Professional Documents
Culture Documents
YUANQIANG CAI
HONGLEI SUN
Elsevier
Radarweg 29, PO Box 211, 1000 AE Amsterdam, Netherlands
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom
50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States
© 2017 Elsevier Inc. All rights reserved.
No part of this publication may be reproduced or transmitted in any form or by any means, electronic
or mechanical, including photocopying, recording, or any information storage and retrieval system,
without permission in writing from the publisher. Details on how to seek permission, further information
about the Publisher’s permissions policies and our arrangements with organizations such as the
Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website:
www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright by the Publisher
(other than as may be noted herein).
Notices
Knowledge and best practice in this field are constantly changing. As new research and experience broaden
our understanding, changes in research methods, professional practices, or medical treatment may
become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in evaluating and
using any information, methods, compounds, or experiments described herein. In using such information
or methods they should be mindful of their own safety and the safety of others, including parties for
whom they have a professional responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume
any liability for any injury and/or damage to persons or property as a matter of products liability, negligence
or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in
the material herein.
All structures made by human beings have to be placed on or in the soil. At the very
beginning, these soil foundations are only subjected to static loads, which are loads that
build up gradually over time or with negligible dynamic effects, also known as monotonic
loads. In the 1930s, due to the rapid development of machine manufacturing and trans-
portation industries, the dynamic interaction between structural foundations and the
underlying soil behavior under the action of cyclic loads started to receive considerable
attention in a number of engineering fields. Cyclic loads are loads which exhibit a degree
of regularity both in their magnitude and frequency. Stress reversals, rate effects and
dynamic effects are the important factors that distinguish cyclic loads from static loads.
Practically speaking, no real cyclic loads exist in nature; however, many kinds of loads
can be simplified into cyclic loads for the convenience of study, analysis and design. For
example, the operation of a reciprocating or a rotary machine typically produces a cyclic
load. The passing of a long train can be considered a cyclic load. Even cars running on a
road on the same line can be simplified as a cyclic load. Cyclic loads acting on the struc-
tures and soil can produce elastic waves in the ground which will act on the surrounding
foundations and soil. These actions can cause environmental and safety problems. As a
result, it is very important to take a deep look into this area to advance the knowledge
regarding the theory of vibrations, the principles of wave propagation, and numerical
methods in finding appropriate solutions for problems of practical interest.
In the past few decades, many studies have been carried out on soil-structure inter-
actions under cyclic loads. Most of them have treated the soil as an elastic or viscoelastic
medium. However, there is underground water in what is considered soil medium, such
that the soil is actually a two-phase medium. Biot [1] pioneered the development of
an elastodynamic theory for a fluid-filled elastic porous medium. Since its publication,
Biot’s theory has had wide applications in the geotechnical professions for analyzing wave
propagation characteristics under cyclic loads. The aim of this book is to provide a tutorial
and a state-of-the-art compilation of the advances in the applications of Biot’s theory.
vii
CHAPTER 1
2 1 1 3
ε γ γ
2 3 6 xx 2 xy 2 zx 7
σ xx τxy τzx 6 7
6 1 7
4 τxy σ yy τyz 5 and 6 1 γ εyy γ 7
6 2 xy 2 yz 7
τzx τyz σ zz 6 7
4 5
1 1
γ γ yz εzz
2 zx 2
2 1 1 3
εrr γ rθ γ zr
2 3 6 2 2 7
σ rr τrθ τzr 6 7
6 1 7
4 τrθ σ θθ τθz 5 and 6 1 γ εθθ γ 7
6 2 rθ 2 θz 7
τzr τθz σ zz 6 7
4 5
1 1
γ γ θz εzz
2 zr 2
where σ ij is the stress tensor, τij is the shear stress, γ ij is the shear strain, and εij is the
strain tensor.
Solutions for Biot’s Poroelastic Theory in Key Engineering Fields © 2017 Elsevier Inc.
http://dx.doi.org/10.1016/B978-0-12-812649-3.00001-0 All rights reserved. 1
2 Solutions for Biot's Poroelastic Theory in Key Engineering Fields
Table 1.1.1 Direction cosines between coordinate axes in Cartesian and cylindrical coordinates
x y z
r cos θ sin θ 0
θ sin θ cos θ 0
z 0 0 1
The transformation rule between the stresses and strains in different coordinates is
σ p0 q0 ¼ lp0 i lq0 j σ ij (1.1.1)
εp0 q0 ¼ lp0 i lq0 j εij (1.1.2)
where σ ij denotes the stresses in Cartesian coordinates, σ p0 q0 stands for the stresses in a new
Cartesian coordinate system after rotation, and lp0 i and lq0 j represent the direction cosines
between two coordinate axes.
Combining Eq. (1.1.1) with Table 1.1.1, the relationships between the stresses in
cylindrical coordinates and Cartesian coordinates can be easily derived:
σ r ¼ σ x cos 2 θ + σ y sin 2 θ + 2τxy sin θ cos θ
σ α ¼ σ x sin 2 θ + σ y cos 2 θ 2τxy sin θ cos θ
τrα ¼ σ y σ x sin θ cos θ + τxy cos 2 θ sin 2 θ
(1.1.3)
τzr ¼ τzx cos θ + τyz sin θ
ταz ¼ τzx sin θ + τyz cos θ
σz ¼ σz
ε ¼ E T ðrÞu (1.1.7)
@σ r 1 @τrθ @τzr σ r σ θ @ 2 ur
+ + + ¼ρ 2
@r r @θ @z r @t
@τrθ 1 @σ θ @τθz 2τrθ @ 2 uθ
+ + + ¼ρ 2 (1.1.10)
@r r @θ @z r @t
@τzr 1 @τθz @σ z τzr @ 2 uz
+ + + ¼ρ 2
@r r @θ @z r @t
The tensor form of the equations of motion in Cartesian coordinates can be written as
where the dots above the symbols denote partial differentiation with respect to time t.
where cijkl are components of a fourth-rank tensor including 81 components. Since the
stress vectors are symmetric, the exchange of the indices i and j does not alter the result.
Noting that the strain vectors are symmetric as well, the same process can be done to the
indices k and l; and then we have the relationships
In addition, since we are considering the adiabatic process, we still have the following
relationship:
cijkl ¼ cklij
Thus, among the 81 components of cijkl, the maximum number of independent ones is
21. For a homogeneous medium, the number of independent components goes down to
2, which are the Lame constant λ and μ, or Young’s modulus E and Poisson’s ratio ν. And
the tensor form of Hooke’s law can be simplified as
two kinds of longitudinal waves (P1 and P2 waves) and one kind of rational wave (the
S wave) exist in the saturated porous medium. The existence of the P2 wave always dis-
tinguishes the two-phase medium from the single-phase one. The basic variables of Biot’s
theory are always the solid skeleton displacement (u) and the average displacement of
pore fluid relative to the solid skeleton (w).
