Você está na página 1de 15

MATLAB-problems:

Momentary electric field amplitude

Make a program in MATLAB where you simulate the momentary electric field amplitude
from a number of coherent point sources in a plane (start with only 2). Place the sources at
constant distance from each other and observe the electric field at different time instances
within the period of the light. Use different colour maps to illustrate the momentary amplitude
of the electric field.
Show also what happens when the distance is increased between the sources.
What change will you obtain if two point sources have opposite phase? Compare with two
point sources with the same phase. Make an image sequence where you see how the field
advances. Start to show the field without the 1/r dependence. Try also to introduce the 1/r
dependence.

Show your solutions by diagrams and/or images.

Some MATLAB-help

1. Hint: Start by creating the coordinate matrices xmat and ymat, that means the x- and y-
coordinates in all sample points. This means that it is easy to express the formulas clearer.
No for-loops are needed. Maybe like this.
N=200; % This gives a horizontal area of 200x200 image elements.
a = 0.05 % sample distance in the plane in micrometers
la=0.6328 % Wavelength of the light in micrometers
c=0.3 % Velocity of light in micrometers/fs;
xvekt = -N/2*a:a:(N/2-1)*a; yvekt = xvekt; [xmat,ymat]=meshgrid(xvekt,yvekt);

Here we created the coordinate matrices.


The distance from a point source at (x1,y1) in the plane can be written as:

Radius1=sqrt((xmat-x1)^2+(ymat-y1)^2);

The field from a point source is now:

E1=K*cos(2*pi*/la)*(radius1-c*t));

Add the fields from the different point sources and calculate the total field Etot for
different values of t. Use for instance 10 different values of t: la/(10*c)…..la/c.
Try also to introduce the 1/r dependence.
Note! To avoid problems with division by zero it may be necessary to mask the matrix.
Example:
mask= (radius>0.2);
It is possible to see the field in an image sequence like:
for q=1:10 % repeat 10 times
t=t+(la/(10*c));image(vekt,yvekt,Etot);
pause(1) % 1 second pause between images.
end
Try this!
Dispersion 1

The oscillator of the 10 Hz lab delivers pulses whose spectra is centered on λ0=800 nm and a
width of Δλ=50 nm. The spectral amplitude can be expressed as:
  ω −ω 
2

E ( ω,0 ) = E 0 e iϕ (ω)
exp  −  0
 
  ∆ω  .
   
Retrieve the shape of the pulse in the time domain for a phase φ(ω)=0.
When a pulse with a broadband frequency goes through a material of a length L whose
refractive index n(ω) changes with ω the phase can be written as ϕ(ω)=L ωn (ω) . This
c
can be written under a Taylor expansion form:
∂ϕ (ω − ω0 ) 2 ∂ 2ϕ
ϕ(ω) = ϕ(ω0 ) + (ω − ω0 ) + + ...
∂ω 2! ∂ω 2
ω0 ω0

1 ∂ϕ
n

Each coefficient α =
n

n! ∂ωn
is a constant that can be obtained for a given material.
ω0

Retrieve the pulse in the time domain for several α1(ω), α2(ω), α3(ω) etc. Discuss the changes
induced by these chirp rates. Plot these results on a graph.
Hint: Typical orders of magnitude for α1(ω) , α2(ω) α3(ω) are 10-15s, 10-28s2, 10-42s3

A typical dispersive media is a SF57 glass plate. Its refractive index, approximated by a
Sellmeier equation fit, is given by (λ in micrometer):
1.81651371 λ2 0.428893641 λ2 1.07186278 λ2
n (λ ) = 1 + 2 + 2 + 2
λ − 0.043704198 λ − 0.0592801172 λ − 121 .419942
Compute the passage of a pulse through 2 mm of this material. Plot the incoming and
outgoing pulses on a graph and estimate their duration.

Show your solutions by diagrams and/or images.

