Calculus II for Engineers

My Students1

Fall 2010

1 Itis based mostly on the textbook, Smith and Minton’s Calculus Early Transcendental Functions, 3rd Ed and it
has been reorganized and retyped by Jae Lee.
Calculus II for Engineers Fall, 2010

Page 2 of 126
 C ONTENTS

9 Parametric Equations and Polar Coordinates 7

9.1 Plane Curves and Parametric Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
9.2 Calculus and Parametric Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
9.3 Arc Length and Surface Area in Parametric Equations . . . . . . . . . . . . . . . . . . . . . 15
9.4 Polar Coordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
9.4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
9.4.2 Conversion from Rectangular to Polar Coordinates . . . . . . . . . . . . . . . . . . 17
9.4.3 Conversion from Polar to Rectangular Coordinates . . . . . . . . . . . . . . . . . . 18
9.4.4 Graph of Polar Equation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
9.5 Calculus and Polar Coordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
9.6 Conic Sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
9.7 Conic Sections in Polar Coordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

10 Vectors and the Geometry of Space 25

10.1 Vectors in the Plane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
10.1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
10.1.2 Representation of Vector in terms of Component . . . . . . . . . . . . . . . . . . . 26
10.1.3 Representation of Vector in terms of Standard Basis . . . . . . . . . . . . . . . . . . 28
10.1.4 Representation of Vector in Polar Form . . . . . . . . . . . . . . . . . . . . . . . . 29
10.1.5 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
10.2 Vectors in Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
10.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
10.2.2 Vectors in R3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
10.2.3 Some Problems from Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
10.3 The Dot Product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
10.3.1 Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
10.3.2 Geometric Interpretation of Dot Product . . . . . . . . . . . . . . . . . . . . . . . . 35
10.3.3 Components and Projections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
10.3.4 Application: Work Done . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
10.4 The Cross Product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
10.4.1 Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
10.4.2 Application: Area, Distance, Volume, Torque . . . . . . . . . . . . . . . . . . . . . 40
10.5 Lines and Planes in Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
10.5.1 Lines in R3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
10.5.2 Planes in R3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
10.6 Surfaces in Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

3
Calculus II for Engineers Fall, 2010

11 Vector–Valued Functions 47
11.1 Vector–Valued Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
11.1.1 Vector–Valued Functions in Plane . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
11.1.2 Vector–Valued Functions in Space . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
11.1.3 Arc Length . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
11.2 The Calculus of Vector–Valued Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
11.2.1 Limit, Continuity and Derivative . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
11.2.2 Interpretation of Derivative . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
11.2.3 Integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
11.3 Motion In Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
11.4 Curvature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
11.4.1 Arc Length Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
11.4.2 Curvature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
11.5 Tangent and Normal Vectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
11.6 Parametric Surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

12 Functions of Several Variables and Partial Differentiation 57

12.1 Functions of Several Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
12.2 Limits nd Continuity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
12.3 Partial Derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
12.3.1 Partial Derivative . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
12.3.2 Higher–Order Partial Derivative . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
12.4 Tangent Planes And Linear Approximations . . . . . . . . . . . . . . . . . . . . . . . . . . 59
12.5 The Chain Rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
12.5.1 Chain Rule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
12.5.2 Implicit Differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
12.6 Gradient and Directional Derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
12.6.1 Functions of Two Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
12.6.2 Functions of Three Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
12.7 Extrema of Functions of Several Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
12.7.1 Definition and Second Derivatives Test . . . . . . . . . . . . . . . . . . . . . . . . 75
12.7.2 Linear Regression: Method of Least Squares . . . . . . . . . . . . . . . . . . . . . 80
12.7.3 Steepest Ascent Algorithm: Estimating Local Extrema without Critical Points . . . . 80
12.7.4 Absolute Extrema . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
12.8 Constrained Optimization and Lagrange Multipliers . . . . . . . . . . . . . . . . . . . . . . 81

13 Multiple Integrals 83
13.1 Double Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
13.1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
13.1.2 Double Integral over a Rectangle . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
13.1.3 Geometrical Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
13.1.4 Double Integral over a General Region . . . . . . . . . . . . . . . . . . . . . . . . 84
13.1.5 Changing Order of Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
13.1.6 Linearity of Double Integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
13.2 Area, Volume, and Center of Mass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
13.2.1 Area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
13.2.2 Volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
13.2.3 Moments and Center of Mass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
13.2.4 Moments of Inertia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
13.3 Double Integrals in Polar Coordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
13.3.1 Polar Coordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

Page 4 of 126
Calculus II for Engineers Fall, 2010

13.3.2 Double Integrals in Polar Coordinates . . . . . . . . . . . . . . . . . . . . . . . . . 93

13.4 Surface Area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
13.5 Triple Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
13.5.1 Formulas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
13.5.2 Applications: Volume, Mass and Center of Mass . . . . . . . . . . . . . . . . . . . 99
13.6 Cylindrical Coordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
13.6.1 Cylindrical Coordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
13.6.2 Evaluating Triple Integrals with Cylindrical Coordinates . . . . . . . . . . . . . . . 103
13.7 Spherical Coordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
13.7.1 Spherical Coordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
13.7.2 Triple Integrals in Spherical Coordinates . . . . . . . . . . . . . . . . . . . . . . . 108
13.8 Change of Variables in Multiple Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

14 Vector Calculus 111

14.1 Vector Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
14.2 Line Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
14.2.1 Line Integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
14.2.2 Application: Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
14.3 Independence of Path and Conservative Vector Fields . . . . . . . . . . . . . . . . . . . . . 113
14.4 Green’s Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
14.5 Curl and Divergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
14.5.1 Curl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
14.5.2 Divergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
14.6 Surface Integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
14.7 The Divergence Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
14.8 Stokes’ Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
14.9 Applications of Vector Calculus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116

15 Appendix: Distances between Various Objects 117

15.1 Two–Dimensional Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
15.1.1 Two Points P(x0 , y0 ) and Q(x1 , y1 ) . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
15.1.2 Line L : y = mx + b and Point P(x0 , y0 ) not lying on L . . . . . . . . . . . . . . . . . 117
15.1.3 Parallel Lines L : y = mx + b and M : y = mx + c . . . . . . . . . . . . . . . . . . . 118
15.2 Three–Dimensional Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
15.2.1 Two Points P(x0 , y0 , z0 ) and Q(x1 , y1 , z1 ) . . . . . . . . . . . . . . . . . . . . . . . . 119
15.2.2 Line L and Point Q(x0 , y0 , z0 ) not lying on L:
L : x = x1 + at, y = y1 + bt, z = z1 + ct . . . . . . . . . . . . . . . . . . . . . . . . 119
15.2.3 Parallel Lines L and M:
L : x = x0 + at, y = y0 + bt, z = z0 + ct and M : x = x1 + a′ s, y = y1 + b′ s, z = z1 + c′ s . 120
15.2.4 Skew Lines L and M (Version 1):
L : x = xL + aLt, y = yL + bLt, z = zL + cLt and
M : x = xM + aM s, y = yM + bM s, z = zM + cM s . . . . . . . . . . . . . . . . . . . . 121
15.2.5 Plane Π : ax + by + cz = d and Point P(x0 , y0 , z0 ) not lying on Π . . . . . . . . . . . 122
15.2.6 Plane Π : ax + by + cz = d and Line L : x = x0 + a′t, y = y0 + b′t, z = z0 + c′t not
intersecting with Π . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
15.2.7 Parallel Planes Π0 : ax + by + cz = d and Π1 : a′ x + b′ y + c′ z = e . . . . . . . . . . . 124
15.2.8 Appendix: Skew Lines L and M (Version 2):
L : x = xL + aLt, y = yL + bLt, z = zL + cLt and
M : x = xM + aM s, y = yM + bM s, z = zM + cM s . . . . . . . . . . . . . . . . . . . . 125

Page 5 of 126
Calculus II for Engineers Fall, 2010

Page 6 of 126
 Chapter 9

Parametric Equations and Polar Coordinates

§9.1 Plane Curves and Parametric Equations.

Given any pair of functions x(t) and y(t) defined on the same domain D, the equations

x = x(t), y = y(t)

are called parametric equations and t is called the parameter.

For each choice of t, the parametric equations specify a point (x, y) = (x(t), y(t)) in the xy–plane. The
collection of all such points is called the graph of the parametric equations. In the case where x(t) and y(t)
are continuous functions and D is an interval of the real line, the graph of the parametric equations is a curve
in the xy–plane and it is called the plane curve.

Example 9.1.1. Sketch the plane curve defined by the parametric equations x = 6−t 2 , y = t/2 for −2 ≤ t ≤ 4.

A NSWER 1 (NAIVE M ETHOD ). We make a table and based on the table we plot the points.

t x y
−2 2 −1
−1 5 −1/2
0 6 0
1 5 1/2
2 2 1
3 −3 3/2 Table 9.1: Graph of x = 6 − t 2 , y = t/2, −2 ≤ t ≤ 4
4 −10 2 □

As t increases from −2 to 4, the point (x, y) moves from (2, −1) to (−10, 2) along the plane curve. This
direction is called the orientation of the plane curve.

A NSWER 2 (E QUATION OF x AND y). As an easier way, we may consider eliminating the parameter t and
getting the equation of x and y.

y = t/2 −→ t = 2y −→ x = 6 − t 2 = 6 − (2y)2 = 6 − 4y2 −→ x = 6 − 4y2 .

Moreover, since −2 ≤ t ≤ 4, so we have −1 ≤ y = t/2 ≤ 2. Thus, we have the following equation with the
interval of y:
x = 6 − 4y2 , −1 ≤ y ≤ 2.
If we sketch its graph, we have exactly same graph as the one in Answer 1 above. □

Example 9.1.2. Sketch the plane curve defined by the parametric equations x = 2 cost, y = 2 sint, for (1)
0 ≤ t ≤ 2π and (2) 0 ≤ t ≤ π and (3) −π /2 ≤ t ≤ π /2.

7
Calculus II for Engineers Fall, 2010

Figure 9.1: Plane curves of x = 2 cost, y = 2 sint with various intervals of t

A NSWER. We eliminate t by using the identity cos2 t + sin2 t = 1 and get the equation of x and y and sketch
its graph.
x y x2 y2
cost = , sint = −→ + = cos2 t + sin2 t = 1 −→ x 2 + y2 = 22 ,
2 2 22 22
which is a circle centered at the origin of radius 2.
(1) 0 ≤ t ≤ 2π : In this case, we have

−2 ≤ x = 2 cost ≤ 2, −2 ≤ y = 2 sint ≤ 2,

which means the plane curve is traversed one whole circle on the xy–plane. Also we observe

1. as t moves from 0 to π /2, (x, y) moves from (2, 0) to (0, 2),

2. as t moves from π /2 to π , (x, y) moves from (0, 2) to (−2, 0),
3. as t moves from π to 3π /2, (x, y) moves from (−2, 0) to (0, −2),
4. as t moves from 3π /2 to 2π , (x, y) moves from (0, −2) to (2, 0).

It implies that the plane curve has the counterclockwise orientation. See the figure 9.1.
(2) 0 ≤ t ≤ π : In this case, we have

−2 ≤ x = 2 cost ≤ 2, 0 ≤ y = 2 sint ≤ 2,

which means the plane curve is traversed top semicircle on the xy–plane. Also we observe

1. as t moves from 0 to π /2, (x, y) moves from (2, 0) to (0, 2),

2. as t moves from π /2 to π , (x, y) moves from (0, 2) to (−2, 0).

It implies that the plane curve has the counterclockwise orientation. See the figure 9.1.
(3) −π /2 ≤ t ≤ π /2: In this case, we have

0 ≤ x = 2 cost ≤ 2, −2 ≤ y = 2 sint ≤ 2,

which means the plane curve is traversed right semicircle on the xy–plane. Also we observe

1. as t moves from −π /2 to 0, (x, y) moves from (0, −2) to (2, 0),

2. as t moves from 0 to π /2, (x, y) moves from (2, 0) to (0, 2).

It implies that the plane curve has the counterclockwise orientation. See the figure 9.1. □
Exercise 9.1.3. Sketch the plane curve defined by the parametric equations x = 2 sint, y = 2 cost, for 0 ≤ t ≤
2π .

Page 8 of 126
Calculus II for Engineers Fall, 2010

Example 9.1.4. Sketch the plane curve defined by the parametric equations.

1. x = 2 cost, y = 3 sint, for 0 ≤ t ≤ 2π

2. x = 2 + 4 cost, y = 3 + 4 sint, for 0 ≤ t ≤ 2π
3. x = 3 cos(2t), y = 3 sin(2t), for 0 ≤ t ≤ 2π
4. x = 3 cos(t/2), y = 3 sin(t/2), for 0 ≤ t ≤ 2π

Figure 9.2: Plane curves of Example 9.1.4

A NSWER. (1) x = 2 cost, y = 3 sint, for 0 ≤ t ≤ 2π : By eliminating t, we have

x y x 2 y2 x 2 y2
cost = , sint = −→ + = cos2 t + sin2 t = 1 −→ + = 1,
2 3 22 32 22 32
which is a ellipse passing through the points (±2, 0) and (0, ±3). In fact, the plane curve is a whole ellipse
and it has the counterclockwise orientation. See the figure 9.2.
(2) x = 2 + 4 cost, y = 3 + 4 sint, for 0 ≤ t ≤ 2π : By eliminating t, we have

x−2 y−3 (x − 2)2 (y − 3)2

cost = , sint = −→ + = cos2 t + sin2 t = 1
4 4 42 42
−→ (x − 2)2 + (y − 3)2 = 42 ,

which is a circle centered at (2, 3) of radius 4. In fact, the plane curve is a whole circle and it has the
counterclockwise orientation. See the figure 9.2.
(3) x = 3 cos(2t), y = 3 sin(2t), for 0 ≤ t ≤ 2π : By eliminating t, we have

x y x2 y2
cos(2t) = , sin(2t) = −→ 2
+ 2 = cos2 (2t) + sin2 (2t) = 1 −→ x 2 + y2 = 32 ,
3 3 3 3

Page 9 of 126
Calculus II for Engineers Fall, 2010

which is a circle centered at the origin of radius 3. In fact, the plane curve is a whole circle and it has the
counterclockwise orientation. But, the circle is traced out twice: first time for 0 ≤ t ≤ π and second time for
π ≤ t ≤ 2π . It is because of 2t in the parametric equations. See the figure 9.2.
(4) x = 3 cos(t/2), y = 3 sin(t/2), for 0 ≤ t ≤ 2π : By eliminating t, we have

x y x2 y2
cos(t/2) = , sin(t/2) = −→ + = cos2 (t/2) + sin2 (t/2) = 1 −→ x2 + y2 = 32 ,
3 3 32 32
which is a circle centered at the origin of radius 3. In fact, the plane curve has the counterclockwise orientation
and it is a top semicircle. It is because of t/2 in the parametric equations. See the figure 9.2. □
Exercise 9.1.5. Sketch the plane curve defined by the parametric equations.

1. x = cos(3t), y = sin(3t), for 0 ≤ t ≤ 2π

2. x = cos(4t), y = sin(4t), for 0 ≤ t ≤ 2π
3. x = cos(t/3), y = sin(t/3), for 0 ≤ t ≤ 2π
4. x = cos(t/4), y = sin(t/4), for 0 ≤ t ≤ 2π

Example 9.1.6 (C HALLENGEABLE). Sketch the plane curve defined by the parametric equations: (1) x =
2 tant, y = 2 sint and (2) x = 2 cost, y = 3 sin(2t).

Figure 9.3: Plane curves of x = 2 tant, y = 5 sint and x = 2 cost, y = 3 sin(2t)

A NSWER. (1) x = 2 tant, y = 5 sint: A simple observation gives

y sint x y x
sint = , = tant = −→ = sint = cost −→ 2y = 5x cost.
5 cost 2 5 2
The given equation sint = y/5 implies the following triangle.

.5

.y

. t.
√
. 52 − y2
√
It also implies cost = 52 − y2 /5 and thus we have
√ √
52 − y2 5x2 5x2
2y = 5x −→ 4y2 = x2 (52 − y2 ) −→ y2 = −→ y=± .
5 x2 + 4 x2 + 4

Page 10 of 126
Calculus II for Engineers Fall, 2010

See the figure 9.3.

(2) x = 2 cost, y = 3 sin(2t): A simple formula says sin(2t) = 2 sint cost and so we have
x x
cost = , y = 3 sin(2t) = 6 sint cost = 6 sint = 3x sint −→ y = 3x sint.
2 2
The given equation cost = x/2 implies the following triangle.

.2
√
. 22 − x2

. t.
.x
√
It also implies sint = 22 − x2 /2 and thus we have
√
22 − x 2 3 √ 2
y = 3x sint = 3x −→ y= x 2 − x2 .
2 2
See the figure 9.3. □

In the challengeable problems as above, we need to use the triangle technique and the basic formulas:

sin(A ± B) = sin A cos B ± cos A sin B cos(A ± B) = cos A cos B ∓ sin A sin B
sin(2A) = 2 sin A cos A cos(2A) = 2 cos2 A − 1 = 1 − 2 sin2 A
( ) ( )
2 A 1 − cos A 2 A 1 + cos A
sin = cos =
2 2 2 2
tan2 A + 1 = sec2 A cos2 t + sin2 t = 1

So far when parametric equations are given, we have sketched its plane curve via two ways: naive method
(table) and elimination method (finding the equation of x and y by eliminating t). Now let us consider the
other way. Suppose we have a graph of x and y or the equation of x and y. Then how can find parametric
equations having the given graph as its plane curve?

Example 9.1.7. Find parametric equations for the line segment joining the point A(1, 2) (initial point) and
the point B(4, 7) (terminal point).

A NSWER 1 (U SING E QUATION OF x AND y). The easiest way may be using the equation of x and y satisfy-
ing the conditions given in the problem, if we can find it. In this problem, it is easy to see that the equation
of the line passing through A and B is found by

5 5 1
y= (x − 1) + 2 −→ y = x+ .
3 3 3

• When we set x = t, we can easily have y = 35 t + 13 , i.e., we have the parametric equations

5 1
x = t, y= t+ .
3 3

Page 11 of 126
Calculus II for Engineers Fall, 2010

Since the orientation of the line segment is from A to B, so as t increases, the point (x, y) on the line should
move from A to B. In our parametric equations, x = t and it should start at A(1, 2). Thus t must start at 1.
Also the line segment should end at B(4, 7) and so t must end at 4. That is, the given line segment should be
represented by the parametric equations,
5 1
x = t, y= t+ , 1 ≤ t ≤ 4.
3 3
3y−1
• When we set y = t, the equation x = 5 implies x = 3t−1
5 and so we have

3t − 1
x= , y = t.
5
Since the orientation of the line segment is from A to B, so as t increases, the point (x, y) on the line should
move from A to B. In our parametric equations, y = t and it should start at A(1, 2). Thus t must start at 2.
Also the line segment should end at B(4, 7) and so t must end at 7. That is, the given line segment should be
represented by the parametric equations,
3t − 1
x= , y = t, 2 ≤ t ≤ 7.
5
There are many other parametric equations for the given line segment. □

A NSWER 2 (U SING F ORM x = a + bt, y = c + dt). Since we are asked to find the parametric equations for
the line segment, we set up
x = a + bt, y = c + dt, α ≤ t ≤ β,
where a, b, c, d and α and β are all constants to be determined. (Each equation on x and y represents a line
on the tx–plane and ty–plane, respectively.)
For the simplest case, let us choose α = 0 and β = 1 and find other constants. Since A(1, 2) is the starting
point of the line segment, x = a + bt should be 1 at t = α , i.e., at t = 0. So we have

1 = x = a + b(0) = a −→ a = 1.

Similarly, since A(1, 2) is the starting point of the line segment, y = c + dt should be 2 at t = α , i.e., at t = 0.
So we have
2 = y = c + d(0) = c −→ c = 2.
By the same argument, since B(4, 7) is the end point of the line segment, x = a + bt and y = c + dt should be
4 and 7, respectively, at t = β , i.e., t = 1. So we have

4 = x = a + b(1) = a + b = 1 + b −→ b=3
7 = y = c + d(1) = c + d = 2 + d −→ d = 5.

Therefore, we find parametric equations

x = 1 + 3t, y = 2 + 5t, 0 ≤ t ≤ 1.

Clearly there are many other parametric equations representing the given line segment. □

Remark 9.1.8. In the Answers above, we have found three parametric equations for the line segment:

3 , for 1 ≤ t ≤ 4
1. x = t, y = 5t+1
2. x = 5 , y = t, for 2 ≤ t ≤ 7
3t−1

3. x = 1 + 3t, y = 2 + 5t, for 0 ≤ t ≤ 1

Page 12 of 126
Calculus II for Engineers Fall, 2010

Figure 9.4: Portion of parabola y = x2 from A(−1, 1) to B(3, 9)

Are they same? They must be same, because they represent the same line segment. But then how can we
prove that those three parametric representations are exactly same?

Example 9.1.9. Find parametric equations for the portion of the parabola y = x2 from A(−1, 1) to B(3, 9).

A NSWER. See the figure 9.4. Since the equation of x and y is given, we just set x = t or y = t.
• Let us set x = t. Then it is easy to see y = t 2 and t moves from t = −1 to t = 3, i.e.,

x = t, y = t 2, −1 ≤ t ≤ 3.
√
• Let us set y = t. Then it is easy to see x2 = t and x = ± t . In this case, we need to be careful. We observe
that t = 0 corresponds to the origin (x, y) = (0, 0). But t = 1 corresponds to (−1, 1) and/or (1, 1).
If t = 1 corresponds to the point (1, 1), then there is no way to trace out the portion between the origin and
the point A(−1, 1).
If t = 1 corresponds to the point A(−1, 1), then it means the graph is traversed from the origin (t = 0) to the
point A (t = 1) as t moves from 0 to 1. But then in order for the graph to pass by the point B(3, 9), it should be
traced out the portion between the origin and the point A(−1, 1) again, i.e., it is traced
√ out the portion twice,
which is not acceptable. For this reason, the parametric equations y = t and x = ± t does not represent the
given portion and so it is not acceptable. □

Exercise 9.1.10. Find parametric equations for the portion of the parabola y = −x2 + 4 from A(−1, 3) to
B(3, −5).

Example 9.1.11. Sketch the plane curve defined by the parametric equations x = t 2 − 2, y = t 3 − t.

Figure 9.5: Plane curve of x = t 2 − 2, y = t 3 − t

Page 13 of 126
Calculus II for Engineers Fall, 2010

A NSWER. The equation of x and y corresponding to the parametric equations is obtained by

√
t 2 = x + 2 −→ t = ± x + 2 −→ y = t 3 − t = ±(x + 2)3/2 ∓ (x + 2)
of which graph is not easy to sketch. So we use the naive method – table. We plot the x–intercepts and
y–intercepts and connect those points.
• x–intercept: A x–intercept is a point whose y–component is 0. So we solve y = 0 for t:
y = t 3 − t = t(t 2 − 1) = t(t − 1)(t + 1) = 0 at t = 0, ±1.
At t = 0, we have (x, y) = (−2, 0). At t = ±1, we have (x, y) = (−1, 0).
• y–intercept: A y–intercept is a point whose x–component is 0. So we solve x = 0 for t:
√ √ √
x = t 2 − 2 = (t − 2 )(t + 2 ) = 0 at t = ± 2 .
√ √ √ √
At t = − 2 , we have (x, y) = (0, − 2 ). At t = 2√, we√have (x, y) = (0, 2 ). √ √
• Signs of x and y: x = t 2 − 2 is negative for t ∈ (− 2 , 2 ) and positive for t ∈ R − [− 2 , 2 ]. (R means
the set of all real numbers.) y = t 3 −t is negative for t ∈ (−∞, −1)∪(0, 1) and positive for t ∈ (−1, 0)∪(1, ∞).
Combining all pieces of the information above, we have the following table.
√ √
t − 2 −1 0 1 2
x + 0 − −1 − −2 − −1 − 0 +
√ √
y − − 2 − 0 + 0 − 0 + 2 +
Based on the table, we have the figure 9.5. □
Exercise 9.1.12. Sketch the plane curve defined by the parametric equations x = t 3 − t, y = t 4 − 5t 2 + 4.
Example 9.1.13. Suppose that a missile is fired toward your location from 500 miles away and follows a
flight path given by the parametric equations
x = 100t, y = 80t − 16t 2 , for 0 ≤ t ≤ 5. (9.1.1)
Two minutes later, you fire an interceptor missile from your location following the flight path
x = 500 − 200(t − 2), y = 80(t − 2) − 16(t − 2)2 , for 2 ≤ t ≤ 7. (9.1.2)
Determine whether the interceptor missile hits its target.

Figure 9.6: Plane curve of Attacking Missile (Red) and Interceptor (Blue)

A NSWER. When two paths meet, they should have the same x and same y component at the same time t. So
we solve the equations for t:
100t = x = 500 − 200(t − 2) −→ 100t = 500 − 200(t − 2) −→ t =3
80t − 16t 2 = y = 80(t − 2) − 16(t − 2)2 −→ t = 3.5
It implies when t = 3, two paths have the same x but different y’s (difference in y is 32). When t = 3.5, they
have the same y but different x’s (difference in x is 150). Therefore, they do not meet at all. See the figure 9.6.
(The figure in the far left–hand side shows the intersection. But in fact they do not meet as in the second
figure.) □

Page 14 of 126
Calculus II for Engineers Fall, 2010

Remark 9.1.14. For the attacking missile following the path (9.1.1), the given interceptor (9.1.2) is useless.
How can we modify (9.1.2) so that the modified interceptor can intercept the attacking missile? The modified
interceptor should follow the path as given in the third one in figure 9.6.

§9.2 Calculus and Parametric Equations.

Skip. Please read the textbook.

§9.3 Arc Length and Surface Area in Parametric Equations.

Skip. Please read the textbook.

Page 15 of 126
Calculus II for Engineers Fall, 2010

§9.4 Polar Coordinates.

□ 9.4.1 Introduction.
A point on the plane can be described in terms of the horizontal and vertical distances from the origin. This
system is referred as the system of rectangular coordinates. See the figure 9.7.

.y .y
.x
.(x, y) .(r, θ )

.y .r

.θ
. .
.
O .x .
O .x

Figure 9.7: Rectangular Coordinates. Figure 9.8: Polar Coordinates.

As an alternative description of a point on the plane, we use (r, θ ), where r is the distance from the origin
to the point and θ is the angle (in radians) measured from the positive x–axis counterclockwise to the ray
connecting the origin and the point. Here r and θ are called the polar coordinates for the point and this system
is referred as the system of polar coordinates. See the figure 9.8.
By the definition,

1. any point on the positive x-axis has θ = 0,

2. any point on the positive y-axis has θ = π /2,
3. any point on the negative x-axis has θ = π ,
4. any point on the negative y-axis has θ = 3π /2.

See the figure 9.9.

.y
.θ = π /2
.B(3, π /2)

. .
.θ = π .O .θ = 0 .D(2, π ) .O .A(2, 0) .x

.θ = 3π /2 . (−3, π /2)
C

Figure 9.9: Angles θ = 0, θ = π /2, θ = π , θ = 3π /2 Figure 9.10: Plots of points in Example 9.4.1

Example 9.4.1. Plot the points with the indicated polar coordinates (r, θ ) and determine the corresponding
rectangular coordinates (x, y) for (1) (r, θ ) = A(2, 0), (2) (r, θ ) = B(3, π /2), (3) (r, θ ) = C(−3, π /2) and (4)
(r, θ ) = (2, π ).

Page 16 of 126
Calculus II for Engineers Fall, 2010

A NSWER. (1) (r, θ ) = A(2, 0): Since θ = 0, the point A should be on the positive x–axis and by r = 2, it
should be located 2 units away from the origin. Thus, it corresponds to (x, y) = (2, 0).
(2) (r, θ ) = B(3, π /2): Since θ = π /4, the point B should be on the positive y–axis and by r = 3, it should be
located 3 units away from the origin. Thus, it corresponds to (x, y) = (0, 3).
(3) (r, θ ) = C(−3, π /2): The angle is the same as in (2), but a negative value of r indicates that the point is
located 3 united in the opposite direction. So, the point C should be on the line corresponding to θ = −π /2
(which is same as θ = 3π /2) with the distance 3, i.e., it corresponds to (x, y) = (0, −3).
(4) (r, θ ) = D(2, π ): Since θ = π , the point D should be on the negative x–axis and by r = 2, it should be
located 2 units away from the origin. Thus, it corresponds to (x, y) = (−2, 0). See the figure 9.10. □
Example 9.4.2. Find a polar coordinate representation (r, θ ) of the rectangular point (x, y) = (1, 1).

√ . When√we draw a line from the origin to the point (x, y) = (1, 1), we can measure the length
A NSWER
r = 12 + 12 = 2 of the line segment and the angle θ = π /4 from the positive√axis axis to the line segment.
Thus, the point (x, y) = (1, 1) has the polar coordinate representation
√ (r, θ ) = ( 2 , π /4). We notice that we
can specify the same point by using a negative value of r, r = − 2 , with the angle√ θ = π /4 + π = 5π /4. So
the point (x, y) = (1, 1) has another polar coordinate representation (r, θ ) = (− 2 , 5π /4). In fact, all of the
polar points are (√ π ) ( √ π )
(r, θ ) = 2 , + 2nπ and (r, θ ) = − 2 , + π + 2nπ . □
4 4
Remark 9.4.3. Each point in the plane has infinitely many polar coordinates representations. For a given
angle θ , the angles θ ± 2π , θ ± 4π and so on, all correspond to the same ray. For convenience we use the
notation θ + 2nπ (for any integer n) to represent all of these possible angles.

□ 9.4.2 Conversion from Rectangular to Polar Coordinates.

Is there a better way to convert a point given in rectangular coordinates to the one in polar coordinates? We
may use the following relationship between (x, y) and (r, θ ):
y
r 2 = x 2 + y2 , tan θ = .
x
Example 9.4.4. Find all polar coordinate representations (r, θ ) for the rectangular points (1) (x, y) = (2, 3)
and (2) (x, y) = (−3, 1).
A NSWER. (1) (x, y) = (2, 3): By the formula, we have
√ y 3
r2 = x2 + y2 = 4 + 9 = 13 −→ r = ± 13 and tan θ = = −→ θ ≈ 0.98.
x 2
√
Since (x, y) = (2, 3) is in the first quadrant, so it is basically represented by (r, θ ) = ( 13 , 0.98). Thus, all
polar representations are
(√ ) ( √ )
(r, θ ) = 13 , 0.98 + 2nπ and (r, θ ) = − 13 , 0.98 + π + 2nπ ,

where n is any integer.

(2) (x, y) = (−3, 1): By the formula, we have
√ y 1
r2 = x2 + y2 = 9 + 1 = 10 −→ r = ± 10 and tan θ = = −→ θ ≈ −0.32.
x −3
√
Since (x, y) = (−3, 1) is in the second quadrant, so it is basically represented by (r, θ ) = (− 10 , −0.32).
Thus, all polar representations are
( √ ) (√ )
(r, θ ) = − 10 , −0.32 + 2nπ and (r, θ ) = 10 , −0.32 + π + 2nπ ,

where n is any integer. □

Page 17 of 126
Calculus II for Engineers Fall, 2010

□ 9.4.3 Conversion from Polar to Rectangular Coordinates.

Now conversely, for a point (r, θ ) given in polar coordinates, how can we find its rectangular representation
(x, y)? We use the following equations:

x = r cos θ , y = r sin θ ,

which is very important and worth to be memorized.

Example 9.4.5. Find the rectangular coordinates (x, y) for the polar points (1) (r, θ ) = (3, π /6) and (2)
(r, θ ) = (−2, 3).

A NSWER. (1) (r, θ ) = (3, π /6): By the formula, we have

√
π 3 3 π 3
x = r cos θ = 3 cos = , y = r sin θ = 3 sin = .
6 2 6 2
√
Thus, the point corresponds to (x, y) = (3 3 /2, 3/2).
(2) (r, θ ) = (−2, 3): By the formula, we have

x = r cos θ = −2 cos 3 ≈ 1.98, y = r sin θ = −2 sin 3 ≈ −0.28.

Thus, the point corresponds to (x, y) = (1.98, −0.28). □

□ 9.4.4 Graph of Polar Equation.

Let us refresh our memory on a graph of an equation/a function. A graph means a set of all points satisfying
the equation or the points obtained by the function. For instance, the graph G of the line y = mx + b on the
plane means the following set

G = { (x, y) : x and y satisfies y = mx + b }

= { (x, mx + b) : x is any real number } .

Also the graph G of the function y = f (x) means the set

G = { (x, y) : x and y satisfies y = f (x) }

= { (x, f (x)) : x is any real number } .

In this topic, we sketch the graphs of the polar equations r = constant, θ = constant and r = f (θ ), some
function of θ .

Example 9.4.6. Sketch the graphs of (1) r = 2 and (2) θ = π /3.

A NSWER 1 (U SING D EFINITION ). (1) r = 2: According to the definition of the polar coordinates, the graph
G of r = 2 means the following set

G = { (r, θ ) : r and θ satisfies r = 2 }

= { (r, θ ) : the point (r, θ ) is 2 units away from the origin } .

Since there is no condition on θ , when we collect all points 2 units away from the origin, we have a circle
centered at the origin of radius 2. Thus, the graph of r = 2 is such a circle.
(2) θ = π /3: According to the definition of the polar coordinates, the graph G of θ = π /3 means the set

G = { (r, θ ) : r and θ satisfies θ = π /3 }

= { (r, θ ) : ray connecting the origin to point (r, θ ) has the angle π /3 from the positive x–axis } .

Page 18 of 126
Calculus II for Engineers Fall, 2010

.y

.r = 2 .2 .θ = π /3

.π /3
.
.−2 .2 .x

.−2

Figure 9.11: Graphs of r = 2 and θ = π /3.

Since there is no condition on r, when we√collect all points satisfying the condition, the graph of θ = π /3
√ slope tan(π /3) = 3 and passing through the origin, i.e., in rectangular coordinates,
becomes a line with the
it corresponds to y = 3 x.
We should be careful. The graph of θ = π /3 is in fact the same as the graph of θ = π /3 + π , because there
is no condition on r. For example, (r, θ ) = (−2, π /3) should be on the graph of θ = π /3 and as we know, it
is also on the graph of θ = π /3 + π . See the figure 9.11. □
A NSWER 2 (U SING E QUATION OF x AND y). Let us use the formula to convert the polar equation to the
rectangular equation.
(1) r = 2:
x 2 + y2 = r 2 = 22 −→ x 2 + y2 = 22 ,
which is a circle centered at the origin of radius 2.
(2) θ = π /3:
y π √ √
= tan θ = tan = 3 −→ y = 3 x,
x 3
√
which is a straight line with the slope 3 and passing through the origin. See the figure 9.11. □
As we can see, using the equation of x and y is much easier to sketch the graph of the given polar equations.
So whenever the rectangular representation of the polar equation is available and it is easy to sketch, we will
follow this method.