Considering the concept of effective stress of the saturated mixture, the relationship
between effective stress, total stress and pore pressure can be expressed as
where σ ij0 is the effective stress tensor, σ ij is the total stress tensor, δij is the Kronecker delta,
and α is the Biot constant that depends on the geometry of material voids. For the most
part, in soil mechanics problems, α 1 can be assumed. The relationship between total
stress and effective stress becomes
which corresponds to the classical effective stress definition by Terzaghi. Thus the tensor
form of the equations of motion for a fully saturated poroelastic medium becomes (omit-
ting the body force):
where the dots above the symbols denote partial differentiation with respect to time t;
thus u€i is the acceleration of the solid part, wi is the fluid displacement relative to
the solid part, and w€i is the fluid acceleration relative to the solid part. For fully saturated
porous media (no air trapped inside), density is equal to ρ ¼ nρf + ð1 nÞρs , where n is
the porosity, and ρs and ρf are the soil particle and water densities, respectively.
For the pore fluid, the equation of momentum balance can be expressed as
where the parameter b ¼ ρfg/kD, where kD is the Darcy permeability of the soil medium
and g is the gravity; p is the pore water pressure.
According to the classical effective stress definition by Terzaghi in Eq. (1.1.17), the
constitutive Eq. (1.1.13) becomes
The final equation is the mass conservation of the fluid flow, which is expressed by
p_ ¼ αM e_ + M ς_ (1.1.22)
where
ς ¼ wi, i (1.1.23)
Basic Equations and Governing Equations 7
Substituting Eqs. (1.1.21)–(1.1.23) into Eqs. (1.1.19), (1.1.20), a u-w formulation can be
obtained as follows:
μui, jj + λ + α2 M + μ uj, ji + αMwj, ji ¼ ρu€i + ρf w€i (1.1.24)
αMuj, ji + Mwj, ji ¼ ρf u€i + mw€i + bw_ i (1.1.25)
Zienkiewicz et al. [2] proposed a simplified u-p formulation in the context of finite ele-
ment analysis of the liquefaction of saturated sand soil. By neglecting the second time
derivatives of the relative fluid displacement from the original Biot [3] u-w formulation,
the u-p formulation is deduced for reducing the primary variables in the sense of finite
element calculation; there are 3 + 3 and 3 + 1 nodal variables in three-dimensional analysis
for the u-w and u-p formulation, respectively. Moreover, the solid displacement (u) and
the pore fluid pressure ( p) are always the main concern.
By simply neglecting the pore fluid relative acceleration terms, i.e., w€i in Eqs. (1.1.19),
(1.1.20), it is easy to eliminate w_ i using Eqs. (1.1.20), (1.1.22) leaving the ui and p as pri-
mary variables. This simplified formulation is economical and convenient in numerical
analysis, for there are only 3 + 1 nodal variables in the three-dimensional analysis. The
equation set becomes
σ ij, j ¼ ρu€i (1.1.26)
ρf 1
u_i, i + u€i p, i ¼ p=M
_ (1.1.27)
b b ,i
@ur ur @uθ @uz @wr wr @wθ @wz
e¼ + + + , ς¼ + + +
@r r r@θ @z @r r r@θ @z
and r2 denotes the Laplacian operator which is given by
@2 1 @ 1 @2 @2
r2 ¼ + + +
@r 2 r @r r 2 @θ 2
@z2
Γ sD \ Γ tN ¼ ∅ (1.1.31)
>
: w p
ΓD \ ΓN ¼ ∅
CHAPTER 2
By using the Fourier transform given in Eqs. (2.1.1), (1.1.24), (1.1.25), (1.1.21), and
(1.1.22) are transformed from partial differential equations to ordinary differential equa-
tions. The governing equations in the x direction are given as
Solutions for Biot’s Poroelastic Theory in Key Engineering Fields © 2017 Elsevier Inc.
http://dx.doi.org/10.1016/B978-0-12-812649-3.00002-2 All rights reserved. 9
10 Solutions for Biot's Poroelastic Theory in Key Engineering Fields
@ θ~ @ ζ~
μr2 u~x + λ + α2 M + μ αM ¼ ω2 ρ~ux ω2 ρf w~x (2.1.5)
@x @x
@ θ~ @ ζ~
αM M ¼ ω2 ρf u~x ω2 mw~x + iωbw~x (2.1.6)
@x @x
@~ p @ ζ~ @ θ~
¼ M αM (2.1.7)
@x @x @x
Substituting Eq. (2.1.7) into Eq. (2.1.5) yields
@ θ~ @~
p
μr2 u~x + ðλ + μÞ
α f ¼ ω2 ρ~
ux ω2 ρf w~x (2.1.8)
@x @x
Substituting Eq. (2.1.7) into Eq. (2.1.6) gives
@~pf
¼ ω2 ρf u~x ω2 mw~x + iωbw
~x (2.1.9)
@x
and then the following equation can be obtained from Eq. (2.1.9):
@~pf
ω2 ρf u~x
w~x ¼ @x (2.1.10)
iωb ω2 m
Then substituting Eq. (2.1.10) into Eq. (2.1.9) yields
@ θ~ @~
p
μr2 u~x + ðλ + μÞ + ω2 ðρ ϑρf Þ~
ux ðα ϑÞ f ¼ 0 (2.1.11)
@x @x
ω2 ρf
where ϑ ¼ 2 :
ω m iωb
Similarly, the following equations in the y and z direction can be obtained:
@ θ~ @~
p
μr2 u~y + ðλ + μÞ + ω2 ðρ ϑρf Þ~
uy ðα ϑÞ f ¼ 0 (2.1.12)
@y @y
@~
pf
