Applied Math.

III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

PART II: VECTOR CALCULUS [VECTOR ANALYSIS]

Vector calculus (also called vector analysis) deals with calculus (derivatives and integrals) of
functions of the form: 𝑓: 𝐷 ⊆ ℝ → ℝ𝑛 and : 𝐷 ⊆ ℝ𝑚 → ℝ𝑛 . Such functions are called vector
valued functions or, briefly, vector functions In this part of the course (chapter 3 and 4), we
study the differential calculus and integral calculus of vector valued functions.

Chapter 3: Vector Differential Calculus

3.1 Introduction
3.1.1 Introduction to Vector Functions
Definition 1: A vector-valued function (of a real variable) is a function that assigns a vector
to every real number in a certain set I of real numbers. More precisely, a vector valued function
is a function whose domain is a subset of ℝ and whose range is a subset of ℝ𝑛 for some natural
number n greater than or equal to 2.

A vector valued function is usually denoted by a bold faced Latin alphabet such as F, G, H, r, u,
v etc. or by a Latin alphabet together with a small picture of arrow over it, such as 𝐹⃗ , 𝐺⃗ , 𝐻
⃗⃗ , 𝑟⃗,
etc. The independent variable of a vector valued function is usually denoted by 𝑡, because in
most applications it indicates time. For example: Suppose one wants to follow (study) the motion
of a particle (say a satellite) moving in space for the time duration of 2 hrs(say over time interval
[0, 2]). If P is the point at which the particle is located at time t and O is a fixed point on earth,
→ →
then OP is a variable vector and the association t → OP defines a vector-valued function.

Figure 1.vector function

Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

In this course we consider mainly three dimensional vector valued functions. A three
dimensional vector valued function F is given by a formula of the type
F(t ) = f1 (t )i + f 2 (t ) j + f 3 (t )k .The functions f1 , f 2 , and f 3 are called the component functions

of F.

Example 1:

(1) Let r(t ) = (t + 1)i + (t 2 − 5) j + (t 2 + 2)k ,

a) Find the domain r.

b) Find r (1) and sketch it.
Solution:

a) The domain of r is ℝ.
b) 𝒓(1) = (1 + 1)𝒊 + (12 − 5)𝒋 + (12 + 2)𝒌 = 2𝒊 − 4𝒋 + 3𝒌

(2) The function Let 𝐹⃗ : [0, 1] → ℝ3 given by 𝐹⃗ (𝑡) ≔ 𝒊 is a constant vector function.

Graph and Trace of Vector Valued Functions:

Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

⃗⃗⃗⃗⃗⃗⃗⃗⃗=〈𝑓1 (𝑡), 𝑓2 (𝑡), 𝑓3 (𝑡)〉. The set

Let 𝐹⃗ : 𝐼 ⊆ ℝ → ℝ𝑛 be given by 𝐹(𝑡)

⃗⃗⃗⃗⃗⃗⃗⃗⃗): 𝑡 ∈ 𝐼}
Graph 𝐹⃗ : = {(𝑡, 𝐹(𝑡)

is called the graph of 𝐹⃗ . It is impossible to draw the graph of F, because it requires 4 axes to do
so. But we can draw its range. The graph of the range of 𝐹⃗ is called the trace of 𝐹⃗ . The trace is
usually a curve and we denote it by 𝒞. As t increases 𝐹⃗ moves on 𝒞 in a unique direction. This
direction is called the orientation (positive direction) of 𝐹⃗ . One way of sketching the trace of a
vector function is by using the following procedures.

Plotting the Trace of a Vector Function r(t): Procedures: t r(t)

1. Select sample values for the parameter t from its domain. t1 r(t1 )
2. For each selected value of t , evaluate r(t).
t 2 r(t 2 )
3. Plot the terminal points of r(t), for each r(t) in step 2.
4. Connect the plotted points with a curve to visualize the trace of the vector function. ⋮

t n r(t n )

Figure 2.Trace of a vector Function

We can consider r(t) as the position vector of a point P moving in space and 𝒞 as the trajectory
of the point. Recall that a vector ⟨𝑎, 𝑏, 𝑐⟩ is called a position vector of a point 𝑃(𝑎, 𝑏, 𝑐)

if it starts at the origin and ends at the point (𝑎, 𝑏, 𝑐).

Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

Example 2: Let

F(t ) = cos ti + sin tj

Sketch the trace of 𝐹⃗ and indicate its orientation.

Solution: The curve traced out by F entirely lies in the xy-plane because the kth component of F
is 0. Let (x, y) be a point on the trace. Then (x, y) = F(t) for some t. This implies that

𝑥 = 𝑐𝑜𝑠𝑡
{
𝑦 = 𝑠𝑖𝑛𝑡

⇒ 𝑥 2 + 𝑦 2 =1. Thus every point (x, y) on the trace of F is a point of the standard unit circle 𝑥 2 +
𝑦 2 =1. Conversely, each point on the standard unit circle is a point on the trace of F, since the
domain of F is ℝ. Hence the trace of F is the standard unit circle. More over the orientation of
the trace is in anticlockwise (ACCW) direction as suggested by the following sample values.
⃗⃗⃗⃗⃗⃗⃗⃗⃗‖=1 for each t, hence each vector in the range has unit magnitude, so they
(Note also that ‖𝐹(𝑡)
lie on the standard unit circle).

t 0 𝜋 𝜋 3𝜋 2𝜋
2 2
F(t) (1, 0) (0,1) (-1,0) (0, -1) (1, 0)

Figure 3.Trace of F (t) =cost i +sint j

Example 3: Sketch the trace of G(t ) = − cos ti + sin tj, t  0,   and indicate its orientation.
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

Solution: Using similar arguments as Example 1 we see that the trace is the upper half of
standard unit circle with clockwise orientation.

Figure 4.Trace of G(t ) = − cos ti + sin tj, t  0,  

Example 4: Let r (t ) = a cos ti + b sin tj + ctk for t  0 and positive constants a, b and c. Sketch
the curve 𝒞 determined by r (t), and indicate the orientation.

Solution: If (𝑥, 𝑦, 𝑧) is a point on the trace, then there exist 𝑡 ≥ 0 such that

x = a cost, y = a sin t , z = bt ; .If we eliminate the parameter t in the first two equations we

x2 y2 x2 y2
obtain + = 1 .Thus every point 𝑃(𝑥, 𝑦, 𝑧) on 𝒞 is on the elliptical cylinder + = 1 . If
a2 b2 a2 b2
t varies from 0 to 2 , the point 𝑃 starts at (a,0,0) and moves up ward while making one
revolution around the cylinder. This curve is called a cylindrical helix.

Figure 5.cylinderical helix

Likewise, a conical helix may be defined as a spiral on a conic surface.

Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

Given a trace 𝒞 of an unknown vector valued function; we are sometimes asked to find a vector-
valued function F such that C is the trace of F. In fact, such function is not unique.

Example 5: Find a vector valued function F given that its trace is the line segment with
orientation from (2, 2, 2) to (0, 0, 0 ).

Example 6: A circle of radius r rolls along the x axis in the positive direction at the rate of 1
radian per unit of time. Let P be the point on the circle that is at the origin at time 0. Find a
vector-valued function F that traces out the curve described by P.

Figure 6: Cycloid

Solution: As can be seen from Fig 7, the vector F (t) from the origin to P at time t can be written
as a + b, where a points from the origin to the center C of the circle, and b points from C to P.
To express a in terms of t, we observe that at time t the circle has rolled a horizontal distance rt
along the x axis, so the i component of a is rt. Since C is r units from the x axis, the j component
of a is r. Consequently, a = rti + rj . In order to express b in terms of t, we first use the
hypothesis that the circle rolls at the rate of 1 radian per unit of time, which implies that t can be

interpreted as the number of radians through which CP has rotated since time 0. Therefore at

time t the vector CP makes an angle of (3 / 2 − t ) radians with the positive x-axis (Fig.5). Since
the length of b is the radius of the circle, we conclude that
3 3
b = CP = r cos( − t )i + r sin( − t ) j. consequently
2 2
3 3
F(t ) = a + b = ( rti + rj )+ r cos( − t )i + r sin( − t ) j.
2 2
= r (t − sin t )i + r (1 − cos t ) j .
 Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

Figure 7.Cycloid

A cycloid is the curve traced out by a point on a circle as it rolls along a straight line. It is an
example of roulette. Roulette is a curve generated by tracing the trajectory of a point on a curve
rolling on another curve.

Certain traces are defined by two or more formulas:

Example 6: Parameterize the closed curve determined by the line 𝑦 = − 𝑥 and the parabola
𝑦 = 𝑥2.

3.1.2 Limit and Continuity of Vector Valued Functions

A. Limit of Vector Valued Functions

→ →
Definition 3: A vector valued function of real variable t, F (t ) is said to have the limit l as t
→
approaches t0 , if u (t ) is defined on some neighborhood of t0 (possibly except at t0 ) and

→ →
lim F (t ) − l = 0
t →t 0
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

→ →
F (t ) - l

Constant

vector,
Variable
vector,

Figure 8.Limit of a vector Function

Theorem 1: Let F(t ) = f1 (t )i + f 2 (t ) j + f 3 (t )k. Then F has a limit at a if and only if

f1 , f 2 , and f 3 have limits at a. In that case

    
lim F(t ) = lim f1 (t ) i + lim f 2 (t ) j + lim f 3 (t ) k
t →a t →a t →a t →a

Example 7: Find

 sin t 
lim  i + e t j + (t + 2 )k 
t →0
 t 

Solution:

lim 
t →0
 sin t
 t
 
i + e t j + (t + 2 )k  =  lim
sin t 
  t →0 t 
( ) (
i + lim e j + lim (t + 2 ) k
t →0
t
t →0
)
= i + j + 2k.

Theorem 2 (Limit Rules): Let F and G be vector-valued functions, and let f and g be real-
valued functions. Assume that lim F (t ) and lim G (t ) exist and that lim f (t ) exists and
t →a t →a t→ a

lim g ( s ) = a. Then
s →b
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

lim (F + G )(t ) = lim F(t ) + lim G (t )

a. t →a t →0 t →0

lim (F − G )(t ) = lim F(t ) − lim G (t )

b. t →a t →0 t →0

lim fF(t ) = lim f (t ) lim F(t )

c. t →a t →a t →a

lim (F.G )(t ) = lim F(t )  lim G (t )

d. t →a t →a t →a

lim (F  G )(t ) = lim F(t )  lim G (t )

e. t →a t →a t →a

lim (F  g )( s) = lim F(t )

f. s →b t →a

sin( t − 1) t +3 t−2
Example8: Let F(t ) = i+ j + cos tk and G (t ) = (t 2 + 1)i − j − t 2 + 1k
t −1 t−2 t +3

Find lim (F  G )(t ) and lim (F  G )(t )

t →1 t →1

Solution: For each of the required limits we have a choice of methods. We can (1) find the
product of F and G and take the limit of the product, or (2) find lim F (t ) and lim G (t ), take the
t →1 t →1

product of these limits, and apply Definition 3.3.In our case for lim (F  G )(t ) and lim (F  G )(t )
t →1 t →1

 sin( t − 1) t +3 
we use method (2). Since lim F(t ) = lim  i+ j + cos tk  = i − 4 j − k
t →1 t →1
 t −1 t−2 

 t−2  1
and lim G (t ) = lim  (t 2 + 1)i − j − t 2 + 1k  = 2i + j − 2k .
t →1 t →1
 t +3  4

 
Hence lim (F.G )(t ) = lim F(t )  lim G (t ) = (i − 4 j − k )   2i + j − 2k  = 1 + 2
1
t →1 t →1 t →1
 4 

 
lim (F  G )(t ) = lim F(t )  lim G (t ) = (i − 4 j − k )   2i + j − 2k 
1
And
t →1 t →1 t →1
 4 

16 2 + 1 33
= i − (2 − 2 ) j + k .
4 4
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

B. Continuity of Vector Valued Functions

Definition 4: A vector-valued function F is said to be continuous at a if

lim F(t ) = F(a) .

t →a

Example 9:

Theorem 3: If 𝑭(t) = 𝑓1 (𝑡)𝒊 + 𝑓2 (𝑡)𝒋 + 𝑓3 (𝑡)𝒌 , then F is continuous at a if and only

if 𝑓1 (𝑡)𝒊 + 𝑓2 (𝑡)𝒋 + 𝑓3 (𝑡)𝒌 are continuous at a.

Continuity on an interval is defined in the usual way.

3.1.3 Derivative of Vector Valued Functions

Definition 5: Let F be a vector-valued function. The derivative of F is the vector-valued
function F' defined by
F(t + h) − F(t )
F' (t ) = lim
h →0 h

for every t whenever the limit exists.

 Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

𝐹 (𝑡 + ℎ) − 𝐹(𝑡)

F (t + h)

𝐹 (𝑡)

Figure 9 .Derivative of vector function

Note: The derivative vector F' (t) is “tangent” to the trace of F at F (t) and it has the same
direction as the trace.

Justification:

Case1: If h>0, then the curve is traced from F (t) to F(t + h) and F' (t) has the same direction as
F(t + h) - F(t). (Because F' (t) is the limit of multiples of F (t + h) - F (t) by a positive
number, 1/h)

Case2: If h<0, the curve is traced out from F (t + h) to F (t). Since the derivative is the limit of
multiples of F (t + h) - F (t) by the negative number 1/h, so F' (t) and F (t + h) - F (t)
have opposite directions. Moreover, the trace and F (t + h) - F (t) also have opposite directions.
So, the derivative vector and the trace have same direction.

The next theorem states that to find F ' (t ) we differentiate each component of F (t).

Theorem 4: If F(t ) = f1 (t )i + f 2 (t ) j + f 3 (t )k and f1 , f 2 , and f 3 are differentiable, then

F' (t ) = f1 ' (t )i + f 2 ' (t ) j + f 3 ' (t )k .

Proof: By definition 3.5

Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

F(t + h) − F(t )
F' (t ) = lim
h →0 h

[ f 1 (t + h)i + f 2 (t + h) j + f 3 (t + h)k ] − [ f 1 (t )i + f 2 (t ) j + f 3 (t )k ]
= lim
h →0 h
 f (t + h) − f 1 (t ) f (t + h) − f 2 (t ) f (t + h) − f 3 (t ) 
= lim  1 i+ 2 j+ 3 k
h →0
 h h h 

Taking the limit of each component gives us the conclusion of the theorem.

If F ' (t ) exists, we say that F is differentiable at t. If I is an interval, then we say that F is

differentiable on I if the components of F are differentiable on I.

Example 9: Let F(t ) = ln t i + e −3t j + t 2 k.

a) Find the domain of F and determine where F is continuous.

b) Find F' (t ) and F" (t ).
Solution:

a) Since ln t is not defined for t  0, the domain of F is the set of positive real numbers.
Moreover, F is continuous throughout its domain, since each component determines a
continuous function.
b) Applying theorem (3.6), we have
1 1
F' (t ) = i − 3e −3t j + 2tk and F" (t ) = − 2
i + 9e −3t j + 2k.
t t

Theorem 5 (Differentiation Rules):

Let F, G, and f be differentiable at a, and let g be differentiable at b with g (b) = a. Then the
following hold.

a. (F + G )' (a) = F' (a) + G ' (a)

b. (F − G )' (a) = F' (a) − G ' (a)

c. ( fF)' (a) = f ' (a)F(a) + f (a)F' (a)

d. (F  G )' (a) = F' (a)  G (a) + F(a )  G ' (a)

Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

e. (F  G )' (a) = F' (a)  G (a) + F(a)  G ' (a)

f. (F  g )' (b) = F' ( g (b)) g ' (b) = F' (a) g ' (b) .