Example 9.4.7. Find the polar equation(s) corresponding to the hyperbola x2 − y2 = 9.

Figure 9.12: Hyperbola x2 − y2 = 9

A NSWER. The formula implies

x = r cos θ , y = r sin θ

Page 19 of 126
Calculus II for Engineers Fall, 2010

( )
−→ 9 = x2 − y2 = r2 cos2 θ − r2 sin2 θ = r2 cos2 θ − sin2 θ = r2 cos(2θ )
√
−→ r2 cos(2θ ) = 9 −→ r2 = 9 sec(2θ ) −→ r = ±3 sec(2θ ) .
√ √
The graph of the polar equations r = ±3 sec(2θ ) is the hyperbola x2 − y2 = 9. Specifically, √r = 3 sec(2θ )
corresponds to the right branch of the hyperbola,√because the points (r, θ ) satisfying r = 3 sec(2θ ) should
correspond to positive x = r cos θ > 0. r = √−3 sec(2θ ) corresponds to the left branch of the hyperbola,
because the points (r, θ ) satisfying r = −3 sec(2θ ) should correspond to negative x = r cos θ < 0. See the
figure 9.12. □

Example 9.4.8. Sketch the graph of the polar equation r = sin θ .

Figure 9.13: Graph of r = sin θ

A NSWER (U SING E QUATION OF x AND y). By the formula, we have

r = sin θ −→ r2 = r sin θ −→ x2 + y2 = r2 = r sin θ = y −→ x2 + y2 − y = 0.

We recall the formula on the completing the square: a quadratic function f (x) = ax2 +bx+c has the following
vertex form, ( )
b 2 b2 − 4ac
2
f (x) = ax + bx + c = a x + − .
a 4a
Using the form, we can rewrite the result
( )
1 2 1
x +y −y = 0
2 2
−→ x + y−
2
= 2,
2 2

of which graph is a circle centered at the point (0, 1/2) of radius 1/2. Therefore, the graph of r = sin θ is the
circle centered at the point (0, 1/2) of radius 1/2. See the figure 9.13. □

Example 9.4.9 (Limaçon1 ). Sketch the graph of the polar equation r = 3 + 2 cos θ .

A NSWER 1 (U SING R ECTANGULAR E QUATION ). Just as we did in Example 9.4.6, let us try to find the
corresponding rectangular equation of x and y. By the formulas, we have

r = 3 + 2 cos θ −→ r2 = 3r + 2r cos θ
1 Inmathematics, limaçons (pronounced with a soft c), also known as limaçons of Pascal, are heart–shaped mathematical
curves. The cardioid is a special case, with a cusp. They arise in polar coordinates in the form r = a + b cos θ which in Cartesian
coordinates is (x2 + y2 − bx)2 = a2 (x2 + y2 ).

Page 20 of 126
Calculus II for Engineers Fall, 2010

Figure 9.14: Graph of r = 3 + 2 cos θ

√
−→ x2 + y2 = r2 = 3r + 2r cos θ = ±3 x2 + y2 + 2x
√ ( 2 )2
−→ x2 + y2 − 2x = ±3 x2 + y2 −→ x + y2 − 2x = 9(x2 + y2 )

of which graph is not easy to sketch. It is because of the constant term 3. Without 3, r = 2 cos θ is easily
converted to the rectangular equation, which is (x − 1)2 + y2 = 1. For this reason, we do not follow this way
to sketch the graph. □
A NSWER 2 (NAIVE M ETHOD – TABLE ). We make the following table and based on the table, we sketch
the graph.
Interval cos θ r = 3 + 2 cos θ
[0, π /2] Decreases from 1 to 0 Decreases from 5 to 3
[π /2, π ] Decreases from 0 to −1 Decreases from 3 to 1
[π , 3π /2] Increases from −1 to 0 Increases from 1 to 3
[3π /2, 2π ] Increases from 0 to 1 Increases from 3 to 5
See the figure 9.14. □

Example 9.4.10 (Cardioid). Sketch the graph of the polar equation r = 2 − 2 sin θ .

A NSWER. We make the following table and based on the table, we sketch the graph.
Interval sin θ r = 2 − 2 sin θ
[0, π /2] Increases from 0 to 1 Decreases from 2 to 0
[π /2, π ] Decreases from 1 to 0 Increases from 0 to 2
[π , 3π /2] Decreases from 0 to −1 Increases from 2 to 4
[3π /2, 2π ] Increases from −1 to 0 Decreases from 4 to 2
See the figure 9.15. □

Page 21 of 126
Calculus II for Engineers Fall, 2010

Figure 9.15: Graph of r = 2 − 2 sin θ

Exercise 9.4.11. Sketch the graph of the polar equation r = 1 − 2 sin θ .

Example 9.4.12 (Four–Leaf Rose). Sketch the graph of the polar equation r = sin(2θ ).
A NSWER. We make the following table and based on the table, we sketch the graph. Since sin(2θ ) = 0 at
2θ = 0, π , 2π , 3π , 4π , i.e., θ = 0, π /2, π , 3π /2, 2π . So the graph of sin(2θ ) is given as in the figure 9.16. The
graph shows 8 intervals as in the given table below.

Interval r = sin(2θ ) Quadrant

[0, π /4] Increases from 0 to 1 First
[π /4, π /2] Decreases from 1 to 0 First
[π /2, 3π /4] Decreases from 0 to −1 Fourth
[3π /4, π ] Increases from −1 to 0 Fourth
[π , 5π /4] Increases from 0 to 1 Third
[5π /4, 3π /2] Decreases from 1 to 0 Third
[3π /2, 7π /4] Decreases from 0 to −1 Second
[7π /4, 2π ] Increases from −1 to 0 Second
Figure 9.16: Graph of r = sin(2θ )
See the figure 9.17. □

§9.5 Calculus and Polar Coordinates.

Skip. Please read the textbook.

§9.6 Conic Sections.

Skip. Please read the textbook.

§9.7 Conic Sections in Polar Coordinates.

Skip. Please read the textbook.

Page 22 of 126
Calculus II for Engineers Fall, 2010

Figure 9.17: Graph of r = sin(2θ )

Page 23 of 126
Calculus II for Engineers Fall, 2010

Page 24 of 126
 Chapter 10

Vectors and the Geometry of Space

§10.1 Vectors in the Plane.

□ 10.1.1 Introduction.
The term vector is used to indicate a quantity (such as displacement or velocity or force) that has both mag-
nitude and direction. Vectors only impart magnitude and direction. Vectors don’t impart any information
about where the quantity is applied. This is an important idea to always remember in the study of vectors.
In a graphical sense vectors are represented by directed line segments. The length of the line segment is the
magnitude of the vector and the direction of the line segment is the direction of the vector. However, because
vectors don’t impart any information about where the quantity is applied, any directed line segment with
the same length and direction will represent the same vector. We denote a vector by printing a letter in
boldface (v) or by putting an arrow above the letter (→−
v ). In the lecture notes, usually I will use the arrowed
vector.
The zero vector, denoted by 0, has length 0. It is the only vector with no specific direction.
Suppose a particle moves along a line segment from point A to point B. The corresponding displacement
−→
vector has initial point A (the tail) and terminal point B (the tip) and we indicate this vector by writing AB.
−→ −→ −
→
The length of the vector AB (referred as the magnitude of the vector AB) is denoted by ∥AB∥.
When discussing vectors, we refer to real numbers as scalars. It is very important that you begin now to
carefully distinguish between vector and scalar quantities.
(0) Equality: two vectors u and v are identical, i.e., u = v if and only if u and v have the same length and
same direction.
(1) Addition: Consider two vectors u and v, where u has the initial point A and the terminal point B and v
−→ −→
has the initial point B and the terminal point C. Simply, u = AB and v = BC. Then the addition or the sum of
u and v is defined as follows:
−
→ − → − →
u + v = AB + BC = AC.
In other words, if we add two vectors, the result is also a vector (called the resultant vector) and it is made
with the initial point of one vector and the terminal point of the other vector. Just like numbers, we can switch
the order, i.e.,
−
→ − → − → − → − →
v + u = BC + AB = AB + BC = AC.

Figure 10.1: Sum of Two Vectors

We summarize as follows:

Definition 10.1.1. If u and v are vectors positioned so the initial point of u is at the terminal point of v, then
the sum u + v is the vector from the initial point of u to the terminal point of v.

This definition is sometimes called the Triangle Law. See the figure 10.1.
(2) Multiplication: We will study two important products on vectors, which are called the dot product and
cross product. Since they are very important, we will study them later in Section 10.3 and 10.4. Here, we

25
Calculus II for Engineers Fall, 2010

discuss the scalar multiplication, i.e., multiplying a vector by a scalar (= a real number). If we multiply a
vector u by a scalar 3, then the result is 3u, which is also a vector. (We don’t write anything between 3 and
u.) The vector 3u has the same direction as the vector u, but the length of the vector 3u is three time longer
than the length of the original vector u. However, if we multiply a vector v by a scalar −1, then the resulting
vector −v has the opposite direction as the original vector v and same length as the vector v.

Figure 10.2: Difference of Two Vectors

(3) Subtraction: The subtraction u − v can be understood as the sum of u and −v, where −v is the vector
multiplied by a scalar −1. Therefore u − v can be handled by the addition and multiplication above. See the
figure 10.2.
(4) Division: There is no division between vectors. We cannot divide a vector by a vector. We cannot divide
a scalar (i.e., a real number) by a vector.

Figure 10.3: Regular Hexagon

Example 10.1.2. A regular hexagon1 is given as in the figure 10.3. When two vectors a and b are given,
express other vectors in terms of a and b.

□ 10.1.2 Representation of Vector in terms of Component.

Since the location of the initial point is irrelevant, we typically draw vectors with their initial point located at
the origin. Such a vector is called a position vector. For the position vector a with initial point at the origin
O(0, 0) and terminal point at the point A(a1 , a2 ), we denote the vector by
−→
a = OA = ⟨a1 , a2 ⟩ .

We call a1 and a2 the components of the vector a; a1 is the first component and a2 is the second component.
Be careful to distinguish between the point (a1 , a2 ) and the position vector ⟨a1 , a2 ⟩. The magnitude of the
position vector a follows directly from the Pythagorean Theorem. We have
√
∥a∥ = ∥ ⟨a1 , a2 ⟩ ∥ = a21 + a22 .

For two position vectors a = ⟨a1 , a2 ⟩ and b = ⟨b1 , b2 ⟩, the operations are defined as follows:

1. Zero Vector: The zero vector 0 is expressed as 0 = ⟨0, 0⟩.

1a hexagon with six sides of equal length

Page 26 of 126
Calculus II for Engineers Fall, 2010

2. Equality: a = b is equivalent to say that a1 = b1 and a2 = b2 .

3. Sum: a + b = ⟨a1 , a2 ⟩ + ⟨b1 , b2 ⟩ = ⟨a1 + b1 , a2 + b2 ⟩.
4. Scalar Multiplication with scalar c (= real number): ca = c ⟨a1 , a2 ⟩ = ⟨ca1 , ca2 ⟩.
5. Difference: a − b = ⟨a1 , a2 ⟩ − ⟨b1 , b2 ⟩ = ⟨a1 − b1 , a2 − b2 ⟩.

As we can see above, all the operations above are done componentwise.
Observations:

1. If a vector a satisfies ∥a∥ = 0, then the vector a should be the zero vector, i.e., a = 0.
2. For a scalar c and a vector a = ⟨a1 , a2 ⟩, we observe
√
∥ca∥ = ∥c ⟨a1 , a2 ⟩ ∥ = ∥ ⟨ca1 , ca2 ⟩ ∥ = (ca1 )2 + (ca2 )2
√ √ √ √
2 2 2
= c (a1 + a2 ) = c 2 a1 + a2 = |c| a21 + a22 = |c|∥a∥.
2 2

That is, simply we deduce

∥ca∥ = |c|∥a∥.
Please do not forget the absolute of c.
3. When c = −1 in the observation 2 above, we get

∥ − a∥ = | − 1|∥a∥ = ∥a∥, i.e., ∥ − a∥ = ∥a∥.

Example 10.1.3. For vectors a = ⟨3, 2⟩ and b = ⟨−2, 5⟩, compute (1) a + b, (2) 4a, (3) 3a + 2b, (4) 2a − 3b
and (5) ∥2a − 3b∥.

Definition 10.1.4. Two vectors having the same direction or opposite direction are called parallel.

For two vectors a and ca, where c is a nonzero scalar, they are parallel. When c is positive, a and ca have
the same direction. When c is negative, a and ca have the opposite direction. Hence, we deduce that two
(nonzero) position vectors a and b are parallel if and only if b = ca, for some scalar c. In this event, we say
that b is a scalar multiple of a.

Example 10.1.5. Determine whether the given pair of vectors is parallel: (1) a = ⟨2, 3⟩ and b = ⟨4, 5⟩ and
(2) a = ⟨2, 3⟩ and b = ⟨−4, −6⟩.

Example 10.1.6. Find the constants a and b such that two vectors u = ⟨2a, a⟩ and v = ⟨b, 3⟩ are parallel and
the magnitude of a is 5.

We denote that the set of all position vectors in two–dimensional space by

V2 = { ⟨x, y⟩ : x, y ∈ R } .

Theorem 10.1.7. For any vectors u, v and w in V2 , and any scalars a and b in R, the following hold:

1. Commutativity: u + v = v + u
2. Associativity: u + (v + w) = (u + v) + w
3. Zero Vector: u + 0 = u
4. Additive Inverse: u + (−u) = 0
5. Distributive Law: a(u + v) = au + av
6. Distributive Law: (a + b)u = au + bu
7. Multiplication by 1: (1)u = u
8. Multiplication by 0: (0)u = 0

Page 27 of 126
Calculus II for Engineers Fall, 2010

Figure 10.4: Position Vectors

−→ −→
For any two points A(a1 , a2 ) and B(b1 , b2 ), we can consider two position vectors OA = ⟨a1 , a2 ⟩ and OB =
⟨b1 , b2 ⟩. By the Triangle Law or definition of sum of vectors 10.1.1, we have
−→ −→ −→
AB = OB − OA = ⟨b1 , b2 ⟩ − ⟨a1 , a2 ⟩ = ⟨b1 − a1 , b2 − a2 ⟩ ,
−
→
where the last equality comes by the component wise operations. That is, the vector AB is obtained by
−
→
AB = ⟨b1 − a1 , b2 − a2 ⟩ .
See the figure 10.4.
Example 10.1.8. Find the vector represented by the directed line segment with initial point A(2, −3) and
terminal point B(−2, 1).

□ 10.1.3 Representation of Vector in terms of Standard Basis.

We recall that any point (a, b) on the plane can be represented as
(a, b) = (a, 0) + (0, b) = a(1, 0) + b(0, 1),
where (1, 0) and (0, 1) work as the base. Thanks to the Theorem 10.1.7 above, we can express a vector
v = ⟨a, b⟩ in the exactly same way:
v = ⟨a, b⟩ = ⟨a, 0⟩ + ⟨0, b⟩ = a ⟨1, 0⟩ + b ⟨0, 1⟩ ,
where ⟨1, 0⟩ and ⟨0, 1⟩ are called the standard basis vectors. Since they are very important, we denote them
as follows:
i = ⟨1, 0⟩ , and j = ⟨0, 1⟩ .
Hence, the vector v = ⟨a, b⟩ is represented by
v = ⟨a, b⟩ = a ⟨1, 0⟩ + b ⟨0, 1⟩ = ai + bj, i.e., v = ai + bj.
Here we call a and b the horizontal and vertical components of v, respectively.
Definition 10.1.9. Any vector u with the magnitude 1, i.e., ∥u∥ = 1, is called a unit vector.
We observe that ∥i∥ = 1 and ∥j∥ = 1. That is, the standard basis vectors are unit vectors. Unit vectors are
very important and for any nonzero vectors, we can always find a unit vector with the same direction.
Theorem 10.1.10. For any nonzero position vectors v, a unit vector having the same direction as v is given
by
1 v
u= v= .
∥v∥ ∥v∥
Since 1/∥v∥ is a scalar, (1/∥v∥)v and v have the same direction and 1/∥v∥ makes the magnitude of (1/∥v∥)v
to be 1. (A vector’s magnitude is sometimes called its norm.) The process of dividing a nonzero vector by its
magnitude is sometimes called normalization. We will see this process again later in Chapter 12 Functions
of Several Variables and Partial Differentiation.
Example 10.1.11. Find the unit vector in the direction of the vector v = 2i − j.

Page 28 of 126
Calculus II for Engineers Fall, 2010

□ 10.1.4 Representation of Vector in Polar Form.

Skip. Please read the textbook.

□ 10.1.5 Applications.
Whenever two or more forces are acting on an object, the net force/resultant force acting on the object is
simply the sum of all of the force vectors.
Example 10.1.12. At a certain point during a jump, there are two principal forces acting on a sky diver:
gravity exerting a force of 180 pounds straight down and air resistance exerting a force of 180 pounds up and
30 pounds to the right. What is the net force acting on the sky diver?
Example 10.1.13. An airplane has an airspeed of 400 mph. Suppose that the wind velocity is given by the
vector w = ⟨20, 30⟩. In what direction should the airplane head in order to fly due west (i.e., in the direction
of the unit vector −i = ⟨−1, 0⟩)?

Figure 10.5: Hanging Crate

Example 10.1.14. Two ropes are attached to a large crate. See the figure 10.5. Suppose that that rope A
exerts a force of ⟨−164, 115⟩ pounds on the crate and rope B exerts a force of ⟨177, 177⟩ pounds on the crate.
If the crate weighs 275 pounds, what is the net force acting on the crate? Based on your answer, which way
will the crate move?
Example 10.1.15. The thrust of an airplane’s engines produces a speed of 600 mph in still air. The wind
velocity is given by ⟨−30, 60⟩. In what direction should the airplane head to fly due west?

Figure 10.6: Velocities Around Airplane

Page 29 of 126
Calculus II for Engineers Fall, 2010

A NSWER. See the figure 10.6. Let us define velocities (vectors).

1. a = ⟨x, y⟩: Velocity of Thrust

2. w = ⟨−30, 60⟩: Wind Velocity
3. ⟨0, −c⟩: Net Force, where c is a positive real number.

Then, the net force comes from the sum of all forces:

⟨0, −c⟩ = a + w = ⟨x, y⟩ + ⟨−30, 60⟩ = ⟨x − 30, y + 60⟩ −→ x = 30, y = −c − 60.

√
By the given information, ∥a∥ = x2 + y2 = 600:

6002 = x2 + y2 = 302 + (−c − 60)2

√ √
c = −60 ± 6002 − 302 = −60 ± 30 399 = 539.25 and − 659.25.
√
Since c should be positive, so we choose c = −60 + 30 399 and thus
√ ⟨ √ ⟩ ⟨ √ ⟩
x = 30, y = −c − 60 = −30 399 −→ a = 30, −30 399 = 30 1, − 399
( ) √ ( √ )
−1 y −1 −30 399 −1
tan = tan = tan − 399 = −1.52078 rad = −87.134◦ . □
x 30

Page 30 of 126
Calculus II for Engineers Fall, 2010

§10.2 Vectors in Space.

□ 10.2.1 Introduction.
We now extend several ideas from the two–dimensional Euclidean space, R2 to the three–dimensional Eu-
clidean space, R3 . The three–dimensional Euclidean space is formed by adding an perpendicular2 axis to
the plane. The three coordinate planes in R3 (the xy–plane, the yz–plane and the xz–plane) divide space into
eight octants. The first octant is the one with x > 0, y > 0 and z > 0.
Example 10.2.1. Plot the points (1, 2, 3) and (3, −2, −6).
A NSWER. See the figure 10.7. □

Figure 10.7: Point (1, 2, 3)

In two–dimensional analytic geometry, the graph of an equation involving x and y is a curve in R2 . In
three–dimensional analytic geometry, an equation in x, y and z represents a surface in R3 .
Example 10.2.2. What surfaces in R3 are represented by the following equations? (1) z = 3, (2) y = 5.

Figure 10.8: Planes z = 3 and y = 5

A NSWER. See the figure 10.8.

(1) The surface of z = 3 is the set of points,
{ (x, y, 3) : any x, y ∈ R } ,
which represents the plane parallel to the xy–plane and passing through the point (0, 0, 3).
(2) The surface of y = 5 is the set of points,
{ (x, 5, z) : any x, z ∈ R } ,
which represents the plane parallel to the xz–plane and passing through the point (0, 5, 0). □
2 perpendicular, orthogonal: intersecting at or forming right (= 90◦ ) angles

Page 31 of 126
Calculus II for Engineers Fall, 2010

When an equation is given, we must understand from the context whether it represents a curve in R2 or a
surface in R3 . For instance, y = 5 represents a plane in R3 , but of course y = 5 can also represent a line in
R2 . In general in R3 , if k is a constant, then x = k represents a plane parallel to the yz–plane, y = k is a plane
parallel to the xz–plane, and z = k is a plane parallel to the xy–plane.

Example 10.2.3. Describe and sketch the surfaces in R3 represented by the equations (1) x − y = 0, (2)
x + y = −2.

Figure 10.9: Planes x − y = 0 and x + y = −2

A NSWER. See the figure 10.9.

(1) The surface of x − y = 0, i.e., y = x, is the set of points,

{ (x, x, z) : any x, z ∈ R } ,

which represents the plane parallel to the z–axis and passing through the point (0, 0, 0).
(2) The surface of x + y = −2, i.e., y = −x − 2, is the set of points,

{ (x, −x − 2, z) : any x, z ∈ R } ,

which represents the plane parallel to the z–axis and passing through the point (0, −2, 0). □
The familiar formula for the distance between two points in a plane is easily extended to the following three–
dimensional formula.

Theorem 10.2.4 (D ISTANCE BETWEEN T WO P OINTS IN T HREE –D IMENSIONAL S PACE). The distance
d(P1 , P2 ) between the points P1 (x1 , y1 , z1 ) and P2 (x2 , y2 , z2 ) is
√
d(P1 , P2 ) = (x2 − x1 )2 + (y2 − y1 )2 + (z2 − z1 )2 .

Example 10.2.5. Find the distance from the point P(2, −1, 7) to the point Q(1, −3, 5).

Example 10.2.6. Find an equation of a sphere with radius r and center C(a, b, c).

Hint. By definition, a sphere is the set of all points P(x, y, z) whose distance from C is r. Use the distance
formula above. □

Example 10.2.7. Show that x2 + y2 + z2 + 4x − 6y + 2z + 6 = 0 is the equation of a sphere and find its center
and radius.

Hint. Complete the squares in each variable. □

Page 32 of 126
Calculus II for Engineers Fall, 2010

□ 10.2.2 Vectors in R3 .
As in two dimensions, vectors in three–dimensional space have both direction and magnitude. We again
visualize vectors as directed line segments joining two points. A vector v is represented by any directed
line segment with the appropriate magnitude and direction. The position vector a with terminal point at
A(a1 , a2 , a3 ) (and initial point at the origin) is denoted by a = ⟨a1 , a2 , a3 ⟩.
We denote the set of all three–dimensional position vectors by
V3 = { ⟨x, y, z⟩ : x, y, z ∈ R } .
 Magnitude: A position vector v = ⟨a, b, c⟩ has the magnitude
√
∥v∥ = ∥ ⟨a, b, c⟩ ∥ = a2 + b2 + c2 ,
by the distance formula in three dimensions.
 The vector with initial point at P(a1 , a2 , a3 ) and terminal point at Q(b1 , b2 , b3 ) corresponds to the position
−→
vector PQ = ⟨b1 − a1 , b2 − a2 , b3 − a3 ⟩.
 Arithmetic: for vectors u = ⟨u1 , u2 , u3 ⟩ and v = ⟨v1 , v2 , v3 ⟩, we have

1. u + v = ⟨u1 , u2 , u3 ⟩ + ⟨v1 , v2 , v3 ⟩ = ⟨u1 + v1 , u2 + v2 , u3 + v3 ⟩.

2. u − v = ⟨u1 , u2 , u3 ⟩ − ⟨v1 , v2 , v3 ⟩ = ⟨u1 − v1 , u2 − v2 , u3 − v3 ⟩.
3. cu = c ⟨u1 , u2 , u3 ⟩ = ⟨cu1 , cu2 , cu3 ⟩, where c is a scalar.

 For a scalar c and a vector v ∈ V3 , we have

∥cv∥ = |c|∥v∥,
just like the case of the vectors in the plane.
 The zero vector 0 in V3 is defined by 0 = ⟨0, 0, 0⟩.
Theorem 10.2.8. For any vectors u, v and w in V3 , and any scalars a and b in R, the following hold:

1. Commutativity: u + v = v + u
2. Associativity: u + (v + w) = (u + v) + w
3. Zero Vector: u + 0 = u
4. Additive Inverse: u + (−u) = 0
5. Distributive Law: a(u + v) = au + av
6. Distributive Law: (a + b)u = au + bu
7. Multiplication by 1: (1)u = u
8. Multiplication by 0: (0)u = 0

 Standard Basis Vectors: A vector v = ⟨a, b, c⟩ is expressed by

v = ⟨a, b, c⟩ = ⟨a, 0, 0⟩ + ⟨0, b, 0⟩ + ⟨0, 0, c⟩ = a ⟨1, 0, 0⟩ + b ⟨0, 1, 0⟩ + c ⟨0, 0, 1⟩ .
We denote the vectors ⟨1, 0, 0⟩, ⟨0, 1, 0⟩, and ⟨0, 0, 1⟩ by i, j and k, respectively, that is,
i = ⟨1, 0, 0⟩ , j = ⟨0, 1, 0⟩ , k = ⟨0, 0, 1⟩ .
We call them the standard basis vectors in V3 . Using them, the vector v = ⟨a, b, c⟩ is expressed by
v = ⟨a, b, c⟩ = ai + bj + ck.
 Normalization: For a nonzero vector v ∈ V3 , a unit vector in the same direction as v is given by
1
u= v.
∥v∥

Page 33 of 126
Calculus II for Engineers Fall, 2010

Example 10.2.9. Find the unit vector in the direction of the vector v = 2i − j − 2k.

Example 10.2.10. Find a unit vector in the same direction as ⟨1, −2, 3⟩ and write ⟨1, −2, 3⟩ as the product of
its magnitude and a unit vector.

□ 10.2.3 Some Problems from Exercises.

 Give an equation (e.g., z = 0) for the indicated figure: (1) xy–plane, (2) x–axis.
−→ −→
 Find the displacement vectors PQ and QR and determine whether the points P(2, 3, 1), Q(0, 4, 2) and
R(4, 1, 4) are colinear3 .
 Use vectors to determine whether the points (2, 1, 0), (4, 1, 2) and (4, 3, 0) form an equilateral4 triangle.
 Use vectors to determine whether the points (1, −2, 1), (−2, −1, 2), (2, 0, 2) and (−1, 1, 3) form a square.

Figure 10.10: Hanging Crate

 Two ropes are attached to a 500 pound crate. See the figure 10.10. Rope A exerts a force of ⟨10, −130, 200⟩
pounds on the crate, and rope B exerts a force of ⟨−20, 180, 160⟩ pounds on the crate. If no further ropes are
added, find the net force on the crate and the direction it will move. If a third rope C is added to balance the
crate, what force must this rope exert on the crate?
 The thrust of an airplane’s engine produces a speed of 700 mph in still air. The plane is aimed in the
direction of ⟨6, −3, 2⟩, but its velocity with respect to the ground is ⟨580, −330, 160⟩ mph. Find the wind
velocity.

3 on the same line

4 having all sides or faces equal

Page 34 of 126
Calculus II for Engineers Fall, 2010

§10.3 The Dot Product.

□ 10.3.1 Definition.
Definition 10.3.1. The dot product of two vectors a = ⟨a1 , a2 , a3 ⟩ and b = ⟨b1 , b2 , b3 ⟩ in V3 is defined by

a · b = ⟨a1 , a2 , a3 ⟩ · ⟨b1 , b2 , b3 ⟩ = a1 b1 + a2 b2 + a3 b3 .

Likewise, the dot product of two vectors a = ⟨a1 , a2 ⟩ and b = ⟨b1 , b2 ⟩ in V2 is defined by

a · b = ⟨a1 , a2 ⟩ · ⟨b1 , b2 ⟩ = a1 b1 + a2 b2 .

Be careful: The dot product gives a scalar, not a vector. That is, a · b is a real number, not a vector.
Example 10.3.2. Compute the dot product a · b: (1) a = ⟨1, 2, 3⟩ and b = ⟨5, −3, 4⟩, (2) a = 2i − 5j and
b = 3i + 6j.
It is easy to see that for the standard basis vectors i, j and k, we have

i · i = 1 = j · j = k · k, and i · j = 0 = j · k = k · i.

The dot product obeys many of the laws that hold for ordinary products of real numbers. These are stated in
the following theorem.
Theorem 10.3.3. If a, b, and c are vectors and s is a scalar, then

1. Commutativity: a · b = b · a
2. Distributive Law: a · (b + c) = a · b + a · c
3. (sa) · b = s (a · b) = a · (sb)
4. 0·a = 0
5. a · a = ∥a∥2

Example 10.3.4. For two vectors a and b, expand (a + b) · (a + b) and (a − b) · (a + b).

A NSWER. Using the Theorem 10.3.3, we have

∥a + b∥2 = (a + b) · (a + b) = (a + b) · a + (a + b) · b = a · a + b · a + a · b + b · b
= ∥a∥2 + a · b + a · b + ∥b∥2 = ∥a∥2 + 2a · b + ∥b∥2
(a − b) · (a + b) = (a − b) · a + (a − b) · b = a · a − b · a + a · b − b · b
= ∥a∥2 − a · b + a · b − ∥b∥2 = ∥a∥2 − ∥b∥2 . □

□ 10.3.2 Geometric Interpretation of Dot Product.

Let θ ∈ [0, π ] be the angle between two nonzero vectors a and b. Then we have the following formula which
is used by physicists as the definition of the dot product.
Theorem 10.3.5.
a·b
a · b = ∥a∥∥b∥ cos θ , cos θ = ,
∥a∥∥b∥
where 0 ≤ θ ≤ π .
Notice

1. a and b have the same direction if and only if θ = 0.

2. a and b have opposite directions if and only if θ = π .

Page 35 of 126
Calculus II for Engineers Fall, 2010

3. a · b = 0 if and only if θ = π /2. We say that a and b are orthogonal or perpendicular if θ = π /2.

We consider the zero vector 0 to be orthogonal to every vector.

Example 10.3.6. If the vectors a and b have lengths 4 and 6, respectively, and the angle between them is
π /3, find a · b.
Example 10.3.7. Find the angle between the vectors: (1) a = ⟨2, 1, −3⟩ and b = ⟨1, 5, 6⟩, (2) a = ⟨2, 2, −1⟩
and b = ⟨5, −3, 2⟩.
Example 10.3.8. Determine whether the following pairs of vectors are orthogonal: (1) a = ⟨1, 3, −5⟩ and
b = ⟨2, 3, 10⟩ and (2) a = ⟨4, 2, −1⟩ and b = ⟨2, 3, 14⟩.
Example 10.3.9. Show that 2i + 2j − k is perpendicular to 5i − 4j + 2k.
The following two results provide us with some powerful tools for comparing the magnitudes of vectors.
Theorem 10.3.10 (C AUCHY–S CHWARTZ I NEQUALITY). For any vectors a and b,

|a · b| ≤ ∥a∥∥b∥

Theorem 10.3.11 (T RIANGLE I NEQUALITY). For any vectors a and b,

∥a + b∥ ≤ ∥a∥ + ∥b∥

Exercise 10.3.12. For nonzero vectors, a and b, show that c = ∥a∥b + ∥b∥a bisects the angle between a and
b.

□ 10.3.3 Components and Projections.

Figure 10.11: Component and Projection

−→ −
→
Consider two vectors PQ and PR. Let S be the foot of the perpendicular line segment from R to the line
−→ −→
containing PQ. See the figure 10.11. Then, the length of the line segment PS is called the component of PR
−→ −→
onto PQ and denoted by Comp− → PR. (When we put a stick on the desert slantly, we see the shadow of the
PQ
−
→
stick on the desert. We can think of the line segment PS as the shadow of the directed stick PR.) Let θ be the
−→ −
→ −
→ −→
angle between PQ and PR. Then since the line segment PS is the shadow of PR onto PQ, by the elementary
trigonometry, we observe
−
→ −
→ −
→
length of PS = ∥PR∥ cos θ , i.e., Comp−→ PR = ∥PR∥ cos θ .
PQ
−
→ −→
Using the position vectors a = PR and b = PQ, the component of a onto b is given by

Compb a = ∥a∥ cos θ ,

where 0 ≤ θ ≤ π is the angle between a and b. The Theorem 10.3.5 implies

a·b a·b a·b
Compb a = ∥a∥ cos θ = ∥a∥ = , i.e., Compb a =
∥a∥∥b∥ ∥b∥ ∥b∥

Page 36 of 126
Calculus II for Engineers Fall, 2010

which is very important and worth to be memorized. Since ∥b∥ is a scalar, we can rewrite the equation as
( )
b
Compb a = a ·
∥b∥
Since b/∥b∥ is a unit vector in the direction of b, the component of a onto b can be understood as the dot
product of a and the unit vector in the direction of b.
−→ −→
Let us go back to the argument with PQ and PR. We assign a direction to the line segment PS. Since the
−→ −→
line segment is on the line containing the vector PQ, so we can consider the vector PS, which is called the
−→ −→ −→ −
→ −
→
projection of PR onto PQ and denoted by Proj− → PR, i.e., PS = Proj−→ PR. Then we have
PQ PQ
−
→ −
→ −
→
length of the line segment PS = ∥PS∥, i.e., ∥PS∥ = Comp−→ PR.
PQ
−
→ −→ −→
Since PS is parallel to PQ, so by multiplying the unit vector in the direction of PQ , we get
−→ (− → −→ ) −→ −→ −→
−→ ( −→) PQ −→ PR · PQ PQ → PR · PQ −→
−
PS = Comp− → PR
PQ −→ , Proj−→ PR =
PQ −→ −→ , Proj−→ PR = −
PQ → PQ.
∥PQ∥ ∥PQ∥ ∥PQ∥ ∥PQ∥2
−
→ −→
Using the position vectors a = PR and b = PQ, the projection of a onto b is given by
a·b
Projb a = b.
∥b∥2
In summary,

1. Component of a along b is a scalar given by

a·b
Compb a = .
∥b∥
2. Projection of a along b is a vector, which is parallel to b, and given by
a·b
Projb a = b.
∥b∥2
In some books, the component Compb a is called the scalar projection of a onto b. The projection Projb a is
called the vector projection of a onto b.
Example 10.3.13. Find the component and projection of a along b: (1) a = ⟨2, 3⟩ and b = ⟨−1, 5⟩, (2)
a = ⟨1, 1, 2⟩ and b = ⟨−2, 3, 1⟩.
Example 10.3.14. Show that, in general, Compb a ̸= Compa b and Projb a ̸= Proja b. When does the equality
hold?

