Exponential Decay and Lack of

Analyticity for the System of the

Kirchhoff–Love Plates and
Membrane-Like Electric Network
arXiv:1908.04826v1 [[Link]] 13 Aug 2019

Equation with fractional partial damping

Fredy Maglorio Sobrado Suárez∗
Department of Mathematics, Federal University of Technology of Paraná, Brazil
and
Filomena Barbosa Rodrigues Mendes
Department of Electrical Engineering, Federal University of Technology of Paraná, Brazil

Abstract

The emphasis in this paper is on the Coupled System of a Kirchhoff-Love

Plate Equation with the Equation of a Membrane-like Electrical Network,
where the coupling is of higher order given by the Laplacian of the displace-
ment velocity γ∆ut and the Laplacian of the electric potential field γ∆vt ,
here only one of the equations is conservative and the other has dissipative
properties. The dissipative mechanism is given by an intermediate damping
(−∆)θ vt between the electrical damping potential for θ = 0 and the Laplacian
of the electric potential for θ = 1. We show that S(t) = eBt is not analytic for
θ ∈ [0, 1], however S(t) = eBt decays exponentially for 0 ≤ θ ≤ 1.

Key words and phrases: Electric Network Equation, Kirchhoff-Love Plates,

Lack of Analiticity, Exponential Decay.

1 Introduction
In the literature there are several mathematical models that model a single electrical
network connecting piezoelectric actuators and/or transducers, see for example [20],
Corresponding author.
∗

Email address: fredy@[Link] (Fredy Maglorio Sobrado Suárez), filomena@[Link]

(Filomena Barbosa Rodrigues Mendes).

1
[22] and [21]. In Particular in [22], equations (2b) and (2c). We have, for example,
the equations of a second order electric transmission line with zero order or second
order dissipation:
(S,Z)-network: second-order network with zeroth-order dissipation

vtt − β2 ∆v + δ0 vt = 0 (1)

(S,S)-network: second-order network with second-order dissipation

vtt − β2 ∆v − δ2 ∆vt = 0. (2)

Here we denote by v(x, t) the time-integral of the electric potential difference between
the nodes and the ground.
The paper deal with the following coupled system of the Kirchhoff-Love Plates
and Membrane-Like Eletric Network deduced in [20]:

utt + α∆2 u − γ∆vt = 0, x ∈ Ω, t > 0, (3)

vtt − β∆v + γ∆ut + δvt + δγ∆u = 0, x ∈ Ω, t > 0, (4)

satisfying the boundary conditions

u = ∆u = 0, v = 0, x ∈ ∂Ω, t > 0, (5)

and prescribed initial data

u(x, 0) = u0 (x), ut (x, 0) = u1 (x), v(x, 0) = v0 (x), vt (x, 0) = v1 (x), x ∈ Ω. (6)

Here we denote by u(x, t) the transversal displacements of the plates and v(x, t) is
time- integral of the electric potential difference between the nodes and the ground,
at time t > 0 and x ∈ Ω ⊂ Rn the domain with smooth boundary ∂Ω, the coefficient
DP 1
α := 2 2
= 2
MP l ω π
in which :
DP := JI (2µL + λL ), is bending stiffness, µL and λL are the Lamê moduli and
3
JI := 2h3 . 2h is the thickness of the thin plate is assumed to be small compared
with the l length the diameter of the plate, MP := 2ρl2 h the total
q mass of the plate,
π DP
ρ is the mass density and the characteristic pulsation ω := l M P
, the coefficient δ
is positive and is defined as
RN
δ :=
LN ω
where RN and LN are respectively the net-inductance and net-resistance. The cou-
pling coefficient r
gme 1
γ :=
lω MP CN

2
where CN is net-capacitance per unit area of the plate, gme is eletro-mechanical
coupling coefficient of the constitutive equation of the properties of the single PZT
actuator:   

 M 11 
 gmm

0 −gme  w,11 

M22 =  0 gmm −gme  w,22 (7)
 Q 
  gme gme gee
 . 
ψ
d2
where gmm is the actuator bending stiffness, gee is capacitance. Equation (7) estab-
lishes that the actuators can supply electrically induced moments only in two fixed
material orthogonal directions and that they cannot supply ‘mixed’ moments; the
orthogonal system introduced accounts for this directionality. Moreover we have
assumed that the piezoelectric stiffnesses in e1 and e2 are equal; this assumption
seems reasonable when using PZT actuators exploiting Poisson effect. where Mii
and w,ii are the piezo-electrically induced bending moments and curvatures, while
.
Q
d 2 and ψ are the charge per unit area and voltage between the actuator plates.
1
β := .
LN CN l2 ω 2
The systems (3)-(4) they set up indirect damping mechanisms which arise, not from
insertion of damping terms into the original equations describing the mechanical mo-
tion, but by coupling those equations to further equations describing other processes
in the structure.
In this paper we study the asymptotic behavior and analyticity of the abstract
model with fractional damping (−∆)θ vt to 0 ≤ θ ≤ 1 (9)-(10), being that the
equation (10) models the electric network and that in the particular case for θ = 0
and θ = 1 are the mathematical models (1) and (2) of [20]. But with a single
coupling term in each of the equations of the system, given by γAvt and −γAut
respectively. For more details on modeling see [20, 21].
Let Ω a bounded set in Rn with smooth boundary and given the operator:
A : D(A) ⊂ L2 (Ω) → L2 (Ω), where
A = −∆, D(A) = H 2 (Ω) ∩ H01 (Ω). (8)
It is known that this operator given in (8) is selfadjoint, positive and has compact
inverse. Using this operator, given the system (3)–(6), can be written in an ab-
stract way as follows: u, v defined in some suitable subespace of X = L2 (Ω) that
1
contemplates the boundary conditions (5) (for example D(A 2 ) = H01 (Ω))) such that
satisfies the equations
utt + αA2 u + γAvt = 0, x ∈ Ω, t > 0, (9)
vtt + βAv − γAut + δAθ vt = 0, x ∈ Ω, t > 0, (10)
the initial data
u(0) = u0 , ut (0) = u1 , v(0) = v0 , vt (0) = v1 , (11)
and satisfying the boundary conditions
u = Au = 0, v = 0, x ∈ ∂Ω, t > 0, (12)

