|
|
|
[Contents] | [Index] |
Cooperation with: C. Lai (European Centre for Training and Research in Earthquake Engineering, Pavia, Italy), R. Lancellotta, S. Foti (Politecnico di Torino, Italy)
Description:
The project is the continuation of the theoretical research on
surface waves on the basis of the model by Wilmanski [5]. Here, the
dispersion relation for surface waves on an impermeable boundary
of a fully
saturated
poroelastic medium is investigated numerically in the whole range
of frequencies. To this aim the linear simplified model of a two-component
poroelastic medium is used. Similarly to the classical Biot
model, it is a continuum mechanical model but it is much simpler due to the
lack of coupling of stresses.
The general dispersion relation has been ascertained, and results of
an earlier work [6] for the high and low frequency approximations
have been recalled. The main part of the project concerns
numerical aspects: first we
indicate the applied numerical procedure and then we illustrate the
numerical results for the normalized velocities and attenuations of the
Rayleigh and Stoneley waves, [1, 2]. It is known that surface
modes of propagation in linear models result from the combination of bulk
modes. Physically, this means that at any point of the boundary classical
longitudinal and shear waves combine into the Rayleigh wave which must be
slower than both bulk waves. The presence of the second longitudinal bulk
wave P2 yields the existence of the second surface mode--the Stoneley wave
which should be slower than the P2 wave--the slowest of bulk waves. Both
quantities, velocities and attenuations, are shown for different values of
the bulk permeability coefficient , in different ranges of
frequencies. A decay of the Rayleigh wave velocity, mentioned in [3], has been confirmed in the range of small frequencies in spite of
the lack of static coupling between components. Moreover, we compare the
behavior of the two types of surface waves with the behavior of two bulk
waves: P1 and P2.
Here, we only show a part of the project, namely the governing equations, the boundary conditions, and the numerical results for a chosen value of the permeability coefficient:
Model
Within the linear model of a two-component poroelastic saturated
medium the process is described by the macroscopic fields
x, t
- partial mass density of the fluid,
vF
x, t
- velocity of the fluid,
vS
x, t
- velocity of the skeleton,
eS
x, t
- symmetric tensor of small deformations
of the skeleton and the porosity n. These fields satisfy the following set
of linear equations
![]() ![]() ![]() ![]() ![]() ![]() |
|||
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
|||
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
|||
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
|||
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Here
,
, n0 denote constant reference values of
partial mass densities, and porosity, respectively, and
,
,
,
,
,
,
,
are constant material
parameters. The first one describes the macroscopic compressibility of the
fluid component, the next two are macroscopic elastic constants of the
skeleton,
is the coupling constant,
is the coefficient of
bulk permeability,
is the relaxation time, and
,
describe equilibrium and nonequilibrium changes of porosity, respectively.
For the purpose of this work we assume
= 0.
Boundary conditions
In order to determine surface waves in a saturated poroelastic
medium we need conditions for z = 0. In the general case of a boundary
between a saturated porous material and a fluid the boundary conditions were
formulated by Deresiewicz & Skalak. We quote them here in a slightly
modified form and for an impermeable boundary
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
|||
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Results for velocities and attenuations
![]() |
Conclusions
RAYLEIGH
·The velocity of propagation of this wave lies in the interval
determined by the limits
0 and
. The high frequency limit is approx. 4.7 % higher than the low
frequency limit. The wave is always slower than the S wave. As a function
of
it possesses an inflection point and it is slightly
nonmonotonous.
·This nonmonotonicity appears in the range of small frequencies. The
velocity possesses in this range a minimum whose size is very small.
Interestingly, the minimum value remains constant for the different values
of . This means that the decay is not driven by the diffusion. Such a
behavior is also observed within Biot's model.
·The attenuation of this wave grows from zero for = 0 to
infinity as
. In the range of large frequencies
it is linear (a constant positive quality factor). This means that it is a
leaky wave.
STONELEY
·The velocity of this wave grows monotonically from the zero value
for = 0 to a finite limit which is slightly smaller than the
velocity of the P2 wave. The growth of the velocity of this wave in the
range of low frequencies is much steeper than the one of Rayleigh waves
similarly to the growth of the P2 velocity.
·Both the velocity and attenuation of the Stoneley wave approach zero
as
.
·The attenuation of the Stoneley wave grows monotonically to a finite
limit for
(zero quality factor). It is slightly
smaller than the attenuation of P2 waves. Consequently, in contrast to the
claims in the literature, the Stoneley wave is attenuated.
References:
|
|
|
[Contents] | [Index] |