Chapter18 MatrixMethodsinParaxialOptics
LectureNotesforModernOpticsbasedon Pedrotti&Pedrotti&Pedrotti Instructor:NayerEradat Instructor: Nayer Eradat Spring2009
2/20/2009
Eradat,SJSU,MatrixMethodsinParaxialOptics
Matrixmethodsinparaxialoptics Matrix methods in paraxial optics
Describingasinglethicklensintermsofitscardinalpoints. Describingasingleopticalelementwitha2x2matrix. Analysisoftrainofopticalelementsbymultiplicationof2x2matricesdescribing eachelement. Computerraytracingmethods,amoresystematicapproach
2/20/2009
MatrixMethodsinParaxialOptics
Cardinalpointsandcardinalplanes
Wedefinesixcardinalpointsontheaxisofathicklensfromwhichitsimaging propertiescanbededuced. properties can be deduced. Planesnormaltotheaxisatthecardinalpointsarecalledcardinalplanes. Cardinalpointsandplanesinclude Firstandsecondsetoffocalpointsand focalplanes. First and second set of focal points and focal planes. Firstandsecondprincipalpointsand principalplanes.Theraysdeterminingthe focalpointschangedirectionattheirintersectionwiththeprincipalplanes.
2/20/2009
MatrixMethodsinParaxialOptics
Cardinalpointsandcardinalplanes
Firstandsecondnodalpointsand nodalplanes.Nodalpointsofathicklensorany opticalsystempermitcorrectiontotheraythataimsthecenterofthelens.Anyray optical system permit correction to the ray that aims the center of the lens. Any ray thataimsthefirstnodalpointemergesfromthesecondnodalpointundeviatedbut slightlydisplaced.
2/20/2009
MatrixMethodsinParaxialOptics
Cardinalpointsand Cardinal points and cardinalplanes
Allthedistancesthataredirectedtotheleftarenegative()anddirectedtotheright arepositive(+)bythesignconvention.Noticethatfocaldistancesarenotmeasured are positive (+) by the sign convention. Notice that focal distances are not measured fromthevertices
2/20/2009
MatrixMethodsinParaxialOptics
1 nL n ' nL n ( nL n )( nL n ' ) t = f1 nR2 nR1 nnL R1 R2 n' for n = n ' then f 2 = f1 f1 n Location of the principal planes: f2 = r= nL n ' n n f1t ; s = L f 2t nL R2 nL R1
Basicequationsforthethicklens
The positions of the nodal points: n' n n' n n n v = 1 + L t f1 ; w = 1 + L t f2 n nL R2 n ' nL R1 Image and object distances and lateral magnification: ns f f 1 + 2 = 1 and m = i so si n ' so The sign convention is as usual (real + and virtual ) as long as the distances are measured relative to their corresponding principla planes. For an ordinary thin lens in air: n = n ' = 1 and r = v, s = w we arive at the usual thin lens equations: s 1 1 1 and m = i and f = f 2 = f1 + = s o si f so
2/20/2009 MatrixMethodsinParaxialOptics 6
Thematrixmethodsinparaxialoptics
Foropticalsystemswithmanyelementsweuseasystematicapproachcalledmatrix method. Wefollowtwoparametersforeachrayasitprogressesthroughtheopticalsystem. Arayisdefinedbyitsheightanditsdirection(theangleitmakeswiththeoptical ) axis). Wecanexpressy7 and7intermsofy1 and1 multipliedbythetransfermatrixof thesystem.
2/20/2009
MatrixMethodsinParaxialOptics
Thetranslationalmatrix Consider simple tanslationof a ray in a homogeneous medium.