3
where Ω be a bounded open set of Rn with smooth boundary ∂Ω.
In the last decades, many researchers have focused their efforts in the study
of the asymptotic stability of several coupled systems with indirect damping (This
terminology was used by Russell in his work [24]). Systems of two coupled equations
as wave-wave, plate-plate or plate-wave equations with indirect damping inside of
their domains, or on their boundaries, were studied by several authors. We going
are briefly mention some of them:
Alabau et al. in [2]. They considered an abstract evolution equations given by:
utt + A2 u + αv = 0, x ∈ Ω, t > 0,
vtt + A1 v + Bvt + αu = 0, x ∈ Ω, t > 0,
in which Ω be a bounded open set of Rn with smooth boundary ∂Ω and A1 , A2 are
self-adjoint positive linear operators in Hilbert space and B is a bounded operator.
When A1 = ∆, A2 = ∆2 and B is the identity operator, this system models can
be obtained by coupling two partial differential equations of different orders. Con-
sider for instance the following wave-Petrowsky system, where β > 0, with partial
3/2
frictional damping. For this case, they showed that, if 0 < |α| < CΩ and
v0 ∈ H 3 (Ω) ∩ H02 (Ω), u0 ∈ H 6 (Ω) ∩ H03 (Ω),
v1 ∈ H 2 (Ω) ∩ H01 (Ω), u1 ∈ H 4 (Ω) ∩ H02 (Ω).
then the energy of the solution satisfies, for every t > 0, the estimate
Z
(|∂t v|2 + |∇v|2 + |∂t u|2 + |∆u|2 )dx
Ω
C
≤ (kv0 k23,Ω + ku0k6,Ω + kv1 k22,Ω + ku1 k24,Ω )
t
In this direction other results can be found in [4, 5, 6, 9, 13, 19].
Alabau et al. [4] (see also [2, 1, 5]) considered an abstract system of two coupled
evolution equations with applications to several hyperbolic systems satisfying hybrid
boundary conditions. They have shown the polynomial decay of their solutions
using energy method and multiplicative techniques. Tebou [27] considered a coupled
system of wave-plate equations with indirect damping mechanisms. He showed this
system is not exponentially stable when the damping acts either in the plate equation
or in the wave equation and a polynomial decay of the semigroup was showed.
Recently, Guglielmi [9] considered an abstract hyperbolic system which generalizes
some wave-plate equations. When the wave equation is frictionally damped, he
proved that this system is not exponentially stable and a polynomial decay was
obtained. No result about the optimal decay rate was provided. Many other papers
were published in this direction, some of them can be viewed in [5, 15, 16, 23, 27, 29].
Now we will mention some concrete problems that motivated the elaboration of
this paper:
Dell’Oro et al. in [10]. They considered the abstract system given by:
utt + γAutt + A2 u − Aσ φ = 0, x ∈ Ω, t > 0,
φt + Aφ + Aσ ut = 0, x ∈ Ω, t > 0,

4
where Ω be a bounded open set of Rn with smooth boundary ∂Ω and A is a positive
self-adjoint operator. When A = −∆ this system models a thermoelastic plate.
They showed that the semigroup of this system is exponentially stable if and only
if σ ≥ 1. Moreover, when 1/2 ≤ σ < 1, they proved that the semigroup decays
polynomially to zero as t−1/(4−4σ) for initial data in the domain of the semigroup
generator, and such a decay rate is optimal. In this same work, they also showed
that, for the case γ = 0 and 0 ≤ σ < 1/2, the semigroup decays polynomially with
the optimal rate t−1/(1−2σ) . Similar works can be found in [3, 6, 12, 16, 25, 26, 28].
Finally a more recent result with fractional dissipation was published this year,
see Oquendo-Suárez(2019)[14]. An abstrac system of two coupled plate equations
was studied