□ 10.3.4 Application: Work Done.

One use of projections occurs in physics in calculating work. The work done by a constant force F in moving
an object through a distance d is defined as W = Fd, but this applies only when the force is directed along
−
→
the line of motion of the object. Suppose, however, that the constant force is a vector F = PR, pointing in
−→
some other direction. If the force moves the object from P to Q, then the displacement vector is d = PQ.
The work W done by this force is defined to be the product of the component of the force F along d and the
distance moved:
F·d
W = (Compd F) ∥d∥ = ∥d∥ = F · d, i.e., W = F · d.
∥d∥
Thus the work W done by a constant force F is the dot product F · d, where d is the displacement vector.

Page 37 of 126
Calculus II for Engineers Fall, 2010

Example 10.3.15. A wagon is pulled a distance of 100 m along a horizontal path by a constant force of 70
N. The handle of the wagon is held at an angle of 35◦ above the horizontal. Find the work done by the force.

Example 10.3.16. You exert a constant force of 40 pounds in the direction of the handle of the wagon. If
the handle makes an angle of π /4 with the horizontal and you pull the wagon along a flat surface for 1 mile
(5280 feet), find the work done.

Example 10.3.17. A force is given by a vector F = 3i + 4j + 5k and moves a particle from the point P(2, 1, 0)
to the point Q(4, 6, 2). Find the work done.

Example 10.3.18. A car makes a turn on a banked road. If the road is banked at 10◦ , show that a vector
parallel to the road is ⟨cos 10◦ , sin 10◦ ⟩. If the car has weight 2000 pounds, find the component of the weight
vector along the road vector. This component of weight provides a force that helps the car turn.

Example 10.3.19. In the sequel of the previous example, find the component of the weight vector along the
road vector for a 2500–pound car on a 15◦ bank.

Page 38 of 126
Calculus II for Engineers Fall, 2010

§10.4 The Cross Product.

□ 10.4.1 Definition.
We start with the definition from the “Linear Algebra”:

Definition 10.4.1. (1) The determinant of 2 × 2 matrix of real numbers is a number defined by
 
 a a 
 1 2 
  = a1 b2 − a2 b1 .
 b1 b2 

(2) The determinant of 3 × 3 matrix of real numbers is a number defined by

 
       
 a1 a2 a3   b b   b b   
 
 b1 b2 b3  = a1  2 3  − a2  1 3  + a3  b1 b2 

   c2 c3   c1 c3   c1 c2 
 
 c1 c2 c3 
= a1 (b2 c3 − b3 c2 ) − a2 (b1 c3 − b3 c1 ) + a3 (b1 c2 − b2 c1 ) .

Now we define the cross product of two vectors.

Definition 10.4.2. For two vectors a = ⟨a1 , a2 , a3 ⟩ and b = ⟨b1 , b2 , b3 ⟩ in V3 , we define the cross product (or
vector product) of a and b to be
 
 
 i j k       
   a2 a3   a1 a3   a1 a2 
a × b =  a1 a2 a3  =  i− j+ k
   b2 b3   b1 b3   b1 b2 
 b1 b2 b3 
= (a2 b3 − a3 b2 ) i − (a1 b3 − a3 b1 ) j + (a1 b2 − a2 b1 ) k.

Notice that a × b is also a vector in V3 . To compute a × b, you must write the components of a in the second
row and the components of b in the third row: the order is important! The cross product is defined only for
vectors in V3 . There is no corresponding operation for vectors in V2 .

Example 10.4.3. Compute a × b: (1) a = ⟨1, 3, 4⟩ and b = ⟨2, 7, −5⟩, (2) a = i + 2j + 3k and b = 4i + 5j + 6k.

Theorem 10.4.4. For any vector a in V3 , we have

a × a = 0, and a × 0 = 0.

Theorem 10.4.5. For any vectors a and b in V3 , a × b is orthogonal to both a and b.

We recall the Remark 3 stating that two vectors u and v are orthogonal if and only if u · v = 0. A simple
computation shows
(a × b) · a = 0 = (a × b) · b.
Hence, the Theorem 10.4.5 above is proven.
Two different points determines one line, because there is only one line passing through two different points.
Similarly, two different lines determines one plane, because there is only one plane containing those two
different lines. Given two vectors, we can consider two different lines containing each of the vectors and
those two lines determines one plane. Since the vector a × b is orthogonal to both a and b, so we can observe
that the vector a × b is orthogonal to the plane determined by the vectors a and b. But, given a plane, out
of which side of the plane does a × b point? A simple computation shows i × j = k, where i, j and k are the
standard basis vectors of V3 . It illustrates the right–hand rule: If the fingers of your right hand curl in the

Page 39 of 126
Calculus II for Engineers Fall, 2010

Figure 10.12: Right–Hand Rule

direction of a rotation (through an angle less than 180◦ ) from a to b, then your thumb points in the direction
of a × b. See the figure 10.12.
By the right–hand rule or a simple computation, for the standard basis vectors we deduce
i × j = k, j × k = i, k × i = j,
j × i = −k, k × j = −i, i × k = −j.
Since the cross product does not follow several of the rules you might expect a product to satisfy, you might
ask what rules the cross product does satisfy. We summarize these in the Theorem below.
Theorem 10.4.6. For any vectors a, b and c in V3 and any scalar s, the following hold:

1. Anti–Commutativity: a × b = − (b × a)
2. (sa) × b = s (a × b) = a × (sb)
3. Distributive Law: a × (b + c) = a × b + a × c
4. Distributive Law: (a + b) × c = a × c + b × c
5. Scalar Triple Product: a · (b × c) = (a × b) · c
6. Vector Triple Product: a × (b × c) = (a · c) b − (a · b) c

Now that we know the direction of the vector a × b, the remaining thing we need to complete its geometric
description is its length ∥a × b∥. This is given by the following theorem.
Theorem 10.4.7. If 0 ≤ θ ≤ is the angle between a and b, then
∥a × b∥ = ∥a∥∥b∥ sin θ .
Since a vector is completely determined by its magnitude and direction, we can now say that a × b is the
vector that is perpendicular to both a and b, whose orientation is determined by the right–hand rule, and
whose length is ∥a∥∥b∥ sin θ . In fact, that is exactly how physicists define a × b.
Corollary 10.4.8. Two nonzero vectors a and b in V3 are parallel if and only if a × b = 0.

□ 10.4.2 Application: Area, Distance, Volume, Torque.

Area
If a and b are represented by directed line segments with the same initial point, then they determine a paral-
lelogram5 with base ∥a∥, altitude ∥b∥ sin θ . See the figure 10.13. The parallelogram has the area A obtained
by
Area A = (base)(altitude) = ∥a∥∥b∥ sin θ = ∥a × b∥.
Thus we have the following way of interpreting the magnitude of a cross product.
5a quadrilateral whose opposite sides are both parallel and equal in length

Page 40 of 126
Calculus II for Engineers Fall, 2010

Figure 10.13: Parallelogram formed by a and b

Theorem 10.4.9. The parallelogram formed by a and b has the area

A = ∥a × b∥.

Example 10.4.10. Find a vector perpendicular to the plane passing through the points P(1, 4, 6), Q(−2, 5, −1)
and R(1, −1, 1).

Example 10.4.11. Find the area of the triangle with vertices P(1, 4, 6), Q(−2, 5, −1) and R(1, −1, 1).

Example 10.4.12. Find the area of the parallelogram with two adjacent sides formed by the vectors a =
⟨1, 2, 3⟩ and b = ⟨4, 5, 6⟩.

Distance from a Point to a Line

Let d represent the distance from the point Q to the line through the points P and R. From elementary
trigonometry, we have that
−→
d = ∥PQ∥ sin θ
−→ −
→
where θ is the angle between PQ and PR. By the Theorem 10.4.7, we have
−→ − → −→ −
→ −
→
∥PQ × PR∥ = ∥PQ∥PR∥ sin θ = ∥PR∥d.

Solving this for d, we get the following formula.

Theorem 10.4.13. The distance from the point Q to the line through the points P and R is given by
−→ − →
∥PQ × PR∥
d= −→ .
∥PR∥

Example 10.4.14. Find the distance from the point Q(1, 2, 1) to the line through the points P(2, 1, −3) and
R(2, −1, 3).

Volume

Figure 10.14: Parallelepiped formed by a, b and c

Page 41 of 126
Calculus II for Engineers Fall, 2010

For any three non–coplanar vectors a, b and c, we consider the parallelepiped formed using the vectors as
three adjacent edges. See the figure 10.14. We choose a and b to form the base of the parallelogram. Then
the area of the base is A = ∥a × b∥. See the topic 10.4.2. If θ is the angle between c and a × b, then the height
of the parallelepiped is h = ∥c∥| cos θ |. (We must use | cos θ | instead of cos θ in case θ > π /2.) Therefore
the volume V of the parallelepiped is

Volume V = (area of base)(height) = Ah = ∥a × b∥∥c∥| cos θ | = |c · (a × b) |.

Thus we have proven the formula.

Theorem 10.4.15. The volume of the parallelepiped determined by the vectors a, b, and c is the magnitude
of their scalar triple product:
V = |c · (a × b)| .

Since we can choose b and c to form the base of the parallelogram, by the same argument, the volume is
obtained by
V = |a · (b × c)| .
That is, under the fixed products, we can rearrange the vectors:

V = |a · (b × c)| = |a · (c × b)| = |b · (c × a)| = |b · (a × c)| = |c · (a × b)| = |c · (b × a)| .

If we use the formula above and discover that the volume of the parallelepiped determined by a, b, and c is
0, then the vectors must be coplanar6 .
 Fast Computation: For a = ⟨a1 , a2 , a3 ⟩, b = ⟨b1 , b2 , b3 ⟩ and c = ⟨c1 , c2 , c3 ⟩, we have
 
 
 c1 c2 c3 
 
c · (a × b) =  a1 a2 a3  .
 
 b1 b2 b3 

Example 10.4.16. Find the volume of the parallelepiped with three adjacent edges formed by the vectors
a = ⟨1, 2, 3⟩, b = ⟨4, 5, 6⟩ and c = ⟨7, 8, 0⟩.

Example 10.4.17. Use the scalar triple product to show that the vectors a = ⟨1, 4, −7⟩, b = ⟨2, −1, 4⟩, and
c = ⟨0, −9, 18⟩ are coplanar.

Torque

Figure 10.15: Wrench with Torque

6 lying in the same plane

Page 42 of 126
Calculus II for Engineers Fall, 2010

The idea of a cross product occurs often in physics. In particular, we consider a force F acting on a rigid body
at a point given by a position vector r. For instance, if we tighten a bolt by applying a force to a wrench, we
produce a turning effect. See the figure 10.15. The torque τ (relative to the origin) is defined to be the cross
product of the position and force vectors:
τ = r×F
and measures the tendency of the body to rotate about the origin. The direction of the torque vector indicates
the axis of rotation. The magnitude of the torque vector is

∥τ ∥ = ∥r × F∥ = ∥r∥∥F∥ sin θ ,

where θ is the angle between the position and force vectors. Observe that the only component of F that can
cause a rotation is the one perpendicular to r, that is, ∥F∥ sin θ . The magnitude of the torque is equal to the
area of the parallelogram determined by r and F.

Example 10.4.18. If you apply a force of magnitude 25 pounds at the end of a 15–inch–long wrench, at an
angle of π /3 to the wrench, find the magnitude of the torque applied to the bolt. What is the maximum torque
that a force of 25 pounds applied at that point can produce?

Example 10.4.19. A bolt is tightened by applying a 40 N force to a 0.25 m wrench with an angle of 75◦
between the wrench and the force vector. Find the magnitude of the torque about the center of the bolt.

Page 43 of 126
Calculus II for Engineers Fall, 2010

§10.5 Lines and Planes in Space.

In this section, we study the lines and planes in R3 and find the equations of a line and a plane.

□ 10.5.1 Lines in R3 .
We recall that a line in the xy–plane is determined when a point (a, b) on the line and the direction m of the
line (its slope or angle of inclination) are given. The equation of the line can then be written using the point–
slope form: y − b = m(x − a). Similarly, a line L in three–dimensional space is determined when we know
a point P1 (x1 , y1 , z1 ) on L and the direction of L. In three dimensions the direction of a line is conveniently
described by a vector.
Let us look for the line L that passes through the point P1 (x1 , y1 , z1 ) and that is parallel to the position vector
−→
a = ⟨a1 , a2 , a3 ⟩. For any other point P(x, y, z) on L, we observe the vector P1 P is parallel to a. It implies
−→
P1 P = ta,
−→
where t is some scalar. Since P1 P = ⟨x − x1 , y − y1 , z − z1 ⟩, so the equation implies

⟨x − x1 , y − y1 , z − z1 ⟩ = ⟨ta1 ,ta2 ,ta3 ⟩ −→ x = x1 + ta1 , y = y1 + ta2 , z = z1 + ta3 .

We call the last equations parametric equations for the line L, where t is the parameter.
Another way of describing the line L is to eliminate the parameter t from the parametric equations. If none
of a1 , a2 , or a3 is 0, we can solve each of these equations for t, equate the results, and obtain
x − x1 y − y1 z − z 1
= = .
a1 a2 a3
which is called symmetric equations of the line L.

Example 10.5.1. Find an equation of the line L through the point (1, 5, 2) and parallel to the vector ⟨4, 3, 7⟩.
Also, determine where the line L intersects the yz–plane.

Example 10.5.2. Find an equation of the line L passing through the points P(1, 2, −1) and Q(5, −3, 4).

Example 10.5.3. (1) Find parametric equations and symmetric equations of the line that passes through the
points P(2, 4, −3) and Q(3, −1, 1).
(2) At what point does this line intersect the xy–plane?

Definition 10.5.4. Let L and M be two lines in R3 , with parallel vectors vL and vM , respectively.

1. The lines L and M are parallel if and only if vL and vM are parallel.
2. If L and M intersect, then

(1) the angle between lines L and M is same as the angle between parallel vectors vL and vM ,
(2) the lines L and M are orthogonal to each other if and only if the vectors vL and vM are orthogonal to
each other.

Example 10.5.5. Show that the lines L and M defined by

L : x − 2 = −t, y − 1 = 2t, z − 5 = 2t
M : x − 1 = s, y − 2 = −s, z − 1 = 3s

are not parallel, yet do not intersect.

Definition 10.5.6. Nonparallel, nonintersecting lines are called skew lines.

Page 44 of 126
Calculus II for Engineers Fall, 2010

Note that it’s fairly easy to visualize skew lines. Draw two planes that are parallel and draw a line in each
plane (so that it lies completely in the plane). As long as the two lines are not parallel, these are skew lines.
Example 10.5.7. Show that the lines and with parametric equations

L : x = 1 + t, y = −2 + 3t, z = 4 − t
M : x = 2s, y = 3 + s, z = −3 + 4s

are skew lines, that is, they do not intersect and are not parallel (and therefore do not lie in the same plane).

□ 10.5.2 Planes in R3 .
Although a line in space is determined by a point and a direction, a plane in space is more difficult to
describe. A single vector parallel to a plane is not enough to convey the “direction” of the plane, but a vector
perpendicular to the plane does completely specify its direction. Thus a plane in space is determined by a
point in the plane and a vector that is orthogonal to the plane. This orthogonal vector is called a normal
vector of the plane.
Let us find an equation of the plane which passes through a point P1 (x1 , y1 , z1 ) and orthogonal to the vector
a = ⟨a, b, c⟩. Let P(x, y, z) be an arbitrary point in the plane. Then since P and P1 are both points in the plane,
−→
the vector P1 P = ⟨x − x1 , y − y1 , z − z1 ⟩ lies in the plane and so, must be orthogonal to a. By the formula on
the dot product, hence, we deduce
−→
0 = a · P1 P = ⟨a, b, c⟩ · ⟨x − x1 , y − y1 , z − z1 ⟩ = a (x − x1 ) + b (y − y1 ) + c (z − z1 ) .

That is, the equation of the plane passing through the point (x1 , y1 , z1 ) with normal vector ⟨a, b, c⟩ is

a (x − x1 ) + b (y − y1 ) + c (z − z1 ) = 0.

Example 10.5.8. Find an equation of the plane containing the point (1, 2, 3) with normal vector ⟨4, 5, 6⟩, and
sketch the plane.
Example 10.5.9. Find an equation of the plane through the point (2, 4, −1) with normal vector ⟨2, 3, 4⟩. Find
the intercepts and sketch the plane.
Note that if we expand out the equation of the plane, we get

0 = a (x − x1 ) + b (y − y1 ) + c (z − z1 ) = ax + by + cz − (ax1 + by1 + cz1 ) .

Since − (ax1 + by1 + cz1 ) is a constant, when we let d = − (ax1 + by1 + cz1 ), the equation above becomes
simply
ax + by + cz + d = 0.
where d = − (ax1 + by1 + cz1 ). We refer to this last equation as a linear equation in the three variables x, y
and z. Conversely, it can be shown that if a, b, and c are not all 0, then the linear equation above represents a
plane with normal vector ⟨a, b, c⟩.
Example 10.5.10. Find the plane containing the three points P(1, 2, 2), Q(2, −1, 4) and R(3, 5, −2).
Example 10.5.11. Find an equation of the plane that passes through the points P(1, 3, 2), Q(3, −1, 6), and
R(5, 2, 0).
In three dimensions, two planes are either parallel or they intersect in a straight line. Suppose that two planes
having normal vectors a and b, respectively, intersect. Then the angle between the planes is the same as the
angle between a and b. With this in mind, we say

1. the planes are parallel whenever their normal vectors are parallel and

Page 45 of 126
Calculus II for Engineers Fall, 2010

2. the planes are orthogonal whenever their normal vectors are orthogonal.

Example 10.5.12. Find an equation for the plane through the point (1, 4, −5) and parallel to the plane defined
by 2x − 5y + 7z = 12.
Example 10.5.13. Describe and draw the planes y = 3 and y = 8.
Example 10.5.14. (1) Find the angle between the planes x + y + z = 1 and x − 2y + 3z = 1.
(2) Find symmetric equations for the line of intersection L of these two planes.
Example 10.5.15. Find the line of intersection of the planes x + 2y + z = 3 and x − 4y + 3z = 5.

Distance between Plane and Point

Let us find the distance between the plane ax + by + cz + d = 0 and a point P0 (x0 , y0 , z0 ) which is not on the
plane. We choose a point P1 (x1 , y1 , z1 ) lying on the plane. Then we observe the distance between P0 and the
−→
plane is simply the absolute value of the component of the vector P1 P along the normal vector a, i.e.,
 
 −−→
distance between plane and P0 = Compa P1 P0  .

Computing it, we get

−−→
−−→ P1 P0 · a ⟨x0 − x1 , y0 − y1 , z0 − z1 ⟩ · ⟨a, b, c⟩
Compa P1 P0 = =
∥a∥ ∥ ⟨a, b, c⟩ ∥
a(x0 − x1 ) + b(y0 − y1 ) + c(z0 − z1 ) ax0 + by0 + cz0 − (ax1 + by1 + cz1 )
= √ = √ .
a2 + b2 + c2 a2 + b2 + c2
Since P1 (x1 , y1 , z1 ) is on the plane ax + by + cz + d = 0, we get

ax1 + by1 + cz1 + d = 0, i.e., d = − (ax1 + by1 + cz1 ) ,

which implies
−−→
−−→ P1 P0 · a ax0 + by0 + cz0 − (ax1 + by1 + cz1 ) ax0 + by0 + cz0 + d
Compa P1 P0 = = √ = √
∥a∥ a2 + b2 + c2 a2 + b2 + c2
  |ax + by + cz + d|
 −−→ 0 0 0
Compa P1 P0  = √ .
a + b + c2
2 2

Thus, simply, the formula for the distance D between the plane ax + by + cz + d = 0 and a point P0 (x0 , y0 , z0 )
which is not on the plane can be written as
|ax0 + by0 + cz0 + d|
D= √ .
a2 + b2 + c2
Example 10.5.16. Find the distance between the parallel planes:
 Π1 : 2x − 3y + z = 6 and Π2 : 4x − 6y + 2z = 8.
 Π1 : 10x + 2y − 2z = 5 and Π2 : 5x + y − z = 1.
Example 10.5.17. The lines L : x = 1 + t, y = −2 + 3t, z = 4 − t and M : x = 2s, y = 3 + s, z = −3 + 4s are
skew. Find the distance between them.

§10.6 Surfaces in Space.

Presentation via Mathematica and Webpage: Interactive Gallery of Quadric Surfaces
(http://www.math.umn.edu/˜rogness/quadrics/index.shtml)
Please read the textbook.

Page 46 of 126
 Chapter 11

Vector–Valued Functions

§11.1 Vector–Valued Functions.

We briefly discuss some examples of vector–valued functions.

□ 11.1.1 Vector–Valued Functions in Plane.

Definition 11.1.1. A vector–valued function r(t) in the two–dimensional space is a mapping from its domain
D ⊂ R to its range R ⊂ V3 , so that for each t ∈ D, r(t) = v for exactly one vector v ∈ V2 . We can always write
a vector–valued function as
r(t) = f (t)i + g(t)j,
for some scalar functions f and g (called the component functions of r).

Example 11.1.2. Sketch a graph of the curve traced out by the endpoint of the two–dimensional vector–
valued function
r(t) = (t + 1)i + (t 2 − 2)j.

⟨ ⟩
t r(t) = t + 1,t 2 − 2
−3 ⟨−2, 7⟩
−2 ⟨−1, 2⟩
−1 ⟨0, −1⟩
0 ⟨1, −2⟩
1 ⟨2, 2⟩
2 ⟨3, 7⟩
3 ⟨4, 14⟩
4 ⟨5, 23⟩ Figure 11.1: Graph of r(t) = (t + 1)i + (t 2 − 2)j

A NSWER. We make the table as below and connect all the endpoints of the vectors in the table. Then
we obtain the graph of the vector–valued function r(t). As t increases, the endpoints of the vector–valued
functions move from the second quadrant via the third and fourth quadrant to the first quadrant. This direction
of increasing values of t is called the orientation of the curve. See the figure 11.1. □
A NSWER. Another way to get the graph is using x and y as follows: let x = t + 1 (first component of r(t))
and y = t 2 − 2 (second component of r(t)). Then, we get t = x − 1 and putting this into the equation of y
implies
y = t 2 − 2 = (x − 1)2 − 2, i.e., y = (x − 1)2 − 2,
of which graph is exactly same as the one in the figure 11.1. In this solution, it is not easy to determine the
orientation of the curve compared to the solution above. □

Example 11.1.3. Sketch a graph of the curve traced out by the endpoint of the vector–valued function

r(t) = 4 costi − 3 sintj, t ∈ R.

47
Calculus II for Engineers Fall, 2010

Figure 11.2: Graph of r(t) = 4 costi − 3 sintj

A NSWER. One can use the table as the Example 11.1.2. But I prefer using x and y. The curve can be written
parametrically as
x = 4 cost, y = −3 sint, t ∈ R.
We deduce an equation of x and y by getting rid of the parameter t as follows:
( x )2 ( y )2 ( x )2 ( y )2
2 2
+ = cos t + sin t = 1, i.e., + = 1,
4 3 4 3
of which graph is an ellipse. See the figure 11.2. □

□ 11.1.2 Vector–Valued Functions in Space.

Definition 11.1.4. A vector–valued function r(t) in the three–dimensional space has the domain D ⊂ R and
the range R ⊂ V3 . We can write it as
r(t) = f (t)i + g(t)j + h(t)k,
for some scalar functions f , g and h (called the component functions of r).
Example 11.1.5. Plot the curve traced out by the vector–valued function r(t) = sinti − 3 costj + 2tk, t ≥ 0.

Figure 11.3: Graph of r(t) = sinti − 3 costj + 2tk

A NSWER. The curve is given parametrically by

x = sint, y = −3 cost, z = 2t, t ≥ 0.
While most curves in three dimensions are difficult to recognize, you should notice that there is a relationship
between x and y here, namely,
( y )2
2
x + = sin2 t + cos2 t = 1.
3

Page 48 of 126
Calculus II for Engineers Fall, 2010

But since z = 2t, z will increase as t increases and so, the curve will wind its way up the cylinder, as t
increases. We call this curve an elliptical helix. See the figure 11.3. □
Example 11.1.6. Plot the curve traced out by the vector–valued function
r(t) = ⟨3 + 2t, 5 − 3t, 2 − 4t⟩ , t ∈ R.

Figure 11.4: Graph of r(t) = ⟨3 + 2t, 5 − 3t, 2 − 4t⟩

A NSWER. Notice that the curve is given parametrically by

x = 3 + 2t, y = 5 − 3t, z = 2 − 4t, t ∈ R.
You should recognize these equations as parametric equations for the straight line parallel to the vector
⟨2, −3, −4⟩ and passing through the point (3, 5, 2). See the figure 11.4. □

□ 11.1.3 Arc Length.

 C ALCULUS I FOR E NGINEERS:
The curve of y = f (x) on [a, b] has the arc length
ˆ b√
s= 1 + [ f ′ (x)]2 dx.
a
 A RC L ENGTH OF V ECTOR –VALUED F UNCTION IN P LANE:
A vector–valued function r(t) = ⟨ f (t), g(t)⟩ for t ∈ [a, b] has the arc length
ˆ b√
s= [ f ′ (t)]2 + [g′ (t)]2 dt.
a
 A RC L ENGTH OF V ECTOR –VALUED F UNCTION IN S PACE:
A vector–valued function r(t) = ⟨ f (t), g(t), h(t)⟩ for t ∈ [a, b] has the arc length
ˆ b√
s= [ f ′ (t)]2 + [g′ (t)]2 + [h′ (t)]2 dt.
a
Example 11.1.7. Find the arc length of the curve traced out by the endpoint of the vector–valued function
⟨ ⟩
r(t) = 2t, lnt,t 2 , 1 ≤ t ≤ e.
A NSWER. First, notice that for x(t) = 2t, y(t) = lnt and z(t) = t 2 , we have
1
x′ (t) = 2, y′ (t) = , z′ (t) = 2t,
t
and the curve is traversed exactly once for 1 ≤ t ≤ e. (To see why, observe that x = 2t is an increasing
function.) By the formula above, we have
√
ˆ 2 ( )2
1
s= 22 + + (2t)2 dt = e2 . □
1 t

Page 49 of 126
Calculus II for Engineers Fall, 2010

§11.2 The Calculus of Vector–Valued Functions.

□ 11.2.1 Limit, Continuity and Derivative.

Definition 11.2.1. For a vector–valued function r(t) = ⟨ f (t), g(t), h(t)⟩, the limit of r(t) as t approaches a is
given by
⟨ ⟩
lim r(t) = lim ⟨ f (t), g(t), h(t)⟩ = lim f (t), lim g(t), lim h(t)
t→a t→a t→a t→a t→a

provided all of the indicated limits exist. If any of the limits indicated on the right–hand side of the equation
above fail to exist, then limt→a r(t) does not exist.
⟨ ⟩
Example 11.2.2. Find limt→0 t 2 + 1, 5 cost, sint .

Definition 11.2.3. The vector–valued function r(t) = ⟨ f (t), g(t), h(t)⟩ is continuous at t = a whenever

lim r(t) = r(a)

t→a

(i.e., whenever the limit exists and equals the value of the vector–valued function).

Theorem 11.2.4. A vector–valued function r(t) = ⟨ f (t), g(t), h(t)⟩ is continuous at t = a if and only if all of
f , g and h are continuous at t = a.
⟨ ⟩
Example 11.2.5. Determine for what values of t the vector–valued function r(t) = e5t , ln(t + 1), cost is
continuous.

Definition 11.2.6. The derivative r′ of the vector–valued function r is defined by

r (t + h) − r(t)
r′ (t) = lim ,
h→0 h

for any values of t for which the limit exists. When the limit exists for t = a, we say that r is differentiable at
t = a.

Theorem 11.2.7. Let r(t) = ⟨ f (t), g(t), h(t)⟩ and suppose that the components f , g and h are all differentiable
for some value of t. Then r is also differentiable at that value of t and its derivative is given by
⟨ ⟩
r′ (t) = f ′ (t), g′ (t), h′ (t)
⟨ ⟩
Example 11.2.8. Find the derivative of r(t) = sin(t 2 ), ecost ,t lnt .

Theorem 11.2.9. Suppose that r and s are differentiable vector–valued functions, f is a differentiable scalar
function and a and b are any scalar constants. Then,

1. Linearity: [ar + bs]′ = ar′ + bs′

2. Product Rule (Scalar Function Multiple): [ f r]′ = f ′ r + f r′
3. Product Rule (Dot Product): [r · s]′ = r′ · s + r · s′
4. Product Rule (Cross Product): [r × s]′ = r′ × s + r × s′
5. Chain Rule: [r( f )]′ = f ′ r′ ( f )

Page 50 of 126
Calculus II for Engineers Fall, 2010

□ 11.2.2 Interpretation of Derivative.

We say that the curve traced out by the vector–valued function r(t) = ⟨ f (t), g(t), h(t)⟩ on an interval I is
smooth if r′ is continuous on I and r′ (t) = 0, except possibly at any endpoints of I. Notice that this says that
the curve is smooth provided f ′ , g′ and h′ are all continuous on I and f ′ (t), g′ (t) and h′ (t) are not all zero at
the same point in I.
⟨ ⟩
Example 11.2.10. Is the plane curve traced out by the vector–valued function r(t) = t 3 ,t 2 smooth?

We refer to r′ (a) as a tangent vector to the curve C at the point corresponding to t = a. Be sure to observe
that r′ (a) lies along the tangent line to the curve at t = a and points in the direction of the orientation of C.

Example 11.2.11. For r(t) = ⟨− cos 2t, sin 2t⟩, plot the curve traced out by the endpoint of r(t) and draw the
position vector and tangent vector at t = π /4.

A NSWER. See the figure 11.5. □

Figure 11.5: Graph of r(t) = ⟨− cos 2t, sin 2t⟩

Theorem 11.2.12. ∥r(t)∥ = constant if and only if r(t) and r′ (t) are orthogonal, for all t.

The Theorem says that the path traced out by r(t) lies on a circle centered at the origin if and only if the
tangent vector is orthogonal to the position vector at every point on the curve. Likewise, in three dimensions,
it says that the curve traced out by r(t) lies on a sphere centered at the origin if and only if the tangent vector
is orthogonal to the position vector at every point on the curve.

□ 11.2.3 Integral.

Definition 11.2.13. The vector–valued function R(t) is an antiderivative of the vector–valued function r(t)
whenever R′ (t) = r(t).

Definition 11.2.14. If R(t) is any antiderivative of r(t), the indefinite integral of r(t) is defined to be
ˆ
r(t) dt = R(t) + c,

where c is an arbitrary constant vector.

As in the scalar case, R(t) + c is the most general antiderivative of r(t). (Why is that?) Notice
ˆ ˆ ⟨ˆ ˆ ˆ ⟩
r(t) dt = ⟨ f (t), g(t), h(t)⟩ dt = f (t) dt, g(t) dt, h(t) dt

That is, you integrate a vector–valued function by integrating each of the individual components.

Page 51 of 126
Calculus II for Engineers Fall, 2010

´⟨2 t 2
⟩
Example 11.2.15. Evaluate the indefinite integral t + 2, sin(2t), 4te dt.

Definition 11.2.16. For the vector–valued function r(t) = ⟨ f (t), g(t), h(t)⟩, we define the definite integral of
r(t) on the interval [a, b] by
ˆ ˆ ⟨ˆ ˆ ˆ ⟩
b b b b b
r(t) dt = ⟨ f (t), g(t), h(t)⟩ dt = f (t) dt, g(t) dt, h(t) dt
a a a a a

Theorem 11.2.17. Suppose that R(t) is an antiderivative of r(t) on the interval [a, b]. Then,
ˆ b
r(t) dt = R(b) − R(a)
a
´1⟨ ⟩
0 sin(π t), 6t + 4t dt.
Example 11.2.18. Evaluate 2

§11.3 Motion In Space.

Skip. Please read the textbook.

Page 52 of 126
Calculus II for Engineers Fall, 2010

§11.4 Curvature.

□ 11.4.1 Arc Length Function.

Recall that the curve traced out by the endpoint of the vector–valued function r(t) = ⟨ f (t), g(t), h(t)⟩, for
a ≤ t ≤ b, has the arc length
ˆ b√
s= [ f ′ (t)]2 + [g′ (t)]2 + [h′ (t)]2 dt.
a
We define the arc length parameter s(t) to be the arc length of that portion of the curve from u = a up to u = t.
That is, ˆ √ ˆ
u=t u=t
s(t) = [ f ′ (u)]2 + [g′ (u)]2 + [h′ (u)]2 du = ∥r′ (u)∥ du.
u=a u=a

Example 11.4.1. Find an arc length parameterization of the circle of radius 4 centered at the origin.

A NSWER. Note that one parameterization of this circle is

C : x = f (t) = 4 cost, y = g(t) = 4 sint, 0 ≤ t ≤ 2π .

It gives the vector–valued function r(t) = ⟨4 cost, 4 sint⟩ on [0, 2π ]. In this case, the arc length from u = 0 to
u = t is given by
ˆ t√ ˆ t√ ˆ t
s(t) = ′ 2 ′ 2
[ f (u)] + [g (u)] du = 2 2
[−4 sin u] + [4 cos u] du = 4 1 dt = 4t.
0 0 0

That is, t = s/4, so that an arc length parameterization for C is

s s
C : x = 4 cos , y = 4 sin , 0 ≤ s ≤ 8π . □
4 4
Exercise 11.4.2. Reparametrize the helix r(t) = ⟨cost, sint,t⟩ with respect to arc length measured from
(1, 0, 0) in the direction of increasing t.

□ 11.4.2 Curvature.
Consider the smooth curve C traced out by the endpoint of the vector–valued function r(t). Recall that for
each t, r′ (t) can be thought of as both the velocity vector and a tangent vector, pointing in the direction of
motion (i.e., the orientation of C). Notice that

r′ (t)
T(t) =
∥r′ (t)∥

is also a tangent vector, but has length one (∥T(t)∥ = 1). We call T(t) the unit tangent vector to the curve
C. That is, for each t, T(t) is a tangent vector of length one pointing in the direction of the orientation of C.

⟨ ⟩
Figure 11.6: Graph of r(t) = t 2 + 1,t

Page 53 of 126
Calculus II for Engineers Fall, 2010

⟨ ⟩
Example 11.4.3. Find the unit tangent vector to the curve determined by r(t) = t 2 + 1,t .

