Midterm Practice Problems

1. Optical alignment
Aligning an optical system often requires precise control over the position of a laser beam. One
way to achieve this position control is by inserting a glass plate into the beam path, and in fact
you can purchase these types of “tweaker” devices for fiber optic system alignment (see below left).
Tilting the glass plate will cause the incoming laser beam to be displaced by a small amount, as
illustrated in the diagram (below right).

(a) Find an expression for the displacement of the laser (D) as a function of the angle of the
tweaker plate from normal incidence (θ), the index of refraction of the plate (n), and its
thickness (T ). Note: you may assume that n is real and independent of frequency in the
problems in this problem set, and you may assume the index of air is ≈ 1.
(b) Find an approximate expression for D valid in the limit of small angles θ. Compare the exact
and approximate expressions by plotting the relative displacement (D/T ) for n = 1.5 (glass)
as a function of angle θ (you are welcome to use plotting software).
(c) Find the maximum displacement for a practical device with T = 3mm, n = 1.5, and θmax =
±5 degrees. It should be clear why this is called a “tweaker” and not a “beam translator”!
(d) How much power would you lose out of your laser beam if you operated the tweaker in (c)
at θmax ? Give your answer as percentage of power lost, and assume the laser is polarized
perpendicular to the plane of incidence.
(e) Discuss the optical path taken through the plate in light of Fermat’s principle. Consider the
optical path between the point at the start of the input beam and the point at the end of
the output beam (this path is not straight!). Why is it nevertheless the path of least time?

2. SELFOC
Graded index optics play an important role in modern optical technology, from imaging to com-
munications. One useful tool is a cylindrical rod whose index of refraction decreases away from
the center (marketed under the trade name “SELFOC”) – it is used in lenses, endoscopes, and
other applications. Similar ideas are used to create graded index optical fibers. To understand the
principle of operation, consider a rectangular piece of dielectric material whose index of refraction
varies only in the y direction, with the value
2 y 2 /2
n(y) = n0 e−α (1)
Suppose that a ray of light traveling in the z direction enters the slab at y = y0 , z = 0 (i.e. entering
the slab normal to its surface), and assume that the paraxial approximation is appropriate – this
means that you can assume the angle that the ray makes with the z direction is always very small.
 (a) What is the trajectory y(z) that the light beam follows inside the slab? Sketch the beam’s
trajectory. Based on your answer, under what conditions (i.e. constraints on the given
constants α, y0 , n0 ) would you expect the paraxial approximation to hold?
(b) If the slab has length L (so the light exits at z = L), at what angle θf from the z-axis does the
ray propagate away from the slab (in air, nair ≈ 1)?1 Please use approximations valid when
the paraxial approximation holds. Sketch the trajectory taken by the light as it propagates
through and then exits the slab (i.e. illustrate your answers).
3. Brewster plates
In class, we showed that the Fresnel reflection coefficient for external reflections with polarization
in the plane of incidence (“P-polarization”) vanished for some angle of incidence θP . Similarly,
internal reflection of P-polarized light has a vanishing reflection coefficient at some other angle of
incidence θP0 6= θP . In this problem, you will consider a beam of P-polarized light incident on a
glass plate (e.g. a microscope slide).
(a) Find an expression for θP and θP0 in terms of the index of air na and the index of glass ng .
(It will be ugly.) What are the values of θP and θP0 for na = 1 and ng = 1.5? What do you
notice about the sum of these values?
(b) Show that θP and θP0 are “compatible” for light going through a parallel plate. Let me explain
what I mean: consider a beam of P-polarized light going through a glass plate (with parallel
surfaces and some thickness d) at an angle of incidence θP . Show that the reflection vanishes
for both the external surface (as the beam enters the plate, air to glass) and the internal
surface (as the beam exits the plate, glass to air).2
(c) For a glass (ng = 1.5) Brewster plate (oriented at the appropriate angle for lossless trans-
mission of P-polarized light), what fraction of incident power would be lost if the beam
was instead S-polarized? This hints at how Brewster plates are used in laser cavities to
impose polarization-dependent loss. Since only the mode with the largest round trip gain
(and lowest round trip loss) will lase, the Brewster plate thereby stabilizes the polarization
of the intra-cavity field to P-polarized (this suppresses polarization fluctuations that could
otherwise occur).