ρ1 utt + γ1 Autt + β1 A2 u + αv = 0, x ∈ Ω, t > 0,

ρ2 vtt + γ2 Avtt + β2 A2 v + αu + κAθ vt = 0, x ∈ Ω, t > 0,

where Ω be a bounded open set of Rn with smooth boundary ∂Ω and one of these
equations is conservative and the other has fractional dissipative properties given
by Aθ vt , where 0 ≤ θ ≤ 1 and A = −∆. It is shown that the semigroup decays
polynomially with a rate that depends on θ and some relations between the structural
coefficients of the system. Have also shown that the rates obtained are optimal.
The remainder of this paper is organized as follows: in section 2, we study the
well-posedness of the system (9)-(12) through the semigroup theory. We left our
main results for the last two sections. In section 3 was introduced the study the
asymptotic behavior of the semigroup S(t) = eBt , where we show the exponential
decay to 0 ≤ θ ≤ 1 and finally subdivide section 4 into two subsections for using
hipotese to operator given in (8) is selfadjoint, positive and has inverse compact,
show the lack of analyticity of the semigroup S(t) = eBt : in the subsection we show
for 0 ≤ θ < 1 and in subsection 4.2 we show for the case θ = 1.

2 Well-Posedness of the System

In this section we will use the semigroup theory for assure the existence and unique-
ness of strong solutions for the system (9)-(12) where the operator A is given by
(8). Before this, let us see some preliminary results. It is important recalling that A
is a positive self-adjoint operator with compact inverse on a complex Hilbert space
X = L2 (Ω). Therefore, the operator Aθ is self-adjoint positive for θ ≥ 0, bounded
for θ ≤ 0, and using the embedding

D(Aθ1 ) ֒→ D(Aθ2 ),

is continuous for θ1 > θ2 . Here, the norm in D(Aθ ) is given by kukD(Aθ ) := kAθ uk,
u ∈ D(Aθ ), where k · k denotes the norm in the Hilbert space X. Some of these
spaces are: D(A1/2 ) = H01 (Ω), D(A0 ) = L2 (Ω) and D(A−1/2 ) = H −1 (Ω).
Now, we will use a semigroup approach to study the well-posedness of the
system (9)-(12). Taking w = ut , vt = z and considering U = (u, v, w, z) and

5
U0 = (u0 , v0 , u1 , v1 ), the system (9)–(12), can be written in the following abstract
framework
d
U(t) = BU(t), U(0) = U0 , (13)
dt
where the operator B is given by
 
2 θ
BU := w, z, −αA u − γAz, −βAv + γAw − δA z , (14)

for U = (u, v, w, z). This operator will be defined in a suitable subspace of the
phase space
1
H := D(A) × D(A 2 ) × D(A0 ) × D(A0 ).
It’s a Hilbert space with the inner product
1 1
hU1 , U2 i := αhAu1, Au2 i + βhA 2 v1 , A 2 v2 i + hw1 , w2 i + hz1 , z2 i,
for Ui = (ui , vi , wi , zi ) ∈ H, i = 1, 2. In these conditions, we define the domain of B
as
n 1
o
D(B) := U ∈ H : (w, z) ∈ D(A) × D(A 2 ),(−αAu, βv − δAθ−1 z) ∈ [H 2 (Ω)]2 .

To show that the operator B is the generator of a C0 - semigroup we invoke a

result from Liu-Zheng’ book.

Theorem 1 (see Theorem 1.2.4 in [11]) Let A be a linear operator with domain
D(A) dense in a Hilbert space H. If A is dissipative and 0 ∈ ρ(A), the resolvent set
of A, then A is the generator of a C0 - semigroup of contractions on H.

Let us see that the operator B satisfies the conditions of this theorem. Clearly,
we see that D(B) is dense in H. Effecting the internal product of BU with U, we
have
RehBU, Ui = −δkAθ/2 zk2 dx, ∀ U ∈ D(B), (15)
that is, the operator B is dissipative.
To complete the conditions of the above theorem, it remains to show that 0 ∈
ρ(B). Let F = (f1 , f2 , f3 , f4 ) ∈ H, let us see that the stationary problem BU = F
has a solution U = (u, v, w, z). From the definition of the operator B given in (14)
, this system can be written as
w = f1 , αA2 u = −[γAf2 + f3 ], (16)
z = f2 , βAv = γAf1 − δAθ f2 − f4 . (17)
This problem can be placed in a variational formulation: to find t = (u, v) such that
1
b(t, z) = hh, zi, ∀ z = (z1 , z2 ) ∈ D(A) × D(A 2 ). (18)
where
h = (−[γAf2 + f3 ], γAf1 − δAθ f2 − f4 ) ∈ D(A0 ) × D(A0 ) and
1 1
b(t, z) := αhAu, Az1 i + βhA 2 v, A 2 z2 i.

6
 1
The Proof of the coercivity of this sesquiline form in Hilbert space D(A)×D(A 2 )
is immediate, now, applying the Lax-Milgram Theorem and taking into account the
first equations of (16)-(17) we have a unique solution U ∈ H. As this solution
satisfies the system (16)-(17) in a weak sense, from these equations we can conclude
that U ∈ D(B). Finally, as
1
b(t, t) = αkAuk2 + kA 2 vk2

From second equations of (16)-(17), applying Cauchy-Schwarz and Young inequal-

ities to the second member of this inequality, for ε > 0 there exists Cε > 0, such
that
1
αkAuk2 + βkA 2 vk2 ≤ Cε kF k2 .

