Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 1

These notes were written to outline the major topics covered in Auburn Univer-
sity’s Calculus III course based on Stewart’s 7th Edition Calculus text. It progresses
through most of the sections in Chapters 12 through 16, but Chapter 15 is reorganized
slightly to introduce the Jacobian before introducing alternate coordinate systems.
In addition, sections 15.5 and 15.6 are omitted entirely to match Auburn’s course
syllabus.
The purpose of these notes is not to replace any calculus or analysis textbook, but
rather to be used as a guide/outline for students and instructors covering the topics
in a Calculus III course.
As such, when deemed necessary, mathematical rigor is abandoned for the sake of
simplicity or brevity. (Many theorems actually only apply to “nice” functions, usually
requiring some level of continuity or differentiability.) Since for many applications of
interest the relevant functions are “nice”, students should be able to use these notes
as a “good-enough” resource for working on computational problems, particularly the
accompanying study problems.

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 2

12.1 Three-Dimensional Coordinate Systems

• Distance between points in 3D space

p
D = (x2 − x1 )2 + (y2 − y1 )2 + (z2 − z1 )2

• Simple planes in 3D Space

x = a, y = b, z = c

• Spheres in 3D Space
(x − x0 )2 + (y − y0 )2 + (z − z0 )2 = a2

12.2 Vectors

• Definition of a Vector

– A vector v = →−
v is a mathematical object which stores length (magnitude)
and direction, and can be thought of as a directed line segment.
– Two vectors with the same length and direction are considered equal, even
if they aren’t in the same position.
– We often assume the initial point lays at the origin.

• Component Form
The vector with initial point at (0, 0, 0) and terminal point at (vx , vy , vz ) is
represented by
hvx , vy , vz i
• 2D and 3D Vectors

ha, bi = ha, b, 0i
• Position Vector
If P = (a, b, c) is a point, then P = ha, b, ci is its position vector.
We assume (a, b, c) = ha, b, ci.
• Vector Between Points
The vector from P1 = (x1 , y1 , z1 ) to P2 = (x2 , y2 , z2 ) is
−−→
P1 P2 = P1 P2 = hx2 − x1 , y2 − y1 , z2 − z1 i

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 3

• Length of a Vector
q
|v| = | hv1 , v2 , v3 i | = v12 + v22 + v32

• The Zero Vector

→
−
0 = 0 = h0, 0, 0i

• Vector Operations

– Addition

hv1 , v2 , v3 i + hu1 , u2 , u3 i = hv1 + u1 , v2 + u2 , v3 + u3 i

– Scalar Multiplication

k hv1 , v2 , v3 i = hkv1 , kv2 , kv3 i

• Vector Operation Properties

1. u + v = v + u
2. (u + v) + w = u + (v + w)
3. u + 0 = u
4. u + (−u) = 0
5. 0u = 0
6. 1u = u
7. a(bu) = (ab)u
8. a(u + v) = au + av
9. (a + b)u = au + bu

• Unit Vectors

– A unit vector or direction is any vector whose length is 1.

– Standard unit vectors
∗ i = h1, 0, 0i
∗ j = h0, 1, 0i
∗ k = h0, 0, 1i

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 4

– Standard Unit Vector Form:

hvx , vy , vz i = vx i + vy j + vz k
– Length-Direction Form:
v
v = |v|
|v|

12.3 The Dot Product

• Dot Product
u · v = hu1 , u2 , u3 i · hv1 , v2 , v3 i = u1 v1 + u2 v2 + u3 v3

• Angle between vectors

u·v
cos θ =
|u||v|
• Alternate Dot Product formula
u · v = |u||v| cos θ

• Orthogonal Vectors
– u, v are orthogonal if u · v = 0
– u, v are orthogonal if the angle between them is π
2
= 90◦
– 0 is orthogonal to every vector
• Dot Product Properties
1. u · v = v · u
2. (cu) · v = u · (cv) = c(u · v)
3. u · (v + w) = u · v + u · w
4. u · u = |u|2
5. 0 · u = 0
• Projection Vector  
u·v v
projv (u) =
|v| |v|
• Work
W = F · D = |F||D| cos θ

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 5

12.4 The Cross Product

• Determinants

– 2x2 Determinant

a b
= ad − bc
c d
– 3x3 Determinant
a1 a2 a3
b2 b3 b b b b
b1 b2 b3 = a1 − a2 1 3 + a3 1 2
c2 c3 c1 c3 c1 c2
c1 c2 c3

b2 b3 b b b b
= a1 + a2 3 1 + a3 1 2
c2 c3 c3 c1 c1 c2

= (a1 b2 c3 + a2 b3 c1 + a3 b1 c2 ) − (a3 b2 c1 + a1 b3 c2 + a2 b1 c3 )

• Cross Product
i j k  
u2 u3 u3 u1 u1 u2
u × v = u1 u2 u3 = , ,
v2 v3 v3 v1 v1 v2
v1 v2 v3

= hu2 v3 − u3 v2 , u3 v1 − u1 v3 , u1 v2 − u2 v1 i

Shortcut “long multiplication” method:

h u1 , u2 , u3 i
×h v1 , v2 , v3 i
h u2 v3 − u3 v2 , u3 v1 − u1 v3 , u1 v2 − u2 v1 i

• Right-Hand Rule

– A method for determining a special orthogonal direction used throughout

mathematics and physics in 3D space, with respect to an ordered pair of
vectors u, v
– u × v is orthogonal to both u, v according to the Right-Hand Rule.

