Jim Lambers

MAT 169
Fall Semester 2009-10
Lecture 32 Notes

These notes correspond to Section 9.2 in the text.

Tangents of Parametric Curves

When a curve is described by an equation of the form 𝑦 = 𝑓 (𝑥), we know that the slope of the
tangent line of the curve at the point (𝑥0 , 𝑦0 ) = (𝑥0 , 𝑓 (𝑥0 )) is given by
𝑑𝑦
= 𝑓 ′ (𝑥).
𝑑𝑥
However, if the curve is defined by parametric equations

𝑥 = 𝑓 (𝑡), 𝑦 = 𝑔(𝑡),

then we may not have a description of the curve as a function of 𝑥 in order to compute the slope
of the tangent line in this way. Instead, we apply the Chain Rule to obtain
𝑑𝑦 𝑑𝑦 𝑑𝑥
= .
𝑑𝑡 𝑑𝑥 𝑑𝑡
Solving for 𝑑𝑦/𝑑𝑥 yields
𝑑𝑦
𝑑𝑦 𝑑𝑡
= 𝑑𝑥
.
𝑑𝑥 𝑑𝑡
This allows us to express 𝑑𝑦/𝑑𝑥 as a function of the parameter 𝑡.
Example The slope of the tangent to the spiraling curve defined by

𝑥 = 𝑡 sin 𝑡, 𝑦 = 𝑡 cos 𝑡,

which is shown in the Lecture 31 notes, is given by

𝑑𝑦
𝑑𝑦 𝑑𝑡 cos 𝑡 − 𝑡 sin 𝑡
= 𝑑𝑥
= .
𝑑𝑥 𝑑𝑡
sin 𝑡 + 𝑡 cos 𝑡

At the point (𝜋/2, 0), which corresponds to 𝑡 = 𝜋/2, the slope of the tangent is

cos 𝜋2 − 𝜋2 sin 𝜋2 0− 𝜋
2 ⋅1 𝜋
𝑚= 𝜋 𝜋 𝜋 = 𝜋 =− .
sin 2 + 2 cos 2 1+ 2 ⋅0 2

1
From the point-slope form of the equation of a line, we see the equation of the tangent line of the
curve at this point is given by
𝜋( 𝜋)
𝑦−0=− 𝑥− .
2 2
□
We know that a curve defined by the equation 𝑦 = 𝑓 (𝑥) has a horizontal tangent if 𝑑𝑦/𝑑𝑥 = 0,
and a vertical tangent if 𝑓 ′ (𝑥) has a vertical asymptote. For parametric curves, we also can identify
a horizontal tangent by determining where 𝑑𝑦/𝑑𝑥 = 0. This is the case whenever 𝑑𝑦/𝑑𝑡 = 0,
provided that 𝑑𝑥/𝑑𝑡 = 0, thus excluding the case where 𝑑𝑦/𝑑𝑥 is the indeterminate form 0/0.
Similarly, the tangent line is vertical whenever 𝑑𝑥/𝑑𝑡 = 0, but 𝑑𝑦/𝑑𝑡 ∕= 0.
Example Consider the unit circle, which can be parametrized by the equations

𝑥 = cos 𝑡, 𝑦 = sin 𝑡, 0 ≤ 𝑡 < 2𝜋.

The slope of the tangent at any point on the circle is given by

𝑑𝑦
𝑑𝑦 𝑑𝑡 cos 𝑡
= 𝑑𝑥
= = − cot 𝑡.
𝑑𝑥 𝑑𝑡
− sin 𝑡

A horizontal tangent occurs whenever cos 𝑡 = 0, and sin 𝑡 ∕= 0. This is the case whenever 𝑡 = 𝜋/2
or 𝑡 = 3𝜋/2. Substituting these parameter values into the parametric equations, we see that the
circle has two horizontal tangents, at the points (0, 1) and (0, −1).
A vertical tangent occurs whenever sin 𝑡 = 0, and cos 𝑡 ∕= 0. This is the case whenever 𝑡 = 0 or
𝑡 = 𝜋. Substituting these parameter values into the parametric equations, we see that the circle
has two vertical tangents, at the points (1, 0) and (−1, 0). □
It is important to note that unlike a curve defined by 𝑦 = 𝑓 (𝑥), a point on the curve may have
more than one tangent line, because a parametric curve is allowed to intersect itself.
Example Consider the curve defined by the parametric equations

𝑥 = 𝑡2 , 𝑦 = (𝑡2 − 4) sin 𝑡.

