only three parameters. It is also indicated from the above parametric study that the classical viscoelastic model may overestimate the development of long-term settlement (Stage II) and fail to describe the attenuation creep of soil mass accurately. Using the fractional viscoelastic models, we can obtain the best fitting effect of the actual soil deformation characteristics by selecting appropriate values of the fractional differential order and the coefficient of viscosity. In geotechnical design, this model can provide more accurate settlement prediction and thus save construction costs of foundations or embankments. Fig. 6 Settlement-time relationship at r = 1 m under different α values Fig. 5 Settlement-time relationship at r = 1 m under different η values
  • 11. Settlement analysis of viscoelastic foundation under vertical line load using a fractional Kelvin-Voigt model 77 4.3 Determination of the influential distance The determination of influential distance due to loading is another major engineering concern. Fig. 7 shows the settlement results when the value of r0 is set to 10, 15 and 20 m while α is fixed to be 0.5. It is shown that with the increase of r0, the time-dependent and ultimate settlement (t = 1000 d) will increase but the settlement profile is not affected significantly. If the influential distance of settlement is overestimated (like r0 = 20 m), a larger settlements will be predicted. 5. Conclusions Based on the theory of viscoelasticity and fractional calculus, a fractional Kelvin-Voigt viscoelastic model is developed to account for the time-dependent behaviour of soil foundations under concentrated line load. Analytical solution of displacements in the foundation was derived using Laplace transforms. The fractional theory shows its potential in modeling the long-term foundation deformation. The presented analytical procedure can be extended to solve other geotechnical problems. From a parametric study, the fractional order of derivative differential operator and the coefficient of viscosity exert a significant influence on the time-dependent settlements of foundation. Compared with the KVM, the FKVM can accurately capture the actual deformation of soil foundations under loading. However, the empirical relationship between the model parameters and soil properties is an important issue to be solved, which requires the support of abundant measurement data from laboratory and field experiments. Careful attention should be paid to the predetermined influential distance, as well. Fig. 7 Settlement distributions at the foundation surface under different predetermined r0 values
  • 12. 78 Hong-Hu Zhu, Lin-Chao Liu, Hua-Fu Pei and Bin Shi Acknowledgements The financial supports provided by the National Basic Research Program of China (973 Program) (Grant No. 2011CB710605) and National Key Technology R&D Program of China (2012BAK10B05) are gratefully acknowledged. References Atanackovic, T. M. and Stankovic B. (2004), “Stability of an elastic rod on a fractional derivative type of foundation”, J. Sound Vib., 277(1-2), 149-161. Bagley, R. L. and Torvik, P. J. (1983), “Fractional calculus-a different approach to the analysis of viscoelastically damped structures”, Am. Inst. Aeronaut. Astronaut. J., 21(5), 741-748. Bagley, R. L. and Torvik, P. J. (1986), “On the fractional calculus model of viscoelastic behavior”, J. Rheol., 30(1), 133-155. Bjerrum, L. (1967), “Engineering geology of Norwegian normally-consolidated marine clays as related to settlement of buildings”, Géotechnique, 17(2), 81-118. Christie, I. F. (1964), “A re-appraisal of Merchant's contribution to the theory of consolidation”, Géotechnique, 14(4), 309-320. Dikmen, Ü. (2005), “Modeling of seismic wave attenuation in soil structures using fractional derivative scheme”, J. Balkan Geophys. Soc., 8(4), 175-188. Gemant, A. (1936), “A method of analyzing experimental results obtained from elasto-viscous bodies”, J. Appl. Phys., 7, 311-317. Justo, J. L. and Durand, P. (2000), “Settlement-time behaviour of granular embankments”, Int. J. Numer. Anal. Meth. Geomech., 24(3), 281-303. Kaliakin, V. N. and Dafalias, Y. F. (1990), “Theoretical aspects of the elastoplastic–viscoplastic bounding surface model for cohesive soils”, Soil Found., 30(3), 11–24. Li, G. G., Zhu, Z. Y. and Cheng, C. J. (2001), “Dynamical stability of viscoelastic column with fractional derivative constitutive relation”, Appl. Math. Mech. -Engl. Ed., 22(3), 294-303. Liu, L. C., Yan, Q. F. and Sun, H. Z. (2006), “Study on model of rheological property of soft clay”, Rock Soil Mech., 27(S1), 214-217. (in Chinese) Miller, K. S. and Ross, B. (1993), An Introduction to the Fractional Calculus and Fractional Differential Equations, Wiley, New York . Padovan, J. (1987), “Computational algorithms for FE formulations involving fractional operators”, Comput. Mech., 2(4), 271-287. Paola, M. Di., Marino, F. and Zingales M. (2009), “A generalized model of elastic foundation based on long- range interactions: Integral and fractional model”, Int. J. Solids Struct., 46(17), 3124-3137. Rossikhin, Y. A. and Shitikova M. V. (2010), “Application of fractional calculus for dynamic problems of solid mechanics: novel trends and recent results”, Appl. Mech. Rev., 63(1), 010801, 1-52. Timoshenko, S. P. and Goodier, J. N. (1970), Theory of Elasticity, McGraw-Hill, New York. Welch, S. W. J., Rorrer, R. A. L. and Ronald, G. D. Jr. (1999), “Application of time-based fractional calculus methods to viscoelastic creep and stress relaxation of materials”, Mech. Time-Depend. Mater., 3(3), 279-303. Yin, J. H. and Graham, J. (1994), “Equivalent times and one-dimensional elastic viscoplastic modelling of time- dependent stress-strain behaviour of clays”, Can. Geotech. J., 31(1), 42-52. Zhang, W. and Shimizu, N. (1998), “Numerical algorithm for dynamic problems involving fractional operators”, J. Soc. Mech. Eng. Int. J. Ser. C., 41(3), 364-370. Zhu, H. H., Liu, L. C. and Ye, X. W. (2011), “Response of a loaded rectangular plate on fractional derivative viscoelastic foundation”, J. Basic Sci. Eng., 19(2), 271-278. (in Chinese) CC