Energy decay for Maxwell s equations with Ohm s law on partially cubic domains

Description

Niveau: Secondaire, Lycée, Terminale
Energy decay for Maxwell?s equations with Ohm?s law on partially cubic domains Kim Dang Phung Yangtze Center of Mathematics, Sichuan University, Chengdu 610064, Sichuan Province, China. E-mail: Abstract .- We prove a polynomial energy decay for the Maxwell?s equations with Ohm?s law on partially cubic domains with trapped rays. Keywords .- Maxwell?s equation; decay estimates; trapped ray. 1 Introduction The problems dealing with Maxwell?s equations with nonzero conductivity are not only theoretical interesting but also very important in many industrial applications (see e.g. [3], [7], [8]). Let be a bounded open connected region in R3, with a smooth boundary @ . We suppose that is simply connected and @ has only one connected component. The domain is occupied by an electromagnetic medium of constant electric permittivity o and constant magnetic permeability o. Let E and H denote the electric and magnetic ?elds respectively. The Maxwell?s equations with Ohm?s law are described by 8 >>>>< >>>>: o@tE curlH + E = 0 in [0;+1) o@tH + curlE = 0 in [0;+1) div (oH) = 0 in [0;+1) E = H = 0 on @ [0;+1) (E;H) (; 0) = (Eo;Ho) in .

• conductivity has


• scalar wave


• polynomial energy


• wave operator


• interpolation estimate


• has only


• unique solution


• estimate becomes


• energy decay

Subjects

• Academic document
• Education
• Secondary scrool
• Première
• Scalar field
• D'Alembert operator

Informations

• Published by: profil-zyan-2012
• Reads: 13
• Language: English

Exrait

Energy

decay for Maxwells equations with on partially cubic domains 

Ohmslaw

Kim Dang Phung Yangtze Center of Mathematics, Sichuan University, Chengdu 610064, Sichuan Province, China. _ g_phung@ya o.fr E-mail: hokim dan

Abstract .-We prove a polynomial energy decay for the Maxwells equations with Ohms law on partially cubic domains with trapped rays.

Keywords .-Maxwells equation; decay estimates; trapped ray.

1 Introduction

The problems dealing with Maxwells equations with nonzero conductivity are not only theoretical interesting but also very important in many industrial applications (see e.g. [3], [7], [8]). Letbe a bounded open connected region inR 3 , with a smooth boundary@. We suppose that is simply connected and@has only one connected component. The domainis occupied by an electromagnetic medium of constant electric permittivity" o and constant magnetic permeability o . LetEandH Thethe electric and magnetic elds respectively.denote  Maxwells equations with Ohms law are described by 8 " o @ t EcurlH+E= 0in[0;+1) > <  o @ t H (v + i c d u o rl H E )0= =0in in    [0 0[ ; ; + + 1 1 ) ) (1.1) E=H= 0on@[0;+1) > : (E; H) (;0) = (E o ; H o )in. Here,(E o ; H o )are the initial data in the energy spaceL 2 () 6 anddenotes the outward unit normal vector to@. The conductivity is such that2L 1 ()and0. It is well-known that when the conductivity is identically null, then the above system is conservative and whenis bounded from below by a positive constant, then an exponential energy decay rate holds for the Maxwells equations with Ohms law in the energy space. The situation becomes more delicate when we only assume that

(x)constant >08x2!

for some non-empty connected open subset!of that the condition. Observediv (" o E) = 0in [0;+1)does not appear because the free divergence is not preserved by the Maxwells equations with Ohms law. Here we know that the above system is dissipative and its energy tends to zero in large time. However, we would like to establish the energy decay rate as well. In the eld of control theory, the exponential energy decay rate of a linear dissipative system is deduced from an observability  This work was supported by the NSF of China under grants 10771149 and 60974035.

1

estimate. Precisely, in order to get an exponential decay rate in the energy space we should have the following observability inequality 9C; T c >080 Z  j(E; H) (; )j 2 dxC Z  +T c Z  jEj 2 dxdt