MATLAB-help

Note the following:


w=w0+[-N:1:N]*dw0 % This is how to sample the angular frequency w0 with 2*N+1 points.
In the t-domain the corresponding sample distance is 2*pi/((2N+1)*dw0). A suitable time
interval will then be:
t=[0:2*pi/((2*N+1)*dw0):2*pi*2*N/((2N+1)*dw0)]
Fourier optics 1

Using the Huygens-Fresnel principle (a short presentation is given in appendix 1), write a
program in MATLAB which calculates the E-field in plane 2 from the E-field which has
propagated a distance L. Below you can find some help and hints. Verify your program by
simulating the diffraction pattern after two holes illuminated by a plane E-field in plane 1.
Check that you obtain the correct size of the pattern!
Below is shown a Gaussian field at plane 1. We have then a certain field amplitude in the
point O. We now introduce a mask in plane 2 so that for all areas where the contribution to
the field in O is out of phase with the contribution from the centre of plane 1. Such a mask is
called a zone plate. This will result in an increase of the field in the point O.

Before a mask is introduced After a mask is introduced

Test your program on this problem and show that the image in plane 2 will give a bright spot
in the centre (the focus of the zone plate).
(Suitable numerical values can be a wavelength of 0.633 µm (He-Ne laser), 1/e2-radius for the
plane 1: σ =500µm and L = 10 cm. Then the zone plate for a = 10µm, N=256 have
approximately the shape given in the figure). See if you can find other focal point for the zone
plate.
The observation of a bright spot (Poissons spot) after the insertion of an obstruction in the
field was observed in an experiment by Arago and was an important verification of the wave
theory presented by Fresnel 1818. See more at page 110 in the textbook (FoP).
Show your solutions by diagrams and/or images.

MATLAB-help

a) Hint: Start by creating the coordinate matrices xmat and ymat, that means the x- and
y-coordinates in all sample points. This means that it is easy to express the formulas
clearer. No for-loops are needed. Maybe like this.

N = 256; % Matix size NxN


a = 10 % sample distance in plane 1 in micrometers
xvekt = -N/2*a:a:(N/2-1)*a; % According to MATLABs definition the element with
number N/2+1 has the coordinate x=0.
yvekt = xvekt; [xmat,ymat]=meshgrid(xvekt,yvekt); % Here we create the coordinate
matrices.
Radie = sqrt(xmat.^2+ymat.^2); % This is as an example the distance from the centre to
all the sample points. Note that all operations are computed element for element
gaussfunk=exp(-radie.^2/sigma^2); % Here we create a Gauss function with the 1/e-
radius sigma. By multiply the field with this function we ensure that the field is rather low
at the edge of the object plane.
pi=3.1416;
L=100000;lambda=0.633;k=2*pi/lambda;
b=lambda*L/(N*a); % This is the sampling distance in plane 2 at the distance L from
plane 1
uvekt=-(N/2)*b:b:(N/2-1)*b; %Vektor i plan 2

b) As the number N is an even number it is not clear what is the central element in plane
1 and plane 2. This is determined by the definition of Fourier transform in
MATLAB. If the functions fft2c and ifft2c are used we will have the central element
(N/2+1, N/2+1) for all planes. To check that you can Fourier transform a constant
function. This will give a delta function in the centre.

testmat=fft2c(ones(N,N)); % The Fourier transform of a NxN-matrix with only ones.

c) Store the following files as fft2c.m and ifft2c.m in the same folder as your other m-
files.

% fft2c.m
function output=fft2c(x);
output=fftshift(fft2(fftshift(x)));

% ifft2c.m
function output=ifft2c(x);
output=fftshift(ifft2(fftshift(x)));
The following code multiplies the field with the appropriate phase factor before the field
is Fourier transformed:
x=E1.*exp(i*k*radie.^2/(2*L)); E2=fft2c(x);

d) If you want to show the intensity in one plane you can use the image operation. All
sample points with a value > 64 are shown white.

colormap(gray) % black and white colormap


image(uvekt,vvekt,intens/max(max(intens))*64) % maximum intensity is white
image(uvekt,vvekt,intens/max(max(intens))*64*10); % saturated image (good if you
want to see faint areas)

where uvekt and vvekt corresponds to xvekt and yvekt but in plane 2.

e) To construct the mask that blocks suitable areas is not so difficult. Start with the matrix
s_mat which gives the distance from every sample point in the plane 1 to the point O.
According to the Huygens-Fresnel principle the contribution to the field in the point O
from the sample points is given by

E_contrib = E_gauss.*exp(i*k*s_mat)

Now we create our mask by only use the sample points which have a contribution with
positive real part (e.g. their phase is less that 90 degrees from the positive real axis (which
means that they are approximately in phase with each other).

