What-If-Not Conic Constructions

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 16

What-If-Not Conic Constructions

T3 International Conference
San Antonio, TX
February 25-27, 2011

Steve Phelps
Madeira High School
Cincinnati, OH
sphelps@madeiracityschools.org
The What-If-Not Game.

I would venture to say that most of us

are familiar with the paper folding construction

of the parabola, as illustrated in the pictures at

right. We begin by placing a point (the focus)

and a line (the directrix) on a piece of paper,

folding this line repeatedly to the point, resulting

in a collection of folds (an envelope) that are all tangent to a parabola with the given focus

and directrix.

What I would like for us to do is to play a

What-If-Not “game” with the parabola

construction (Brown and Walter, 2005). In their

book The Art of Problem Posing (2005), Brown

and Walter share a method of problem posing

based upon changing essential elements of a

problems. I am certainly over-simplifying their process, but by shifting the context and

challenging the givens of a problem, we ask the

question, “What if some attribute were NOT as

it were, but were instead, something else?”

The What-If-Not game I am proposing

consists of changing certain features of the

construction of a parabola, and exploring the

results. Though there are many features of this

construction that we could change, my investigations began when I applied the What-If-Not
strategy to the directrix in this construction. For example, instead of using a line in the

construction, what would happen if we were to

use a square or a circle in the role of the

directrix? What would happen if we were to use

the graph of the sine function, or even the

parabola itself, in the role of the directrix? Of

course, changing the directrix object is not the

only attribute of the construction we can change.

There are other more salient features that when made explicit, you may ask yourself, “Why

didn't I think of that?”

However, by playing this What-If-Not

game with the parabola constructions, we are

able to forge new understandings and new

connections between the parabola construction

and the constructions of the ellipse and the

hyperbola. After playing this game on my own, I

have come to view the constructions of the

parabola, ellipse, and hyperbola as nothing more than different cases of the same

construction. This, in turn, allows me to help my students when we perform these

constructions in class. Ultimately, I might have never noticed these connections without

challenging what I had taken for granted in the parabola construction. Furthermore, by playing

this What-If-Not game, I was led to other strange curves, in particular, Tschirnhausen's Cubic

(http://www-groups.dcs.st-and.ac.uk/~history/Curves/Tschirnhaus.html).
The Paper Folding Parabola Construction.

Shown at right is the paper folding

construction of the parabola modeled on the

Nspire. In the screen shot, the paper folds are

modeled by perpendicular bisectors of point A

on the directrix and point F. Point P is the

point on the perpendicular bisector that is on

the parabola. An observation about the paper

folds or perpendicular bisectors that will be

stated without proof is that each fold or line is tangent to the parabola created by the

construction at point P.

For our purposes of playing the What-If-Not game with this construction, we need to

know how to construct point P. Once we know how to construct point P, we will simply mimic

the construction for shapes other that a line.

The construction of the parabola proceeds as

follows. Begin with any point F and any line l.

Place point A on line l. Point A will drive our

locus in the construction. We next construct a

line perpendicular to l through point A, followed

by constructing the perpendicular bisector of

points F and A. The intersection point P of the

perpendicular to l and the perpendicular

bisector of F and A is the point on the parabola that will be used to create the locus.

In this construction there appears to be essentially four construction attributes that we


could change in order to play our What-If-Not game: (1) There is a point F and some object

that serves as a directrix; (2) there is a point A placed on this directrix object (this will be our

locus driver) and a perpendicular constructed to this object at point A; (3) there is

perpendicular bisector of points A and F constructed; and (4) the intersection point of the

perpendicular and the perpendicular bisector is constructed.

Before we go on, a useful exercise for

Algebra 2 students is to derive the equation of

the parabola formed by the parabola

construction. In the figure at right, the distance

from P to A is y a . The distance from F to P

is   x −0   y −a 
2 2
. Given P is on the

perpendicular bisector of F and A, these two distances are equal, so

  x −0    y −a  =y a
2 2

2 2 2
 x −0   y −a  = y a 

x 2 y 2−2aya2 =y 2 2aya 2

x 2=4ay

x2
=y
4a

which is what most Algebra 2 students would recall as the equation of a parabola.

Furthermore, as a prelude to calculus, the slope of the perpendicular bisector at any point

 
2
x x
x, is the negative reciprocal of the slope of FA, which is − .
4a 2a
Our First What-If-Not Game.

For our first What-If-Not game, we will change a very small part of the parabola

construction by asking the question, “What if

the line constructed at point A was not

perpendicular, but instead, made a 60° angle

with the directrix?” Certainly, in the spirit of

Brown and Walter (2005), the idea of

constructing a perpendicular is a taken-for-

granted assumption in the construction. In the

figure at right, we show the resulting curve with the parabola shown as the dotted curve for

comparison.

