![MathJax Logo](/templates/jsp/_style2/_tandf/pb2/images/math-jax.gif)
ABSTRACT
Using a suitable stochastic version of the compactness argument of [Zhikov VV. On an extension of the method of two-scale convergence and its applications. Sb Math. 2000;191(7–8):973–1014], we develop a probabilistic framework for the analysis of heterogeneous media with high contrast. We show that an appropriately defined multiscale limit of the field in the original medium satisfies a system of equations corresponding to the coupled ‘macroscopic’ and ‘microscopic’ components of the field, giving rise to an analogue of the ‘Zhikov function’, which represents the effective dispersion of the medium. We demonstrate that, under some lenient conditions within the new framework, the spectra of the original problems converge to the spectrum of their homogenisation limit.
COMMUNICATED BY:
1. Introduction
Asymptotic analysis of differential equations with rapidly oscillating coefficients has featured prominently among the interests of the applied analysis community during the last half a century. The problem of understanding and quantifying the overall behaviour of heterogeneous media has emerged as a natural step within the general progress of material science, wave propagation and mathematical physics. In this period several frameworks have been developed for the analysis of families of differential operators, functionals and random processes describing multiscale media, all of which have benefitted from the invariably deep insight and mathematical elegance of the work of V. V. Zhikov. In the present paper, we touch upon two subjects in which his contributions have inspired generations of followers: the stochastic approach to homogenisation, in particular through his collaboration with S. M. Kozlov during the 1980s, and the analysis of differential operators describing periodic composites with high contrast, which started with his fundamental contribution [Citation1].
Our present interest in the context of stochastic homogenisation of high-contrast composites stems from the relationships that have recently been indicated between media with negative material properties (‘metamaterials’), and more generally time-dispersive media, and ‘degenerate’ families of differential operators, where e.g. loss of uniform ellipticity of the symbol is known to lead to non-classical dispersion relations in the limit of vanishing ratio ϵ of the microscopic (l) and macroscopic (L) lengths: . The work [Citation1] has provided an example, in the periodic context, of what one should expect in the limit as
in terms of the two-scale structure of the solution as well as the spectrum of the related differential operator, in the case when the metamaterial is modelled by disjoint ‘soft’ inclusions with low, order
values of the material parameters (say, elastic constants in the context of linearised elasticity), embedded in a connected ‘stiff’ material with material constants of order
. In mathematical terms, the coefficients of the corresponding differential expression alternate between values of different orders in ϵ, where the contrast increases as ϵ gets smaller.
In the present article, we introduce a stochastic framework for the analysis of homogenisation problems with soft inclusions and explore the question on what version of the results of [Citation1] can be achieved in this new framework. In particular, we are interested in the equations that describe the stochastic two-scale limit, in an appropriate sense, of the sequence of solutions to the probabilistic version of a Dirichlet problem in a bounded domain of . Furthermore, we show that the spectra of such problems converge, in the Hausdorff sense, to the spectrum of the limit problem, which we analyse in a setting that models distributions of soft inclusions whose shapes are taken from a certain finite set and whose sizes vary over an interval. To our knowledge, the present manuscript is the first work containing an analysis of random heterogeneous media with high contrast that results in a ‘complete’ Hausdorff-type convergence statement for the spectra of the corresponding differential operators. Various aspects of multiscale analysis of high-contrast media in the stochastic context have been looked at in a handful of papers, e.g. [Citation2–4].
While in the periodic context norm-resolvent convergence results been obtained for high-contrast media, see [Citation5,Citation6], the stochastic case remains open to developments of a similar nature. It is anticipated that the operator-theoretic approach to problems of the kind we discuss in the present article will provide a general description of the types of spectral behaviour that can occur in the real-world applications where it is difficult to enforce periodicity of the microstructure. On the other hand, as we show in the present work, new wave phenomena should be expected in the stochastic setting (e.g. a non-trivial continuous component of the spectral measure of the homogenised operator for a bounded-domain problem), which makes the related future developments even more exciting.
Next, we outline the structure of the paper. In Section 2 we recall the notion of stochastic two-scale convergence, which we use, in Sections 3 and 4, to pass to the limit, as in a family of homogenisation problems with random soft inclusions. In Section 3 we give a formulation of the high-contrast problem we study and provide some auxiliary statements. In Section 4 we describe the limit problem and prove the strong resolvent convergence of the ϵ-dependent family to the limit system of equations. In Section 5 we provide a link between the spectra of the Laplacian operator on realisations of the inclusions and of the corresponding stochastic Laplacian. In Section 6 we prove that sequences of normalised eigenfunctions of the ϵ-dependent problems are compact in the sense of strong stochastic two-scale convergence. Finally, in Section 7 we discuss two examples of the general stochastic setting and describe the structure of the corresponding limit spectrum.
