Файл: Doicu A., Wriedt T., Eremin Y.A. Light scattering by systems of particles (OS 124, Springer, 2006.pdf

ВУЗ: Не указан

Категория: Не указан

Дисциплина: Не указана

Добавлен: 28.06.2024

Просмотров: 776

Скачиваний: 0

ВНИМАНИЕ! Если данный файл нарушает Ваши авторские права, то обязательно сообщите нам.

and

Hs(r) 0

1.4 Scattered Field

 

 

 

 

 

 

= × S hs (r ) g (ks, r, r ) dS(r )

 

 

 

j

 

es (r ) g (ks, r, r ) dS(r ) ,

r Ds

 

k0µs

 

× ×

S

r Di

35

,

where g is the Green function and the surface fields es and hs are the tangential components of the electric and magnetic fields on the particle surface, i.e., es = n × Es and hs = n × Hs, respectively. In the above equations we use a compact way of writing two formulas (for r Ds and for r Di) as a single equation.

A similar result holds for vector functions satisfying the Maxwell equations in bounded domains. With Ei, Hi being a solution to Maxwell’s equations in Di we have

−Ei(r) 0

and

−Hi(r) 0

= × ei (r ) g (ki, r, r ) dS(r )

S

+

j

 

 

 

 

× × hi (r ) g (ki, r, r ) dS(r ) ,

k ε

 

0

i

S

= × hi (r ) g (ki, r, r ) dS(r )

S

 

j

 

 

 

 

× × ei (r ) g (ki, r, r ) dS(r ) ,

k µ

 

0

i

S

r Di r Ds

r Di , r Ds

where ei = n × Ei and hi = n × Hi.

A rigorous proof of these representation theorems on the assumptions Es, Hs C1(Ds) ∩ C(Ds) and Ei, Hi C1(Di) ∩ C(Di) can be found in Colton and Kress [39]. An alternative proof can be given if we accept the validity of Green’s second vector theorem for generalized functions such as the threedimensional Dirac delta function δ(r − r ). To prove the representation theorem for vector fields satisfying the Maxwell equations in bounded domains we use the Green second vector theorem for a divergence free vector field a ( · a = 0). Using the vector identities × × b = b + · b and a · ( · b) = · (a · b) for · a = 0, and the Gauss divergence theorem

we see that

 

 

 

 

D [a · b + b · ( × × a)] dV =

S {n · a ( · b) + n · [a × ( × b)

 

+ ( × a) × b]} dS .

(1.64)

In the above equations, the simplified notations · a and a · b should be understood as ( · a) and a( · b), respectively. Next, we choose an


36 1 Basic Theory of Electromagnetic Scattering

arbitrary constant unit vector u and apply Green’s second vector theorem (1.64) to a(r ) = Ei(r ) and b(r ) = g(ki, r , r)u, for r Di. Recalling that

 

g (k , r , r) + k2g (k , r , r) =

δ (r

r) ,

(1.65)

 

i

i i

 

 

 

and × × Ei = ki2Ei, we see that the left-hand side of (1.64) is

 

 

{Ei (r ) · g (ki, r , r) u + g (ki, r , r) u · [ × × Ei (r )]}

Di

 

dV (r )

 

 

 

 

 

 

×

 

 

 

 

 

 

= − u · Ei (r ) δ (r − r) dV (r ) = −u · Ei(r) .

Di

Taking into account that for r = r,

× × g (ki, r , r) u = ki2g (ki, r , r) u + · g (ki, r , r) u , we rewrite the right-hand side of (1.64) as

{n · Ei [ · g (ki, ·r) u] + n · [Ei × ( × g (ki, ·r) u)

S

+ ( × Ei) × g (ki, ·r) u]} dS

={n · Ei [ · g (ki, ·r) u] + n · [Ei × ( × g (ki, ·r) u)]

S

+ 12 n · [( × Ei) × ( × × g (ki, ·r) u) ki

( × Ei) × · g (ki, ·r) u]} dS .

From Stokes theorem we have

n · { × [Hi · g (ki, ·r) u]} dS = 0 ,

S

whence, using the vector identity ×(αb) = α×b+α ×b and the Maxwell equations, we obtain

S n · Ei [ · g (ki, ·, r) u] dS

 

 

=

1

 

n

·

[(

 

×

E

)

 

 

·

g (k , , r) u] dS .