A NSWER. See the figure 11.6. □

From the figure 11.7, you can see that the tangent vector changes direction very slowly when C is fairly
straight, but it changes direction more quickly when C bends or twists more sharply.
The curvature of C at a given point is a measure of how quickly the curve changes direction at that point.
Specifically, we define it to be the magnitude of the rate of change of the unit tangent vector with respect to
arc length. (We use arc length so that the curvature will be independent of the parametrization.)

Figure 11.7: Unit Tangent Vectors

Definition 11.4.4. The curvature κ of a curve is the scalar quantity

dT
κ =
ds

where T represents the unit tangent vector and s represents the arc length parameter.

Note that, while the definition of curvature makes sense intuitively, it is not a simple matter to compute κ
directly from the definition. So we look for another form. The curvature is easier to compute if it is expressed
in terms of the parameter t instead of s, so we use the Chain Rule to write

dT/ dt ′
dT
=
dT ds
−→ κ = = ∥T (t)∥
dt ds dt ds/ dt ∥ ds/ dt∥

But ds
dt = ∥r′ (t)∥ from the definition of the arc length function, so finally we have

∥T′ (t)∥
κ=
∥r′ (t)∥
Example 11.4.5. Find the curvature of a straight line traced out by the vector–valued function

r(t) = ⟨at + x0 , bt + y0 , ct + z0 ⟩

for some constants a, b, c, x0 , y0 and z0 (where at least one of a, c or e is nonzero).

Example 11.4.6. Show that the curvature of a circle of radius a is 1a .

Notice that the result of the example is consistent with your intuition. First, observe that you should be able to
drive a car around a circular track while holding the steering wheel in a fixed position. (That is, the curvature
should be constant.) Further, the smaller that the radius of a circular track is, the sharper you will need to turn
(that is, the larger the curvature). On the other hand, on a circular track of very large radius, it would seem as
if you were driving fairly straight (i.e., the curvature will be close to 0).
Although the formula on curvature above can be used in all cases to compute the curvature, the formula given
by the following theorem is often more convenient to apply.

Page 54 of 126
Calculus II for Engineers Fall, 2010

Theorem 11.4.7. The curvature of the curve given by the vector–valued function r(t) is

∥r′ (t) × r′′ (t)∥

κ=
∥r′ (t)∥3

Example 11.4.8. Find the curvature of the helix traced out by r(t) = ⟨2 sint, 2 cost, 4t⟩.
⟨ ⟩
Example 11.4.9. Find the curvature of the twisted cubic r(t) = t,t 2 ,t 3 at a general point and at (0, 0, 0).

In the case of a plane curve that is the graph of a function, y = f (x), we can derive a particularly simple for-
mula for the curvature. Notice that such a curve is traced out by the vector–valued function r(t) = ⟨t, f (t), 0⟩,
where the third component is 0, since the curve lies completely in the xy–plane. Applying the formula to this
vector–valued function and putting the parameter t = x back, we obtain

| f ′′ (x)|
κ=
{1 + [ f ′ (x)]2 }3/2

Example 11.4.10. Find the curvature of the parabola y = ax2 + bx + c. Also, find the limiting value of the
curvature as x → ∞.

Example 11.4.11. Find the curvature of the parabola y = x2 at x = 0, x = 1 and x = 2.

§11.5 Tangent and Normal Vectors.

Skip. Please read the textbook.

§11.6 Parametric Surfaces.

Skip. Please read the textbook.

Page 55 of 126
Calculus II for Engineers Fall, 2010

Page 56 of 126
 Chapter 12

Functions of Several Variables and Partial Differentiation

§12.1 Functions of Several Variables.

Skip. Please read the textbook.

§12.2 Limits nd Continuity.

Skip. Please read the textbook.

§12.3 Partial Derivatives.

□ 12.3.1 Partial Derivative.

In general, if is a function of two variables x and y, suppose we let only x vary while keeping y fixed, say y = b,
where b is a constant. Then we are really considering a function of a single variable x, namely, g(x) = f (x, b).
If g has a derivative at a, then we call it the partial derivative of f with respect to x at (a, b) and denote it
by fx (a, b). Thus
fx (a, b) = g′ (a), where g(x) = f (x, b).
By the definition of a derivative, we have

g(a + h) − g(a)
g′ (a) = lim
h→0 h
and so the equation above becomes

f (a + h, b) − f (a, b)
fx (a, b) = lim .
h→0 h
Similarly, the partial derivative of f with respect to y at (a, b), denoted by fy (a, b), is obtained by keeping
x fixed (x = a) and finding the ordinary derivative at b of the function h(y) = f (a, y)

f (a, b + h) − f (a, b)
fy (a, b) = lim .
h→0 h
If we now let the point (a, b) vary, then fx and fy become functions of two variables.

Definition 12.3.1. If f is a function of two variables x and y, its partial derivatives are the functions fx and
fy defined by

f (x + h, y) − f (x, y) f (x, y + h) − f (x, y)

fx (x, y) = lim , fy (x, y) = lim .
h→0 h h→0 h
Remark 12.3.2 (N OTATIONS FOR PARTIAL D ERIVATIVES). If z = f (x, y), then we write

∂f ∂ ∂z
fx (x, y) = fx = = f (x, y) = = f1 = D1 f = Dx f .
∂x ∂x ∂x
∂f ∂ ∂z
fy (x, y) = fy = = f (x, y) = = f2 = D2 f = Dy f .
∂y ∂y ∂y
To compute partial derivatives, all we have to do is remember that the partial derivative with respect to x is
just the ordinary derivative of the function g of a single variable that we get by keeping y fixed. Thus we have
the following rule.

57
Calculus II for Engineers Fall, 2010

Remark 12.3.3 (RULE FOR F INDING PARTIAL D ERIVATIVES OF z = f (x, y)).

1. To find fx , regard y as a constant and differentiate f (x, y) with respect to x.

2. To find fy , regard x as a constant and differentiate f (x, y) with respect to y.

Example 12.3.4. If f (x, y) = x3 + x2 y3 − 2y2 , find fx and fy and fx (1, 2) and fy (1, 2).

A NSWER.
∂f ∂ ∂ ∂
fx = = x3 + (x2 y3 ) − 2 y2 = 3x2 + 2xy3 − 2(0) = 3x2 + 2xy3 ,
∂x ∂x ∂x ∂x
∂f ∂ 3 ∂ 2 3 ∂
fy = = x + (x y ) − 2 y2 = 0 + 3x2 y2 − 4y = 3x2 y2 − 4y,
∂y ∂y ∂y ∂y
2 3
fx (1, 2) = 3(1) + 2(1)(2) = 19, fy (1, 2) = 3(1)2 (2)2 − 4(2) = 4 □

Example 12.3.5. For f (x, y) = 3x2 + x3 y + 4y2 , find fx and fy and fx (1, 0) and fy (2, −1).
( )
Example 12.3.6. Find the partial derivatives fx and fy : (1) f (x, y) = exy + xy and (2) f (x, y) = sin x
1+y .

Example 12.3.7. For a real gas, van der Waals’ equation states that
( )
n2 a
P + 2 (V − nb) = nRT.
V

Here, P is the pressure of the gas, V is the volume of the gas, T is the temperature (in degrees Kelvin), n is
the number of moles of gas, R is the universal gas constant and a and b are constants. Compute and interpret
∂P ∂T
∂ V and ∂ P .

□ 12.3.2 Higher–Order Partial Derivative.

If is a function of two variables, then its partial derivatives fx and fy are also functions of two variables, so
we can consider their partial derivatives ( fx )x , ( fx )y , ( fy )x , and ( fy )y , which are called the second partial
derivatives of f . If z = f (x, y), we use the following notation:

Remark 12.3.8 (N OTATIONS FOR S ECOND PARTIAL D ERIVATIVES).

( )
∂ ∂f ∂2 f ∂ 2z
( fx )x = fxx = f11 = = 2 = 2.
∂x ∂x ∂x ∂x
( )
∂ ∂f ∂ f
2 ∂ 2z
( fx )y = fxy = f12 = = = .
∂y ∂x ∂ y∂ x ∂ y∂ x
( )
∂ ∂f ∂2 f ∂ 2z
( fy )x = fyx = f21 = = = .
∂x ∂y ∂ x∂ y ∂ x ∂ y
( )
∂ ∂f ∂2 f ∂ 2z
( fy )y = fyy = f22 = = 2 = 2.
∂y ∂y ∂y ∂y

Thus the notation fxy or ∂∂y∂fx means that we first differentiate with respect to x and then with respect to y,
2

whereas fyx in computing the order is reversed.

Example 12.3.9. Find the second derivatives of f (x, y) = x3 + x2 y3 − 2y2 .

A NSWER. From the Example 12.3.4 above, we have

fx = 3x2 + 2xy3 , fy = 3x2 y2 − 4y.

Page 58 of 126
Calculus II for Engineers Fall, 2010

Therefore,
∂ ( 2 ) ∂ ( 2 )
fxx = 3x + 2xy3 = 6x + 2y3 , fxy = 3x + 2xy3 = 6xy2
∂x ∂y
∂ ( 2 2 ) ∂ ( 2 2 )
fyx = 3x y − 4y = 6xy2 , fyy = 3x y − 4y = 6x2 y − 4 □
∂x ∂y

Exercise 12.3.10. Find all second–order partial derivatives of f (x, y) = x2 y − y3 + ln x.

Notice in the example above that fxy = 6xy2 = fyx . It turns out that this is true for most, but not all, of the
functions that we will encounter. The following theorem discovered by Alexis Clairaut (1713 – 1765) gives
conditions under which we can assert that fxy = fyx .

Theorem 12.3.11 (C LAIRAUT ’ S T HEOREM). If fxy and fyx are continuous on an open set containing (a, b),
then fxy (a, b) = fyx (a, b).

We can, of course, compute third–, fourth– or even higher–order partial derivatives.

Example 12.3.12. Compute fxxyz if f (x, y, z) = sin(3x + yz).

A NSWER.

fx = 3 cos(3x + yz), fxx = −9 sin(3x + yz)

fxxy = −9z cos(3x + yz), fxxyz = −9 [cos(3x + yz) − y sin(3x + yz)] □

Exercise 12.3.13. For f (x, y) = cos(xy) − x3 + y4 , compute fxyy and fxyyy .

√
Exercise 12.3.14. For f (x, y, z) = xy3 z + 4x2 y, defined for x, y, z ≥ 0, compute fx , fxy and fxyz .

Exercise 12.3.15. The sag in a beam of length L, width w and height h is given by

L4
S(L, w, h) = c
wh3
for some constant c. Show that
∂S 4 ∂S 1 ∂S 3
= S, = − S, = − S.
∂L L ∂w w ∂h h
Use this result to determine which variable has the greatest proportional effect on the sag.

§12.4 Tangent Planes And Linear Approximations.

Skip. Please read the textbook.

Page 59 of 126
Calculus II for Engineers Fall, 2010

§12.5 The Chain Rule.

□ 12.5.1 Chain Rule.

We recall from Calculus I that f (x) = (x2 + 8)5 has the derivative

f ′ (x) = 5(x2 + 8)4 (x2 + 8)′ = 5(x2 + 8)4 (2x) = 10x(x2 + 8)4 .

The general form of the chain rule says that for differentiable functions h and g,
d
[h(g(x))] = h′ (g(x))g′ (x).
dx
We now extend the chain rule to functions of several variables. This takes several slightly different forms,
depending on the number of independent variables, but each is a variation of the already familiar chain rule
for functions of a single variable.
For functions of more than one variable, the Chain Rule has several versions, each of them giving a rule for
differentiating a composite function.

Chain Rule: Case 1

Theorem 12.5.1. Suppose that z = f (x, y) is a differentiable function of x and y, where x and y are both
differentiable functions of t. Then z is a differentiable function of t and
dz ∂ f dx ∂ f dy
= + or z′ (t) = fx x′ (t) + fy y′ (t).
dt ∂ x dt ∂ y dt

Example 12.5.2. If z = x2 y + 3xy4 , where x = sin(2t) and y = cost, find z′ (t) = dz

dt when t = 0.

A NSWER. For z = f (x, y) = x2 y + 3xy4 , we compute

fx = 2xy + 3y4 , fy = x2 + 12xy3 , x′ (t) = 2 cos(2t), y′ (t) = − sint

( ) ( )
z′ (t) = fx x′ + fy y′ = 2xy + 3y4 (2 cos(2t)) + x2 + 12xy3 (− sint)
( ) ( )
= 2 sin(2t) cost + 3 cos4 t (2 cos(2t)) + sin2 (2t) + 12 sin(2t) cos3 t (− sint)
( ) ( )
= 2 cos(2t) 2 sin(2t) cost + 3 cos4 t − sint sin2 (2t) + 12 sin(2t) cos3 t , z′ (0) = 6. □

Exercise 12.5.3. For z = f (x, y) = x2 ey , x(t) = t 2 − 1 and y(t) = sint, find z′ (t) = dz
dt .

To remember the Chain Rule, it’s helpful to draw the tree diagram in the figure 12.1. We draw branches from
the dependent variable z to the intermediate variables x and y to indicate that z is a function of x and y. Then
we draw branches from x and y to the independent variables t. On each branch we write the corresponding
partial derivative. To find ∂∂ zt , we find the product of the partial derivatives along each path from z to t and
then add these products:
dz ∂ f dx ∂ f dy
= +
dt ∂ x dt ∂ y dt
Example 12.5.4. Suppose the production of a firm is modeled by the Cobb–Douglas production function
P(k, l) = 20k1/4 l 3/4 , where k measures capital (in millions of dollars) and l measures the labor force (in
thousands of workers). Suppose that when l = 2 and k = 6, the labor force is decreasing at the rate of 20
workers per year and capital is growing at the rate of $400000 per year. Determine the rate of change of
production.

We can easily extend the Theorem 12.5.1 above to the case of a function f (x, y), where x and y are both
functions of the two independent variables s and t, x = x(s,t) and y = y(s,t).

Page 60 of 126
Calculus II for Engineers Fall, 2010

Figure 12.1: Tree Diagrams, Case 1 and Case 2

Chain Rule: Case 2

Theorem 12.5.5. Suppose that z = f (x, y) is a differentiable function of x and y, where x = x(s,t) and
y = y(s,t) are differentiable functions of s and t. Then

∂z ∂ f ∂x ∂ f ∂y ∂z ∂ f ∂x ∂ f ∂y
= + , and = +
∂s ∂x ∂s ∂y ∂s ∂t ∂x ∂t ∂y ∂t
i.e., zs = fx xs + fy ys , and zt = fx xt + fy yt .
∂z ∂z
Example 12.5.6. If z = f (x, y) = ex sin y, where x = st 2 and y = s2t, find ∂s and ∂t .

A NSWER. For z = f (x, y) = ex sin y, we compute

fx = ex sin y, fy = ex cos y, xs = t 2 , xt = 2st, ys = 2st, yt = s2

2( )
zs = fx xs + fy ys = ex sin y(t 2 ) + ex cos y(2st) = tex (t sin y + 2s cos y) = test t sin(s2t) + 2s cos(s2t)
2( )
zt = fx xt + fy yt = ex sin y(2st) + ex cos y(s2 ) = sex (2t sin y + s cos y) = sest 2t sin(s2t) + s cos(s2t) . □

Exercise 12.5.7. Suppose that z = f (x, y) = exy , x(u, v) = 3u sin v and y(u, v) = 4uv2 . Find the partial deriva-
tives ∂∂ uz and ∂∂ vz .

For the tree diagram in this Case of the Chain Rule, see the figure 12.1.

Exercise 12.5.8. Consider z = f (x, y), where x(r, θ ) = r cos θ and y(r, θ ) = y sin θ . Prove the followings.

1. fr = fx cos θ + fy sin θ .
2. frr = fxx cos2 θ + 2 fxy cos θ sin θ + fyy sin2 θ .
3. fxx + fyy = frr + frr + frθ2θ .

A slightly different use of a change of variables is demonstrated in the following example. An important
strategy in solving some equations is to first rewrite and solve them in the most general form possible. One
convenient approach to this is to convert to dimensionless variables. As the name implies, these are typically
combinations of variables such that all the units cancel. One example would be for an object with (one–
dimensional) velocity v ft/s and initial velocity v(0) = v0 ft/s. The variable V = vv0 is dimensionless because
the units of V are ft/s divided by ft/s, leaving no units. Often, a change to dimensionless variables will
simplify an equation.

Page 61 of 126
Calculus II for Engineers Fall, 2010

Example 12.5.9. An object moves in two dimensions according to the equations of motion: x′′ (t) = 0 and
y′′ (t) = −g, with initial velocity x′ (0) = v0 cos θ and y′ (0) = v0 sin θ and initial position x(0) = y(0) = 0.
(1) Rewrite the equations and initial conditions in terms of the variables
g g g
X = 2 x, Y = 2 y, T = t.
v0 v0 v0
(2) Show that the variables X, Y and T are dimensionless, assuming that x and y are given in feet and t in
seconds.
d2 y
A NSWER. (1) To transform the equations, we first need to rewrite the derivatives x′′ = d2 x
dt 2
and y′′ = dt 2
in
terms of X, Y and T . From the chain rule, we have
( ) ( )
dx dx dT d v20 X d gt v2 dX g dX
= = = 0 = v0 .
dt dT dt dT g dt v0 g dT v0 dT
Again, we must be careful computing the second derivative. We have
( ) ( ) ( ) ( )
d2 x d dx d dX d dX dT d2 X d gt d2 X g d2 X
= = v 0 = v 0 = v 0 = v0 = g .
dt 2 dt dt dt dT dT dT dt dT 2 dt v0 dT 2 v0 dT 2
By the same argument, we have
dy dY d2 y d2Y
= v0 , = g .
dt dT dt 2 dT 2
Hence, the differential equations x′′ (t) = 0 and y′′ (t) = −g become, respectively,
d2 X d2Y d2 X d2Y
g = 0, g = −g, i.e., = 0, = −1.
dT 2 dT 2 dT 2 dT 2
The initial conditions x′ (0) = v0 cos θ and y′ (0) = v0 sin θ become, respectively,
   
dX  dY  dX  dY 
v0 = v0 cos θ , v0 = v0 sin θ , i.e., = cos θ , = sin θ .
dT T =0 dT T =0 dT T =0 dT T =0
The initial value problem is now
 
d2 X d2Y dX  dY 
= 0, = −1, = cos θ , = sin θ , X(0) = 0, Y (0) = 0.
dT 2 dT 2 dT T =0 dT T =0
g
(2) We observe X = v20
x has the units:

ft/s2 ft2 /s2

ft = = 1.
(ft/s)2 ft2 /s2
Similarly Y and T have no units. □

□ 12.5.2 Implicit Differentiation.

We suppose that an equation F(x, y) = 0 of the form defines y implicitly as a differentiable function of x, that
is, y = f (x), where F(x, f (x)) = 0 for all x in the domain of f . If F is differentiable, we can apply the Chain
Rule, Case 1, 12.5.1 to differentiate both sides of the equation F(x, y) = 0 with respect to x. Since both x and
y are functions of x, we obtain
∂F
∂ F dx ∂ F dy
= − ∂x = − ,
dy Fx
+ = 0, simply,
∂ x dx ∂ y dx dx ∂F Fy
∂y
assuming Fy ̸= 0.

Page 62 of 126
Calculus II for Engineers Fall, 2010

Theorem 12.5.10 (I MPLICIT D IFFERENTIATION).

dy
1. For the equation F(x, y) = 0, we can find dx as follows:

dy Fx
=−
dx Fy

∂z ∂z
2. For the equation F(x, y, z) = 0, we can find ∂x and ∂y as follows:

∂z Fx ∂z Fy
=− and =−
∂x Fz ∂y Fz

Example 12.5.11. Find y′ (x) = dy/ dx if x3 + y3 = 6xy.

A NSWER. The given equation can be written as

x3 + y3 − 6xy = 0.

Let F(x, y) = x3 + y3 − 6xy. Then we have

F(x, y) = x3 + y3 − 6xy = 0.

Using the formula above, we get

dy Fx 3x2 − 6y x2 − 2y
=− =− 2 =− 2 . □
dx Fy 3y − 6x y − 2x
∂z ∂z
Exercise 12.5.12. Find ∂x and ∂y if x3 + y3 + z3 + 6xyz = 1.

Page 63 of 126
Calculus II for Engineers Fall, 2010

§12.6 Gradient and Directional Derivatives.

For the animations or plots involved with the topics to be discussed here, please confer the following webpage:
http://www.usd.edu/˜jflores/MultiCalc02/Contents.htm
Explicitly, for the level curves, “Functions of Several Variables” in the webpage may be useful. For the
gradient and directional derivatives, “Directional Derivatives and the Gradient Vector” in the webpage may
be helpful.

□ 12.6.1 Functions of Two Variables.

In Calculus I, we learned the instantaneous rate of change f ′ (a) of the function f (x) of one variable at x = a.
It means the slope of the tangent line to the graph of f (x) at x = a.
In Calculus II, we deal with the function f (x1 , x2 , . . . , xn ) of several variables. For such a function, how can
we define the instantaneous rate of change? What is the meaning of the rate of change?
A graph of y = f (x) is given in the two–dimensional space. At each point (a, f (a)) on the graph, we can
draw one tangent line with the slope f ′ (a).
The surface of z = f (x, y) is given in the three–dimensional space. At each point of the surface, we can
find one tangent plane. Since a plane contains infinitely many lines, at each point on the surface, we can
think of infinitely many tangent lines. We may say the instantaneous rate of change of f (x, y) at the point
P(a, b, f (a, b)) on the surface corresponds to the slope of the tangent line in the case of two–dimensional
space. However, since there are so many tangent lines, we should choose only one by assigning the direction
of the line.
We give a special name to the instantaneous rate of change of f (x, y) at the point P(x, y) and in the direction
of the unit vector u.

Definition 12.6.1. The directional derivative of f (x, y) at the point (a, b) and in the direction of the unit
vector u = ⟨u1 , u2 ⟩ is given by

f (a + hu1 , b + hu2 ) − f (a, b)

Du f (a, b) = lim ,
h→0 h
provided the limit exists.

When u = ⟨1, 0⟩, the directional derivative becomes


f (a + h, b) − f (a, b) ∂ f (x, y) 
Du f (a, b) = lim = = fx (a, b),
h→0 h ∂ x (x,y)=(a,b)

which is the value of the partial derivative of f (x, y) with respect to x at (x, y) = (a, b).
When u = ⟨0, 1⟩, the directional derivative becomes

f (a, b + h) − f (a, b) ∂ f (x, y) 
Du f (a, b) = lim = = fy (a, b),
h→0 h ∂ y (x,y)=(a,b)

which is the value of the partial derivative of f (x, y) with respect to y at (x, y) = (a, b).
In Calculus I, even if we learned the limit definition of the derivative, we did not use the limit definition to
compute the derivative. Rather than the limit definition, we applied for the formulas. By the same argument,
we need formulas to get the directional derivatives easily.

Theorem 12.6.2. Suppose that f is differentiable at (a, b) and u = ⟨u1 , u2 ⟩ is any unit vector. Then we can
write
Du f (a, b) = fx (a, b)u1 + fy (a, b)u2 .

The proof of this formula is based on the chain rule.

Page 64 of 126
Calculus II for Engineers Fall, 2010

Example 12.6.3. Compute the directional derivative at the ⟨ given point

⟩ in the direction of the indicated vector.
(1) f (x, y) = x y − 4y at (2, −1) in the direction of u = √2 , √2 .
3 2 1 1
√
(2) g(x, y) = x2 + y2 at (3, −4) with u in the direction of ⟨3, −2⟩.
(3) h(x, y) = sin(2x − y) at (π , 0) with u in the direction from (π , 0) to (2π , π ).

A NSWER. We use the Theorem 12.6.2 above.

(1) It is easy to get fx = 3x2 y and fy = x3 − 8y and so

1 1 −12 + 16 √
Du f (2, −1) = fx (2, −1) √ + fy (2, −1) √ = √ = 2 2.
2 2 2

(2) The unit vector u in the direction of ⟨3, −2⟩ is obtained as follows:

⟨3, −2⟩ ⟨3, −2⟩

u= = √ .
∥ ⟨3, −2⟩ ∥ 13
Using this unit vector, we get the directional derivative:
x x y y
gx (x, y) = =√ , gy (x, y) = =√ ,
g(x, y) x 2 + y2 g(x, y) x 2 + y2
3 4
gx (3, −4) = , gy (3, −4) = − ,
5 5
3 −2 9+8 17
Du g(3, −4) = gx (3, −4) √ + gy (3, −4) √ = √ = √ .
13 13 5 13 5 13

(3) The vector in the direction from (π , 0) to (2π , π ) is ⟨2π − π , π − 0⟩ = ⟨π , π ⟩. The unit vector u in the
direction of ⟨π , π ⟩ = π ⟨1, 1⟩ is obtained as follows:

π ⟨1, 1⟩ ⟨1, 1⟩
u= = √ .
∥π ⟨1, 1⟩ ∥ 2
Using this unit vector, we get the directional derivative:

hx (x, y) = 2 cos(2x − y), hy (x, y) = − cos(2x − y), hx (π , 0) = 2, hy (π , 0) = −1,

1 1 2−1 1
Du h(π , 0) = hx (π , 0) √ + hy (π , 0) √ = √ = √ . □
2 2 2 2
For convenience, we define the gradient of a function to be the vector–valued function whose components
are the first–order partial derivatives of f . We denote the gradient of a function f by grad f or ∇ f (read “del
f ”).

Definition 12.6.4. The gradient of f (x, y) is the vector–valued function

⟨ ⟩
∂f ∂f ∂f ∂f ⟨ ⟩
∇ f (x, y) = , = i+ j, ∇ f (x, y) = fx , fy = fx i + fy j,
∂x ∂y ∂x ∂y

provided both partial derivatives exist.

Remark 12.6.5.

1. Recall from Section 11.2 The Calculus of VectorValued Functions: For the vector–valued function
r(t) = ⟨ f (t), g(t)⟩ of one variable t, we can get the derivative r′ (t) = ⟨ f ′ (t), g′ (t)⟩, which is the tangent
vector.

Page 65 of 126
Calculus II for Engineers Fall, 2010

2. The definition above is for the vector–valued function f (x, y) of two variables x and y:
⟨ ⟩
∇ f (x, y) = fx , fy .

When using the gradient, the Theorem 12.6.2 for the directional derivative can be simplified as the dot product
of the gradient and the unit vector in the direction of interest. For any unit vector u = ⟨u1 , u2 ⟩,
⟨ ⟩
Du f (x, y) = fx u1 + fy u2 = fx , fy · ⟨u1 , u2 ⟩ = ∇ f (x, y) · u.
Thus we have the following theorem.
Theorem 12.6.6. If f is a differentiable function of x and y and u is any unit vector, then
Du f (x, y) = ∇ f (x, y) · u.
Writing directional derivatives as a dot product has many important consequences.
Example 12.6.7. For f (x, y) = (x − y)2 , compute the directional derivative of f (x, y) at the point (2, 1) in the
direction of the indicated vector.
(1) u in the direction of v = ⟨2, −1⟩.
(2) u in the direction of w = ⟨−2, 1⟩.
A NSWER. First, we compute the gradient of f :
⟨ ⟩
∇ f (x, y) = fx , fy = ⟨2(x − y), −2(x − y)⟩ = 2(x − y) ⟨1, −1⟩ .
So at the point (2, 1), we have ∇ f (2, 1) = 2 ⟨1, −1⟩.
(1) The unit vector u in the direction of v = ⟨2, −1⟩ is obtained as follows:
v ⟨2, −1⟩
u= = √ .
∥v∥ 5
By the Theorem 12.6.6,
⟨2, −1⟩ 6
Du f (2, 1) = 2 ⟨1, −1⟩ · √ =√ .
5 5
(2) The unit vector u in the direction of w = ⟨−2, 1⟩ is obtained as follows:
w ⟨−2, 1⟩
u= = √ .
∥w∥ 5
By the Theorem 12.6.6,
⟨−2, 1⟩ 6
Du f (2, 1) = 2 ⟨1, −1⟩ · √ = −√ . □
5 5
What is the graphical interpretation of the directional derivative? What does it mean by Du f (2, 1) = √65 ?
Simply speaking, it represents the slope of the surface (i.e., slope of the tangent line to the surface) at the
point.
√ surface of z = f (x, y) = (x − y) . In the case of (1) of the example above, the unit
The point (2, 1, 1) is on the 2

vector u is u = ⟨−2, 1⟩ / 5 . Based on these two pieces of information, we consider a plane S such that (i)
it contains the point √(2, 1, 1), (ii) it is perpendicular to the xy–plane and finally (iii) it is parallel to the unit
vector u = ⟨−2, 1⟩ / 5 . In this plane S, we introduce a new coordinate system having the direction of u as
the positive horizontal axis and (2, 1) as the origin.
Then, we have the two–dimensional graph of z = f (x, y) on this plane S with the coordinate system. The
slope of the tangent line to this image graph at the origin is Du f (2, 1) = √65 . For the explicit explanation,
please confer the textbook.
We can use a contour plot to estimate the value of a directional derivative. We refresh our memory on the
level curve and the contour plot of f (x, y).

Page 66 of 126
Calculus II for Engineers Fall, 2010

Definition 12.6.8. A level curve or contour curve of f (x, y) is the (two–dimensional) graph of the equation
f (x, y) = c, for some constant c. A contour plot of f (x, y) is a graph of numerous level curves f (x, y) = c,
for representative values of c.
√
Example 12.6.9. The level curves of the function f (x, y) = 9 − x2 − y2 for c = 0, 1, 2, 3 are
√
9 − x 2 − y2 = c i.e., x2 + y2 = 9 − c2 ,
√
which is a family of concentric circles with center
√ √ (0, 0) and radius 9 − c2 . Since c = 0, 1, 2, 3, we have
total four level curves (circles) with radius 3, 8 , 5 , 0, respectively.

For more information on the level curve and the contour curve, please read Section 12.1 Functions of Several
Variables.
√
Example 12.6.10. Use a contour plot of z = f (x, y) = x2 + y2 to estimate Du f (1, −1) for u = ⟨−3, 4⟩ / 5 .

Figure 12.2: Contour Plot of z = x2 + y2

2 2
Answer. √ √plot of f (x, y) = x + y for c = 0.2, 1, 2, 3 is the set of circles with center (0, 0) and
√ The contour
radius 0.2 , 1, 2 , 3 , respectively. See the figure 12.2. Using the graph, we approximate the directional
△z
derivative by estimating △u , where △u is the distance traveled along the unit vector u.
( ) −→
(1) When △u = 1, the terminal point of u with the initial point P(1, −1) is Q 52 , − 15 . (Why? PQ =
−→
⟨−3, 4⟩ /5, which is in( the same
) direction as u and ∥ PQ∥ = 1.)
The terminal point Q 5 , − 5 is on the level curve z = x2 + y2 = 0.2, because (2/5)2 + (−1/5)2 = 1/5 = 0.2.
2 1

Thus, when traveling along the vector u as much as △u = 1, the vector appears to extend from z = 2 level
curve (where the initial point P lies on) to z = 0.2 level curve (where the terminal point Q lies on). That is,

△z 0.2 − 2
= = −1.8.
△u 1
( √ √
)
2(2+3 6 ) 3−8 6
(2) We observe that the point R 25 , 25 = R (0.7479, −0.6638) is on the z = 1 level curve and
−
→
the vector PR is in the same direction as u. Here,
⟨ ( √ ) ( √ )⟩
−→ 3 −7 + 3 6 −4 −7 + 2 6
PR = , = ⟨0.0418, 0.3361⟩ .
25 25

Page 67 of 126
Calculus II for Engineers Fall, 2010

So, when the vector extends from z = 2 level curve (where the initial point P√lies on) to z = 1 level curve
−
→
(where the initial point R lies on), the distance traveled along u is ∥PR∥ = 7−25 6 ≈ 0.4202. That is, in this
case, we have
△z 1−2
= ≈ −2.3798.
△u 0.4202
(3) If we use the Theorem 12.6.6, we have Du f (1, 1) = ∇ f (1, 1) · u and so Du f (1, 1) = −14/5 = −2.8. □
Keep in mind that a directional derivative gives the rate of change of a function in a given direction.
Remark 12.6.11. We raise some questions.

1. Calculus I: at what point does a given function have the horizontal tangent line and the vertical tangent
line? The answer is involved with the critical numbers and the local extrema.
2. Calculus II: in what direction does a given function have its maximum or minimum rate of increase?

Let us answer to the Question 2. Combining the Theorem 10.3.5 in Section 10.3 with the Theorem 12.6.6,
we deduce

Du f (a, b) = ∇ f (a, b) · u = ∥∇ f (a, b)∥∥u∥ cos θ = ∥∇ f (a, b)∥ cos θ −→ Du f (a, b) = ∥∇ f (a, b)∥ cos θ ,

where θ is the angle between ∇ f (a, b) and the directional unit vector u. Now we can say the followings:

1. The maximum value ∥∇ f (a, b)∥ of Du f (a, b) occurs when cos θ = 1, i.e., θ = 0, i.e., ∇ f (a, b) is in
the same direction as u. It implies
∇ f (a, b)
u= .
∥∇ f (a, b)∥
2. The minimum value −∥∇ f (a, b)∥ of Du f (a, b) occurs when cos θ = −1, i.e., θ = π , i.e., ∇ f (a, b) is
in the opposite direction as u. It implies
∇ f (a, b)
u=− .
∥∇ f (a, b)∥

3. The zero value Du f (a, b) = 0 occurs when cos θ = 0, i.e., θ = π /2, i.e., ∇ f (a, b) is orthogonal to u.
4. Recall that the level curves are curves in the xy–plane on which f is constant. So at any point on the
level curve, the directional derivative becomes zero. Why? For the level curve f (x, y) = c, c constant,
we introduce a parameter t so that the variables x and y become functions of t and thus the equation
turns to be f (x(t), y(t)) =⟨ c. If ⟩we differentiate the equation with respect to t, the Chain Rule implies
fx x′ (t) + fy y′ (t) = 0, i.e., fx , fy · ⟨x′ (t), y′ (t)⟩ = 0. So we deduce
⟨ ⟩ ⟨x′ (t), y′ (t)⟩
Du f (x, y) = fx , fy · u = 0, where u = .
∥ ⟨x′ (t), y′ (t)⟩ ∥
Thus, for this reason, the directional derivative becomes zero.
Since ⟨x′ (t), y′ (t)⟩ is the derivative of ⟨x(t), y(t)⟩, so ⟨x′ (t), y′ (t)⟩ is in fact the tangent vector of the curve.
That is, at any point (x, y), ∇ f (x, y) should be orthogonal to the tangent line to the level curve. (Aside:
In Calculus I and II, “angle between a line/vector and a graph” means “angle between a line/vector and
the tangent line to the graph.” So “∇ f (a, b) is orthogonal to the tangent line to the level curve” can be
simply said “∇ f (a, b) is orthogonal to the level curve”.) See the figure 12.3.