In conclusion of this section, we introduce some notation used throughout the paper. For a Banach space X and its dual we denote by
the corresponding duality. For a Hilbert space H the inner product of
is also denoted by
and, if
by
. For a set
we denote by
its characteristic function, which takes value one on the set
and zero on the complement to
in the appropriate ambient space. For
we denote by
its closure and by
its Lebesgue measure. Further, we use the notation
for the ball in
of radius r with the centre at the origin; Y denotes the cube
with torus topology, where the opposite faces are identified; and
. For an operator
on some Hilbert space, we denote by
its spectrum. Finally, for a Lipschitz open set
we denote by
the (positive) Laplace operator with the Dirichlet boundary condition on
. For
, we denote by
the element of
which satisfies
. For
, by
we denote the kth coordinate vectors.
2. Stochastic two-scale convergence
2.1. Probability framework
Let be a complete probability space. We assume that
is countably generated, which implies that the spaces
,
are separable. For a function
we will sometimes write
for
.
Definition 2.1
A family of measurable bijective mappings
on a probability space
is called a dynamical system on Ω with respect to P if:
;
,
;
is measurable (for the standard σ-algebra on the product space, where on
we take the Lebesgue σ-algebra).
We next define the notion of ergodicity for dynamical systems introduced above.
Definition 2.2
A dynamical system is called ergodic if one of the following equivalent conditions is fulfilled:
f measurable,
.
.
Henceforth we assume that the dynamical system is ergodic.
Remark 2.1
Note that for the condition (b) the implication has to hold, if the symmetric difference between
and B is a null set. It can be shown (see, e.g. [Citation7]) that ergodicity is also equivalent to an a priori weaker implication
For we write
, defining the realisation
. There is a natural unitary action on
associated with
(1)
(1) It can be shown that the conditions of Definition 2.1 imply that this is a strongly continuous group (see [Citation8]). It is often necessary that the set of full measure be invariant in the sense that together with the point ω it contains the whole ‘trajectory’
. This requirement can always be met on the basis of the following simple lemma (see [Citation8, Lemma 7.1]).
Lemma 2.1
Let be a set of full measure in Ω. Then there exists a set of full measure
such that
, and for a given
we have
for almost all
.
For each we define the infinitesimal generator
of the unitary group
by the formula
(2)
(2) where the limit is taken in
. Notice that
are commuting, self-adjoint, closed and densely defined linear operators on the separable Hilbert space
. The domain
of such an operator is given by the set of
-functions for which the limit (Equation2
(2)
(2) ) exists. We consider the set
(3)
(3) and similarly
It is shown by the standard semigroup property that
is dense in
. We also define the space
By the smoothening procedure discussed in [Citation8, p.232] (see also the text before Lemma 3.1 below), it is shown that
is dense in
for all
as well as in
for all k. Furthermore, it is shown that
is separable. Notice that, due to the infinitesimal generator being closed,
can be equivalently defined as the function that satisfies the property
(4)
(4) If
we may also define
for all
. It can be shown that the following identity holds (see [Citation9]):
(5)
(5) Moreover, for a.e.
the function
is the distributional derivative of
a proof of this fact can be found in [Citation9, Lemma A.7].
Following [Citation10], we denote by the seminorm on
given by
By
we denote the completion of
with respect to the seminorm
. The gradient operator
and the operator
are extended uniquely by continuity to mappings from
to
and from
to
respectively. Finally, by a density argument, it is easily seen that
is also the completion of
with respect to
.
2.2. Definition and basic properties
The key property of ergodic systems is the following theorem, due to Birkhoff (for a more general approach, see [Citation11]).
Theorem 2.1
‘Ergodic Theorem’
Let be a probability space with an ergodic dynamical system
on Ω. Let
, and let
be a bounded open set. Then for P-a.e.
one has
(6)
(6) Furthermore, for all
and a.e.
the function
satisfies
. For
one has
weakly in
as
.
The elements ω such that (Equation6(6)
(6) ) holds for every
and bounded open
are refereed to as typical elements, while the corresponding sets
are called typical trajectories. Note that the separability of
implies that almost every
is typical, and in what follows we only work with such ω.
For vector spaces we denote by by
their usual tensor product. We define the following notion of stochastic two-scale convergence, which is a slight variation of the definition given in [Citation12]. In [13], the authors average over the probability space and do not use the Birkhoff Ergodic Theorem. As a consequence, they do not obtain convergence almost everywhere but only in mean, which results in a weaker notion of stochastic two-scale convergence than the one introduced in [12]. In the context of calculus of variations, the first results are obtained in [14,15]. The authors of these papers do not use stochastic two-scale convergence at all, as this was introduced later on, but rely on a formula for non-periodic homogenisation for a.e. ω∈Ω as well as on the ergodic theorem. We shall stay in the Hilbert setting (p=2), as it suffices for our analysis. Finally, we denote by S a bounded open LIpschitz set in
.
Definition 2.3
Let be a typical trajectory and
a bounded sequence in
. We say that
weakly stochastically two-scale converges to
and write
if
(7)
(7)
If additionally we say that
strongly stochastically two-scale converges to u and write
.
Remark 2.2
The convergence of is defined along a fixed typical trajectory and a priori the limit depends on this trajectory. In applications, such as the analysis of the PDE family in Section 4, it often turns out that the limit does not depend on the trajectory chosen. For this reason, and to simplify notation, in what follows we often do not indicate this dependence explicitly.