¼ ω2 ρf u~y ω2 mw~y + iωbw~y (2.1.13)
@y
@ θ~ @~
p
μr2 u~z + ðλ + μÞ uz ðα ϑÞ f ¼ 0
+ ω2 ðρ ϑρf Þ~ (2.1.14)
@z @z
@~ pf
¼ ω2 ρ~
uz ω2 mw~z + iωbw~z (2.1.15)
@z
Taking the derivative of Eqs. (2.1.9), (2.1.13), (2.1.15) with respect to x, y, z, respec-
tively, and adding them together yields
r2 p~f ¼ ω2 ρf θ~ + ω2 mζ~ iωbζ~ (2.1.16)
Solutions for Saturated Soil Under Moving Loads 11
d4 p~f 2 ~f 4
2 d p
4
+ β 1 2ξ2
2η 2
+ ξ + η4 + 2ξ2 η2 β1 ξ2 β1 η2 + β2 p~f ¼ 0 (2.1.26)
dz dz
12 Solutions for Biot's Poroelastic Theory in Key Engineering Fields
d2 u~z 2
μ 2
+ S μξ 2
μη 2
u~z ðλ + μÞðγ 1 Aχ 1 eγ1 z + γ 2 Bχ 2 eγ2 z Þ
dz
+ðα ϑÞðγ 1 Aeγ1 z + γ 2 Beγ 2 z Þ ¼ 0 (2.1.35)
where
S2 ¼ ω2 ðρ ϑρf Þ (2.1.36)
Solutions for Saturated Soil Under Moving Loads 13
The general solution for Eq. (2.1.35) is u~z ¼ Aa1 γ 1 eγ1 z + Ba2 γ 2 eγ2 z and the particular
solution is u~z ¼ Ceγ3 z ; then the solution of u~z can be expressed as
χ 1 ðλ + μÞ α + ϑ χ ðλ + μÞ α + ϑ
a1 ¼ , a2 ¼ 2 2 (2.1.39)
S μL1
2 2 S μL22
By applying Fourier transforms with respect to x and y, Eq. (2.1.12) can be expressed as
d2 u~y
μ 2 + S2 μξ2 μη2 u~y + iη½ðλ + μÞðAχ 1 eγ 1 z + Bχ 2 eγ2 z Þ
dz
ðα ϑÞðAeγ1 z + Beγ 2 z Þ ¼ 0 (2.1.40)
and thus
u~y ¼ iηðAa1 eγ 1 z + Ba2 eγ2 z Þ + iDeγ3 z (2.1.41)
Applying Fourier transforms with respect to x and y into θ ¼ ui, i , the following equation
can be obtained:
@~
uz
θ~ ¼ iξ~
ux + iη~
uy + (2.1.42)
@z
By substituting Eqs. (2.1.37), (2.1.41) into Eq. (2.1.42), the solutions for u~x can be
obtained as
i
u~x ¼ χ 1 + a1 ξ2 L12 Aeγ1 z + χ 2 + a2 ξ2 L22 Beγ 2 z + ðCγ 3 + DηÞeγ 3 z
ξ
(2.1.43)
The constitutive equation in saturated soil can be expressed as
@uz
σ z ¼ λθ + 2μ αpf (2.1.44)
@z
@ux @uz
τxz ¼ μ + (2.1.45)
@z @x
@uy @uz
τyz ¼ μ + (2.1.46)
@z @y
14 Solutions for Biot's Poroelastic Theory in Key Engineering Fields
@~uz
σ z ¼ λθ~ + 2μ
e α~
p (2.1.47)
@z f
@~
ux
eτxz ¼ μ u
+ iξμ~ (2.1.48)
@z z
@~
uy
eτyz ¼ μ u
+ iημ~ (2.1.49)
@z z
Substituting Eqs. (2.1.27), (2.1.32), (2.1.37), (2.1.41), (2.1.43) into Eqs. (2.1.47)–(2.1.49)
yields
where
g1 ¼ χ 1 + a1 2ξ2 L12 (2.1.53)
g2 ¼ χ 2 + a2 2ξ2 L22 (2.1.54)
g3 ¼ χ 1 λ 2μa1 γ 21 α (2.1.55)
g4 ¼ χ 2 λ 2μa2 γ 22 α (2.1.56)
P jx Vt j a, jyj b
σ z ðx, y, tÞ ¼ (2.1.57)
0 jx Vt j > a, jyj > b
8 8πP
>
> sin ξa sin ηb δðξV + a0 Þ ξη 6¼ 0
>
> ξη
>
>
>
>
>
>
< 8πP a sin ηb δða0 Þ ξ ¼ 0,η 6¼ 0
e
σ z ðξ, η, a0 Þ ¼ η (2.1.58)
>
>
>
> 8πP
>
> b sin ξa δðξV + a0 Þ ξ 6¼ 0,η ¼ 0
>
> ξ
>
>
:
8πPab δða0 Þ ξ ¼ 0,η ¼ 0
By substituting the boundary condition into Eqs. (2.1.27), (2.1.50)–(2.1.52), and intro-
ducing the auxiliary spatial coordinate xt ¼ x Vt, the solution in the time domain is
obtained:
ð +∞ ð +∞ ð +∞
1
uz ðxt , y, zÞ ¼ e
σ z ðξ, η, ωÞϕðξ, η, z, ωÞeiξx eiηy eia0 t dξdηdω
ð2π Þ3 ∞ ∞ ∞
ð +∞ ð +∞
1
¼ e
σ z ðξ, ηÞϕðξ,η, z, ξV Þeiξxt eiηy dξdη (2.1.59)
ð2π Þ2 ∞ ∞
where
1 2 2
ϕðξ, η, z, α0 Þ ¼ γ 3 + ξ + η2 ðγ 2 a2 eγ2 z γ 1 a1 eγ 1 z Þ + ðγ 1 g5 γ 2 g6 Þeγ3 z
Δ0
(2.1.60)
2 2
Δ0 ¼ ðg4 g3 Þ γ 3 + ξ + η 2μγ 3 ðγ 1 g5 γ 2 g6 Þ
2
(2.1.61)
infinity in the horizontal direction. The model of pavement on the saturated ground is
shown in Fig. 2.2.1.
The Kirchhoff small-deflection thin-plate theory was employed as by Kim and
Roessert [5]. Denoting the reaction of the ground to the plate by F(x, y, t), the governing
equation of the plate is given as follows:
@ 4 w ðx, y, tÞ @ 4 wðx, y, t Þ @ 4 w ðx, y, t Þ @ 2 wðx, y, t Þ
Dp + 2 + + mb ¼ qðx, y, tÞ + F ðx, y, tÞ
@x4 @x2 @y2 @y4 @t 2
(2.2.1)
where w is the vertical displacement of the plate; mb is the mass density of the plate per
unit area; and q(x, y, t) is the traffic load pressure acting on the surface of the road. Dp is the
flexural rigidity of the plate and can be defined as follows:
Eh3
Dp ¼ (2.2.2)
12 1 νp 2
where E, h, and νp are the elastic modulus, thickness, and Poisson’s ratio of the plate,
respectively.
The traffic load is modeled as four rectangular load pressures expressed as:
(
da d1 da + d1 dw d2 dw + d2
qðx, y, tÞ ¼ Q jx Vt j and jyj (2.2.3)
2 2 2 2
0 otherwise
where Q is the magnitude of the load pressure; d1 and d2 are the loaded lengths of a tire
print in the x and y directions, respectively; dw is the distance between the left and right
wheels; and da is the distance between the front and rear wheels. V is the vehicle speed.
The constitutive equations of the thin plate are given as
2
E @ w @2w
σx ¼ + νp 2 (2.2.4)
1 ν2p @x2 @y
h
Plate y x
Observation point
z
Fig. 2.2.1 Theoretical model of the plate-ground system.