This inequality and the first equations of (16)-(17) imply that kUk ≤ CkF k, then 0
belongs to the resolvent set ρ(B). Consequently, from Theorem 1 we have B is the
generator of a contractions semigroup.
As B is the generator of a C0 -semigroup the solution of the abstract system (13)
is given by U(t) = etB U0 , t ≥ 0. Thus, we have shown the following well-posedness
theorem:

Theorem 2 (see [17]) Let us take initial data U0 in H then there exists only one
solution to the problem (13) satisfying

U ∈ C([0, ∞[; H).

Moreover, if U0 ∈ D(B) then the solution satisfies

U ∈ C([0, ∞[; D(B)) ∩ C 1 (0, T ; H).

3 Stability Results
In this section, we will study the asymptotic behavior of the semigroup of the system
(9)-(12).
First we will use the following spectral characterization of exponential stability
of semigroups due to Gearhart[8](Theorem 1.3.2 book of Liu-Zheng ) and to study
analiticity we will use a characterization of the book of Liu-Zheng (Theorem 1.3.3).

Theorem 3 (see [11]) Let S(t) = eBt be a C0 -semigroup of contractions on a

Hilbert space. Then S(t) is exponentially stable if and only if

ρ(B) ⊇ {iλ/λ ∈ R} ≡ iR (19)

and
lim sup k(iλI − B)−1 k < ∞ (20)
|λ|→∞

holds.

7
Theorem 4 (see [11]) Let S(t) = eBt be C0 -semigroup of contractions on a Hilbert
space. Suppose that
ρ(B) ⊇ {iλ/λ ∈ R} ≡ iR
Then S(t) is analytic if and only if
lim sup kλ(iλI − B)−1 k < ∞. (21)
|λ|→∞

holds.

It follows, C will denote a positive constant that assumes different values in

different places and the coupling coefficient α assumed positive. The results remain
valid when this coefficient is negative.
First, note that if λ ∈ R and F = (f1 , f2 , f3 , f4 ) ∈ H then the solution U =
(u, v, w, z) ∈ D(B) of the stationary system (iλI − B)U = F can be written in the
form
iλu − w = f1 (22)
iλv − z = f2 (23)
iλw + αA2 u + γAz = f3 (24)
iλz + βAv − γAw + δAθ z = f4 , (25)
We have
θ
δkA 2 zk2 = Reh(iλ − B)U , Ui = RehF, Ui ≤ kF kkUk. (26)
From equations (23) and (26), we have
θ
λ2 kA 2 vk2 ≤ C kF kkUk + kF k2 .

(27)
As θ−2
2
≤ 0 ≤ θ2 , taking into account the continuous embeding D(Aθ2 ) ֒→ D(Aθ1 ),
θ2 > θ1 and (26), we obtain
θ−2
kA 2 zk2 ≤ C kF kkUk + kF k2 .

(28)
2 2

kzk ≤ C kF kkUk + kF k . (29)

3.1 Exponential Decay for 0 ≤ θ ≤ 1

In this subsection we show the exponential decay using Theorem (21), let us first
check condition (20).
Now, notice that:
hA2 v, Aσ wi = hA2 v, Aσ (iλu − f1 )i = −iλhAσ v, A2 ui − hA1+σ v, Af1 i
hA2 u, Aσ zi = hA2 u, Aσ (iλv − f2 )i = −iλhA2 u, Aσ vi − hA1+σ u, Af2 i.
Summing up, both equations and taking the real part, we have
Re{hA2 v, Aσ wi + hA2 u, Aσ zi} = −Re{hA1+σ u, Af2 i + hA1+σ v, Af1 i} (30)

We are going to begin the study of the asymptotic behavior of the semigroup etB
with the followings lemmas.

8
Lemma 5 Let 0 ≤ θ ≤ 1 and σ ≤ −1. The solutions of equations (22)-(25) satisfy
the following equality
γα σ+2 2 σ+1
kA 2 wk = γkA 2 zk2 − αRe{hA1+σ u, Af2 i + hA1+σ v, Af1 i}
β
δα 2θ+σ σ+2 α
+ RehA 2 z, A 2 wi − Rehf4 , Aσ+1 wi − Rehf3 , Aσ zi
β β
λα σ σ+2 σ+2 σ−2
+ ImhA 2 z, A 2 wi + λImhA 2 w, A 2 zi.
β

Proof: Applying the product duality to equation (24) with Aσ z and recalling that
the operator A is self-adjoint, we have
σ+1
γkA 2 zk2 = −αhA2 u, Aσ zi − iλhw, Aσ zi + hf3 , Aσ zi.
α σ+1
Similarly, applying the product duality to equation (25) with A w and using
β
the equation (22) we obtain

