Moire Fringes and the Conic Sections

Author(s): M. R. Cullen
Source: The College Mathematics Journal, Vol. 21, No. 5 (Nov., 1990), pp. 370-378
Published by: Mathematical Association of America
Stable URL: http://www.jstor.org/stable/2686902
Accessed: 22/10/2010 14:59

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=maa.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Mathematical Association of America is collaborating with JSTOR to digitize, preserve and extend access to
The College Mathematics Journal.

http://www.jstor.org
 Moir4 Fringes and the Conic Sections
M. Ft. Cullen

Mike Cullen is Professor of Mathematics and formerchairman

at Loyola MarymountUniversityin Los Angeles, and has been a
faculty member at Grinnell College, L.S.U., and New Mexico
Highlands University.He received his Ph.D. in mathematics at
age 24 fromthe Universityof Iowa and has done post-graduate
work in biomathematics at the Universityof Washington. His
publications include research papers in classical complex analy?
sis and two undergraduate books in mathematical biology. He
takes special delight in uncovering nonstandard applications of
undergraduate mathematics.

When two geometric patterns are superimposed, additional geometric patterns

(known as moire patterns) may become visible. In the "Op Art" posters of the
1960's two superimposed screens create additional images that appear to change
rapidly with the slightest movement of the viewer [2, pp. 239-252], [3], [12]. To
create a moire or "watered" silk, two layers of ribbed silk are steam pressed
together and a third pattern emerges which resembles reflections on the surface of
a pool of water [5]. The scientific applications of the mathematical theory of
overlapping geometric patterns range from wave interference phenomena in physics
to the detection of stress patterns in metals and of aberrations in lenses [4], [5],
[11]. Two objects can be positioned precisely by arranging for moire patterns to
appear with any small displacement in alignment [8]. Particular examples will be
given later in the paper.
Figure 1 shows a family of concentric circles that has been superimposed on a
grid of horizontal lines. A third family of oval-shaped curves, the moire pattern of
the two original families of curves, can be clearly seen near the vertical bisector of
the figure. Notice that each individual oval (or moire fringe) appears as the eye
follows successive intersections of lines and circles. The moire fringes arising from

Figure 1

370 THE COLLEGE MATHEMATICS

JOURNAL
such families of lines and circles are indeed conic sections, as we will show. We
first explain exactly how these fringes are formed.

General Moire Fringes

Let f(x, y) = c denote the family of level curves of a function of two variables.
When the collection of values that c assumes is discrete, we call the family a
grating and write fix, y) = cm, where m ranges over a set of consecutive integers
and ck<ck+l. Moire fringes are formed from points of intersection of two gratings.
If, for a particular m and n, point P is an intersection point of fix, y) = am and
gix,y) = bn, we will record (m,n) as the grating coordinates of P. In Figure 2,
A = (m + 1, n-l), B = (m-1, n + 1), C = (m - 1, n-l), and D =
<m + l,n + l>. Note that A,P, and B each have m+nas the sum of their grating
coordinates. Thus a moire fringe through A,P, and B may be found by solving the
system

fix,y) =am

gix,y) =bn

= a, a fixed
subject to the condition that m + n integer. The set of all such curves,
generated from different values of a, is a moire pattern. Likewise, the difference
between the first and second coordinates of C, P, and D is m ?n, and so a second
moire fringe may be found by solving the same system subject to the condition
m ? n = a.
The moire fringe with the smaller diagonal distance (either AB or CD in Figure
2) is the fringe that the eye perceives. The relative relationship between these
diagonal distances may change as one grating slides over the other, which results in
some dramatic op art effects. Note that AB is the smaller diagonal when the angle
9 in Figure 2 is acute. Since 6 is equal to the angle between gradient vectors - V/
and Vg, it follows that cos0= -(V/)-(Vg)/(||-V/|| ||Vg||). We may therefore
conclude that the moire pattern corresponding to m +n=a is visible in those
regions for which

df dg df dg
-+-<0
dx dx dy dy

and the moire pattern corresponding to m?n=a is visible in those regions for

8(x,y) = bn+ K DJ(x,y)

= am + 1
g(x,y) = bn
= am
fix,y)

fix,y) = am_x
B

Figure 2

VOL. 21, NO. 5, NOVEMBER1990 371

which

df dg df dg
-+->0
dx dx dy dy

[11, pp. 14-17].