mask= ( real(E_contrib)>0 );

f) To construct a mask that creates two holes is not so difficult. Here we simulate a hole with
the radius 10 sample distances and positioned at a distance from the center of 5 sample
distances in the x-direction and 8 sample distances in the y-direction :
x0=5*a;y0=8*a;
radius1 = sqrt((xmat-x0).^2+(ymat-y0).^2);
mask=(radius1<10*a); E_gauss=gaussfunc;
E1=E_gauss*mask;

Fourier optics 2

Using the Huygens-Fresnel principle, write a program in MATLAB which calculates the E-
field in plane 2 from the E-field which has propagated a distance L. At the end of this paper
you can find some help and hints. Verify your program by simulating the diffraction pattern
after a rectangular opening illuminated by a plane E-field in plane 1.

Far field and grating:

Determine the condition for the distance L for which you can use the Fraunhofer
approximation, which means that we are in the far field of the field in plane 1. Test your
condition by using a periodic field (a grating) according to the hints given below. The field is
an amplitude modulated grating with slit like regions that block the light.
At which distance do you believe that you are in the far field. Compare with your analytic
condition. If we are in the far field, what happens with the intensity and its distribution on the
screen as L increases?
From an amplitude modulated grating it is easy to make a phase grating (use exp(i*..)). Make
a grating that has the phase zero or π. Is there a difference compared with the amplitude
modulated grating?
λ
Compare also your results with the grating equation sin θ q = sin θi + q ⋅ (FoP 2.4-10)
Λ
where Λ is the period of the grating and q is a positive or negative integer. Note that the
angle θ q is found from the position of the order in the second plane divided by the distance L.
Note that the grating equation does not tell anything about the intensities of the different
orders. Which orders are missing in your simulations?

Verify your solutions by diagrams and/or images.


MATLAB-help

a) to d) + the following:
g) To construct a mask that blocks suitable areas is not so difficult. Here we simulate a slit
with the width 10 sample distances placed at the distan 40 sample distances from the
center:
x0=40*a;% x position at 40*a
mask=(abs(x_mat-x0)<(10*a)); E_gauss=gaussfunc;
E1=E_gauss*mask;

h) To create a grating in the x direction where all periods consist of equally broad slits with
the value zero or 1, can be simulated with the following code:
n_period=2^4; % The number of periods in the matrix shall be 2^integer so that
each period holds an integer number of pixels
period_in_pixels=N/n_period;
borderwidth_in_pixels=period_in_pixels/2;
one_period=[1*ones(1,borderwidth_in_pixels) ,0*ones(1, borderwidth_in_pixels)];
gitter_mat=repmat(one_period,N,n_period);
colormap(gray);image(xvekt,yvekt,gitter_mat*64)
(Two make a phase grating replace 1 and 0 with appropriate values)

E1=gitter_mat;

Fourier optics 3

1. Using the Huygens-Fresnel principle, write a program in MATLAB which calculates the
E-field in plane 2 from the E-field which has propagated a distance L. At the end of this
paper you can find some help and hints. Verify your program by simulating the diffraction
pattern after a double slit illuminated by a plane E-field in plane 1.

2. Modify your Huygens-Fresnel program so that you use the two step method (TSM see
appendix 2).
Attach your code. Check your program by placing plane 2 quite near plane 1. Let the field
in plane 1 be obtained by a circular mask which blocks a region around the centre. This
field is easy to construct by using the variable radius and the symbol >. Let the sampling
distance in plane 2 be the same as in plane 1 as the field will not diverge appreciably
when propagating the short distance (choose L2 ≈ L1 ? L ). Check your program by
propagating the distance from zero and then increase the distance a little so that the field
in plane 2 deviates from that in plane1. Can you see the ringing in the transformation from
bright to dark? This is what you will se behind a laser illuminated object. (Note that
particles in the liquid filled part of your eye will also cause similar ringings on the retina.
Plane 2 is in this case the retina).
Check now your program with the following. Let the mask have a radius of 400 µm. as (a
= 10, N = 256). Check at what distance behind the mask you will see the “Poisson spot”.
This is a bright spot in the middle of the shadow behind the mask. Make an intensity plot
along the u- or v-axis that clearly shows the spot. As a check of you program it will
appear at a distance of about 1 cm behind the mask. The experimental evidence of the
Poisson spot made Fresnel the winner of a prize 1818. Poisson had claimed that Fresnel
was wrong with his wave theory as this would predict such an impossible spot.
Show your solutions by diagrams and/or images.

