Multivariable Calculus

Hechen Hu

November 25, 2017

ii
Contents

1 Vectors in Rn Space 1
1.1 Definition and Properties . . . . . . . . . . . . . . . . . . . . 1
1.2 Addition/Subtraction . . . . . . . . . . . . . . . . . . . . . . 1
1.3 Scalar Multiplication . . . . . . . . . . . . . . . . . . . . . . . 1
1.4 Dot Product . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.5 Direction Angles\Cosines . . . . . . . . . . . . . . . . . . . . 2
1.6 Orthogonal Projections . . . . . . . . . . . . . . . . . . . . . . 2
1.7 Cross Product . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.8 Scalar triple product . . . . . . . . . . . . . . . . . . . . . . . 3

2 Lines and Planes 5

2.1 Equations of Lines . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 Equations of Planes . . . . . . . . . . . . . . . . . . . . . . . 5
2.3 Angle between Planes . . . . . . . . . . . . . . . . . . . . . . 5
2.4 Distance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.5 Cylindrical and Spherical Coordinates . . . . . . . . . . . . . 6

3 Quadric Surfaces 9
3.1 Traces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.2 Type of Quadric Surfaces . . . . . . . . . . . . . . . . . . . . 9

4 Calculus of Vector-Valued Functions 11

4.1 Orientation/Direction of its graph . . . . . . . . . . . . . . . 11
4.2 Domain and Natural Domain . . . . . . . . . . . . . . . . . . 11
4.3 Radius Vector/Position Vector . . . . . . . . . . . . . . . . . 11
4.4 Vector Form of A Line Segment . . . . . . . . . . . . . . . . . 11
4.5 Calculus of Vector-Valued Functions . . . . . . . . . . . . . . 12
4.6 Arc Length . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
4.7 Arc Length as A Parameter . . . . . . . . . . . . . . . . . . . 15
4.8 Unit Tangent, Normal, and Binormal Vectors . . . . . . . . . 16
4.9 Curvature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.10 Motion Along A Curve . . . . . . . . . . . . . . . . . . . . . . 18
4.11 Projectile Motion . . . . . . . . . . . . . . . . . . . . . . . . . 19

iii
iv CONTENTS

5 Multivariable Calculus 21
5.1 Definition of Multivariable functions and their graph . . . . . 21
5.2 Properties of Sets in 2-d and 3-d space . . . . . . . . . . . . . 21
1

Vectors in Rn Space

1.1 Definition and Properties

Vector is a geometrical object that has both magnitude and direction. Ex-
amples include force and velocity.
Properties in in nth dimension vector space for Euclidean vector where ei is
the basis vector for the ith axis(for convenience we will us i, j, k denote the
basis vector in a 3-d space):

1.2 Addition/Subtraction

n
X
a±b= (ai ± bi )ei (1.1)
i=1

1.3 Scalar Multiplication

n
X
ka = kai .e.i (1.2)
i=1

1.4 Dot Product

n
X
a·b= ai ∗ bi
i=1
= kak kbk cos θ (1.3)

1
2 1. VECTORS IN RN SPACE

(Result is a scalar) Dot product has the following properties:

(a) a·b=b·a (Commutative)
(b) a · (b + c) = a · b + a · c (Distributive)
(c) k(a · b) = (ka) · b = a · (kb) (Associative)
2
(d) a · a = kak
(e) 0·a =0

1.5 Direction Angles\Cosines

The directional angles α, β and γ between the vector v and basis vectors i,
j, k in a 3-d space satisfy the following equations:
v·i v1
cos α = = (1.4)
kvk kik kvk
v·j v2
cos β = = (1.5)
kvk kjk kvk
v·k v3
cos γ = = (1.6)
kvk kkk kvk
According to their definitions we can also see that:

v1 2 + v1 2 + v1 2
cos2 α + cos2 β + cos2 γ = √ 2
v1 2 + v1 2 + v1 2
=1 (1.7)

1.6 Orthogonal Projections

The orthogonal projection of v on an arbitrary non-zero vector b can be
written as:
v·b
projb v = b (1.8)
kbk2
Moreover, we can see that v−projb v is the vector component of v orthogonal
to b.