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 6

• Cross Product Magnitude

|u × v| = |u||v| sin θ

The area of the parallelogram determined by u, v is |u × v|.

• Parallel Vectors
– u, v are parallel if u × v = 0
– u, v are parallel if the angle between them is 0 = 0◦ or π = 180◦
– 0 is parallel to every vector
• Cross Product Properties
1. (ru) × (sv) = (rs)(u × v)
2. u × (v + w) = u × v + u × w
3. (v + w) × u = v × u + w × u
4. v × u = −(u × v)
5. 0×u=0
6. u×u=0
• Standard Unit Vector Cross Products
1. i × j = k
2. j × k = i
3. k × i = j
The standard unit vectors are known as a “right handed frame”.
• Torque

→
−τ =r×F
→
−
| τ | = |r||F| sin θ
• Triple Scalar (or “Box”) Product

u1 u2 u3
(u × v) · w = v1 v2 v3
w1 w2 w3
Its absolute value |(u × v) · w| gives the volume of a parallelpiped determined
by the three vectors.

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 7

12.5 Equations of Lines and Planes

• Vector Equation and Parametric Equations for a Line

r(t) = P0 + tv

x = x0 + At, y = y0 + Bt, z = z0 + Ct
for −∞ < t < ∞
• Symmetric Equations for a Line
x − x0 y − y0 z − z0
= =
A B C

• Line Segment joining a pair of points

r(t) = P0 + t(P1 − P0 ) = (1 − t)P0 + tP1

for 0 ≤ t ≤ 1
• Distance from a Point to a Line
|PS × v|
d=
|v|

• Equation for a Plane

A(x − x0 ) + B(y − y0 ) + C(z − z0 ) = 0

Ax + By + Cz = D

• Line of Intersection of Two Planes

r(t) = P0 + t(n1 × n2 )

• Angle of Intersection of Two Planes

n1 · n2
cos θ =
|n1 ||n2 |

• Distance from a Point to a Plane

|PS · n|
d=
|n|

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 8

12.6 Cylinders and Quadratic Surfaces

• Sketching surfaces

– To sketch a 3D surface, sketch planar cross-sections

∗ z = c is parallel to xy plane
∗ y = b is parallel to xz plane
∗ x = a is parallel to yz plane

• Cylinders

– A cylinder is any surface generated by considering parallel lines passing

through a planar curve.
– A 3D surface defined by a function of only two variables results in a cylin-
der.

• Quadric Surfaces

– A quadric surface is any surface defined by a second degree equation of

x, y, z.
– Most helpful to consider the cross-sections in each of the coordinate planes.

• Ellipsoids

– Cross-sections in the coordinate planes include

∗ Three ellipses

• Elliptical Cone

– Cross-sections in the coordinate planes include

∗ Two double-lines
∗ One point (with parallel ellipses)

• Elliptical Paraboloid

– Cross-sections in the coordinate planes include

∗ Two parabolas
∗ One point (with parallel ellipses)

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 9

• Hyperbolic Paraboloid

– Cross-sections in the coordinate planes include

∗ Two parabolas (with parallel parabolas)
∗ One double line (with parallel hyperbolas)

• Hyperboloid of One Sheet

– Cross-sections in the coordinate planes include

∗ Two hyperbolas
∗ One ellipsis (with parallel ellipses)

• Hyperboloid of Two Sheets

– Cross-sections in the coordinate planes include

∗ Two hyperbola
∗ One empty cross-section (with parallel ellipses)

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 10

13.1 Vector Functions and Space Curves

• Curves, Paths, and Vector Functions

– A position function maps a moment in time to a position on a path. It

can be defined with parametric equations

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

or with a vector function

r(t) = hx(t), y(t), z(t)i

– x(t), y(t), z(t) are called component functions

• Vector Function Limits

D E
lim r(t) = lim f (t), lim g(t), lim h(t)
t→a t→a t→a t→a

• Continuity of Vector Functions

– The function r(t) is continuous if

lim r(t) = r(a)

t→a

for all a in its domain.

– r(t) is continuous exactly when f (t), g(t), h(t) are all continuous.

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 11

13.2 Derivatives and Integrals of Vector Functions

• Derivatives of Vector Functions

dr r(t + ∆t) − r(t)
= r0 (t) = lim = hf 0 (t), g 0 (t), h0 (t)i
dt ∆t→0 ∆t
– r(t) is differentiable if r0 (t) is defined for every value of t is in its domain.
– r0 (a) is a tangent vector to the curve where t = a
– The tangent line to a curve at t = a:

l(t) = r(a) + tr0 (a)

• Differentiation Rules for Vector Functions

d
[C] = 0
dt
d
[cu(t)] = cu0 (t)
dt
d
[f (t)C] = f 0 (t)C
dt
d
[u(t) ± v(t)] = u0 (t) ± v0 (t)
dt
d
[f (t)u(t)] = f (t)u0 (t) + f 0 (t)u(t)
dt
d
[u(t) · v(t)] = u(t) · v0 (t) + u0 (t) · v(t)
dt
d
[u(t) × v(t)] = u(t) × v0 (t) + u0 (t) × v(t)
dt
du d du df
= [u(f (t))] = u0 (f (t))f 0 (t) =
dt dt df dt
• Derivative of a Constant Length Vector Function

– If |r(t)| = c always, then

r(t) · r0 (t) = 0
– Thus the derivative of a constant length vector function is perpindicular
to the original.