MATLAB-help

a) – d) and g) above + the following

i) Equation (2.31) will be in MATLAB:


u2 +v 2
a 2 − jk 2 L2   x2 + y 2 
− jk

E2 ( u , v ) = 2 e ⋅ ifft 2c  g ( ξ ,η ) ⋅ fft 2c  E1 ( x, y ) e 2 L1 
b   

The sampling distances will be a, b and c in plane 1, plane 2 and plane t respectively.
For plane t and 2 we have
λ λ
c= ⋅ L1 and b = ⋅ L2
N ⋅a N ⋅c
This means that we can also write
λ L
b= L2 = 2 ⋅ a
λ L1
N⋅ L1
N ⋅a

Thus it is possible to scale the resolution in plane 2 by changing the ratio L2 / L1 .

Fourier optics 4

1. In figure 4.4-5 (FoP page 131) there is a drawing of a typical 4f set-up. By introducing a
spatial filter in the diffraction plane it is possible to change the image obtained in the
image plane. The light distribution in the diffraction plane is approximately the Fourier
transform of the function f(x,y) in the object plane. The image obtained in the image plane
is approximately the invers Fourier transform of the frequency distribution in the
diffraction plane.
2. Write a program in MATLAB which calculates the Fourier transform of an image.
Replace L with f and skip the phase factor (write also a program which calculates the
inverse Fourier transform. Use an image of the size 256x256 pixels and check that you get
back the same image when you make the operation ℑℑ−1 ( g ( x, y )) .
3. Test your code by performing low pass and high pass filtering. Create a cross grating and
make a vertical and horizontal filtering according to figure FoP4.4-6c. Calculate the
frequency limits of the filters that you use. Are the values and results reasonable?

Verify your solutions by diagrams and/or images.

MATLAB-help

a) – d) above + the following


j) To make the Fourier transform of an image you can create it for instance in the program
Paint. Use the function IMREAD:
BILD =IMREAD(‘testbild.bmp’,’bmp’) reads the image file “testbild.bmp” to a matrix
“BILD”. Then you obtain the Fourier transform in the Fourier plane with the scale given
by the focal length.

k) To make a low pass filtering you can multiply the Fourier transform of the object with a
mask (blocking low frequency values, see e) and g)) and then make an invers Fourier
transform to obtain the filtered image.
l) If you want to show the intensity in one plane you can use the following image operation.
All sample points with a value > 64 are shown white.

colormap(gray) % black and white colormap


image(uvekt,vvekt,intens/max(max(intens))*64) % maximum intensity is white
image(uvekt,vvekt,intens/max(max(intens))*64*10) % saturated image (good if you
want to see faint areas)

where uvekt and vvekt corresponds to xvekt and yvekt but in plane 2. Similarly you can
also look at an image in plane 3.

Appendix 1
The Huygens-Fresnel principle
Here is an other presentation of how to find the wave field in the image plane generated from
an object in another plane. (See Sec. 4.1C). Note that we use here different coordinate systems
for the different planes.
We know that if we consider the diffraction obtained behind two point openings we can treat
the problem by assuming that the light comes from two point sources. Similarly if we want to
find the light distribution behind a double slit we can think of the slit as consisting of two
arrays of point sources. According to the Huygens-Fresnel principle we can assume that every
point of a given wave front is the source of secondary wavelets. Now we can use the
superposition principle so that the resulting new wave front is a superposition of all these
wavelets if we take into account both the amplitude and the phase of the individual wavelets.
The resulting diffraction pattern observed behind e.g. a double slit is the intensity distribution
resulting from the superposition of all the wavelets coming from all point sources that build
up the double slit.
Now we will look at a more general case (figure A2-1):
P la n e 1 P la n e 2
E 1 E 2

I n c id e n t fie ld
(x ,y ) S p h e r ic a l w a v e

ρ
(u ,v )
P o in t s o u r c e s

Figure A2-1 The Huygens-Fresnel principle. All points (x,y) in plane 1 emits a spherical
wave to plane 2.