Translation from point 0 to 1 with paraxial approximation: T l ti f i t t ith i l i ti
1 = 0 and y1 = y0 + L tan 0 = y0 + L 0
We rewrite the equations: y1 = (1) y0 + ( L ) 0 y1 = 1 = ( 0 ) y0 + (1) 0 1 1 L y0 0 1 0
ray-transfer matrix for translation
2/20/2009
MatrixMethodsinParaxialOptics
Refractionmatrix
Consider refraction of a ray at a spherical interface (paraxial approximation): Ray coordinates before refraction ( y, ) and ray coordinates after refraction ( y', ') y y and = = R R Paraxial form of Snell's law: n = n ' '
' = ' = '
'=
y n y y n = + R n' R R n' 1 n n ' = 1 y + R n' n' The approximate linear equations: y' = (1) y + ( 0 ) 1 1 n 1 R n' 0 n n '
y ' 1 n n = ' = 1 y + ' n' R n '
y 1 0 0 n n'
Ray-transfer matrix for refraction
y ' If R we have transfer matrix for refration by plane interface: = '
2/20/2009 MatrixMethodsinParaxialOptics
y
9
Refraction by a plane
Thereflectionmatrix
Consider refraction of a ray at a spherical interface (paraxial approximation): Ray coordinates before refraction ( y, ) and ray coordinates after refraction ( y', ') y y and = = R R Goal: connect ( y', ') to ( y, ) by a ray transfer matrix for reflection by a concave mirror
' = ' = '
Sign convention for the angles: ( + ) pointing upward and ( ) pointing downward
= + = +
y y and ' = ' = ' R R To eliminate and ' we use = ' we get
' = +
y 2y = + R R q The desired equations become: y' = (1) y + ( 0 ) 1 2 R 0 y 1
10
y ' = 2 ' = y + (1) ' R
2/20/2009
Ray-transfer matrix for reflection
MatrixMethodsinParaxialOptics
Thethicklensandthinlensmatrices ThegeneralRaytransfermatrix
Goal: G l construct a matrix that represents a thick l i h hi k lens with two diff ih different material on each side of it. i l h id f i In traversing the lens the ray undergoes two refractions and one translation for which we have derived the matrices. The radii of curvature are ( + ) in this example. The symbolic equations are: y0 y1 = M 1 for the first reflection 0 1 y y0 y2 y1 3 = M 3M 2 M1 = M 2 for the translation 2 1 3 M : Transfer matrix 0 of the entire lens y3 y2 = M 3 for the second reflectiont 2 3 The i di id l Th individual matrix operates on the ray i the same order in which the optical actins i fl i h in h d i hi h h i l i influense the ray. h No comutative propery for multiplication of matricies. Only associative property holds. M=M 3 M 2 M 1 = ( M 3 M 2 ) M 1 = M 3 ( M 2 M 1 ) M 2 M 3 M 1 Generalizing the matrix relationship for any number of translating, reflecting, refracting surfaces: yf y0 = M N M N 1 M 2 M 1 with M = M N M N 1 f 0 M : Transfer matrix
of the entire lens
M 2 M 1 ray transfer matrix for the optical system.
2/20/2009
MatrixMethodsinParaxialOptics
11
Thethicklensandthinlensmatrices
Goal: Applying the results for a thick lens Let R represent a reflaction matrix and T represent tsranslation M=R 2 TR1 the ray-transfer matrix for a thick lens can be written as: 0 0 1 1 1 t M = nL n ' nL n nL n 0 1 n' n ' R2 nL R1 nL y For a thin lens t 0 in one environment ( n = n ') the ray-transfer matrix becomes 1 0 0 0 1 1 1 0 M = nL n nL n nL n = nL n 1 1 0 1 1 n R n nL n R2 R1 nR2 L 1 We exprss the lower left hand element in terms of the focal length: 1 nL n 1 1 = the lansmaker's formula f n R2 R1 1 0 = A B M = 1 1 C D f the ray-transfer matrix for a thin lens also known as the ABCD matrix.
2/20/2009 MatrixMethodsinParaxialOptics 12
2/20/2009
MatrixMethodsinParaxialOptics
13
Significanceofsystemmatrixelements
y f A B y0 = f C D 0 a) If D = 0 f = Cy0 independent of 0 All the rays leavoing the input plane will have the same angle at the output plane. Input plane is on the first focal plane. b) If A = 0 y f = B 0 means yf is independent of y0 that means all the rays departing input plane have the same height at the output plane. This means output plane is the second focal plane. c) If B = 0 y f = Ay0 All the points leaving the input plane at hight y0 will arrive the output plane at height yf output plane is image of the input plane. A=y f / y0 corresponds to linear magnification. d) If C = 0 a f = D 0 independent of y0 p y Input rays of all in one direction will produce output rays all in another direction. This is called (thelescopic system).