Circular and Line Gratings

The radii of the circles in Figure 1 increase by a fixed amount p called the pitch of
the circular grating. The distance between the horizontal lines is kp, for some
constant A > 0, and we will call this distance the pitch of the lines. Thus the
corresponding system may be written as

x2+y2 (mp)
=
y n{kp),

where m and n range over the integers. From the partials condition we see that a
fringe is visible in the upper half plane y > 0 when we impose the condition
m-n=a. Although it can be proved algebraically that this fringe is a conic (by
solving the corresponding system), it is much easier to give a demonstration that
uses the polar definition of a conic: A conic is the set of all points P such that the
ratio of the distance from P to a fixed point F to the distance from P to a fixed
line / is a constant (called the eccentricity of the conic).
Let / denote the fixed horizontal line y= -a(Xp) and let P = (0,0) be the
fixed point. If P=(m,n), then, as illustrated in Figure 3, d(P,F) = mp and
- n =
d(p, /) = n(Xp) + a(\p). Using the condition m a, it follows that the ratio

mp
d(P,F)_ ~
d(P,/) kp(n + a)

is constant and so P lies on a conic with eccentricity 1/A. A similar argument

(which uses the condition m + n = a) can be given for a fringe in the lower half
plane y < 0. We may therefore conclude:
1. If the pitch of the circular grating equals the pitch of the lines (i.e., A = 1), a
family of parabolas is formed with common focus at (0,0) and directrices y = -ap,
for a an integer.

y = n(Xp)

y= -a(\p)

Figure 3

372 THE COLLEGE MATHEMATICS

JOURNAL
 2. If the pitch of the lines is greater than that of the circles (i.e., A > 1), a family
of ellipses is formed as in Figure 1. Each ellipse has one focus at (0,0) and
eccentricity 1/A.
3. If the pitch of the lines is less than that of the circles (i.e., 0 < A < 1), a family
of hyperbolas is formed. Each hyperbola has one focus at (0,0) and eccentricity
1/A.
The third case is illustrated in Figure 4. These three cases can be presented to a
class using transparencies and an overhead projector. The images are quite
pleasing to observe and it is not necessary to align the circles and lines carefully to
see the moire fringes.

Figure 4

Two Circular Gratings

If the centers of two identical circular gratings each of pitch p are positioned at a
distance of 2c from each other, the equations of the two families of curves can be
written as
=
(x + c)2+y2 (mp)2
and
=
(x-c)2+y2 (np)2,

where m and n range over the positive integers. If Fx =(-c,0), = (c,0), and
P2
P = (jc, y), then the equations may be expressed as

d{P,Fx)=mp

d(P,F2)=np.
The condition m ?n=a is equivalent to mp ? np=ap or

d(P,Fi)-d(P9F2)=ap.

Thus the corresponding moire fringes are hyperbolas with foci at Fx and F2.
Likewise the condition m + n = a leads to a family of ellipses with foci at F{ and
F2. The hyperbolas can be clearly seen in Figure 5 as the partials condition
predicts. The ellipses are more apparent when the distance between the centers is
increased.

VOL. 21, NO. 5, NOVEMBER1990 373

 Figure 5

We can also consider circular gratings for which the distance between successive
circles is not constant. For example, in a regular Fresnel zone grating, the area of
each ring is equal to the area of the center circle. Thus, if rn denotes the radius of
the nth circle,
= irp2
irrZ+l-irrZ
or
2 2 2

where rx =p. It follows that r2 = np2 and so the family of circles can be repre?
sented by x2 +y2 = np2. If the centers of two identical Fresnel zone gratings are
positioned 2 c units apart, the corresponding system of equations can be written as

(x + c) -\-y2= mp2

(x-c) -\-y2= np2.

In this case it is easier to solve the system directly to discover moire fringes. If we
impose the condition m-n=a and subtract the two equations, we obtain 4xc =
ap2 or x = Thus, as shown in Figure 6, one set of moire fringes is
ap2/(4c).

Figure 6

374 THE COLLEGE MATHEMATICS

JOURNAL
 = a and add the
composed of vertical lines. If we impose the condition m + n
two equations, we obtain the family of circles x2-\-y2 = ? 1c2). Set r2 =
\{ap2
- 2c2) and note that = r2 +
\{ap2 r2+1 \p2. The area of each ring is therefore
\irp2. Thus the second moire pattern is itself a special Fresnel zone grating (see
Figure 7). As the centers of the original two Fresnel zone gratings are moved
apart, you will also see additional circular moire fringes that correspond to side
conditions of the form m + fin = a.