Figure 1: (Left) For a source located a distance L in front of the vertex of the mirror (which is cylindrically
symmetric about the optical axis), find the form of the surface y(x) that will perfectly collimate the outgoing
light. A random set of rays are drawn to illustrate the collimated beam.

4. Collimating with mirrors

If you look inside a flashlight, you’ll notice a curved mirror is used to direct all the light from
the filament in the forward direction. In physics language, the mirror is there to convert a point
1
If you were unable to obtain an answer to (a), you can still get partial credit for this part: suppose that you know
y(z), and find an expression for θf in terms of y(z).
2
Note that this is very useful! Transparent plates oriented at Brewster’s angle are commonly used for lossless windows
- for example, you can see commercially available Brewster windows at http://www.thorlabs.com/newgrouppage9.cfm?
objectgroup_id=919.
 source (the lamp) into a collimated beam (parallel outgoing rays). Show that the optimal surface
for such a mirror is a paraboloid, and find an equation describing the surface of a mirror that
would perfectly collimate a point source located a distance L from the closest point on the mirror
surface (see diagram below).
5. Object-image distance
Consider a convex thin lens with focal length f (in the paraxial approximation), which is being
used to create a real image of some object. By varying the distance between the object and the
lens, what is the minimum distance between the object and (real) image that you can obtain?
Describe/draw the situation in which the minimum distance occurs.
6. Numerical aperture of a hyperbolic lens

Figure 2: (Left) An illustration of a hyperbolic lens profile appropriate for collimating a source located at
x = −1. The other side of the lens would be flat. (Right) The definition of numerical aperture for a lens is
N A = nm sin θmax , where nm is the medium index and nl is the lens index.

A hyperbolic curve approaches a linear asymptote at large values of x and y. Consider the case
of a hyperbolic lens designed from a material with refractive index n to optimally collimate light
from a source S located a distance L from the front of the lens (in air, with index ≈ 1).
(a) Use physical principles to find the slope of the asymptote – what is the physical limitation
that corresponds to the asymptote? (You can check your result against a mathematical
expression.)
(b) The numerical aperture of a lens is defined as N A = nm sin θmax , where nm is the index of
the medium in which the lens sits (i.e. not the lens material itself!) and θmax is the angle
between the optical axis and the largest-angle ray captured by the lens (see p215 of the text
if this definition doesn’t make sense to you). For collecting light (e.g. in a microscope) the
numerical aperture is a figure of merit, since the light collection efficiency scales as NA2 .
What would be the limiting numerical aperture of this hyperbolic lens (i.e. the numerical
aperture in the limit you made the lens infinitely large)?
(c) If you want to design a high-NA aspheric lens with N A = 0.8 to collimate a source located
5 millimeters beyond the lens, what would be required if you used this hyperbolic design?
Specify the minimum index material which could work, and estimate the size of the lens
required. (You can give approximate values or e.g. ”the lens must be bigger than...”) Does
this seem feasible?
7. Reflections in a spoon
Find a curved convex and concave mirror – a shiny spoon will work.
 (a) Look at yourself in both surfaces (convex and concave), and describe the images you see
(note: make sure that your face is more than a focal length away from the surface – this will
happen for sure if you’re using a spoon). Try to disregard the abberations that occur from
the funny spoon shape, and focus on what is happening from a near-spherical section of the
spoon. Draw a ray diagram that qualitatively explains both images (e.g. size, orientation).
(b) Holding the spoon (or other curved mirror) concave side toward you, put your finger just in
front of the surface (maybe 1-2 millimeters away). What does the image of the front of your
finger look like? How is it different from what you see when the object is far from the spoon?
Draw ray diagrams that explain what is going on.
(c) How could you estimate the focal length of the concave side of the spoon?
8. Reflecting a flower
A spherical mirror is used to create a real image of a flower on a screen 10 cm behind the flower.
(a) Is the mirror concave or convex, and is the image on the observation screen right-side up or
up-side down? Draw a ray diagram to support your conclusion.