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 12

• Antiderivatives of Vector Functions

– If R0 (t) = r(t), then R(t) is an antiderivative of r(t).

Z
– The indefinite integral r(t) dt is the collection of all the antiderivatives
of r(t). Z
r(t) dt = R(t) + C
Z Z Z Z 
r(t) dt = x(t) dt, y(t) dt, z(t) dt

• Definite Integrals
Z b Z b Z b Z b 
r(t) dt = x(t) dt, y(t) dt, z(t) dt
a a a a
Z b
r(t)dt = [R(t)]ba = R(b) − R(a)
a

• Differential Vector Equations

– If we know r0 (t) and r(a) for some t = a, then

Z t
r(t) = r0 (t) dt + r(a)
a

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 13

13.3 Arc Length and Curvature

• Arc Length along a Space Curve

Z b Z b
r(t + ∆t) − r(t)
L= lim dt = |r0 (t)| dt
a ∆t→0 ∆t a

• Arclength Parameter Z t
s(t) = |r0 (u)|du
0
ds
= |r0 (t)|
dt
• Unit Tangent Vector
dr
T(s) =
ds
dr/dt
T(t) =
|dr/dt|
• Curvature
dT
κ(s) =
ds
2
|dT/dt| | dr × d 2r |
κ(t) = = dt dr 3dt
|dr/dt| | dt |
For y = f (x):
|f 00 (x)|
κ(x) =
[1 + (f 0 (x))2 ]3/2
• Principal Unit Normal Vector
dT/ds
N(s) =
|dT/ds|

dT/dt
N(t) =
|dT/dt|
• Binormal Unit Vector
B=T×N
The triple T, N, B forms a right-handed frame.

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 14

13.4 Motion in Space: Velocity and Acceleration

• Position, Velocity, and Acceleration

– Position: r(t)
– Velocity: v(t) = r0 (t) = dr
dt
ds
– Speed: v(t) = |v(t)| = dt
v(t)
– Direction: T(t) = |v(t)|

– Acceleration: a(t) = v0 (t) = r00 (t)

• Ideal Projectile Motion

a(t) = h0, −gi
v(t) = hv0 cos α, −gt + v0 sin αi
 
1 2
r(t) = (v0 cos α)t, − gt + (v0 sin α)t
2

• Tangental and Normal Components of Acceleration

 2   2
ds ds
a= 2
T+κ N + 0B
dt dt

– Tangental component
d2 s
aT = = v0
dt2
– Normal component
 2
ds
q
aN = κ = κv 2 = |a|2 − a2T
dt

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 15

14.1 Functions of Several Variables

• Functions of Two Variables

– A function f of two variables is a rule which assigns a real number

f (x, y) to each pair of real numbers (x, y) in its domain

dom(f ) ⊆ R2

The set of values f takes on is its range

ran(f ) = {f (x, y) : (x, y) ∈ dom(f )}

– The level curve for each k ∈ ran(f ) is given by the equation

f (x, y) = k

– The graph of f is a surface in 3D space which visualizes the function,

given by the equation z = f (x, y).

• Functions of Three Variables

– A function f of three variables is a rule which assigns a real number

f (x, y, z) to each pair of real numbers (x, y, z) in its domain

dom(f ) ⊆ R3

The set of values f takes on is its range

ran(f ) = {f (x, y, z) : (x, y, z) ∈ dom(f )}

– The level surface for each k ∈ ran(f ) is given by the equation

f (x, y, z) = k

• Alternate Forms

– We may also consider functions of the form f (x1 , x2 , . . . ) = f (P ) = f (r).

– If P = (x, y) and r = hx, yi, then f (x, y) = f (P ) = f (r).
– If P = (x, y, z) and r = hx, y, zi, then f (x, y, z) = f (P ) = f (r).

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 16

14.2 Limits and Continuity

• Limits

– If the value of the function f (P ) becomes arbitrarily close to the number

L as vectors P close to P0 are plugged into the function, then the limit
of f (P ) as P approaches P0 is L:

lim f (P ) = L
P →P0

– For functions of two or three variables:

lim f (x, y) = L
(x,y)→(x0 ,y0 )

lim f (x, y, z) = L
(x,y,z)→(x0 ,y0 ,z0 )

• Showing a Limit DNE

– In order for a limit limP →P0 f (x, y) to exist, the values of f must approach
L no matter which direction we approach P0 .
– Choose y = g(x) and y = h(x) where P0 lays on both graphs. If

lim f (x, g(x)) 6= lim f (x, h(x))

x→x0 x→x0

then lim f (x, y) DNE.

P →P0

– Or choose x = g(y) and x = h(y) where P0 lays on both graphs. If

lim f (g(y), y) 6= lim f (h(y), y)

y→y0 y→y0

then lim f (x, y) DNE.