Example 10: Let F(t ) = arctan ti + 5k and G (t ) = i + ln tj − 2tk. Find (F  G )' (t ) and
(F  G )' (t ).
Solution: Two methods are available for finding each required derivative. We can (1) calculate
(F  G )(t ) and (F  G )(t ) and then differentiate it, or (2) find F' (t ) and G ' (t ) and then and use
Theorem 3.7 to find (F  G )' or (F  G )'. We use method (1) to find (F  G )' (t ).

After calculating that (F  G )(t ) = arctan t − 10t , we conclude that

1
(F  G )' (t ) = − 10.
t +1
2

To find (F  G )' (t ), we use method (2). To that end we first find that

1 1
F' (t ) = i and G ' (t ) = j − 2k
t +1
2
t

then

i j k
1 2t ln t
F' (t )  G (t ) = 2 0 0 = 2 j+ 2 k
t +1 t +1 t +1
1 ln t − 2t

and

i j k
5 arctan t
F(t )  G ' (t ) = arctan t 0 5 = − i + 2 arctan tj + k
1 t t
0 −2
t

Now Theorem 3.7(e) yields

(F  G )' (t ) = F' (t )  G (t ) + F(t )  G ' (t )

Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

5  2t   ln t arctan t 
= − i+ 2 + 2 arctan t  j +  2 + k.
t  t +1   t +1 t 

An elementary but significant consequence of Theorem 3.7 is the following.

Corollary: Let F be differentiable on an interval I, and assume that there is a number c such that
F(t ) = c for t in I. Then

F (t )  F' (t ) = 0 , for every t in 𝐼.

Proof: If F(t ) = c for some constant c, then

2
F (t )F (t ) = F(t ) = c2

d d
and [F(t )F(t )] = c 2 = 0 which implies
dt dt

F(t )F' (t ) + F ' (t )F(t ) = 0

2[F(t )F ' (t )] = 0.

Observe in this case that F (t) and F ' (t ) are orthogonal (if F is nonzero).

3.1.4 Definite and Indefinite Integral of Vector-Valued Functions

Similar to the case of the derivative of a vector-valued function, we find the integral of a vector-
valued function F in terms of its component functions. Note that there are several notions of
integral for vector functions. The next chapter is devoted to integrals of vector functions.

Definition 6: Let F(t ) = f1 (t )i + f 2 (t ) j + f 3 (t )k , where f1 , f 2 , and f 3 are integrable on [a, b].

Then we define

b
b  b  b 
a)  F(t )dt =   f1 (t )dt  i +   f 2 (t )dt  j +   f 3 (t )dt k
a a  a  a 

b)  F(t )dt =  f (t )dt i +  f

1 2   f (t )dt k .
(t )dt j + 3
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

Note: If G ' (t ) = F(t ), then 𝑮(𝑡) is and anti-derivative or 𝑭(𝑡). The next result is analogous to
the fundamental theorem of calculus.

Theorem 6: If 𝑮 (𝑡) is and anti-derivative of 𝑭 (𝑡) on [𝑎, 𝑏], then

 F(t )dt = G(t ) = G(b) − G(a).

b
a
a

2
Example 11: Find  F(t )dt if F(t ) = 12t 3 i + 4e 2t j + (t + 1) −1 k.
0

Solution: Finding an anti-derivative for each component of F(t), we obtain

G(t ) = 3t 4 i + 2e 2t j + ln( t + 1)k.

Since G ' (t ) = F(t ), it follows from Theorem 3.10 that

 F(t )dt = G (2) − G (0)

0

= (48i + 2e 4 j + ln 3k ) − (0i + 2 j + 0k )
= 48i + 2(e 4 − 1) j + ln 3k.

The theory of indefinite integrals of vector-valued functions is similar to that developed

for real-valued functions. Hence if G(t) is and antiderivative or F(t), then every antiderivative
has the form G(t) + c for some (constant) vector c, and we write

 F(t )dt = G (t ) + c where G ' (t ) = F(t ).

Indeed, since the velocity v is the derivative of the position vector r and the acceleration a is the
derivative of the velocity v, it follows that

 v(t )dt =  r' (t )dt = r(t ) + cand  a(t )dt =  v' (t )dt = v(t ) + c
Example 12: An object has initial position r0 and initial velocity v 0 and undergoes a constant

acceleration − gk . Show that the position of the object at any time t is given by
1
r (t ) = − gt 2 k + tv 0 + r0
2
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

Solution: By hypothesis a(t ) = − gk. Successive integration yield

v(t ) =  a(t )dt =  − gkdt = − gtk + c (1)

1
and r (t ) =  v(t )dt =  (− gtk + c)kdt = − gt 2 k + ct + c1 (2)
2
where c and c1 are constant vectors. By assumption v(0) = v 0 , so that by letting t = 0 in (1) we

find that c = v 0 . Similarly, r(0) = r0 , so that c1 = r0 by (2).Therefore (2) can be written as

1
r (t ) = − gt 2 k + tv 0 + r0.
2

Exercise 3.1:

1. The traces of which of the following vector functions pass through the point (1, 4)?
a) 𝐹⃗ (𝑡) = 〈𝑡 2 , 𝑡 + 3〉 b) 𝐹⃗ (𝑡) = 〈𝑡 2 , 3 − 𝑡〉 c) 𝐹⃗ (𝑡) = 〈𝑡 2 , 𝑡 − 3〉 d) 𝐹⃗ (𝑡) =
〈𝑡 − 3, 𝑡 2 〉
2. Sketch the curve traced out by the vector valued function. Indicate the direction in which
the curve is traced out.
a) F(t) = 〈−5 + 𝑡, −4 + 2𝑡〉
b) F(t) = 〈−5 + 𝑡, −4 + 2𝑡〉; 0 ≤ 𝑡 ≤ 1
c) F(t) = 〈−5 + 𝑡, −4 + 2𝑡〉; 0 ≤ 𝑡
d) r(t) = (t - 1)i + t2 j
e) F(t ) = 4 cos ti + 4 sin tj + 5k for 0  t  
f) F(t ) = 4 cos ti + 4 sin tj + 5k for 0  t  4
g) F(t ) = 4 cos ti + 4 sin tj + tk for 0  t  
h) F(t ) = 4 cos ti + 4 sin tj + tk for 0  t  4
i) r(t) = 〈3 𝑐𝑜𝑠𝑡, 2 𝑠𝑖𝑛 𝑡〉
j) r(t) = 〈𝑠𝑖𝑛(𝑡), 𝑡, 𝑐𝑜𝑠(𝑡)〉

k) F(t ) = 3 cos ti − 4 sin tj + tk for 0  t  

Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

3. Find a vector-valued function whose trace is the ellipse of major diameter 10 parallel to
the y-axis and minor diameter 4 parallel to the z-axis, centered at (5, 2, 1).
4. Find a vector function for the curve of intersection of 𝑥 2 + 𝑦 2 = 9 and 𝑦 + 𝑧 = 2.
5. Find a vector equation of the space curve that is the intersection of the cylinder 𝒚𝟐 +
𝒛𝟐 = 𝟐 and the plane 𝒂𝒙 + 𝒃𝒚 + 𝒄𝒛 = 𝒅, where 𝒂 ≠ 𝟎.
6. Show that the trace of 𝒓(𝒕) = 〈 𝟏 + 𝒕𝟐 , 𝟏 + 𝒕𝟐 , 𝒕〉 , does not intersect the plane
−𝟐𝒙 + 𝟑𝒚 + 𝒛 = 𝟏.
7. For which values of t is r (t) = 〈𝒔𝒊𝒏 𝒕, − 𝒄𝒔𝒄 𝒕, 𝒕𝒂𝒏 𝒕〉 continuous?
8. Draw the trace of r and the vectors r (π) and 𝒓′ (𝝅) where r (t) =〈𝟑 𝒄𝒐𝒔𝒕, 𝟐 𝒔𝒊𝒏 𝒕〉.
t −1
9. Find the derivative of F  g , where F (t ) = ln t i − 4e 2t
j + k and g (t ) = t
t
10. Find a vector function that represents the curve of intersection of the two surfaces 𝑥 2 +
𝑦 2 + 3𝑧 2 = 5(𝑦 ≥ 0) and 𝑥 2 + 𝑧 2 = 4.

Answer #2f: The curve is a circle in the z = 5 plane which has a radius of 4 and a center at (0, 0,
5), traversed twice counterclockwise.

Answer #2h: The curve is a helix on the cylinder 𝑥 2 + 𝑦 2 = 16 which climbs from the point
(4,0,0) to the point (4,0,4)

3.2Curves and Their Lengths

A. Curves
Definition 1: A space curve (or just a curve) 𝒞 is the range of a continuous vector-valued
function r (t ) = x(t )i + y (t ) j + z (t )k on an interval 𝐼 of real numbers.

Note the difference between trace and curve; a curve is a continuous trace and its domain is an
interval. If 𝒞 is the trace of the vector function 𝒓, we say 𝒞 is parameterized by r or r is a
parameterization of 𝒞. Thus 𝑃(𝑥, 𝑦, 𝑧) is a point in 𝒞 if and only if there exist 𝑡 ∈ 𝐼 such that

x = x(t)
{ 𝑦 = 𝑦(𝑡)
𝑧 = 𝑧(𝑡)
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

These equations are called parametric equations for the curve.A curve 𝒞 can also be represented
as the intersection of two surfaces:

G(x, y, z) = 0
{
H(x, y, z) = 0

This is sometimes called Cartesian form of 𝒞. Thus, a curve can be represented in vector form,
parametric form and Cartesian form. (Contrast this with the different forms of equations of a line
in space).

Since we are accustomed to thinking of the graph of any continuous real valued function
as a curve, we now verify that Definition 3.11 is broad enough to include such curves.

Example 1: Let 𝑓 be a continuous real-valued function on an interval I. Show that the graph of
𝑓 is a curve, and find a parameterization of that curve.

Figure 10.Graph of y=f(x) as a curve

Solution: The vector-valued function r defined by r (t ) = ti + f (t ) j for all t in I is continuous and
traced out the graph of f. Thus the graph of f is the range of r, so by Definition 3.11 the graph of
f is a curve.

Example 2: Write the equations of the curve/line/ containing the point P0(x 0, y 0, z 0) and
parallel to the vector 𝑣 =< 𝑎, 𝑏, 𝑐 >.
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

Solution: Let P(x, y, z) be an arbitrary (moving) point on the line. Let r0 and r be position vectors
of P0 and P respectively. Then

→ → −→
r (t ) = r0 + P0 P (Vector form)

 <x(t), y(t), z(t)> = (x 0, y 0, z 0) + t(a, b, c)

 x (t) = x 0+ at , y(t) = y 0 + bt and z(t)= z 0 + ct (parametric form)

* A curve can have many possible parameterizations, i.e., a parameterization of a curve is not
unique.
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

Definition 2 (types of curves): Let 𝐼 = [𝑎, 𝑏] be an interval. A curve 𝒞 with

parameterization 𝒓(𝑡) = 𝑥(𝑡)𝒊 + 𝑦(𝑡)𝒋 + 𝑧(𝑡)𝒌; 𝑡 ∈ 𝐼 is said to be a:

a) simple curve if for every t1, t2 ∈ (𝑎, 𝑏) with t1 ≠ t2 we have 𝑟(𝑡1 ) ≠ 𝑟( 𝑡2 ); that is, r
is one to one on the interior of I.
b) closed curve if r (a) = r (b) and r is one to one on (𝑎, 𝑏) except possibly at a finite
number of points.
c) smooth curve if :
i) r’ exist on I
ii) r’ is continuous on I and
iii) r ' (t )  0 for each 𝑡 interior point of I.
d) piecewise smooth curve if I is composed of a finite number of closed subintervals on
each of which r is smooth
Intuitively (informally), a curve is:

a) simple if it does not cross itself (but its end points can coincide).
b) closed if its initial and terminal points coincide and in addition if it does not cross itself
except possibly at a finite number of points.
c) smooth if it is composed of one piece with no sharp corners (Figure 12.24(a)).
d) piecewise smooth if it is composed of one piece and has at most a finite number of sharp
corners.

Figure11.Smooth and non-smooth curve

Example 3: Show that the standard unit circle is smooth.

Fig.7
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

Solution: The circle can be parameterized by

r (t ) = cos ti + sin tj for 0  t  2 .

The function r is differentiable on [0,2 ], and

r ' (t ) = − sin ti + cos tj for 0  t  2 .

Therefore r ' is continuous on [0,2 ], and

r' (t ) = (− sin t ) 2 + (cos t ) 2 = 1

from this we conclude that r ' (t )  0 for each t in [0,2 ]. It follows that the circle is smooth.

Example 4: Show that the curve

r (t ) = r (t − sin t )i + r (1 − cos t ) j for − 2  t  2

which parameterizes two arches of a cycloid, is piecewise smooth.

Solution: We find that r ' (t ) = r (1 − cos t )i + r sin tj

Even though r ' is continuous, r is not smooth because r ' (0) = 0. However, r ' (t )  0 if t is not
− 2 ,0, or 2 . Therefore r is smooth on [−2 ,0] and on [0,2 ], and hence r is piecewise smooth.

B. Length of a Curve (Arc)

To find the length of a simple curve 𝒞 we may proceed as follows. We inscribe in 𝒞 a
broken line of n chords joining the two end points of 𝒞 as shown in Fig.8. This we do for each
positive integer n in an arbitrary way but so that the maximum chord-length approaches zero as n
approaches infinity. The lengths of the chords can be obtained from Pythagoras theorem. If the
sequence of these lengths l1 , l2 , . . . is convergent to a limit l, then 𝒞 is said to be rectifiable, and
l is called the length of 𝒞.

If 𝒞 is not simple but consists of finitely many rectifiable simple curves, the length of C is
defined to be the sum of the lengths of those curves. We now define the length of an arc as
follows.
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

Let 𝒞 be a piecewise smooth curve with parameterization r on I and let t0 ∈ I. Let P0 and P be the
points of 𝒞 corresponding to t0 and t0 + ∆𝑡. Then if ∆𝑡 is small the length dL of 𝒞 between P0 and
P can be approximated by ∆𝑟 = 𝑟(𝑡 + ∆𝑡) − 𝑟(𝑡). In turn ∆𝑟 can be approximated by 𝑟 ′ (𝑡)𝑑𝑡,
∆𝑟
because 𝑟 ′ (𝑡) = lim 𝑑𝑡. Thus, dL ≈ 𝑟 ′ (𝑡)𝑑𝑡. This motivated the following definition.
𝑡→∞

Figure 12.: Length of a curve

Definition 3: Let 𝓒 be a curve with a piecewise smooth parameterization r defined on[𝒂, 𝒃], then
the length L of 𝓒 is defined by
b b b
dr
L =  r (t ) dt =  dt =  r (t )  r (t )dt
a a
dt a

whose value is independent of the choice of the parametric representation.