 

 

k2

S

 

 

i

 

×

 

i ·

 

i

 

 

 

 

 

 

 

 

 

 

Finally, using the vector identity a ·(b ×c) = (a ×b) ·c, the symmetry relationg(ki, r , r) = − g(ki, r , r), the Maxwell equations and the identities


1.4 Scattered Field

37

[ × g (ki, r , r) u] · [n (r ) × Ei (r )]

= { × g (ki, r , r) [n (r ) × Ei (r )]} · u

and

[ × × g (ki, r , r) u] · [n (r ) × Hi (r )]

= { × × g (ki, r , r) [n (r ) × Hi (r )]} · u ,

we arrive at

 

 

 

 

 

 

!

 

 

 

 

−u · Ei(r) = u · ×

S n (r ) × Ei (r ) g (ki, r, r ) dS(r )

 

+

j

 

 

 

"

 

 

× × n (r ) × Hi (r ) g (ki, r, r ) dS(r ) .

k ε

 

 

0

i

S

 

Since u is arbitrary, we have established the Stratton–Chu formula for r Di. If r Ds, we have

u · Ei (r ) δ (r − r) dV (r ) = 0 ,

Di

and the proof follows in a similar manner. For radiating solutions to the Maxwell equations, we see that the proof is established if we can show that

SR {n · [Es × ( × g (ks, ·, r) u)

 

 

 

+

1

(

 

×

E

)

(

 

g (k ,

, r) u)" dS

0 ,

k2

 

 

 

s

 

×

× ×

s ·

 

 

 

s

 

 

 

 

 

 

 

 

 

 

 

as R → ∞, where SR is a spherical surface situated in the far-field region. To prove this assertion we use the Silver–M¨uller radiation condition and the general assumption Im{ks} ≥ 0.

Alternative representations for Stratton–Chu formulas involve the free space dyadic Green function G instead of the fundamental solution g [228]. A dyad D serves as a linear mapping from one vector to another vector, and in general, D can be introduced as the dyadic product of two vectors: D = a b. The dot product of a dyad with a vector is another vector: D · c = (a b) · c = a(b · c) and c · D = c · (a b) = (c · a)b, while the cross product of a dyad with a vector is another dyad: D×c = (a b)×c = a (b×c) and c × D = c × (a b) = (c × a) b. The free space dyadic Green function

is defined as

 

 

 

 

 

 

 

 

1

 

 

G (k, r, r ) =

I +

 

g (k, r, r ) ,

 

k2


38 1 Basic Theory of Electromagnetic Scattering

where I is the identity dyad (D · I = I · D = D). Multiplying the di erential equation (1.65) by I and using the identities

× × Ig = g − I g ,

× × ( g) = 0 ,

gives the di erential equation for the free space dyadic Green function

× ×

G

(k, r, r ) = k2

G

(k, r, r ) + δ (r − r )

I

.

(1.66)

The Stratton–Chu formula for vector fields satisfying the Maxwell equations in bounded domains read as

−Ei(r)

=

ei (r )

 

×

 

 

(ki, r, r ) dS(r )

 

 

 

G

 

 

 

0

 

S

·

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

r

Di

 

 

 

 

 

 

 

 

 

+jk0µi

hi (r ) · G (ki, r, r ) dS(r ) ,

r

Ds

 

 

 

S

 

 

 

 

 

 

 

 

 

and this integral representation follows from the second vector-dyadic Green theorem [220]:

 

a

 

 

 

 

 

(

× ×

a)

·

 

 

 

dV

 

 

 

D

D

D

 

· × ×

 

 

 

 

 

 

 

=

 

a

 

 

 

 

+ (

×

a)

×

 

 

 

dS ,

n

 

 

D

 

D

 

S

·

× ×

 

 

 

 

 

 

 

applied to a(r ) = Ei(r ) and D(r ) = G(ki, r , r), the di erential equation

for the free space dyadic Green function, and the identity (b×D) = (a×b)· D. For radiating solutions to the Maxwell equations, we use the asymptotic behavior of the free space dyadic Green function in the far-field region

r

 

 

 

 

 

 

1

 

 

× × G (ks, r, r ) + jksG (ks, r, r ) = o

 

 

, as r → ∞ ,

r

 

r

to show that the integral over the spherical surface vanishes at infinity.

Remark. The Stratton–Chu formulas are surface-integral representations for the electromagnetic fields and are valid for homogeneous particles. For inhomogeneous particles, a volume-integral representation for the electric field can be derived. For this purpose, we consider the nonmagnetic domains Ds and Di (µs = µi = 1), rewrite the Maxwell equations as

× Et = jk0Ht , × Ht = jk0εtEt in Dt , t = s, i ,

and assume that the domain Di is isotropic, linear and inhomogeneous, i.e.,

εi = εi(r). The Maxwell curl equation for the magnetic field Hi can be written as