Figure 7

Physical Applications

A ripple tank creates moire fringes from the superposition of the crests of two
wave trains. A train of horizontal waves may be generated by dipping a ruler into
the surface of the water with a certain frequency, while circular waves are easily
formed by dipping a pencil into the tank. If wave trains are generated simultane?
ously from two such sources, the moire fringes correspond to positions where
double crests are formed. These appear as bright regions when the wave pattern is
projected onto the screen of the ripple tank. Excellent ripple tank photos are given
in many physics texts (see, e.g. [7]).
Moire patterns can be used for the precise linear and angular positioning of
objects. Typically, one grating is fixed to an object which is then viewed through an
identical transparent master grating. If the previously aligned object is displaced, a
moire pattern will be visible. For example, to obtain a continuous record of the
movement of a building during an earthquake, a high speed camera can record the
moire patterns formed when a line grating on a wall is viewed through a second
line grating [1].
When two identical line gratings of pitch p are superimposed, the moire pattern
is again a family of parallel lines with pitch p/(2 sin(0/2)), where 6 is the acute
angle that the lines make with one another. Note that when 6 is near zero, the
moire fringes are far apart and the slightest displacement of the gratings causes
the fringes to move rapidly (see Figure 8). Aside from alignment problems in
optics, this key property can be used to detect loose thermal-protection tiles on the
space shuttle [10], to align cloth while being heat set [9], and to detect movement
of teeth [6]!

VOL 21, NO. 5, NOVEMBER1990 375

 Figure 8

If the two line gratings are placed on a circular grating of the same pitch, three
moire patterns are visible?two families of parabolas formed from the circular
grating and each line grating, and a third family formed by the circular grating and
the linear fringes of the two line gratings (see Figures 9 and 10). When 0 < 60?, A
= > 1 and so this third family is a family of ellipses. When 0 = 60?,
l/(2sin(0/2))
the pitch of the moire pattern is again p, and a third family of parabolas is formed.
This provides us with a visual way of positioning the two line gratings at an angle
of 60?. The method can be accommodated to an angle 0O other than 60? by
=
adjusting the pitch (A \ sec 0O) of the second line grating.

Figure 9
0<6O?

376 THE COLLEGE MATHEMATICS

JOURNAL
 Figure 10
0>6O?

Projects for Students

The following moire fringe problems are well within the capabilities of most
calculus students. Students can experiment with transparencies or computers to
anticipate or verify their results:
1. Find the moire fringes that result from the two families of lines

y = mp

y = 5X + nirp)

for m and n integers and s and r fixed real numbers. (The second family of lines
has pitch (r cos 6)p, where s = tan 0.)
2. Find the moire fringes that result when the Fresnel zone grating x2+y2 = mp2
is placed on a grid of horizontal lines y = niXp).
3. The centers Fx and F2 of circular gratings of pitches p and Xp, respectively
are positioned at a distance of 2c from one another. Show that the moire fringes
may be described as {P:AdiP,Fl)?diP,F2) is constant}. When A # 1, these
curves are not conies. Experiment with circular gratings to discover the shapes of
these curves.
4. The moire pattern corresponding to m + n = a is visible in those regions for
which

df dg df dg
-+-<0.
dx dx dy dy

Use this criterion to explain why the centers of the circular gratings in Figure 5
must be pulled apart in order for the elliptical fringes to be seen. Perform a similar
analysis for the Fresnel zone plates in Figure 6.

VOL 21, NO. 5, NOVEMBER1990 377

References

1. R. Chang and C. Lin, Test the highbuildingvibrationand the deformationduringan earthquakeby

highspeed camera withmoire fringetechnique,Proceedingsof SPIE-The InternationalSocietyfor
OpticalEngineering497 (1984) 36-39.
2. M. Gardner,MartinGardner'sSixthBook of MathematicalGamesfromScientific American,W. H.
Freeman,New York, 1971.
3. G. Oster,Optical art, AppliedOptics4 (11), (1965) 1359-1369.
4. _, Moire patterns,Encyclopediaof Scienceand Technology, McGraw-Hill,1985,pp. 299-301.
5. G. Oster and Y. Nishijima,Moire patterns,Scientific American208 (5), (1963) 54-63.
6. R. L. Patersonet a!., A techniqueforreal timemeasurementof tootheruption,Proceedingsof the
NinthAnnual Conferenceof theIEEE Engineering in Medicineand BiologySociety2 (1987) 737-8.
7. PhysicalScience StudyCommittee,Physics,D.C. Heath, Lexington,MA, 1960,Chapters17 and 18.
8. C. A. Sciammarella,The moiremethod?a review,Experimental Mechanics22 (11) (1982) 418-433.
9. Thomas P. Sheahen, Opportunitiesto apply physicsto energyconservationproblemsin process
industries:examples fromtextilesand papermaking,Proceedingsof Conferenceon PhysicsCareers,
Employment and Education,State College PA, 1-3, 1978, pp. 201-206.
10. W. L. Snow, et al., Dynamic moire methodsfordetectionof loosened space shuttletiles, NASA,
Report No: NASA-TM-83205,Sept. 1981, 17 pages.
11. P. S. Theocaris, MoireFringesin StrainAnalysis,PergamonPress,Elmsford,NY, 1969.
12. W. Witschi,Moires, Computers& MathematicswithApplications12B (1-2), 1986, pp. 363-378.