(One way to remember this definition is that ‖𝐫′ (t) ‖is speed and so its integral is distance along
the path from 𝑡 = 𝑎 to 𝑡 = 𝑏. which is equivalent to arc length).

Example 4: Show that the circumference L of a circle of radius r is L = 2𝝅r.

Example 5: Find the length of the circular helix

r (t ) = a cos ti + a sin tj + btk for 0  t  2 .

Solution: Since r (t ) = −a sin ti + a cos tj + bk, and r (t ) = (−a sin t ) 2 + (a cos t ) 2 + b 2

Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

From Definition 3.14 we have

2 2
l= 
0
(−a sin t ) 2 + (a cos t ) 2 + b 2 dt = 
0
a 2 + b 2 dt = 2 a 2 + b 2 .

C. Arc length Parameterization

Let 𝓒 be a curve with parameterization r(t), t ∈ [a, b].
If we replace the fixed upper limit b in the above definition with a variable upper limit t, the
integral becomes a function of t, say, s(t); denoting the variable of integration by u, we

t
 dr 
have s(t ) =  r (u ) du where  r = .
a  du 

s(t) is called the arc length function of 𝓒. (Note that s(t) is the length of the portion of 𝓒
between r(a) and r(t)). By the first fundamental theorem of calculus, we get that

s/ (t) = ‖𝐫 ′ (t) ‖

 s/ (t)  0. But r/ (t) cannot be zero as C must be smooth. So, s/ (t) > 0, for all t  0 .

s is strictly increasing.

s has an inverse (we have t = f-1(s))

Hence, t can be regarded as a function of s.

Thus 𝓒 can be parameterized as r/ (t(s)). This is called the arc length parameterization of 𝓒.

The arc length s may serve as a parameter in parametric representation of curves. We

shall see that this use of s as a parameter simplifies various formulas in the subsequent sections.

* A curve can have many parameterizations. The most widely used parameters are time, angle,
and arc length. For instance, if 𝒞 is a circle with radius 1 centered at the origin, then its
parameterization in terms of angle  can be <r cos  , r sin  >. If its angular velocity is w, then
its parameterization in terms of time t is <r cosw  , r sin w  >. Its arc length parameterization is
given below.
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

Example 6: Give an arc length parameterization for the circle

r (t ) = a cos ti + a sin tj

we have r (t ) = −a sin ti + a cos tj, r  r = a 2 , and therefore,

t
s(t ) =  adt * = at.
0

Hence t(s) =s/a and a representation of the circle with the arc length s as a parameter is

s s
r ( s ) = a cos i + a sin j .
a a

The circle is oriented in the counterclockwise sense, which corresponds to increase values of s.

D. Motion in Space: Velocity and Acceleration

The most important physical applications of derivatives of vector-valued functions arise in the
study of motion. As an object moves through space, the coordinates x, y, and z of its location are
functions of time. Let us assume that these functions are twice differentiable. Then we define its
position, velocity, speed, acceleration, distance and displacement as follows.
❖ Position at time t: r (t ) = x(t )i + y (t ) j + z (t )k

dr dx dy dz
❖ Velocity at time t: v(t ) = = i+ j+ k
dt dt dt dt
2 2 2
 dx   dy   dz 
❖ Speed at time t: v (t ) =   +   +  
 dt   dt   dt 

dv d 2 x d2y d 2z
❖ Acceleration at time t: a(t ) = = 2 i + 2 j+ 2 k
dt dt dt dt
t2

❖ Distance moved over time interval [t 1, t2] : S=  v(t ) dt

t1

❖ Displacement over time interval [t 1, t2] = r(t2) – r(t1)

The position vector r (t) is often called the radial vector or radius (this is why the symbol r is
used).
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

Example 7: The position vector of a point P moving on xy-plane is

r(t ) = (t 2 + t )i + t 3 j for 0  t  2

a) Find the velocity and acceleration of P at time t.

b) Sketch the path 𝒞 of the point, together with v(1) and a(1)
Solution:

a) the velocity and acceleration are given by

v(t ) = r' (t ) = (2t + 1)i + 3t 2 j
a(t ) = v' (t ) = 2i + 6tj

b) The parametric equations for C are

x = t 2 + t, y = t3

t 0 0.5 1 1.5 2

x 0 0.75 2 3.75 6

y 0 0.125 1 3.375 8

The coordinates of several points (𝑥, 𝑦) on C are listed in the above table. This gives the

sketch of C. At 𝑡 = 1 ,the point location of P is at (2,1) with

v(1) = r ' (1) = 3i + 3j and a(1) = r" (1) = 2i + 6 j

Exercise 3.2

𝑡2 2 2
1. Where is r (t) = 〈𝑡 3 − 𝑡 2 , 2
− 3 𝑡, cos (𝑡 − 3)〉 not smooth?

2. Find the position function r (t) of a moving body given that the acceleration is a (t) =
〈e−3t , t, sin t 〉 , initial velocity is v(0) = 〈4, − 2, 4〉, and initial position is r(0) = 〈0, 4, −2〉.
3. A baseball is hit from a height of 3 feet with initial speed 120 feet per second and at an
angle of 31 degrees above the horizontal. Find a vector-valued function describing the
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

position of the ball t seconds after it is hit. To be a home run, the ball must clear a wall that
is 385 feet away and 6 feet tall. Determine whether this is a home run.
4. A particle moves through 3-space in such a way that its velocity is 𝒗(𝑡) = 𝑡𝒊 + 𝑡 2 𝒋 +
𝑡 3 𝒌. If the particle’s initial position at time t = 0 is (1, 2, 3), what is the particle’s position
when t = 1?
5. Show that the velocity vector of a particle moving on a sphere is always perpendicular to
the radial vector.
6. The position function of a particle is given by r(t) = ht 2 , 5t, t2 − 16ti, t ≥ 0. When is the
speed of the particle a minimum?
7. Find the arc length of the curve described parametrically by r(t) =
𝜋
〈𝑐𝑜𝑠𝑡, 𝑠𝑖𝑛 𝑡, 𝑙𝑛 (𝑐𝑜𝑠𝑡) 〉from t = 0 to 𝑡 = .
4

8. Find the arc length parameterization of the curve described parametrically by

r(t) = 〈𝑒 𝑡 𝑐𝑜𝑠𝑡, 𝑒 𝑡 𝑠𝑖𝑛 𝑡〉 measured from 𝑡0 = 0.
2
9. Find parametric equations for the tangent line to the curve 𝒓 = 〈𝑒 𝑡 , 𝑡𝑒 𝑡 , 𝑡𝑒 𝑡 〉 at the point
(1,0,0).
10. Where on the curve 𝑟(𝑡) = 〈2𝑡, 3 sin(2𝑡) , 3cos (2𝑡)〉 are we after traveling for a distance
𝜋√10
of 3
starting from t=0?

3.3Tangents, Normals, Binormals, Curvature and Torsion

A. Tangents, Normals and Binormals
Let 𝒞 be a curve with parameterization r and P = r (t). Then the vector r / (t ) is tangent to 𝓒 at P
r (t )
and its direction is the same as the orientation of 𝒞. The unit vector along this vector is .
r (t )

We denote this vector by T and call it the unit tangent vector to 𝒞 at P. Note that if 𝒞 is a space
curve and P is a point on it, then there are infinitely many vectors normal to 𝒞 at r (P). Among
them the one which indicates the direction to which the curve bends at that point is called the
unit principal normal vector to 𝒞 at P.
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

Coming back to the case of T (t), since T has constant length and it is not 0, so the derivative T'
(t) is orthogonal to T (t) (see the corollary of page 11). When T' (t )  0 , we can divide T' (t ) by

T' (t ) to obtain a unit vector orthogonal to T(t ) , called the unit principal normal vector N:

Definition 1: Let 𝒞 be a smooth curve with parameterization (𝑡), 𝑡𝜖[𝑎, 𝑏] . Let 𝑡 be an interior
point of 𝐼 and 𝑃 = 𝑟(𝑡). The vector

′
a. ⃗⃗⃗⃗⃗⃗⃗⃗⃗ = 𝑟 (𝑡) is called the unit tangent vector to 𝒞 at 𝑃 (or at t).
𝑇(𝑡) ‖𝑟 ′ (𝑡)‖

′
b. ⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗ = 𝑇 (𝑡) is called the (principle) unit normal vector to 𝒞 at 𝑃 (or at t).
𝑁(𝑡) ‖𝑁 ′ (𝑡)‖

c. ⃗⃗⃗⃗⃗⃗⃗⃗⃗ = ⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝐵(𝑡) 𝑇(𝑡) × ⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝑁(𝑡) is called the unit bi-normal vector to 𝒞 at 𝑃 (or at t).

Example 1: Consider the circle of radius r parameterized by

r (t ) = r cos ti + r sin tj for 0  t  2 .

     
Then calculate T  , N   and B  .
3 3 3

Solution: Since

r (t ) = −r sin ti + r cos tj

we have r (t ) = (−r sin t ) 2 + (r cos t ) 2 = r.

Therefore the tangent vector is given by

r (t )
T(t ) = = − sin ti + cos tj.
r (t )

Consequently

  3 1
T  = − i + j.
3 2 2
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

Example 2: Find the parametric equations of the tangent line L to the curve
 
r (t ) = cos ti + sin tj for 0  t  2 , at the point r  .
3

  2
Solution: Since L contains r   = (r(1/2) , r( ) ) =: (x0, y0) and is parallel to the vector
3 2

  3 1 − 3 r 2 1
T  = − i + j , so the equation is X= x0 + a t = r/2 + t, Y= y0 + b t = + t,
3 2 2 2 2 2
where t is a parameter.

The three vectors T, N and B form a right handed system of mutually perpendicular vectors
called the moving trihedral.

Definition 2: Let 𝓒 be a curve and 𝑃 be a point on 𝓒. The plane determined by

a. T and N is called the osculating plane at 𝑃.

b. N and B is called the normal plane at 𝑃 and

c. T and B is called the rectifying plane.

If you are traveling along a curve 𝒞 in space, there are three important vectors at each position
of your journey: the vector representing your forward direction, the direction in which your path
is turning, N and the direction in which your motion “twists” out of the plane formed by T and N
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

Figure 13: T, N, B and their planes

B. Curvature
For a smooth curve 𝒞, ‖𝐓 (t) ‖ is always 1, but the direction can vary from point to point,
according to the bending nature of the curve. The rate of change of T with respect to arc length s
 dT 
  is related to the rate at which the curve bends.
 ds 

ΔT is zero

ΔT is small

ΔT is greatest

Mark points A and B on each of the three curves above such that the arc AB has same length in
the three cases. Then draw unit tangent vectors to each curve at each of the points A and B. For
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

the same change S in arc length you see different change in the direction of T for the three
cases. Change in the direction of T is highest for curve3, medium for curve 2 and the lowest for
curve 1. In case 1, in fact there is no change in T. Curvature of C at point P is the rate at which T
changes per unit of length along the curve at P.

Definition 3: Let 𝓒 be a curve, represented by a three times continuously differentiable vector-

valued function r of s, where s is the arc length. Then

𝜅(𝑠) = ‖𝑇 ′ (𝑠)‖ (3)

is called the curvature of 𝓒.

Remark (Alternative Formula I for curvature): If 𝒓(𝑡) is an a parameterization of 𝒞, we get

the following alternative formula:

𝑟 ′ (𝑡)
𝜅 ′ (𝑠) = ‖‖𝑟 ′ (𝑡)‖‖

dT dT
dT dT ds dT dt
=  = = dt
dt ds dt ds ds dr
dt dt

This suggests the curvature of a curve C can be given as

T
dT
 ( s) = T' ( s) = = =  (t ).
ds r

* Curvature measures the degree to which a curve bends at a point. It is the measure of the
sharpness of a curve. It measures how much a curve deviates from linearity at a point. Curvature
is a nonnegative number, it can be zero. Curvature of C doesn’t depend the parameterization
used to describe C.

Example 3: Show that the curvature of a straight line at any point is 0.

Example 4: Show that the curvature of a circle with radius r at any point is 1/r.
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

Example 5: Prove that the maximum curvature of an ellipse is at the end points of the major
axis.
Definition 4: Let  (t ) be the curvature of a smooth curve 𝒞 at P=r(t) . Then:

1
a. the number  (t ) = is called the radius of curvature at t (or at P).
 (t )

b. The circle with radius  (t ) and center on the concave side of 𝒞 and that has the same
tangent line at P as 𝒞 is called the circle of curvature. It is also called osculating circle and it
is the circle of best fit to 𝒞 at P.
c The center of this circle is called the center of curvature at t.

For eg: A car moving on a curved track: at any instant it can be thought to be moving on a circle
of radius  .

Theorem 2 (Alternative Formula II for Curvature): Let 𝑟⃗ (𝑡) = 〈𝑥(𝑡), 𝑦(𝑡), 𝑧(𝑡)〉 be a vector-
valued function that traces the smooth curve C. Then the curvature of C at the point r⃗ (t) is also
given by the formula:

‖𝑟′(𝑡) × 𝑟′′(𝑡)‖
𝜅(𝑡) =
‖𝑟 ′ (𝑡)‖3

Applications of Curvature:
- Curvature helps engineers design safe and comfortable roads, especially when dealing with
curves.
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

B. Torsion
Curvature is measure of departure from linearity and torsion is measure of departure from
planarity.

Change in the bi-normal B amounts geometrically to “twist” in the curve, from the plane of T &
N. This change in B is clearly directed towards  N and is perpendicular to T and B.
−→
−d B →
Definition 5: Let B be the bi-normal vector of a curve 𝓒 at a point P. The number  = .N
ds
is called the torsion of 𝓒 at P.

*  measures the extent to which the curve twists (from the plane of N and T) at each point in
space. If P is a train climbing up a curved track, the rate at which the head light turns from side
to side per unit distance is the curvature of the track. The rate at which the engine turns up and
down is the torsion.

Example 6: Let r (t ) = a cos ti + a sin tj + btk , be a smooth parameterization of a helix. Find

torsion of the helix.

Solution: a) If we write the parameterization of the helix in say u we have

r ' (u ) = −a sin ui + a cos uj + buk and thus

t t
s(t ) =  r' (u) du =  a 2 + b 2 = t a 2 + b 2 .
0 0

Consequently the arc length and r(s) are respectively given by

s s s
s = t a 2 + b 2 and r ( s) = a cos i + a sin j + b k , where K = a 2 + b 2
K K K

a s a s b
b) Since T( s ) = r ' ( s) = − sin i + cos j + k ,
K K K K K
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

a s a s
T' ( s ) = r" ( s ) = − 2
cos i − 2 sin j and curvature is given by the formula
K K K K
 (s) = T' (s) = r" (s) we have

a a
 ( s) = T' ( s) = T' ( s)  T' ( s) = = 2 .
K 2
a + b2

C) To find torsion first we need to find N( s ) and B' ( s) since the torsion is given by the formula
 ( s) = −N( s)  B' ( s). By formula

T' ( s ) s s
N( s ) = = − cos i − sin j and
 K K

b s b s a
B ( s ) = T( s )  N ( s ) =
sin i − cos j + k which implies
K K K K K
b s b s
B' ( s ) = 2 cos i + 2 sin j
K K K K

b b
Thus  ( s ) = − N ( s )  B' ( s ) = = 2 .
K 2
a + b2

Properties

• A plane curve with non-vanishing curvature has zero torsion at all points.
Conversely, if the torsion of a regular curve is identically zero then this curve
belongs to a fixed plane.