γα σ+2 2 iλα σ σ+2 δα 2θ+σ σ+2

kA 2 wk = αhA2 v, Aσ wi + hA 2 z, A 2 wi + hA 2 z, A 2 wi
β β β
α
− hf4 , Aσ+1 wi.
β
Now, to get the conclusion of this Lemma it is sufficient to perform the subtraction
of these last two equations,take the real part and using the identity (30). ✷

Taking σ = −2, in Lemma(5), we have

γα
kwk2 = γkA 2 zk2 − αRe{hA−1 u, Af2 i + hA−1 v, Af1 i}
−1

β
δα α
+ RehAθ−1 z, wi − Rehf4 , A−1 wi − Rehf3 , A−2 zi (31)
β β
α
+ Imhz, A λwi + λImhw, A−2zi,
−1
β

From equation (24), we have A−1 λw = iαAu + iγz − iA−1 f3 , therefore

α α
Imhz, A−1 λwi = Imhz, iαAu + iγz − iA−1 f3 i
β β
α2 θ 2−θ αγ α
= − RehA 2 z, A 2 ui − kzk2 + Rehz, A−1 f3 i
β β β
2
α θ 2−θ α
≤ − RehA 2 z, A 2 ui + Rehz, A−1 f3 i. (32)
β β

Substituting (32) into (31), we have

9
 γα
kwk2 ≤ γkA 2 zk2 − αRe{hA−1 u, Af2i + hA−1 v, Af1 i}
−1

β
δα α
+ RehAθ−1 z, wi − Rehf4 , A−1 wi − Rehf3 , A−2 zi (33)
β β
2
α θ 2−θ α
− RehA 2 z, A 2 ui + Rehz, A−1 f3 i + ImhA−2 λw, zi.
β β

On the other hand of the equation (24), we have A−2 λw = iαu+iγA−1z−iA−2 f3 ,

therefore

ImhA−2 λw, zi = Imhiαu + iγA−1 z − iA−2 f3 , zi

θ
zk2 + RehA−2 f3 , zi
−θ −1
= −αRehA 2 u, A 2 zi − γkA 2

θ
u, A 2 zi + RehA−2 f3 , zi.
−θ
≤ −αRehA 2 (34)

Now, substituting (34) into (33), we have

γα
kwk2 ≤ γkA 2 zk2 − αRe{hA−1 u, Af2 i + hA−1 v, Af1 i}
−1

β
δα α
+ RehAθ−1 z, wi − Rehf4 , A−1 wi − Rehf3 , A−2 zi (35)
β β
α2 θ 2−θ α θ
− RehA 2 z, A 2 ui + Rehz, A−1 f3 i − αRehA 2 u, A 2 zi
−θ

β β
−2
+RehA f3 , zi.

Applying Cauchy-Schwarz and Young inequalities, taking into account the con-
θ
tinuous embedding D(Aθ2 ) ֒→ D(Aθ1 ), θ2 > θ1 , −1 2
< θ2 , θ − 1 ≤ and using
2
estimative (26) we have, for ε > 0, there existe kε > 0, such that

2−θ
kwk2 ≤ C{kF kkUk} + εkwk2 + εkA uk2 + εkA uk2 .
−θ
2 2 (36)

On the other hand, by effecting the product duality of (24) by A−θ u, we have
2−θ 2−θ
uk2 = hw, A−θ (iλu)i − γhA ui + hf3 , A−θ ui
−θ
αkA 2 2 z, A 2

2−θ
wk2 + hw, A−θ f1 i − γhA ui + hf3 , A−θ ui
−θ −θ
= kA 2 2 z, A 2

Taking real part and applying Cauchy-Schwarz and Young inequalities, taking
into account the continuous embedding, − 2θ ≤ 2θ , we have
2−θ
uk2 ≤ C{kF kkUk} + kA wk2.
−θ
kA 2 2 (37)

Substituting (37) into (36) and , taking into account the continuous embedding,
− θ2 ≤ 2−θ
2
, we have

kwk2 ≤ C{kF kkUk} for 0 ≤ θ ≤ 1. (38)

10
 Taking the duality product between equation (24) and u and using the equation
(22), we obtain
αkAuk2 = −γhz, Aui + kwk2 + hw, f1i + hf3 , ui. (39)

Applying Cauchy-Schwarz and Young inequalities, taking into account the contin-
uous embedding D(Aθ2 ) ֒→ D(Aθ1 ), θ2 > θ1 , −1 2
< θ2 , 0 ≤ θ2 and using estimatives
(26) and (38) we have, for ε > 0, there existe kε > 0, such that

αkAuk2 ≤ C{kF kkUk} for 0 ≤ θ ≤ 1. (40)

Similarly, applying the duality product to equation (25) with v and using the
equation (23), we have
1 θ θ
βkA 2 vk2 = γhAw, vi + kzk2 − δhA 2 z, A 2 vi + hz, f2 i + hf4 , vi. (41)