Using the above described principle we have:


1) Every point (x,y) in plane 1 is the source of a spherical wave .
2) The amplitude of the wave is proportional to the amplitude of the incoming wave in
the position (x,y).
3) The phase of the wave is the same as the incoming wave in position (x,y).
We assume that the light between the two planes can be described by a scalar theory (we
neglect polarisation effects) and that the field is monochromatic.
Now we can calculate the field in the position (u,v) in the plane 2 due to the source in position
(x,y) in the original plane 1:
e jk ρ
⋅ Const ⋅ E1 ( x, y ) e jϕ ( x , y ) (2.16)
ρ
This lead to the following total field in plane 2 in the position (u,v):

e jk ρ ( x , y ,u ,v )
∫ ∫ ρ ( x, y, u, v ) ⋅ E1 ( x, y )dxdy
E2 (u, v ) = Const ⋅
−∞
(2.17)

This integral is not so easy to calculate in the general case but in the case of light propagating
at small angles to the optical axis and where the distance between plane 1 and plane 2 is long
compared to the dimensions of object space and image space (se figure A2-2), we can make
the following approximations:

y v ( u ,v )
x u
( x ,y )
ρ
z
x ,y ,u ,v < < L

L
Figure A2-2 The paraxial approximation.
The distance ρ in the denominator is approximately equal to L. For the phase in the
nominator we need to make a better approximation:
( x − u )2 ( y − v )2 ( x − u)2 ( y − v )2
ρ = L2 + ( x − u )2 + ( y − v )2 = L ⋅ 1 + + ≈ L ⋅ (1 + + )=
L2 L2 2 L2 2 L2
x 2 + y 2 u 2 + v 2 xu + yv
=L+ + −
2L 2L L
2 2
∞ x2 + y 2
1 jkL jk u 2+Lv xu + yv
jk − jk
⇒ E2 ( u , v ) =
jkL
e ⋅e ∫ ∫ E ( x, y ) ⋅ e
−∞
1
2L
e L
dxdy (2.18)

1 u2 + v2
The constant comes from Maxwell’s equations. This factor and the factor e jk 2 L are
jkL
not of interest if we are only considering the intensity of the field. In this case both factors can
be replaced by 1.
x2 + y2
Note! The integral is the Fourier transform of the function f ( x, y ) = E ( x, y ) ⋅ e jk 2L .
1 1
x2 + y2
If the distance L is very long we have e jk 2 L ≈ 1 . In this case we have the Fraunhofer
approximation where E2 is called the far field of E1. In the Fraunhofer approximation the
integral is just the Fourier transform of E1. The limit for Fraunhofer approximation is given by
N F′ = 1 (FoP 4.2-2) .

Appendix 2
The two steps method for propagation
p la n e 1 p la n e 2 p la n e t
y v η
E 1(x ,y ) x E 2(u ,v ) u
know n
E t ( ζ , ηζ )
to c a l-
c u l a te

L 1 ( lo n g )

L 2 (lo n g )
L
c a n b e s m a ll o r n e a r z e r o

Figure A3-1 The geometry used in the two steps method for propagation

We have now explained how we can obtain the field in one plane when we know the field in
another plane. The limitations in our description were that we don’t have polarisation effects
and that the distance between the planes is long compared to distances within both planes.
Now we will discuss a way to circumvent the problem with short distances between the
planes. We will introduce the two steps method of propagation. In this case we use an extra
plane, plane t. We study the field in this plane that is due to propagation from both plan 1 and
plane 2 (see figure A3-1). From Huygens-Fresnel principle we calculate the field Et in two
ways. We get:
u2 +v 2 ∞ x2 + y 2 xξ + yη
1 jkL1 jk jk − jk
Et (ξ ,η ) = e ⋅e ∫−∞ ∫ E1 ( x, y ) ⋅ e
2 L1 L1 2 L1
e dxdy
jkL1
1 4 4 4 4 44 2 4 4 4 4 4 43 (2.29)
 jk
x2 + y 2 
 
more compact = FT  E1 ( x , y )⋅e 2 L1 
 ( x , y )→( ξ ,η ) 
 

We can also calculate the field Et from the unknown field E2 in plane 2:
u2 +v 2 ∞ u2 +v 2 xξ + yη
1 jkL2 jk jk − jk
Et (ξ ,η ) = e ⋅e ∫ ∫ E ( u, v ) ⋅ e
2 L2 2 L2 L2
2 e dudv
jkL2
1 4 4 4 4 44 2 4 4 4 4 4 43
−∞
(2.30)
 jk
u2 + v2 
 
= FT  E2 ( u ,v )⋅e 2 L2 
 ( u ,v )→( ξ ,η ) 
 