The question now to ask is, “What is this

curve?” Considering the figure at right, we

drop a perpendicular from P to the directrix l,

meeting the directrix at B. This makes BP the

distance from P to the directrix. With P being

on the perpendicular bisector of A and F makes

FP = AP. However, applying our knowledge of

special right triangles, BP will always be

 3 AP . Therefore, the distance from F to P will always be 2 times the distance from P to
2 3
the line l.

This ratio of distances is nothing more than eccentricity. With the eccentricity being

greater than 1, the curve is a hyperbola. The figures on the next page show the different
curves generated for different angle values. For

1
each, the eccentricity is .
sin

It would seem logical that if you decided

to conduct a thought experiment, if angles less

than 90° would produce a hyperbola, then

angles greater than 90° would produce an

ellipse. However, if you notice that in each of

the figures shown on the previous page and on

this page, the perpendicular bisector intersects

the locus curve in two distinct places. This

“other place” would correspond to the

supplement of the angle.

In the figure below right, you will find a

hyperbola constructed with a line making a

130° angle with the directrix. Notice that is

could also be constructed with the line making

a 50° angle. Regardless, as the angle

approaches 90°, the parabola serves as the

limiting case of the hyperbola construction.


Our Second What-If-Not Game.

For our second What-If-Not game, we

begin to explore the construction using different

objects as the directrix by asking the question,

“What if the directrix was not a line, but was a

circle?” Many of you may be familiar with the

paper folding construction of the ellipse. Some

of you may use a paper plate for this. Others

may use a large circle drawn on patty paper. Regardless, approaching this construction

though a What-If-Not point of view provides insight into the construction that can not be found

through paper folding alone.

With our directrix object being a circle

and a point F placed inside the circle, we place

point A on the circle, then we construct a

perpendicular to the circle through A. However,

a perpendicular to the circle is nothing more

than a line through the center of the circle.

Therefore, we construct a line through the center of the circle O and point A. Next, we

construct the perpendicular bisector of points A and F. The intersection point P of this line and

line OA is on the ellipse.

A proof that point P describes an ellipse goes something like this: Since P is on the

perpendicular bisector of F and A, FP = AP. Therefore, FP + PO = AP + PO = AO which is the

radius of the circle, which is constant. Therefore, the sum of the distances from P to F and

from P to O is constant, meaning that P lies on an ellipse with foci F and O.


By a simple change in the parabola construction – changing the directrix object from a

line to a circle – we gain new insight into the construction of another conic. In fact, the

parabola and the conic are constructed in exactly the same way, with the exception of the

directrix object.

Our Third What-If-Not Game.

For our third What-If-Not game, we are

going to play with this ellipse construction and

ask the question, “What if point F was not

inside the circle but outside the circle?” Indeed,

from looking at the figure to the right, it appears

to produce a hyperbola.

We begin again just as we did with the

ellipse construction, but instead, we start with a

circle and point F outside the circle. We place

point A on the circle, and construct a

perpendicular to the circle at A, which as

before, will be a line through the center of the

circle. Let point P be the intersection point of

this line and the perpendicular bisector of points A and F. Point P is on the hyperbola.

A proof of this goes something like this: Point P being on the perpendicular bisector of

A and F means that AP = PF. Noticing that | OP – PF | = | OP – AP | = | OA | which is just the

radius of the circle, which is constant. Therefore, the difference of the distances from P to F

and from P to O is constant, meaning that P lies on a hyperbola with foci F and O.
Our Fourth What-If-Not Game.

For our fourth and final game, we ask

the question, “What if the the directrix was not

a line, but was a parabola?” As shown at right,

we can barely make out the parabola

construction with a focus F and a directrix l.

Point P is the point that creates the parabola.

We are going to repeat this construction using

the parabola as the directrix object – in a way, a kind of recursive construction. This

construction produces the envelope of lines all tangent to the thicker curve known as

Tschirnhausen's Cubic.

The construction of this curve proceeds

as follows: Begin with a focus F and a line l and

a point A on the line as shown at right. Point A

will drive our locus. Construct the parabola as

before.

Using the parabola as our directrix, we

do not need to place a point on the parabola.

Instead, we will instead use point P used to construct the locus of the parabola as the point

on parabola. We next need to construct a line (line n) perpendicular to the parabola at P. To

do this, we need only construct a line perpendicular to m (the perpendicular bisector of F and

A) at P. Remember, the perpendicular bisectors are tangent to the parabola at P. We next

construct the perpendicular bisector of P and F (line k) and construct the intersection Q of this

line and line n. Point Q is equidistant from the parabola and the focus F.
Tschirnhausen's Cubic is the Negative Pedal.

By playing our What-If-Not game with the

parabola construction and treating the parabola

itself as the directrix object, we have stumbled

upon Tschirnhausen's Cubic. More formally,

Tschirnhausen's Cubic is the negative pedal of a

parabola with respect to the focus.

What is a negative pedal? Start with a

curve (a circle in the figures at right) and a fixed

point (point F in the figures at right). Place a point

on the curve (point A) and draw a line

perpendicular the line containing the fixed point F

and the point A on the curve. The envelope of

these lines as the point moves along the curve in

the negative pedal of the curve.

What are the implications of this definition?

It turns out that we have been constructing

negative pedals throughout this paper.