Subtracting (41) from (39) and taking the real part, we have

1
βkA 2 vk2 = αkAuk2 + γRe{hiλAv − Af2 , ui + hiλAu − Af1 , vi} − kwk2
θ θ
−δRehA 2 z, A 2 vi + Rehz, f2 i + Rehf4 , vi − Rehw, f1i − Rehf3 , ui
≤ αkAuk2 + γλIm{hAv, ui + hu, Avi} − γRe{hf2 , Aui + hAf1 , vi}
θ θ
−δRehA 2 z, A 2 vi + Rehz, f2 i + Rehf4 , vi − Rehw, f1i − Rehf3 , ui

Now, as Im{hAv, ui+hu, Avi} = 0 and 2θ ≤ 12 , using the estimative (40) and applying
Cauchy-Schwarz inequality and Young inequality and continuous embedding we have
the inequality
1
βkA 2 vk2 ≤ C {kF kkUk} for 0 ≤ θ ≤ 1. (42)

Therefore, estimates (29), (38), (40) and (42), condition (20) the Theorem(3) is
verified for 0 ≤ θ ≤ 1.
Now let’s show condition (19). It’is prove that iR ⊂ ρ(B) by contradiction, then
we suppose that iR 6⊂ ρ(B). As 0 ∈ ρ(B), we consider the highest positive number
λ0 such that the ] − iλ0 , iλ0 [⊂ ρ(B) then iλ0 or −iλ0 is an element of the spectrum
σ(B). We Suppose iλ0 ∈ σ(B) (if −iλ0 ∈ σ(B) the proceeding is similar). Then, for
0 < δ < λ0 there exist a sequence of real numbers (λn ), with δ ≤ λn < λ0 , λn → λ0 ,
and a vector sequence Un = (un , vn , wn , zn ) ∈ D(B) with unitary norms, such that

k(iλn − B)Un k = kFn k → 0,

as n → ∞. From (40) and (42) for 0 ≤ θ ≤ 1, we have
αkAun k2 ≤ C kFn kkUn k + kFn k2 ,


βkA1/2 vn k2 ≤ C kFn kkUn k + kFn k2 .



In addition to the estimates and (29) and (38) for 0 ≤ θ ≤ 1, we have

kwn k2 + kzn k2 → 0.

11
 Consequently,

αkAun k2 + βkA1/2 vn k2 + kwn k2 + kzn k2 → 0.

Therefore, we have kUn k → 0 but this is absurd, since kUn k = 1 for all n ∈ N. Thus,
iR ⊂ ρ(B).
This completes the proof of condition (19) of the Theorem(3).

4 S(t) = eBt is not analytic for θ ∈ [0, 1]

In this section we will show that S(t) is not analytic for θ ∈ [0, 1]. The proof will
be divided into two subsections: In the first subsection (4.1) we show the lack of
analyticity for 0 ≤ θ < 1 and in subsection (4.2) we test the lack of analyticity for
θ = 1.
The proof of both subsections is made showing that the condition (21) of Theo-
rem(4) do not check.
The spectrum of operator A = −∆ defined in (8) is constituted by positive
eigenvalues (σn ) such that σn → ∞ as n → ∞. For n ∈ N we denote with en an
unitary L2 -norm eigenvector associated to the eigenvalue σn , that is
(
ken kL2 (Ω) = 1, for 0 ≤ θ < 1, n ∈ N,
Aen = σn en , Aθ en = σnθ en , 1 (43)
kA 2 en kL2 (Ω) = 1, for θ = 1, n ∈ N.

4.1 Lack of analiticity for θ ∈ [0, 1)

Theorem 6 The semigroup associated to system (9)-(12), is not analytical for θ ∈
[0, 1).

Proof: We will show the condition (21) of the Theorem(4). Consider the eigenval-
ues and eigenvectors of the operator A as in (8) and (43) respectively.
Let Fn = (0, 0, −en , 0) ∈ H. The solution U = (u, v, w, z) of the system (iλI −
B)U = Fn satisfies w = iλu, z = iλv and the following equations

λ2 u − αA2 u − iλγAv = en ,
λ2 v − βAv + iγAu − iλδAθ v = 0.

Let us see whether this system admits solutions of the form

u = µen , v = νen ,

for some complex numbers µ and ν. Then, the numbers µ, ν should satisfy the
algebraic system
 2
λ − ασn2 µ − iλγσn ν = 1,

(44)
 2 θ

iλγσn µ + λ − βσn − iδσn λ ν = 0. (45)

12
 On the other hand solving the system (44)-(45), we find that

p2,n (λ2 ) − iδσnθ λ


µ= , (46)
p1,n (λ2 )p2,n (λ2 ) − γ 2 λ2 σn2 − iδσnθ λp1,n (λ2 )

where

p1,n (λ2 ) := λ2 − ασn2 and p2,n (λ2 ) = λ2n − βσn . (47)

Taking sn = λ2 = λ2n and considering the polynomial

qn (sn ) := p1,n (sn )p2,n (s) − γ 2 σn2 sn

= s2n − [(α + γ 2 )σn2 + βσn ]sn + αβσn3 .