1.7 Cross Product

 
i j k
a × b =det a1 a2 a3 
 
b1 b2 b3
= (a2 b3 − a3 b2 )i − (a1 b3 − a3 b1 )j + (a1 b2 − a2 b1 )k
= kak kbk sin θn (1.9)
1.8. SCALAR TRIPLE PRODUCT 3

(n is the vector that perpendicular to both a and b and its direction is

decided by the right hand rule in a right-handed coordinate system.)
(Result is a vector that is orthogonal to both a and b)

At the same time, we can see that

i×j=k j × i = −k (1.10)
j×k=i k × j = −i (1.11)
k×i=j i × k = −j (1.12)

What’s more, the area A of the parallelogram that has a and b as adjacent
sides is:
A = ka × bk (1.13)
Thus, a × b = 0 if a and b are parallel vectors. More useful properties of
cross product:

(a) a × b = −b × a (Anti − Commutative)

(b) a × (b + c) = a × b + a × c (Distributive)
(c) (b + c) × a = b × a + c × a
(d) k(a × b) = (ka) × b = a × (kb) (Associative)
(e) a×0=0×a =0
(f ) a·a =0

1.8 Scalar triple product

 
a1 a2 a3
(a · b × c) = det  b1 b2 b3  (1.14)
 
c1 c2 c3

If we switch two rows of this matrix, the product will be multiplied

by −1.
The absolute value of scalar triple product will give us the volume of the
parallelepiped that has a, b, c as adjacent edges. Therefore, a · (b × c) =
0 iff they lie on the same plane.
4 1. VECTORS IN RN SPACE
2

Lines and Planes

2.1 Equations of Lines

The line in 3-d space that passes through the point P0 (x0 , y0 , z0 ) and is
parallel to the non-zero vector v =< a, b, c >= ai+bj+ck has equations:

x = x0 + at, y = y0 + bt, z = z0 + ct (P arametric) (2.1)

l =< x0 , y0 , z0 > +t < a, b, c > (V ector) (2.2)

If two lines doesn’t intercept or parallel to each other in a 3-d space, they
are skew.

2.2 Equations of Planes

Definition: A vector perpendicular to a plane is called a normal to that
plane.
A plane which passing through P0 (x0 , y0 , z0 ) and having n =< a, b, c >as
its normal has equations:

a(x − x0 ) + b(y − y0 ) + c(z − z0 ) = 0 (P oint − N ormalf orm)

(2.3)
ax + by + cz + d = 0 (d = −ax0 − by0 − cz0 )(Generalf orm)
(2.4)

2.3 Angle between Planes

For two planes that have n1 and n2 as its normal, the acute angle between
them θ can be obtained from the following equation:

|n1 · n2 |
cos θ = (2.5)
kn1 k kn2 k

5
6 2. LINES AND PLANES

2.4 Distance

The distance D between a point P0 (x0 , y0 , z0 ) and the plane ax + by +

cz + d = 0 is:

|ax0 + by0 + cz0 + d|

D= √ (2.6)
a2 + b2 + c2

Example: Distance between two skew lines:

Solution Find two parallel planes passing through the two skew lines by
taking the cross product of their direction vectors, and find the distance
between these two planes.

Example: Find the equation of the plane that passes through the
line of intersection of the planes P1 : x − z = 1 and P2 : y + 2z = 3
and is perpendicular to the plane P3 : x + y − 2z = 1.

Solutions: Find the direction vector for the line of intersection of P1 and
P2 by taking the cross product of their normal vectors. Then take the cross
product of the direction vector of the line and the normal vector of P3 to
find the normal vector of the desired plane. Then find a point on the line of
intersection and solve for the plane.

2.5 Cylindrical and Spherical Coordinates

Definition. The three coordinates (r, θ, z) of a point P are defined as:

r is the Euclidean distance from the z-axis to the point P .
θ is the angle between x-axis and the line from the origin to the projection
of P on xy-plane.
The height z is the signed distance from the xy-plane to the point P .

