Professional Documents
Culture Documents
1. Introduction
A basic linear model, developed in [19], for describing the transverse vibration
of beams is given by two coupled partial differential equations
wtt = (K(wx ))x ,
on (0, l) R+
(1.1)
I tt (x, t) = (EIx )x + K(wx ).
Here, t is the time variable and x the space coordinate along the beam, the length of
which is l, in its equilibrium position. The function w is the transverse displacement
of the beam and is the rotation angle of a filament of the beam. The coefficients
, I , E, I and K are the mass per unit length, the polar moment of inertia of a
cross section, Youngs modulus of elasticity, the moment of inertia of a cross section
and the shear modulus respectively. The natural energy of the beam is
Z
1 l
2
2
2
2
|t w| + I |t | + EI |x | + K |x w | )dx.
(1.2)
E(t) =
2 0
The aim of this paper is to study two types of stabilization of this system, the
internal and the boundary stabilization. Let us mention some results about the
stabilizability of the Timoshenko beam. The case of two boundary control force
has already been considered by Kim and Renardy [4] for the Timoshenko beam.
They proved the exponential decay of the energy E(t) by using a multiplier technique and provided numerical estimates of the eigenvalues of the operator associated
to this system, and by Lagnese & Lions [11] for the study of the exact controllability, Taylor [7] studied the boundary control of system (1.1) with variable physical
characteristics. Recently Shi & Feng [16] established the exponential decay of the
energy E(t) with locally distributed feedback (two feedback).We will first prove that
it is possible to stabilize uniformly (with respect to the initial data) this beam,
2000 Mathematics Subject Classification. 35L55, 93C05, 93C20, 93D15, 93D20, 93D30.
Key words and phrases. Timohsenko beam, strong stability, non-uniform stability,
uniform stability, Lyapunov function.
c
2003
Southwest Texas State University.
Submitted September 2, 2002. Published March 16, 2003.
1
EJDE2003/29
which is assumed to be clamped at the two ends of (0, l), by using a unique locally
distributed feedback
wtt = (K(wx ))x ,
on (0, l) R+
(1.3)
(0, t) = (l, t) = 0.
EI
I .
uk
vk
2
where Yk =
wk H, k = 1, 2. In the sequel we will denote by kY k := (Y, Y ),
fk
the norm in the energy space. The system 1.3 can be written as
t Y (t) = LY (t),
where
w(t)
t w(t)
Y (t) :=
(t) ,
t (t)
xx
L :=
K
I x
I
0
0
0
0
K
x
0
EI
I xx
K
I
0
0
I
b(x)
with D(L) := ((H 2 (]0, l[) H01 (]0, l[)) H01 (]0, l[))2 .
Proposition 2.1 ([17],[16]). The operator (L, D(L)) generates a C0 -semigroup of
contractions (eLt )t0 on H.
EJDE2003/29
e t
re (e ) = e
(2.2)
+
t R .
(2.3)
is often called the type and e the essential type of the semigroup. A third real
number which plays an essential role in stability theory is the spectral abscissa s(L)
of L defined by
s(L) = sup {Re : (L)} .
These three real numbers are related as follows [5, Theorem 3.6.1., p. 107]:
(L) = max (e (L), s(L)) .
Clearly, the uniform stability of eLt is equivalent to (L) < 0. We are now ready
to state our first result.
Theorem 2.2. Assume that b C([0, l]) and
b(x) b > 0
(2.4)
Then:
eLt is asymptotically stable.
EI
Lt
If K
is non uniformly stable.
6= I , then e
Before giving the proof of this theorem we need to recall the following result.
Theorem 2.3 (Benchimol [2]). Let L be a maximal linear operator in a complex
Hilbert space H and assume that:
(a) L has a compact resolvent.
(b) L does not have purely imaginary eingenvalues.
Then eLt is strongly stable.
Theorem 2.2. To prove strong stability of eLt , it remains to verify the properties
(a) and (b) of theorem 2.3.
Property (a) follows at once from the compactness of the imbedding D(L) in H,
a consequence of Rellichs theorem.