P →P0

• Limit Laws

lim (f (P ) ± g(P )) = lim f (P ) ± lim g(P )

P →P0 P →P0 P →P0

lim (f (P ) · g(P )) = lim f (P ) · lim g(P )

P →P0 P →P0 P →P0

lim (kf (P )) = k lim f (P )

P →P0 P →P0

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 17

lim f (P )
f (P ) P →P0
lim =
P →P0 g(P ) lim g(P )
P →P0
 r/s
r/s
lim (f (P )) = lim f (P )
P →P0 P →P0

• Computing Limits

– Variables not involved in a limit may be eliminated:

lim f (x) = lim f (x)

P →P0 x→x0

– Due to the Limit Laws, many limits follow the “just plug it in” rule.
– If plugging in results in a zero in a denominator, use factoring, perhaps
with conjugates.
– L’Hopital’s Rule does not apply for multiple variable limits.

• Continuity

– A function f (P ) is continuous if lim f (P ) = f (P0 ) for all points P0 in

P →P0
its domain.
– If a multi-variable function is composed of continuous single-variable func-
tions, then it is also continuous.

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 18

14.3 Partial Derivatives

• Partial Derivatives

– For a function f of two variables (x, y):

∂f f (x + h, y) − f (x, y)
= fx (x, y) = lim
∂x h→0 h
∂f f (x, y + h) − f (x, y)
= fy (x, y) = lim
∂y h→0 h
– To compute partial derivatives with respect to a variable, treat all other
variables as constants and differentiate as normal.
– Functions of more than two variables behave similarly. For T (x, y, z):

∂T T (x, y, z + h) − T (x, y, z)
= Tz (x, y, z) = lim
∂z h→0 h

• Higher Order Partial Derivatives

∂ 2f
 
∂ ∂f
= = (fy )x = fyx
∂x∂y ∂x ∂y
∂ 2g
 
∂ ∂g
= = (gz )z = gzz
∂z 2 ∂z ∂z

• Mixed Derivative Theorem

– For many naturally occuring functions:

fxy = fyx

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 19

14.4 Tangent Planes and Linear Approximations

• Tangent Plane to z = f (x, y) at (a, b, f (a, b))

z − f (a, b) = fx (a, b)(x − a) + fy (a, b)(y − b)

• Linearization of f (x, y) at (a, b)

L(x, y) = f (a, b) + fx (a, b)(x − a) + fy (a, b)(y − b)

• Differentiability and a Sufficient Condition

– A multi-variable function f is differentiable at a point if its linearizara-

tion approximates the value of the function near that point.
– If fx , fy exist near (a, b) and are continuous at (a, b), then f is differentiable
at (a, b).

• Linear Approximation
If f is differentiable at (a, b), then

f (x, y) ≈ f (a, b) + fx (a, b)(x − a) + fy (a, b)(y − b)

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 20

14.5 The Chain Rule

• Gradient Vector Function

 
∂f ∂f
∇f (x, y) = hfx (x, y), fy (x, y)i = ,
∂x ∂y
 
∂f ∂f ∂f
∇f (x, y, z) = hfx (x, y, z), fy (x, y, z), fz (x, y, z)i = , ,
∂x ∂y ∂z

• Nested Functions

– If f is a function of r = hx, y, zi, and x, y, z are functions of t, then we say

x, y, z are itermediate variables and may consider the following composed
function of t:
f (r(t)) = f (x(t), y(t), z(t))
– If f is a function of r = hx, y, zi, and x, y, z are functions of s = ht, u, vi,
then we say x, y, z are itermediate variables and may consider the following
composed function of t, u, v:

f (r(s)) = f (x(t, u, v), y(t, u, v), z(t, u, v))

• Chain Rule

– For functions of the form f (r(t)) = f (x(t), y(t), z(t)):

df dr ∂f dx ∂f dy ∂f dz
= ∇f · = + +
dt dt ∂x dt ∂y dt ∂z dt

– For functions of the form f (r(s)) = f (x(t, u, v), y(t, u, v), z(t, u, v)):

∂f ∂r ∂f ∂x ∂f ∂y ∂f ∂z
= ∇f · = + +
∂t ∂t ∂x ∂t ∂y ∂t ∂z ∂t

• Differentiation by Substitution

– The multi-variable Chain Rule can be avoided by “plugging in” functions

and using single-variable calculus.

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 21

• Total Derivative

– If f is a function of x, y, z, and y, z are also functions of x, then

df dr ∂f ∂f dy ∂f dz
= ∇f · = + +
dx dx ∂x ∂y dx ∂z dx

• Implicit Differentiation

– If f (x, y) = c defines y as a function of x, then

dy ∂f /∂x fx
=− =−
dx ∂f /∂y fy

• Tree Diagram for the Chain Rule

– The tree diagram for the chain rule can be used to generate the chain rule.
– It also holds for multiple levels of intermediate variables.

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 22

14.6 Directional Derivatives and the Gradient Vector

• Directional Derivative

– The directional derivative of f for the unit vector u is

Du f = ∇f · u

– The maximum value of Du f at a fixed point P0 is |∇f (P0 )|, which occurs
∇f (P0 )
when u = |∇f (P0 )|
.