• The curvature and the torsion of a helix are constant. Conversely, any space curve
with constant non-zero curvature and constant torsion is a helix. The torsion is
positive for a right-handed helix and is negative for a left-handed one.

Alternative Formulas for Torsion (Read)

C. Tangential and Normal Components of velocity and acceleration

Note that if s and t are any two nonzero and nonparallel vectors, then we can resolve any
other vector x in the plane of s and t as
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

x = x ss + x t t

To see this draw the line through the tail A of x parallel to s and another line through the
head B of x parallel to t. Let these lines meet at C. Then x = ⃗⃗⃗⃗⃗⃗
𝐴𝐶 + 𝐶𝐵⃗⃗⃗⃗⃗⃗ . But 𝐴𝐶
⃗⃗⃗⃗⃗⃗ ∥ 𝒔 and 𝐶𝐵
⃗⃗⃗⃗⃗⃗
∥ t. So there exist scalars xs and xt such that x = x s s + x t t . Since the unit tangent vector T
and unit normal vector N at any point on a smooth curve C are nonparallel, so any vector b in
the plane determined by T and N can be written as:

b = b TT + b N N

Let r denotes the position vector of an object moving along C and assume that T and N exist.
Then v(t) = r/ (t) = || 𝒓′(𝑡) || T = ||𝑣(𝑡)|| T = ||𝑣(𝑡)|| + 0N

Hence, the tangential component of velocity is 𝑣𝑇 = || 𝑣(𝑡)|| and the normal component of
velocity is 𝑣𝑁 = 0.

On the other hand, the tangential and normal components of acceleration a (t) are obtained by
differentiating r/ (t).

Thus a (t) = = ‖𝑣(𝑡)‖′ ′T(t) + ‖𝐯(t) ‖ 𝑻′ (𝑡) =‖𝑣(𝑡)‖′ 𝐓(t) + ‖𝐯(t) ‖||𝑻 (𝑡)||′ N(t)

 The tangential component of acceleration and the normal component of acceleration are given
respectively by:

a T = ‖𝑣(𝑡)‖′ and a N = ||𝒗(𝑡)|| ||𝑻 (𝑡)||′

Example 7: A particle moves in the xy- plane with parametric equations

3 2
𝑥(𝑡) = 𝑡
{ 2
4
𝑦(𝑡) = 𝑡 3
3

Find the tangential and normal components of acceleration vector when t = 1.

41 12
Answer: aT (1) = 5
and aN (1) =
5

D. Centripetal and centrifugal forces (Read)

Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

Exercises 3.3

𝜋
⃗⃗⃗⃗⃗⃗⃗⃗⃗ = 𝑐𝑜𝑠𝑡 𝑖 + 𝑠𝑖𝑛𝑡 𝑗 – 𝑘 at 𝑡 = .
1. Find 𝑟′, 𝑻, 𝑵 and 𝑩 for 𝑟(𝑡)
2

⃗⃗⃗⃗⃗⃗⃗⃗⃗ = 𝑡 𝑖 + 𝑡 2 𝑗 + 𝑡 3 𝑘 at (1,1,1).
2. Find the equations of the rectifying plane 𝑟(𝑡)
3. Find the length of the curve r (t ) = e t i + e − t j + 2tk , for 0  t  1.
4. Give the arc length function parameter representation of the helix

r (t ) = 2 cos ti + 2 sin tj + 5tk , t  0.

−− 

5. Find the curvature of the trace of r (t ) = (t2 + 4) i + 2t j

at 𝑡 = 0.

6. Find the point on the graph of 𝑦 = 𝑒 𝑥 at which the curvature is maximum.

7. Find the angle between the curves 〈𝑡, 1 − 𝑡, 3 + 𝑡 2 〉 and 〈3 − 𝑡, 𝑡 − 2, 𝑡 2 〉 where they
meet. (The angle between two curves at a point is the angle between their tangent
vectors.)
8. Let C be the curve with parameterization 𝑥 = 2𝑡, 𝑦 = 𝑡 2 − 1; −1 ≤ 𝑡 ≤ 2. Find the
slopes of the ta ngent line and normal line to C at P(x, y).
9. Find a formula for the tangent T (t) to the circular helix r (t ) = 2 cos ti + 2 sin tj + 3tk .

10. Show the Frenet formula N' = B − T holds. (Hint: Differentiate N = B  T ).

3.4 Scalar Fields and Vector Fields: Gradient, Laplacian, Divergence and Curl
A. Scalar and Vector Fields: Definitions and Examples
Definition 1: A scalar field is a function that assigns a real number to every point in a certain
region 𝐷 of space. (i.e., it is a function of the form f : D  R 3 → R ). The domain 𝐷 can also be

a subset of R 2 .

For instance:
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

• Temperature at a point 𝑃 in space,

• Elevation of a point 𝑃 on earth above sea level,
• Distance between an arbitrary point 𝑃 and a fixed point P0 in space and
• Air pressure at a point 𝑃 in atmosphere
are scalar fields.
Example 1: The distance f (P) of any point P from a fixed point P0 in space is a scalar field
whose domain of definition D is the whole space.

f (P) = f ( x, y, z) = ( x − x0 ) 2 + ( y − y0 ) 2 + ( z − z 0 ) 2 ,where

x, y, z and x0 , y0 , z 0 are the coordinates of P and P0, respectively.

Example 2: The temperature T with in a body B is a scalar field, namely, the temperature field
in B. The function T may depend on time or other parameters.

Definition 2: A vector field is a function that assigns a vector to every point in a certain region
→
D of space. (i.e., it is a function of the form F : D  R 3 → R 3 ).

The following physical quantities are vector fields.

• Force Fields (gravitational field of the earth, etc)

• Velocity Fields( the velocity of a point in a flowing fluid, tangential velocity of a point on
a rotating region body, etc)
• Electric fields
• Magnetic Fields
• Gradient Fields
• Slope Fields /directional fields/ of an ODE
A 3D vector field is given by a formula of the type

𝐹⃗ (𝑥, 𝑦, 𝑧) = 𝑀(𝑥, 𝑦, 𝑧)𝒊 + 𝑁(𝑥, 𝑦, 𝑧)𝒋 + 𝑃(𝑥, 𝑦, 𝑧)𝒌

The vector field 𝑭 is said to be continuous if component functions 𝑀, 𝑁 and 𝑃 are continuous,
differentiable if M,N and P are differentiable, so on.
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

Example 3: The velocity vector 𝑉(𝑷) of a rotating body 𝐵 with angular velocity 𝜔 constitute a
vector field, the so called velocity field of the rotation. If we introduce a Cartesian coordinate
system having the origin on the axis of rotation then

V ( x, y, z ) = W  ( xi + yj + zk )

and if W points in the positive z-direction then W = k and

i j k
V = 0 0  =  (− yi + xj)
x y z

Example 4: Let a particle A of mass M be fixed at a point P0 and let a particle B of mass m be
free to take up various positions P in space. Then A attracts B. According to Newton’s law of
gravitation the corresponding gravitational force 𝑭(𝒙, 𝒚, 𝒛) directed from P to P0 and its
1
magnitude is proportional to where r is the distance between P and P0 say,
r2
GMm
F ( x, y , z ) = 2
, where G = 6.67x10-8 cm3/gm.sec2.
r

Hence 𝐹(𝑥, 𝑦, 𝑧) defines a vector field in space. If we introduce Cartesian coordinate system
such that P0 has the coordinates x0 , y0 , z 0 and P has the coordinates x, y, z then by Pythagorean
Theorem

r= ( x − x0 ) 2 + ( y − y 0 ) 2 + ( z − z 0 ) 2

assuming that r > 0 and introducing

u = ( x − x0 )i + ( y − y0 ) j + ( z − z 0 )k

 1 
we have u = r , and  − u  is a unit vector in the direction of F. The minus sign indicates that
 r 
F is directed from P to P0 thus
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

 1  GMm
F( x, y, z) = F ( x, y, z)  − u  = − 3 u
 r  r

x − x0 y − y0 z−z
= −c 3
i−c 3
j−c 3 0 k.
r r r

This vector function describes the gravitational force acting on B.

It is impossible to sketch the graph of a vector field as it requires 6 axes. But it is possible to
→
represent a vector field pictorially by drawing F ( x, y, z ) for every point (x, y, z) (by making the
→
tail of F ( x, y, z ) at (𝑥, 𝑦, 𝑧)).

Example 5: Represent 𝐹⃗ (𝑥, 𝑦) ≔ 𝑥𝒊 + 𝑦𝒋 pictorially.

Figure 14.Plot of the vector field 𝑭 = 𝒙𝒊 + 𝒚𝒋

→ → →
Example 6: Represent F ( x, y, z ) = y i − x j pictorially.
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

Figure 15: Plot of the vector field 𝑭 = 𝒚𝒊 − 𝒙𝒋

Note: Figures 12 and 13 are plotted by a free online vector field plotter
(https://www.wolframalpha.com).

Vector field illustrations are useful for visualizing and understanding various phenomena,
particularly those involving movement or forces. These illustrations can help in fields like
meteorology (e.g., visualizing wind patterns, illustrating ocean currents, etc.).

B. Gradient, Laplacian, Divergence and Curl

In this section we study four kinds of "derivatives" of scalar/vector fields, namely gradient,
divergence, Laplacian and curl. We study the definition, properties and applications (physical
significance) of each. These quantities have applications in fluid dynamics, potential theory, etc.

Definition 3(gradient): Let f be a scalar field. The gradient of f , written grad f or f , is the
vector field defined by:

f f f
f = grad f= i+ j+ k
x y z

*The operator

  
 := i+ j+ k
x y z
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie (PhD)

is a vector differential operator and it is called the del or nabla operator. (An operator is a
function which operates on (or modifies) another function, i.e., an "operator" is a function which
acts on functions to produce other functions. A vector operator assigns a vector valued function
to every function in its domain).

Example 7: For the scalar field ∅ (x, y) = x4yz, calculate

a) gradient of 𝜙 b) grad 𝜙 (−1, 1, 5

Answer: a) grad 𝜙 = 〈4𝑥 3 𝑦𝑧, 𝑥 4 𝑧, 𝑥 4 𝑦〉 and b) grad 𝜙 (−1, 1, 5) =
〈−20, 5, 1〉

Theorem 1[Algebra Rules for Gradient]: Let 𝑓 and 𝑔 be scalar fields and k be a
constant. Then

a. Constant multiple rule: (kf ) = kf

b. Sum rule: ( f + g ) = f + g
c. Difference Rule: ( f + g ) = f − g
d. Product Rule: ( f + g ) = gf + fg
f gf − fg
e. Quotient Rule: ( g ) = g2

Definition 4(Laplacian): Let 𝑓 be a scalar field.

2 f 2 f 2 f
a) The expression  f = f =
2
+ + is called the Laplacian of 𝑓.
x 2  y 2 z 2

2 f 2 f 2 f
b) The PDE + + =0 is called 𝑳𝒂𝒑𝒍𝒂𝒄𝒆’𝒔 𝒆𝒒𝒖𝒂𝒕𝒊𝒐𝒏.
x 2  y 2 z 2
c) A scalar field satisfying Laplace’s equation is called harmonic.

Physical Significance of Gradient:

Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie(PhD)

The gradient of a scalar field at a point is a vector whose magnitude is equal to the
maximum rate of change of the scalar field at that point and whose direction shows the
direction for maximum rate of increase. For example, consider a point on earth. You have
infinitely many directions to go from that point on the surface of the earth. The steepest
direction is the direction of the gradient (at that point) of the elevation above sea level
function.

Flux is the amount (number, mass, volume, etc) of something/field/ crossing a surface. It
depends on field strength, area of the surface and angle between the surface and the field.
Divergence is flux per unit area; i.e., it is flux density at a point.

Definition 5 (divergence ): Let F( x, y, z ) = M ( x, y, z )i + N ( x, y, z ) j + P( x, y, z )k be a

M N P
vector field such that , , and exist, then the divergence of F, denoted div F or
x y z
  F , is the function defined by

M N P
divF =   F( x, y, z ) = ( x, y, z ) + ( x, y , z ) + ( x, y, z ). (4)
x y z

Example 8: If F is a constant vector field, then 𝑑𝑖𝑣 𝑭 = 0.

Example 9: If 𝑭 = < 𝑥, 𝑦 >, then 𝑑𝑖𝑣 𝑭 = 2.
Example 10: If 𝑭 = < −𝑥, −𝑦 >, then 𝑑𝑖𝑣 𝑭 = −2.
Example 11: Given F = ( x 2 y 3 − z 4 )i + 4 x 5 y 2 zj − y 4 z 6 k , find 𝑑𝑖𝑣 𝑭, 𝑑𝑖𝑣 𝐹 (2, 2, 0) and
𝑑𝑖𝑣 (0, 2, 2) .

Solution:
 2 3  
divF =   F = ( x y − z 4 )i + (4 x 5 y 2 ) j + (− y 5 z 6 )k
x y z
= 2 xy3 + 8 x 5 yz − 6 y 4 z 5 .

𝑑𝑖𝑣 𝐹 (1, 2, 0) = 32 and

𝑑𝑖𝑣 (0, 2, 1) = −96.

41
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie(PhD)

Theorem 2[properties of Divergence]: Let 𝑭 and 𝑮 be vector fields, 𝑓 be a scalar

filed and 𝑘 be a constant. Then

a) div (kF) = k div F

b) div(F + G) = div F + div G
c) div(F - G)= div(F) – Div(G)
d) div(fF) = grad(f)F + f div(F)
e) div(F×G) = curl F. G – F. curl F

From (a) and (b) we see that divergences a linear function.

Physical Geometrical Interpretation/Significance of Curl:

Divergence measures how much a vector field diverges (spreads away) from a point. If
div (P) is positive, then vectors outgoing from a small area around P are larger than
vectors going inwards.

Circulation is the amount of force that pushes along a closed path. Curl is circulation per
unit area. It is microscopic circulation; i.e., circulation at a point.

Definition 6 (curl): The curl of a vector field F( x, y, z ) = Mi + Nj + Pk is the vector field

 P N   M P   N M 
curlF =   F =  − i +  −  j +  − k.
 y z   z x   x y 

The formula for curl F can be written in the following determinant form.

i j k
  
curlF =   F =
x y z
M N P

* Curl is also called rotor and curl (F) is also denoted by rot F.