EJDE2003/29
Suppose the conclusion of (b) is false, then L have a purely imaginary eingenvalues i. Let U1 the eigenvectors associated to i. Using the definition of L it
follows that LU1 = iU1 if and only if
Kxx u Kx v = 2 u,
EI.xx v + Kx u Kv ib(x)v = I 2 v,
u = v = 0.
0,l
(2.5)
0,l
(2.6)
v = u = 0 on ]b0 , b1 [,
u = v = 0.
0,l
0,l
0 0
0
0
0 0
0
0
D=
0 0
0
0
K
0 0 I 0
EI
I
is based on the
This operator is compact in the energy space H. Then re (L) = re (L D). Now we
calculate the essential spectral radius of L1 := L D , for this reason we transform
the equations by introducing the following variables:
s
s
K
K
p=
x u + t u, q =
x u + t u,
s
s
EI
EI
=
x v + t v, =
x v + t v.
I
I
which are solutions of the system
p
p
p
t
q + Kx q + C q = 0,
(2.7)
EJDE2003/29
p0
K=
0
0
K/
0
0
0
q
I
K
0
2
0
EI
b(x)
K
2I
2IK
q
C = 2I
I
0
K
2
EI
b(x)
K
2IK
2I
2I
0
,
p0
EI/I
q
I
EI
b(x)
2I
q
I
K
2
EI
b(x)
2I
K
2
EI
L1 t
When K
)=
6= I , we can apply a result in [6, p 324] which implies that re (e
L2 t
re (e ) = exp(t) where
0
0
0
0
K/ p 0
0
0
0 b(x) 0
0
0
EI/I
0
2I
x +
p0
L2 =
0
0
0
0
0
0
0
K/
p
b(x)
0
0
0 2I
0
0
0
EI/I
s
s
b(x)
b(x)
EI
EI
x = ( +
),
x = ( +
),
I
2I
I
2I
with the boundary conditions
(p + q)(0, t) = ( + )(0, t) = (p + q)(l, t) = ( + )(l, t) = 0.
Then we have
r
p(x) = c1 exp(
x),
K
r
r
q(x) = c1 exp(
x),
K
Z x
I
b(y)
(x +
dy)),
(x) = c2 exp(
EI
0 2I
r
Z x
I
b(y)
(x) = c2 exp(
(x +
dy)),
EI
0 2I
where c1 and c2 are in R\{0}. Using the boundary conditions, we see that one of
the following two equations must hold for non-trivial solutions,
r
r
exp(
l) exp(
l) = 0,
K
K
r
r
Z l
Z l
I
b(y)
I
b(y)
exp(
(l +
dy)) exp(
(l +
dy) = 0,
EI
2I
EI
0
0 2I
EJDE2003/29
(
0
b(x)
)dx,
2I
and thus
= sup{Re(), (L2 )} = 0.
Then re (eL2 t ) = 1. Using the property r(eLt ) re (eLt ) we deduce, the non uniform
EI
stability of the semigroup associated to the Timoshenko beam if K
6= I .
3. Uniform stability
In this section our main result is the following theorem.
Theorem 3.1. Assuming that b(x) satisfies (2.4), eLt is exponentially stable if
K
EI
= I .
The proof of the uniform stability is based on the construction of a Lyapunov
function (Y (t)) satisfying the following inequalities:
i) There exist two positives constants, d0 , d1 such that
2
d0 kY k (Y ) d1 kY k
Y D(L)
(3.1)
t (Y (t)) d2 kY (t)k
(3.2)
Proof. It follows from (3.1) that kY (t)k d11 (Y (t)). Using (3.2) then we
have
d2
t (Y (t)) (Y (t))
d1
thus
d2
(Y (t)) exp( t)(Y ) .
d1
Using (3.1), we have kY (t)k2 dd10 exp( dd21 t)kY k2
Lemma 3.3. For any positive constant 1 and scalar functions h, c such that:
h(0) > 0, h(l) < 0, hx < 0 on [b0 , b1 ], hx > 0 on [0, l]\[b0 , b1 ], and cxx < 0 on [0, l],
we have the following statements:
i)
Z l
Z
Z
EI l
I l
EI
t
I t .h(x)x dx =
hx (x )2 dx
hx (t )2 dx +
[h.(x )2 ]l0
2
2 0
2
0
0
Z l
Z l
(b(x)t ).h(x)x dx + K
(wx ).hx dx,
0
ii)
Z
Z
I t .c(x)dx = EI
t
0
EI
c(x ) dx +
2
2
l
2
cxx () dx + I
0
c(t )2 dx.