We summarize the answers in the following Theorem.

Theorem 12.6.12. Suppose that f is a differentiable function of x and y at the point (a, b). Then

1. the maximum rate of change of f at (a, b) is ∥∇ f (a, b)∥, occurring in the direction of the gradient,

Page 68 of 126
Calculus II for Engineers Fall, 2010

Figure 12.3: Level Curves of f (x, y) = 5 + x2 + 2y, Yellow: Tangent Line at (1, −3), Blue: Gradient of f (x, y)
at (1, −3)

2. the minimum rate of change of f at (a, b) is −∥∇ f (a, b)∥, occurring in the opposite direction of the
gradient,
3. the rate of change of f at (a, b) is 0 in the direction orthogonal to ∇ f (a, b),
4. the gradient ∇ f (a, b) is orthogonal to the level curve f (x, y) = c at the point (a, b), where c = f (a, b).

Remember that the directional derivative corresponds to the rate of change of the function f (x, y) in the given
direction.
Example 12.6.13. Find the directions of maximum and minimum change of f at the given point and the
values of the maximum and minimum rates of change.
(1) f (x, y) = y2 e4x at (3, −1)
(2) g(x, y) = x cos(3y) at (−2, π ).
A NSWER. (1) We compute the gradient vector first:
⟨ ⟩ ⟨ ⟩
∇ f (x, y) = fx , fy = 4y2 e4x , 2ye4x = 2ye4x ⟨2y, 1⟩ , ∇ f (3, −1) = −2e12 ⟨−2, 1⟩ .
By the Theorem 12.6.12, we deduce
(i) the direction of maximum change is ∇ f (3, −1) = −2e 12 ⟨−2, 1⟩ and the maximum rate of change along
√
this direction is ∥∇ f (3, −1)∥ = ∥ − 2e12 ⟨−2, 1⟩ ∥ = 2e12 5 ,
(ii) the direction of minimum change is √ −∇ f (3, −1) = 2e12 ⟨−2, 1⟩ and the minimum rate of change along
this direction is −∥∇ f (3, −1)∥ = −2e12 5 .
(2) We compute the gradient vector first:
⟨ ⟩
∇g(x, y) = gx , gy = ⟨cos(3y), −3x sin(3y)⟩ , ∇g(−2, π ) = ⟨cos(3π ), 6 sin(3π )⟩ = ⟨−1, 0⟩ .
By the Theorem 12.6.12, we deduce
(i) the direction of maximum change is ∇g(−2, π ) = ⟨−1, 0⟩ and the maximum rate of change along this
direction is ∥∇g(−2, π )∥ = ∥ ⟨−1, 0⟩ ∥ = 1,
(ii) the direction of minimum change is −∇g(−2, π ) = ⟨1, 0⟩ and the minimum rate of change along this
direction is −∥∇g(−2, π )∥ = −1. □
Note that the gradient vector ∇ f (a, b) gives the direction of fastest increase (i.e., maximum rate of change)
of f .
Example 12.6.14. Find the direction of maximum increase of the function f (x, y) = 3x − x3 − 3xy2 from the
point A(0.6, −0.7) and sketch the path of steepest ascent.

Page 69 of 126
Calculus II for Engineers Fall, 2010

A NSWER. By the Theorem 12.6.12, the maximum increase occurs in the direction ∇ f (0.6, −0.7):
⟨ ⟩ ⟨ ⟩ ⟨ ⟩
∇ f (x, y) = fx , fy = 3 − 3x2 − 3y2 , −6xy = 3 ⟨1, 0⟩ − 3 x2 + y2 , 2xy ,
∇ f (0.6, −0.7) = 3 ⟨1, 0⟩ − 3 ⟨0.85, −0.84⟩ = ⟨0.45, 2.52⟩ .
The unit vector u in the direction ∇ f (0.6, −0.7) is
∇ f (0.6, −0.7) ⟨0.45, 2.52⟩
u= = = ⟨0.1758, 0.9844⟩ .
∥∇ f (0.6, −0.7)∥ ∥ ⟨0.45, 2.52⟩ ∥
We observe the point A0 (0.6, −0.7) is on the level curve z = f (x, y) = 0.702. A simple computation shows for
−−→
the point A1 (0.6623, −0.5920) on the level curve z = f (x, y) = 1, A0 A1 = ⟨0.0623, 0.1080⟩ is in the direction
of u. We do the same argument with the new point A1 and find A2 . Repeating the steps several times, we
deduce the following table:
f (x, y) = c A0 A1 A2 A3 A4 A5
0.702 (0.6, −0.7)
1 (0.6623, −0.5920)
1.3 (0.7429, −0.4824)
1.6 (0.8437, −0.3614)
1.9 (1.0045, −0.1821)
2 (1, 0)

Connecting each point on each level curve, we have the path of steepest ascent which remains perpendicular
to each level curve through which it passes. Finding an equation for the path of steepest ascent is challenging.
See the figure 12.4. □

Figure 12.4: At each point Ai , the gradient is perpendicular to each level curve

Remark 12.6.15 (A SIDE). In the Example above, we used the level curves between which the distance was
−−−−→
fixed 0.3. Due to the fixed distance, each vector Ai Ai+1 does not have the magnitude 1. Suppose that the
distance between given level curves L and M is more than 1. Then, how can we find the path of steepest
ascent between L and M? I will explain this in class. If you take a look at the exercises in the textbook, you
will see the problems asking to find the path of steepest ascent from the given contour plot (i.e., plot of level
curves). From the starting point A0 on the level curve L0 , we move to the direction of the gradient and find
the point A1 on the level curve L1 . Keeping this argument, we can get the desired path.

Page 70 of 126
Calculus II for Engineers Fall, 2010

□ 12.6.2 Functions of Three Variables.

Most of the results above extend easily to functions of any number of variables.
Definition 12.6.16. The directional derivative of f (x, y, z) at the point (a, b, c) and in the direction of the
unit vector u = ⟨u1 , u2 , u3 ⟩ is given by
f (a + hu1 , b + hu2 , c + hu3 ) − f (a, b, c)
Du f (a, b, c) = lim ,
h→0 h
provided the limit exists.
The gradient of f (x, y, z) is the vector–valued function
⟨ ⟩
∂f ∂f ∂f ∂f ∂f ∂f
∇ f (x, y, z) = , , = i+ j+ k,
∂x ∂y ∂z ∂x ∂y ∂z
provided all the partial derivatives are defined.
We extend the Theorem 12.6.6 on the directional derivative of functions of two variables as follows:
Theorem 12.6.17. If f is a differentiable function of x, y and z and u is any unit vector, then

Du f (x, y, z) = ∇ f (x, y, z) · u.

As for the functions of two variables, we have

Du f (x, y, z) = ∇ f (x, y, z) · u = ∥∇ f (x, y, z)∥∥u∥ cos θ = ∥∇ f (x, y, z)∥ cos θ ,

where θ is the angle between the vectors ∇ f (x, y, z) and u. Also by the same argument, the direction of
maximum increase (i.e., fastest increase) at any point is given by the gradient at that point.
Example 12.6.18. The temperature at a point (x, y, z) is given by

T (x, y, z) = 200e−x
2 −3y2 −9z2
,

where T is measured in ◦ C (centigrade) and x, y, z in meters.

(1) Find the rate of change of temperature at the point P(2, −1, 2) in the direction toward the point Q(3, −3, 3).
(2) In what direction does the temperature increase fastest at P(2, −1, 2)?
(3) Find the maximum rate of increase at P(2, −1, 2).
A NSWER. Since the main idea to solve the problems is the gradient, we first compute the gradient of T :
⟨ ⟩
∇T (x, y, z) = Tx , Ty , Tz = T (x, y, z) ⟨−2x, −6y, −18z⟩ = −2T (x, y, z) ⟨x, 3y, 9z⟩
∇T (1, −2, 1) = −2T (1, −2, 1) ⟨1, −6, 1⟩ = −400e−22 ⟨1, −6, 1⟩ .

• Useful technique: We recall that g(x) = eh(x) has the derivative

( )′
g′ (x) = eh(x) = eh(x) h′ (x) = g(x)h′ (x).
−→
(1) To get the desired rate of change, we need the unit vector u in the direction of the vector PQ = ⟨1, −2, 1⟩,
−→
PQ ⟨1, −2, 1⟩ ⟨1, −2, 1⟩
u = −→ = = √ .
∥PQ∥ ∥ ⟨1, −2, 1⟩ ∥ 6

Thus by the Theorem 12.6.17, the rate of change of T at P(1, −2, 1) in the direction of u is as follows:
⟨1, −2, 1⟩ 5600
Du T (1, −2, 1) = ∇T (1, −2, 1) · u = −400e−22 ⟨1, −6, 1⟩ · √ =− √ .
6 6

Page 71 of 126
Calculus II for Engineers Fall, 2010

(2) The direction of the fastest increase is ∇T (1, −2, 1) = −400e−22 ⟨1, −6, 1⟩, exactly like the result stated
in the Theorem 12.6.12 on the function of two variables.
(3) Again, exactly like the result stated in the Theorem 12.6.12 on the function of two variables, the maximum
rate of increase is the norm of ∇T (1, −2, 1):
√
∥∇T (1, −2, 1)∥ = −400e−22 ⟨1, −6, 1⟩ = 400e−22 ∥ ⟨1, −6, 1⟩ ∥ = 400 38 e−22 ≈ 6.8782 × 10−7 . □

We generalize the argument on the level curve in Remark 12.6.11 (4) to the function of three variables.
Recall that for some constant c, the equation z = f (x, y) = c defines a level curve of the function z = f (x, y)
with the height/level c on the xy–plane.
For some constant k, the equation w = f (x, y, z) = k defines a level surface of the function w = f (x, y, z) with
the height/level k in the xyz–space. Suppose that u is any unit vector lying in the tangent plane to the level
surface w = f (x, y, z) = k at a point (a, b, c) on the level surface. Then by the same argument developed in
Remark 12.6.11, we have
0 = Du f (a, b, c) = ∇ f (a, b, c) · u.
This occurs when ∇ f (a, b, c) is orthogonal to u. Since u is any unit vector on the tangent plane, it implies
∇ f (a, b, c) should be the normal vector to the tangent plane at the point (a, b, c). Now that we have the normal
vector to the tangent plane and the point (a, b, c) on the plane is given, we can find the equation of the tangent
plane at the point (a, b, c):

0 = ∇ f (a, b, c) · ⟨x − a, y − b, z − c⟩
⟨ ⟩
= fx (a, b, c), fy (a, b, c), fz (a, b, c) · ⟨x − a, y − b, z − c⟩
= fx (a, b, c)(x − a) + fy (a, b, c)(y − b) + fz (a, b, c)(z − c),
i.e., 0 = fx (a, b, c)(x − a) + fy (a, b, c)(y − b) + fz (a, b, c)(z − c).

Recall from Section 10.5 Lines and Planes in Space that the equation of the plane containing a point (a, b, c)
with normal vector ⟨α , β , γ ⟩ (meaning this vector is normal/orthogonal/perpendicular to the plane) is

0 = ⟨α , β , γ ⟩ · ⟨x − a, y − b, z − c⟩ = α (x − a) + β (y − b) + γ (z − c),
i.e., 0 = α (x − a) + β (y − b) + γ (z − c).

We summarize the argument in the following Theorem.

Theorem 12.6.19. Suppose that f (x, y, z) has continuous partial derivatives at the point (a, b, c) and also
∇ f (a, b, c) ̸= 0. Then ∇ f (a, b, c) is a normal vector to the tangent plane to the surface f (x, y, z) = k, at the
point (a, b, c). Further, the equation of the tangent plane is

0 = fx (a, b, c)(x − a) + fy (a, b, c)(y − b) + fz (a, b, c)(z − c).

Definition 12.6.20. The normal line to the level surface at the point (a, b, c) on the surface is the line
passing through the⟨ point (a, b, c) and perpendicular ⟩ to the tangent plane, i.e., parallel to the normal vec-
tor ∇ f (a, b, c) = fx (a, b, c), fy (a, b, c), fz (a, b, c) .

We observe that the normal line to the level surface of w = f (x, y, z) passing through the point (a, b, c) has the
following parametric equations:

x = a + fx (a, b, c)t, y = b + fy (a, b, c)t, z = c + fz (a, b, c)t. (12.6.1)

Example 12.6.21. Find the equation of the tangent plane and the normal line at the point (−2, 1, −3) to the
ellipsoid
x2 z2
+ y2 + = 3.
4 9

Page 72 of 126
Calculus II for Engineers Fall, 2010

A NSWER. The ellipsoid is the level surface (with k = 3) of the function

x2 z2
f (x, y, z) = + y2 + .
4 9
Therefore, we have
⟨ ⟩ ⟨ ⟩
⟨ ⟩ x 2z 2
∇ f (x, y, z) = fx , fy , fz = , 2y, , and ∇ f (−2, 1, −3) = −1, 2, − .
2 9 3
The Theorem 12.6.19 gives the equation of the tangent plane at (−2, 1, −3) as
2
−1(x + 2) + 2(y − 1) − (z + 3) = 0, i.e., 3x − 6y + 2z + 18 = 0.
3
By the equation (12.6.1) just above the example, parametric equations of the normal line are
2
x = −2 − t, y = 1 + 2t, z = −3 − t.
3
See the figure 12.5. □

x2 2
Figure 12.5: Ellipsoid 4 + y2 + z9 = 3, Tangent Plane and Normal Line
As the last technique in this section, we use the argument developed for the tangent plane to the level surface
w = g(x, y, z) = k to find the tangent plane to the surface of z = f (x, y).
The surface of z = f (x, y) can be viewed as the level surface of the new function w = g(x, y, z) = f (x, y) − z
with the height/level 0. By the Theorem 12.6.19, for the new function w = g(x, y, z), the equation of the
tangent plane at the point (a, b, f (a, b)) on the level surface is obtained as follows:
0 = ∇g(a, b, f (a, b)) · ⟨x − a, y − b, z − f (a, b)⟩
⟨ ⟩
= gx (a, b, f (a, b)), gy (a, b, f (a, b)), gz (a, b, f (a, b)) · ⟨x − a, y − b, z − f (a, b)⟩
= gx (a, b, f (a, b))(x − a) + gy (a, b, f (a, b))(y − b) + gz (a, b, f (a, b))(z − f (a, b)).
But, since gx = fx , gy = fy and gz = −1, so the equation becomes
0 = fx (a, b)(x − a) + fy (a, b)(y − b) − (z − f (a, b)),
i.e., z = fx (a, b)(x − a) + fy (a, b)(y − b) + f (a, b).

Page 73 of 126
Calculus II for Engineers Fall, 2010

Example 12.6.22.√Find an equation of the tangent plane to the surface of each function at (1, 1, 2).
(1) z = f (x, y) = 9√− 3x2 − 2y2
(2) z = g(x, y) = 4 − −x2 − y2 + 8x + 6y − 8

A NSWER. We start with the gradient of f and g:

⟨ ⟩ ⟨3x, 2y⟩ ⟨3, 2⟩ ⟨3, 2⟩
∇ f (x, y) = fx , fy = − , ∇ f (1, 1) = − =− ,
f (x, y) f (1, 1) 2
⟨ ⟩ ⟨x − 4, y − 3⟩ ⟨−3, −2⟩ ⟨3, 2⟩
∇g(x, y) = gx , gy = , ∇g(1, 1) = =− .
4 − g(x, y) 4 − g(1, 1) 2
√
• Useful technique: We recall that h(x) = k(x) has the derivative
(√ )′ k′ (x) k′ (x)
′
h (x) = k(x) = √ = .
2 k(x) 2h(x)

(1) By the argument preceding this example, the equation of the tangent plane is
3 3
z = fx (1, 1)(x − 1) + fy (1, 1)(y − 1) + f (1, 1) = − (x − 1) − (y − 1) + 2 = − (x − 3) − y.
2 2
(2) By the argument preceding this example, the equation of the tangent plane is
3 3
z = gx (1, 1)(x − 1) + gy (1, 1)(y − 1) + g(1, 1) = − (x − 1) − (y − 1) + 2 = − (x − 3) − y,
2 2
√
which is exactly same as the one for (1) z = f (x, y) = ⟨9 − 3x ⟩ − 2y .
2 2
How do we get this result? It’s because ∇ f (1, 1) = − 2 , 1 = ∇g(1, 1). Thus, f and g share the tangent
3

plane z = − 32 (x − 3) − y, i.e., 3x + 2y + 2z = 9 at the point (1, 1, 2). See the figure 12.6. □

Figure 12.6: Upper Ball: Surface of w = g(x, y), Lower Ellipsoid: Surface of z = f (x, y)
Just as it is important to constantly think of ordinary derivatives as slopes of tangent lines and as instantaneous
rate of change, it is crucial to keep in mind at all times the interpretations of gradients. Always think of
gradients as vector–valued functions whose values specify the direction of maximum increase of a function
and whose values provide normal vectors (to the level curves in two dimensions and to the level surfaces in
three dimensions).

Page 74 of 126
Calculus II for Engineers Fall, 2010

§12.7 Extrema of Functions of Several Variables.

□ 12.7.1 Definition and Second Derivatives Test.

Figure 12.7: Surface of z = f (x, y) = xe−x

2 /2−y3 /3+y
, Blue Dot: Local Maximum, Red Dot: Local Minimum

We start with the surface z = f (x, y) = xe−x /2−y /3+y . We observe that the graph has the peak at the blue
2 3

dot (1, 1, e1/6 ) and a valley at the red dot (−1, 1, −e1/6 ). Such points are referred as local extrema, which are
mathematically defined as follows.

Definition 12.7.1. We call f (a, b) a local maximum of f if there is an open disk R centered at (a, b), for
which f (a, b) ≥ f (x, y) for all (x, y) ∈ R. Similarly, f (a, b) is called a local minimum of f if there is an open
disk R centered at (a, b), for which f (a, b) ≤ f (x, y) for all (x, y) ∈ R. In either case, f (a, b) is called a local
extremum of f .

The idea here is the same as the local extrema for the functions of one variable, i.e., if f (a, b) ≥ f (x, y) for all
(x, y) “near” (a, b), we call f (a, b) a local maximum. If we look at the peak and valley carefully, we observe
at both points, the tangent plane is horizontal to the xy–plane. It implies that at these points, the gradient of f
should be the zero vector. For this reason, we define the critical point of f (x, y) as follows. (We can think of
this argument as the generalization of the one on the critical number for a function of one variable.)

Definition 12.7.2. The point (a, b) is a critical point of the function f (x, y) if the following conditions are
satisfied:

1. (a, b) is the domain of f , and

2. either fx (a, b) = 0 and fy (a, b) = 0, or
3. one or both of fx (x, y) and fy (x, y) do not exist at (a, b).

The second condition above corresponds to f ′ (c) = 0 in Calculus I. The third condition above corresponds to
the one that f ′ (c) is undefined in Calculus I.
Recall from the Calculus I that the critical numbers are the candidates for local extrema:

1. Even if f (x) has a critical number x = c, f (x) may or may not have a local extremum at x = c.(Examples:
f (x) = x2 has a critical number x = 0 and the local minimum at x = 0. g(x) = x3 has a critical number at
x = 0 but it does not have a local extremum at x = 0.)
2. If f (x) has a local extremum at x = c, then x = c should be a critical number of f (x).

For the function f (x, y) of two variables, we can develop the exactly same argument.

Theorem 12.7.3. If f (x, y) is a local extremum at (a, b), then (a, b) must be a critical point of f (x, y).

Page 75 of 126
Calculus II for Engineers Fall, 2010

Example 12.7.4. Find all critical points of f (x, y) = xe−x

2 /2−y3 /3+y
and analyze each critical point graphi-
cally.
A NSWER. We start with the first partial derivatives:

fx = (1 − x2 )e−x fy = x(1 − y2 )e−x

2 2−y3 /3+y 2 2−y3 /3+y
, .

 Useful technique: g(x) = h(x)ek(x) has the derivative,

( )′ ( )
g′ = hek = h′ ek + hk′ ek = ek h′ + hk′ .

Since exponentials are always positive, we get

(1) fx = (1 − x2 )e−x /2−y /3+y = 0 −→ 1 − x2 = 0 −→ x = ±1.
2 3

(2) fy = x(1 − y2 )e−x /2−y /3+y = 0 −→ x(1 − y2 ) = 0 −→ x = 0 or y = ±1.

2 3

But if x = 0, then fx ̸= 0 and so there are no critical points with x = 0. That is, all critical points are
combinations of x = ±1 and y = ±1:

(1, 1), (1, −1), (−1, 1), (−1, −1).

From the figure 12.7, we observe that at (1, 1), f (x, y) has the local maximum value f (1, 1) = e1/6 , while at
(−1, 1), f (x, y) has the local minimum value f (−1, 1) = −e1/6 .
What can we say at (1, −1) and (−1, −1)? See the figure 12.8.
(1) At (1, −1): Observe the intersection of the graph of f (x, y) and the plane x = 1. The intersection is a curve
from the point A to the point B on the plane x = 1. The curve has the local minimum (red dot) at y = −1.
That is, along the curve connecting A and B on the plane x = 1, the point (1, −1) gives the local minimum.
But if we consider the intersection of the graph of f (x, y) and the plane y = −1, then the intersection is a
curve from the point C to the point D on the plane y = −1. The curve has the local maximum (red dot) at
x = 1. That is, along the curve connecting C and D on the plane y = −1, the point (1, −1) gives the local
maximum.
Thus we can say that the local extrema at (1, −1) depends on the curves (one giving the local maximum at
(1, −1) and one giving the local minimum at (1, −1)). In this case, we have the local maximum and the local
minimum at (1, −1) simultaneously and the point (1, −1) is called the saddle point.
(2) At (−1, −1): We have the exactly same feature as at (1, −1). That is, (−1, −1) is also a saddle point. □

Figure 12.8: Saddle Point S, Local Maxima A and B, Local Minima C and D
The formal definition of the saddle point is given as follows:

Page 76 of 126
Calculus II for Engineers Fall, 2010

Definition 12.7.5. The point P(a, b, f (a, b)) is a saddle point of z = f (x, y) if the following conditions are
satisfied:

1. (a, b) is a critical point of f ,

2. every open disk centered at (a, b) contains

(1) points (x, y) in the domain of f for which f (x, y) < f (a, b) and also
(2) points (x, y) in the domain of f for which f (x, y) > f (a, b).

How to find the saddle point? How to determine the local extrema? Above, we explained by using the graph.
However, in general, it is not easy to find such important points by using the graph. Just as we developed
and used the First Derivative Test and the Second Derivative Test on the function of a single variable in
Calculus I, we introduce the Second Derivatives Test on the function of two variables as follows:
Theorem 12.7.6 (S ECOND D ERIVATIVES T EST). Suppose that f (x, y) has continuous second–order partial
derivatives in some open disk containing the point (a, b) and that fx (a, b) = fy (a, b) = 0. Define the discrim-
inant D for the point (a, b) by

D(a, b) = fxx (a, b) fyy (a, b) − [ fxy (a, b)]2 .

1. If D(a, b) > 0 and fxx (a, b) > 0, then f has a local minimum at (a, b).
2. If D(a, b) > 0 and fxx (a, b) < 0, then f has a local maximum at (a, b).
3. If D(a, b) < 0, then f has a saddle point at (a, b).
4. If D(a, b) = 0, then no conclusion can be drawn.

How to understand the Theorem?

Part I. In order to have D(a, b) > 0, we should have fxx (a, b) fyy (a, b) > 0. It occurs only when
(1) fxx (a, b) > 0 and fyy (a, b) > 0 or (2) fxx (a, b) < 0 and fyy (a, b) < 0.
 Case (1) fxx (a, b) > 0 and fyy (a, b) > 0: In this case, we observe that the surface z = f (x, y) is concave
upward in the plane y = b (from fxx (a, b) > 0) and also concave upward in the plane x = a (from fyy (a, b) > 0).
Hence, the surface will look like an upward–opening paraboloid near the point (a, b) and thus f (x, y) has the
local minimum at (a, b).
 Case (2) fxx (a, b) < 0 and fyy (a, b) < 0: In this case, we observe that the surface z = f (x, y) is concave
downward in the plane y = b (from fxx (a, b) < 0) and also concave downward in the plane x = a (from
fyy (a, b) < 0). Hence, the surface will look like an downward–opening paraboloid near the point (a, b) and
thus f (x, y) has the local maximum at (a, b).
Part II. In order to have D(a, b) < 0, fxx (a, b) and fyy (a, b) should have the opposite signs (one positive and
one negative). It implies that f (x, y) has the local minimum and the local maximum at (a, b) simultaneously.
Hence, f (x, y) has the saddle point at (a, b).
Be careful! Having fxx (a, b) > 0 and fyy (a, b) > 0 without having D(a, b) > 0 does not say that f (a, b) is a
local minimum.
Example 12.7.7. Locate and classify all critical points for f (x, y) = xye−x
2 /2−y2 /2
.
A NSWER. Part I. Critical Points.
( ) 2 ( ) 2
fx = y 1 − x2 e−x /2−y /2 , fy = x 1 − y2 e−x /2−y /2 .
2 2

Solving fx = 0, we have x = ±1 or y = 0. Solving fy = 0, we have x = 0 or y = ±1. So, both fx = 0 and

fy = 0 are satisfied simultaneously at the points, (±1, ±1) and (0, 0), which are critical points.
Part II. Classification of Critical Points.

fxx = (−3 + x2 ) f (x, y) = xy(−3 + x2 )e−x

2 /2−y2 /2
,

Page 77 of 126
Calculus II for Engineers Fall, 2010

fyy = (−3 + y2 ) f (x, y) = xy(−3 + y2 )e−x

2 /2−y2 /2
,
(1 − x2 )(1 − y2 )
f (x, y) = (1 − x2 )(1 − y2 )e−x /2−y /2 ,
2 2
fxy =
xy
[ ( ) ( ) ( )2 ] 2 2
D(x, y) = fxx (x, y) fyy (x, y) − fxy
2
(x, y) = − y2 + 1 x4 + −y4 + 5y2 + 2 x2 − y2 − 1 e−x −y .

By the Theorem 12.7.6, we deduce

fxx (0, 0) = 0 = fyy (0, 0), D(0, 0) = −1 < 0 −→ saddle point at (0, 0)
fxx (1, 1) = −2e−1 = fyy (1, −1) < 0, D(1, 1) = 4e−2 > 0 −→ local maximum at (1, 1)
fxx (1, −1) = 2e−1 = fyy (1, −1) > 0, D(1, −1) = 4e−2 > 0 −→ local minimum at (1, −1)
−1 −2
fxx (−1, 1) = 2e = fyy (−1, 1) > 0, D(−1, 1) = 4e >0 −→ local minimum at (−1, 1)
−1 −2
fxx (−1, −1) = −2e = fyy (−1, −1) < 0, D(−1, −1) = 4e >0 −→ local maximum at (−1, −1)

That is, f (x, y) has the local maximum at (1, 1) and (−1, −1) and the local minimum at (1, −1) and (−1, 1)
and the saddle point at (0, 0). See the figure 12.9. □

Figure 12.9: Black Dot: Saddle Point at (0, 0), Blue Dots at Peaks: Local Maximum at (1, 1) and (−1, −1),
Red Dots at Valleys: Local Minimum at (1, −1) and (−1, 1)

Example 12.7.8. Locate and classify all critical points for f (x, y) = 2x2 − y3 − 2xy.
A NSWER. Part I. Critical Points.

fx = 2(2x − y) = 0, fy = −(3y2 + 2x) = 0.

From the first equation, we get y = 2x. Putting this into the second equation, we get
1
3(2x)2 + 2x = 0 −→ x = 0 or x=− .
6
( )
So we have two critical points (0, 0) and − 61 , − 13 .
Part II. Classification of Critical Points.

fxx (x, y) = 4 > 0, fyy (x, y) = −6y, fxy (x, y) = −2,

Page 78 of 126
Calculus II for Engineers Fall, 2010

D(x, y) = fxx (x, y) fyy (x, y) − fxy

2
(x, y) = −4(6y + 1).

( 0), we)have D(0, 0)(= −4 < )0. By the Second Derivatives

(1) At (0, ( ) Test, f (x, y) has the saddle point at (0, 0).
(2) At − 61 , − 13 , we have D − 16 , − 13 = 4 > 0 and fxx − 16 , − 31 = 4 > 0. Thus, by the Second Derivatives
Test, f (x, y) has this local minimum f (0, 0) = − 54
1
at (0, 0). See the figure 12.10. □

Figure 12.10: Black Dot: Saddle Point at (0, 0, 0), Red Dot: Local Minimum at (−1/6, −1/3, −1/54)

Example 12.7.9. Locate and classify all critical points for f (x, y) = x3 − 2y2 − 2y4 + 3x2 y.
A NSWER. Part I. Critical Points.
fx (x, y) = 3x(x + 2y) = 0, fy (x, y) = −4y − 8y3 + 3x2 = 0.
From the first equation, we have x = 0 or x = −2y. Putting these to the second equation, we get
0 = −4y − 8y3 = −4y(2y2 + 1), y = 0,
1
0 = −4y − 8y3 + 3(−2y)2 = −4y(1 + 2y2 − 3y) = −4y(2y − 1)(y − 1), y=
or y = 1.
2
Putting y = 1/2 and y = 1 back to x = −2y, we get x = −1 and x = −2, respectively. So we have the critical
points (0, 0), (−1, 1/2) and (−2, 1).
Part II. Classification of Critical Points.
fxx (x, y) = 6(x + y), fyy (x, y) = −4(6y2 + 1), fxy (x, y) = 6x,
D(x, y) = fxx (x, y) fyy (x, y) − fxy (x, y) = −36x − 24(x + y)(6y2 + 1).
2 2

(1) At (−1, 1/2): We have D(−1, 1/2) = −6 < 0. By the Second Derivatives Test, f (x, y) has the saddle
point at (−1, 1/2).
(2) At (−2, 1): We have D(−2, 1) = 24 > 0 and fxx (−2, 1) = −6 < 0. By the Second Derivatives Test, f (x, y)
has the local maximum at (−2, 1).
(3) At (0, 0): We have D(0, 0) = 0. So the Second Derivatives Test cannot give us any information. However,
we can deduce some information from the function f (x, y) = x3 − 2y2 − 2y4 + 3x2 y. When y = 0, the function
becomes f (x, 0) = x3 , which has an inflection point at x = 0. It shows that there is no local extrema at this
point (0, 0). Hence, the critical point (0, 0) has no significant graphical feature.
As you can see from the figure 12.11, it is not easy to tell the local extrema and saddle point from the graph.
So in this case, it’s good to use the analysis, i.e., the Second Derivatives Test. □

Page 79 of 126
Calculus II for Engineers Fall, 2010

Figure 12.11: Black Dot: Critical Point at (0, 0), Red Dot: Saddle Point at (−1, 1/2), Blue Dot: Local
Maximum at (−2, 1)

□ 12.7.2 Linear Regression: Method of Least Squares.

Skip. Please read the textbook.

□ 12.7.3 Steepest Ascent Algorithm: Estimating Local Extrema without Critical Points.
Skip. Please read the textbook.

□ 12.7.4 Absolute Extrema.

Definition 12.7.10. We call f (a, b) the absolute maximum of f on the region R if f (a, b) ≥ f (x, y) for all
(x, y) ∈ R. Similarly, f (a, b) is called the absolute minimum of f on the region R if f (a, b) ≤ f (x, y) for all
(x, y) ∈ R. In either case, f (a, b) is called an absolute extremum of f .
Recall from Calculus I that for a function of a single variable, if f is continuous on the closed interval [a, b],
then f has the maximum and minimum value on [a, b] and the absolute extrema must occur at either critical
numbers of f or at the endpoints of the interval [a, b]. We can extend this argument to the function of two
variables. We say that a region R ⊂ R2 is bounded if there is a disk that completely contains R.
Theorem 12.7.11 (E XTREME VALUE T HEOREM). Suppose that f (x, y) is continuous on the closed and
bounded region R ⊂ R2 . Then f has both an absolute maximum and an absolute minimum on R. Further, an
absolute extremum may only occur at a critical point in R or at a point on the boundary of R.
If f (a, b) is an absolute extremum of f in R and (a, b) is an interior point of R, then (a, b) is also a local
extremum of f , in which case, (a, b) must be a critical point. This implies that all of the absolute extrema
of a function in region R occur either at critical points or on the boundary of the region. For this reason, in
order to find the absolute extrema on R , we look for (i) the local extrema inside R and (ii) the extrema on the
boundary of R and compare all the extrema.
We take the following steps:

Step 1. Critical Points: Find all critical points of f in the region R.

Step 2. Extrema on Boundary of R: Find the maximum and minimum of f on the boundary of R.
Step 3. Comparison: Compare the values of f at the critical points with the maximum and minimum values
of f on the boundary of R.

Page 80 of 126
Calculus II for Engineers Fall, 2010

Example 12.7.12. Find the absolute extrema of f (x, y) = 5 + 4x − 2x2 + 3y − y2 on the region R bounded by
the lines y = 2, y = x and y = −x.

A NSWER. Step 1. Critical Points.

fx = 4(1 − x), fy = 3 − 2y,

which give x = 1 and y = 3/2. That is, we have only one critical point (1, 3/2) which is inside the region R
with the value f (1, 3/2) = 9.25.
Step 2. Extrema on Boundary of R.
(1) On the boundary y = x on [0, 2], the function becomes f (x, x) = −3x2 + 7x + 5. Let g(x) = f (x, x). We find
the absolute extrema of g(x) on [0, 2]. By Calculus I, g(x) has the absolute maximum g(7/6) = 9.08333 at
x = 7/6 and the absolute minimum g(0) = 5 at x = 5. That is, only on the line y = x, f (x, y) has the absolute
maximum f (7/6, 7/6) = 9.08333 and the absolute minimum f (0, 0) = 5.
(2) On the boundary y = −x on [−2, 0], the function becomes f (x, −x) = −3x2 + x + 5. Let h(x) = f (x, −x).
We find the absolute extrema of h(x) on [−2, 0]. By Calculus I, h(x) has the absolute maximum h(0) = 5 at
x = 0. (Be careful! Even though h′ (x) = 0 at x = 1/6, since 1/6 is not in the domain [−2, 0], so it cannot be
the critical number.) h(x) has the absolute minimum h(−2) = −9 at x = −2. That is, only on the line y = −x,
f (x, y) has the absolute maximum f (0, 0) = 5 and the absolute minimum f (−2, 2) = −9.
(3) On the boundary y = 2 on [−2, 2], the function becomes f (x, 2) = −2x2 + 4x + 7. Let k(x) = f (x, 2). We
find the absolute extrema of k(x) on [−2, 2]. By Calculus I, k(x) has the absolute maximum k(1) = 9 at x = 1
and the absolute minimum k(−2) = −9 at x = −2. That is, only on the line y = 2, f (x, y) has the absolute
maximum f (1, 2) = 9 and the absolute minimum f (−2, 2) = −9.
Step 3. Comparison. We collect all the values found above and compare them.
(1, 3/2) (0, 0) (7/6, 7/6) (−2, 2) (1, 2)
f (x, y) 9.25 5 9.08333 −9 9
Based on the table, we conclude that on the region R, f (x, y) has the absolute maximum f (1, 3/2) = 9.25 at
the critical point (1, 3/2) inside the region R and the absolute minimum f (−2, 2) = −9 at the boundary point
(−2, 2) on the boundary of R. See the figure 12.12. □

Figure 12.12: Left–Hand Side Figure: Region R with Black Dot (1, 3/2) and Red Dot (−2, 2), Right–Hand
Side Figure: Graph of f (x, y) on Region R with Absolute Maximum (Black Dot (1, 3/2, 9.25)) and Absolute
Minimum (Red Dot (−2, 2, −9)).