This curve has two tangents at the point (𝜋 2 , 0). To see this, we first note that 𝑥 = 𝑡2 = 𝜋 when
√
𝑡 = ± 𝜋. Substituting these values into the equation for 𝑦, we obtain 𝑦 = 0, since sin 𝑡 = 0 when
𝑡 = ±𝜋. Therefore, there are two distinct parameter values corresponding to this point on the
curve.
Next, we must compute 𝑑𝑦/𝑑𝑥 for both values of 𝑡. We have
𝑑𝑦
𝑑𝑦 𝑑𝑡 (𝑡2 − 4) cos 𝑡 + 2𝑡 sin 𝑡 𝑡2 − 4
= 𝑑𝑥
= = sin 𝑡 + cos 𝑡.
𝑑𝑥 𝑑𝑡
2𝑡 𝑡

2
Substituting 𝑡 = −𝜋 yields
𝑑𝑦 (−𝜋)2 − 4 𝜋2 − 4
= sin(−𝜋) + cos(−𝜋) = ≈ 1.8684.
𝑑𝑥 −𝜋 𝜋
On the other hand, substituting 𝑡 = 𝜋 yields
𝑑𝑦 𝜋2 − 4 𝜋2 − 4
= sin 𝜋 + cos 𝜋 = − ≈ −1.8684.
𝑑𝑥 𝜋 𝜋
The curve is illustrated in Figure 1. □

Figure 1: Graph of the parametric curve 𝑥 = 𝑡2 , 𝑦 = (𝑡2 − 4) sin 𝑡.

In order to graph curves, it is helpful to know where the curve is concave up or concave down. For
a curve defined by 𝑦 = 𝑓 (𝑥), this is determined by computing its second derivative 𝑑2 𝑦/𝑑𝑥2 = 𝑓 ′′ (𝑥)
and checking its sign. For a parametric curve, we can compute 𝑑2 𝑦/𝑑𝑥2 in the same way as 𝑑𝑦/𝑑𝑥,
by using the Chain Rule. First, we note that
𝑑2 𝑦
( )
𝑑 𝑑𝑦
= .
𝑑𝑥2 𝑑𝑥 𝑑𝑥

3
Then, from the Chain Rule,

𝑑2 𝑦 𝑑𝑥
( ) ( )
𝑑 𝑑𝑦 𝑑 𝑑𝑦 𝑑𝑥
= = 2 .
𝑑𝑡 𝑑𝑥 𝑑𝑥 𝑑𝑥 𝑑𝑡 𝑑𝑥 𝑑𝑡

Solving for 𝑑2 𝑦/𝑑𝑥2 yields ( )

𝑑 𝑑𝑦
𝑑2 𝑦 𝑑𝑥 𝑑𝑥
= .
𝑑𝑥2 𝑑𝑥
𝑑𝑡
To use this formula, one first computes 𝑑𝑦/𝑑𝑥 in terms of 𝑑𝑦/𝑑𝑡 and 𝑑𝑥/𝑑𝑡, as described above.
Then, 𝑑𝑦/𝑑𝑥 is a function of 𝑡, which can be differentiated with respect to 𝑡 in the usual way, before
being divided by 𝑑𝑥/𝑑𝑡 to obtain 𝑑2 𝑦/𝑑𝑥2 .
It is possible to obtain a formula for 𝑑2 𝑦/𝑑𝑥2 that uses only derivatives of 𝑥 and 𝑦 with respect
to 𝑡. By applying the Quotient Rule to differentiate 𝑑𝑦/𝑑𝑥 with respect to 𝑡, we obtain

𝑑𝑥 𝑑2 𝑦 2
𝑑2 𝑦 𝑑𝑡 𝑑𝑡2 − 𝑑𝑦 𝑑 𝑥
𝑑𝑡 𝑑𝑡2
= ( 𝑑𝑥 )3 ,
𝑑𝑥2
𝑑𝑡

although the first formula may be easier to remember.

Example Consider the astroid, illustrated in the Lecture defined by the parametric equations

𝑥 = cos3 𝑡, 𝑦 = sin3 𝑡, 0 ≤ 𝑡 < 2𝜋.

This curve is illustrated in Figure 2. To determine where the curve is concave up or concave down,
we first compute 𝑑𝑦/𝑑𝑥 as a function of 𝑡:
𝑑𝑦
𝑑𝑦 𝑑𝑡 3 sin2 𝑡 cos 𝑡
= 𝑑𝑥
= = − tan 𝑡.
𝑑𝑥 𝑑𝑡
−3 cos2 𝑡 sin 𝑡

Next, we use this to compute 𝑑2 𝑦/𝑑𝑥2 :

( )
𝑑 𝑑𝑦
2
𝑑 𝑦 𝑑𝑡 𝑑𝑥 − sec2 𝑡 1
2
= 𝑑𝑥
= 2
= 4
.
𝑑𝑥 𝑑𝑡
−3 cos 𝑡 sin 𝑡 3 cos 𝑡 sin 𝑡

We conclude that the astroid is concave up whenever sin 𝑡 > 0, which is the case when 𝑦 > 0. It is
concave down whenever sin 𝑡 < 0, which is the case whenever 𝑦 < 0. □