EJDE2003/29
(wx ).cdx
+K
(b(x)t ).c(x)dx.
iii)
l
t
0
=K
1 h(x)x + 1 c(x)dx.
dx +
0
+
0
=
0
l
I t .h(x)t x dx
0
(
0
EI
I
(x )2 hx (t )2 hx )dx
2
2
+
0
EI
[h.(x )2 ]l0
2
I tt .c(x)dx +
I t .c(x)t dx
0
l
Z
(EIxx + K(wx ) b(x)t ).c(x)dx +
=
0
I t .c(x)t dx
0
Z
= EI
c(x )2 dx +
EI
2
cxx ()2 dx + K
(wx ).cdx
0
Z
(b(x)t ).c(x)dx + I
c(t )2 dx
Z
Kx (wx ).f dx +
=
0
t w.(t f )dx
0
Z
=
Z
K(wx ).fx dx +
t w.(t f )dx
0
EJDE2003/29
=
0
t w.(t f )dx
0
1
l
Rl
0
v dx. Then
0
l
=
0
1
l
Z
K(wx )dx +
1 h(x)x + 1 c(x)dx.
0
t w.(t f )dx
0
Z
K l
=
K(wx ).(1 h(x)x + 1 c(x))dx +
1 h(x)x
l 0
0
Z l
Z l
+ 1 c(x)dx.
dx
t w.((xx )1 x (1 h(x)t x + 1 c(x)t ))dx.
Z
Now we introduce the function
Z l
Z l
Z l
1 (Y (t)) := E(t) + 1
I t .h(x)x dx + 1
I t .c(x)dx
(t w.f )dx
0
we have
|t 1 (Y (t))|
Z l
I
a1 b2
a2 b2
1 1 c2
(b(x) 1 hx + 1 I c + 1
+ 1
+
+
)(t )2 dx
2
2
2
2a
2a
3
4
0
Z l
h2
EI
EI
+
(1
hx 1 EIc + 1
)(x )2 dx + 1
[h.(x )2 ]l0
2
2a
2
1
0
Z l
EI
c2
+
(1
cxx + 1
)()2 dx
2
2a2
0
Z l
Z l
+ K
(1 h(x)x + 1 c(x))dx.
dx
0
+(
1 a3 1 a4
+
)
2
2
(t w)2 dx.
EI
[h.(x )2 ]l0 )
2
(t w.f )dx
0
Z l
EI
I
b(x)(t )2 dx + 1 ( (
(x )2 hx (t )2 hx )dx
2
2
0
(K(wx ) b(x)t ).h(x)x dx +
I t .c(x)dx t
I t .h(x)x dx + 1 t
0
EJDE2003/29
Z
Z l
EI l
cxx ()2 dx + K
(wx ).cdx
2 0
0
0
Z l
Z l
(b(x)t ).c(x)dx + I
c(t )2 dx)
Z
+ 1 (EI
c(x )2 dx +
0
l
K
l
Z l
Z
(1 h(x)x + 1 c(x))dx.
dx
0
0
l
Z l
EI
I
2
(1
hx 1 EIc)(x )2 dx
= (b(x) 1 hx + 1 I c)(t ) dx +
2
2
0
0
Z l
Z l
Z l
1
(b(x)t ).h(x)x dx K
(1 h(x)x + 1 c(x))dx.
dx
Z
+ 1
Z
EI
EI
[h.(x )2 ]l0 + 1
2
2
cxx ()2 dx 1
(b(x)t ).c(x)dx
0
Z l
I
EI
2
= (b(x) 1 hx + 2 I c)(t ) dx +
(1
hx 1 EIc)(x )2 dx
2
2
0
0
Z l
Z
Z l
K l
1
(b(x)t ).h(x)x dx
(1 h(x)x + 1 c(x))dx.
dx
l 0
0
0
Z
Z l
EI
EI l
+ 1
[h.(x )2 ]l0 + 1
cxx ()2 dx 1
(b(x)t ).c(x)dx
2
2 0
0
Z l
Z l
1
1 l
1
3
1
2
1 (hx x (xx ) h)t w.t dx
(t w) dx +
(t )2 dx ,
2
2a3 0
0
0
Z
Z
Z
l
1 a4 l
1 l
1
2
1 x (xx ) t w.c(x)t dx
(t w) dx +
(c(x)t )2 dx.