§12.8 Constrained Optimization and Lagrange Multipliers.

Skip. Please read the textbook.

Page 81 of 126
Calculus II for Engineers Fall, 2010

Page 82 of 126
 Chapter 13

Multiple Integrals

§13.1 Double Integrals.

□ 13.1.1 Introduction.
From Calculus I, we recall that the integral of f (x) over an interval I defined by I = { x ∈ R : a ≤ x ≤ b }
is given by
ˆ b ˆ x=b ˆ ˆ
f (x) dx or f (x) dx or f (x) dx or f (x) dx.
a x=a I [a,b]

For a function f (x, y) of two variables x and y, we can consider a double integral of f (x, y). A double integral
of f (x, y) over a region R ⊂ R2 is written by
¨
f (x, y) dA.
R

□ 13.1.2 Double Integral over a Rectangle.

Theorem 13.1.1 (F UBINI ’ S T HEOREM). Suppose that f (x, y) is integrable over the rectangle R defined by
{ }
R = (x, y) ∈ R2 : a ≤ x ≤ b, c ≤ y ≤ d

Then we can write the double integral of f (x, y) over R as either of the iterated integrals:
¨ ˆ [ˆ ] ˆ ˆ ˆ ˆ
x=b y=d x=b y=d b d
f (x, y) dA = f (x, y) dy dx = f (x, y) dy dx = f (x, y) dy dx
R x=a y=c x=a y=c a c
¨ ˆ [ˆ ] ˆ
y=d x=b y=d ˆ x=b ˆ dˆ b
or f (x, y) dA = f (x, y) dx dy = f (x, y) dx dy = f (x, y) dx dy
R y=c x=a y=c x=a c a

˜
Example 13.1.2. Evaluate R (6x
2 + 4xy3 ) dA, where R = { (x, y) : 0 ≤ x ≤ 2, 1 ≤ y ≤ 4 }.

A NSWER. By Fubini’s Theorem 13.1.1, we have

¨ ˆ x=2 [ˆ y=4 ] ˆ x=2 [ ]y=4
2 3
f (x, y) dA = (6x + 4xy ) dy dx = 6x2 y + xy4 y=1
dx
R x=0 y=1 x=0
ˆ x=2 ] [
2 255 2 x=2 3
= (18x + 255x) dx = 6x + x = 558.
x=0 2 x=0
¨ ˆ y=4 [ˆ x=2 ] ˆ y=4
2 3
[ 3 ]x=2
or f (x, y) dA = (6x + 4xy ) dx dy = 2x + 2x2 y3 x=0 dy
R y=1 x=0 y=1
ˆ y=4 [ ]y=4
= (16 + 8y3 ) dy = 16y + 2y4 y=1 = 558. □
y=1
˜
Exercise 13.1.3. Evaluate R (x + y) dA, where R = { (x, y) : 1 ≤ x ≤ 2, 3 ≤ y ≤ 4 }.

83
Calculus II for Engineers Fall, 2010

□ 13.1.3 Geometrical Application.

From Calculus I, we recall that the integral of y = f (x) = 5 over the interval

I={x : 2≤x≤9}

represents the area of the region under the curve y = f (x) = 5. We observe that the region is a rectangle with
the base 9 − 2 = 7 and the height 5. That is,
ˆ 9
5 dx = 7 × 5 = 35,
2

which can be checked by computing the definite integral in the left–hand side.
By the same argument, we observe that the integral of z = f (x, y) = 7 over a rectangle

R = { (x, y) : 2 ≤ x ≤ 5, 9 ≤ y ≤ 15 }

represents the volume of the box formed by the plane z = f (x, y) = 7 over R. We observe that the box has the
length 5 − 2 = 3, width 15 − 9 = 6 and height 7. That is,
¨
7 dA = 3 × 6 × 7 = 126,
R

which can be checked by computing the double integral in the left–hand side. We will discuss more applica-
tions in later Sections.

□ 13.1.4 Double Integral over a General Region.

From this topic, the order of integration is important.

Figure 13.1: Region bounded by y = g1 (x) and y = g2 (x)

Theorem 13.1.4 (T YPE I. R EGION B OUNDED BY y = g1 (x) AND y = g2 (x)). Suppose f (x, y) is continuous
on the region R defined by

R = { (x, y) : a ≤ x ≤ b, g1 (x) ≤ y ≤ g2 (x) }

for continuous functions g1 and g2 , where g1 (x) ≤ g2 (x) for all x ∈ [a, b]. Then
¨ ˆ [ˆ ] ˆ ˆ
x=b y=g2 (x) b g2 (x)
f (x, y) dA = f (x, y) dy dx = f (x, y) dy dx
R x=a y=g1 (x) a g1 (x)

See the figure 13.1. Be careful! When we change the order of the integration, the integral becomes different.
¨ ˆ g2 (x) ˆ b
Number = f (x, y) dA = f (x, y) dx dy = Function of x
R g1 (x) a

Page 84 of 126
Calculus II for Engineers Fall, 2010

Figure 13.2: Region bounded by y = x, y = 0, and x = 4

˜ x2
Example 13.1.5. Evaluate R (4e − 5 sin y) dA, where R is the region bounded by graphs of y = x, y = 0,
and x = 4.

A NSWER. See the figure 13.2. The region R is represented as follows:

R = { (x, y) : 0 ≤ x ≤ 4, 0 ≤ y ≤ x }.

By the Theorem 13.1.4, we deduce

¨ ˆ 4ˆ x ˆ 4[ ]x
x2 x2 2
(4e − 5 sin y) dA = (4e − 5 sin y) dy dx = 4ex y + 5 cos y dx
R 0 0 0 0
ˆ 4( )
2
= 4ex x + 5 cos x − 5 dx = 2e16 + 5 sin 4 − 22. □
0
˜
Exercise 13.1.6. Evaluate R (3x
2 + 2y) dA, where R is the region bounded by graphs of y = 1 − x2 , y = 0.

Figure 13.3: Region bounded by x = h1 (y) and x = h2 (y)

Theorem 13.1.7 (T YPE II. R EGION B OUNDED BY x = h1 (y) AND x = h2 (y)). Suppose f (x, y) is continuous
on the region R defined by

R = { (x, y) : c ≤ y ≤ d, h1 (y) ≤ x ≤ h2 (y) }

for continuous functions h1 and h2 , where h1 (y) ≤ h2 (y) for all y ∈ [c, d]. Then
¨ ˆ [ˆ ] ˆ ˆ
y=d x=h2 (y) d h2 (y)
f (x, y) dA = f (x, y) dx dy = f (x, y) dx dy
R y=c x=h1 (y) c h1 (y)

See the figure 13.3.

Page 85 of 126
Calculus II for Engineers Fall, 2010

Figure 13.4: Region bounded by x = y2 and x = 2 − y

˜
Example 13.1.8. Let R be the region bounded by graphs of x = y2 and x = 2 − y. Write R f (x, y) dA as an
iterated integral.

A NSWER. See the figure 13.4. The region R is represented as follows:

R = { (x, y) : −2 ≤ y ≤ 1, y2 ≤ x ≤ 2 − y }.
By the Theorem 13.1.7, we deduce
¨ ˆ 1 ˆ 2−y
f (x, y) dA = f (x, y) dx dy. □
R −2 y2

Exercise 13.1.9. Find an integral equal to the volume of the solid bounded by the surfaces and evaluate the
integral:
z = 6 − x − y, z = 0, x = 4 − y2 , x = 0.

□ 13.1.5 Changing Order of Integration.

Sometimes we need to change the order of integration in order to evaluate a double integral.
´1´1 2
Example 13.1.10. Evaluate the iterated integral 0 y ex dx dy.

Figure 13.5: Region R

A NSWER. See the figure 13.5. The region R in the double integral is interpreted as follows:
R = { (x, y) : 0 ≤ y ≤ 1, y ≤ x ≤ 1 }.
However, we can interpret the region R in the other way:
R = { (x, y) : 0 ≤ x ≤ 1, 0 ≤ y ≤ x }.
Using this interpretation, the double integral becomes
ˆ 1ˆ 1 ¨ ˆ 1ˆ x ˆ 1[ ]x ˆ 1
x2 x2 x2 x2 2 1
e dx dy = e dA = e dy dx = e y dx = ex x dx = (e − 1) . □
0 y R 0 0 0 0 0 2

Page 86 of 126
Calculus II for Engineers Fall, 2010

Figure 13.6: Region R

´ 1 ´ √y
Example 13.1.11. Change the order of integration: 0 0 f (x, y) dx dy.
A NSWER. See the figure 13.6. The region R in the double integral is interpreted as follows:
√
R = { (x, y) : 0 ≤ y ≤ 1, 0 ≤ x ≤ y }.
However, we can interpret the region R in the other way:
R = { (x, y) : 0 ≤ x ≤ 1, x2 ≤ y ≤ 1 }.
Using this interpretation, the double integral becomes
ˆ 1 ˆ √y ¨ ˆ 1ˆ 1
f (x, y) dx dy = f (x, y) dA = f (x, y) dy dx. □
0 0 R 0 x2
´ 2 ´ ln y
Exercise 13.1.12. Change the order of integration: 1 0 f (x, y) dx dy.

□ 13.1.6 Linearity of Double Integral.

We recall from Calculus I
ˆ b ˆ b ˆ b ˆ b ˆ p ˆ b
(c f (x) + dg(x)) dx = c f (x) dx + d g(x) dx, f (x) dx = f (x) dx + f (x) dx,
a a a a a p

where a ≤ p ≤ b and c and d are any constants. Similarly, we have the following theorem for double integrals.

Figure 13.7: R = R1 ∪ R2

Theorem 13.1.13 (L INEARITY OF D OUBLE I NTEGRALS). Let f (x, y) and g(x, y) be integrable over the
region R ⊂ R2 and let c and d be any constant. Then, the following hold:
˜ ˜ ˜
1. R (c f (x, y) + dg(x, y)) dA = c R f (x, y) dA + d R g(x, y) dA
2. If R = R1 ∪ R2 , where R1 and R2 are non–overlapping regions, then
˜ ˜ ˜
R f (x, y) dA = R1 f (x, y) dA + R2 f (x, y) dA

See the figure 13.7.

Page 87 of 126
Calculus II for Engineers Fall, 2010

§13.2 Area, Volume, and Center of Mass.

□ 13.2.1 Area.
´b ˜
The integral a 1 dx = b − a gives the length of the interval [a, b]. The double integral R 1 dA gives the area
of the region R.
Example 13.2.1. Find an area of the region R bounded by y = x2 and y = x + 2.

y.
.4
.y = x + 2

.2

.y = x2

. .x
.−1 .2

Figure 13.8: Region R bounded by y = x2 and y = x + 2

A NSWER. The region R can be represented as follows:

R = { (x, y) : −1 ≤ x ≤ 2, x2 ≤ y ≤ x + 2 }.

Hence, the area of R is obtained as follows:

¨ ˆ 2ˆ x+2 ˆ 2
9
A= 1 dA = 1 dy dx = (x + 2 − x2 ) dx = .
R −1 x2 −1 2
We note from Calculus I that the last integral represents the area of the region between two curves. That is,
the area of R can be obtained as follows:
ˆ 2 ˆ 2
( ) 9
A= (upper curve − lower curve) dx = x + 2 − x2 dx = . □
−1 −1 2
Exercise 13.2.2. Use a double integral to compute the area of the region bounded by the curves y = 3x,
y = 5 − 2x, and y = 0.

□ 13.2.2 Volume.
We recall from Calculus I that the area under the curve y = f (x) over the interval [a, b] is obtained by the
´b
integral a f (x) dx. By the same argument, we observe that ˜ the volume of the solid under the surface z =
f (x, y) over the region R is obtained by the double integral R f (x, y) dA.
Example 13.2.3. Compute the volume of the solid bounded by x + 2y + 3z = 6 and three coordinate plates.

A NSWER 1. Step 1. Sketch and Setup Double Integral. The equation x + 2y + 3z = 6 represents a plane
passing through (6, 0, 0), (0, 3, 0) and (0, 0, 2). The plane and the three coordinates plates, i.e., x = 0, y = 0
and z = 0, forms a solid given in the figure 13.9. The solid is called a tetrahedron. We observe from the
given equation
x 2y
x + 2y + 3z = 6, i.e., z = 2− − .
3 3

Page 88 of 126
Calculus II for Engineers Fall, 2010

.y

.3 .x + 2y = 6 with z = 0

.R

. .x
.O .6
Figure 13.9: Tetrahedron bounded by x + 2y + 3z = 6 and three coordinate plates and Region R

So the volume of the solid is obtained by the double integral

¨ ( )
x 2y
V= 2− − dA,
R 3 3
where R is the region of the solid in the xy-plane.
Step 2. Setup and Compute Iterated Integral. To express the double integral into the iterated integral, we
need to express the region R in the xy–plane in terms of x and y. When the plane x + 2y + 3z = 6 meets the
xy–plane, the intersection becomes a line of which equation is x + 2y = 6. So the region R in the xy–plane is
the region bounded by x + 2y = 6, x = 0 and y = 0. That is,
{ }
6−x
R = { (x, y) : 0 ≤ x, 0 ≤ y, x + 2y ≤ 6 } = (x, y) : 0 ≤ x ≤ 6, 0 ≤ y ≤
2
or R = { (x, y) : 0 ≤ x, 0 ≤ y, x + 2y ≤ 6 } = { (x, y) : 0 ≤ y ≤ 3, 0 ≤ x ≤ 6 − 2y } .

Therefore, we deduce the following iterated integral from the double integral above:
¨ ( ) ˆ x=6 ˆ y= 6−x ( ) ˆ x=6 ( 2 )
x 2y 2 x 2y x
V= 2− − dA = 2− − dy dx = − x + 3 dx = 6
R 3 3 x=0 y=0 3 3 x=0 12
¨ ( ) ˆ y=3 ˆ x=6−2y ( ) ˆ y=3 ( 2 )
x 2y x 2y 2y
or V = 2− − dA = 2− − dx dy = − 4y + 6 dy = 6 □
R 3 3 y=0 x=0 3 3 y=0 3

Remark 13.2.4. There are two more ways to solve the example above. In the solution above, we used the
region R in the xy–plane in the double integral. However, we CAN use the region P in the yz–plane or the
region Q in the xz–plane. Let us solve the example using the region P in the yz–plane and the region Q in the
xz–plane.

A NSWER 2. U SING REGION P IN yz– PLANE. When we use the region in the yz–plane, we should take the
function of two variables y and z as the integrand of the double integral. That is,

x + 2y + 3z = 6, i.e., x = 6 − 2y − 3z.

So the volume of the solid is obtained by the double integral

¨
V= (6 − 2y − 3z) dA,
P

Page 89 of 126
Calculus II for Engineers Fall, 2010

.z

.2 .z
.x + 3z = 6 with y = 0
.2y + 3z = 6 with x = 0 .2

.P
.y .Q
. . .x
.3 .6

Figure 13.10: Regions P and Q

where P is the region of the solid in the yz–plane.

By setting x = 0 in the given equation x + 2y + 3z = 6, we can obtain the line (boundary of the tetrahedron in
the yz–plane). So the region P in the yz–plane can be expressed as follows (See the figure 13.10.):
{ }
6 − 2y
P = { (y, z) : 0 ≤ y, 0 ≤ z, 2y + 3z ≤ 6 } = (y, z) : 0 ≤ y ≤ 3, 0 ≤ z ≤
3
{ }
6 − 3z
or P = { (y, z) : 0 ≤ y, 0 ≤ z, 2y + 3z ≤ 6 } = (y, z) : 0 ≤ z ≤ 2, 0 ≤ y ≤ .
2
Therefore, we deduce the following iterated integral from the double integral above:
¨ ˆ y=3 ˆ z= 6−2y ˆ y=3 ( 2 )
3 2y
V= (6 − 2y − 3z) dA = (6 − 2y − 3z) dz dy = − 4y + 6 dy = 6
P y=0 z=0 y=0 3
¨ ˆ z=2 ˆ y= 6−3z ˆ z=2 ( 2 )
2 9z
or V = (6 − 2y − 3z) dA = (6 − 2y − 3z) dy dz = − 9z + 9 dz = 6 □
P z=0 y=0 z=0 4
A NSWER 3. U SING REGION Q IN xz– PLANE. When we use the region in the xz–plane, we should take the
function of two variables x and z as the integrand of the double integral. That is,
6 − 3z − x
x + 2y + 3z = 6, i.e., y= .
2
So the volume of the solid is obtained by the double integral
¨
6 − 3z − x
V= dA,
Q 2
where Q is the region of the solid in the xz–plane.
By setting y = 0 in the given equation x + 2y + 3z = 6, we can obtain the line (boundary of the tetrahedron in
the xz–plane). So the region Q in the xz–plane can be expressed as follows (See the figure 13.10.):
{ }
6−x
Q = { (x, z) : 0 ≤ x, 0 ≤ z, x + 3z ≤ 6 } = (x, z) : 0 ≤ x ≤ 6, 0 ≤ z ≤
3
or Q = { (x, z) : 0 ≤ x, 0 ≤ z, x + 3z ≤ 6 } = { (x, z) : 0 ≤ z ≤ 2, 0 ≤ x ≤ 6 − 3z } .
Therefore, we deduce the following iterated integral from the double integral above:
¨ ˆ x=6 ˆ z= 6−x ˆ x=6 ( )
6 − 3z − x 3 6 − 3z − x x2
V= dA = dz dx = 3−x+ dx = 6
Q 2 x=0 z=0 2 x=0 12
¨ ˆ z=2 ˆ x=6−3z ˆ z=2 ( 2 )
6 − 3z − x 6 − 3z − x 9z
or V = dA = dx dz = − 9z + 9 dz = 6 □
Q 2 z=0 x=0 2 z=0 4

Page 90 of 126
Calculus II for Engineers Fall, 2010

As we can see, for the case of the volume of the tetrahedron, there are total 6 ways (3 ways depending on the
region and 2 ways for each region) to solve the problem.

Exercise 13.2.5.  Find the volume of the tetrahedron bounded by the plane 2x + y + z = 2 and the three
coordinates planes.
 Find the volume of the solid lying in the first octant and bounded by the graphs of z = 4 − x2 , x + y = 2,
x = 0, y = 0 and z = 0.
 Find the volume of the solid bounded by the graphs of z = 2, z = x2 + 1, y = 0 and x + y = 2.

□ 13.2.3 Moments and Center of Mass.

We discuss a physical application of double integrals. Consider a thin, flat plate (a lamina) in the shape of
the region R ⊂ R2 . From an engineering standpoint, it is often important to determine where you could place
a support to balance the plate. We call this point the center of mass of the lamina. The center of mass of a
lamina is a point whose x– and y–component are called the moments with respect to the x–axis and y–axis,
respectively. Let us introduce some definitions.

Definition 13.2.6. For a lamina with density function ρ (x, y) in the shape of the region R ⊂ R2 ,

1. the total mass m is obtained by ¨

m= ρ (x, y) dA
R

2. the moments, Mx and My , with respect to the x– and y–axis are, respectively,
¨ ¨
Mx = yρ (x, y) dA, My = xρ (x, y) dA
R R

3. the center of mass is the point (x̄, ȳ), where

˜ ˜
My xρ (x, y) dA Mx yρ (x, y) dA
x̄ = = ˜R , ȳ = = ˜R
m R ρ (x, y) dA m R ρ (x, y) dA

Example 13.2.7. Find the mass and center of mass of a triangular lamina with vertices (0, 0), (1, 0), and
(0, 2) if the density function is ρ (x, y) = 1 + 3x + y.

.y

.2

.2x + y = 2

.R
. .x
.1

Figure 13.11: Triangular Lamina R

Page 91 of 126
Calculus II for Engineers Fall, 2010

A NSWER. The equation of the line connecting (1, 0) and (0, 2) is y = 2 − 2x. So the triangular lamina in the
figure 13.11 can be represented by the region R as follows:
{ }
2−y
R = { (x, y) : 0 ≤ x ≤ 1, 0 ≤ y ≤ 2 − 2x } = (x, y) : 0 ≤ y ≤ 2, 0 ≤ x ≤
2

Let us use the first representation.

(1) Total Mass m:
¨ ˆ x=1 ˆ y=2−2x ˆ x=1 ( ) 8
m= ρ (x, y) dA = (1 + 3x + y) dy dx = 4 1 − x2 dx =
R x=0 y=0 x=0 3

(2) Center of Mass, (x̄, ȳ):

¨ ˆ x=1 ˆ y=2−2x ˆ x=1 (
1 3 3 ) 3
x̄ = xρ (x, y) dA = x (1 + 3x + y) dy dx = x − x3 dx =
m R 8 x=0 y=0 2 x=0 8
¨ ˆ x=1 ˆ y=2−2x ˆ x=1 (
1 3 1 ) 11
ȳ = yρ (x, y) dA = y (1 + 3x + y) dy dx = 7 − 9x − 3x2 + 5x3 dx =
m R 8 x=0 y=0 4 x=0 16
(3 )
Thus, the total mass is m = 8
3 and the center of mass is (x̄, ȳ) = 11
8 , 16 . □

Exercise 13.2.8. Find the mass and the center of mass of the lamina bounded by y = x3 and y = x2 with the
density ρ (x, y) = 4.

□ 13.2.4 Moments of Inertia.

The moment of inertia (also called the second moment) of a particle of mass m about an axis is defined to
be mr2 , where r is the distance from the particle to the axis. We extend this concept to a lamina with density
function ρ (x, y) and occupying a region R by proceeding as we did for ordinary moments.

Definition 13.2.9. The moments of inertia, Ix and Iy , of the lamina about the x–axis and y–axis are, respec-
tively, ¨ ¨
Ix = y2 ρ (x, y) dA, Iy = x2 ρ (x, y) dA.
R R

Example 13.2.10. Find the moments of inertia Ix and Iy for the lamina in the previous example: the triangular
lamina with vertices (0, 0), (1, 0), and (0, 2) and the density function is ρ (x, y) = 1 + 3x + y.

A NSWER.
¨ ˆ x=1 ˆ y=2−2x ˆ x=1 ( ) 8
Iy = x ρ (x, y) dA =
2 2
x (1 + 3x + y) dy dx = 4 x2 − x4 dx = ≈ 0.533
R x=0 y=0 x=0 15
¨ ˆ x=1 ˆ y=2−2x
Ix = y2 ρ (x, y) dA = y2 (1 + 3x + y) dy dx
R x=0 y=0
ˆ x=1 ( )
20 16x3 28
= − 16x + 8x2 + − 4x4 dx = ≈ 1.8667
x=0 3 3 15

A comparison of the two moments of inertia shows that it is much more difficult to rotate the triangular
lamina about the x–axis than about the y–axis. □

Exercise 13.2.11. Find the moments of inertial Ix and Iy for the lamina in the Exercise 13.2.8.

Page 92 of 126
Calculus II for Engineers Fall, 2010

§13.3 Double Integrals in Polar Coordinates.

□ 13.3.1 Polar Coordinates.

We refresh our memory on the polar coordinates from the Section 9.4 Polar Coordinates.

Figure 13.12: Rectangular and Polar coordinates

Theorem 13.3.1 (C ONNECTION BETWEEN P OLAR AND C ARTESIAN C OORDINATES). Suppose a point P
has Cartesian coordinates (x, y) and polar coordinates (r, θ ). Then we have
y
x = r cos θ , and y = r sin θ ⇐⇒ x 2 + y2 = r 2 , and tan θ = .
x

□ 13.3.2 Double Integrals in Polar Coordinates.

Figure 13.13: Region R in polar coordinates

Theorem 13.3.2 (F UBINI ’ S T HEOREM). Suppose that f (r, θ ) is continuous on the region R given by

R = { (r, θ ) : α ≤ θ ≤ β , g1 (θ ) ≤ r ≤ g2 (θ ) } ,

where 0 ≤ g1 (θ ) ≤ g2 (θ ) for all θ ∈ [α , β ]. See the figure 13.13. Then,

¨ ˆ θ =β ˆ r=g2 (θ )
f (r, θ ) dA = f (r, θ )r dr dθ .
R θ =α r=g1 (θ )

Example 13.3.3. Find the area of the region bounded by the curve r = 3 + 2 sin θ .

Page 93 of 126
Calculus II for Engineers Fall, 2010

Figure 13.14: Region bounded by r = 3 + 2 sin θ

A NSWER. For each fixed θ , r ranges from 0 to 3 + 2 sin θ . To go all the way around the cardioid (graph of
r = 3 + 2 sin θ : see the figure 13.14), exactly once, θ ranges from 0 to 2π . By Fubini’s Theorem 13.3.2, we
have
¨ ˆ 2π ˆ 3+2 sin θ ˆ 2π [ 2 ] ˆ
r 3+2 sin θ 1 2π
A= dA = r dr dθ = dθ = (3 + 2 sin θ )2 dθ = 11π . □
R 0 0 0 2 0 2 0

Exercise 13.3.4. Find the area of the region bounded by the curve r = 2 − 2 cos θ .
˜ √
Example 13.3.5. Evaluate R x2 + y2 dA, where R is the disk x2 + y2 ≤ 9.

√
Figure 13.15: Solid under z = x2 + y2 over x2 + y2 ≤ 9

A NSWER. See the figure 13.15. The disk R can be represented in polar coordinates as follows:

R = { (r, θ ) : 0 ≤ r ≤ 3, 0 ≤ θ ≤ 2π } .

Using Fubini’s Theorem 13.3.2 with the representation, we have

¨ √ ˆ θ =2π ˆ r=3 √ ˆ θ =2π ˆ r=3
2 2
x + y dA = r r dr dθ =
2 r2 dr dθ
R θ =0 r=0 θ =0 r=0
ˆ θ =2π [ 3 ] ˆ θ =2π
r r=3
= dθ = 9 dθ = 19π . □
θ =0 3 r=0 θ =0

Page 94 of 126
Calculus II for Engineers Fall, 2010

Figure 13.16: Two Cylinders

˜ −x2 −y2 dA,
Exercise 13.3.6. Evaluate Re where R is the disk x2 + y2 ≤ 4.
Example 13.3.7. Find the volume of the solid bounded by z = 7, z = 0, x2 + y2 = 25, and x2 + y2 = 4.

A NSWER 1 BASIC F ORMULA. We observe that the solid is between two cylinders. See the figure 13.16.
The large cylinder formed by x2 + y2 = 25, z = 0 and z = 7 has the volume
Vlarge = (Base Area) × (Height) = π (52 )(7) = 175π .
The small cylinder formed by x2 + y2 = 4, z = 0 and z = 7 has the volume Vsmall = π (22 )(7) = 28π . Thus,
we get the volume of the solid,
( )
V = Vlarge −Vsmall = 7π 52 − 22 = 147π . □
A NSWER 2 R ECTANGULAR C OORDINATES. When we cut the solid horizontally, the cut sides have the ex-
actly same shape with same area, which is the annulus R 1 . The region R is expressed as
{ }
R = (x, y) : 4 ≤ x2 + y2 ≤ 25
{ √ √ }
= (x, y) : −2 ≤ x ≤ 2, 4 − x2 ≤ y ≤ 25 − x2
{ √ √ }
∪ (x, y) : −2 ≤ x ≤ 2, − 25 − x ≤ y ≤ − 4 − x
2 2 = R1 ∪ R2 ,

where R1 and R2 represent the first and second region, respectively. This separation implies
¨ ¨ ¨
f (x, y) dA = f (x, y) dA + f (x, y) dA.
R R1 R2

From the figure 13.16, we observe that the height of the solid moves from z = 0 to z = 7. Hence, the volume
of the solid is obtained by
¨ ¨ ¨ ˆ x=2 ˆ y=√25−x2 ˆ x=2 ˆ y=−√4−x2
V= (7 − 0) dA = 7 dA + 7 dA = √ 7 dy dx + √ 7 dy dx
R R1 R2 x=−2 y= 4−x2 x=−2 y=− 25−x2
ˆ x=2 [√ √ ] ˆ x=2 [ √ √ ]
=7 25 − x − 4 − x dx + 7
2 2 − 4 − x2 + 25 − x dx
2
x=−2 x=−2
ˆ x=2 [√ √ ]
= 14 25 − x2 − 4 − x2 dx,
x=−2

which is not easy to compute. □

1 region between two circles having the same center and different radii

Page 95 of 126
Calculus II for Engineers Fall, 2010

A NSWER 3 P OLAR C OORDINATES. We use the polar coordinates and represent the region R as follows:
{ }
R = (x, y) : 4 ≤ x2 + y2 ≤ 25 = { (r, θ ) : 0 ≤ θ ≤ 2π , 2 ≤ r ≤ 5 } .
Then, the volume of the solid is
¨ ˆ θ =2π ˆ r=5
V= 7 dA = 7r dr dθ
R θ =0 r=2
ˆ θ =2π [ ]r=5 ˆ θ =2π [ ]
7 7 2 7
= r 2
dθ = (5 − 2 ) dθ = (52 − 22 )2π = 7π (52 − 22 ) = 147π .
2
□
θ =0 2 r=2 θ =0 2 2
As we can see, in this problem, the first method is the easiest one and the method using the rectangular
coordinates is the most difficult one in getting the volume.
Example 13.3.8. Find the volume of the solid below z = 8 − x2 − y2 and above z = 3x2 + 3y2 .

Figure 13.17: Solid below z = 8 − x2 − y2 and above z = 3x2 + 3y2

A NSWER. See the figure 13.17. We first find the intersection of two surfaces, z = 8 − x2 − y2 and z =
3x2 + 3y2 :
8 − x2 − y2 = 3x2 + 3y2 −→ 4x2 + 4y2 = 8 −→ x2 + y2 = 2,
√
which is a circle centered at (0, 0) with radius 2 . It implies that a point (x, y, z) in the solid, precisely, (x, y),
should be in the disk x2 + y2 ≤ 2. That is, as the base of the solid, we can take the region R:
{ }
R = (x, y) : x2 + y2 ≤ 2 .

On this region R, the surface of z = 8 − x2 − y2 is higher than that of z = 3x2 + 3y2 . So, the volume V of the
solid is obtained by
¨ ¨
[ ] [ ]
V= 8 − x − y − (3x + 3y ) dA =
2 2 2 2
8 − 4(x2 + y2 ) dA.
R R

When we use the following representation of R:

{ } { √ √ √ √ }
R = (x, y) : x2 + y2 ≤ 2 = (x, y) : − 2 ≤ x ≤ 2 , − 2 − x2 ≤ y ≤ 2 − x2 ,

the volume V becomes

¨ ˆ √ ˆ √
[ ] x= 2 y= 2−x2 [ ]
V= 8 − 4(x + y ) dA =
2 2
√ √ 8 − 4(x2 + y2 ) dy dx,
R x=− 2 y=− 2−x2

Page 96 of 126
Calculus II for Engineers Fall, 2010

which is not easy to compute. So we use the polar coordinates and express the region R as follows:
{ } { √ }
R = (x, y) : x2 + y2 ≤ 2 = (r, θ ) : 0 ≤ r ≤ 2 , 0 ≤ θ ≤ 2π .

Using this representation, the volume V becomes

¨ ˆ √
[ ] θ =2π ˆ r= 2 [ ]
V= 8 − 4(x2 + y2 ) dA = 8 − 4r2 r dr dθ
R θ =0 r=0
ˆ √
θ =2π ˆ r= 2 [ ]
ˆ θ =2π [ ] √
ˆ θ =2π
4 r= 2
= 8r − 4r3 dr dθ = 4r 2
− r r=0 dθ = 4 dθ = 8π . □
θ =0 r=0 θ =0 θ =0
´ 2 ´ √4−x2
e−x
2 −y2
Example 13.3.9. Evaluate the iterated integral by converting to polar coordinates √ dy dx.
0 − 4−x2

A NSWER. The full expression of the integral is

ˆ 2ˆ √ ˆ √
4−x2 x=2 ˆ y= 4−x2
−x2 −y2
e−x
2 −y2
√ e dy dx = √ dy dx.
0 − 4−x2 x=0 y=− 4−x2

It implies that the region R of the integral is

{ √ √ }
R = (x, y) : 0 ≤ x ≤ 2, − 4 − x ≤ y ≤ 4 − x
2 2 .

If we sketch the region R in the xy–plane, we observe that R represents a quarter circle centered at the origin
with radius 2 lying on the first quadrant of the xy–plane. When using the polar coordinates, R is represented
as follows: { π π }
R = (r, θ ) : 0 ≤ r ≤ 2, − ≤ θ ≤ .
2 2
Hence, the integral becomes in the polar coordinates
ˆ 2ˆ √ ˆ √
4−x2 x=2 ˆ y= 4−x2
−x2 −y2
e−x
2 −y2
√ e dy dx = √ dy dx
0 − 4−x2 x=0 y=− 4−x2
ˆ ˆ ˆ [ ]r=2
θ =π /2 θ =π /2
e−r
r=2 2
−r2
= e r dr dθ = − dθ
θ =−π /2 r=0 θ =−π /2 2
r=0
ˆ θ =π /2
1 − e−4 π (1 − e−4 )
= dθ = . □
θ =−π /2 2 2

§13.4 Surface Area.

Skip. Please read the textbook.

Page 97 of 126
Calculus II for Engineers Fall, 2010

§13.5 Triple Integrals.

□ 13.5.1 Formulas.

Box Q
Theorem 13.5.1 (F UBINI ’ S T HEOREM). Suppose that f (x, y, z) is continuous on the box Q defined by

Q = { (x, y, z) : a ≤ x ≤ b, c ≤ y ≤ d, r ≤ z ≤ s } .