Solutions for Saturated Soil Under Moving Loads 17
2
E @ w @2w
σy ¼ + ν p (2.2.5)
1 ν2p @y2 @x2
2.2.2 Coupling Between the Pavement and Saturated Ground
As the contact surface between the pavement and the poroelastic half-space is assumed to
be smooth and fully permeable, the boundary conditions of the half-space at z ¼ 0 m are
given as follows:
σ z ðx, y, 0, tÞ ¼ F ðx, y, tÞ (2.2.6)
τxz ðx, y, 0, t Þ ¼ 0 (2.2.7)
τyz ðx, y, 0, tÞ ¼ 0 (2.2.8)
pf ðx, y, 0, tÞ ¼ 0 (2.2.9)
uz ðx, y, 0, t Þ ¼ w ðx, y, tÞ (2.2.10)
(2.2.13)
where
1 2 2
ϕðε, η, z, ωÞ ¼ γ 3 + ε + η2 ðγ 2 a2 eγ2 z γ 1 a1 eγ1 z Þ + ðγ 1 g5 γ 2 g6 Þeγ3 z (2.2.14)
Δ0
18 Solutions for Biot's Poroelastic Theory in Key Engineering Fields
In Fig. 2.2.2A for c0 ¼ 0:2, the displacement curves are almost symmetric and have one
peak at the center of the tandem axles. The maximum displacement decreases with
increasing b*. In Fig. 2.2.2B for the high load velocity of c0 ¼ 1:0, the displacements
become much larger than those of c0 ¼ 0:2, and significant ground vibration is observed.
The maximum displacements apparently decrease as b* increases.
The dimensionless maximum vertical displacement against dimensionless load veloc-
ity for different values of flexural rigidity is plotted in Fig. 2.2.3. The displacements are
normalized with respect to the displacement at c0 ¼ 0. The maximum surface displace-
ment response has nearly a static character up to a velocity of c0 ¼ 0:5. However, with a
further increase in load velocity, the displacements increase rapidly and reach a maximum
value near c0 ¼ 0:93 and then decrease rapidly. The critical speed for the poroelastic half-
space is the Rayleigh wave speed, which equals 0.936Vs with a soil Poisson’s ratio of 0.35
according to Richart [6]. Thus, the critical speed of the pavement system is a little smaller
than the Rayleigh wave speed of poroelastic soil. This phenomenon is due to the
dynamic interaction between the plate and poroelastic half-space. It is clearly observed
in Fig. 2.2.3 that the peak values of the normalized maximum displacement curves
decrease as the plate flexural rigidity Dp* increases.
The acceleration responses at the ground surface are investigated in Fig. 2.2.4. For c0 ¼
0:2 in Fig. 2.2.4A, the acceleration curve has one peak at the center of the tandem axle.
0.04
b* = 0.1
0.025 b* = 1
0.03
b* = 0.1 b* = 10
0.020 b* = 1 b* = 100
0.02
b* = 10
u*z(xt,0,0)
u*z(xt,0,0)
0.015 b* = 100
0.01
0.010
0.00
0.005
–0.01
0.000
–30 –20 –10 0 10 20 30 –40 –30 –20 –10 0 10 20 30 40
(A) xt (B) xt
Fig. 2.2.2 Effects of the soil permeability coefficients on the displacement responses. (A) c0 ¼ 0:2;
(B) c0 ¼ 1:0:
20 Solutions for Biot's Poroelastic Theory in Key Engineering Fields
2.0
Dp* = 0.05
1.6 Dp* = 5
D * = 50
1.4 p
1.2
1.0
0.8
0.6
0.0 0.2 0.4 0.6 0.8 1.0 1.2
C
0
The permeability of the soil has little effect on ground acceleration responses. When the
load velocity exceeds the Rayleigh wave speed, as shown in Fig. 2.2.4B, the accelerations
apparently vibrate to a wide extent, and the magnitude of the accelerations is much larger
than that for c0 ¼ 0:2. Thus, as the vehicle speed exceeds the critical speed, the vibrations of
the road increase significantly. The permeability affects the acceleration responses signif-
icantly at this vehicle speed. The acceleration responses decrease as b* increases.
0.015
b* = 0.1
Dimensionless vertical acceleration
0.001
0.010 b* = 1
b* = 10
0.000 0.005 b* = 100
b* = 0.1
0.000
–0.001 b* = 1
b* = 10
–0.005
–0.002 b* = 100
–0.010
–0.003
–30 –20 –10 0 10 20 30 –40 –30 –20 –10 0 10 20 30 40
xt xt
Fig. 2.2.4 Acceleration responses at the surface of the ground. (A) c0 ¼ 0.2; (B) c0 ¼ 1.0.
Solutions for Saturated Soil Under Moving Loads 21
1.0
1.2
0.8 Dp* = 0.05 Dp* = 0.05
0.6 Dp* = 0.5 Dp* = 0.5
0.8
Dp* = 5 Dp* = 5
0.4
s*z(xt,0,0)
Dp* = 50 Dp* = 50
s*z(xt,0,0)
0.4
0.2
0.0 0.0
–0.2
–0.4
–0.4
–0.6 –0.8
–30 –20 –10 0 10 20 30 –40 –20 0 20 40
xt xt
Fig. 2.2.5 The vertical stresses of the plate for different Dp*. (A) c0 ¼ 0.2; (B) c0 ¼ 1.0.
The vertical stresses of the plate generated by the moving traffic were obtained from
the compatibility condition between the plate and the poroelastic half-space. In
Fig. 2.2.5, the vertical stress at the interface between the plate and the half-space is plotted
against xt for different Dp* when c0 ¼ 0:2. Clearly, the magnitude of the stress decreases
rapidly with increasing Dp*. When D∗p ¼ 0:05, the positive peak stresses occurred at the
point at which the load was applied. And negative stress was caused at the center of the
tandem axles because the plate was arched during the load area when the flexural rigidity
of the plate was small. As Dp* increased, the maximum stresses occurred at the center of
the tandem axles, and no negative stress was caused. In Fig. 2.2.5B, the vertical stress for
high load velocity c0 ¼ 1:0 is presented. The vertical stresses of the plate fluctuate signif-
icantly faced with the load and the frequency of fluctuations decreased with increasing
Dp*. Compared with Fig. 2.2.5A, the magnitude of the vertical stress increased slightly
when the load velocity increased to c0 ¼ 1:0. The effect of c0 on the magnitude of the
vertical stress at the contact surface was not pronounced.
Fig. 2.3.1. The track model proposed by Picoux and Le Houedec [8] is introduced in this
chapter.
In order to calculate the receptance of the track-ground system, a unit vertical har-
monic load eiΩt was applied on the rail, which was pointing downwards and located at
x ¼ 0 when t ¼ 0, moving along the rails at speed c. The governing equation for a rail
represented by an Euler beam is written as
@ 4 uR ðx, tÞ @ 2 uR
EI + m R + kP ½uR ðx, t Þ uS ðx, tÞ ¼ eiΩt δðx ct Þ (2.3.1)
@x 4 @t 2
in which uR is the vertical displacement of the Euler beam, EI is the bending stiffness of
the rail beam, mR is the mass of the rail per unit length, kP denotes the spring constant of
the rail pads, and uS is the vertical displacement of the sleepers.