Now, taking qn (sn ) = 0, we have the roots of the polynomial qn are given by
p
± [(α + γ 2 )σn2 + βσn ] ± σn (α + γ 2 )2 σn2 + 2β(γ 2 − α)σn + β 2
sn = . (48)
2

|xn |
Thus, if we introduce the notation xn ≈ yn meaning that lim is a positive
n→∞ |yn |
real number.
Taking sn = s+
n from equation (48), we have

sn ≈ σn2 and λn ≈ σn . (49)

Then
p2,n (sn ) = sn − βσn ≈ σn2 . (50)
From qn (sn ) = 0 in (46), we have

p2,n (λ2n ) − iδλn σnθ p22,n (λ2n )


p2,n (λ2n )
µn = = 2 2 2 + i 2 3 2+θ . (51)
−iδσnθ λn p1,n (λ2 ) γ λn σn δγ λn σn

Therefore
|µn | ≈ |λ|−1−θ (52)

4.2 Lack of Analiticity for θ = 1

1
Let Fn = (0, −A 2 en , 0, 0) ∈ H. The solution U = (u, v, w, z) of the system (iλI −
1
B)U = Fn satisfies w = iλu, z = iλv + A 2 en and the following equations
3
λ2 u − αA2 u − iγλAv = γA 2 en ,
1 3
λ2 v − βAv + iγλAu − iδλAv = iλA 2 en + δA 2 en .

Let us see whether this system admits solutions of the form

u = µen , v = νen ,

13
for some complex numbers µ and ν. Then, the numbers µ, ν should satisfy the
algebraic system
 2 3
λ − ασn2 µ − iγλσn ν = γσn2 ,

(53)
1 3
iγλσn µ + λ2 − βσn − iδλσn ν = iλσn2 + δσn2 .

(54)

On the other hand solving the system (53)-(54), we find that

3 1 5
δσn2 p1,n (λ2 ) + i[λσn2 p1,n (λ2 ) − γ 2 λσn2 ]
ν= , (55)
p1,n (λ2 )p2,n (λ2 ) − γ 2 λ2 σn2 − iδλσn p1,n (λ2 )
where

p1,n (λ2 ) := λ2 − ασn2 and p2,n (λ2 ) = λ2n − βσn . (56)

Taking sn = λ2 = λ2n , denote ν = νn and considering the polynomial

qn (sn ) := p1,n (sn )p2,n (s) − γ 2 σn2 sn

= s2n − [(α + γ 2 )σn2 + βσn ]sn + αβσn3 .

Now, taking qn (sn ) = 0, we have the roots of the polynomial qn are given by
p
± [(α + γ 2 )σn2 + βσn ] ± σn (α + γ 2 )2 σn2 + 2β(γ 2 − α)σn + β 2
sn = . (57)
2
Taking sn = s−
n from equation (57), we have

p
[(α + γ 2 )σn2 + βσn ]2 − [ (α + γ 2 )2 σn4 + 2β(γ 2 − α)σn3 + β 2 σn2 ]2
sn = p
2[(α + γ 2 )σn2 + βσn + (α + γ 2 )2 σn4 + 2β(γ 2 − α)σn3 + β 2 σn2 ]
4αβσn3
= p (58)
2[(α + γ 2 )σn2 + βσn + (α + γ 2 )2 σn4 + 2β(γ 2 − α)σn3 + β 2 σn2 ]

From (58), to obtain

1
sn ≈ σn then λn ≈ σn2 and |p1,n (sn )| ≈ σn2 . (59)

From qn (sn ) = 0 in (55), we have

 2 5 1 3
 [γ λn σn2 − λn σn2 p1,n (λ2n )] + iδσn2 p1,n (λ2n ) 

|νn | =  
δλn σn p1,n (λ2n ) 
 2 1
 (γ + α)σn2 − λ2n

σn2 
=  1 +i  (60)
δσn2 p1,n (λ2n ) λn

Therefore, using estimatives (59) in (60), we obtain

|νn | ≈ σn0 ≈ λ0n . (61)

14
 Finally, for C > 0, the solution Un of the system (iλn − B)U = Fn , satisfies
(
kwn k = |λn |kun k = |λn ||µn |ken k = |λn ||µn | for 0 ≤ θ < 1
kUn k ≥ 1 1 (62)
CkA 2 vk = C|νn |kA 2 en k = C|νn | for θ = 1.

Then, using estimatives (52) and (61) in (62), for κ > 0 we have
(
|λ|−θ > κ|λ|−1 for, 0 ≤ θ < 1.
kUk ≥ κ (63)
|λ|0 > κ|λ|−1 for, θ = 1.

From (63), of the condition (21) the Theorem(4) does not check. This completes
the proof of the lack of analyticity for 0 ≤ θ ≤ 1. ✷

Acknowledgments
This research was partly performed when one authors were visiting the Institute of
Pure and Applied Mathematics (IMPA), the warm hospitality was greatly appreci-
ated. Gratefully acknowledge researcher professor, Felipe Linares for having secured
visiting.

References
[1] F. Alabau, Stabilisation frontière indirecte de systèmes faiblement couplés. C.
R. Acad. Sci. Paris Sér. I Math. 328 (1999), 1015-1020.

[2] F. Alabau, P. Cannarsa and V. Komornik, Indirect internal stabilization of

weakly coupled systems. J. Evolution Equations 2 (2002), 127-150.