Conversion between Rectangular and Cylindrical Coordinates:

Suppose the rectangular and cylindrical coordinate of the same point is

2.5. CYLINDRICAL AND SPHERICAL COORDINATES 7

respectively (x, y, z) and (r, θ, z).

Rectangular to Cylindrical
q
r= x2 + y 2
y
tan θ =
x
z=z
Cylindrical to Rectangular
x = r cos θ
y = r sin θ
z=z

Note that θ is not always tan−1 xy . To find the correct θ, we must know
which quadrant of xy-plane the point lies over.

Definition. The spherical coordinates of a point P are then defined as

follows:
The radius ρ is the Euclidean distance from the origin O to P .
The azimuth θ is the signed angle measured from the azimuth reference
direction to the orthogonal projection of the line segment OP on the xy-
plane.
The inclination φ is the angle between the z-axis and the line segment OP .

Conversion Between Spherical and Other Two kinds of Coor-

dinates:

Suppose the rectangular, cylindrical, and spherical coordinate of the same

point is respectively (x, y, z), (r, θ, z), and (ρ, θ, φ). θ in cylindrical co-
ordinate is the same as θ in spherical coordinate.

Spherical to Rectangular or Cylindrical

r = ρ sin φ
x = r cos θ = ρ sin φ cos θ
y = r sin θ = ρ sin φ sin θ
z = ρ cos φ
Rectangular to Spherical
p q
ρ= r2 + z 2 = x2 + y 2 + z 2
y
tan θ =
x
z z
cos φ = = p
ρ x2 + y2 + z2
8 2. LINES AND PLANES
3

Quadric Surfaces

3.1 Traces
To help graphing a complex surface in a 3-d space, we obtain traces, or the
curves(mesh lines) formed by cutting this surface with well-chosen planes.
Usually, surfaces are built up from traces in planes that are parallel to the
coordinate planes.

3.2 Type of Quadric Surfaces

9
10 3. QUADRIC SURFACES

Name Equation Figure

x2 y2 z2
Ellipsoid a2
+ b2
+ c2
=1

x2 y2 z2
Hyperboloid of one sheet a2
+ b2
− c2
=1

z2 x2 y2
Hyperboloid of two sheets c2
− a2
− b2
=1

x2 y2
Elliptic cone z2 = a2
+ b2

x2 y2
Elliptic paraboloid z= a2
+ b2

y2 x2
Hyperbolic paraboloid z= b2
− a2
4

Calculus of Vector-Valued
Functions

4.1 Orientation/Direction of its graph

The direction a graph of a vector-valued function goes when its parameter,
t, increases is called the orientation or direction of increasing parameter.

4.2 Domain and Natural Domain

The domain of a vector-valued function is the set of all allowable values of
t. The natural domain of a vector-valued function is the intersection of its
component functions’ domain.

4.3 Radius Vector/Position Vector

If a function can be expressed as F (t) =< f (t), g(t), h(t) >, then the
position vector of it at t = k is < f (k), g(k), h(k) >.

4.4 Vector Form of A Line Segment

For two vectorsr0 and r1 that has its initial point at origin, the line passes
through the terminal points of them can be written as:

r = r0 + t(r1 − r0 ) (4.1)

And this is called the two-point vector form o a line.

11
12 4. CALCULUS OF VECTOR-VALUED FUNCTIONS

4.5 Calculus of Vector-Valued Functions

The Calculus of vector-valued functions in 2-d and 3-d space is similar to
"normal" functions:just apply each operator to its component func-
tions and "sum" them up. The definition of integrable, differentiable
and continuous is also similar:each property requires its component
functions have the corresponding property.
The tangent line of the graph at point r(t0 ):

r = t0 + tr0 (t0 ) (4.2)

For the dot product and cross product, which are unique to vector-valued
functions, the derivative is defined as following:
d dr2 dr1
[r1 (t) · r2 (t)] = r1 (t) · · r2 (t) + (4.3)
dt dt dt
d dr2 dr1
[r1 (t) × r2 (t)] = r1 (t) × + × r2 (t) (4.4)
dt dt dt
In 2-d space, the tangent line to a circle is perpendicular to the radius
at the point of tangency. Similarly, in for a vector-valued function, if kr(t)k
is constant for all t, then:

r(t) · r0 (t) = 0 (4.5)

that is, r(t) and r0 (t) are orthogonal for all t.