(Reviewers,continuedfrompage 369.)
Hurley, U. of Connecticut;Martin Isaacs, U. of Wisconsin,Madison; Walther Janous, Innsbruck,
Austria;Charles Jepsen,GrinnellC; Elgin Johnston, Iowa State U.; Charles Jones, GrinnellC;
Zaven Karian, Denison U.; Jerome Keisler, U. of Wisconsin,Madison; Paul Kelly, Santa Barbara,
CA; James R. King, U. of Washington;L. R. King, Davidson C; Roger Kirchner, Carleton C;
David Klein, CaliforniaState U., Northridge;Benjamin G. Klein, Davidson C; Vivian Kraines,
MeredithC; Joseph B. Kruskal, AT&T Bell Laboratories;Lester H. Lange, San Jose State U.; L.
Carl Leinbach, Gettysburg C; Louis Leithold, PacificPalisades, CA; Tung-Po Lin, CaliforniaState
U., Northridge;Peter Lindstrom, North Lake C; Andy Liu, U. of Alberta; Donna Llewellyn,
Georgia Tech.; Robin Lock, St. Lawrence U.; Antonio Lopez, Jr.,Loyola U., New Orleans; Robert
Lopez, Rose-Hulman Institute;B. Pauline Lowman, Western KentuckyU.; William F. Lucas,
ClaremontGraduate School; Roy Luke, Los Angeles Pierce C; Rajindar Singh Luthar, Janesville,
WI; Frederick Luttmann, Sonoma State U.; W. A. J. Luxemburg, Caltech; Joseph Malkevitch,
York C, Jamaica,NY; Colin L. Mallows, AT&T Bell Laboratories;Stephen Maurer, Swarthmore
C; Thomas R. McCutcheon, Los Angeles Pierce C; Richard McIntosh, U. of Regina; B. Arthur
Miller, Mount Allison U.; Charles G. Moore, NorthernArizona U.; Jerry Moreno, JohnCarroll
U.; Frank Morgan, WilliamsC; Roy Myers, PennsylvaniaState U., New Kensington;Richard D.
Neidinger, Davidson C; Yves Nievergelt, Eastern WashingtonU.; Andrew M. Odlyzko, AT&T
Bell Laboratories;Morris Orzech, Queen's U., Kingston;Arnold Ostebee, St. Olaf C; Paul R.
Patten, NorthGeorgia C; Brian Peterson, San Jose State U.; Richard Pfiefer, San Jose State U.;
Stephen Pierce, San Diego State U.; John Poole, FurmanU.; Gerald Porter, U. of Pennsylvania;
Charles Rees, U. of New Orleans; Amy Rocha, San Jose State U.; Edwin Rogers, Siena C; Mary
Rosen, California State U., Northridge;Kenneth Ross, U. of Oregon; Edward Rozema, U. of
Tennessee; Dieter Ruoff, U. of Regina; Hans Sagan, North Carolina State U.; Paul Schaefer,
Honeoye, NY; Mark Schilling, CaliforniaState U., Northridge;George Schnibben, Francis Marion
C; Alan Schoenfeld, U. of California,Berkeley;William C. Schulz, NorthernArizona U.; Jon
Scott, MontgomeryC; J. Arthur Seebach, St. Olaf C; Judy Seery, AT&T Bell Laboratories;
Annie Alexander Selden, Tennessee Technical U.; Lester Senechal, Mount Holyoke C; Donald
R. Sherbert, U. of Illinois; Tatiana Deretsky Shubin, San Jose State U.; David B. Singmaster,
Polytech.of the South Bank, London; Daniel Sloughter, Furman U.; Larry Small, Los Angeles
Pierce G; Donald B. Small, Colby G; Andy Sterrett, Denison U.; J. Suck, Essen, FRG; Alvin
Swimmer,Arizona State U.; Richard Trout, RutgersU.; George Tsintsifas, Thessaloniki,Greece;
Thomas Tucker, Colgate U.; Barbara Turner, CaliforniaState U., Long Beach; Floyd Vest, U. of
NorthTexas; Michelle Wachs, U. of Miami, Coral Gables; Bert Waits, Ohio State U.; William T.
Watkins, Pan AmericanU.; Joan W. Weiss, FairfieldU.; Ronald Wenger, U. of Delaware; John
Williams, U. of Hartford;Stephen Willson, Iowa State U.; Lynn Wilson, AT&T Bell Laboratories;
JeffWitmer, Oberlin G; John W. Wrench, Jr., Frederick,MD; Paul Zorn, St. Olaf G

378 JOURNAL
THE COLLEGE MATHEMATICS