4
 Figure 2: Graph of the astroid 𝑥 = cos3 𝑡, 𝑦 = sin3 𝑡, for 0 ≤ 𝑡 < 2𝜋.

Areas Under Parametric Curves

Recall that the area 𝐴 of the region bounded by the curve 𝑦 = 𝐹 (𝑥), the vertical lines 𝑥 = 𝑎 and
𝑥 = 𝑏, and the 𝑥-axis is given by the integral
∫ 𝑏
𝐴= 𝐹 (𝑥) 𝑑𝑥.
𝑎

Now, suppose that the curve 𝑦 = 𝐹 (𝑥) is also defined by the parametric equations 𝑥 = 𝑓 (𝑡),
𝑦 = 𝑔(𝑡), for 𝛼 ≤ 𝑡 ≤ 𝛽. Furthermore, suppose that 𝑓 (𝛼) = 𝑎 and 𝑓 (𝛼) = 𝑏. If the curve is
traversed only once as 𝑡 increases from 𝛼 to 𝛽, then the area can also be computed by integrating
with respect to 𝑡 as follows:
∫ 𝑏 ∫ 𝛽
𝐴= 𝐹 (𝑥) 𝑑𝑥 = 𝑔(𝑡)𝑓 ′ (𝑡) 𝑑𝑡.
𝑎 𝛼

5
On the other hand, if 𝑡 = 𝛼 corresponds to the right endpoint of the curve, and 𝑡 = 𝛽 corresponds
to the left endpoint, then limits of integration must be reversed:
∫ 𝛼 ∫ 𝛽
′
𝐴= 𝑔(𝑡)𝑓 (𝑡) 𝑑𝑡 = − 𝑔(𝑡)𝑓 ′ (𝑡) 𝑑𝑡.
𝛽 𝛼

Example The upper half-circle with center (0, 0) and radius 1 can be defined by the parametric
equations 𝑥 = cos 𝑡, 𝑦 = sin 𝑡, for 0 ≤ 𝑡 ≤ 𝜋. Because 𝑡 = 0 corresponds to the right endpoint of
this curve, and 𝑡 = 𝜋 corresponds to the left endpoint, the area bounded by the upper half-circle
and the 𝑥-axis is given by

sin 2𝑡 𝜋
∫ 0 ∫ 0 ∫ 𝜋 ∫ 𝜋 
2 2 1 − cos 2𝑡 𝑡 𝜋
𝐴= sin 𝑡(− sin 𝑡) 𝑑𝑡 = − sin 𝑡 𝑑𝑡 = sin 𝑡 𝑑𝑡 = 𝑑𝑡 = − = ,
𝜋 𝜋 0 0 2 2 4 0  2

which, as expected, is half of the area of the circle. □

6
Summary
∙ The slope of the tangent line of a parametric curve defined by parametric equations 𝑥 = 𝑓 (𝑡),
𝑦 = 𝑔(𝑡) is given by 𝑑𝑦/𝑑𝑥 = (𝑑𝑦/𝑑𝑡)/(𝑑𝑥/𝑑𝑡).

∙ A parametric curve has a horizontal tangent wherever 𝑑𝑦/𝑑𝑡 = 0 and 𝑑𝑥/𝑑𝑡 ∕= 0. It has a
vertical tangent wherever 𝑑𝑥/𝑑𝑡 = 0 and 𝑑𝑦/𝑑𝑡 ∕= 0.

∙ The concavity of a parametric curve at a point can be determined by computing 𝑑2 𝑦/𝑑𝑥2 =

𝑑(𝑑𝑦/𝑑𝑥)/𝑑𝑡/(𝑑𝑥/𝑑𝑡), where 𝑑𝑦/𝑑𝑡 is best represented as a function of 𝑡, not 𝑥. The curve is
concave up when 𝑑2 𝑦/𝑑𝑥2 is positive, and concave down if it is negative.

∙ A parametric curve 𝑥 = 𝑓 (𝑡), 𝑦 = 𝑔(𝑡) can have two tangents at a point (𝑥0 , 𝑦0 ) on its graph,
if there are two distinct values of the parameter 𝑡, 𝑡1 and 𝑡2 , such that 𝑓 (𝑡1 ) = 𝑓 (𝑡2 ) = 𝑥0
and 𝑔(𝑡1 ) = 𝑔(𝑡2 ) = 𝑦0 .

∙ The area of the region bounded by the parametric curve 𝑥 = 𝑓 (𝑡), 𝑦 = 𝑔(𝑡), the 𝑥-axis, the
line 𝑥 = 𝑎, and the line 𝑥 = 𝑏, where 𝑓 (𝛼) = 𝑎 and 𝑔(𝛽) = 𝑏, is the integral from 𝛼 to 𝛽 of
𝑔(𝑡)𝑓 ′ (𝑡) 𝑑𝑡, provided that the curve is only traversed once as 𝑡 increases from 𝛼 to 𝛽.