Note also that, by density, the set of admissible test functions g in (Equation7(7)
(7) ) can be extended to
.
In the next proposition, we collect the properties of stochastic two-scale convergence that we use in the present work.
Proposition 2.1
The following properties of stochastic two-scale convergence hold.
Let
be a bounded sequence in
. Then there exists a subsequence
not relabelled
and
such that
.
If
then
.
If
is a bounded sequence with
in
for some
then
.
If
is uniformly bounded and
strongly in
for some
and
is bounded in
with
for some
then
.
Let
be a bounded sequence in
. Then on a subsequence (not relabelled)
in
and there exists
such that
Let
be a bounded sequence in
such that
is bounded in
. Then there exists
such that on a subsequence
(8)
(8)
Proof.
In view of analogies with the periodic case, we just give a sketch of the proof. A proof of (a) can be found in [Citation12, Lemma 5.1]. For the proof of (b), we take an arbitrary and calculate
We obtain the claim by approximating
with functions
. The proof of (c), (d) is straightforward. The proof of (e) goes in the same way as in the periodic case, by the duality argument.
First, one proves that if is such that
then there exists
such that
. One then proceeds in the same way as in the periodic case (see [Citation16]). In order to show the claim (f), take the subsequence such that
, where
and
, where
. We choose the test functions of the form
, where
and
, and using integration by parts we conclude
from which the claim follows by a density argument, in view of the property (Equation4
(4)
(4) ).
3. Problem formulation and auxiliary statements
Let be a bounded open Lipschitz set. We take
such that
and for each
consider its ‘realisation’
We assume that the following conditions are satisfied.
Assumption 3.1
For a.e. one has
(9)
(9) where:
are open connected sets with Lipschitz boundary;
One has
There exist
such that
There exists a sequence of disjoint bounded domains
such that
and
such that for all
the following extension property holds: for all
there exists
satisfying
It is easily seen that Assumption 3.1 holds for the examples given in Section 7.1. Denote by Λ the set of typical elements satisfying the conditions listed in Assumption 3.1, and for all
define
as the union of all components
that are subsets of S and stay sufficiently far from its boundary, in the sense that there exists
such that
(10)
(10) We denote the complement of the set
by
and the corresponding set indicator functions by
and
.
For each we consider the following Dirichlet problem in S: for
and
find
such that
(11)
(11) where
with a symmetric and positive-definite matrix
.
For all we also define the Dirichlet operator
in
corresponding to the differential expression
e.g. by considering the bilinear form
It is well known that the spectrum of is discrete. The following subspace of
will play an essential role in our analysis:
Notice that as a consequence of Ergodic Theorem (Theorem 2.1) one has
i.e.
consists of
-functions that vanish on
. Henceforth we assume that
without mentioning it explicitly.
The next two lemmas use a standard smoothening (or ‘mollification’) procedure, which we now describe. We take and (cf. [Citation8, p.232]) choose a nonnegative even function
with
and write
for all
. For each
we define the mollification
of g by
Notice that
(12)
(12) from which we infer that
Arguing by induction, we show that
and if
then
. Before we state and prove the lemmas, we introduce additional notation. We define the space
as well as the sets
Also, for all
we define the set
By using the density of
in
it can be seen that for all
the set
is measurable. Notice that for each fixed
where m is large enough, there exist constants
such that
(13)
(13) In the next lemma we assume that a relaxed version of the right inequality in (Equation13
(13)
(13) ) holds uniformly in ω.
Lemma 3.1
Suppose that for a.e. there exists a sequence of positive values
converging to zero, such that
Then the set
is dense in
.
Proof.
Using Ergodic Theorem and the assumption of the lemma, it can be shown that as
. To prove the density, it suffices to approximate
, where
by a function from
for which we use the above mollification procedure. Notice that for
small enough, one has
. It remains to check
as
, but this follows from the strong continuity of the group
see (Equation1
(1)
(1) ):
as required.
Notice that, by the standard Poincaré inequality, for each there exists C>0 such that
(14)
(14) In the following lemma we impose this condition uniformly.
Lemma 3.2
Assume that for a.e. there exists a constant C>0 such that
(15)
(15) and that (Equation14
(14)
(14) ) is satisfied for all k and large enough m. Then the set
is dense in
.
Proof.
We take and define
. Notice that as a consequence of (Equation5
(5)
(5) ),
and
. It is easily seen that
in
as
. Thus we can assume, without loss of generality, that
. We define
. It can be seen from the proof of Lemma 3.1 that
in
as
. Notice that for a.e.
we have, for
,
and therefore
Notice also that for a.e. there exist
such that for all
where m is sufficiently large, we have
where we have used (Equation12
(12)
(12) ), (Equation14
(14)
(14) ), (Equation15
(15)
(15) ) and Young's inequality. Using the Ergodic Theorem we conclude that there exists C>0 such that
from which the claim follows.
4. Limit equations and two-scale resolvent convergence
We define the quadratic form
and denote by
the completion of
with respect to the seminorm
. The proof of the following lemma is straightforward.