It is noted by Knothe and Grassie [9] and Grassie et al. [10] that the effect of the dis-
creteness of the rail supports may be neglected when the dominating frequency content
of the response is not in the vicinity of the so-called pinned-pinned resonance frequency.
This frequency is normally located in the range 700–1000 Hz, depending on the rail
properties and the sleeper distance. The main frequency range of interest for the percep-
tion of ground vibration is about 5–80 Hz; therefore it is suitable to represent the sleepers
by a continuous mass:
@ 2 uS ðx, tÞ
mS + kP ½uS ðx, tÞ uR ðx, t Þ ¼ FS ðx, tÞ (2.3.2)
@t 2
where mS is the mass of the sleeper per unit length and FS is the load between the sleepers
and the ballast.
The ballast was first considered by the Cosserat model (Suiker et al. [11]). At the top
and bottom of the ballast, the system can be written:
Rail
Sleeper
Ballast
x
Poroelastic half-space
z
Fig. 2.3.1 Theoretical model of the vehicle-track-ground coupling system.
Solutions for Saturated Soil Under Moving Loads 23
mB @ 2 uS ðx, tÞ @ 2 uB ðx, tÞ
2 + + kB ½uS ðx, tÞ uR ðx, tÞ ¼ FS ðx, tÞ (2.3.3)
6 @t 2 @t 2
mB @ 2 uS ðx, tÞ @ 2 uB ðx, tÞ
+ 2 + kB ½uS ðx, t Þ + uB ðx, tÞ ¼ FB ðx, t Þ (2.3.4)
6 @t 2 @t 2
where mB is the mass of the ballast per unit length, kB is the spring constant between bal-
last and sleepers, FB is the ballast load on the soil, and uB is the vertical displacement of the
ballast.
where
1 h 2 2 i
ϕðε, η, z*, ΩÞ ¼ γ 3 + ε + η2 γ 2 a2 eγ2 z* γ 1 a1 eγ 1 z* + ðγ 1 g5 γ 2 g6 Þeγ3 z*
Γ
2
gi ¼ Ei + 2ε bi 2 Fi i ¼ 1, 2, g3 ¼ λ*E1 2γ 21 F1 α, g4 ¼ λ*E2 2γ 22 F2 α
g5 ¼ g1 + 2η2 F1 , g6 ¼ g2 + 2η2 F2 , Γ ¼ ðg4 g3 Þ γ 23 + ξ2 + η2 2γ 3 ðγ 1 g5 γ 2 g6 Þ
In Eq. (2.3.12), F~B ðε, ωÞ remains unknown and will be resolved in the next section.
∗
α2 ðξ, ωÞα4 ðξ, ωÞ2 α2 ðξ, ωÞα3 ðξ, ωÞα5 ðξ, ωÞ α4 ðξ, ωÞF~B ðξ, ωÞk∗p
u~*R ðξ, ωÞ ¼
α1 ðξ, ωÞα4 ðξ, ωÞ2 α1 ðξ, ωÞα3 ðξ, ωÞα5 ðξ, ωÞ + α5 ðξ, ωÞk*p 2
(2.3.18)
The displacement of the rail and ground in the time domain can be expressed as follows
by introducing an auxiliary spatial coordinate x∗t ¼ x* c0 τ, and the time-domain results
are obtained by the FFT algorithm.
ð
1 ∞ *
u*R ðx*, τÞ ¼ u~*R ðξ,Ω* ξc0 Þeiξxt dξ eiΩ*τ (2.3.19)
2π ∞
ð ð
1 ∞ ∞ *
u*z ðx*, y*, z*, τÞ ¼ 2 u~* ðξ, η, Ω* ξc0 Þeiηy* eiξxt dηdξ eiΩ*τ (2.3.20)
4π ∞ ∞ z
Eqs. (2.3.19), (2.3.20) can also be expressed as:
∗ iΩ*τ
u*R ðx*, τÞ ¼ uΩ
R xt e (2.3.21)
∗ iΩ*τ
u*z ðx*, y*, z*, τÞ ¼ uΩ
z xt , y*, z* e (2.3.22)
Eqs. (2.3.21), (2.3.22) denote that, in the auxiliary spatial coordinate, the displacements
of the track-ground system are harmonic and have the same vibration frequency as the
dynamic load.
1
Z0V ðΩÞ ¼ KV Ω2 MV BP0 ðΩÞ (2.3.24)
The receptance between the jth and kth wheelsets within a vehicle is denoted by ΔW jk (W
means wheelset), where j, k ¼ 1, 2, …, N; N is the number of wheelsets of the vehicle.
ΔWjk denotes the displacement amplitude of the jth wheelset due to a unit vertical har-
monic load with an excitation frequency Ω exerted at the kth wheelset. The displacement
vector of the wheelset in the vehicle is expressed as
0 T
Z0W ðΩÞ ¼ ZW1 0
ðΩÞ, ZW2 0
ðΩÞ, ZW3 0
ðΩÞ, …, ZWN ðΩÞ (2.3.25)
and
T
P0 ðΩÞ ¼ P10 ðΩÞ, P20 ðΩÞ, P30 ðΩÞ, …, PN0 ðΩÞ (2.3.26)
is the wheel-rail force vector for a vehicle.
The displacement vector of the wheelsets is part of that for the corresponding vehicle.
Therefore, it can be written as
Eq. (2.3.29) gives the receptance matrix at the wheelsets for a single vehicle. Suppose
there are N1 identical vehicles being considered; the total number of the wheel-rail loads
is M ¼ N1 N . The receptance matrix at the wheelsets for the train, denoted by ΔT (where
T stands for train), is given by
2 W 3
Δ ⋯ 0
6 7
ΔT ¼ diag ΔW , …, ΔW ¼ 6 4 ⋮ ⋮ ⋮ 7
5 (2.3.30)
W
0 ⋯ Δ
The elements of matrix ΔT are denoted by ΔT lk, where k,l ¼ 1, 2, …, M.
Thus, the receptance at the jth wheel-rail contact point due to a unit load at the kth
wheel-rail contact point on the rail is determined by
Ω ∗
ΔRjk ¼ u R ljk (2.3.31)
Solutions for Saturated Soil Under Moving Loads 27
where
ljk∗ ¼ a∗j a∗k (2.3.32)
is the dimensionless distance between the two contact points.
The complex amplitudes of the displacements at the wheel-rail contact points on the
rails are given by
∗
z0 R ðΩ*Þ ¼ ΔR P0 *ðΩ*Þ (2.3.33)
where
2 3
ΔR
11 Δ12 ⋯ Δ1M
R R
6 R 7
6 Δ21 ΔR
22 ⋯ Δ2M
R
7
R6
Δ ¼6 7 (2.3.34)
7
4 ⋮ ⋮ ⋮ ⋮ 5
ΔR
M1 ΔM2 ⋯ ΔMM
R R
z0 *R ðΩ*Þ ¼ ðz0 *R1 ðΩ*Þ, z0 *R2 ðΩ*Þ, z0 *R3 ðΩ*Þ, …, z0 *RM ðΩ*ÞÞ
T
(2.3.35)
Eq. (2.3.35) represents the displacement vector of the rail at the wheel-rail contact points
observed in the auxiliary spatial coordinate.