• Normal Vector to Level Curves and Surfaces

– The gradient vectors ∇f are normal vectors to the level curves f (x, y) = k
for every (x, y) in the domain of f .
– The gradient vectors ∇f are normal vectors to the level surfaces f (x, y, z) =
k for every (x, y, z) in the domain of f .

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 23

14.7 Maximum and Minimum Values

• Local Maximum and Minimum Values

– Let f be a function of many variables defined near the point P0 .

∗ f has a local maximum f (P0 ) at P0 if f (P0 ) is the largest value of
f near P0
∗ f has a local minimum f (P0 ) at P0 if f (P0 ) is the smallest value of
f near P0

• Critical Points
– If P0 is a point in the domain of f and
∇f (P0 ) = 0 or ∇f (P0 ) DNE
then P0 is called a critical point.
– Critical points occur when the tangent plane is horizontal or DNE.
– The local maximum and minimum values of a function always occur at
critical points.
• Saddle Points

– Not every critical point gives a local extreme value.

– The saddle points of f are the critical points which don’t yield local
extreme values.

• Discriminant Function
– The discriminant of f with variables x, y is the function
fxx fxy 2
fD = = fxx fyy − fxy
fyx fyy

• Second Derivative Test for Local Extreme Values of f (x, y)

Let (a, b) be a critical point of of f where ∇f is defined.
– If fD (a, b) > 0 and fxx (a, b) < 0, then f (a, b) is a local maximum.
– If fD (a, b) > 0 and fxx (a, b) > 0, then f (a, b) is a local minimum.
– If fD (a, b) < 0, then f has a saddle point at (a, b).
– If fD (a, b) = 0, then the test is inconclusive.

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 24

• Absolute Maximum and Minimum Values

– Let f be a function of many variables.

∗ f has an absolute maximum f (P0 ) at P0 if f (P0 ) is the largest value
in the range of f
∗ f has an absolute minimum f (P0 ) at P0 if f (P0 ) is the smallest
value in the range of f
– Every continuous function of many variables with a closed and bounded
domain has an absolute maximum and minimum value.

• Finding Absolute Max/Min of f (x, y) on a Closed and Bounded Region D

– The following points are candidates for giving the absolute extrema:
∗ Critical points of f within D.
∗ Critical points for a function which gives part of the boundary of D.
∗ Corners of D.
– Plug each of these into f (x, y). The largest of these is the absolute maxi-
mum, and the smallest of these is the absolute minimum.

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 25

14.8 Lagrange Multipliers

• The Method of Lagrange Multipliers

– The Method of Lagrange Multipliers says that if f is a function of

many variables which has an absolute max/min value on the restriction
g(P ) = k where ∇g 6= 0, then the absolute max/min occurs at a point P
where
∇f (P ) = λ∇g(P ) and g(P ) = k
for some real number λ.
– If two constraints g(P ) = k and h(P ) = l are given, then the absolute
max/min occurs where

∇f (P ) = λ∇g(P ) + µ∇h(P ) and g(P ) = k and h(P ) = l

for some real numbers λ, µ.

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 26

15.1 Double Integrals over Rectangles

• Double Integral

– We define the double integral of a function f (x, y) over a region R to be

x n
X
f (x, y) dA = lim f (xn,i , yn,i )∆An,i
n→∞
R i=1

where for each positive integer n we’ve defined a way to partition R into
n pieces
∆Rn,1 , ∆Rn,2 , . . . , ∆Rn,n
where ∆Rn,i has area ∆An,i , contains the point (xn,i , yn,i ), and

lim max(∆An,i ) = 0
n→∞

– Since for f (x, y) ≥ 0,

n
X
f (xn,i , yn,i )∆An,i
i=1

is an approximation of the volume under z = f (x, y) and over R, the

double integral is used to define the precise volume.
– If f is not always positive, then the double integral represents net volume:
volume above the xy-plane minus volume below the xy-plane.

• Midpoint Rule for Approximating Rectangular Double Integrals

– For the rectangle

R : a ≤ x ≤ b, c ≤ y ≤ d
we may approximate the double integral by partitioning the rectangle into
a grid of m × n rectangular pieces all with area ∆A and evaluating:
x m X
X n
f (x, y) dA ≈ f (xi , yj )∆A
R i=1 j=1

where (xi , yj ) is the midpoint of the i × j rectangle.

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 27

15.2 Iterated Integrals

• Volume as Integral of Area

– If A(x) is the area of a solid’s cross-section, then the solid’s volume is

Z b
V = A(x) dx
a

• Iterated Integrals over Rectangles

– A double integral over a rectangle

R : a ≤ x ≤ b, c ≤ y ≤ d

can be expressed as the iterated integrals:

x Z bZ d Z d Z b
f (x, y) dA = f (x, y) dy dx = f (x, y) dx dy
R a c c a

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 28

15.3 Double Integrals over General Regions

• Double Integrals over Nonrectangular Regions

– For Type I regions which may be expressed as

R : a ≤ x ≤ b, g1 (x) ≤ y ≤ g2 (x)

a double integral over R may be expressed as the iterated integral:

x Z b Z g2 (x)
f (x, y) dA = f (x, y) dy dx
R a g1 (x)