Lemma 4.1
For each there exists
such that
or, equivalently,
(16)
(16) In particular, one has
.
Remark 4.1
It follows from the observations in [Citation12, p.265–266] that if the following extension property is satisfied for a.e. : for all
there exists
and a sequence of functions
such that
where
is a constant independent of u and ϵ, then the matrix
is positive definite.
Notice that under Assumption 3.1, the extension property in Remark 4.1 is satisfied. We define the space
which is clearly a direct sum, naturally embedded in
. Before stating the next theorem we prove a simple lemma that implies that gives norm bounds for each component of H by the norm in
.
Lemma 4.2
Let and
such that
on
. Then there exists a constant C>0 such that
where we use the natural embedding
.
Proof.
By Cauchy-Schwartz inequality we have
and hence
(17)
(17) It remains to bound
by
, which is done by the triangle inequality:
By we denote the orthogonal projection. For
we have
Theorem 4.1
Under Assumption 3.1, let and suppose that
is a bounded sequence in
such that
. For each
consider the solution
to (Equation11
(11)
(11) ). Then for a.e.
one has
where
satisfy
(18)
(18)
(19)
(19)
Remark 4.2
The system (Equation18(18)
(18) ) and (Equation19
(19)
(19) ) is understood in the weak sense:
(20)
(20) Noting that
is a closed subspace of
and bearing in mind Lemma 4.2, it follows by the Lax-Milgram lemma that for all
the problem (Equation20
(20)
(20) ) has a unique solution in
. Its solutions for the right-hand sides
and
coincide. The solution of the Equation (Equation19
(19)
(19) ) has the form
(21)
(21) where
is the solution of (Equation19
(19)
(19) ) obtained by setting
and
is the solution of (Equation19
(19)
(19) ) obtained by setting
and f=0. Substituting (Equation21
(21)
(21) ) into (Equation18
(18)
(18) ), we obtain an equation on
.
Proof.
The proof follows a standard argument. First, by (Equation11(11)
(11) ), there exists a constant C>0 such that
(22)
(22) For each
we extend
using Assumption 3.1, to a sequence
which is bounded in
. From Proposition 2.1 we infer that there exist
,
,
such that on a subsequence we have
(23)
(23) To obtain the Equation (Equation18
(18)
(18) ), we take test functions of the form
in (Equation11
(11)
(11) ), where
,
and
. In the limit as
we obtain
(24)
(24) Setting
it follows that
and the characterisation (Equation16
(16)
(16) ) yields
a.e.
. Taking arbitrary
in (Equation24
(24)
(24) ), we obtain the ‘macroscopic’ part (Equation18
(18)
(18) ) of the limit problem. The ‘microscopic’ part (Equation19
(19)
(19) ) is obtained by taking test functions of the form
in (Equation11
(11)
(11) ), where
,
. The convergence of the whole sequence can be deduced by uniqueness of the solution of the system (Equation18
(18)
(18) )–(Equation19
(19)
(19) ).
Remark 4.3
The following observation was made in [Citation1] in the periodic setting. The formulation (Equation20(20)
(20) ) can be interpreted from the operator-theoretic point of view. Namely, we define a positive-definite operator
on a dense linear subset of
(which is a dense subset of H under the condition of Lemma 3.1), as follows. One takes
and defines the domain
as the set of solutions of (Equation20
(20)
(20) ) obtained for varying
. To see that
is dense in H, take the solutions
for
respectively. Setting
,
as the test function in the equation for
and
,
as the test function in the equation for
yields
Thus, if
then necessarily
which implies g=0. The operator
defined by
is a bounded linear mapping between Hilbert spaces, where the norm on
is given by
We shall need the following statement for the convergence of spectra of the operators associated with (Equation11(11)
(11) ). It is proved in the same way as the previous theorem, and we omit the proof.
Proposition 4.1
Under Assumption 3.1, let and suppose that
is a bounded sequence in
such that
. For each
let
be the solution of
(25)
(25) Then for a.e.
one has
where
is the solution of the problem
(26)
(26)
Remark 4.4
Theorem 4.1 and Proposition 4.1 are still valid if, instead of a fixed we take a sequence
such that
and
, for Theorem 4.1, i.e.
for Proposition 4.1, where
. Notice that
splits into the spectra of scaled Laplace operators on each inclusion contained in
(27)
(27) where
is defined in (Equation10
(10)
(10) ). Notice that there exists C>0 such that for all
the solution
of (Equation11
(11)
(11) ) satisfies
(28)
(28) and similarly the solution of (Equation25
(25)
(25) ) satisfies
(29)
(29)
In what follows we denote by the operator generated by the bilinear form
(30)
(30) As a consequence of Proposition 4.1 and Remark 4.4, we have the following statement.
Corollary 4.1
Assume that Assumption 3.1 holds. Then
Proof.