or simply, in virtue of a semigroup property, 9C; T c >0 Z  j(E o ; H o )j 2 dxC Z 0 T c Z  Ej 2 d jxdt for any initial data(E o ; H o )in the energy spaceL 2 () 6  can also look for establishing the. We above observability inequality for any initial data in the energy space intersecting suitable invariant subspaces but not with the conditiondivE o = 0in. Such estimate is established in [11] under the geometric control condition of Bardos, Lebeau and Rauch [2] for the scalar wave operator and when the conductivity has the property that(x)constant >0for allx2!and(x) = 0for allx2n! now, we consider a subset. From!that the geometric control condition for thesuch scalar wave operator or other assumptions based on the multiplier method fail. In such geometry, we do not hope an exponential energy decay rate in the energy space. Our geometry (described precisely in Section 3) presents parallel trapped rays and can be compared to the one in [12] or in [4],[10] for the two dimensional case. It generalises the cube (see [8]) and therefore explicit and analytical results are harder to obtain. Our main result gives a polynomial energy decay with regular initial data. Our proof is based on a new kind of observation inequality (see (4.33) below) which can also be seen as an interpolation estimate. It relies with the construction of a particular solution for the operator i@ s +h  @ t 2  inspired by the gaussian beam techniques. the dispersion property for the one Also dimensional Schrödinger operator will play a key role.

The plan of the paper is as follows. In the next section, we recall the known results about the Maxwells equations with Ohms law that will be used in the following. Section 3 contains the statement of our main result, while Section 4 is concerned with its proof. In Section 5, we present the interpolation estimate, while Section 6 includes its proof. Finally, two appendix are added dealing with inequalities involving Fourier analysis.

2 The Maxwells equations with Ohms law

We begin to recall some well-known results concerning the Maxwells equations with Ohms law: well- posedness, energy identity, standard orthogonal decomposition and asymptotic behaviour in time of the energy of the electromagnetic eld.

2.1 Well-posedness of the problem

Let us introduce the spaces V=L 2 () 3  n G2L 2 () 3 ; divG= 0; G j@ = 0 o , W= n (F; G)2L 2 () 6 ; curlF2L 2 () 3 ; F j@ = 0; divG= 0; G j@ = 0;curlG2L 2 () 3 o .

2

(2.1.1)

(2.1.2)

It is well-known that if(E o ; H o )2 V, there is a unique weak solution(E; H)2C 0 ([0;+1);V). Further, if(E o ; H o )2 W, there is a unique strong solution(E; H)2C 0 ([0;+1);W)\C 1 ([0;+1);V). Let us dene the functionals of energy E(t 2 1 =) Z   " o jE(x; t)j 2 + o jH(x; t)j 2  dx, (2.1.3) E 1 (t 2 ) 1 = Z   " o j@ t E(x; t)j 2 + o j@ t H(x; t)j 2  dx. (2.1.4) We can easily check that the energyEis a continuous positive non-increasing real function on[0;+1) and further for any initial data(E o ; H o )2 W, d d t E(t) + Z  x)jE.5) ( (x; t)j 2 dx= 0, (2.1 and for anyt 2 > t 1 0, Z t 1 t 2 Z  (x)jE(x;6) E(t 2 ) E(t 1 ) + t)j 2 dxdt= 0, (2.1. E 1 (t 2 ) E 1 (t 1 ) + Z t 1 t 2 Z  (x)j@ t E(x; t)j 2 dxdt= 0. (2.1.7)

2.2

Orthogonal decomposition

BothEand o Hcan be described, by means of the scalar and vector potentialspandAwith the Coulomb gauge, in an unique way as follows. Proposition 2.1-.For any initial data(E o ; H o )2 W, there is a unique(p; A)2C 1  [0;+1); H 01 ()   C 2  [0;+1); H 1 () 3  such that(E; H)the solution of (1.1) the Maxwells equations with Ohms law satises

  E o = H  = r c p ur  lA @ t A 8 : < " o  o @ t 2 A+ curl curlA= o (" o @ t A r di p  + v  A  = = E 0 ) 0n n n o i i   @   [ 0 0[ [ ; ;+ + 0;+ 1 1) ) 1)

(2.2.1)

(2.2.2)

and we have the following relations kEk 2 L2() 3 =krpk 2 L2() 3 +k@ t Ak 2 L2() 3 ,(2.2.3) k" o @ t rpk L 2 () 3  kEk L 2 () 3 ,(2.2.4) 9c >0kAk 2 L2() 3 ckcurlAk 2 L2() 3 .(2.2.5) Further, sincecurlH2L 2 () 3 ,curl curlA2L 2 () 3 anddivA2H 01 (). The proof is essentially given in [11, page 121] from a Hodge decomposition and is omitted here. Now, the vector eldAhas the nice property of free divergence and satises a second order vector wave equation with homogeneous boundary conditionA= divA= 0and with a second member inC 1  [0;+1); L 2 () 3  bounded by2 o kEk L2() 3  the sake of simplicity, we assume from now. For that" o  o 1. =

3