– For Type II regions which may be expressed as

R : h1 (y) ≤ x ≤ h2 (y), a ≤ y ≤ b

a double integral over R may be expressed as the iterated integral:

x Z b Z h2 (y)
f (x, y) dA = f (x, y) dx dy
R a h1 (y)

• Finding Limits of Integration

1. Sketch the region and label bounding curves
2. Determine if the region is Type I or Type II by identifying (I) bottom/top
curves y = g1 (x), y = g2 (x) or (II) left/right curves x = h1 (y), x = h2 (y).
For Type I:
3. Use the leftmost and rightmost x-values in the region a, b to complete the
iterated integral:
x Z b Z g2 (x)
f (x, y) dA = f (x, y) dy dx
R a g1 (x)

For Type II:

3. Use the bottommost and topmost y-values in the region c, d to complete
the iterated integral:
x Z d Z h2 (y)
f (x, y) dA = f (x, y) dx dy
R c h1 (y)

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 29

• Swapping Variables of Integration

– You can only swap the order of integration of an iterated integral by draw-
ing the region and reinterpreting it as a region of the opposite Type.

• Additivity
If R can be split into two regions R1 , R2 , then
x x x
f (x, y) dA = f (x, y) dA + f (x, y) dA
R R1 R2

• Average Value of Two-Variable Functions

– The average value of a two-variable function f over a region R is defined

to be x
1
f (x, y) dA
Area of R
R

• Area as a Double Integral

– The area of a region R in the plane is

x x
A= dA = 1 dA
R R

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 30

15.7 Triple Integrals

• Triple Integral

– We define the triple integral of a function f (x, y, z) over a solid D to be

y n
X
f (x, y, z) dV = lim f (xn,i , yn,i , zn,i )∆Vn,i
n→∞
D i=1

where for each positive integer n we’ve defined a way to partition D into
n pieces
∆Dn,1 , ∆Dn,2 , . . . , ∆Dn,n
where ∆Dn,i has volume ∆Vn,i , contains the point (xn,i , yn,i , zn,i ), and

lim max(∆Vn,i ) = 0
n→∞

• Iterated Integral for Rectangular Boxes

– The triple integral over the rectangular box

D : a1 ≤ x ≤ a2 , b1 ≤ y ≤ b2 , c1 ≤ z ≤ c2

can be expressed as the iterated integrals:

y Z a2 Z b2 Z c2
f (x, y, z) dV = f (x, y, z) dz dy dx
D a1 b1 c1

Z b2 Z c2 Z a2 Z a2 Z c2 Z b2
= f (x, y, z) dx dz dy = f (x, y, z) dy dz dx = . . .
b1 c1 a1 a1 c1 b1

• Iterated Integral for Generated Solids

– If the solid D is determined by the bottom/top surfaces

h1 (x, y) ≤ z ≤ h2 (x, y)

and has shadow R in the xy-plane, then a triple integral over D can be
expressed as:
y x Z h2 (x,y)
" #
f (x, y, z) dV = f (x, y, z) dz dA
D R h1 (x,y)

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 31

– In general:
y x Z top surface 
f (x, y, z) dV = f (x, y, z) d dA
D R bottom surface

where  is chosen from x, y, z to be the “up” orientation.

• Additivity
If D can be split into two regions D1 , D2 , then
x x x
f (x, y, z) dV = f (x, y, z) dV + f (x, y, z) dV
D D1 D2

• Average Value of Three-Variable Functions

– The average value of a three-variable function f over a solid D is defined

to be y
1
f (x, y, z) dV
Volume of D
R

• Volume as a Triple Integral

– The volume of a solid D in space is

y y
V = dV = 1 dV
D D

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 32

15.10 Change of Variables in Multiple Integrals

• Transformations

– Two similar regions in 2D space can be transformed by a “nice” pair of

functions

r(u, v) = r(s) = hx(s), y(s)i = hx(u, v), y(u, v)i

that map points in a uv plane to the xy plane.

– Two similar solids in 3D space can be transformed by a “nice” triple of
functions

r(u, v, w) = r(s) = hx(s), y(s), z(s)i = hx(u, v, w), y(u, v, w), z(u, v, w)i

that map points in a uvw space to the xyz space.

• The Jacobian

– The Jacobian of a 2D transformation given by r(u, v) is the determinant

∂x ∂x
∂(x, y) ∂r ∂u ∂v
rJ (u, v) = = = ∂y ∂y
∂(u, v) ∂s ∂u ∂v

– The Jacobian of a 3D transformation given by r(u, v, w) is the determinant

∂x ∂x ∂x
∂(x, y, z) ∂r ∂u
∂y
∂v
∂y
∂w
∂y
rJ (u, v, w) = = = ∂u ∂v ∂w
∂(u, v, w) ∂s ∂z ∂z ∂z
∂u ∂v ∂w

• 2D Substitution

– Suppose that the region R in the xy-plane is the result of applying the
transformation r(u, v) to the region G in the uv-plane.
– Then it follows that
x x
f (x, y) dx dy = f (x(u, v), y(u, v))|rJ (u, v)| du dv
R G

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 33

• Unit Square and Triangle

– The unit square in the uv plane with vertices (0, 0), (1, 0), (1, 1), and (0, 1)
is useful for substitution problems involving parallelograms.
– The unit triangle in the uv plane with vertices (0, 0), (1, 0), and (1, 1) is
useful for substitution problems involving triangles.

