Материал: Weber H., Herziger G., Poprawe R. (eds.) Laser Fundamentals. Part 1 (Springer 2005)(263s) PEo

Внимание! Если размещение файла нарушает Ваши авторские права, то обязательно сообщите нам

124

3.1.7 Beam propagation in optical systems

[Ref. p. 131

Srotated cyl.

with

A =

0 1

1 0

and

= R−1 Scyl R = A B

C D

, B =

0 0

, C =

− cos2 θ/fx

− sin θ cos θ/fx

, D =

1 0

0 0

− sin θ cos θ/fx

− sin2 θ/fx

0 1

O

− sin θ cos θ/fx

− sin2 θ/fx + 1/qyy

Q−1 =

− cos2 θ/fx + 1/qxx

− sin θ cos θ/fx .

Therefore, the output field

k

cos2 θ

1

sin

θ cos θ

u O (r) = exp −i

−

+

x2 − 2

xy +

2

fx

qxx

fx

− sin2 θ + 1 y2

fx qyy

is a general astigmatic Gaussian beam with a mixing term between the coordinates x and y.

3.1.7.3 Waist transformation

Often, the transfer of the beam waist is required for instance for focusing of laser light. Then, the following algorithms are much more simple than the q-parameter algorithm.

3.1.7.3.1 General system (fundamental mode)

In Table 3.1.20 the waist transformation for a general system is given.

3.1.7.3.2 Thin lens (fundamental mode)

The formulae (3.1.123)–(3.1.126) are further simplified using the focal length f for the thin lens only, see Table 3.1.21.

Remark : Discussion of equation (3.1.127):

The right-hand-side term of (3.1.127) containing z0 represents the modification introduced by the Gaussian beam optics to the thin-lens equation ((3.1.95), t 0) shown in Fig. 3.1.42.

In Fig. 3.1.43 the relation of the Gaussian waist transfer to the thin-lens equation of geometrical optics for di erent influences of di raction is shown.

Main modifications of the geometrical optics:

– No “image distance” is at infinity.

–

For z = f (point P ) the image is at z = f (transfer of the object-side focal plane to the image-side

focal plane after (3.1.130), not ∞).

–

If a target z -position is given, then two starting z-positions are possible.

Example 3.1.17. Given for Fig. 3.1.42: z = 1179 mm, w0 = 0.22 mm, λ = 1.06 µm; it follows z = 109 mm, w0 = 0.02 mm, θ = 0.96◦ , and z0 = 1.21 mm. The second right-hand term of (3.1.127) translates the Gaussian waist image by 0.16 mm in comparison with the geometrical optical image towards the lens.

Landolt-B¨ornstein

New Series VIII/1A1

Ref. p. 131]

3.1 Linear optics

125

Table 3.1.20. Waist transformation for a general system.

Given

Solution

– ABCD-matrix of the system,

z

=

−

B)(Cz + D)

ACz2

–

waist w0,

C2z0

+ (Cz + D)

0 for

(Az + 2

−2

C = 0 ,

–

wavelength λ , including z0 = π w02/λ ,

Az + B

for

C = 0 ,

–

distance z to the input plane of the system.

−

D

(3.1.123)

Cz + A

z0

z 0

z0’

z0

= z0

=

,

Optical system

Cz + D

C2z02 + (Cz + D)2

w

0

characterized by its

’

w ’

(3.1.124)

0

z

A B

- matrix

w0 =

z

C D

λ 0

,

(3.1.125)

z

z ’

π

Θ0

= π z0 .

(3.1.126)

λ

Fig. 3.1.41. Waist transformation by an optical

system.

The beam parameter product is invariant:

Asked : Waist w0 and distance z to the output plane

of the system including z .

= w0 Θ0

= λ/π .

0

w0 Θ0

Table 3.1.21. Waist transformation by a thin lens.

Given

Solution

–

Focal length f of the lens,

1

1

1

2

+

=

+

z0

,

(3.1.127)

–

wavelength λ ,

z [z2 + z02

2

/λ ,

z

z

f

− zf ]

–

waist w0 , including z0 = π w0

see Fig. 3.1.42,

–

distance z to the input plane of the system.

w0 = w0

f

,

(3.1.128)

0

0

z02

+ (z − f )2

w

w ’

π w

2

f

f

z0

=

0

.

(3.1.129)

Image point

λ

of paraxial optics

If z = f , then

z

z ’

z

= f

and

w0 =

w0f

.

(3.1.130)

Fig. 3.1.42. Waist transformation by a thin lens.

z0

Asked : Waist w0 and distance z to the output plane of the system and z0 .

Landolt-B¨ornstein

New Series VIII/1A1

126

3.1.7 Beam propagation in optical systems

[Ref. p. 131

4

Geometrical optics

3

z0 /f = 3

2

f 1

z0 /f = 1

P

'/

z0 /f = 0.5

0

z

z0 /f = 0.3

-1

Geometrical optics

z0 /f = 0.2

Fig. 3.1.43. Relation of the Gaussian waist transfer (full

-2

-3

lines) to the thin-lens equation (dashed) of geometrical

-3

-2

-1

0

1

2

3

4

optics for di erent influences of di raction (wavelength λ

z / f

respectively z0).

3.1.7.4 Collins integral

For Fresnel’s approximation of di raction in paraxial systems see [68Goo, 71Col, 78Loh, 94Roe]. It was generalized to the propagation of field distributions in ABCD-described systems by [70Col, 76Arn, 05Gro2, 05Hod].