Take and
and define
. As a consequence of Remark 4.4, the sequence of solutions of (Equation25
(25)
(25) ) converges weakly two-scale to the solution of (Equation26
(26)
(26) ), which is a resolvent equation. Moreover, (Equation27
(27)
(27) ) and (Equation29
(29)
(29) ) imply the existence of C>0 such that
and therefore
.
5. Spectral completeness for inclusions
Next we prove that
We shall use the assumptions of Lemma 3.2 as well as assume that for each
there exists
such that for a.e.
the following implication holds:
(31)
(31) Notice that, by regularity theory, the above condition is satisfied for a fixed
and
whenever the boundary
is sufficiently regular. In what follows we use a sequence
that is dense in
where the constant
is defined in Assumption 3.1.
We will now define a sequence of random variables that is invariant for all whose realisation is such that the shape that contains the origin is the same. For
define the set
Lemma 5.1
For every is measurable.
Proof.
Notice that
(32)
(32) Since for each fixed
there is a countable set of lines satisfying the property (Equation32
(32)
(32) ), the set
is measurable.
We define the random variables
Notice that
whenever
and also, due to the assumption,
for a.e.
. We denote by F the random vector
(33)
(33) For a.e.
,
we define the set
Furthermore, we introduce the set
, which is a translation of the set
containing the origin:
Finally, we define a characteristic function of the translation of the set
and a measurable function of ω taking values in
:
(34)
(34) Notice that for a.e.
one has
.
Lemma 5.2
For every the function
taking values in
is measurable with respect to the Borel σ-algebra on
.
Proof.
Firstly notice that
(35)
(35) is a measurable mapping taking values in the set
, with Borel σ-algebra. To check this notice that for each
the set
is measurable: the related proof is similar to that of Lemma 5.1. Further, for
the norm
is written as a limit of Riemann sums, and each Riemann sum can be written in terms of a finite number of
and values of function
. Thus
is measurable. Since the topology in
is generated by the balls of the form
, where
and
we have that the mapping given by (Equation35
(35)
(35) ) is measurable. The claim follows by using the fact that the convolution is a continuous (and thus measurable) operator from
to
.
Notice that by construction
is a dense subset of
for a.e.
(see also the proof of Lemma 3.2). For
we introduce the following subset of
(36)
(36) For
and a.e.
we also define
as follows:
Finally, for every
and
we define the random variable
(37)
(37)
Lemma 5.3
For every and
, the function
is measurable.
Proof.
We use Lemma 5.2 and the fact that is a continuous map from
to
and
is a measurable function, since
Lemma 5.4
For the set
is measurable.
Proof.
The claim follows by observing that
Now we are going to define a measurable mapping from to the subspace
. We set it to be an
-projection onto
of a specially chosen function of x and ω. We need the following measurability lemma.
Lemma 5.5
Assume that is a measurable function taking values in
for a.e.
. Then the
-distance
is a measurable map.
Proof.
The claim follows from the formula
For and
we define a measurable map
as follows:
where
(38)
(38) and
is the minimal value of m in (Equation38
(38)
(38) ) Notice that in this way for a.e.
the
-projection of
on
is not zero. We also define the random variable
in the following way:
By invoking the measurability of
see Lemma 5.1, it is easily seen that R is indeed measurable. Next, for
, l>0 we define the random variable
by
where, for all
is the corresponding minimal valueFootnote1 of m, and
is the ball with the centre at D and radius
see (5.1). We also define
Notice that in this way
is the value at the origin (taking into account for
the relative position of the origin with respect to the shape) of the (unique)
-projection of
onto
. As a consequence of (Equation31
(31)
(31) ), we have
Notice that by construction
if
. We are ready for the proof of main statement.
Theorem 5.1
Under Assumption 3.1, the assumption of Lemma 3.2 and (Equation31(31)
(31) ), one has
Proof.
We take . There are two possibilities:
There exists
such that
has zero probability. In this case we denote
For all
the set
has positive probability.
In the case (a), by the continuity of probability, we conclude that where (cf. (Equation36
(36)
(36) ))
By Lemma 2.1 and Corollary 4.1 we infer that
In particular, we conclude that
.
In the case (b) we construct a Weyl sequence showing that . To this end, we define
Then, by the above construction and using Ergodic Theorem, one has
It follows from the above that
consists of exactly those
that satisfy the property (b). The set
is closed, hence its complement is a countable union of open disjoint intervals. Every element of such an interval
satisfies the property (a) with
and therefore
. Using Lemma 2.1, we obtain
The claim follows since there is only countable number of such intervals.
6. Convergence of spectrum
In our analysis we keep in mind the examples set in Section 7, for which it is shown that . In the present section we assume that this holds, as well as the conclusion of Theorem 5.1, i.e.
(39)
(39) We are interested in approximating the spectra
of the operators
(see Section 3) by the spectrum
of the limit operator. We claim that
for a.e.
where the convergence is understood in the Hausdorff sense:
For all
there are
such that
.
If
and
, then
.
We prove this claim by adapting the argument of [Citation17]. First, we introduce the notion of strong resolvent convergence.