• 3D Substitution

– Suppose that the solid D in xyz space is the result of applying the trans-
formation r(u, v, w) to the region H in uvw space.
– Then it follows that y
f (x, y, z) dx dy dz
D
y
= f (x(u, v, w), y(u, v, w), z(u, v, w))|rJ (u, v, w)| du dv dw
H

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 34

15.4 Double Integrals in Polar Coordinates

• Integrating over Regions expressed using Polar Coordinates

– The polar coordinate transformation

r(r, θ) = hr cos θ, r sin θi

from polar G into Cartesian R yields

x x
f (x, y) dA = f (r cos θ, r sin θ) r dr dθ
R G

15.8 Triple Integrals in Cylindrical Coordinates

• Cylindrical Coordinates

– The cylindrical coordinate transformation

r(r, θ, z) = hr cos θ, r sin θ, zi

from cylindrical H into Cartesian D yields

y y
f (x, y, z) dV = f (r cos θ, r sin θ, z) r dr dθ dz
D H

15.9 Triple Integrals in Spherical Coordinates

• Spherical Coordinates

– The spherical coordinate transformation

r(ρ, φ, θ) = hρ sin φ cos θ, ρ sin φ sin θ, ρ cos φi

from spherical H into Cartesian D yields

y y
f (x, y, z) dV = f (ρ sin φ cos θ, ρ sin φ sin θ, ρ cos φ) ρ2 sin φ dρ dφ dθ
D H

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 35

16.1 Vector Fields

• Vector Fields

– A vector field assigns a vector to each point in 2D or 3D space.

F = F(r) = F(x, y) = hP (x, y), Q(x, y)i = hP (r), Q(r)i = hP, Qi

F = F(r) = F(x, y, z) = hP (x, y, z), Q(x, y, z), R(x, y, z)i = hP (r), Q(r), R(r)i = hP, Q, Ri

• Gradient Vector Field

– The gradient vector field ∇f (x, y) = hfx (x, y), fy (x, y)i assigns vectors
whose directions are normal to level curves and whose magnitudes are
equal to the maximal directional derivative at the point.
– The gradient vector field ∇f (x, y, z) = hfx (x, y, z), fy (x, y, z), fz (x, y, z)i
assigns vectors whose directions are normal to level surfaces and whose
magnitudes are equal to the maximal directional derivative at the point.

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 36

16.2 Line Integrals

• Common Curve Parametrizations

– A line segment beginning at P0 and ending at P1

r(t) = P0 + t(P1 − P0 ), 0 ≤ t ≤ 1

– A circle centered at the origin with radius a

r(t) = ha cos t, a sin ti , 0 ≤ t ≤ 2π (counter-clockwise)

r(t) = ha sin t, a cos ti , 0 ≤ t ≤ 2π (clockwise)

– A planar curve given by y = f (x) from (x0 , y0 ) to (x1 , y1 )

r(t) = ht, f (t)i , x0 ≤ t ≤ x1 (for x0 ≤ x1 )

r(t) = h−t, f (−t)i , −x0 ≤ t ≤ −x1 (for x0 ≤ x1 )

• Line Integrals with Respect to Arclength

– We define the line integral with respect to arclength of a function of

many variables f (r) along a curve C to be
Z n
X
f (r) ds = lim f (rn,i )∆sn,i
C n→∞
i=1

where for each positive integer n we’ve defined a way to partition C into
n pieces
∆Cn,1 , ∆Cn,2 , . . . , ∆Cn,n
where ∆Cn,i has length ∆sn,i , contains the position vector rn,i , and

lim max(∆sn,i ) = 0
n→∞

– If r(t) is a parametrization of C for a ≤ t ≤ b, then

Z Z t=b
ds
f (r) ds = f (r(t)) dt
t=a dt
C

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 37

• Line Integrals with Respect to Variables

– Similarly, we can find the line integral with respect to a variable for
a function of many variables f (r) along a curve C:
Z Z t=b
dx
f (r) dx = f (r(t)) dt
t=a dt
C

– Similar defintions hold for y, z.

• Line Integrals of Vector Fields

– The line integral of a vector field is defined to be the line integral

with respect to arclength of the dot product of the vector field F(r) =
hP (r), Q(r), R(r)i with the unit tangent vector T(r) to the curve.
Z
F(x, y, z) · T(x, y, z) ds
C

– There are several ways to write and evaluate line integrals of vector fields:
Z Z Z
F · T ds = F · dr = hP, Q, Ri · h dx, dy, dzi
C C C
Z Z b 
dx dy dz
= P dx + Q dy + R dz = P (r(t)) + Q(r(t)) + R(r(t)) dt
C a dt dt dt
Z b
dr
= F(r(t)) · dt
a dt
• Additivity
Let C1 + C2 represent the curve taken by moving along C1 followed by moving
along C2 .
Z Z Z
f (r) ds = f (r) ds + f (r) ds
C1 +C2 C1 C2
Z Z Z
f (r) dx = f (r) dx + f (r) dx
C1 +C2 C1 C2
Z Z Z
F · dr = F · dr + F · dr
C1 +C2 C1 C2