3.1.7.4.1 Two-dimensional propagation

In Table 3.1.22 the propagation in rotational symmetric systems and simple astigmatic systems is given.

Table 3.1.22. Propagation in rotational symmetric systems and simple astigmatic systems.

Given

Solution

– ABCD-matrix of the optical

Field U O(x2) in the output plane (Collins integral ):

system (see Tables 3.1.11

U O(x2) =

i

and 3.1.12),

e−

ikL

– field distribution in the in-

λ B

put plane U I(x),

∞

d x1 U I(x1) exp −i

k

2

2

.

– path length along the opti-

×

Ax1

− 2x1x2 + Dx2

(3.1.131)

2B

cal axis L.

−∞

Example 3.1.18. The waist of a Gaussian beam is given with U I(x1) = exp (−x21/w2I) in the input plane. The system consists of a thin lens with the focal length f followed by a free-space propagation by distance z. The ABCD-matrix is calculated from Fig. 3.1.34 and Table 3.1.11:

C D

−1/f

1

A B

=

1 − z/f

z .

∞

x2

i

−

−

k

z

1

2

−

−

λ z

w2I

2z

f

U O(x2) =

e−ikL

d x1 exp

1

exp i

1

x2

2x1x2 + x2 .

−∞

Landolt-B¨ornstein

New Series VIII/1A1

Ref. p. 131]

3.1 Linear optics

127

The result is an output Gaussian intensity distribution with the waist radius

w O =

w I

,

1 +

w2

2

π I

λ f

π w O w I

2

the waist position z = zwaist = f

, and z O =

π w2O/λ .

λ f

For inclusion of displacements and misalignments in Collins Integral see [96Tov].

3.1.7.4.2 Three-dimensional propagation

In Table 3.1.23 the propagation in in general astigmatic systems is given.

Table 3.1.23. Propagation in general astigmatic systems.

Given

Solution

– S : matrix of

the optical sys-

Field U O(r2) in the output plane (Collins integral ):

tem, see Table 3.1.15 and

−

i exp(−i kL)

(3.1.99) with

U O(r2) =

d r1U I (r1)

λ

√

det B

A B

k

S = C D

,

× exp −i

r1 B−1 A r1 − 2 r1 B−1 r2 + r2 D B−1 r2

(3.1.132)

2

– field distribution in the input

with det B the determinant and B−1 the inverse of the matrix B .

plane: U I(r1) , where r1 is the

Examples in [70Col, 05Gro2, 05Hod].

position vector in the input

plane.

3.1.7.5Gaussian beams in optical systems with stops, aberrations, and waveguide coupling

3.1.7.5.1Field distributions in the waist region of Gaussian beams including stops and wave aberrations by optical system

Classical cases of optical system design are given in [99Bor, 80Hof, 86Haf]. [82Wag, 95Gae] use the calculation of the field distribution in the image by a stop and wave aberrations in the exit pupil.

The analog is modeled for Gaussian beams on the exit pupil in the following references:

–focused Gaussian beams with aberrations and stops: see [69Cam, 71Sch],

–obscuration of a rotationally symmetrical Gaussian beam including longitudinal focal shift: see [82Car, 86Sta],

–extended systematic discussion of di raction with stops, obscuration, and aberrations: see [86Mah, 01Mah],

–spherical aberration: see [98Pu].

Landolt-B¨ornstein

New Series VIII/1A1

128

3.1.7 Beam propagation in optical systems

[Ref. p. 131

3.1.7.5.2 Mode matching for beam coupling into waveguides

The calculation of the excitation coe cient of an eigenmode in a waveguide (output mode) by the incident mode (input mode) at the surface of the waveguide is described in Table 3.1.24.

This task occurs

–if a laser beam is formed by an optical system and coupled afterwards into an optical fiber,

–if a laser beam of a master oscillator is to be coupled into a power amplifier,

–in the case of waveguide-waveguide coupling especially fiber-fiber coupling or coupling between semiconductor lasers.

Solutions are available in commercial optical design programs.

Table 3.1.24. Definitions for waveguide coupling.

Given

Solution

–

Incident beam (emitted by a laser

Coupling coe cient (power relation):

(and) transformed by an optical system):

O IO O IO

Einput (x, y) .

η =

.

(3.1.133)

–

Waveguide with an eigenmode field the

N I N O

coupling to which is asked: Eoutput (x, y) .

Overlap integral :

Plane of mode

O IO =

∞

∞

(3.1.134)

x

d x

d y E I(x, y) E O(x, y) .

matching

−∞

−∞

Einput ( x)

Eoutput ( x)

Normalization:

z

∞

∞

N I =

d x

d y E I(x, y) E I (x, y) .

(3.1.135)

Waveguide

−∞

−∞

Fig. 3.1.44. Mode matching.

Normalization:

Asked: Part of power transmitted into the waveguide (fiber, laser, integrated optical waveguide).

∞

∞

N O =

d x d y E O(x, y) E O(x, y) .

(3.1.136)

−∞

−∞

E ective antireflection layers are assumed to be on the waveguide.

3.1.7.5.3 Free-space coupling of Gaussian modes

For the case that a Gaussian output waist of a source waveguide and a Gaussian input waist of a receiver waveguide are separated by air, the coupling of both waveguides is generally treated in [64Kog]. Higher-order modes are also included. The approximation of small misalignments (o set and tilt) is given in Table 3.1.25, large o sets and tilts are treated in [64Kog, 91Wu].

Landolt-B¨ornstein

New Series VIII/1A1

Источник: https://files.student-it.ru/previewfile/280942