4.6 Arc Length

In a 2-d space, the arc length L of a parametric curve x = x(t), y =
y(t), (a ≤ t ≤ b) can be given as:
s
Z b
dx dy
L= ( )2 + ( )2 dt (4.6)
a dt dt
Lemma. In a 2-d space, the arc length L of a function y = f (x) that itself
and its derivative is continuous on [a, b] is:
Z
L= ds (4.7)

where
s
dy
ds = 1+( )2 dx if y = f (x), a ≤ x ≤ b
dx
s
dx
ds = 1+( )2 dy if x = g(y), c ≤ y ≤ d
dy
4.6. ARC LENGTH 13

Proof. As we can see in the figure below, the arc length is the sum of distance
between n consecutive points when n → ∞

Arc Length L can be written as:

n q
X
L = lim (xi+1 − xi )2 + (yi+1 − yi )2
n→∞
i=1

Additionally, we can see that

q q
(xi+1 − xi )2 + (yi+1 − yi )2 = ∆x2 + ∆yi2

According to Mean Value Theorem, there exists an x̄ such that

∆yi = f 0 (x̄i )∆x

Thus
q q
∆x2 + ∆yi2 = ∆x2 + ∆yi2
q
= ∆x2 + (f 0 (x̄i )∆x)2
q
= 1 + [f 0 (x̄i )]2 ∆x
14 4. CALCULUS OF VECTOR-VALUED FUNCTIONS

The exact length of the given curve is

n q
X
L = lim (xi+1 − xi )2 + (yi+1 − yi )2
n→∞
i=1
Xn q
= lim 1 + [f 0 (x̄i )]2 ∆x
n→∞
i=1
Z bq
= 1 + [f 0 (x)]2 dx
a
s
Z b
dy
= 1+( )2 dx
a dx

Now we can prove Theorem (4.6):

Proof. Recall that x = x(t), y = y(t), therefore

s
Z b
dy
L= 1+( )2 dx
a dx
v
Z bu dy 2
t1 + ( dt ) dx dt
u
= dx
a ( dt )2 dt
v
Z bu dy 2
t1 + ( dt ) dx dt
u
= dx
a ( dt )2 dt
s
Z b
1 dx dy dx
= dx
( )2 + ( )2 dt
a | | dt dt dt
dt

dx
If we assume that dt
≥ 0,then
s
Z b
dx dy
L= ( )2 + ( )2 dt
a dt dt

Analogously, the arc length L of a smoothly parametrized function(have

a continuously turning tangent vector) in 3-d space is
s
Z b
dx dy dz
L= ( )2 + ( )2 + ( )2 dt
a dt dt dt
Z b
dr
= dt (4.8)
a dt
4.7. ARC LENGTH AS A PARAMETER 15

4.7 Arc Length as A Parameter

Sometime it would be more convenient to replace t with s, which is the
length of arc measured along the curve from some fixed reference point.
There are three steps:
Step 1. Select an reference point.
Step 2. Choose one direction from the reference point as the positive di-
rection.
Step 3. Change the length s to a "signed" length, which means s is positive
if s "moves along the curve" to its positive direction.
Note that there are infinitely many different arc length parameterizations.

Theorem 4.7.1. Chain Rule Let r(t) be a vector-valued function in 2-

d/3-d space that is differentiable with respect to t. If t = g(τ ) is a change
of parameter in which g is differentiable with respect to τ , then r(g(τ )) is
differentiable with respect to τ and

dr dr dt
= (4.9)
dτ dt dτ

A change in parameter is smooth if r(g(τ )) is smooth and r(t) is smooth.

dt dt
For all τ , dτ > 0 is called a positive change of parameter while dτ < 0 is
called a negative change of parameter.