2
2a4 0
0
0
Then
|t 1 (Y (t))|
Z l
I
a1 b2
a2 b2
1 1 c2
(b(x) 1 hx + 1 I c + 1
+ 1
+
+
)(t )2 dx
2
2
2
2a3
2a4
0
10
EJDE2003/29
EI
h2
hx 1 EIc + 1
)(x )2 dx
2
2a
1
0
Z l
EI
EI
c2
+ 1
[h.(x )2 ]l0 +
(1
cxx + 1
)()2 dx
2
2
2a2
0
Z
Z l
Z l
K l
1 a3 1 a4
+
(1 h(x)x + 1 c(x))dx.
dx + (
+
) (t w)2 dx.
l 0
2
2
0
0
Z
(1
Lemma 3.5. For scalar positive constants ai (i = 5, 6), and scalar functions k, d
such that k(0) > 0, k(l) < 0, kx < 0 on [b0 , b1 ], kx > 0 on [0, l]\[b0 , b1 ], and d > 0
on [0, l], we have: i)
Z l
Z l
Z l
K
a5
1
t
t w.k(x)wx dx
( kx + K )(wx )2 dx + K
(k(x)x )2 dx
2
2
2a
5 0
0
0
Z l
Z
t
t w.d(x)wdx =
0
and iii)
t
Z
Z
(t .( wx ) t w.x )dx
0
l
K
a6
( +
)( wx )2 dx +
I
2I
(1 +
0
K
I b2
)(t )2 dx + [x .wx ]l0 .
2a6
Z
=
Z
tt w.k(x)wx dx +
t w.k(x)t wx dx
0
l
Z
Kx (wx ).k(x)wx dx +
=
0
t w.k(x)t wx dx
0
Z
=
0
K
kx (wx )2 dx
2
Z
Kx .k(x)wx dx
K
kx (t w)2 dx + [k.(wx )2 ]l0
2
2
Z
I t .h(x)x dx
t
0
(
0
K
a5
1
kx + K )(wx )2 dx + K
2
2
2a5
K
kx (t w)2 dx + [k.(wx )2 ]l0 .
2
2
Z
0
(k(x)x )2 dx
EJDE2003/29
11
0
l
Kx (wx ).d(x)wdx
0
0
Z l
=
0
d(x)(t w)2 dx
d(x)(t w)2 dx
iii) Multiplying the first equation in (1.3) by x and the second equation of system
(1.3) by wx , we obtain
Z l
t
(t .( wx ) t w.x )dx
0
l
Z
tt .( wx )dx +
t .(t t wx )dx
0
l
Z
tt w.x dx
0
l
t wt x dx
0
EI
xx .( wx )dx
I
=
0
K
( wx )2 dx
I
t .(t t wx )dx
+
0
0
K
b(x)
t .( wx )dx
I
0
Z l
K
x (wx ).x dx
t wt x dx.
EI
I ,
=
0
K
( wx )2 dx
I
Z
0
b(x)
t .( wx )dx +
I
(t )2 dx
K
[x .wx ]l0 .
0
0
0
Therefore,
Z l
t
(t .( wx ) t w.x )dx
Z
0
l
K
a6
( +
)( wx )2 dx +
I
2I
(1 +
0
K
I b2
)(t )2 dx + [x .wx ]l0 .
2a6
t w.(k(x)wx d(x)w)dx.
+ 3
0
(3.3)
12
EJDE2003/29
(t .( wx ) t w.x )dx
t (Y (t)) =t 1 (Y (t)) + 2 t
0
t w.(k(x)wx d(x)w)dx.