Eliminating Et and moving the factors before the integrals to the right side, we get:
   
2 2 2 2
u +v ξ 2 +η 2  1 1  x +y
jk jk
 E (u, v ) ⋅ e 2 L2  L2 jk ( L1 − L2 ) jk 2  L1 − L2   E ( x , y ) ⋅ e 2 L1 
FT  2 = e e ⋅ FT  1  (2.31)
 (u, v ) → (ξ ,η )  1L1 4 4 44 2 4 4 4 43  ( x, y ) → (ξ ,η ) 
g ( ξ ,η )

The field E2(u,v) is now found by making the invers Fourier transform of both sides and then
u2 +v 2
− jk
multiply by the factor . With this method it is possible to study the propagation of
2 L2
e
light in small steps near the object.

Answers to the problems:

P1-2: 16 m 2
P1-3: b) 20˚
P1-4: θ < 4.8°
P1-6:

yb = 3.0 cm
F

b = -2.0 cm,
P1-8: f = 24 cm in air and f = 71 cm if immersed in water.
P1-9: a) 181 mm to the left of the first surface b) virtual
P1-10: f = 4 cm
P1-11: f = -10 cm, R2 = 5 cm, y(final image) ≈ -2.6 cm, upside down
P1-12 : a) 25 cm b) -10 cm
P1-13 : a) 27 cm b) 80 cm to the right of the lens. The image is upside down and enlarged 2x.
P1-14 : 750 m
P1-15 : a) -33 b) 80 cm x 1.2 m
P1-16 :
+ 2 , 5 c m + 1 , 5 c m
F o c a l p l a n e

- 2 , 5 c m - 1 , 5 c m

P1-17 : 15 cm
P1-18 : NA = 0.21, θ a = 12˚
P1-19 : NA = 1
P1-20 : 0.03 mm
P1-21 : See FoP Figure 1.3-5
P1-22 : NA = 0.49, period = 20,7 mm, length = 10,4 mm
P1-23 . Graded-index fiber: NA = 0.2104, Step-index fiber: NA = 0.2098
 d 0 dz 

 1 ∫
0 n( z ) 
P1-24: M =  
0 1 
 sin α d 
cos α d
P1-29: M =  α 
 
 −α sin α d cos α d 

P2-1: Note! In this problem the unit for the field amplitude should be mV/m instead of mV.
π
A ≈ 1.5 mV/m, T = 20ns, λ = 6 m, arg(A) = −3 ,+ ,since the wave travels to the left.
2
π
P2-2: a) A ≈ 3.5 mV/m, T = 20ns, λ = 6 m, arg(A) = , − ,since the wave travels to the
3
right. In the right region we have A ≈ 2.8 mV, T = 20 ns, λ = 4.2 m, n = 1.43, arg(A) =
1.23 ⋅ π
b) u = 3.5 ⋅ e − iπ / 3 mV/m
P2-3: a) 488,2 nm b) n = 1,698 c) 3.6 ⋅ 10−14 Hz
P2-4: a << 4 cm, θ m = 0,04 rad , N F = 2.5
P2-5: 5,4 mV
P2-6: a) 6 slits b) 0.12 mm c) 0.06 mm d) 2.5 cm
P2-7: a) 3 b) 1.4 U
P2-8: 1.16 μm
P2-9: 0.18 nm
P2-10: a) −22° b) 13.9 pm
P2-12: 128 pm and 144 pm
P2-13: 0.29 mm
2 r I / I =1
P2-16: V = ; 1 2
r +1
P2-17: a) ∆p = 2d cos θ b) 0.72˚c) n = 1.00049
P4-1: 1.5sinc( f x / 2) ⋅ sinc(3 ⋅ f y )
P4-2: a) G ( f x , f y ) ⋅ exp { − j ⋅ 2π ( f x ⋅ 0.4 + f y ⋅ 2.3)} b) ( f x ⋅ 0.4 + f y ⋅ 2.3) = 0, ±1, ±2,......
P4-3: x = λ f / c
P4-4: 4 mm
P4-6:a) x ≤ −2 : 0, −2 ≤ x ≤ −1: 4 + 2 x, −1 ≤ x ≤ 0 : 5 + 3x,0 ≤ x ≤ 1: 5 − 3x,1 ≤ x ≤ 2 : 4 − 2 x.
b) 0 ≤ x ≤ 1: x,1 ≤ x ≤ 2 : 3 x − 2, 2 ≤ x ≤ 3 : 4,3 ≤ x ≤ 4 :16 − 4 x .
3
3x x
P4-7: γ ( x ) = 1 − + when -1 < x < 1, γ ( x ) = 0 when x > 1
2 2
P4-8: a) a-f, c-d, e-i, b is the correlation of h and g
b) c is a sum of two object: a triangle and a pattern of small openings. d is the sum of
the FT of a triangle and the FT of the openings
c) First take the FT of each function. Multiply the result and take IFT of the new
result. This is the convolution of the original functions. This seams complicated but
is done quickly by a FFT(Fast Fourier Transform) algorithm.
P4-9: The lines will lean 45° and the distance will change to 1 / 2 times what it was before.
P4-10:The incoherent system is the best.
P4-11: 0.54 mm
P4-12
a ) F T o f f i g u r e F T3 - 5o af
b 3)
f ig u r e - b
O p e n in g f o r t h e
z e r o s p . f r e q u e n c y