[3] F. Alabau-Boussouira, Indirect boundary stabilization of weakly coupled hy-

perbolic systems. SIAM J. Control Optim. 41 (2002), 511-541.

[4] F. Alabau, P. Cannarsa, and R. Guglielmi, Indirect stabilization of weakly

coupled system with hybrid boundary conditions. Mathematical Control and
Related Fields 4 (2011), 413-436.

[5] F. Alabau-Boussouira and M. Léautaud, Indirect stabilization of locally coupled

wave-type systems. ESAIM Control Optim. Calc. Var. 18 (2012), 548-582.

[6] A. Bátkai, K-J. Engel, J. Prüs and R. Schnaubelt, Polynomial stability of op-
erator semigroups. Math. Nachr. 279, No. 13–14, (2006), 1425 – 1440.

[7] K.J. Engel and R. Nagel, One-parameter semigroups for linear evolution equa-
tions, Springer 2000.

[8] Gearhart. Spectral theory for contraction semigroups on Hilbert spaces. Trans.
Amer. Math. Soc. 236 (1978), 385-394.

15
 [9] R. Guglielmi, Indirect stabilization of hyperbolic systems through resolvent
estimates. Evol. Equ. Control Theory 6 (2015), 59-75.

[10] F. Dell’Oro, J.E.M. Rivera and V. Pata, Stability properties of an abstract

system with applications to linear thermoelastic plates. J. Evol. Equ. 13 (2013),
777-794.

[11] Z. Liu and S. Zheng, Semigroups associated with dissipative systems, Chapman
& Hall CRC Research Notes in Mathematics, Boca Raton, FL, 398 (1999).

[12] Z. J. Han and Z. Liu, Regularity and stability of coupled plate equations with
indirect structural or Kelvin-Voigt damping, ESAIM: Control, Optimisation
and Calculus of Variation (2018).

[13] J. Hao, Z. Liu and J. Yong, Regularity analysis for an abstract system of cou-
pled hyperbolic and parabolic equations. Journal of Differential Equations 259
(2015), 4763-4798.

[14] H.P. Oquendo and F.M.S. Suárez, Exact decay rates for coupled plates with
partial fractional damping. Zeitschrift für angewandte Mathematik und Physik-
ZAMP(online)V1, (2019), pp 70-88.

[15] H.P. Oquendo and R.P. Raya, Best rates of decay for coupled waves with dif-
ferent propagation speeds. ZAMP Z. Angew. Math. Phys. 68 (2017), Art. 77.

[16] H.P. Oquendo, Frictional versus Kelvin-Voigt damping in a transmission prob-

lem. Math. Meth. Appl. Sci. 40 (2017), 7026-7032.

[17] A. Pazy, Semigroups of Linear Operators and Applications to Partial Differen-

tial Equations, Applied Mathematical Sciences 44, Springer, (1983).

[18] S. Mansouri, Boundary Stabilization of Coupled Plate Equations. Palestine

Journal of Mathematics 2 (2013), 233-242.

[19] F.M.S. Suárez and H.P. Oquendo, Optimal decay rates for partially dissipa-
tive plates with rotational inertia. Acta Applicandae Mathematicae(online) V1,
(2019), pp 1–16.

[20] S. Vidoli and F. dell’Isola, Vibrations Control in Plates Uniformly Distributed

PZT Actuators Interconnected Via Electric Networks. Eur. J. Mech. A/Solids
V20, (2001), 435-456.

[21] S. Alessandroni, F. dell’Isola and M. Porfiri, A Revival of Electric Analogs

for Vibrating Mechanical System Aimed to Their Efficient Control by PZT
Actuators. International Journal of Solids and Structures 39, (2002), 5295–5324.

[22] C. Maurini, F. dell’Isola and D. Del Vescovo, Comparison of Piezoelectronic

networks acting as distributed vibration absorbers. Mechanical Systems and
Signal Processing 18, (2004), 1243–1271.

16
[23] M. Renardy, On localized Kelvin-Voigt damping. ZAMM Z. Angew. Math.
Mech. 84 (2004), 280-283.

[24] D.L. Russell, A general framework for the study of indirect damping mecha-
nisms in elastic systems. J. Math. Anal. Appl. 173 (1993), 339-358.

[25] Y. Shibata, On the exponential decay of the energy of a linear thermoelastic

plate. Comp. Appl. Math. 13 (1994), 81-102.

[26] L. Tebou, Stabilization of some coupled hyperbolic/parabolic equations. Dis-

crete Contin. Dyn. Syst. Ser. B 14 (2010), 1601-1620.

[27] L. T. Tebou, Energy decay estimates for some weakly coupled Euler-Bernoulli
and wave equations with indirect damping mechanisms. Mathematical Control
and Related Fields 2 (2012), 45-60.

[28] L. Tebou, Simultaneous stabilization of a system of interacting plate and mem-

brane. Evol. Equ. Control Theory 2 (2013), 153-172.

[29] L. Tebou, Indirect stabilization of a Mindlin-Timoshenko plate. J. Math. Anal.

Appl. 449 (2017), 1880-1891.

17