+ 3 t
0
(b(x) 1 hx + 1 I c(x) + 1
+ 1
+
2
2
2
2a3
0
2
2
I b
1 c
+
+ 2 (1 +
))(t )2 dx
2a4
2a6
Z l
(h(x))2
3
EI
hx 1 EIc + 1
+K
(k(x))2 )(x )2 dx
+
(1
2
2a
2a
1
5
0
Z l
Z l
EI
(c(x))2
K
a6
+
(1
cxx + 1
)()2 dx + 2
( +
)( wx )2 dx
2
2a2
I
2I
0
0
Z
Z
Z l
l
K l
+ 3
K(wx ).x (d(x)w)dx +
(1 h(x)x + 1 c(x))dx.
dx
l 0
0
0
K
EI
K
3
l
2 l
2 l
+ 2 | [x .wx ]0 | + 1
[h.(x ) ]0 +
[k.(wx ) ]0
2
2
Z l
K
a5
+ 3
( kx + K )(wx )2 dx
2
2
0
Z l
3
1 a3 1 a4 2
+
(
kx 3 d(x) + (
+
) )(t w)2 dx .
2
2
2
0
Using Cauchy Shwartzs inequality and Youngs inequality, we obtain
Z
Z
Z l
l
Ka7 l
K
2
K(wx ).(d(x)wx )dx
(wx ) dx +
(d(x)wx )2 dx
2
2a7 0
0
0
where a7 > 0,
Z
Z
Z
2
l
Ka7 l
Kd c0 l
2
K(wx ).dx wdx
(wx ) dx +
(wx )2 dx
2
2a7
0
0
0
where d = max(dx ),
Z
Z l
K l
(1 h(x)x + 1 c(x))dx.
dx
l 0
0
EJDE2003/29
K l 2 1
l 2
(h(x)x )2 dx +
(
0
13
Kl2 1 (2 + c2 (x)) 2
) dx.
2
Then
|t (Y (t))|
I
a1 b2
a2 b2
1
hx + 1 I c(x) + 1
+ 1
+
2
2
2
2a3
0
2
2
1 c
I b
+
+ 2 (1 +
))(t )2 dx
2a4
2a6
Z l
3
EI
(h(x))2
hx 1 EIc + 1
+K
(k(x))2
+
(1
2
2a
2a
1
5
0
l 2 1
Kl2 1 (2 + c2 )
(h(x))2 + c0 (
))(x )2 dx
+K
2l
2
Z l
EI
(c(x))2
+
(1
cxx + 1
)()2 dx
2
2a
2
0
Z l
2 K
2 a6
+
(
+
+ 3 Ka7 )( wx )2 dx
I
2I
0
EI
K
EI
K
+ (2
+ 1
h(l))(x (l))2 + (2
1
h(0))(x (0))2
2
2
2
2
K3
K
K3
K
+
k(l))(wx (l))2 + (2
k(0))(wx (0))2
+ (2
2
2
2
2
Z l
2
K
a5
Kd c0
Kd2
+ 3
( kx + K
+
+
)(wx )2 dx
2
2
2a7
2a7
0
Z l
3
1 a3 1 a4 2
+
(
kx 3 d(x) + (
+
) )(t w)2 dx,
2
2
2
0
Z
(b(x) 1
where c0 is the Poincarre constant and c = max(c(x)) on [0, l], then we choose i
(i = 1, 2, 3) sufficiently small and we choose the constant ai (i = 1 . . . 7) such that
a2 b2
I b2
a1 b2
+ 1
+ 2
< 0,
2
2
2a6
I
1 1 c2
1 hx + 1 I c(x) +
+
+ 2 < 0,
2
2a3
2a4
EI
(h(x))2
3
l 2 1
1
hx (x) 1 EIc(x) + 1
+K
(k(x))2 + K
(h(x))2 < 0,
2
2a1
2a5
2l
2 K
2 a6
3 Ka7
+
+
< 0,
I
2I
2
3
1 a3 1 a4 2
kx 3 d(x) + (
+
) < 0,
2
2
2
2
EI
(c(x))
cxx +
< 0,
2
2a2
b(x) + 1
K
a5
Kd c0
Kd2
kx + K
+
+
< 0,
2
2
2a7
2a7
14
EJDE2003/29
Then (Y (t)) satisfies (3.2). The inequality (3.1) follows from the Cauchy Schwartz
inequality. Thus the system is exponentially stable.