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 38

• Effects of Curve Orientation

Let −C represent the curve taken by moving along C in the opposite direction.
Z Z
f (r) ds = + f (r) ds
C −C
Z Z
f (r) dx = − f (r) dx
C −C
Z Z
F · dr = − F · dr
C −C

• Work

– If F is a vector field representing the force applied to an object as it is

moved over a smooth curve C, then the work done by the force over that
curve is given by Z Z b
dr
F · dr = F(r(t)) · dt
C a dt

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 39

16.3 The Fundamental Theorem for Line Integrals

• The Fundamental Theorem

– If C is any smooth curve beginning at the point A and ending at the point
B, then Z
∇f · dr = [f ]B
A = f (B) − f (A)
C

– If C is any smooth curve which is closed (begins and ends at the same
point), then Z
∇f · dr = 0
C

• Conservative Fields

– We say F = hM, N, P i is a conservative field if there is a potential func-

tion f such that ∇f = F.
– Line integrals of conservative fields are said to be path independent since
for any curve C beginning at A and ending at B:
Z Z
F · dr = ∇f · dr = [f ]BA = f (B) − f (A)
C C

– We can prove a field is conservative by finding its potential function or

showing it satisfies the Component Test:
∂P ∂Q ∂Q ∂R ∂R ∂P
= , = , =
∂y ∂x ∂z ∂y ∂x ∂z

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 40

16.4 Green’s Theorem

• Green’s Theorem

– Let C be the boundary of the region R oriented counter-clockwise, and

F(x, y) be a two-dimensional vector field.
Z x  ∂Q ∂P 
F · dr = − dA
C ∂x ∂y
R

– Due to Green’s Theorem, we can find the area of R using a line integral:
x Z
A= 1 − 0 dA = x dy
R C

x Z
A= 0 − (−1) dA = −y dx
R C

x 1  1 Z
1 1
A= − − dA = x dy − y dx
2 2 C 2 2
R

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 41

16.5 Curl and Divergence

• Gradient Operator  
∂ ∂ ∂
∇= , ,
∂x ∂y ∂z

• Curl

– The curl of a vector field is another vector field:

 
∂R ∂Q ∂P ∂R ∂Q ∂P
curl F = ∇ × F = − , − , −
∂y ∂z ∂z ∂x ∂x ∂y

– By the Component Test, if F is conservative, then curl F = 0.

• Divergence

– The divergence of a vector field is the scalar function:

∂P ∂Q ∂R
div F = ∇ · F = + +
∂x ∂y ∂z

– For any vector field, the divergence of curl is always zero.

div curl F = 0

• Green’s Theorem Alternate Forms

– If F is a two-dimensional vector field, and n is the outward unit normal

vector field for a counter-clockwise closed curve C:
Z x
F · T ds = (curl F) · k dA
C D
Z x
F · n ds = div F dA
C D

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 42

16.6 Parametric Surfaces and Their Areas

• Parametric Surface Equations

r(u, v) = hx(u, v), y(u, v), z(u, v)i

• Common Parametric Surfaces

– The plane determined by the point P0 and vectors v1 and v2 can be

parametrized by
r = P0 + uv1 + vv2
– The surface z = f (x, y) can be parametrized by

r = hx, y, f (x, y)i

– A surface determined by a cylindrical coordinate equation can be parametrized

by substituting into
r = hr cos θ, r sin θ, zi
– A surface determined by a spherical coordinate equation can be parametrized
by substituting into

r = hρ sin φ cos θ, ρ sin φ sin θ, ρ cos φi

• Surface Area

– If G is the region in the uv plane which maps onto the surface S by the
parametric equations r(u, v), then the surface area of S is:
x
|ru × rv | dA
G

where ru = hxu , yu , zu i and rv = hxv , yv , zv i.

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 43

16.7 Surface Integrals

• Surface Integral

– If G is the region in the uv plane which maps onto the surface S by the
parametric equations r(u, v), then the surface integral of f (r) along S is:
x x
f (r) dσ = f (r(u, v))|ru × rv | dA
S G

• Surface Orientation

– The orientation of a surface is determined by a continuous unit normal

vector field n on the surface.
– The Möbius strip is an example of a non-orientable surface.

• Surface Integral of Vector Field

– If G is the region in the uv plane which maps onto the surface S by the
parametric equations r(u, v), and n is the unit normal vector field giving
the orientation of S, then the surface integral of the vector field F along
S is: x x x
F · d~σ = F · n dσ = F · (ru × rv ) dA
S S G

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013

 Stewart’s Calculus Chapter 12-16 — Lecture Notes Page 44

16.8 Stokes’ Theorem

• Stokes’ Theorem

– Let C give the counter-clockwise oriented boundary of a surface S.

x Z
curl F · d~σ = F · dr
S C

16.9 Divergence Theorem

• Divergence Theorem

– Let S give the outward-oriented boundary surface of the solid D.

x y
F · d~σ = div F dV
S D

http://github.com/StevenClontz/Stewart-12to16 Last updated on August 29, 2013