2/20/2009 MatrixMethodsinParaxialOptics 14
Example18.3
2/20/2009
MatrixMethodsinParaxialOptics
15
Locationofcardinalpointsforanopticalsystem
Since the system ray-transfer matix explains the optical properties of an optical system we expect a relationship between the system matrix and location of the cardinal points. Input and output planes define limits of an optical system. We define distances locating six cardinal planes with respect to the input and output planes. F1 and F2 are at f1 and f 2 from the principal points at H1 and H 2 F1 and F 2 are at p and q from the reference input and output planes r and s are disances of the reference input and output planes from the principal points at H1 and H 2 p p p p p p v and w are disances of the reference input and output planes from the nodal points at N1 and N 2 Sign convention:
( + ) distance measured to the
right of a reference plane
( - ) distance measured to the
left of a reference plane
2/20/2009
MatrixMethodsinParaxialOptics
16
Input ray ( y 0 , 0 ) and output ray ( y f , 0 ) from figure (a) y f = Ay0 + B 0 D y0 = 0 0 = Cy0 + D 0 C For small angles 0 = y0 y D p= 0 = p 0 C
Locationofcardinalpoints
p is negative that means it is to the left of the input reference plane.
0 =
f1 = f1 =
yf f1 yf = ( Ay0 + B 0 ) = AD B C
AD BC Det ( M ) n0 1 = = = f1 nf C C C n0 nf
We used: Det ( M ) = AD BC = n 1 D 0 C nf =r
r = p f1 =
2/20/2009
MatrixMethodsinParaxialOptics
17
Locationofcardinalpoints
Using fi U i figure (b) we can fi d q, f 2 , s find A 1 A 1 ; s= ; f2 = q s = C C C Using figure (c) we can find v, w q=
0 = f = =
y0 v Notice y 0 is negative i.e. below the optical axis
= Cy0 + D
y0
1 D = v C
( n0 / n f ) A D 1 v= and w = C C
2/20/2009
MatrixMethodsinParaxialOptics
18
1) When initial and final material have the same index of refraction then r = v and s = w i.e. ) principal points and nodal points coinside 2) When initial and final material have the same index of refraction then first and second focal lengths are equal f1 = f 2 3) The separation of the principal points is the same as separation of the nodal points or r s = v w
2/20/2009 MatrixMethodsinParaxialOptics 19
Examples:towthinlensesinairseparatedbyadistanceL
2/20/2009
MatrixMethodsinParaxialOptics
20
Examples:towthinlensesinairseparatedbyadistanceL
Focal lengths of the lenses f A , f B , Assume the input and output reference planes are located on the lenses lenses. The system transfer matrix includes two thin-lens matrices and a translation matrix. L 1 L 0 0 1 1 fA 1 L 1 = M = 1 0 1 1 1 1 1 1 L 1 fB fB 1 fB fB f A f A 1 1 st oca e gt o t e syste : a d t e seco d oca e gt o t e syste : First focal length of the system: f1 = and the second focal length of the system: f 2 = = f eq C C L 1 1 1 = + f eq f A f B f A f B f eq The first principal point and nodal point: r = v = L fB The second principal point and f eq nodal point: s = w = d l i t L fA
2/20/2009
MatrixMethodsinParaxialOptics
21
Example
2/20/2009
MatrixMethodsinParaxialOptics
22
Raytracing
Limitinganalysisofopticalsystemstoparaxialraysisanoversimplificationoftheproblem Limiting anal sis of optical s stems to para ial ra s is an o er simplification of the problem andignoreseffectofaberrations. Raytracingisfollowingtheactualpathofeachraythroughthesystemusinglawsof reflectionandrefraction.Traditionallyitisdonebyhandandgraphicallybuttodayitisall computerized. Weintroducearaytracingtechniquethatisoftenlimitedtomeridionalrays. Meridionalraysaretheraysthatpassthroughtheopticalaxisofthesystem. Meridionalraystendtostayinthemeridionalplanesasthelawsofrefraction/reflection Meridional rays tend to stay in the meridional planes as the laws of refraction/reflection requirethem. Thislimitsourtreatmenttoa2dimensionalspace. Skewraysaretheonesthatcontributetotheimageanddonotpasstheopticalaxis. Analysisoftheskewraysrequirea3dimensionaltreatment. Understandingaberrationsrequireanalysisofthenonparaxialraysandskewrays. Designofthecomplicatedlenssystemsrequireknowledgeandexperiencewithraytracing techniquesandoptimizingperformanceofthesystembychangingsystemparametersand techniques and optimizing performance of the system by changing system parameters and arrivingataperfectperformance.