Theorem 4.7.2. Let C be the graph of a smooth vector-valued function r(t)

in 2-d or 3-d space, and let r(t0 ) be any point on C. Then the following
formula defines a positive change of parameter from t to s, where s is an
arc length parameter having r(t0 ) as its reference point:
Z t
dr
s= du (4.10)
t0 du

Theorem 4.7.3. If C is the graph of a smooth vector-valued function r(t)

in 2-d or 3-d space, where t is a general parameter, and if s is the arc
length parameter for C defined by Theorem 2, then for every value of t the
tangent vector has length
dr ds
= (4.11)
dt dt
Proof. This can be derived from applying the Fundamental Theorem of Cal-
culus to Theorem 2.
16 4. CALCULUS OF VECTOR-VALUED FUNCTIONS

Theorem 4.7.4. If C is the graph of a smooth vector-valued function r(t)

in 2-d or 3-d space, where s is the arc length parameter, then for every value
of s the tangent vector to C has length

dr
=1 (4.12)
ds
Proof. Let t = s in Theorem 3.

Theorem 4.7.5. If C is the graph of a smooth vector-valued function r(t)

in 2-d or 3-d space, and if
dr
=1 (4.13)
dt
for every value of t, then t is an arc length parameter that has its reference
point at the point on C where t = 0.

Proof. The formula

Z t
dr
s= du
0 du
defines an arc length parameter for C with reference point r(0). Note that

dr
=1
dt
by hypothesis. Thus the formula can be rewrite as
Z t
s= du = t − 0 = t
0

4.8 Unit Tangent, Normal, and Binormal Vectors

Definition. The unit tangent of a smooth vector-valued function r(t) in
2-d space or 3-d space that points in the direction of increasing parameter
can be expressed as:
r0 (t)
T(t) = 0
kr (t)k
and it’s called the unit tangent vector to C at t.

Note that for all smooth parameterization which induce the same direc-
tion have the same unit tangent vector.

Recall that if a vector-valued function r(t) has constant norm, then r(t)
and r0 (t) are orthogonal. Because T(t) has constant norm 1, so T(t) and
4.9. CURVATURE 17

T0 (t) are orthogonal. This implies that T0 (t) is perpendicular to the tan-
gent line to C at t, so we say that T0 (t) is normal to C at t. If T0 (t) 6= 0,
then
T0 (t)
N(t) =
kT0 (t)k
is the principle unit normal vector, or simply unit normal vector to C at t
and points in the same direction as T0 (t).

The unit normal vector always points toward the concave side of C in 2-d
space.

According to Theorem 4.7.4, kr0 (t)k = 1. Thus

T(s) = r0 (s)
and consequently
r00 (s)
N(s) =
kr00 (s)k
Definition. The binormal vector to C at t can be defined as
B(t) = T(t) × N(t)
r0 (t) × r00 (t)
=
kr0 (t) × r00 (t)k
that is, the cross product of its unit tangent vector and unit normal vector
and the direction of binormal vector is determined by the right-hand rule.
kB(t)k = 1 since kT(t) × N(t)k = kT(t)k kN(t)k sin(π/2) = 1.

In terms of arc length parameteriation, it can be expressed as

r0 (s) × r00 (s)
B(s) =
kr00 (s)k
Together with unit tangent vector and unit normal vector, the binormal
vector define three mutually perpendicular planes that point through that
point – the TB-plane (called the rectifying plane), the TN-plane (called
the osculating plane), and the NB-plane (called the normal plane). The
coordinate system(right-hand) system determined by these three vectors is
called the TNB-frame.

4.9 Curvature
Definition. If C is a smooth curve in 2-d space or 3-d space that is
parametrized by arc length, then the curvature of C, denoted by κ = κ(s),
is defined by
dT
κ(s) = = r00 (s)
ds
18 4. CALCULUS OF VECTOR-VALUED FUNCTIONS

Theorem 4.9.1. If r(t) is a smooth vector-valued function in 2-d space

or 3-d space, then for each value of t at which T(t) and r00 (t) exist, the
curvature κ can be expressed as

kT0 (t)k
κ(t) =
kr0 (t)k
kr0 (t) × r00 (t)k
=
kr0 (t)k3

Definition. If at P κ 6= 0, then there exists a unique circle Ω passing

through P on the concave side of C such that the curvature of Ω is κ and
it shares a common tangent vector with C. Then Ω is called a osculating
circle or circle of curvature at P , 1/κ is called the radius of curvature at
P , and the center of Ω is called the center of curvature at P .