Definition 6.1
Let and
be the operators acting on
and on
respectively. We say that
strongly two-scale resolvent converge to
and write
if
It can be shown that the property (a) is satisfied if we have strong two-scale resolvent convergence (see the proof of [Citation17, Proposition 2.2]). Theorem 4.1 shows that the following implication holds:
It can be shown that this is equivalent to strong two-scale resolvent convergence (see [Citation17, Proposition 2.8]) and thus the property (a) is satisfied.
In order to prove (b), we start from the eigenvalue problem for the operator (it has a compact resolvent and its spectrum is discrete), i.e.
(40)
(40) If we have that
and
, then we would also have
. However, the problem would be if u=0, because then
. The next lemma tells us if
then necessarily the sequence of eigenvalues is compact with respect to the strong two-scale convergence and thus s belongs to the point spectrum of the operator
, since then necessarily
.
Theorem 6.1
Suppose that (Equation39(39)
(39) ) holds and that for each
satisfy (Equation40
(40)
(40) ). If
then for a.e.
the sequence
is compact in the sense of strong two-scale convergence.
Proof.
satisfies
We use Assumption 3.1 and for each ϵ extend
, denoting the extensions by
. Notice that there exists C>0 such that
(41)
(41) The difference
satisfies:
(42)
(42) From the estimate (Equation41
(41)
(41) ) we see that
is weakly compact in
and thus there exists
such that
which immediately implies
. Furthermore, as a consequence of (Equation27
(27)
(27) ), (Equation29
(29)
(29) ) and (Equation39
(39)
(39) ), the following estimate holds for some
Therefore, from Proposition 4.1 and Remark 4.4 we conclude that
, where the limit z satisfies
(43)
(43) We also consider the problem
(44)
(44) In the same way as before we conclude that for some C>0:
Analogously, we conclude that
which satisfies
(45)
(45) By testing (Equation42
(42)
(42) ) with
and (Equation44
(44)
(44) ) with
we conclude
Finally, by testing (Equation43
(43)
(43) ) with
and (Equation45
(45)
(45) ) with
and integrating over S we conclude
which completes the proof.
7. Spectrum of the limit operator: examples
This section is devoted to the description of the spectrum of the limit operator. Since it crucially depends on the intrinsic properties of the microscopic part of the operator and the properties of the probability space, it does not seem feasible (at least at the current stage of research in this area) to provide a characterisation of the spectrum in a general setting. We shall consider several interesting, from the point of view of applications, examples of probability spaces and configurations of soft inclusions. The general example of a finite number of shapes of randomly varying size is described in Section 7.1. Then we consider the case of a single shape of fixed size in Section 7.2, and the case of a single shape of randomly varying size in Section 7.3, for which we provide the full description of the spectrum of the limit operator with the proofs. The characterisation of the spectrum in the general case of Section 7.1 is analogous to the case of a single shape considered in Section 7.3.
7.1. The setting of finite number of shapes of varying size
Let be a sequence of independent and identically distributed random vectors taking values in
where
and
is an appropriate probability space. We also assume that we have a finite number of shapes
that represent the inclusions, where the first and the second components of
model the shape and the size, respectively. We also set
. On
there is a natural shift
, which is ergodic. We next state the discrete analogue of Lemma 2.1.
Lemma 7.1
Assume that is a set of full measure. Then there exists a subset
of full measure such that for each
we have
.
We treat as a probability space with Lebesgue measure dy and the standard algebra
of Lebesgue measurable sets, and define
On Ω we define a dynamical system
. We consider the set
. It is easily seen that
is measurable. For a fixed
the realisation
consists of the inclusions
. Next, we describe the generators
in the present example. Taking
and using the above lemma, note that there exists a subset of full measure
such that for all
and
we have
. It is clear that for
one has
Using this fact and the statement following (Equation5
(5)
(5) ), we infer that
and
(46)
(46)
7.2. Simple example
In this section we set so that
and, by a standard procedure, see e.g. [Citation18], identify the elements of the probability space
with sequences
whose components
take values in the two-element set
. Let
be an open subset of Y whose closure is contained in Y (‘soft inclusion’). The value 0 or 1 of
corresponds to the absence or the presence of the inclusion in the ‘shifted cell’
respectively. We also set
Then, for a given
the realisation
is the union of the sets (‘inclusions’)
over all
such that
. For this example the space
consists of all functions of the form
(47)
(47) It is also important to understand how one applies the stochastic gradient. For a function
we have (see (Equation46
(46)
(46) ))
(48)
(48) Consider formally the spectral problem for the limit operator:
(49)
(49)
(50)
(50) We write the solution to the ‘microscopic’ Equation (Equation50
(50)
(50) ) in the form
, where (recall Remark 4.2)
(51)
(51) In other words, v is given by (Equation47
(47)
(47) ) with
satisfying
(52)
(52) whenever
such that
and
otherwise.