(b) If the absolute value of the mirror radius of curvature is |R|, what is the distance between the
mirror and the flower (the real one)? (You can give your answer in terms of the object-image
distance d = 10 cm, and you might want to make sure that your answer is positive!).
9. Matrix methods with thin lenses
For matrix problems, you are welcome to use software for matrix multiplication such as Mathe-
matica or Maple or Wolfram Alpha. If you do so, please attach a printout of your computer-aided
calculations, either referenced in your problem set or commented/annotated to explain what you
are doing. Note: printouts without any explanation will not get full credit. It is also entirely
reasonable to do all calculations by hand.
In this problem, you will apply matrix methods to a system of two thin lenses. Note that for
a thin spherical lens (two spherical surfaces in the thin-lens and paraxial approximations), the
system matrix is  
1
Athin lens = 1 − f , (2)
0 1
where f is the focal length of the lens. (You may use this result!)
(a) Using matrix techniques, find the system matrix for a lens system composed of two thin
lenses of focal length f1 (on the left – i.e. the first lens encountered by a ray) and focal
length f2 (on the right – the second lens encountered by a ray) separated by a distance d.
The lenses are in air (nair = 1). Show all work.
(b) Using matrix methods, find the front focal length of this optical system. (This means: Choose
an appropriate input or output ray, use the system matrix to find the corresponding output
or input ray, and propagate rays as needed with additional transfer matrices to find the front
focal length of this optical system.)
10. Mitigating aberrations
Below are two diagrams of a lens being used to image a monochromatic point source onto a screen.
(a) Identify which of the five “third order” aberrations could possibly occur in each setup.
(b) In which situation would you expect that lens aberrations will play a smaller role? Give
three different reasons why your choice will have lower aberrations.
(In both cases, you can assume that the drawing is approximately to scale, and the lensmakers
equation is satisfied, so an image of the point source will form on the screen).

11. Aberrations
In each of the following images, a point source (e.g. a distant star) is imaged through an imperfect
optical system. Identify which aberration is responsible for the images, and explain what else you
can know about the image (e.g whether the object is on or off the optical axis, in what direction,
how coherent the light is, etc.)

12. *Amplitude and phase modulation

Consider a sinusoidal optical disturbance (measured at fixed location) whose amplitude is period-
ically modulated by a small fraction,
EAM = E0 (1 +  cos(ωM t)) cos ω0 t, (3)
as compared to a sinusoidal field (again measured at fixed position) whose phase is periodically
modulated by a small amount
EφM = E0 cos (ω0 t +  cos ωM t). (4)
(a) Show that, in the limit that   1, EAM and EφM can both be written as the real part
of a sum of three complex exponential terms with different frequencies. Be careful – the
operation of taking the real part does not commute with multiplying components together
– i.e. Re[gf ] 6= Re[g]Re[f ]. Find the frequencies and amplitudes of the components in each
case.
(b) Sketch the three phasors which sum together to form the amplitude modulated beam (working
in the frame rotating at the carrier frequency ω0 ). Explain how the phasor evolution leads
to a modulated amplitude without affecting the phase of the wave. (This will probably be
easiest to explain if you put the tails of both small phasors at the tip of the dominant (carrier)
phasor).
(c) Sketch the three phasors which sum together to form the phase modulated beam (again,
working in the frame rotating at the carrier frequency ω0 ). Explain how the phasor evolution
leads to a change in phase without affecting the amplitude of the field (to leading order in
).