42
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie(PhD)

Example 12 (positive/ACCW curl): For the vector field 𝐺(𝑥, 𝑦) = {−𝑦, 𝑥}, div
𝐺(𝑥, 𝑦) = 0 𝑎𝑛𝑑 curl 𝐺(𝑥, 𝑦) = 2𝒌.

Example 13(negative/CCW curl): For the vector field G(x, y) = {y, -x}, div G(x, y) =
0 and curl G(x, y) = - 2k.

Example 14: If = < 𝑥, 𝑥 >, 𝑐𝑢𝑟𝑙 𝐹 = 𝑘.

Example 15: If F( x, y, z ) = xy 2 z 4 i + (2 x 2 y + z ) j + y 3 z 2 k , find   F.

Solution: Applying (1), we have

i j k
  
F =
x y z
xy 2 z 4 (2 x y + z )
2
y3z2

= (3 y 2 z 2 − 1)i + 4 xy 2 z 3 j + (4 xy − 2 xyz 4 )k.

Theorem 3 [properties of curl]: Let 𝐹 and 𝑮 be differentiable vector fields, 𝑓 be a

differentiable scalar field and k be a constant.

a) Curl (kF) = k curl (F)

b) Curl (F + G) = curl (F) + curl (G)

c) Curl (F - G) = curl (F) - curl (G)

d) Curl (fF) = f curl (F) + grad (f) ×F

e) Curl (F×G) = F div G – G div F + (G.𝛻)F – (F.𝛻)G

From (a) and (b) we see that curl is a linear function.

Properties of Combinations of Gradient, Divergence and Curl: Higher Order

Derivatives

43
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie(PhD)

a) 𝑐𝑢𝑟𝑙 (𝑔𝑟𝑎𝑑 𝑓) =0 , for any twice continuously differentiable 𝑓: ℝ3 → ℝ.

b) 𝑑𝑖𝑣 (𝑐𝑢𝑟𝑙 𝑭) = 0, for any vector field F.
c) 𝑑𝑖𝑣 (𝑔𝑟𝑎𝑑 𝑓) = ∇2 𝑓
Definition 7: A vector field F is said to be
a. solenoidal if 𝒅𝒊𝒗 (𝑭) = 0
b. irrotational if 𝒄𝒖𝒓𝒍 (𝑭) = 0
c. conservative if there exists a scalar field 𝒇 such that 𝒈𝒓𝒂𝒅 (𝒇) = F. In this case 𝒇
is called a potential function for 𝑭.

Theorem 4 [The Curl Criterion for a Conservative Vector Field]:

a. Any conservative vector field F: 𝑈 → ℝ has zero curl.

b. The converse of (a) is true if U is simply connected domain.

Theorem 5 (Helmholtz Decomposition Theorem): Any vector field on can be expressed

as the sum of a conservative vector field and a solenoidal vector field.

Example 16: For what values of 𝑎 and 𝑏 the vector field 𝐹⃗ (𝑥, 𝑦, 𝑧) = (𝑥 + 𝑧)𝑖 +
𝑎(𝑦 + 𝑧)𝑗 + 𝑏(𝑥 + 𝑦)𝑘 is conservative? Answer: 𝑎 = 𝑏 = 1.

Note: Gradient, Laplacian, divergence and curl can be defined in general coordinates.
The reader can refer reference books for the definition of these vector derivatives in
curvilinear coordinates (cylindrical, spherical, parabolic, elliptic, etc.).

Exercise 3.4

1. Compute the divergence of 𝑮(𝑥, 𝑦, 𝑧) = 2𝑥 3 𝒊 − 3𝑥𝑦𝒋 + 3𝑥 2 𝑧𝒌? Answer: The

divergence of the vector field 9𝑥 2 − 3𝑥
2. Compute the curl of 𝑭(𝑥, 𝑦, 𝑧) = 𝑥 2 𝒊 + 𝑥𝑦𝑧𝒋 − 𝑧𝒌 at the point (2, 1, −2).
3. Prove that 𝑑𝑖𝑣(𝛻𝑓 × 𝛻𝑔) = 0.

44
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie(PhD)

4. Select the irrotational vector field (i.e., whose curl is zero)

(𝑎)𝑦𝑧𝑖 − 2𝑥𝑧𝑗 + 𝑥𝑦𝑧𝑘 , (𝑏)𝑦𝑧𝑖 + 𝑥𝑧𝑗 + 𝑥𝑧𝑘 ,
(𝑐)𝑧𝑖 – 𝑧 2 𝑗 + 𝑦𝑧𝑘 , (𝑑) 𝑦𝑖 + (𝑥 − 𝑧)𝑗 − 𝑦𝑘 .
5. For what values of 𝑏 and 𝑐 will 𝑭 = (𝑦 2 + 2𝑐𝑧𝑥)𝒊 + 𝑦(𝑏𝑥 + 𝑐𝑧)𝒋 + (𝑦 2 +
𝑐𝑥 2 )𝒌 is the gradient of some scalar field𝑓?
Answer: The values of b and c are both 2.
6. Find 𝑎 and b so that 𝐹⃗ (𝑥, 𝑦, 𝑧) = 〈𝑥 + 2𝑦 + 4𝑧, 2𝑎𝑥 + 𝑏𝑦 − 𝑧, 4𝑥 − 𝑦 + 2𝑧〉 is both
solenoidal and irrotational.
Answer: 𝑎 = 1 and 𝑏 = −3
7. Is the vector field 𝑭 = (𝑥, 𝑦, 5𝑥)conservative?
Answer: The curl is (0, −5,0) ≠ (0,0,0). Therefore the vector field is not conservative.
−𝑦 𝑥
8. Let 𝑭(𝑥, 𝑦) =(𝑥 2+𝑦 2 , ). Show that 𝑐𝑢𝑟𝑙(𝑭) = 0 but 𝑓 is non-conservative. Explain
𝑥 2 +𝑦 2

why this does not contradict the curl criterion for conservativeness.
9. Find a potential function 𝑓 for the field 𝑭 = 5𝑥 𝒊 + 8𝑦 𝒋 + 5𝑧 𝒌.
5
Answer: A potential function for the given vector field is 𝑓(𝑥, 𝑦, 𝑧) = 𝑥 2 + 4𝑦 2 +
2
5 2
𝑧 +𝐶
2

45
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie(PhD)

Chapter 4: Line Integrals and surface Integrals

In this chapter we consider two generalizations of definite integrals (namely line

integrals and surface integrals) and the four main theorems of vector integral calculus
(namely the Fundamental theorem of Line Integrals, Green’s Theorem, Divergence
Theorem and Stokes’s Theorem). These theorems are generalizations of the Fundamental
theorem of Calculus. From the outset, the author lets you know that unlike the integrals
you have learnt before (single integral, double integral and triple integrals); there are
several notations for line integrals and surface integrals.

Line Integrals Surface Integrals

- ∫𝐶 𝑓(𝑥, 𝑦, 𝑧)𝑑𝑠 (C is a curve,) - ∬𝑺 𝒇(𝒙, 𝒚, 𝒛)𝒅𝒔 (S is a surface, not necessarily

- ∫𝐶 𝐹⃗ (𝑥, 𝑦, 𝑧). ⃗⃗⃗⃗⃗

𝑑𝑟 a plane region)

- ∫𝐶 𝐹⃗ (𝑥, 𝑦, 𝑧). 𝑇
⃗⃗ ds - ∬𝑺 𝐹⃗ (𝑥, 𝑦, 𝑧). ⃗⃗⃗⃗⃗
𝒅𝒔

- ∫𝐶 𝐹⃗ (𝑥, 𝑦, 𝑧). 𝑛⃗⃗ ds - ∬𝑺 𝐹⃗ (𝑥, 𝑦, 𝑧). ⃗𝒏

⃗⃗ ds

𝐵(𝑏 ,𝑏 ,𝑏 )
- ∫𝐴(𝑎 1,𝑎 2,𝑎 3) 𝐹⃗ (𝑥, 𝑦, 𝑧). ⃗⃗⃗⃗⃗
𝑑𝑟
1 2 3

- ∮𝐶 𝑓(𝑥, 𝑦, 𝑧) 𝑑𝑠

- ∳ 𝐹⃗(𝑥, 𝑦, 𝑧). 𝑑𝑟
⃗⃗
C
(contour integral in positive sense)

- ∲𝐹⃗(𝑥, 𝑦, 𝑧). 𝑑𝑟
⃗⃗⃗⃗⃗

C
( contour integral in negative sense)
- ∫𝐶 𝐹⃗ (𝐫(𝒕))𝑑𝑡 (line integral whose value
is a vector)

46
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie(PhD)

In branches of mathematics such as differential geometry, Riemannian geometry and

tensor calculus there is, even, a generalization of surface integration to higher
dimensional surfaces (manifolds). This is called surface integration over manifolds.

4.1 Line integrals

Line integrals are generalizations of definite integrals. Definite integrals are
defined over intervals (straight curves) but line integrals can be defined over non-
straight curves. A line integral is also called curve integral, path integral, or
curvilinear integral. In some cases it is also called contour integral (circulation
integral). Curve integral would have been the most appropriate term, but line integral
is more widely used.

Line integrals have various forms:

a) line integrals of scalar fields with respect to arc length:

C
 f ( x, y, z)ds
b) line integrals of scalar fields with respect to 𝑥, 𝑦 and 𝑧 (aka line integrals in
differential forms):

 Mdx ,  Ndy ,  Pdz ,  Mdx + Ndy + Pdz

C C C C

→ →
c) line integrals of vector fields (work integral)  F ( x, y, z). dr
C

→ →
It can also be written as  F ( x, y, z ). T ds
C

47
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie(PhD)

4.1.1 Line integrals of Scalar Fields with respect to Arc Length

A. Motivation: mass of a wire (or a thin rod)

Consider a piece of wire having the shape of a smooth curve 𝒞 with end points 𝐴 and 𝐵,
lying in xyz-space. Let 𝑓(𝑥, 𝑦, 𝑧) be the mass per unit length (length density) of the wire
at the point 𝑃(𝑥, 𝑦, 𝑧) on 𝒞. What is the total mass of the wire?

̂𝑃𝑘 be the portion of 𝒞 between

Let 𝒫: = {𝑃1 , … , 𝑃𝑛 }be a partition of 𝒞 as shown. Let 𝑃𝑘−1
the points 𝑃𝑘−1 and 𝑃𝑘 (we call this the kth sub-curve). Let ∆𝑆𝑘 be the length of the kth
sub-curve. Let (𝑥𝑘∗ , 𝑦𝑘∗ , 𝑧𝑘∗ ) be a point on the kth sub-curve. The mass ∆𝑚𝑘 of the portion
of the wire between Pk-1 and Pk is approximated by

∆𝑚𝑘 ≈ 𝑓(𝑥𝑘∗ , 𝑦𝑘∗ , 𝑧𝑘∗ ) ∆𝑆𝑘 (1)

Thus the total mass 𝑚 of the wire is approximated by

𝑚 ≈ ∑𝑛𝑘=0 𝑓(𝑥𝑘∗ , 𝑦𝑘∗ , 𝑧𝑘∗ ) ∆𝑆𝑘 =: 𝐽𝑛 (2)

If the length || 𝒫|| of the longest sub curve approaches 0 as n tends to infinity, then the
total mass m of the wire is

𝑚 = lim ∑𝑛𝑘=0 f(xk∗ , yk∗ , zk∗ ) ∆Sk (3)

𝑛→∞

The limit on the RHS of the above equation can represent many other quantities (not only
mass).We call it the line integral of 𝑓 along C from A to B with respect to arc length.

48
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie(PhD)

B. Definition

Definition 1: Let 𝒞 be a smooth curve of finite length (with end points A and B and
oriented from A to B). Let 𝑓 be a scalar field defined at each point of 𝒞. Let 𝒫 ≔
{𝑃1 , 𝑃2 … 𝑃𝑛 } be a partition of 𝒞 as shown in the figure. Then the line integral of 𝑓 along
𝒞 with respect to arc length s is denoted and defined by

∫𝒞 𝑓(𝑥, 𝑦, 𝑧)𝑑𝑠 = lim ∑𝑛𝑘=1 𝑓(𝑥𝑘 , 𝑦𝑘 , 𝑧𝑘 ) ∆𝑆𝑘 (4)

𝑛→∞

where ∆𝑆𝑘 is the length of the portion of C between Pi-1 and P i,

(𝑥𝑘 , 𝑦𝑘 , 𝑧𝑘 ) is any point on 𝒞 between 𝑃𝑘−1 & 𝑃𝑘 i and

Provided that ||𝛥𝑆|| : = max {∆𝑆1,… ∆𝑆𝑛 } tends to 0 as n → 

Figure 1: A Partition of a Finite smooth curve

* NB. n →  doesn’t imply every S i → 0 . But the converse is true. Moreover ∆𝑠𝑘

→ 0 bi-implies ||P|| tends to 0.

49
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie(PhD)

C. Evaluation:

The basic idea is to convert into definite integral. If C is parameterized by a

differentiable vector valued function

r(t) = x(t)i + y(t)j + z(t)k, a ≤ t ≤ b

𝑑𝑠 = ‖𝑟 ′ (𝑡)‖𝑑𝑡 = √[𝑥 ′ (𝑡)]2 + [𝑦 ′ (𝑡)]2 + [𝑧 ′ (𝑡)]2

Thus

∫ 𝑓(𝑥, 𝑦, 𝑧)𝑑𝑠 = ∫ 𝑓(𝑥(𝑡), 𝑦(𝑡), 𝑧(𝑡))√[𝑥 ′ (𝑡)]2 + [𝑦 ′ (𝑡)]2 + [𝑧 ′ (𝑡)]2 𝑑𝑡

𝒞 𝑎

Note: If r(t) = x(t)i + y(t)j + z(t)k, a ≤ t ≤ b is a parameterization of C then, the

partition 𝒫 of C induces a partition {t0,…,tn} of [a, b] and since ∆sk ≈ ||∆rk||≈

r ' (t k ) t k so
*

n n

 f ( x k , y k , z k )s k ≈  f (r (t k )) r ' (t k ) t k
* * * * *

k =1 k =1

The sum in RHS is the Riemann sum of the function g(t): = f (r (t )) r ' (t ) corresponding

to the partition {t0,…,tn} of [a, b].

n n

 f ( x k , y k , z k )s k = lim  f (r (t k )) r ' (t k ) t k

* * * * *
But lim
n → n →
k =1 k =1

Hence


C
f ds =  f (r (t )) r ' (t ) dt
a

50
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie(PhD)

Example-1: Evaluate

 f ds
C

Where f = xy4 and C is the right half of the circle x2 + y2 = 16 and have counter
clockwise orientation.

8192
Ans.
5

Question: What if the orientation is ACW?

D. Properties

a)  kf = k  f
C C

b)  f g =  f g
C C C

c) f =f+f
C C1 C2
(where C= C1 U C2 and C is smooth)

Further properties:

- If f is continuous at every point on C and if C is smooth, then f

C
exist.

- f
C
is parameterization independent , i.e. any smooth parameterization of C

yields the same value of  f as long as the parameterization of the curve C is

C

traced out exactly once as t increases from a to b.