Definition. Let φ(s) be the angle measured counterclockwise from the

direction of the positive x-axis to the unit tangent vector T in terms of
arc length. Then
dφ
κ(s) = | |
ds

4.10 Motion Along A Curve

Definition. If a particle moves along a smooth vector-valued function r(t),
then the velocity of this particle is

ds
v(t) = r0 (t) = T(t)
dt
where v(t) is the tangent vector at t. The direction of v(t) gives the instan-
taneous direction of motion at t. The magnitude of v(t) is the instantaneous
rate of change of arc length as a function of time, or just simply speed.

Definition. The acceleration of this particle with velocity of v(t) is a(t) =

v0 (t) = r00 (t). What’s more

d2 s ds
a(t) = 2
T(t) + κ(t)( )2 N(t)
dt dt
If the particle is travelling on a circle with radius r, the speed v0 is constant,
v2
and ka(t)k = r0 .

Example 4.1. Over a time interval [t1 , t2 ], the distance traveled by a par-
ticle is Z t2
s= kv(t)k dt
t1
4.11. PROJECTILE MOTION 19

d2 s
Definition. If aT = dt2
and aN = κ(t)( ds
dt
)2 , then

a = aT T + aN N

and aT is called the tangential component of acceleration and aN is called

normal component of acceleration . aT T is the tangential vector component
of acceleration and aN N is called the normal vector component of acceler-
ation

Example 4.2.
v·a
aT =
kvk
kv × ak
aN =
kvk
kv × ak
κ=
kvk3

In 2-d space, the cross product can be computed by viewing r(t) =<
x(t), y(t) > as r(t) =< x(t), y(t), 0 > in 3-d space.

Example 4.3.
kak2 = aT2 + aN
2

for all smooth position functions.

4.11 Projectile Motion

Theorem 4.11.1. For a projectile motion

gt2
r(t) = − j + tv0 + s0 j
2
gt2
= (− + s0 )j + tv0
2
where g is the gravitational acceleration, v0 is the initial velocity, and s0 is
the initial height.
20 4. CALCULUS OF VECTOR-VALUED FUNCTIONS
5

Multivariable Calculus

5.1 Definition of Multivariable functions and their

graph
Definition. A real-valued function f of n real variables is a mapping that
assign an ordered pair of n real numbers to a real number in D ⊂ Rn .

Definition. If f is a function of two variables and k is a real number, then

the level curve of f of height k is {(x, y)|f (x, y) = k}. A collection of
many level curves for a function f all drawn in the same xy-plane is called
a contour plot/map of f .

5.2 Properties of Sets in 2-d and 3-d space

Definition. For r > 0 and a point P , the open ball of radius r centered at
P means the set of all points whose distance to P is less than r and denotes
Br (P ).

Suppose D ⊂ R2 or D ⊂ R3 .

Definition. A point P is said to be an interior point of D if ∃r >

0(Br (P ) ⊂ D).

Definition. A point P is said to be an boundary point of D if ∀r >

0((Br (P ) ∩ D 6= ∅) ∧ (Br (P ) \ D 6= ∅))

Definition. For a set D in 2-d and 3-d space, the set of all its interior points
is called the interior of D. The set of all its boundary points is called the
boundary of D.

Definition. A set is to be said open if it contains none of its boundary

points, and when it contains all of its boundary points, it’s called a closed
set.

21
22 5. MULTIVARIABLE CALCULUS

However, some sets are neither open or close.

Example 5.1. A set D in n-space is both open and closed iff D = Rn or

D = ∅.

Definition. A subset D of Rn is said to be bounded if ∃r > 0∃P (D ⊂

Br (P )). That is, the set is contained in some set with finite radius. It’s
unbounded if it’s not bounded.

Definition. A point P is said to be the accumulation point of some D ⊂ Rn

if Br (P ) ⊂ D ∧ ∀r > 0(Br (P ) \ D \ {P } 6= ∅).