2/20/2009
MatrixMethodsinParaxialOptics
23
Raytracing
Goal: followin a meridional ray through a single spherical refracting surface. n, n ' : Indexes of refraction; R radius of curvature; A origin of the ray. , ' : angle with optical axis before and after refraction. O point of intersection with the optical axis; P with the refracting surface; I with the optical axis after refraction. I & O are conjugate poits with distances s , s ' from the vertices. Q: perpendicular distance from the vertex, V, to the incident ray. , ' : angles of incidence and refraction. Sign convention: distances to the left of vertex and to the right of V are + above the optical axis + and below are . From left to right, the angles have the same sign as the slopes. Input parameters for each ray: h : elevation, : a g e, and put pa a ete s o eac ay: e evat o , angle, a d D: distance from the vertex parallel to the optical axis. From figure we write: s = D h Q ; in OBV sin = Q = s sin tan s R sin + Q sin = R
a+Q R a in VNC sin = R in PMC sin =
= sin 1
Q + sin R
n sin At P n sin = n 'sin ' ' = sin 1 n' in CPI = ' ' ' = ' +
2/20/2009 MatrixMethodsinParaxialOptics 24
Raytracing
Q': perpendicular distance from the vertex, V, to the refracted ray. a' R Q ' = R sin ' sin ' ( ) Q ' a ' In PLC sin ( ' ) = R In CMV sin ( ') = In ITV sin ( ') = Q' Q ' s' = sin ' s'
Now we have the new values for the refracted ray ', Q ', s ' which prepare us for the next refraction in the sequence. y But before that we need to calculate effect of the transfer by t in the material with index n '. V M Q' Q In V2 MV1 sin ( 2 ) = 1 = 1 2 Q2 = Q '1 + t sin 2 t t And 2 = ' We W need to modify th equations dt dif the ti for the special cases: 1) Incident ray is parallel to the optical axis. 2) Surface is plane with an infinite radius of curvature.
2/20/2009
MatrixMethodsinParaxialOptics
25
2/20/2009
MatrixMethodsinParaxialOptics
26
The rays are parallel to the axis so we use the second column of the table
Exampleraytracing
Do D a ray t trace f t rays through for two th h a rapid landscape photographic lens of three elements. The parallel rays enter the lens from a distant object at altitudes of 1 and 5 mm above the optical axis. The lens specifications are: R1 = 120.8 R2 = 34.6 R3 = 96.2 R4 = 51 2 51.2 t1 = 6 t2 = 2 t3 = 3 n1 = 1.521 n2 = 1.581 n3 = 1.514 1 514
Input First surface : n = 1, n ' = 1.521 =0 h = 1 or 5 R = -120.8 Second surface : t =6 n = 1.581 R = -34.6 34 6
Results ray at h=1 Q =1
Results ray at h=5 Q=5
' = 0.16250
s ' = 352.66 Q ' = 1.0000 Q = 1.0170
' = 0.81280
s ' = 352.53 Q ' = 5.0010 Q = 5.0861
' = 0.22020
s ' = 264.59 264 59 Q ' = 1.0170 Q = 1.0247
' = 1.10410
s ' = 264.03 264 03 Q ' = 5.0876 Q = 5.1261
Third surface : t=2 n = 1.514 R = -96.2 Final surface : t =3 n = 1.581 R = -51.2 51 2
' = 0.20300
s ' = 289.26 Q ' = 1.0247 Q = 1.0353
' = 1.01780
s ' = 288.58 Q ' = 5.1260 Q = 5.1793
' = 0.28830
s ' = 205.72 Q ' = 1.0353
' = 1.45200
s ' = 203.91 Q ' = 5.1672
27
2/20/2009
MatrixMethodsinParaxialOptics
There is no common focus s ' = 1.8mm
2/20/2009
MatrixMethodsinParaxialOptics
28