51
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie(PhD)

- f
C
is orientation independent, i.e., if –C is the same curve as C but with opposite

orientation, then

 fds =  fds
−C C

In view of property (c) above we make the following definition.

Definition 2 (line integral over piecewise smooth curve): Let 𝒞 be a piecewise smooth
curve with smooth pieces 𝒞 1,…, 𝒞 n. We define

 f = f
C C1
+…+  f
Cn

Example-2 (three dimensional): Evaluate

 xyzds
C

Where C is the helix given by 𝒓(𝑡) : = < 𝑐𝑜𝑠𝑡, 𝑠𝑖𝑛𝑡, 3𝑡 >, 0 ≤ 𝑡 ≤ 4𝜋. Ans. -3√10

Example-3(piecewise smooth): Evaluate  f ds

C

Where 𝑓 = 4𝑥 3 and C is the union of the following curves.

C1: the line segment with end points (-2, -1) and (0, -1)

C2: part of the curve y = x3 - 1 between (0, -1) and (1, 0)

C3: the line segment with end points (1, 0) and (1, 2)

Ans. -8 + 2/27 (103/2 - 1)

Example 4 (application): Find the mass of the piece of wire described by the curve
x2+y2=1 with density function 𝑓(𝑥, 𝑦) = 3 + 𝑥 + 𝑦.

52
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie(PhD)

Solution: The curve is the standard unit circle. It can be parameterized by the vector
function 𝒓(𝑡) =< 𝑐𝑜𝑠 (𝑡), 𝑠𝑖𝑛(𝑡) > with 0 ≤ 𝑡 ≤ 𝜋. We have 𝑥 (𝑡) = 𝑐𝑜𝑠(𝑡) and
y(t)=sin(t), so x'(t)=-sin(t) and 𝑦′(𝑡) = 𝑐𝑜𝑠 (𝑡). The mass is given by the formula

The term in the square root is 1, hence we have

4.1.2 Line Integrals in Differential Form

Definition 3: Let C be a smooth curve (oriented from A to B) of finite length. Let 𝑓 be a

scalar field defined at each point of C. Let {P1, P2… P n} be a partition of C and ⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝑃𝑖−1 𝑃𝑖 =
( xi , y i , z i ). Then we define the line integrals of f(x, y, z) along C from point A to
point B with respect to x , y and z as:
n

 f ( x, y, z )dx = lim
n →
 f ( x , y , z )x
1
i i i i
C

 f ( x, y, z )dy = lim
n →
 f ( x , y , z )y
1
i i i i
C

 f ( x, y, z )dy = lim
n →
 f ( x , y , z )z
1
i i i i
C

The three line integrals often come together and give a line integral in differential form.

Definition 4: Let M, N and P be continuous functions of x, y and z. Then we define:

 M dx + Ndy + Pdz =  Mdx +  Ndy +  Pdz

C C C C

53
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie(PhD)

Evaluation: The line integrals above are evaluated by expressing 𝑥, 𝑦, 𝑧, 𝑑𝑥, 𝑑𝑦, 𝑑𝑧 in
terms of 𝑡 where 𝑡 is the parameter in a suitable parameterization of 𝒞. The result is an
ordinary single integral.

Remark [orientation dependence]: There is one important difference between the

integral in differential form and the line integral of previous subsection. If – C is the same
curve as C, but with opposite orientation then

 M dx + Ndy + Pdz = -  M dx + Ndy + Pdz =

−C C

Reason: The integral above changes with changes in 𝑥(𝑡), 𝑦(𝑡) and 𝑧(𝑡). Why?

But changes in x(t),y(t) and z(t) doesn’t affect the arc length element ds. Why?

Example 5: Evaluate  (3 + x + y)dy, where C is the standard unit circle oriented

C

a) Counter clockwise (ccw). Ans. 𝜋

b) Clockwise (cw). Ans. −𝜋

Example 6: Find ∫𝐶 𝑦𝑑𝑥 + 𝑧𝑑𝑦 where C is the part of the helix

𝒓(𝑡) = 𝑠𝑖𝑛 𝑡 𝒊 + 𝑐𝑜𝑠 𝑡 𝒋 + 𝑡𝒌 ; 0 < 𝑡 < 2𝜋

Solution: We have 𝒓′(𝑡) = 𝑐𝑜𝑠 𝑡 𝒊 − 𝑠𝑖𝑛 𝑡 𝒋 + 𝒌 so that 𝑦 𝑑𝑥 +

𝑧 𝑑𝑦 = (𝑐𝑜𝑠2𝑡 − 𝑡 𝑠𝑖𝑛 𝑡)𝑑𝑡

2𝜋
This leads us to the integral ∫0 (𝑐𝑜𝑠 2 𝑡 − 𝑡𝑠𝑖𝑛𝑡)𝑑𝑡 with a little bit of effort (using
integration by parts) we get −𝜋 .

Example 7: Evaluate ∫𝐶 𝑦𝑑𝑥 + 𝑧𝑑𝑦 + 𝑥𝑑𝑧 where C has the parameterization x = t, y = t2

, z = t3; 0 ≤ t ≤1.

Ans. 89/60

54
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie(PhD)

Example 8: Evaluate ∫𝒞 𝑥𝑑𝑥 + 𝑦𝑑𝑦 + 𝑧 2 𝑑𝑧 when 𝒞 is the helix parameterized by

𝒓( 𝑡) = 𝑐𝑜𝑠( 𝑡) 𝒊 + 𝑠𝑖𝑛( 𝑡)𝒋 + 𝑡 𝒌 for t in[ 0,2𝜋] .

8𝜋 3
Ans. 3

4.1.3 Line Integrals of Vector Fields(with respect to displacement)

A. Motivation (Work done by a variable force )

Let 𝐹⃗ (𝑥, 𝑦, 𝑧) ∶ = 𝑀 𝒊 + 𝑁𝒋 + 𝑃 𝒌 be a force field defined at each point of a region in

xyz-space and let 𝒞 be a curve in this space with end points A and B. What is the total
work done by F in moving a point from A to B along 𝒞?

Note that if a constant force F moves an object through any vector displacement S, then
the work done is given by:

𝑾 = 𝑭. 𝑺 = ‖𝑭‖‖𝑺‖ 𝐜𝐨𝐬 𝜽

Figure 2: Work done by a constant force F

If a variable force moves an object through a small vector displacement ∆r, then the
small amount of work ∆w done by the force is approximately the product of the distance
traveled ‖∆𝒓‖and the cosine component of the force. (Because the force is approximately
constant on small displacement)

This implies that ∆w ≅ ‖𝑭‖‖𝑺‖ 𝐜𝐨𝐬 𝜽 = 𝑭(𝑥, 𝑦) · ∆𝒓

Subdivide 𝒞 into n small sections as shown. Let 𝒫: = {P1, …, Pn}be a partition of 𝒞 as

⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
shown. Let ∆r k: =P ∗ ∗ ∗ ̂
k−1 Pk . Let (𝑥𝑘 , 𝑦𝑘 , 𝑧𝑘 ) be a point on Pk−1 Pk ,i.e., arc Pk−1 Pk . The

work ∆Wk done by F in moving the point from Pk-1 to 𝑃𝑘 along 𝒞 is approximated by

55
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie(PhD)

∆𝑊𝑘 ≈ F (𝑥𝑘∗ , 𝑦𝑘∗ , 𝑧𝑘∗ ) ∆𝒓𝑘

Thus the total work W is approximated by

𝑊 ≈ ∑𝑛𝑘=0 𝐅 (𝑥𝑘∗ , 𝑦𝑘∗ , 𝑧𝑘∗ ) ∙ ∆𝒓𝑘 =: 𝐽𝑛

If the length ‖𝒫‖of the longest sub arc approaches 0 as n tends to infinity, the total work
W done by F is given by:

𝑊 = lim ∑𝑛𝑘=0 𝐅(xk∗ , yk∗ , zk∗ ) . ∆𝐫𝐤

𝑛→∞

B. Definition:

Definition 5: Let 𝒞 be a smooth curve of finite length with end points A and B and
oriented from A to B. Let F be a vector field defined on 𝒞. Let 𝒫 ={P1, P2… P n} be a
̂𝑃𝑘 be the portion of 𝒞 between Pk-1 and
partition of 𝒞 as shown in the figure 1. Let 𝑃𝑘−1
𝑃𝑘 . Then the line integral of F along 𝒞 (from A to B with respect to displacement)
is denoted and defined by:

n
∫𝒞 𝑭. 𝒅𝒓 = lim
n→
 F (x , y
k =1
k k , z k )  rk ;

where rk := P
⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗ ̂
k−1 Pk and (𝑥𝑘 , 𝑦𝑘 , 𝑧𝑘 ) 𝜖 𝑃𝑘−1 𝑃𝑘 (provided that ‖𝒫‖ tends to 0 as n →  ).

NB. Both F and 𝒅𝒓 are vectors.

C. Evaluation
If 𝒓(𝑡), 𝑎 ≤ 𝑡 ≤ 𝑏 is a parameterization of 𝒞, then the line integral of F along 𝒞 is given
by:

56
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie(PhD)

b
∫𝒞 𝑭. 𝒅𝒓 =  F (r (t ))  r ' (t )dt
a

If 𝑭 = 𝑀 𝑖 + 𝑁 𝑗 + 𝑃 𝑘 is a vector field and 𝒞 is the curve with parameterization 𝒓 =

𝑥 𝑖 + 𝑦 𝑗 + 𝑧𝑘, t in [𝑎, 𝑏], then  F.dr =  Mdx + Ndy + Pdz

C C

(i.e., the line integral above and the line integral in differential form are equivalent).

Note that this gives us another method for evaluating line integrals of vector fields.

Justification:

We can rewrite r'(t) dt as

𝑑𝑟 𝑑𝑥 𝑑𝑦 𝑑𝑧
𝑑𝑡
𝑑𝑡 = [ 𝑑𝑡 𝑖 + 𝑑𝑡 𝑗 + 𝑑𝑡 𝑘] 𝑑= 𝑑𝑥 𝒊 + 𝑑𝑦𝒋 + 𝑑𝑧𝒌

So that if 𝑭 = 𝑀𝒊 + 𝑁𝒋 + 𝑃𝒌, then 𝑭. 𝒅𝒓 = 𝑭. 𝒓′(𝑡)𝑑𝑡 = 𝑀 𝑑𝑥 + 𝑁 𝑑𝑦 + 𝑃 𝑑𝑧.

Definition 6 (Line Integral of a vector field over piecewise smooth curve): Let C be a
piecewise smooth curve with smooth pieces C1,…, Cn. We define

 F.dr =  F
C C1
+…+  F
Cn

D. Properties

a) F = −F
C −C
(Where C and –C are same curve but have different

orientations)

57
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie(PhD)

b)  kf = k  f
C C

c)  f g =  f g
C C C

Example 9: Evaluate  F.dr where 𝑭(𝑥, 𝑦, 𝑧) = 𝑥𝑧 𝑖– 𝑦𝑧 k and C is the line segment

C

from

a) (−1, 2, 0) to (3, 0, 1). Ans. 3

b) (3, 0, 1) to (−1, 2, 0). Ans. −3

Example 10(application): Find the total work done in moving an object along the helix
C given by 𝒓( 𝑡) = 𝑐𝑜𝑠( 𝑡) 𝒊 + 𝑠𝑖𝑛( 𝑡)𝒋 + 𝑡 𝒌 for t in [ 0,2𝜋] through the vector field
𝑭 ( 𝑥, 𝑦, 𝑧) = < 𝑦, 𝑧 2 , 𝑥 2 >.

Answer: 2𝜋 + 4𝜋 2

Exercise 4:1

1. Several paths with same end points. Evaluate the integral ∫𝒞 𝑭. 𝒅𝒓 where 𝑭 =
⟨𝑥𝑦,−𝑦 2 ⟩ and : 𝒓 = 〈𝑡, 𝑡 𝑛 〉, 0 ≤ 𝑡 ≤ 1, where 𝑛 = 1,2,3, …
2.

4.2 Path independence of Line Integrals: The Fundamental theorem of Line

Integrals
A. Path Dependence of line integrals

58
Applied Math. III Part II: Vector Calculus Compiled by: Tadesse Bekeshie(PhD)

Let 𝐷 be a domain of the vector field F. Let A and B be points in 𝐷, 𝐶 be a curve in 𝐷

joining A to B. Then the value of the line integral  F (r ).dr generally depends on not only
C

on the end points A and B, but also on the curves joining A and B. However there are
exceptions

For example: consider  ydx + xdy (Where C is any curve joining (0, 0) to (1, 1).
C

Consider the 4 paths in the diagram.

1 1

1
1
1

1
Figure 18: Four different paths from (𝟎, 𝟎) to (𝟏, 𝟏)
1

The given line integral is the same along all paths joining (0,0) to (1, 1).

To proceed we need to recall the definitions of some topological notions about curves and
regions.

A path (also called a curve) 𝒞 with parameterization 𝒓(𝑡), 𝑎 ≤ 𝑡 ≤ 𝑏 is said to be:

a) closed if its initial and terminal points coincide (i.e., if 𝑟⃗(𝑎) = 𝑟⃗(b) ) and if 𝒓 is one
to one on open interval (𝑎, 𝑏) except possibly at finitely many points.
b) simple if does not touch or cross itself on the open interval (𝑎, 𝑏).

A (plane or space) region 𝐷 is said to be

59
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

a) open if it does not contain any of its "boundary" points.

b) connected if every two points of D can be joined by a finite number of line segments
that are joined end to end and whose points belong to D.
c) simply connected if it is a domain and if every closed curve in D can be continuously
shrunk to any point in D without leaving D.. (Read also doubly connected, triply
connected and multiply connected.)
Example: The region between two concentric spheres is simply connected but the
region between two concentric circles is not simply connected. Similarly, torus
(doughnut shaped 3D figure) is not simply connected.

Definition 1: Let 𝐷 be a connected region in R3. The line integral  F (r ).dr is said to be path
C

independent in D if  F (r ).dr =  F (r ).dr , for any two paths C1 and C2 in 𝐷 with the same
C1 C2

initial and terminal points.

Figure 4: Path Independence of Line Integrals

B. The Fundamental theorem of Line Integrals

Theorem 1[The Fundamental Theorem of Line Integrals]

60
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

Let C be an oriented curve with initial point A(𝑎1 , 𝑎2 , 𝑎3 ) and terminal point 𝐵(𝑏1 , 𝑏2 , 𝑏3 ).
Let 𝑓 be a function of three variables that is differentiable at every point on C, and assume

that grad 𝑓 is continuous on C. Then  gradf .dr = f ( B) − f ( A)

C

B
Notation: we write  f (r ).dr =  f  dr
C A

Remark: The Fundamental Theorem of Line Integrals says that if 𝑭 is the gradient of a
differentiable function 𝑓 and has domain D, then the value of a line integral ∫𝐶 𝑭. 𝑑𝑟 depends
only on the initial and terminal points of the oriented curve 𝐶 in 𝐷.

Proof:

Part I. Assume that 𝑭 = 𝑔𝑟𝑎𝑑 𝑓 and C is a path from 𝐴 to 𝐵 in D.