We label the eigenvalues of the operator in (Equation52(52)
(52) ) in the increasing order, where we repeat multiple eigenvalues, so that
and
are, respectively, the eigenvalues whose eigenfunctions
have non-zero integral over
and the eigenvalues whose eigenfunctions
have zero integral over
. Following [Citation1], we write the solution to (Equation52
(52)
(52) ) via the spectral decomposition
(53)
(53) and thereby
(54)
(54) Substituting the obtained representation for
into the ‘macroscopic’ equation (Equation49
(49)
(49) ) yields
(55)
(55) where
(56)
(56) is a stochastic version of the ‘Zhikov function’ β in [Citation1]. Assume for the moment that
. Then the intervals where
are the ‘spectral bands’ of
and additionally a Bloch-type spectrum is given by
. The set
corresponds to the gaps in the spectrum of
.
In the setting of this paper, namely, for a bounded domain instead of each spectral band
lying to the left of
we have a ‘band’ of discrete spectrum: a countable set of eigenvalues
(57)
(57) with the accumulation point at the right end
of each band, where
are the eigenvalues of the operator
defined by the form
(58)
(58) The Bloch-type spectrum of
consists of eigenvalues
of infinite multiplicity with eigenfunctions of the form
with
and
(59)
(59) Summarising, the spectrum of
is given by
(60)
(60)
7.3. More advanced example
Here we allow the inclusions to randomly change size, so that
. By analogy with the previous section, we assume that
consists of sequences
such that
. We also assume that the restriction to to
of the probability measure on
is absolutely continuous with respect to Lebesgue measure. As before, consider
and denote by
, where
is the centre of Y, the ‘scaled inclusion’, requiring that
, in order for the extension property in Assumption 3.1 to hold. The values 0 or
of
correspond to the absence of an inclusion or the presence of the inclusion
in the cell
respectively. Furthermore, define
. Then a realisation
is the union of the inclusions
for all
where in the case
we set
. The space
consists of functions of the form
(61)
(61) Consider the spectral problem for
namely
and separate the variables, as in Section 7.2:
where the function v satisfies
(62)
(62) The stochastic gradient is given by
and therefore the problem (Equation62
(62)
(62) ) is equivalent to
(63)
(63) For each
the eigenvalues
and (orthonormal) eigenfunctions
of the operator
acting in
are obtained by scaling the eigenvalues and eigenfunctions of
acting in
, in particular,
,
. Therefore, the formula (Equation53
(53)
(53) ) with
replaced by
gives the solution to
(64)
(64) If
and the set
has gaps, then for
the solution to (Equation63
(63)
(63) ) is given by (Equation61
(61)
(61) ), where the functions
solve (Equation64
(64)
(64) ) with
. Substituting it into the spectral problem for (Equation49
(49)
(49) ) yields the problem (Equation55
(55)
(55) ) with the Zhikov-type function β given by (cf. (Equation56
(56)
(56) ))
(65)
(65) The integral in (Equation65
(65)
(65) ) is well defined for
and the description of the spectrum on the intervals where
follows Section 7.2.
Theorem 7.1
Under the assumptions of the current subsection, the spectrum of is given by
where for each k, the values
are solutions to
see (Equation57
(57)
(57) ). The point spectrum of the operator
is given by
.
It is clear that if the set has gaps, then
also has gaps. We are going to prove the theorem in several steps formulated in the following lemmas. We begin by studying the spectrum of the ‘microscopic’ part of the limit operator.
Lemma 7.2
The spectrum of the operator (see (Equation30
(30)
(30) )) is given by
and does not contain eigenvalues of
.
Proof.
Let for some
and assume that
is an eigenfunction corresponding to λ, i.e.
. (For
argument is similar.) Then v is of the form (Equation61
(61)
(61) ), where
in
whenever
. But λ is only an eigenvalue of the operator
acting in
if
, hence
(66)
(66) It remains to observe that
is a set of measure zero in Ω and hence
. The second claim of the lemma follows.
Now we show that by constructing a Weyl sequence. Without loss of generality we can assume that
. For small enough
we choose an
-function
such that
and
e.g. we can choose
to be equal to a constant proportional to
on
. Consider the sequence
We have
and
hence
It follows that
is a Weyl sequence for
.
It remains to prove that λ is in the resolvent set whenever . Assume the contrary and let
then the resolvent equation
has a unique solution given by (Equation61
(61)
(61) ) with
solving
. Moreover, since
we have
and it follows immediately that
which concludes the proof.
Next, we focus on the spectrum of .
Lemma 7.3
The inclusion holds.
Proof.
The proof of the inclusion repeats the related part of the proof of the Lemma 7.2. Namely, for
,
,
we define a Weyl sequence
where
and
is given by
with an arbitrary fixed
and
defined as in Lemma 7.2. In order to show that
suppose that
for some
and
. Assume, to the contrary, that there exists a bounded resolvent
i.e. the system (Equation18
(18)
(18) )–(Equation19
(19)
(19) ) has a unique solution for all
. For
the second equation reads
Then
must necessary be of the form
, where v is of the form (Equation61
(61)
(61) ) and
in
i.e.
(67)
(67) which clearly blows up as
. We show that the corresponding v is not an element of
leading to a contradiction. Indeed, using the identity
one has
where the last integral diverges.
Lemma 7.4
If for some
then λ is an eigenvalue of
.
Proof.