The rail irregularities are presented by a sinusoidal profile of amplitude A. The profile
of the rail irregularities is given by
zðxÞ ¼ Aeið2π=λ1 Þx (2.3.36)
where λ1 denotes the wavelength. The process was assumed to be linear, so that a dis-
placement input was generated at the excitation frequency f ¼ c=λ1 . The angular fre-
quency is obtained by Ω ¼ 2πc=λ1 . At time t, the lth wheelset arrives at x ¼ al + ct;
thus the displacement input of the rail profile at the lth wheel-rail contact point is
zl ðt Þ ¼ z0l ðΩÞeiΩt ¼ Aeið2π=λ1 Þðal + ctÞ ¼ Aeið2π=λ1 Þal eiΩt (2.3.37)
The coupling of a wheelset with rail is illustrated in Fig. 2.3.2, where zWl0 (Ω)eiΩt denotes
the displacement of the lth wheelset.
The wheel and rail deform locally according to the Hertz theory under the action of
the contact force. Thus the wheel and rail are coupled by a Hertz spring. Provided that
the dynamic contact force is a small fraction of the axle load, the Hertz spring can be taken
to be linear. The stiffness of the Hertzian contact spring is denoted by kHl. It is also
assumed that the wheelset is always in contact with the rail. Thus
z0Wl ðΩÞ ¼ z0Rl ðΩÞ + z0l ðΩÞ + P 0 ðΩÞ=kHl (2.3.38)
From Eqs. (2.3.21) to (2.3.28), the following relations can be derived:
28 Solutions for Biot's Poroelastic Theory in Key Engineering Fields
Z⬘Wl(Ω)
P⬘l(Ω)
KHI
P⬘l(Ω) Z⬘l(Ω)
XM
z0Wl ðΩÞ ¼ ΔTlk Pk0 ðΩÞ (2.3.39)
k¼1
X
M
z0Rl ðΩÞ ¼ ΔR 0
lk Pk ðΩÞ (2.3.40)
k¼1
The unknown item Pl0 (Ω*) can be obtained by solving Eq. (2.3.41). The displacement of
the ground and the rails at the exciting frequency Ω* are given by superposition.
X
M
u∗R ðx*, τÞ ¼ uΩ 0
R ðxt * a*l ÞPl ðΩ*Þe
iΩ*τ
(2.3.42)
l¼1
X
M
u∗z ðx*, τÞ ¼ uΩ 0
z ðxt * al *, y*,z*ÞPl ðΩ*Þe
iΩ*τ
(2.3.43)
l¼1
0.125–24 m. The parameters for the vehicle model are listed in Table 2.3.1 and the
parameters for the track are given in Table 2.3.2, and correspond to those used in Sheng
and Jones [12]. The parameters of the saturated poroelastic half-space are presented in
Table 2.3.3, which are the same as those used in Cai et al. [4]. The parameters were
selected according to Tables 2.3.1–2.3.3 if not denoted in the figures.
The roughness-induced dynamic wheel-rail forces are shown in Fig. 2.3.3 against the
excitation frequency f* (f * ¼ c0 =λ∗1 ). It is shown that the magnitude of the wheel-rail
forces increased as f* increased and reached a maximum value at the critical frequency
around f * ¼ 0:55, then decreased as f* increased further. With a close inspection of
Fig. 2.3.3, one can also see that the magnitude of the dynamic wheel-rail forces have
a local peak at frequency 0.02. This frequency is close to the natural frequency of the
suspended mass on the suspension, as noted in Sheng and Jones [12].
In order to study the attenuation of the ground vibration, the magnitudes of accel-
erations for both axle and dynamic loads are presented against y* in Fig. 2.3.4. In
Fig. 2.3.4A, when c0 ¼ 0:2, the quasistatically induced acceleration decreases rapidly as
y* increases, and is nearly zero at y* ¼ 5. The dynamically induced acceleration is much
larger at y* ¼ 0 and decreases more slowly. The dynamically induced component of
vibration is the dominant one for the acceleration response of the free-field.
Table 2.3.3 Dimensionless parameters for fully water-saturated poroelastic soil medium
Parameter Value
Lame constant, λ* 2
Water density, ρf* 0.53
Parameter of soil structure, m* 1.5625
Hysteretic damping ratio, D 0.02
Ratio between the fluid viscosity and the intrinsic permeability, b* 10
The parameter for the compressibility of the soil particle, α 0.97
The parameter for the compressibility of the fluid, M* 12
In Fig. 2.3.4A, the acceleration response for the elastic half-space soil medium is also pre-
sented. The acceleration responses predicted by the elastic soil model are larger than those
predicted by the poroelastic soil model for both axle and dynamic loads at y* ¼ 0, and the
acceleration responses attenuate more slowly in the elastic soil medium. The free-field
acceleration response predicted by the elastic soil model is larger than that predicted
by the poroelastic one. In Fig. 2.3.4B, when c0 ¼ 0:6, the quasistatically induced accel-
eration is larger than the dynamically induced acceleration at y* ¼ 0, but the dynamically
induced vibration component dissipates more slowly and gives rise to the majority of the
vibration for the free-field off the track. For the case of an elastic soil medium, it can be
seen that the acceleration response dissipates more slowly in the elastic soil. Using the
elastic soil medium model will overestimate the acceleration response under the free-field
condition for c0 ¼ 0:6. In Fig. 2.3.4C, for c0 ¼ 1:0, the quasistatically induced accelera-
tions are much larger and are the dominant component for the ground vibration at
0.0016
0.0014
Magnitude of the wheel-rail force
0.0012
0.0010
0.0008
0.0006
0.0004
0.0002
0.0000
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Excitation frequency
Fig. 2.3.3 Dynamic wheel-rail force at different excitation frequencies.
Solutions for Saturated Soil Under Moving Loads 31
0.0025 0.007
0.002
0.0005
0.001
0.0000 0.000
0 10 20 30 40 0 10 20 30 40
y* y*
0.025
0.020
Axle load for poroelastic half-space
Axle load for elastic half-space
0.015 Dynamic load for poroelastic half-space
Dynamic load for elastic half-space
0.010
0.005
0.000
0 10 20 30 40
y*
(C) c0 = 1.0
Fig. 2.3.4 Ground surface accelerations against the distance. (A) c0 ¼ 0.2; (B) c0 ¼ 0.6; (C) c0 ¼ 1.0.
various y*. The acceleration response predicted by the elastic soil model is apparently
smaller than that predicted by the poroelastic soil model.