Let 𝒓(𝑡) = < 𝑥(𝑡), 𝑦(𝑡), 𝑧(𝑡) > be a parameterization of C.

 gradf .dr =   f
C C
x , f y , f z    dx, dy, dz 

= f
C
x dx + f y dy + f z dz

 f x x' (t )dt + f y y' (t )dt + f z z ' (t )dt

b
=
a

 [ f x x' (t ) + f y y' (t ) + f z z ' (t )]dt

b
=
a

d

b
= [ f ( x(t ), y(t ), z (t ))]dt [chain rule]
a
dt

= 𝑓(𝑥(𝑏), 𝑦(𝑏), 𝑧(𝑏)) − 𝑓(𝑥(𝑎), 𝑦(𝑎), 𝑧(𝑎))

= 𝑓(𝐵) – 𝑓(𝐴)

61
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

Part II. The proof of the converse is difficult; hence it is omitted.

C. Conditions for Path Independence of Line Integrals

If, for a vector field F, there exists a scalar field 𝑓 satisfying 𝑭 = 𝑔𝑟𝑎𝑑 𝑓, then 𝑭 is called a
conservative vector field and 𝑓 is called a potential function of 𝑭.

Definition 2:

a) Let 𝑀, 𝑁 and 𝑃 be scalar fields. An expression of the type

𝑀 𝑑𝑥 + 𝑁𝑑𝑦 + 𝑃 𝑑𝑧

is called a differential form or a Pfaffian form.

b) If in particular, there is a scalar field 𝑓 such that 𝑀 = 𝑓𝑥 , 𝑁 = 𝑓𝑦 , 𝑃 = 𝑓𝑧 , then
the expression in (a) is called exact differential form.

We now define some topological notions which we need in the sequel.

Theorem 2[A Test for exactness of a differential form and conservativeness of a vector field]

Let 𝑭 = 𝑀𝒊 + 𝑁𝒋 + 𝑃𝒌 be a vector field whose components have continuous first order partial
derivatives. If the domain of 𝐹 is simply connected, then 𝑀𝑑𝑥 + 𝑁𝑑𝑦 + 𝑃𝑑𝑧 is exact (i.e., F is
conservative) if and only if 𝑀𝑦 = 𝑁𝑥 , 𝑁𝑧 = 𝑃𝑦 , 𝑀𝑦 = 𝑃𝑥

Proof:

Part I. If F is conservative

There exists a scalar field 𝑓 such that 𝐹 = < 𝑓𝑥, 𝑓𝑦, 𝑓𝑧 >

This implies that 𝑀 = 𝑓𝑥 , 𝑁 = 𝑓𝑦 and 𝑃 = 𝑓𝑧 . Hence 𝑀𝑦 = 𝑓𝑥𝑦 = 𝑓𝑦𝑥 = 𝑁𝑥 .

The rest can be proved analogously.

Part II. This requires Stokes theorem hence it is omitted.

62
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

Theorem 3 (Sufficient and Necessary Condition for path Independence)

Let 𝐷 be a connected region. Consider the following five statements:

1. 𝑭 = 𝑀𝒊 + 𝑁𝒋 + 𝑃𝒌 is a conservative vector field on 𝐷.

2.  F (r ).dr
C
is independent of path in 𝐷.

3.  F (r ).dr
C
=0 for every closed oriented curve C lying in 𝐷 (i.e.,  F .dr = 0 )
C

4. 𝑭. 𝒅𝒓 is an exact differential form on 𝐷.

5. 𝑐𝑢𝑟𝑙 𝑭 = 𝟎 on 𝐷.

Then

a) (1)  (2)  (3)

b) (4)  (5)
c) (1) ⇒ (4). [Hence, by (a), (2) and (3) also imply (4).We can make similar statements
using 1,2,3 and 5.]
d) If the domain of 𝑭 is a simply connected region, then (1)  (2)  (3)  (4)⇔ (5).

Example 11: Show that ∫𝐶(𝑦 2 − 6𝑥𝑦 + 6)𝑑𝑥 + (2𝑥𝑦 − 3𝑥 2 )𝑑𝑦 is path independent in ℝ2
and evaluate from (-1, 0) to (3, 4).

Ans. -36

Example 12: Show that ∫𝐶(𝑦 + 𝑦𝑧)𝑑𝑥 + (𝑥 + 3𝑧 2 + 𝑥𝑧)𝑑𝑦 + (9𝑦𝑧 2 + 𝑥𝑦 − 1)𝑑𝑧 is path
independent in ℝ2 and evaluate from (−1, 0) to (3, 4).

Ans. 194

Exercise 4.2:

4.3 Green’s Theorem and its Applications

63
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

Green’s theorem relates the line integral of a two dimensional vector field 𝑭 = 𝑀𝒊 + 𝑁𝒋
over the boundary C of a plane region R with the double integral of the scalar field f: = Nx –
My. This theorem is due to an English self-taught mathematician, George Green (1973 -
1841). The theorem is useful in

a) evaluating line integrals by double integrals

b) evaluating double integrals by line integrals
c) solving some geometrical and physical problems

Orientation of the boundary curve of a region: The boundary C of a plane region R is said
to have positive orientation or positive direction if R is kept to its left as we walk on C in this
direction.

Theorem 1 [Green’s Theorem]: Let 𝑅 be a closed bounded region in the xy-plane whose
boundary C is a piecewise smooth curve with positively orientation. Let 𝑀, 𝑁, 𝑀𝑦 and 𝑁𝑥 be
continuous everywhere in some domain containing R. Then

 Mdx + Ndy =  ( N
C R
x − M y )dA

Proof:

64
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

Case 1: If R is a simple region (i.e., R is both vertically and horizontally simple)

Thus there exists functions g1(x), g2 (x), h1(x), h2(x) and constants 𝑎, 𝑏, 𝑐 and 𝑑 such that

𝑅 = {(𝑥, 𝑦): 𝑎 ≤ 𝑥 ≤ 𝑏 and 𝑔1 (𝑥) ≤ 𝑦 ≤ 𝑔2 (𝑥)} = {(𝑥, 𝑦): 𝑐 ≤ 𝑦 ≤ 𝑑 and ℎ1 (𝑦) ≤

𝑥 ≤ ℎ2 (𝑦)}

It suffices to prove

i)  Mdx =  − M
C R
y dA

ii)  Ndy =  N
C R
x dA

To show (i) we need a parameterization of C. C is the union of the graphs C1 and C2 of 𝑔1 (𝑥)
and 𝑔2 (𝑥) for 𝑥 in [𝑎, 𝑏]. If we parameterize C1 and C2 by 𝑟1 (𝑡) = < 𝑡, 𝑔1 (𝑡) > and
𝑟2 (𝑡) = < 𝑡, 𝑔2 (𝑡) > , then C1has the same orientation as C but C2 has opposite orientation.
Hence

 Mdx =  Mdx +  Mdx

C C1 −C2
=  M (t , g (t ))dt −  M (t , g
C1
1
C2
2 (t )) dt

65
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

On the other hand

b  g 2 (t ) 
 M dy dx = − M (t , g1 (t )) dt + M (t , g 2 (t )) dt
R y
M dA = a  g ( x) y   
 1  C1 C2

Hence

 Mdx =  − M
C R
y dA

Similarly we can show

 Ndy =  N
C R
x dA

Case 2: If R is not a simple region. (Exercise)

Example 13 : Use Green’s theorem to evaluate

 (x − y 2 )dx + (2 y − x)dy
2

Where C consists of the boundary of the region R in the first quadrant bounded by the graphs
11
of y = x2 and y = x3. Ans. − 420

Example 14 (Verification of Green’s Theorem):

Verify Green’s Theorem for ∫𝐶(𝑦 2 − 7𝑦)𝑑𝑥 + (2𝑥𝑦 + 2𝑥)𝑑𝑦 where C is the circle 𝑥 2 +
𝑦 2 = 1.

Calculating area with Green’s Theorem

a) If a simple closed curve C in xy-plane and the region R it encloses satisfy the hypotheses
of Green’s theorem, then

66
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

1
𝑎𝑟𝑒𝑎(𝑅) =  xdy =  − ydy =
C C
2 C
xdy − ydx

b) If a simple closed curve C in the polar coordinate plane and the region R it encloses
satisfy the hypotheses of Green’s theorem, then

𝑎𝑟𝑒𝑎(𝑅) = r d
2

Example 15: Use Green’s theorem to find the area of the region R bounded by the ellipse
𝑥2 𝑦2
𝑎2
+ 𝑏2 = 1. Ans. 𝜋𝑎𝑏

Alternative Forms of Green’s Theorem

Green’s theorem in the plane can be expressed in two other forms called vector forms of
Green’s theorem. First we define the terms flux and circulation.

Definition 9: Let C be a smooth curve in the domain of a continuous vector field 𝑭 = 𝑀𝑖 + 𝑁𝑗 in

the plane, T be the unit tangent vector to C and 𝒏 be the outward pointing unit normal vector on C.
Then the line integral:

a) ∫𝐶 𝐹⃗ ∙ 𝑛⃗⃗𝑑𝑠 is called the flux of F across C.

b) ∫𝐶 𝐹⃗ ∙ 𝑇
⃗⃗𝑑𝑠 is called the circulation of F along C.

Flux and circulation in space will be defined in section 4.5.

Thus flux is the integral of the normal component of 𝑭 and circulation is the integral of the
tangential component of 𝐹 on a closed curve.

Flux Integral: ∫𝐶 𝐹⃗ ∙ 𝑇
⃗⃗𝑑𝑠 (The Integral of tangential component of F wrt arc length)

Circulation Integral: ∫𝐶 𝐹⃗ ∙ 𝑛⃗⃗𝑑𝑠 (The Integral of normal component of F wrt arc length.

67
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

Theorem 2 [Green’s Theorem in Vector Forms]

Let R and C be as in Green’s theorem, T a unit tangent vector, and n the outer unit normal
vector of C. Then Green’s Theorem can be written in the following two alternative forms.

a) ∮𝐶 𝑭. 𝐧𝒅𝒔 = ∬𝑅 𝑑𝑖𝑣(𝑭)𝑑𝐴 (Green’s Theorem in Vector Form I)

(The outward flux of a field 𝑭 = 𝑴𝑖 + 𝑵𝑗 across a simple closed curve C equals the
double integral of 𝑑𝑖𝑣(𝐹) over the region R enclosed by C.)

b) ∮𝐶 𝑭. 𝐓𝒅𝒔 = ∬𝑅 𝐶𝑢𝑟𝑙(𝑭). 𝒌 𝑑𝐴 (Green’s Theorem in Vector Form II )

(The circulation of a field 𝑭 = 𝑀𝒊 + 𝑁𝒋 around a simple closed curve C in the direction of

C equals the double integral of the k-component of curl 𝐹 over the region R enclosed by C.)

Remark:

a) Green’s Theorem in Vector form I is also called flux-divergence form or normal form of
Greens theorem or divergence theorem in the plane.

b) Green’s Theorem in Vector Form II is also called tangential form of Green’s Theorem or
Stokes’s Theorem in the plane.

Remark: We now have three ways of evaluating line integrals of the form  F (r ).dr
C

𝑏
✓ Parameterize C and use the formula: ∫𝑎 [𝑭(𝒓(𝑡)) . 𝒓′(𝑡)]𝑑𝑡.
✓ Use the Fundamental Theorem of Line Integrals (provided that F is a gradient).
✓ Use Green’s Theorem (provided that C is closed).

68
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

Exercise 4.3

1. Very Green’s theorem for the vector field 𝐹⃗ (𝑥, 𝑦) ≔ 𝑦𝒊 + 3𝑥𝒋 and the path given by
𝑥 2 + 𝑦 2 = 4 on 𝑥𝑦-plane.
2. (line integral by double integral)Use Green’s theorem to evaluate ∮𝐶 𝐹⃗ . ⃗⃗⃗⃗⃗
𝑑𝑟 where 𝐹⃗ ≔
1
〈𝑥𝑦, 𝑥 2 + 𝑥𝑦〉and C is the upper half of the ellipse 𝑥 2 + 4𝑦 2 = 1 and the X-axis on the
2

interval [−1,1] oriented positively.

[𝑑𝑥−𝑑𝑦]
3. (line integral by double integral )Calculate the integral ∮𝐶 using Green’s theorem.
𝑥+𝑦

The contour C is the boundary of the square with the

vertices 𝐴(1,0), 𝐵(0,1), 𝐷(−1,0), 𝐸(0, −1).
4. (double integral by line integral)Use Green’s theorem to evaluate
5. (Geometric Application): Use Green’s Theorem to calculate the area of the
region R bounded by the asteroid: 𝑥 = acos3 𝑡 , 𝑦 = asin3 𝑡 , 0 ≤ 𝑡 ≤ 2𝜋.
6. Evaluate ∮𝐶 𝑥𝑦𝑑𝑥 + 𝑥 2 𝑑𝑦 , where C is the rectangle with vertices
(0, 0), (3, 0), (3, 1), (0, 1) oriented counter-clockwise.
2
7. Compute: ∮𝐶[2𝑦 + √1 + 𝑥 5 ]𝑑𝑥 + [5𝑥 + 𝑒 𝑦 ]𝑑𝑦 ; where 𝐶 ∶= 𝑥 2 + 𝑦 2 = 4.
8. (Green’s theorem with multiple boundary components):

69
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

4.4 Surface Integrals

4.4.1 Preparation for Surface Integrals
With line integrals, we integrate over curves in space; with surface integrals we integrate
over surfaces in space. As a preparation for surface integrals, we should first discuss about
parametric representations for surfaces and orientations of surfaces. As tangent vectors can
be used to indicate the orientation of curves, normal vectors are used to define surface
orientations.
A. Parametric Equations of Surfaces
A surface Σ in 𝑥𝑦𝑧 −space can be described by

a) a Cartesian equation of the form f (x, y, z) = 0 or

b) a vector equation of the form 𝐫(u, v) = x(u, v) 𝐢 + y(u, v)𝐣 + y(u, v)𝐤
(where (u , v ) is a point in some region R on uv-plane).In this case 𝐫 is called a
parameterization of the surface and R is called the parameter domain. The vector
equation above can be expressed as a system of three scalar equations:
𝑥 = 𝑥(𝑢, 𝑣)
𝑦 = 𝑦(𝑢, 𝑣)
𝑧 = 𝑧(𝑢, 𝑣)
{(𝑢, 𝑣) ∈ 𝑅 ⊆ ℝ2

Note: If u0 and 𝑣0 are constants, then 𝐫(u0 , v) and 𝑟(𝑢, 𝑣0 ) represent curves lying on Σ.