F r o m lin e s in f ig u r e 3 - 5 a .

c) 9 mm d) f = 6.8 cm
P4-13: 108
P4-14: a) 1.52 μm b) −68,36°, −30.96°, −5.71°,18.39° and 46.92°
P6-1: If orthogonal then no retardation is obtained.
P6-3: I 0 / 2, I 0 / 4 and I0 / 8
P6-4: 72˚
P6-5: a) in the positive x-direction, left elliptical polarisation
b) in the positive y-direction, right elliptical polarisation
c) in the positive z-direction, left elliptical polarization
P6-6: a) right circularly polarized b) left circularly polarized c) linearly polarized d) right
elliptically polarized e) right elliptically polarized
2N
  π 
P6-7: a) 0 b) 0.25I 0 c) 0.42I 0 d) I 0  cos   e) I 0
  2 N 
P6-8: 56.3˚
P6-9: TE: 0.092 TM: 0.0085 For unpolarized light 0.05
P6-10: 29.7˚
P6-11: a) 56.3˚ b) 90˚ This value is independent of n
P6-12: a) 8.27 % b) at least 15 plates
P6-14: a) 13 ⋅ 10−6 m
b)

O p t i c a l a x i s O p t i c a l a x i s

α β

I 45˚ Circular 45˚ Linear

II 45˚ Circular 0˚ Linear

III 0˚ Linear 45˚ Circular

IV 90˚ Linear 20˚ Elliptic 2 0 °

Long axis sin 20°


For the elliptic polarization: = °
= tan 20° = 0.36
Short axis cos 20

P6-16: If δ = ∆n ⋅ d then I = I 0 sin ( δ / 2 )
λ0
P6-17: a) 85 mm b) 2.0 mm c) linear at 45˚ and at 90˚, I = 0
P6-18: Se FoP Figure 6.6-3, the angle between the beams is 11.4˚
P6-19: 0.0096
P6-20: See FoP Figure 6.6-3, the angle between the beams is 5.8˚
P5-1: This is a pulse of width 2 τ (exp(-1)-point). The puls is moving in the positive z-
direction and the light is linearly polarized along the x-axis.
P5-2:
Relation 1 2 3 4
a) P = ε 0 χ E − a∇ × E Yes No Yes No
b) P + aP = ε 0 E
2
No No No Yes

∂2P ∂P Yes Yes No Yes


c) a1 + a2 + P = ε0 χ E
∂t 2
∂t

{ 2 2
}
d) P = ε 0 a1 + a2 exp  − ( x + y )  E Yes No Yes No
1) Linearity 2) Dispersion 3) Spatial variation* 4) Homogeneity
*Variation of the refractive index with position
P5-3: 2 ⋅ 1016 photons/s
P5-4: a) 7.96 W/m 2 b) 77.4 V/m and 2.58 ⋅ 10−7 Vs/m2
P5-5: 2.4 kV/m
P5-6: 0.22 kV/m and 0.73 μT
P5-7: a) 5 ⋅ 1014 Hz b) 0.6 μm c) 0.39 μm d) 1.95 ⋅ 108 m/s e) 1.538
P5-8: a) 488.2 nm b) n = 1.698 c) 20 μT d) 28 kW/m 2
P5-9: a) 8.46 ⋅ 10−6 s b) 8.1 ns

Você também pode gostar