Therefore, for vehicle speeds below the Rayleigh wave speed, the free-field
acceleration response is dominated by the dynamic load. But when the vehicle speed
approaches the Rayleigh wave speed of the half-space, the axle load is the dominant
one for the acceleration response of the free-field. Using the elastic soil medium model
will overestimate the acceleration response level of the free-field for the train speed below
Rayleigh wave speed, while it will underestimate the response level for the train speed
above the Rayleigh wave speed.
In Fig. 2.3.5, the train-induced excess pore water pressures are studied. In Fig. 2.3.4A,
when c0 ¼ 0:2, the quasistatically induced pore water pressure reaches a positive peak at
the load point and a negative peak behind the load. The dynamically induced pore water
pressure fluctuates along xt* and the maximum value is larger than that for the quasista-
tically induced one. When the vehicle speed exceeds the critical speed, as shown in
Fig. 2.3.5B, the quasistatically induced pore water pressure becomes much larger and
is the dominant component for the pore water pressure responses.
32 Solutions for Biot's Poroelastic Theory in Key Engineering Fields
0.0010
0.006
0.0008
0.005
0.0006 Axle load
0.004
Dynamic load Axle load
0.0004
P*(x*t,0,1)
P*(x*t,0,1)
0.003 Dynamic load
0.0002 0.002
0.0000 0.001
–0.0002 0.000
–0.0004 –0.001
–0.0006 –0.002
–40 –20 0 20 40 –40 –20 0 20 40
x* x*
(A) t (B) t
302 V1 Rail
30B
V3
31A V4
17.36 m
V5
H6
V8 V7
Fig. 2.4.1 Cross-section arrangement of the sensors for the vibration test.
Solutions for Saturated Soil Under Moving Loads 33
1.0
Velocity (cm/s)
0.5
0.0
–0.5
–1.0
1 2 3 4 5 6 7 8
Time (s)
Fig. 2.4.2 Measured velocity at the sleeper (V1) under a train speed of 145.1 km/h maximum value
1.046, minimum value 0.906.
0.06
0.05
0.04
Velocity (cm/s)
0.03
0.02
0.01
0.00
–0.01
–0.02
–0.03
1 2 3 4 5 6 7 8
Time (s)
Fig. 2.4.3 Measured velocity at the embankment (V3) under a train speed of 145.1 km/h maximum
value 0.058, minimum value 0.025.
0.04
0.02
Velocity (cm/s)
0.00
–0.02
–0.04
–0.06
1 2 3 4 5 6 7 8
Time (s)
Fig. 2.4.4 Measured velocity at the free-field (V7) under a train speed of 145.1 km/h maximum value
0.026, minimum value 0.053.
34 Solutions for Biot's Poroelastic Theory in Key Engineering Fields
2.0
1.5
1.0
Velocity (cm/s)
0.5
0.0
–0.5
–1.0
–1.5
0.3
0.2
0.1
Velocity (cm/s)
0.0
–0.1
–0.2
–0.3
–0.4
2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
Time (s)
Fig. 2.4.6 Measured velocity at the embankment (V3) under a train speed of 335.1 km/h maximum
value 0.245, minimum value 0.348.
0.02
0.01
Velocity (cm/s)
0.00
–0.01
–0.02
–0.03
–0.04
–0.05
2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
Time (s)
Fig. 2.4.7 Measured velocity at the free-field (V8) under a train speed of 335.1 km/h maximum value
0.020, minimum value 0.031.
Solutions for Saturated Soil Under Moving Loads 35
Measured velocity
Calculated velocity
1.0
0.5
Velocity (cm/s)
0.0
–0.5
–1.0
1 2 3 4 5 6 7 8
Time (s)
Fig. 2.4.8 Comparison between the calculated and measured velocity at the sleeper (V1) under a train
speed of 145.1 km/h.
Measured velocity
Calculated velocity
1.0
0.5
Velocity (cm/s)
0.0
–0.5
–1.5
1 2 3 4 5 6 7 8
Time (s)
Fig. 2.4.9 Comparison between the calculated and measured velocity at the sleeper (V1) under a train
speed of 335.1 km/h.
36 Solutions for Biot's Poroelastic Theory in Key Engineering Fields
The measured results follow. When the train runs at the speed of 145.1 km/h, the
measured velocity at the sleepers, embankment, and free field are presented in
Figs. 2.4.2–2.4.4. The peak velocity at the sleeper was about 1.0 cm/s, and then
decreased as the distance from the railway centerline increased. When the train speed
increased to 335.1 km/h, the measured velocity at the sleepers, embankment, and free
field are presented in Figs. 2.4.5–2.4.7, and it can be seen that the velocity at the sleeper
was amplified to 1.5 cm/s, then decreased as the distance increased.
In order to validate the correctness of the theoretical model proposed in the previous
section, the comparison between the calculated and measured velocity at the sleeper
(V1) under a train speed of 145.1 km/h and 335.1 km/h is shown in Figs. 2.4.8 and
2.4.9, respectively. It is shown that the calculated and measured results agreed well.
Based on the in situ test results, the following results were obtained:
(1) The velocity of the track was intensified significantly as the train speed increased; the
velocity amplitude of the track at 350 km/h was about two times what it was at
120 km/h.
(2) The vibration decayed rapidly along the transverse direction; the vibration velocity
at the free-field was significantly smaller than that at the sleeper of the track.
CHAPTER 3
Solutions for Biot’s Poroelastic Theory in Key Engineering Fields © 2017 Elsevier Inc.
http://dx.doi.org/10.1016/B978-0-12-812649-3.00003-4 All rights reserved. 37
38 Solutions for Biot's Poroelastic Theory in Key Engineering Fields
P (t)
Soil surface
m r0
h
o
r
H
q
z
Rigid bedrock
@σ rr σ rr σ θθ @σ rz @pf
+ + ¼ ρu€r + ρf w€r (3.1.1)
@r r @z @r
@σ rz σ rz @σ zz @pf
+ + ¼ ρu€z + ρf w€z (3.1.2)
@r r @z @z
@pf ρ
¼ ρf u€r + f w€r + bw_ r (3.1.3)
@r n
@pf ρ
¼ ρf u€z + f w€z + bw_ z (3.1.4)
@z n
The constitutive relations for a homogeneous poroelastic material can be expressed
according to Eq. (1.1.21) as
@ur
σ rr ¼ λe + 2μ (3.1.5)
@r
@uz
σ zz ¼ λe + 2μ (3.1.6)
@z
ur
σ θθ ¼ λe + 2μ (3.1.7)
r
@ur @uz
σ rz ¼ μ + (3.1.8)
@z @r
@ u_r u_r @ u_z @ w_ r w_ r @ w_ z
+ + + + + ¼0 (3.1.9)
@r r @z @r r @z
The motion under consideration was assumed to be time-harmonic with the factor eiωt,
which was henceforth suppressed from all expressions for brevity. It was convenient to
use nondimensional forms of length parameters including the coordinates by selecting the
Problems for Vibrations of Foundations 39
radius of the foundation as a unit of length. Stresses and pore pressure can also be nor-
malized into the nondimensional forms, with respect to the shear modulus μ of the bulk
material. All variables were replaced by the nondimensional quantities, but the previous
notations were used for convenience. In addition, the following nondimensional fre-
quency and material properties were introduced:
rffiffiffi
ρ λ ρ br0
a0 ¼ r0 ω, λ* ¼ , ρ* ¼ f , b* ¼ pffiffiffiffiffi , M* ¼ M=μ (3.1.10)
μ μ ρ ρμ
3.1.2 Boundary Conditions for Vertical Vibrations
For the vibration problem of a rigid foundation perfectly bonded to the surrounding soil
in the vertical direction, as shown in Fig. 3.1.1, it was assumed that the contact surface
between the foundation base and the poroelastic soil was smooth and fully permeable.