Let , and denote by
be the corresponding eigenfunction of
. Since
the problem
has a solution given by (Equation61
(61)
(61) ), (Equation67
(67)
(67) ). Therefore
is the eigenfunction of
corresponding to λ.
Lemma 7.5
A point λ belongs to the resolvent set of if
and
i.e.
or
and
.
Proof.
We claim that the problem (Equation18(18)
(18) ) and (Equation19
(19)
(19) ) has bounded resolvent. Indeed, suppose that
and write (Equation19
(19)
(19) ) in the form
Since λ is not in the spectrum of
, the latter has a bounded resolvent at λ and
, where
is as in (Equation61
(61)
(61) ) and
. In particular,
Substituting the expression for
in (Equation18
(18)
(18) ) we obtain
For
the operator
is invertible and
from which the claim follows.
Proposition 7.1
The set does not contain eigenvalues of the operator
.
Proof.
Assume that , for some
and
, is an eigenvalue of
, i.e. there exists
such that
(68)
(68) Suppose that
for some
then
, where v solves
(69)
(69) Arguing as for the second inclusion of Lemma 7.3, we see that (Equation69
(69)
(69) ) has no
-solution for the given λ. It follows that
and therefore
is an eigenfunction of
which cannot be true by Lemma 7.2.
Now we assume that . Arguing as above, for
we have
, where v solves (Equation69
(69)
(69) ). The solution exists and is given by (Equation61
(61)
(61) ), (Equation67
(67)
(67) ). Substituting
into (Equation68
(68)
(68) ) we see that
must satisfy
which cannot be true since
.
Finally, if then we argue as above for the case
again arriving at a contradiction.
This completes the proof of Theorem 7.1.
Disclosure statement
No potential conflict of interest was reported by the authors.
Additional information
Funding
Notes
1. The function is an ‘approximate eigenfunction’ for
see (Equation37
(37)
(37) ).
References
- Zhikov VV. On an extension of the method of two-scale convergence and its applications. (Russian) Mat Sb. 2000;191(7):31–72. Translation in Sb Math. 2000;191(7–8):973–1014.
- Bellieud M. Homogenization of stratified elastic composites with high contrast. SIAM J Math Anal. 2017;49(4):2615–2665. doi: 10.1137/15M1012050
- Bouchitté G, Bourel C, Manca L. Resonant effects in random dielectric structures. ESAIM Contr Optim Calculus Variations. 2015;21:217–246. doi: 10.1051/cocv/2014026
- Bourgeat A, Mikelić A, Piatnitski A. On the double porosity model of a single phase flow in random media. Asymptotic Anal. 2003;34:311–332.
- Cherednichenko K, Cooper S. Resolvent estimates for high-contrast homogenisation problems. Arch Ration Mech Anal. 2016;219(3):1061–1086. doi: 10.1007/s00205-015-0916-4
- Cherednichenko K, Cherdantsev M, Cooper S. Extreme localisation of eigenfunctions to one-dimensional high-contrast periodic problems with a defect. 2017. arXiv:1702.03538.
- Cornfeld IP, Fomin SV, Sinai YG. Ergodic Theory. New York: Springer-Verlag; 1982. (Grundlehren der Mathematischen Wissenschaften; 245). Translated from the Russian by A. B. Sossinskii.
- Zhikov VV, Kozlov SM, Oleĭnik OA. Homogenization of differential operators and integral functionals. Berlin: Springer-Verlag; 1994.
- Duerinckx M, Gloria A. Stochastic homogenization of nonconvex unbounded integral functionals with convex growth. Arch Ration Mech Anal. 2016;221(3):1511–1584. doi: 10.1007/s00205-016-0992-0
- Sango M, Woukeng J-L. Stochastic two-scale convergence of an integral functional. Asymptot Anal. 2011;73(1–2):97–123.
- Ackoglu MA, Krengel U. Ergodic theorems for superadditive processes. J Reine Angew Math. 1981;323:53–67.
- Zhikov VV, Pyatnitskiĭ AL. Homogenization of random singular structures and random measures. Izv Ross Akad Nauk Ser Mat. 2006;70(1):23–74. doi: 10.4213/im554
- Bourgeat A, Mikelić A, Wright S. Stochastic two-scale convergence in the mean and applications. J Reine Angew Math. 1994;456:19–51.
- Dal Maso G, Modica L. Nonlinear stochastic homogenization. Ann Mat Pura Appl. 1986;144(4):347–389. doi: 10.1007/BF01760826
- Dal Maso G, Modica L. Nonlinear stochastic homogenization and ergodic theory. J Reine Angew Math. 1986;368:28–42.
- Allaire G. Homogenization and two-scale convergence. SIAM J Math Anal. 1992;23(6):1482–1518. doi: 10.1137/0523084
- Zhikov VV. Gaps in the spectrum of some elliptic operators in divergent form with periodic coefficients. (Russian) Algebra i Analiz. 2004;16(5):34–58. Translation in St Petersburg Math J. 2005;16(5):773–790.
- Shiryaev AN. Probability. New York: Springer-Verlag; 1996.