Then we can write the triple integral over Q as a triple iterated integral:
˚ ˆ z=s ˆ y=d ˆ x=b ˆ z=s ˆ x=b ˆ y=d
f (x, y, z) dV = f (x, y, z) dx dy dz = f (x, y, z) dy dx dz
Q z=r y=c x=a z=r x=a y=c
ˆ x=b ˆ z=s ˆ y=d ˆ x=b ˆ y=d ˆ z=s
= f (x, y, z) dy dz dx = f (x, y, z) dz dy dx
x=a z=r y=c x=a y=c z=r
ˆ y=d ˆ z=s ˆ x=b ˆ y=d ˆ x=b ˆ z=s
= f (x, y, z) dx dz dy = f (x, y, z) dz dx dy.
y=c z=r x=a y=c x=a z=r

Moreover, if f (x, y, z) = g(x)h(y)k(z), then

˚ (ˆ ) (ˆ ) (ˆ )
x=b y=d z=s
f (x, y, z) dV = g(x) dx h(y) dy k(z) dz .
Q x=a y=c z=r
˝ y sin z dV ,
Example 13.5.2. Evaluate the triple integral Q 2xe where Q is the box defined by

Q = { (x, y, z) : 1 ≤ x ≤ 2, 0 ≤ y ≤ 1, 0 ≤ z ≤ π } .

A NSWER. By Fubini’s Theorem, we have

˚ ˆ z=π ˆ y=1 ˆ x=2
y
2xe sin z dV = 2xey sin z dx dy dz
Q z=0 y=0 x=1
(ˆ x=2 ) (ˆ y=1 ) (ˆ z=π )
=2 x dx y
e dy sin z dz = 6(e − 1). □
x=1 y=0 z=0
˝
Exercise 13.5.3. Evaluate the triple integral Q (x + y + z) dV , where Q is the box defined by

Q = { (x, y, z) : 0 ≤ x ≤ 1, −1 ≤ y ≤ 2, 0 ≤ z ≤ 3 } .
˝
Exercise 13.5.4. Evaluate the triple integral Q xyz2 dV , where Q is the box defined by

Q = { (x, y, z) : 0 ≤ x ≤ 1, 0 ≤ y ≤ 2, 0 ≤ z ≤ 3 } .

Solid Q bounded by Upper and Lower Surfaces

Theorem 13.5.5 (F IXED O NE VARIABLE). Suppose the solid Q can be written in the form

Q = { (x, y, z) : (x, y) ∈ R, g1 (x, y) ≤ z ≤ g2 (x, y) } ,

where R is some region in the xy–plane and where g1 (x, y) ≤ g2 (x, y) for all (x, y) in R. Then the triple integral
over Q is expressed as follows:
˚ ¨ ˆ z=g2 (x,y)
f (x, y, z) dV = f (x, y, z) dz dA.
Q R z=g1 (x,y)

Page 98 of 126
Calculus II for Engineers Fall, 2010

Figure 13.18: Tetrahedron bounded by 2x + y + z = 4 and three coordinates plates

˝
Example 13.5.6. Evaluate Q 6xy dV , where Q is the tetrahedron bounded by the planes x = 0, y = 0, z = 0
and 2x + y + z = 4.

A NSWER. See the figure 13.18. Let R be the region of Q on the xy–plane. Then the height of the tetrahedron
turns to be z = 4 − y − 2x, i.e., when (x, y) of (x, y, z) ∈ Q moves over R, the z component moves from z = 0
to z = 4 − y − 2x. So, the triple integral becomes
˚ ¨ ˆ z=4−y−2x ¨
6xy dV = 6xy dz dA = 6xy(4 − y − 2x) dA,
Q R z=0 R

which is computed by the method in Section 13.1 Double Integrals. Explicitly, the region R is expressed as
follows:
R = { (x, y) : 0 ≤ x ≤ 2, 0 ≤ y ≤ 4 − 2x } .
Using this expression, the triple integral becomes
˚ ¨ ˆ x=2 ˆ y=4−2x
6xy dV = 6xy(4 − y − 2x) dA = 6xy(4 − y − 2x) dy dx. □
Q R x=0 y=0
˝
Exercise 13.5.7. Evaluate Q z dV , where Q is the tetrahedron bounded by the planes x = 0, y = 0, z = 0
and x + y + z = 1.

□ 13.5.2 Applications: Volume, Mass and Center of Mass.

Volume
˝
The volume V of a solid Q in the 3–dimensional space is obtained by the triple integral V = Q dV .

Example 13.5.8. Find the volume of the tetrahedron Q bounded by x = 0, x = 2y, z = 0 and x + 2y + z = 2.

A NSWER. See the figure 13.19. We take the region R of Q on the xy–plane of Q as our base of the tetrahedron
Q. The lower boundary of Q is the plane z = 0 and the upper boundary is the plane x + 2y + z = 2, that is,
z = 2 − x − 2y. It implies
˚ ¨ ˆ z=2−x−2y
V= dV = dz dA.
Q R z=0

Page 99 of 126
Calculus II for Engineers Fall, 2010

Figure 13.19: Tetrahedron bounded by x = 0, x = 2y, z = 0 and x + 2y + z = 2

From Section 13.1 Double Integrals, the region R of Q on the xy–plane is expressed as follows:
{ x x}
R= (x, y) : 0 ≤ x ≤ 1, ≤ y ≤ 1− .
2 2
Therefore, we get
˚ ¨ ˆ z=2−x−2y ˆ x=1 ˆ y=1−x/2 ˆ z=2−x−2y
1
V= dV = dz dA = dz dy dx = . □
Q R z=0 x=0 y=x/2 z=0 3

Mass and Center of Mass

Theorem 13.5.9. Suppose a solid Q has the mass density ρ (x, y, z). Then

1. Total mass m is given by

˚
m= ρ (x, y, z) dV.
Q

2. It has the center of mass (x̄, ȳ, z̄), where

Myz Mxz Mxy

x̄ = , ȳ = , z̄ = ,
m
˚ m m˚ ˚
Myz = xρ (x, y, z) dV, Mxz = yρ (x, y, z) dV, Mxy = zρ (x, y, z) dV,
Q Q Q
˚
1
i.e., x̄ = ˝ xρ (x, y, z) dV,
Q ρ (x, y, z) dV Q
˚
1
ȳ = ˝ yρ (x, y, z) dV,
Q ρ (x, y, z) dV Q
˚
1
z̄ = ˝ zρ (x, y, z) dV.
Q ρ (x, y, z) dV Q

√ 13.5.10. Find the center of mass of the solid of constant mass density ρ bounded by the graphs of
Example
z = x2 + y2 and z = 4.

Page 100 of 126

Calculus II for Engineers Fall, 2010

√
Figure 13.20: Solid bounded by z = x2 + y2 and z = 4

A NSWER. See the figure 13.20. Notice that the projection R of the solid onto the xy–plane is the disk of
radius 4 centered at the origin:
{ √ √ }
R = (x, y) : −4 ≤ x ≤ 4, − 16 − x2 ≤ y ≤ 16 − x2 .
√
Further, for each (x, y) ∈ R, z ranges from the cone (z = x2 + y2 ) up to the plane z = 4. So, the total mass
m of the solid is given by
˚ ¨ ˆ ˆ ˆ √ ˆ
y= 16−x2
z=4 x=4 z=4
64π
m= ρ (x, y, z) dV = ρ √ dz dA = ρ √ √ dz dy dx = ,
Q R z= x2 +y2 x=−4 y=− 16−x2 z= x2 +y2 3

which will be easily computed when we use the Cylindrical Coordinates (Section 13.6). We leave the center
of mass to the students. Practice and Exercise. □

Page 101 of 126

Calculus II for Engineers Fall, 2010

§13.6 Cylindrical Coordinates.

□ 13.6.1 Cylindrical Coordinates.

The cylindrical coordinates can be viewed as the extension of the polar coordinates into the three–coordinates
system. To be precise, in the cylindrical coordinate system, a point P(x, y, z) in three–dimensional space
is represented by the ordered triple (r, θ , z), where r and θ are polar coordinates of the projection of P onto
the xy–plane and z is the directed distance from the xy–plane to P. See the figure 13.21. To convert from
cylindrical to rectangular coordinates, we use the following relationship.

Figure 13.21: Cylindrical Coordinate System

Theorem 13.6.1 (R ELATION B ETWEEN R ECTANGULAR AND C YLINDRICAL C OORDINATES). A point

P(x, y, z) in the rectangular coordinates system corresponds to a point (r, θ , z) in the cylindrical coordinates
system by the following equations:

x = r cos θ , y = r sin θ , z = z.

So when we convert from rectangular to cylindrical coordinates, we use

y
r 2 = x 2 + y2 , tan θ = , z = z.
x
Example 13.6.2. Write the given equation in cylindrical coordinates: (1) x2 + (y − 3)2 = 9 and (2) z =
√
x 2 + y2 .
A NSWER. (1) In cylindrical coordinates, x = r cos θ and y = r sin θ . So the given equation becomes

(r cos θ )2 + (r sin θ − 3)2 = 9, r2 cos2 θ + r2 sin2 θ − 6r sin θ + 9 = 9,

r2 − 6r sin θ = 0, r = 6 sin θ .

(2) In cylindrical coordinates, r2 = x2 + y2 and so the equation becomes

√
z = r2 = |r|. □

Example 13.6.3. Write the given equation in rectangular coordinates: z = r.

A NSWER. From the relation above, we have

z2 = r2 = x2 + y2 , i.e., z2 = x2 + y2 . □

Exercise 13.6.4. Write the given equation in cylindrical coordinates: (1) z = e−x
2 −y2
and (2) x + 2y + 3z = 4.

Page 102 of 126

Calculus II for Engineers Fall, 2010

□ 13.6.2 Evaluating Triple Integrals with Cylindrical Coordinates.

We can use cylindrical coordinates to simplify certain triple integrals. Suppose we can write the solid Q as

Q = { (r, θ , z) : (r, θ ) ∈ R, k1 (r, θ ) ≤ z ≤ k2 (r, θ ) } , (13.6.1)

where k1 (r, θ ) ≤ k2 (r, θ ) for all (r, θ ) in the region R of the xy–plane defined by

R = { (r, θ ) : α ≤ θ ≤ β , g1 (θ ) ≤ r ≤ g2 (θ ) } , (13.6.2)

where 0 ≤ g1 (θ ) ≤ g2 (θ ) for all θ ∈ [α , β ]. Then we can write

˚ ¨ (ˆ z=k2 (r,θ )
)
f (r, θ , z) dV = f (r, θ , z) dz dA.
Q R z=k1 (r,θ )

Using the polar coordinates, we express the integral in the full form as below.

Theorem 13.6.5 (R EPRESENTATION OF T RIPLE I NTEGRAL IN C YLINDRICAL ˝ C OORDINATES). For Q and

R given in equations (13.6.1) and (13.6.2) above, the triple integral Q f (r, θ , z) dV can be expressed as
follows:
˚ ¨ (ˆ )
k2 (r,θ ) ˆ ˆ ˆ θ =β r=g2 (θ ) z=k2 (r,θ )
f (r, θ , z) dV = f (r, θ , z) dz dA = f (r, θ , z)r dz dr dθ .
Q R k1 (r,θ ) θ =α r=g1 (θ ) z=k1 (r,θ )

˝
Example 13.6.6. Set up the triple integral Q f (x, y, z) dV in cylindrical coordinates, where Q is the solid
above the xy–plane and below z = 9 − x2 − y2 .

A NSWER. See the figure 13.22. In cylindrical coordinates, z = 9 − x2 − y2 becomes z = 9 − r2 . We observe

the surface of z = 9 − x2 − y2 is the paraboloid and so Q has the circular base region. It implies
{ }
Q = (r, θ , z) : (r, θ ) ∈ R, 0 ≤ z ≤ 9 − r2 , R = { (r, θ ) : 0 ≤ θ ≤ 2π , 0 ≤ r ≤ 3 } .

Hence, the triple integral can be set up as follows:

˚ ¨ (ˆ 2
)
9−r ˆ θ =2π ˆ r=3 ˆ z=9−r2
f (r, θ , z) dV = f (r, θ , z) dz dA = f (r, θ , z)r dz dr dθ . □
Q R 0 θ =0 r=0 z=0

√
Figure 13.22: Solid between the xy–plane and z = 9 − x2 − y2 and Solid between z = x2 + y2 and z = 2

´ x=2 ´ y=√4−x2 ´ 2
Example 13.6.7. Evaluate √ √ (x2 + y2 )3/2 dz dy dx.
x=−2 y=− 4−x2 z= x2 +y2

Page 103 of 126

Calculus II for Engineers Fall, 2010

A NSWER. See the figure 13.22. We find the solid Q first in the rectangular coordinates:
{ √ √ √ }
Q = (x, y, z) : −2 ≤ x ≤ 2, − 4 − x2 ≤ y ≤ 4 − x2 , x2 + y2 ≤ z ≤ 2 .
We introduce the region R on xy–plane as follows:
{ √ }
Q = (x, y, z) : (x, y) ∈ R, x2 + y2 ≤ z ≤ 2 ,
{ √ √ }
R = (x, y) : −2 ≤ x ≤ 2, − 4 − x ≤ y ≤ 4 − x
2 2 .

Now we express Q and R in cylindrical coordinates: (Observe that R is a disk x2 + y2 ≤ 4)

Q = { (r, θ , z) : (r, θ ) ∈ R, r ≤ z ≤ 2 } , R = { (r, θ ) : 0 ≤ θ ≤ 2π , 0 ≤ r ≤ 2 } .
Using this representation, the integral becomes
ˆ x=2 ˆ y=√4−x2 ˆ 2
2
√ √ (x + y2 )3/2 dz dy dx
x=−2 y=− 4−x2 z= x2 +y2
ˆ θ =2π ˆ r=2 ˆ 2 ˆ θ =2π ˆ r=2 ˆ 2
= 2 3/2
(r ) r dz dr dθ = r4 dz dr dθ
θ =0 r=0 z=r θ =0 r=0 z=r
θ =2π ˆ r=2 [
ˆ ˆ θ =2π ˆ r=2
]2
= r4 z z=r dr dθ = r4 (2 − r) dr dθ
θ =0 r=0 θ =0 r=0
ˆ θ =2π [ 5 6 ]r=2 ˆ θ =2π
2r r 32 64π
= − dθ = dθ = . □
θ =0 5 6
r=0 θ =0 15 15
˝ √
2 2
Example 13.6.8. Set up and evaluate the triple integral Q ze x +y dV in the appropriate coordinate sys-
tem, where Q is the solid inside x2 + y2 = 4, outside x2 + y2 = 1 and between z = 0 and z = 3.

Figure 13.23: Solid inside x2 + y2 = 4, outside x2 + y2 = 1 and between z = 0 and z = 3

A NSWER. See the figure 13.23. We use the cylindrical coordinate system. In cylindrical coordinates, x2 +
y2 = 4 and x2 + y2 = 1 become r2 = 4, i.e., r = ±2, and r2 = 1, i.e., r = ±1, respectively. So the region Q
inside x2 + y2 = 4, outside x2 + y2 = 1 and between z = 0 and z = 3 can be expressed as follows:
Q = { (r, θ , z) : (r, θ ) ∈ R, 0 ≤ z ≤ 3 } , R = { (r, θ ) : 0 ≤ θ ≤ 2π , 1 ≤ r ≤ 2 } .
Hence, the triple integral can be set up as follows:
˚ √ ¨ (ˆ 3 ) ˆ θ =2π ˆ r=2 ˆ z=3
x2 +y2
ze dV = r
ze dz dA = zer r dz dr dθ
Q R 0 θ =0 r=1 z=0
(ˆ z=3 ) (ˆ r=2 ) (ˆ θ =2π )
9
= z dz er r dr 1 dθ = e2 (2π ) = 9π e2 . □
z=0 r=1 θ =0 2

Page 104 of 126

Calculus II for Engineers Fall, 2010

˝ √
Exercise 13.6.9. Evaluate Q x2 + y2 dV , where Q is the region that lies inside the cylinder x2 + y2 = 16
and between the planes z = −5 and z = 4.
˝
Exercise 13.6.10. Evaluate Q x2 dV , where Q is the solid that lies within the cylinder x2 + y2 = 1, above
the plane z = 0, and below the cone z2 = 4x2 + 4y2 .

Exercise 13.6.11. Evaluate the iterated integral after changing coordinate systems:
´3 ´0 ´ x2 +z2 ( 2 2 )
 −3 −√9−x2 0 x + z dy dz dx.
´ 1 ´ √1−x2 ´ 4 ( 2 )
 0 0 1−x −y
2 2 x + y2 dz dy dx.

Page 105 of 126

Calculus II for Engineers Fall, 2010

§13.7 Spherical Coordinates.

□ 13.7.1 Spherical Coordinates.

The spherical coordinate system simplifies the evaluation of triple integrals over regions bounded by spheres
or cones.
We start with a point P(x, y, z) in the rectangular coordinate system, i.e., the xyz–space. It is easy to put the
point P(x, y, z) in the rectangular coordinate system, which is shown in Figure 1. For the point P(x, y, z), we
can deduce three quantities

1. distance between O and P, denoted by ρ ,

2. angle from the positive z–axis to the line OP, denoted by ϕ ,
3. angle from the positive x–axis to the line OQ, denoted by θ .

See the figure 13.24. These three quantities forms the spherical coordinate system. The combination
of these three quantities, (ρ , ϕ , θ ), corresponds to the point P(x, y, z) and so we may say (ρ , ϕ , θ ) is the
representation of the point P(x, y, z) in the spherical coordinate system.

Figure 13.24: Spherical Coordinate System

Since ρ represents the distance, it should satisfy ρ ≥ 0. Since ϕ represents the angle from the positive z–axis
to the line OP, it should satisfy 0 ≤ ϕ ≤ π . We take 0 ≤ θ ≤ 2π as the range of the θ .
Let us find the specific relationship between the rectangular coordinates and the spherical coordinates.

Conversion from Spherical to Rectangular Coordinates. Suppose a point P is given in the spherical
coordinates, P(ρ , ϕ , θ ). Let us find its representation in the rectangular coordinates P(x, y, z).
We observe that the angle between OP and OQ is π2 − ϕ . It implies that the angle between OP and PQ is ϕ .
(Why?) So, from the triangle OPQ, we get

∥OQ∥ = ∥OP∥ sin ϕ , and ∥PQ∥ = ∥OP∥ cos ϕ ,

where ∥OP∥, ∥OQ∥ and ∥PQ∥ mean the lengths of the line segment OP, OQ and PQ, respectively.
It is easy to see z = ∥PQ∥, i.e.,

z = ∥PQ∥ = ∥OP∥ cos ϕ , z = ∥OP∥ cos ϕ .

Again from the triangle having the angle θ on the xy–plane, we observe

x = ∥OQ∥ cos θ , and y = ∥OQ∥ sin θ .

When we put ∥OQ∥ = ∥OP∥ sin ϕ into the equations, we get

x = ∥OQ∥ cos θ = ∥OP∥ sin ϕ cos θ , and y = ∥OQ∥ sin θ = ∥OP∥ sin ϕ sin θ .

Page 106 of 126

Calculus II for Engineers Fall, 2010

Since the length of OP is ∥OP∥ = ρ , hence, we finally deduce

x = ∥OP∥ sin ϕ cos θ = ρ sin ϕ cos θ ,
y = ∥OP∥ sin ϕ sin θ = ρ sin ϕ sin θ ,
z = ∥OP∥ cos ϕ = ρ cos ϕ .
That is, in the rectangular coordinates, we have P (ρ sin ϕ cos θ , ρ sin ϕ sin θ , ρ cos ϕ ).

Conversion from Rectangular to Spherical Coordinates. Suppose a point P is given in the rectangular
coordinates, P(x, y, z). Let us find its representation in the spherical coordinates P(ρ , ϕ , θ ).
Since ρ represents the length of OP, we get
√ √
ρ = ∥OP∥ = x2 + y2 + z2 , i.e., ρ = x2 + y2 + z2 .
Since z is equal to the length of PQ, we get
√ z
z = ∥PQ∥ = ∥OP∥ cos ϕ = x2 + y2 + z2 cos ϕ , i.e., cos ϕ = √ ,
x2 + y2 + z2
( )
z
i.e., 0 ≤ ϕ = cos−1 √ ≤ π.
x2 + y2 + z2
From the triangle having the angle θ on the xy–plane, we observe
( )
√ x x
x = ∥OQ| cos θ = x2 + y2 cos θ , cos θ = √ , 0 ≤ θ = cos−1 √ ≤ 2π .
x + y2
2 x + y2
2

That is,
√
0≤ρ = x2 + y2 + z2 ,
( )
z
0 ≤ ϕ = cos−1 √ ≤ π,
x2 + y2 + z2
( ) ( )
x y
0 ≤ θ = cos−1 √ = sin−1 √ ≤ 2π .
x 2 + y2 x 2 + y2
We summarize as follows:
Theorem 13.7.1 (R ELATIONSHIP B ETWEEN R ECTANGULAR AND S PHERICAL C OORDINATES). A point
P(x, y, z) in the rectangular coordinate system corresponds to a point (ρ , ϕ , θ ) in the spherical coordinate
system by the following equations:
x = ρ sin ϕ cos θ , y = ρ sin ϕ sin θ , z = ρ cos ϕ .
It is worthwhile to memorize the equation,
√
ρ = x2 + y2 + z2 , or ρ 2 = x2 + y2 + z2 .
Example 13.7.2. Find rectangular coordinates for the point described by (ρ , ϕ , θ ) = (8, π /4, π /3) in spheri-
cal coordinates.

A NSWER. By the Theorem 13.7.1 above, we get

π π √ π π √ π √
x = 8 sin cos = 2 2 , y = 8 sin sin = 2 6 , z = 8 cos = 4 2.
4 3 4 3 4
√ √ √
Hence, we get (x, y, z) = (2 2 , 2 6 , 4 2 ). See the figure 13.25. □

Page 107 of 126

Calculus II for Engineers Fall, 2010

Figure 13.25: Spherical Coordinate System

√
Exercise 13.7.3. The point is given (0, 2 3 , −2) in rectangular coordinates. Find spherical coordinates for
this point.
Example 13.7.4. Rewrite the equation of the cone z2 = x2 + y2 in spherical coordinates.
A NSWER. Putting x = ρ sin ϕ cos θ , y = ρ sin ϕ sin θ , and z = ρ cos ϕ into the equation, we get
(ρ cos ϕ )2 = (ρ sin ϕ cos θ )2 + (ρ sin ϕ sin θ )2 ,
( )
simply, 0 = ρ 2 cos2 ϕ − sin2 ϕ = ρ 2 (cos ϕ − sin ϕ ) (cos ϕ + sin ϕ )
−→ ρ = 0 or cos ϕ − sin ϕ = 0 or cos ϕ + sin ϕ = 0
π 3π
−→ ρ = 0 or ϕ= or ϕ= . □
4 4

□ 13.7.2 Triple Integrals in Spherical Coordinates.

We can use spherical coordinates to simplify a triple integral.
Theorem ˝13.7.5 (R EPRESENTATION OF T RIPLE I NTEGRAL IN S PHERICAL C OORDINATES). The triple
integral Q f (r, θ , z) dV can be expressed as follows:
˚ ˚
f (x, y, z) dV = f (ρ sin ϕ cos θ , ρ sin ϕ sin θ , ρ cos ϕ )ρ 2 sin ϕ dρ dϕ dθ .
Q Q

Note that dV = ρ 2 sin ϕ dρ dϕ dθ .

˝ 2 2 2 3/2
Example 13.7.6. Evaluate Q e(x +y +z ) dV , where Q is the unit ball:
{ }
Q = (x, y, z) : x2 + y2 + z2 ≤ 1 .
A NSWER. In spherical coordinates, Q has the representation
{ }
Q = (ρ , ϕ , θ ) : ρ 2 ≤ 1 = { (ρ , ϕ , θ ) : 0 ≤ ρ ≤ 1, 0 ≤ ϕ ≤ π , 0 ≤ θ ≤ 2π } .
Hence, the triple integral can be set up and computed as follows:
˚ ˚ ˆ θ =2π ˆ ϕ =π ˆ ρ =1
ρ3 4π (e − 1)
(x2 +y2 +z2 )3/2
eρ ρ 2 sin ϕ dρ dϕ dθ =
3
e dV = e dV = . □
Q Q θ =0 ϕ =0 ρ =0 3
Remark 13.7.7. It would have been extremely awkward to evaluate the integral in the Example 13.7.6 above
using rectangular coordinates rather than spherical coordinates. In rectangular coordinates the iterated inte-
gral would have been
˚ ˆ x=1 ˆ y=√1−x2 ˆ z=√1−x2 −y2
2 2 2 3/2 2 2 2 3/2
e(x +y +z ) dV = √ √ e(x +y +z ) dz dy dx.
Q x=−1 y=− 1−x2 z=− 1−x2 −y2

Page 108 of 126

Calculus II for Engineers Fall, 2010

Example 13.7.8. Use spherical coordinates to find the volume of the solid that lies above the cone z =
√
x2 + y2 and below the sphere x2 + y2 + z2 = z.

√
Figure 13.26: Solid between cone z = x2 + y2 and sphere x2 + y2 + z2 = z

A NSWER. See the figure

√ 13.26. We express the equations of cone and sphere in the spherical coordinate
system. (1) Cone z = x2 + y2 becomes ϕ = π4 . (Why?) (2) Sphere x2 + y2 + z2 = z becomes ρ = cos ϕ .
(Why?) So, the solid Q has the representation:
{ π }
Q = (ρ , ϕ , θ ) : 0 ≤ θ ≤ 2π , 0 ≤ ϕ ≤ , 0 ≤ ρ ≤ cos ϕ .
4
Using this representation, the volume of the solid Q is obtained as follows:
˚ ˆ θ =2π ˆ ϕ =π /4 ˆ ρ =cos ϕ
π
V= dV = ρ 2 sin ϕ dρ dϕ dθ = . □
Q θ =0 ϕ =0 ρ =0 8

Example 13.7.9. Find the volume lying inside the sphere x2 + y2 + z2 = 2z and inside the cone z2 = x2 + y2 .
A NSWER. We express the equations of cone and sphere in the spherical coordinate system. (1) Cone z =
√
x2 + y2 becomes ϕ = cos ϕ . (Why?) (2) Sphere x2 + y2 + z2 = 2z becomes ρ = 2 cos ϕ . (Why?) So, the
solid Q has the representation:
{ π }
Q = (ρ , ϕ , θ ) : 0 ≤ θ ≤ 2π , 0 ≤ ϕ ≤ , 0 ≤ ρ ≤ 2 cos ϕ .
4
Using this representation, the volume of the solid Q is obtained as follows:
˚ ˆ θ =2π ˆ ϕ =π /4 ˆ ρ =2 cos ϕ
V= dV = ρ 2 sin ϕ dρ dϕ dθ = π . □
Q θ =0 ϕ =0 ρ =0

´ 2 ´ √4−x2 ´ √4−x2 −y2 ( )

Example 13.7.10. Evaluate the triple iterated integral −2 0 0 x2 + y2 + z2 dz dy dx.

A NSWER. See the figure 13.27. Let us find the expression of the integral solid Q from the given triple
integral: { √ }
√
Q = (x, y, z) : −2 ≤ x ≤ 2, 0 ≤ y ≤ 4 − x2 , 0 ≤ z ≤ 4 − x2 − y2 .
Can you imagine the shape of the solid Q? When we ignore the z–component of the points in Q, we observe
a upper–half disk x2 + y2 ≤ 4. Explicitly,
{ √ } { √ }
Q = (x, y, z) : (x, y) ∈ R, 0 ≤ z ≤ 4 − x2 − y2 , R = (x, y) : −2 ≤ x ≤ 2, 0 ≤ y ≤ 4 − x2

Page 109 of 126

Calculus II for Engineers Fall, 2010

√
Figure 13.27: Solid bounded by z = 4 − x2 − y2 and y = 0 and z = 0

and R is the upper–half disk x2 + y2 ≤ 4. That is, the projection of Q onto the xy–plane is the upper–half disk
x2 + y2 ≤ 4. √
From the expression of Q, we also observe that the solid is under the curve z = 4 − x2 − y2 , which is a upper
sphere of the ball centered at the origin with the radius 4, because the equation implies z2 = 4 − x2 − y2 , i.e.,
x2 + y2 + z2 = 4.
Now, we use the spherical coordinates and express the solid and the integral:
{ π }
Q = (ρ , ϕ , θ ) : 0 ≤ θ ≤ π , 0 ≤ ϕ ≤ , 0 ≤ ρ ≤ 2 . (Why?)
2
Hence, we get
ˆ ˆ √ ˆ √4−x2 −y2 ˚ ˚
2 4−x2 ( ) ( )
2 2
x +y +z 2
dz dy dx = 2
x +y +z 2 2
dV = ρ 2 dV
−2 0 0 Q Q
ˆ θ =π ˆ ϕ =π /2 ˆ ρ =2 ( ) 32π
= ρ 2 ρ 2 sin ϕ dρ dϕ dθ = .□
θ =0 ϕ =0 ρ =0 5

§13.8 Change of Variables in Multiple Integrals.

Skip. Please read the textbook.

Page 110 of 126

 Chapter 14

Vector Calculus

§14.1 Vector Fields.

Skip. Please read the textbook.

§14.2 Line Integrals.

□ 14.2.1 Line Integral.

We start with the evaluation theorem on how to compute the line integral.
Theorem 14.2.1 (E VALUATION T HEOREM).

1. Function of Two Variables. Suppose that f (x, y) is continuous in a region D containing the curve C
and that C is described parametrically by (x(t), y(t)) for a ≤ t ≤ b, where x(t) and y(t) have continuous
first derivatives. Then,
ˆ ˆ t=b √
f (x, y) ds = f (x(t), y(t)) [x′ (t)]2 + [y′ (t)]2 dt.
C t=a

2. Function of Three Variables. Suppose that f (x, y, z) is continuous in a region D containing the curve
C and that C is described parametrically by (x(t), y(t), z(t)) for a ≤ t ≤ b, where x(t), y(t) and z(t) have
continuous first derivatives. Then,
ˆ ˆ t=b √
f (x, y, z) ds = f (x(t), y(t), z(t)) [x′ (t)]2 + [y′ (t)]2 + [z′ (t)]2 dt.
C t=a

Example 14.2.2. Find the mass of a spring in the shape of the helix defined parametrically by x = 2 cost,
y = t, z = 2 sint, for 0 ≤ t ≤ 6π with density ρ (x, y, z) = 2y.
A NSWER. Parametrically, the density becomes ρ (x, y, z) = 2y = 2t. The arc length element ds is given by
√ √
ds = [x′ (t)]2 + [y′ (t)]2 + [z′ (t)]2 dt = 5 dt.

Hence, the mass of the spring is obtained as follows:

ˆ ˆ t=6π √ √
m = ρ (x, y, z) ds = 2t 5 dt = 36π 2 5 . □
C t=0
´
Example 14.2.3. Evaluate the line integral C (2x2 − 3yz) ds, where C is the curve defined parametrically by
x = cost, y = sint, z = cost, with 0 ≤ t ≤ 2π .
A NSWER. Parametrically, the integrand becomes f (x, y, z) = 2x2 − 3yz = 2 cos2 t − 3 sint cost. The arc length
element ds is given by
√ √
ds = [x′ (t)]2 + [y′ (t)]2 + [z′ (t)]2 dt = 1 + sin2 t dt.

Hence, the line integral becomes

ˆ ˆ t=2π ( )√
(2x − 3yz) ds =
2
2 cos2 t − 3 sint cost 1 + sin2 t dt ≈ 6.9922. □
C t=0

111
Calculus II for Engineers Fall, 2010

Theorem 14.2.4 (S EPARATION). Suppose that f (x, y, z) is a continuous function in some region D containing
the oriented curve C. Then, if C is piecewise–smooth, with C = C1 ∪C2 ∪ · · · ∪Cn , where C1 , C2 , . . . , Cn are
all smooth and where the terminal point of Ci is the same as the initial point of Ci+1 , for i = 1, 2, . . . , n − 1,
we have
´ ´
1. −C f (x, y, z) ds = C f (x, y, z) ds.
´ ´ ´ ´
2. C f (x, y, z) ds = C1 f (x, y, z) ds + C2 f (x, y, z) ds + · · · + Cn f (x, y, z) ds.

Theorem
´ 14.2.5 (L ENGTH OF C URVE). For any piecewise–smooth curve C (in two or three dimensions),
C 1 ds gives the arc length of the curve C.

Theorem 14.2.6 (E VALUATION T HEOREM). Suppose that f (x, y, z) is continuous in a region D containing
the curve C and that C is described parametrically by (x(t), y(t), z(t)) for a ≤ t ≤ b, where x(t), y(t) and z(t)
have continuous first derivatives. Then,
ˆ ˆ t=b
f (x, y, z) dx = f (x(t), y(t), z(t))x′ (t) dt
C t=a
ˆ ˆ t=b
f (x, y, z) dy = f (x(t), y(t), z(t))y′ (t) dt
C t=a
ˆ ˆ t=b
f (x, y, z) dz = f (x(t), y(t), z(t))z′ (t) dt
C t=a
´
Example 14.2.7. Compute the line integral C (4xz + 2y) dx, where C is the line segment (1) from (2, 1, 0) to
(4, 0, 2) and (2) from (4, 0, 2) to (2, 1, 0).

A NSWER. We solve only (1): First, we parameterize the line segment from (2, 1, 0) to (4, 0, 2) as follows:

x = 2 + (4 − 2)t = 2 + 2t, y = 1 + (0 − 1)t, z = 0 + (2 − 0)t = 2t, 0 ≤ t ≤ 1.

Then the integrand becomes

4xz + 2y = 4(2 + 2t)(2t) + 2(1 − t) = 16t 2 + 14t + 2.

The element dx becomes

dx = x′ (t) dt = 2 dt.
Hence, the line integral is obtained as follows:
ˆ ˆ t=1 ( ) 86
(4xz + 2y) dx = 16t 2 + 14t + 2 2 dt = . □
C t=0 3

Remark 14.2.8 (V ERY I MPORTANT N OTATION). For the convenience, we usually write
ˆ ˆ ˆ ˆ
f (x, y, z) dx + g(x, y, z) dy + h(x, y, z) dz = f (x, y, z) dx + g(x, y, z) dy + h(x, y, z) dz.
C C C C

□ 14.2.2 Application: Work.

Theorem 14.2.9 (W ORK D ONE). The work done W by the force field F(x, y, z) in moving a particle along
the curve C is obtained as follows: ˆ
W = F(x, y, z) · dr,
C
where dr = ⟨ dx, dy, dz⟩ is the element of the parametric curve C.

Page 112 of 126

Calculus II for Engineers Fall, 2010

Example 14.2.10. Compute the work done by the force field F(x, y, z) = ⟨4y, 2xz, 3y⟩ acting on an object as
it moves along the helix defined parametrically by x = 2 cost, y = 2 sint and z = 3t, from the point (2, 0, 0)
to the point (−2, 0, 3π ).