The boundary conditions on z ¼ 0 and z ¼ H/r0 can then be written in nondimen-
sional quantities as follows:
uz ðr, 0Þ ¼ U0 , 0 r 1 (3.1.11)
σ zz ðr, 0Þ ¼ 0, 1 < r ∞ (3.1.12)
σ rz ðr, 0Þ ¼ 0, 0 r ∞ (3.1.13)
pf ðr, 0Þ ¼ 0, 0 r ∞ (3.1.14)
uz ðr,H=r0 Þ ¼ 0, 0 r 1 (3.1.15)
ur ðr,H=r0 Þ ¼ 0, 0 r 1 (3.1.16)
wz ðr,H=r0 Þ ¼ 0, 0 r 1 (3.1.17)
where U0 ¼ U/r0 with U denoting the vertical displacement of the rigid foundation.
an ideal elastic soil. In the Baranov-Novak method, the soil was divided into two inde-
pendent parts and it was assumed that the soil reaction at the base of the foundation was
equal to that of a foundation resting on the surface of the soil, while the lateral soil reac-
tion was evaluated independently. In this paper, several similar assumptions were intro-
duced: (I) the soil at the side of the foundation is composed of a series of infinitesimally
thin independent poroelastic layers. In each infinitesimally thin poroelastic layer, the
influence of the gradient of σ zz and pf in the vertical direction and the influence of radial
displacement on Rs(t) are neglected. (II) The dynamic reaction at the base of the foun-
dation is independent of depth of embedment.
where Jμ is the Bessel function of the first kind of the μth order, and ξ is the Hankel trans-
form parameter.
The following general solutions can be obtained for a zero-order Hankel transform of
uz, wz, σ zz, and pf, and the first-order Hankel transform of σ rz and ur, respectively:
p~f0 ¼ χ 1 A1 eqz + χ 1 B1 eqz + A2 eξz + B2 eξz (3.1.21)
u~1r ¼ χ 2 A1 eqz + χ 2 B1 eqz + χ 3 A2 eξz + χ 3 B2 eξz + A3 esz + B3 esz (3.1.22)
u~z0 ¼ χ 4 A1 eqz χ 4 B1 eqz + χ 3 A2 eξz χ 3 B2 eξz + χ 5 A3 esz χ 5 B3 esz (3.1.23)
1 qz ξz ξz
w~z0 ¼ κ1 u~z0 qχ A
1 1 e + qχ B
1 1 e qz
ξA 2 e + ξB 2 e (3.1.24)
ρ*a0 2
σ~zz
0
¼ γ 1 A1 eqz + γ 1 B1 eqz 2ξχ 3 A2 eξz 2ξχ 3 B2 eξz 2ξA3 esz 2ξB3 esz (3.1.25)
σ~1rz ¼ γ 2 A1 eqz γ 2 B1 eqz 2ξχ 3 A2 eξz + 2ξχ 3 B2 eξz + γ 3 A3 esz γ 3 B3 esz (3.1.26)
where A1, A2, A3, B1, B2, and B3 are arbitrary functions to be determined from the
boundary and continuity conditions of the given problem. The variables κ 1, q, s, χ i
(i ¼ 1, 2, 3, 4, 5), and γ j (j ¼ 1, 2, 3) appearing in the preceding equations are known
Problems for Vibrations of Foundations 41
functions of the material properties and the frequency of excitation, and are given in
Appendix B.
By applying the Hankel transform to Eqs. (3.1.12)–(3.1.17), together with Eqs. (3.1.21)–
(3.1.26), the boundary conditions can be rewritten in a matrix form:
Qa¼f (3.1.27)
Q11 Q12
where a ¼ fA1 , A2 , A3 , B1 , B2 , B3 g; f ¼ σ~zz ðξ, 0Þ,0,0,0,0,0 ; and Q ¼
T T 0
Q21 Q22
denotes a 6 6 matrix whose elements Q11, Q12, Q21, and Q22 are functions of ξ and
are given in Appendix C.
The solution of Eq. (3.1.27) is
φ∗11 0 φ∗ φ∗
A1 ¼ σ~zz ðξ, 0Þ, A2 ¼ 12 σ~0zz ðξ, 0Þ, A3 ¼ 13 σ~0zz ðξ, 0Þ (3.1.28)
jφj jφj jφj
ψ ∗11 0 ψ∗ ψ∗
B1 ¼ σ~zz ðξ, 0Þ, B2 ¼ 12 σ~0zz ðξ, 0Þ, B3 ¼ 13 σ~0zz ðξ, 0Þ (3.1.29)
jψ j jψ j jψ j
1 1
where φ ¼ Q11 Q12 Q22 Q21 ; ψ ¼ Q12 Q11 Q21 Q22 ; jφj and jψ j denote the determi-
nants of φ and ψ, respectively; φ11 , φ12 , φ13 and ψ ∗11 , ψ ∗12 , ψ ∗13 denote the alge-
∗ ∗ ∗
braic complements of φ11, φ12, φ13 and ψ 11, ψ 12, ψ 13, corresponding to φ and ψ,
respectively.
Substitution of Eqs. (3.1.28), (3.1.29) into Eq. (3.1.23) results in the following
relationship:
u~z0 ðξ, 0Þ ¼ pðξÞ~
σ 0zz ðξ, 0Þ (3.1.30)
ξpðξÞ
σ 0zz ðξ, 0Þ, H ðξÞ ¼
where BðξÞ ¼ ξ~ 1 and ν is Poisson’s ratio of the solid matrix.
1ν
It can be proved that lim ξpðξÞ ¼ ð1 νÞ, i:e:, lim H ðξÞ ¼ 0.
ξ!∞ ξ!∞
The dual integral equations can be reduced to a Fredholm integral equation of the
second kind by employing the following integral representation:
Another random document with
no related content on Scribd:
SECONDARY OR REMOTE RESULTS OF
LIVER DISEASE.