Furthermore, any of the paper folding

constructions of the conics is the construction of a

negative pedal. The negative pedal of a line and a

fixed point not on the line is a parabola. The negative pedal of a circle and a fixed point inside

the circle is an ellipse. The negative pedal of a circle and a fixed point outside the circle is a

hyperbola. And, the negative pedal of a parabola and its focus is Tschirhausen's Cubic.
This definition also answers another What-If-Not question for us: “What if the

perpendicular bisector were not a bisector, but any perpendicular?” As seen in the three

pictures of the ellipse construction on the previous page, changes in this attribute of the

construction have no effect on the resulting curve. The ellipse is the negative pedal of the

circle, regardless of the perpendicular we decide to use.

Writing an Equation for Tschirnhausen's Cubic.

The challenge of identifying this curve that was produced by playing this What-If-Not

game led to the next challenge of finding an equation that describes this curve. It is prudent

(and just good problem solving) to begin by finding the equation Tschirnhausen's Cubic for a

particular parabola. We can then work to find

an equation for the Cubic based on the location

of the focus F.

For the figure shown at the right, we will

find a parametric equation that describes the

locus. Since point A is driving the locus, we will

let point A have the coordinates (t, -2). Since

the parabola has a focus (0, 2) and a directrix

x2
y = -2, the parabola must have the equation y = , and point P must have the coordinates
8

 
2
t t 4
t, . The slope of line m is which makes the slope of line n equal to − . The equation
8 4 t

of line n, which we will need to find the coordinates of Q, in point-slope form using P, I

4 t2
(1) y =−  x −t   .
t 8
 
2
t t 16
To find the equation of line k, we will need the midpoint of F and P, which is , . The
2 16

t 2−16 8t
slope of the segment connecting F and P is , making the slope of k equal to − 2 .
8t t −16

Therefore, the equation of k, in point-slope form, is

(2) y =−
8t
 
t 2 −16
x −
t
2

t 2 16
16
.

We take equations (1) and (2) and solve

the resulting system shown below to find the

parametric coordinates of point Q using the

NspireCAS as shown in the screen shot at

right.

4 t2
y =−  x−t 
t 8
y =− 2
8t
t −16  
t
x− 
2
t 2 16
16

The solution, difficult to see in the screen shot, is

− t⋅ t 2 −48  3t 216


t4≠ 0 and t−4≠ 0 and t ≠ 0 and x = and y =
64 16

The graph of the curve is shown at right,

as is the graph of the parabola. The picture

captures the image at t = 5.85.


The General Equation of Tschirnhausen's Cubic.

The general equation of Tschirnhausen's

Cubic is a worthwhile exercise and an excellent

example of the appropriate use of CAS. We will

begin as shown in the figure at right. Warning:

There are a lot of holes to filled by the reader

during this derivation!

The equation of line n in point-slope form is

2
2a
(1) y =−  x−t   t .
t 4a

4⋅t⋅a
The perpendicular bisector of F and P has a slope of − 2 2 and passes through
t −4a

 
2 2
t t 4a
the midpoint of F and A, which has the coordinates , . Therefore, the equation of
2 8a

the perpendicular bisector is

(2) y =−
4⋅t⋅a
t 2 −4a2  
x−
t
2

t 24a2
4a
.

Solving the system formed by (1) and

(2) is shown in the screen shot at right.

−t⋅ t 2 −12a 2 
x t =
(3) 16a2
3t 2 4a2
y t =
8a
The screen shot of the graph at right

shows Tschirnhausen's Cubic for a focus (0,3)

and directrix y = -3. The screen shot captures

the construction at t = 7.59.

Where Did What-If-Not Lead Us?

Let us look back at what playing this

What-If-Not game did for us. First and foremost, playing this game shed new light on the

paper folding constructions of the three conics. In fact, by exploring these constructions under

the What-If-Not lens allowed us to see that the constructions were all just special cased of the

same construction. They all involve a circle (in the case of a parabola, a circle of infinite

radius) and a point not on the circle. They all involved perpendicular lines and perpendicular

bisectors. The What-If-Not game allowed us to see these constructions in a new light.

Beyond seeing the conics and their constructions in a new light, we stumbled upon a

new curve that we may never would have found if not for playing this game. By changing

some taken-for-granted assumptions about the parabola construction, we decided to let the

parabola be the directrix object. This led to a Tschirnhausen's Cubic. By exploring

Tschirnhausen's Cubic, we were lead to a new organizing principle: the negative pedal of a

curve. By exploring the conic constructions in light of this new organizing principle, we were

able to see that certain attributes of the constructions could be changed without altering the

resulting curve.

By applying the What-If-Not strategies to pose new problems to explore, we did not just

learn something new, we learned something new about something old. And it all started by

playing a game.
References

Brown, S. I. & Walter, M. I. (2005). The art of problem posing. Hillsdale, NJ: Lawrence

Erlbaum.

O'Connor, J. J. & and Robertson, E. F. (1997). “Tschirnhaus' Cubic.” From The MacTutor

History of Mathematics archive. http://www-groups.dcs.st-

and.ac.uk/~history/Curves/Tschirnhaus.html

Weisstein, Eric W. "Tschirnhausen Cubic." From MathWorld--A Wolfram Web Resource.

http://mathworld.wolfram.com/TschirnhausenCubic.html

You might also like