A NSWER. By the formula, the work done is given by

ˆ ˆ
W = F(x, y, z) · dr = 4y dx + 2xz dy + 3y dz.
C C

We observe the point (2, 0, 0) occurs when t = 0 and the point (−2, 0, 3π ) occurs when t = π . Hence with
dx = −2 sint dt, dy = 2 cost dt and dz = 3 dt, the line integral becomes
ˆ
W = 4y dx + 2xz dy + 3y dz
ˆCt=π
= 4(2 sint)(−2 sint) dt + 2(2 cost)(3t)(2 cost) dt + 3(2 sint)(3) dt
t=0
ˆ t=π
= [4(2 sint)(−2 sint) + 2(2 cost)(3t)(2 cost) + 3(2 sint)(3)] dt
t=0
ˆ t=π
[ ]
= −16 sin2 t + 24t cos2 t + 18 sint dt = 36 − 8π + 6π 2 . □
t=0

Example 14.2.11. Compute the work done by the force field F(x, y) = ⟨y, −x⟩ acting on an object as it moves
along the parabola y = x2 − 1 from (1, 0) to (−2, 3).

A NSWER. By the formula, the work done is given by

ˆ ˆ
W = F(x, y) · dr = y dx − x dy.
C C

We use x = t and y = t 2 − 1 as parametric equations for the curve, with t ranging from t = 1 to t = −2. In
this case, dx = dt and dy = 2t dt. Hence the work is
ˆ ˆ t=−2
[ 2 ]
W = y dx − x dy = (t − 1)1 − t(2t) dt = 6. □
C t=1

§14.3 Independence of Path and Conservative Vector Fields.

Skip. Please read the textbook.

§14.4 Green’s Theorem.

Skip. Please read the textbook.

Page 113 of 126

Calculus II for Engineers Fall, 2010

§14.5 Curl and Divergence.

In this section, we briefly discuss the curl and divergence, which are generalizations of the notion of deriva-
tive applied to vector fields.

□ 14.5.1 Curl.
Definition 14.5.1. The curl of the vector field F(x, y, z) = ⟨F1 (x, y, z), F2 (x, y, z), F3 (x, y, z)⟩ is the vector field
defined by ( ) ( ) ( )
∂ F3 ∂ F2 ∂ F1 ∂ F3 ∂ F2 ∂ F1
Curl F = − i+ − j+ − k,
∂y ∂z ∂z ∂x ∂x ∂y
defined at all points at which all the indicated partial derivatives exist.
We memorize the equation Curl F = ∇ × F:
 
 i k  (
 j ) ( ) ( )
 ∂ ∂ ∂  ∂ F3 ∂ F2 ∂ F1 ∂ F3 ∂ F2 ∂ F1

∇×F =  = − i+ − j+ − k = Curl F.
 ∂x ∂y ∂z  ∂y ∂z ∂z ∂x ∂x ∂y
 
 F1 F2 F3 
⟨ ⟩
Example 14.5.2. Compute Curl F for (1) F(x, y, z) = x2 i − 3xyj + xk and (2) F(x, y, z) = xy2 , 3y2 z2 , 2x − zy3 .
A NSWER. From the definition, we need the following derivatives:
∂ F1 ∂ F1
1. Differentiate F1 with respect to only y and z: ∂y and ∂z
∂ F2 ∂ F2
2. Differentiate F2 with respect to only x and z: ∂x and ∂z
∂ F3 ∂ F3
3. Differentiate F3 with respect to only x and y: ∂x and ∂y

(1) We compute those partial derivatives mentioned above:

∂ F1 ∂ ( 2) ∂ F1 ∂ ( 2)
= x = 0, = x = 0,
∂y ∂y ∂z ∂z
∂ F2 ∂ ∂ F2 ∂
= (−3xy) = −3y, = (−3xy) = 0,
∂x ∂x ∂z ∂z
∂ F3 ∂ ∂ F3 ∂
= (x) = 1, = (x) = 0.
∂x ∂x ∂y ∂y
Putting them into the definition, we have
( ) ( ) ( )
∂ F3 ∂ F2 ∂ F1 ∂ F3 ∂ F2 ∂ F1
Curl F = − i+ − j+ − k
∂y ∂z ∂z ∂x ∂x ∂y
= (0 − 0) i + (0 − 1) j + (−3y − 0) k = −j − 3k.
(2) We compute those partial derivatives mentioned above:
∂ F1 ∂ ( 2) ∂ F1 ∂ ( 2)
= xy = 2xy, = xy = 0,
∂y ∂y ∂z ∂z
∂ F2 ∂ ( 2 2) ∂ F2 ∂ ( 2 2)
= 3y z = 0, = 3y z = 6y2 z,
∂x ∂x ∂z ∂z
∂ F3 ∂ ( ) ∂ F3 ∂ ( )
= 2x − zy3 = 2, = 2x − zy3 = −3zy2 .
∂x ∂x ∂y ∂y
Putting them into the definition, we have
( ) ( ) ( )
∂ F3 ∂ F2 ∂ F1 ∂ F3 ∂ F2 ∂ F1
Curl F = − i+ − j+ − k
∂y ∂z ∂z ∂x ∂x ∂y
( )
= −3zy2 − 6y2 z i + (0 − 2) j + (0 − 2xy) k = −9y2 zi − 2j − 2xyk. □

Page 114 of 126

Calculus II for Engineers Fall, 2010

Example 14.5.3. Compute the curl for (1) F(x, y, z) = xi + yj and (2) G(x, y, z) = yi − xj.

A NSWER. A simple computation shows

Curl F = ⟨0, 0, 0⟩ = 0, Curl G = ⟨0, 0, −2⟩ = −2k. □

Definition 14.5.4. If the vector field F satisfies ∇ × F = 0, i.e., Curl F = 0, at a point, then we say that the
vector field F is irrotational at that point. (That is, the fluid does not tend to rotate near the point.)

□ 14.5.2 Divergence.
Another important measure on the behavior of a fluid is the divergence.

Definition 14.5.5. The divergence of the vector field F(x, y, z) = ⟨F1 (x, y, z), F2 (x, y, z), F3 (x, y, z)⟩ is the scalar
function defined by
∂ F1 ∂ F2 ∂ F3
Div F(x, y, z) = + + ,
∂x ∂y ∂z
defined at all points at which all the indicated partial derivatives exist.

We memorize the equation Div F = ∇ · F:

⟨ ⟩
∂ ∂ ∂ ∂ F1 ∂ F2 ∂ F3
∇·F = , , · ⟨F1 , F2 , F3 ⟩ = + + = Div F.
∂x ∂y ∂z ∂x ∂y ∂z
Example 14.5.6. Compute the divergence for the given vector field:

1. F(x, y, z) = xez i + yz2 j + (x + y)k.

⟨ ⟩
2. F(x, y, z) = y2 , x2 ez , cos(xy) .

A NSWER. By the definition, we have

∂ F1 ∂ F2 ∂ F3 ∂ ∂ ( 2) ∂
1. Div F = + + = (xez ) + yz + (x + y) = ez + z2 + 0 = ez + z2 .
∂x ∂y ∂z ∂x ∂y ∂z
∂ F1 ∂ F2 ∂ F3 ∂ ( 2) ∂ ( 2 z) ∂
2. Div F = + + = y + x e + (cos(xy)) = 0 + 0 + 0 = 0. □
∂x ∂y ∂z ∂x ∂y ∂z
Example 14.5.7. Compute the divergence for F(x, y, z) = xi + yj and G(x, y, z) = yi − xj.

A NSWER. By the definition, we have

Div F = 2, Div G = 0. □

Definition 14.5.8. If the vector field F satisfies ∇ · F = 0, i.e., Div F = 0, throughout some region D, then we
say that the vector field F is source–free or incompressible.

Example 14.5.9. If f (x, y, z) is a scalar function and F(x, y, z) is a vector field, determine whether each
operation is a scalar function, a vector field, or undefined: (1) ∇ × (∇F), (2) ∇ × (∇ · F), (3) ∇ · (∇ f ).

We can now summarize a number of equivalent properties for three–dimensional vector fields.

Theorem 14.5.10 (C ONSERVATIVE V ECTOR F IELDS). Suppose F(x, y, z) = ⟨F1 (x, y, z), F2 (x, y, z), F3 (x, y, z)⟩
is a vector field whose components F1 , F2 and F3 have continuous first partial derivatives throughout all of
R3 . Then the followings are equivalent1 :

1. F(x, y, z) is conservative.
1 equal in amount or value; like, equal, same; being essentially equal to something

Page 115 of 126

Calculus II for Engineers Fall, 2010

´
2. C F · dr is independent of path.
´
3. C F · dr = 0 for every piecewise–smooth closed curve C.
4. ∇ × F = 0.
5. F(x, y, z) is a gradient field (i.e., F = ∇ f for some potential function f ).

§14.6 Surface Integrals.

Skip. Please read the textbook.

§14.7 The Divergence Theorem.

Skip. Please read the textbook.

§14.8 Stokes’ Theorem.

Skip. Please read the textbook.

§14.9 Applications of Vector Calculus.

Skip. Please read the textbook.

Page 116 of 126

 Chapter 15

Appendix: Distances between Various Objects

§15.1 Two–Dimensional Space.

□ 15.1.1 Two Points P(x0 , y0 ) and Q(x1 , y1 ).

.Q(x1 , y1 )

.distance d

.
.P(x0 , y0 )
−→ −→
Using the vector PQ, the distance d between P and Q is the magnitude of the vector PQ:
√
−→
d = ∥PQ∥ = ∥ ⟨x1 − x0 , y1 − y0 ⟩ ∥ = (x1 − x0 )2 + (y1 − y0 )2

Theorem 15.1.1. The distance d between two points P(x0 , y0 ) and Q(x1 , y1 ) is obtained by
√
d = (x1 − x0 )2 + (y1 − y0 )2 .

□ 15.1.2 Line L : y = mx + b and Point P(x0 , y0 ) not lying on L.

.P(x0 , y0 ) .Line L : y = mx + b

.90◦
.distance d .Q(x1 , y1 )

.
The line L : y = mx + b can be parameterized by
L : x = t,
y = mt + b,
−→
which is parallel to the vector v = ⟨1, m⟩. Two vectors PQ = ⟨x1 − x0 , y1 − y0 ⟩ and v = ⟨1, m⟩ should be
perpendicular, i.e.,
x1 − x0 + (y1 − y0 )m = 0. (15.1.1)
Since Q(x1 , y1 ) lies on the line L, so
y1 = mx1 + b. (15.1.2)
Solving two equations (15.1.1) and (15.1.2) for x1 and y1 , we have
x0 + (y0 − b)m b + (x0 + my0 )m
x1 = , y1 = .
1 + m2 1 + m2
Thus, the distance d between P and Q (which is the distance between P and the line L) is the magnitude of
−→
the vector PQ and it is computed by
√
−→ |mx0 + b − y0 |
d = ∥PQ∥ = ∥ ⟨x1 − x0 , y1 − y0 ⟩ ∥ = (x1 − x0 )2 + (y1 − y0 )2 = √ .
1 + m2

117
Calculus II for Engineers Fall, 2010

Theorem 15.1.2. The distance d between the line L : y = mx + b and the point P(x0 , y0 ) not lying on L is
obtained by
|mx0 + b − y0 |
d= √ .
1 + m2
Example 15.1.3. Find the distance between the line y = 2x + 3 and the point P(0, 5).

A NSWER. By the Theorem 15.1.2, the distance d is computed by

|mx0 + b − y0 | |2(0) + 3 − 5| 2
d= √ = √ =√ . □
1 + m2 1 + 22 5

□ 15.1.3 Parallel Lines L : y = mx + b and M : y = mx + c.

.Line L : y = mx + b .Line M : y = mx + c

.P(x0 , y0 )
.90◦ .distance d
.90◦ .Q(x1 , y1 )

.
Applying the Theorem (15.1.2) to the point P(x0 , y0 ) and the line M : y = mx + c, we deduce the distance,

|mx0 + c − y0 | |c − b|
d= √ =√ ,
1 + m2 1 + m2
where y0 = mx0 + b is used.

Theorem 15.1.4. The distance d between parallel lines L : y = mx + b and M : y = mx + c is obtained by

|c − b|
d=√ .
1 + m2
Example 15.1.5. Find the distance between two parallel lines L : y = 2x + 3 and M : y = 2x + 7.

A NSWER. By the Theorem 15.1.4, the distance d is computed by

|c − b| |7 − 3| 4
d=√ =√ =√ . □
1 + m2 1 + 22 5

Page 118 of 126

Calculus II for Engineers Fall, 2010

§15.2 Three–Dimensional Space.

□ 15.2.1 Two Points P(x0 , y0 , z0 ) and Q(x1 , y1 , z1 ).

.Q(x1 , y1 , z1 )

.distance d

.
.P(x0 , y0 , z0 )
−→ −→
Using the vector PQ, the distance d between P and Q is the magnitude of the vector PQ:
√
−→
d = ∥PQ∥ = ∥ ⟨x1 − x0 , y1 − y0 , z1 − z0 ⟩ ∥ = (x1 − x0 )2 + (y1 − y0 )2 + (z1 − z0 )2

Theorem 15.2.1. The distance d between two points P(x0 , y0 , z0 ) and Q(x1 , y1 , z1 ) is obtained by
√
d = (x1 − x0 )2 + (y1 − y0 )2 + (z1 − z0 )2 .

□ 15.2.2 Line L and Point Q(x0 , y0 , z0 ) not lying on L:

L : x = x1 + at, y = y1 + bt, z = z1 + ct.

.Q(x0 , y0 , z0 ) .Line L : x = x1 + at, y = y1 + bt, z = z1 + ct

.distance d
.90◦ .R(x2 , y2 , z2 )

. .P(x1 , y1 , z1 )
We choose two points P(x1 , y1 , z1 ) and R(x2 , y2 , z2 ) lying on the line L. Then by the formula on the cross
product, we have
−→ − →
−→ − → −→ − → −→ ∥PQ × PR∥
∥PQ × PR∥ = ∥PQ∥∥PR∥ sin θ = ∥PR∥(distance d), d= −→ . (15.2.1)
∥PR∥
−
→
The vector v = ⟨a, b, c⟩ is parallel to the line L and points P and R are on the line L. Thus, PR is parallel to
the vector v = ⟨a, b, c⟩ and so
−
→
PR = sv,
for some scalar s. It implies that the formula (15.2.1) on the distance d can be simplified by
(−→ )
−→ − → ∥s PQ × v ∥ ∥− →
∥PQ × PR∥ PQ × v∥
d= −→ = = . (15.2.2)
∥PR∥ ∥sv∥ ∥v∥
From the formula (15.2.2), we observe that to compute the distance, we need to choose only one point P on
the line L and use the vector v parallel to the line L.
Theorem 15.2.2. The distance d between the line L and the point Q(x0 , y0 , z0 ) not lying on L : x = x1 +at, y =
y1 + bt, z = z1 + ct is obtained by
−→
∥PQ × v∥
d= ,
∥v∥
where v = ⟨a, b, c⟩ and P is any point on the line L.

Page 119 of 126

Calculus II for Engineers Fall, 2010

Example 15.2.3. Find the distance between the line L and the point Q(3, −1, 7) not lying on L:
L : x = 2 − 2t, y = 2 + t, z = −1 + 6t.
A NSWER. The vector v = ⟨−2, 1, 6⟩ is parallel to the line L. When we choose a point P(2, 2, −1) from the
−→
line L by setting t = 0, we obtain the vector PQ = ⟨1, −3, 8⟩. By the Theorem (15.2.2), the distance between
the line L and the point Q is obtained by
−→ √
∥PQ × v∥ ∥ ⟨1, −3, 8⟩ × ⟨−2, 1, 6⟩ ∥ ∥ ⟨−26, −22, −5⟩ ∥ 1185
d= = = = √ . □
∥v∥ ∥ ⟨−2, 1, 6⟩ ∥ ∥ ⟨−2, 1, 6⟩ ∥ 41

□ 15.2.3 Parallel Lines L and M:

L : x = x0 + at, y = y0 + bt, z = z0 + ct and M : x = x1 + a′ s, y = y1 + b′ s, z = z1 + c′ s.

.Line L : x = x0 + at, y = y0 + bt, z = z0 + ct

.Q(x1 , y1 , z1 )
.90◦ .distance d
.P(x0 , y0 , z0 ) .90◦

. .Line M : x = x1 + a′ s, y = y1 + b′ s, z = z1 + c′ s
We fix one of the lines and choose a point from the other line. Note that there exists a particular scalar k such
that v′ = ⟨a′ , b′ , c′ ⟩ = k ⟨a, b, c⟩ = kv by the condition that L and M are parallel.
(1) Let us fix the line L and choose a point Q(x1 , y1 , z1 ) from the line M. We choose a point P(x0 , y0 , z0 ) from
−→
the line L and obtain the vector PQ. By the Theorem 15.2.2, the distance between the line L and the point Q
is obtained by
−→
∥PQ × v∥
d= ,
∥v∥
which is exactly same as the Theorem 15.2.2.
(2) Let us fix the line M and choose a point P(x0 , y0 , z0 ) from the line L. We choose a point Q(x1 , y1 , z1 ) from
−→
the line M and obtain the vector PQ. By the Theorem 15.2.2, the distance between the line M and the point
P is obtained by (−→ )
−→ ′ ∥k PQ × v ∥ ∥− →
∥PQ × v ∥ PQ × v∥
d= = = ,
∥v′ ∥ ∥kv∥ ∥v∥
which is exactly same as the Theorem 15.2.2.
For this reason, we can use either the vector parallel to the line L or the vector parallel to the line M in the
Theorem 15.2.2. If we choose P from the line L, then we should choose Q from the other line M and vice
versa.
Theorem 15.2.4. The distance d between parallel lines L : x = x0 + at, y = y0 + bt, z = z0 + ct and M : x =
x1 + a′ s, y = y1 + b′ s, z = z1 + c′ s is obtained by
−→ −→
∥PQ × v∥ ∥PQ × v′ ∥
d= , or d= ,
∥v∥ ∥v′ ∥
where v = ⟨a, b, c⟩. If we choose P from the line L, then we should choose Q from the other line M and vice
versa.
Example 15.2.5 (M IDTERM E XAM P ROBLEM). Find the distance between two lines L and M:
L : x = 1 + 2t, y= 3, z = −1 − 4t,
M : x = 2 − s, y = 2, z = 3 + 2s.

Page 120 of 126

Calculus II for Engineers Fall, 2010

A NSWER. We observe that the lines L and M are parallel to the vector v = ⟨1, 0, −2⟩. We choose points
−→
P(1, 3, −1) and Q(2, 2, 3) from each line L and M, respectively, and form the vector PQ = ⟨1, −1, 4⟩ and use
the Theorem 15.2.4:
−→ √
∥PQ × v∥ ∥ ⟨1, −1, 4⟩ × ⟨1, 0, −2⟩ ∥ ∥ ⟨2, 6, 1⟩ ∥ 41
d= = = = √ . □
∥v∥ ∥ ⟨1, 0, −2⟩ ∥ ∥ ⟨1, 0, −2⟩ ∥ 5

□ 15.2.4 Skew Lines L and M (Version 1):

L : x = xL + aLt, y = yL + bLt, z = zL + cLt and
M : x = xM + aM s, y = yM + bM s, z = zM + cM s.

.P(x0 , y0 , z0 )
.Line L : x = xL + aLt, y = yL + bLt, z = zL + cLt
.90◦

.distance d .Line M : x = xM + aM s, y = yM + bM s, z = zM + cM s

.90◦
.Q(x1 , y1 , z1 )

.
Let PL (xL , yL , zL ) and QM (xM , yM , zM ) be the points on L and M given in the parametric equations of L and M
and let vL = ⟨aL , bL , cL ⟩ and vM = ⟨aM , bM , cM ⟩.
Let P(x0 , y0 , z0 ) and Q(x1 , y1 , z1 ) be points on L and M, respectively, such that the line passing through both
P and Q is perpendicular to both lines L and M. Then the length of the line segment of PQ or the magnitude
−→
of the vector PQ becomes the distance between skew lines L and M.
We observe
−→
(1) The vector PQ is perpendicular to both lines L and M, i.e., perpendicular to both vectors vL = ⟨aL , bL , cL ⟩
−→
and vM = ⟨aM , bM , cM ⟩. It implies that PQ is parallel to the cross product vL × vM , i.e.,
−→
PQ = k (vL × vM ) ,

where k is a scalar.
(2) If we introduce a unit vector, we observe
−→ vL × vM vL × vM
PQ = ± (distance d between P and Q) = ±d . (15.2.3)
∥vL × vM ∥ ∥vL × vM ∥

Suppose P(x0 , y0 , z0 ) and Q(x1 , y1 , z1 ) are obtained when t = t0 and s = s0 , i.e.,

x0 = xL + aLt0 , y0 = yL + bLt0 , z0 = zL + cLt0 ,

x1 = xM + aM s0 , y1 = yM + bM s0 , z1 = zM + cM s0 .
−→
Then, the vector PQ is obtained by
−→
PQ = ⟨x1 − x0 , y1 − y0 , z1 − z0 ⟩
= ⟨xM + aM s0 − xL − aLt0 , yM + bM s0 − yL − bLt0 , zM + cM s0 − zL − cLt0 ⟩
= ⟨xM − xL , yM − yL , zM − zL ⟩ + s0 ⟨aM , bM , cM ⟩ − t0 ⟨aL , bL , zL ⟩
−−−→
= PL QM + s0 vM − t0 vL .

Page 121 of 126

Calculus II for Engineers Fall, 2010

Combining with the result (15.2.3), we deduce

vL × vM −→ −−−→ vL × vM −−−→
±d = PQ = PL QM + s0 vM − t0 vL −→ ±d = PL QM + s0 vM − t0 vL .
∥vL × vM ∥ ∥vL × vM ∥
Applying the dot product with vL × vM on both sides, we get
vL × vM −−−→
±d · (vL × vM ) = PL QM · (vL × vM ) + s0 vM · (vL × vM ) − t0 vL · (vL × vM )
∥vL × vM ∥
∥vL × vM ∥2 −−−→ −−−→
±d = PL QM · (vL × vM ) , ±d∥vL × vM ∥ = PL QM · (vL × vM ) .
∥vL × vM ∥
Therefore, we deduce the formula on the distance d between skew lines L and M:
−−−→
|PL QM · (vL × vM ) |
d=
∥vL × vM ∥
Theorem 15.2.6. The skew lines L and M given by
L : x = xL + aLt, y = yL + bLt, z = zL + cLt, and M : x = xM + aM s, y = yM + bM s, z = zM + cM s
has the distance −−−→
|PL QM · (vL × vM ) |
d=
∥vL × vM ∥
where vL = ⟨aL , bL , cL ⟩, vM = ⟨aM , bM , cM ⟩ and PL (xL , yL , zL ) and QM (xM , yM , zM ).
Example 15.2.7. Find the distance between skew lines L and M:
L : x = 3t, y = 9 − t, z = 2 + t,
M : x = −6 − 3s, y = −5 + 2s, z = 10 + 4s.
A NSWER. The lines L and M are skew. (Check this out by yourself!)
We observe that the lines L and M are parallel to the vectors vL = ⟨3, −1, 1⟩ and vM = ⟨−3, 2, 4⟩, respectively.
The lines L and M also pass through the points PL (0, 9, 2) and QM (−6, −5, 10), respectively, and so we have
−−−→
the vector PL QM = ⟨−6, −14, 8⟩.
Now, the Theorem 15.2.6 implies
−−−→
|PL QM · (vL × vM ) | | ⟨−6, −14, 8⟩ · (⟨3, −1, 1⟩ × ⟨−3, 2, 4⟩) |
d= =
∥vL × vM ∥ ∥ ⟨3, −1, 1⟩ × ⟨−3, 2, 4⟩ ∥
  
  
 −6 −14 8  
1     270 √
=  3 −1 1   = √ = 3 30 . □
∥ ⟨−6, −15, 3⟩ ∥  


 3 30
 −3 2 4  
determinant absolute

□ 15.2.5 Plane Π : ax + by + cz = d and Point P(x0 , y0 , z0 ) not lying on Π.

.P(x0 , y0 , z0 )

.distance D .Plane Π : ax + by + cz = d

.90◦
.Q(x1 , y1 , z1 )

.
Page 122 of 126
Calculus II for Engineers Fall, 2010

−→
The vector PQ is parallel to the normal vector n = ⟨a, b, c⟩ of the plane Π. So there exists a unique1 scalar s0
such that
−→
PQ = s0 n −→ x1 = x0 + as0 , y1 = y0 + bs0 , z1 = z0 + cs0 .

Since Q(x1 , y1 , z1 ) is on the plane Π, we have

d = ax1 + by1 + cz1 = a(x0 + as0 ) + b(y0 + bs0 ) + c(z0 + cs0 )

( )
= ax0 + by0 + cz0 + s0 a2 + b2 + c2 = ax0 + by0 + cz0 + s0 ∥n∥2
d − (ax0 + by0 + cz0 )
s0 = .
∥n∥2
−→
The distance D between the plane Π and the point P(x0 , y0 , z0 ) is the magnitude of the vector PQ and so we
have  
 d − (ax0 + by0 + cz0 ) 
−→
D = ∥PQ∥ = ∥s0 n∥ = |s0 |∥n∥ =   ∥n∥ = |d − (ax0 + by0 + cz0 ) | .
∥n∥ 2  ∥n∥

Theorem 15.2.8. The distance D between the plane Π : ax + by + cz = d and the point P(x0 , y0 , z0 ) not lying
on Π is obtained by
|ax0 + by0 + cz0 − d| |ax0 + by0 + cz0 − d|
D= = √ .
∥n∥ a2 + b2 + c2

Example 15.2.9. Find the distance between the plane Π : 2x − 3y + z = 6 and the point P(1, 0, 0) not lying
on Π.

A NSWER. The plane Π has the normal vector n = ⟨2, −3, 1⟩ and the Theorem 15.2.8 implies

|ax0 + by0 + cz0 − d| |2(1) − 3(0) + 1(0) − 6| 4

D= = √ =√ . □
∥n∥ 14 14

□ 15.2.6 Plane Π : ax + by + cz = d and Line L : x = x0 + a′t, y = y0 + b′t, z = z0 + c′t not intersecting with
Π.
We choose the point P(x0 , y0 , z0 ) from the line L and use the Theorem 15.2.8.

Theorem 15.2.10. The distance D between the plane Π : ax + by + cz = d and line L : x = x0 + a′t, y =
y0 + b′t, z = z0 + c′t not intersecting with Π is obtained by

|ax0 + by0 + cz0 − d| |ax0 + by0 + cz0 − d|

D= = √ .
∥n∥ a2 + b2 + c2

Example 15.2.11. Find the distance between the plane Π : 2x − 3y − 3z = 6 and the line L : x = 3t, y =
9 + t, z = 2 + t not intersecting with Π.

A NSWER. The plane Π has the normal vector n = ⟨2, −3, −3⟩ and the line L passes through the point
P(0, 9, 2). The Theorem 15.2.10 implies
√
|ax0 + by0 + cz0 − d| |2(0) − 3(9) − 3(2) − 6| 39
D= = √ =√ . □
∥n∥ 22 22

1 radically distinctive and without equal; the single one of its kind

Page 123 of 126

Calculus II for Engineers Fall, 2010

□ 15.2.7 Parallel Planes Π0 : ax + by + cz = d and Π1 : a′ x + b′ y + c′ z = e.

.Plane Π0 : ax + by + cz = d

. (x0 , y0 , z0 )
P
◦
.distance D .90
.Plane Π1 : a′ x + b′ y + c′ z = e

9. 0◦
.Q(x1 , y1 , z1 )

.
We fix one of the planes and choose a point from the other plane. Note that there exits a particular scalar k
such that n′ = ⟨a′ , b′ , c′ ⟩ = k ⟨a, b, c⟩ = kn by the condition that Π0 and Π1 are parallel. It implies
e
Π1 : a′ x + b′ y + c′ z = e ⇐⇒ Π1 : kax + kby + kcz = e ⇐⇒ Π1 : ax + by + cz = .
k
(1) Let us fix the plane Π0 having the normal vector n = ⟨a, b, c⟩ and choose a point Q(x1 , y1 , z1 ) from the
plane Π1 . Then the Theorem 15.2.8 implies
|ax1 + by1 + cz1 − d| |e/k − d| |e/k − d|
D= = =√ .
∥n∥ ∥n∥ a2 + b2 + c2
(2) Let us fix the plane Π1 having the normal vector n′ = ⟨a′ , b′ , c′ ⟩ = kn and choose a point P(x0 , y0 , z0 ) from
the plane Π0 . Then the Theorem 15.2.8 implies
|a′ x0 + b′ y0 + c′ z0 − e| |kax0 + kby0 + kcz0 − e| |ax0 + by0 + cz0 − e/k| |d − e/k|
D= = = = .
∥n′ ∥ k∥n∥ ∥n∥ ∥n∥
As we can notice, two results in (1) and (2) are exactly same. Thus, we conclude
|d − e/k| |d − e/k|
D= =√ .
∥n∥ a2 + b2 + c2
Theorem 15.2.12. The distance D between two parallel planes Π0 : ax +by+cz = d and Π1 : a′ x +b′ y+c′ z =
e is obtained by
|d − e/k| |d − e/k|
D= =√ .
∥n∥ a2 + b2 + c2
If a′ = a, b′ = b and c′ = c, then k = 1. The Theorem 15.2.12 gives a corollary.
Corollary 15.2.13. The distance D between two planes having the same normal vectors, i.e., Π0 : ax + by +
cz = d and Π1 : ax + by + cz = e, is obtained by
|d − e| |d − e|
D= =√ .
∥n∥ a + b2 + c2
2

Example 15.2.14. Find the distance between parallel planes Π0 : 2x − 3y + z = 6 and Π1 : 4x − 6y + 2z = 8.

A NSWER. The planes Π0 and Π1 have parallel normal vectors n0 = ⟨2, −3, 1⟩ and n1 = ⟨4, −6, 2⟩ with
n1 = 2n0 . The Theorem 15.2.12 implies
|d − e/k| |6 − 8/2| 2
D= √ =√ =√ . □
2
a +b +c2 2 22 + (−3)2 + 12 14

Page 124 of 126

Calculus II for Engineers Fall, 2010

□ 15.2.8 Appendix: Skew Lines L and M (Version 2):

L : x = xL + aLt, y = yL + bLt, z = zL + cLt and
M : x = xM + aM s, y = yM + bM s, z = zM + cM s.
.P(x0 , y0 , z0 )
.Line L : x = xL + aLt, y = yL + bLt, z = zL + cLt
.90◦

.distance d .Line M : x = xM + aM s, y = yM + bM s, z = zM + cM s

.90◦
.Q(x1 , y1 , z1 )

.
Let us develop an argument using the formula on the distance between a plane and a point not lying on the
plane.
Let PL (xL , yL , zL ) and QM (xM , yM , zM ) be the points on L and M given in the parametric equations of L and M
and let vL = ⟨aL , bL , cL ⟩ and vM = ⟨aM , bM , cM ⟩.
Let P(x0 , y0 , z0 ) and Q(x1 , y1 , z1 ) be points on L and M, respectively, such that the line passing through both
P and Q is perpendicular to both lines L and M, i.e., perpendicular to vL and vM . Then we can find a plane Π
−→
containing the line L and having the vector PQ as its normal vector. Once we find the equation of the plane
Π, we use the Theorem 15.2.8 on the distance between the plane Π and any point on the line M.
−→ −→
Since PQ is perpendicular to both vL and vM , so PQ is parallel to the cross product vL × vM , i.e.,
−→
PQ = s (vL × vM ) ,
−→
for some scalar s. It implies that the plane Π containing the line L and having PQ as its normal vector is
expressed by the equation,
Π : (vL × vM ) · ⟨x − xL , y − yL , z − zL ⟩ = 0,
where the point PL (xL , yL , zL ) is chosen arbitrary from the line L.
When we arbitrarily choose the point QM (xM , yM , zM ) on the line M, the Theorem 15.2.8 implies the distance
between Π and the point QM ,
−−−→
| (vL × vM ) · ⟨xM − xL , yM − yL , zM − zL ⟩ | |PL QM · (vL × vM ) |
D= = ,
∥vL × vM ∥ ∥vL × vM ∥
which is exactly same as the Theorem 15.2.6.
Let us rewrite the result above. If vL × vM = ⟨a, b, c⟩, then the equation of the plane Π becomes
Π : a(x − xL ) + b(y − yL ) + c(z − zL ) = 0 −→ Π : ax + by + cz = d,
where d = axL + byL + czL . The result above becomes
−−−→
|PL QM · (vL × vM ) | |axM + byM + czM − d|
D= = √ .
∥vL × vM ∥ a2 + b2 + c2
Theorem 15.2.15. The skew lines L and M given by
L : x = xL + aLt, y = yL + bLt, z = zL + cLt, and M : x = xM + aM s, y = yM + bM s, z = zM + cM s
has the distance −−−→
|PL QM · (vL × vM ) |
d=
∥vL × vM ∥
where vL = ⟨aL , bL , cL ⟩, vM = ⟨aM , bM , cM ⟩ and PL (xL , yL , zL ) and QM (xM , yM , zM ).

Page 125 of 126

Calculus II for Engineers Fall, 2010

Personally, I do not recommend this argument on finding the distance between skew lines. Please try to use
the Theorem 15.2.6 from the Subsection 15.2.4 Distance between Skew Lines (Version 1).

Example 15.2.16. Find the distance between skew lines L and M:

L : x = 3t, y = 9 − t, z = 2 + t,
M : x = −6 − 3s, y = −5 + 2s, z = 10 + 4s.

A NSWER. The lines L and M are skew. (Check this out by yourself!)
We observe that the lines L and M are parallel to the vectors vL = ⟨3, −1, 1⟩ and vM = ⟨−3, 2, 4⟩, respectively.
The lines L and M also pass through the points PL (0, 9, 2) and QM (−6, −5, 10), respectively, and so we have
−−−→
the vector PL QM = ⟨−6, −14, 8⟩.
The plane Π containing the line L and having vL × vM = ⟨−6, −15, 3⟩ as its normal vector is obtained by

Π : −6(x − 0) − 15(y − 9) + 3(z − 2) = 0 −→ Π : 2x + 5y − z = 43.

Applying the Theorem 15.2.15 with the point QM (−6, −5, 10) on the line M, we have the distance

|axM + byM + czM − d| |2(−6) + 5(−5) − 10 − 43| 90 √

D= √ = √ = √ = 3 30 ,
a2 + b2 + c2 ∥ 22 + 52 + (−1)2 30

which is exactly same result as in the Example 15.2.7 on page 122. □

Page 126 of 126