Acknowledgements. The authors wish to thank the anonymous referee for his
or her valuable suggestions. A. Soufyane would like to thank Dr. A. Benabdallah
for her help.
References
[1] Afilal M. and Ammar Khodja F., Stability of coupled second order equations.
Pr
epublications de lEquipe de Math
ematiques de Besancon, 1997.
[2] Benchimol C.D, A note on week Stabilizability of contractions semigroups, SIAM J. Control
and Optimization, 16, pp 373.-379, 1978.
[3] Gohberg and Krein M., Introduction to the Theory of Linear non Selfadjoint Operators,
Translations of Math. Monographs, Vol 18, Amer. Math. Soc. Providence, R. I., 1969.
[4] Kim J. U. and Renardy Y., Boundary control of the Timoshenko beam, SIAM J. Control
and Optimization, 25, 6 pp 1417-1429, 1987.
[5] van Neerven J., The Asymptotic Behaviour of a Semigroup of Linear Operators, Birkh
auser
Verlag, 1996.
[6] Neves A. F., Ribeiro H. S. and Lopes O., On the spectrum of Evolution operators generated
by hyperbolic systems, J. Functional Analysis, 67, 3, 1986.
[7] Taylor S. W., A smoothing property of a hyperbolic system and boundary controllability
Control of partial differential equations (Jacksonville, FL, 1998). J. Comput. Appl. Math.
114(2000), no.1, 23-40.
[8] Kapitonov B. V., Uniform stabilization and exact controllability for a class of coupled
hyperbolic systems. Comp. Appl. Math., Vol 15, No 3, pp 199-212, 1996.
[9] Komornik V., Exact Controllabillity And Stabilization. The Multiplier Method, Masson,
1994.
[10] Lagnese J. E., Boundary Stabilization of Thin Plates. SIAM, 1989.
[11] Lagnese J. E. and Lions J. L., Modelling Analysis and Control of Thin Plates, Masson,
1988.
[12] Morgul O., Boundary Control of a Timoshenko Beam Attached to a rigid Body: Planar
motion. Internat. J. Control 54, No. 4, pp 763-791, 1991.
[13] Morgul O., Dynamic Boundary Control of a Timoshenko Beam. Automatica J. IFAC 28,
No. 6, pp 1255-1260, 1992.
[14] Pazy A., Semigroup of linear Operators and Applications to Partial Differential Equations,
Applied Mathematical Sciences, Vol. 44, Springer -Verlag, 1983.
[15] Russell D. L., A General Framework for the study of Indirect Damping Mechanisms in
Elastic Systems. Journal of Mathematical Analysis And Applications 173, 339-358, 1993.
[16] Shi D. H. and Feng D. X., Exponential decay of Timoshenko Beam with locally distributed
feedback. Proceeding of the 99IFAC World Congress, Beijing, Vol F.
[17] Soufyane A., Stabilisation de la poutre deTimoshenko.C.R. Acad.Sci. Paris,Ser.I, 328, pp
731-734, 1999.
[18] Soufyane A., Uniform stability of coupled second order equations. Electron. J. Diff. Eqns.,
Vol. 2001(2001),No. 25, pp 1-10.
[19] Timoshenko S., On the correction for shear of the differential equation for transverse vibrations of prismatic bars, Philisophical magazine 41, pp 744-746, 1921.
[20] Zhuang Y. and Baras J. S., Distributed control of a Timoshenko beam. In Proceeding of
the Third International Conference on Adaptative Structures, San Diago, California, Nov.
9-11, 1992.
Abdelaziz Soufyane
Math & Comp Sciences department, UAE University, PO Box 17551 Al Ain, UAE
E-mail address: Soufyane.Abdelaziz@uaeu.ac.ae
Ali Wehbe
Universite Libanaise Faculte des Sciences, Section 1, Hadath, Beyrouth, Lebanon
E-mail address:
ali wehbe@yahoo.fr