70
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

Figure 5: Parameterization of a surface

Example 1(Parameterizations of some common surfaces)

a) (Parameterizing the Sphere 𝑥 2 + 𝑦 2 + 𝑧 2 = 𝑅 2 ):

Method I: Using Spherical Coordinates. In spherical coordinate system, every point
𝑃(𝑥, 𝑦, 𝑧) in space can be described by the set of numbers (𝜌, 𝜙, Θ) and the coordinates
of P in rectangular and spherical system are related by:
𝑥 = 𝜌 𝑠𝑖𝑛 𝜙 𝑐𝑜𝑠𝜃, 𝑦 = 𝜌 𝑠𝑖𝑛 𝜙 𝑠𝑖𝑛 𝜃 𝑎𝑛𝑑 𝑧 = 𝜌𝑐𝑜𝑠 𝜙
But for every point P(x,y,z) on the sphere with equation 𝑥 2 + 𝑦 2 + 𝑧 2 = 𝑅 2 , we have
𝜌 = 𝑅 and hence the sphere can be represented by the parametric equarions :
𝑥 = 𝑅 𝑠𝑖𝑛 𝜙 𝑐𝑜𝑠𝜃, 𝑦 = 𝑅 𝑠𝑖𝑛 𝜙 𝑠𝑖𝑛 𝜃 𝑎𝑛𝑑 𝑧 = 𝑅𝑐𝑜𝑠 𝜙.
The values of 𝜃 ranges from 0 to 2𝜋 and the values of 𝜙 ranges from 0 to 𝜋. If we let u
= 𝜃 and v= 𝜙 we get the set of equations :
𝑥 = 𝑎 𝑐𝑜𝑠𝑢 sin 𝑣
{ 𝑦 = 𝑎 𝑠𝑖𝑛𝑢 sin 𝑣
𝑧 = 𝑎 𝑐𝑜𝑠𝑣
Thus, a parameterization of the sphere with radius 𝑅 and center at the origin is:

⃗⃗⃗⃗⃗⃗⃗⃗⃗ = 𝑅 𝑐𝑜𝑠 𝑢 sin 𝑣 𝒊 + 𝑅 sin 𝑢 𝑠𝑖𝑛 𝑣 𝒋 + 𝑅 𝑐𝑜𝑠 𝑣 𝒌; R: 0 ≤ u ≤2𝜋, 0 ≤ v ≤ π.

𝑟(𝑡)

71
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

Method 2 (via latitudes and longitudes): Read!

b) (Any surface of the form z = f(x, y)): A parameterization of the surface z = f(x, y) with
domain R is :
𝑟⃗(𝑢, 𝑣) = 𝑢 𝒊 + 𝑣 𝒋 + 𝑓(𝑢, 𝑣)𝒌 ; (𝑢, 𝑣) 𝜖 𝑅.
Surfaces of the form 𝑦 = 𝑓(𝑥, 𝑧) and 𝑥 = 𝑓(𝑦, 𝑧) can be parameterized in analogous
ways.
c) (Planes)
Method I: A parameterization of the plane 𝑎𝑥 + 𝑏𝑦 + 𝑐 𝑧 = 𝑑 is (𝑟⃗(𝑢, 𝑣)) = 𝑢 𝑖 + 𝑣𝑗 +
((𝑑 − 𝑎𝑢 − 𝑏𝑣))/𝑐)𝑘 (if 𝑐 ≠ 0). The parameter domain is ℝ2 (the entire uv-plane).
Method II: There is also another standard way to parameterize planes in linear algebra.
Read it!
A surface that is part of the plane 𝑎𝑥 + 𝑏𝑦 + 𝑐 𝑧 = 𝑑 can be parameterized using the
same parametric equations given above but with a different parametric domain. For
example, a parameterization of the triangular planar surface with vertices
⃗⃗ where the parameter
(1,0,0), (0,1,0) and (0,0,1) is 𝑟⃗(𝑢, 𝑣)) = 𝑢 𝑖⃗ + 𝑣𝑗⃗ + (1 − 𝑢 − 𝑣)𝑘
domain is the triangular region in uv-plane with vertices (0,0), (1,0) and (1,1).

: Figure 21: Portion of the Plane 𝒙 + 𝒚 + 𝒛 = 𝟏 in the first octant

72
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

d) (Cylinder ): A parameterization of the cylinder 𝑥 2 + 𝑦 2 = 𝑎2 ; −1 ≤ 𝑧 ≤ 1 is:

𝑟⃗(𝑢, 𝑣) = 𝑎 𝑐𝑜𝑠 𝑢𝒊 + 𝑎 sin 𝑢 𝒋 + 𝑣 𝒌; R: 0 ≤ u ≤ 2π, -1 ≤ v ≤ 1.

e) (Cone): A parameterization of the cone 𝑧 = √𝑥 2 + 𝑦 2 ; 0 ≤ 𝑧 ≤ ℎ is:

𝑟⃗(𝑢, 𝑣) = 𝑢 𝑐𝑜𝑠 𝑣𝒊 + 𝑢 𝑠𝑖𝑛 𝑣 𝒋 + 𝑢𝒌; 𝑅: 0 ≤ 𝑢 ≤ ℎ, 0 ≤ 𝑣 ≤ 2𝜋.
f) (Surfaces of Revolution): If the graph z = f(x), a ≤ x ≤ b is rotated about the z-axis
then the resulting surface has a parameterization:
𝑟⃗(𝑢, 𝑣) = 𝑢 𝑐𝑜𝑠 𝑣𝒊 + 𝑢𝑠𝑖 𝑛 𝑣𝒋 + 𝑓(𝑢)𝒌; 0 ≤ 𝑢 ≤ 𝑏, 0 ≤ 𝑣 ≤ 2𝜋.

𝑥2 𝑦2 𝑧2
g) (Ellipsoid): A parameterization of the ellipsoid 𝑎2
+ 𝑏2
+ 𝑐2
= 1 is:

⃗⃗⃗⃗(𝑢, 𝑣 = 𝑎 𝑐𝑜𝑠 𝑢 sin 𝑣 𝑖 + 𝑐 𝑠𝑖𝑛 𝑢 𝑠𝑖𝑛 𝑣 𝑗 + 𝑐𝑐𝑜𝑠𝑣 𝑘; R: 0 ≤ u ≤2𝜋, 0 ≤ v ≤ π.

𝑟)
h) (Torus): A torus is a 3-D figure formed by rolling a rectangle into a cylinder and
bending the cylinder until its bases meet. Equivalently we can obtain it by
rotating a circle C in a plane about a line perpendicular to the plane. The
parameterization of the torus obtained by rotating the circle with center (R, 0,
0) and radius a about y axis is:
r (u, v) = ((R + a cos u) cos v) i + ((R + r cos u) sin v) j + ((a sin u) k

Figure 7: Torus

73
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

B. Tangent Planes and Normal vectors to Surfaces; Surface Area

The tangent vectors of all the curves on a surface S through a point P of S form a plane,
called the tangent plane of S at P. If S has an edge or a cusp (like a cone), S cannot have a
tangent plane at such a point. A vector perpendicular to the tangent plane is called a normal
vector of S at P.

Theorem 1 [Normal vector of a Parametric Surface]: Let

𝑟⃗(𝑢, 𝑣)=〈𝑥(𝑢, 𝑣), 𝑦(𝑢, 𝑣), 𝑧(𝑢, 𝑣)〉; (𝑢, 𝑣) 𝜖 𝑅 ⊆ ℝ2 be a parameterization of a surface, Σ.
Define

𝜕𝑥 𝜕𝑦 𝜕𝑧
𝑟𝑢 𝑣): = 〈𝜕𝑢 (𝑢, 𝑣), 𝜕𝑢 (𝑢, 𝑣), 𝜕𝑢 (𝑢, 𝑣)〉 ,
⃗⃗⃗⃗(𝑢,

𝜕𝑥 𝜕𝑦 𝜕𝑧
⃗𝑟⃗⃗(𝑢,
𝑣 𝑣): = 〈𝜕𝑣 (𝑢, 𝑣), 𝜕𝑣 (𝑢, 𝑣), 𝜕𝑣 (𝑢, 𝑣)〉 and

⃗⃗ (𝑢, 𝑣): = 𝑟⃗⃗⃗⃗(𝑢,

𝑁 𝑢 𝑣) × ⃗⃗⃗⃗(𝑢,
𝑟𝑣 𝑣)

⃗⃗(𝑢0 , 𝑣0 ) ≠ ⃗0⃗, then 𝑁

If 𝑁 ⃗⃗ (𝑢0 , 𝑣0 ) is normal to the surface Σ at 𝑟⃗(𝑢0 , 𝑣0 ).

Justification: Assume that the curves 𝑢 = 𝑐𝑜𝑛𝑠𝑡 and 𝑣 = 𝑐𝑜𝑛𝑠𝑡 intersect at the point ( 𝑟⃗
(𝑢0 , 𝑣0 )) at a nonzero angle. Hence the partial derivatives 𝑟𝑢 and 𝑟𝑣 at P are linearly
independent tangential vectors. Then 𝑟𝑢 and 𝑟𝑣 span the tangent plane of Σ at P. Hence their
cross product gives a normal vector N of Σ at P.

74
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

Example 2 [normal vector to a parametric surface]: Find the surface unit normal 𝑛⃗⃗ and
the equation of the tangent plane to the cylinder:

𝑟⃗⃗⃗ (𝑢, 𝑣) = 〈3 𝑐𝑜𝑠 (𝑢), 3 𝑠𝑖𝑛 (𝑢), 𝑣〉 𝑎𝑡 𝑟(𝜋, 2) = (3, 0, 2).

Answer: 𝑛⃗⃗ = 〈−1,0,0〉 and the equation of the tangent plane is 𝑥 = −1.

Example 3[Equation of tangent plane to a surface]: Find an equation of the tangent plane
to the parametric surface given by: 𝑟⃗(𝑢, 𝑣)= 〈𝑢 + 𝑣, 3𝑢2 , 𝑢 − 𝑣〉at the point(2, 3, 0).

Answer: 3𝑥 − 𝑦 + 3𝑧 = 3.

Theorem 2[Surface Area of a Parametric Surface]: The surface area 𝑆 of the surface Σ
with parameterization 𝑟⃗(𝑢, 𝑣)=〈𝑥(𝑢, 𝑣), 𝑦(𝑢, 𝑣), 𝑧(𝑢, 𝑣)〉; (𝑢, 𝑣) 𝜖 𝑅 ⊆ ℝ2 is given by

S = ∬R‖r⃗⃗⃗⃗(u,
u v) × ⃗⃗⃗⃗(u,
rv v)‖dA;

(provided that Σ is traced exactly once as (𝑢, 𝑣) ranges over all points of 𝑅.).

75
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

Example 4[surface area of a parametric surface]: Find the area of the parametric surface
given by: 𝑟⃗(𝑢, 𝑣)=〈𝑢𝑣, 𝑢 + 𝑣, 𝑢 − 𝑣〉 ;𝑢2 + 𝑣 2 ≤ 1.
𝜋
Answer: 3
(6√6 − 2 √2)

C. Topology of Surfaces: Smoothness and Orientation

Definition 1 [smooth and piecewise smooth surface]: A surface 𝒮 with parameterization

𝑟⃗(𝑢, 𝑣)=〈𝑥(𝑢, 𝑣), 𝑦(𝑢, 𝑣), 𝑧(𝑢, 𝑣)〉; (𝑢, 𝑣) 𝜖 𝑅 ⊆ ℝ2 is said to be

a) smooth if 𝑟⃗⃗⃗⃗(𝑢,
𝑢 𝑣) 𝑎𝑛𝑑 ⃗𝑟⃗⃗(𝑢,
𝑣 𝑣) are continuous and ⃗⃗⃗⃗(𝑢,
𝑟𝑢 𝑣) × ⃗⃗⃗⃗(𝑢,
𝑟𝑣 𝑣) is never zero
for every (𝑢, 𝑣) in the parameter domain.
b) piecewise smooth if it is a union of finitely many smooth surfaces
Intuitively speaking, a surface is smooth if there is no sharp point on the surface. For instance
a cone has a sharp point, hence it is not smooth.

Example 5 [smooth and piecewise smooth surfaces]: Sphere is smooth and cube is
piecewise smooth.

Definition 2 [orientable surface, oriented surface and orientation of a surface]:

A surface Σ is said to be:

a) orientable (two sided) if there exists a continuous unit normal field 𝑛⃗⃗ (x, y, z) defined on
𝒮. Roughly speaking, a surface is orientable if it has two sides (like inside and outside,
left side and right side, upper side and lower side, etc). If a surface is orientable it can be
can be penetrated in two opposite directions (that is, there are two opposite directions of
flow through the surface). Note that, for every orientable surface, there are exactly two
continuous unit normal fields that are negative of each other, say 𝑛⃗⃗ (x, y, z) and
−𝑛⃗⃗ (x, y, z).
b) oriented if one direction of flow through 𝒮 has been chosen as positive direction. Often a

76
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

surface is oriented by choosing a normal vector field that indicates the required direction.
The chosen direction of flow (or normal vector field) is called an orientation of the
surface. (Read also what we mean by orientation of a piecewise smooth surface).

Example 6 (orientable and non-orientable surface):

Most surfaces we encounter in the physical world are orientable. Spheres, planes, and tori are
orientable, for example. Every smooth closed surface is orientable. But Möbius strips,
Roman surfaces, real projective planes, and Klein bottles are non-orientable.

Figure 8: Mobius Strip

Definition 3 [simple and closed surfaces]: A surface 𝒮 is said to be:

a) simple if it never touches or intersects itself.
b) closed if it encloses a solid.

Definition 4(orientation of boundary of a surface): Let 𝒮 be an oriented non-closed

surface with surface normal/orientation/ 𝑛⃗⃗ . Let 𝒞 be the boundary curve of S with unit
⃗⃗. The direction of 𝑇
tangent vector 𝑇 ⃗⃗ is called the positive direction (orientation) of C if 𝑛⃗⃗ × 𝑇
⃗⃗
points into S.

Informally, the positive direction of 𝒞 is the direction a person has to walk on 𝒞 in such a
way that his/her head points in the direction of the surface normal while keeping the surface
to the left.

77
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

4.4.2 Surface Integrals of a Scalar Field

The surface integrals we consider are the following types:

a) surface integral of scalar field:  f ( x, y, z)dydz


b) Surface integrals in differential forms:
- Reduced Surface integrals:  Mdydz ,  Ndxdz ,  Pdxdy
  
- Surface integral in differential form:  Mdydz +  Ndxdz +  Ndxdy
  

c) Surface integral of vector field (Flux Integral): ∬Σ 𝐹⃗ (𝑥, 𝑦, 𝑧) ∙ 𝑛⃗⃗ 𝑑𝑠

A. Motivation: Mass of a Surface with given variable density

Let Σ be a surface in xyz space. Suppose that the density per unit area of the surface is given
at each point (𝑥, 𝑦, 𝑧) of  by the function 𝑓(𝑥, 𝑦, 𝑧).What is the mass of the surface?

Let {Σ 1 , …, Σn} be a partition of  and S k be the area of Σk. Let (x*k, y*k, z* k ) be any

point in Σk . Then 𝑓 (𝑥𝑘∗ , 𝑦𝑘∗ , 𝑧𝑘∗ ) is the density of 𝑓 at (x* k, y* k, z* k ) . Then, if S k is very,

very small, then the mass m k of the subsurface S k of  can be approximated as

m k  f(x k, y k, z k) S k

(Because density is mass per unit area, for a two dimensional body and because the density
is almost constant on S k , as S k is almost a point by assumption). The total mass M of the

surface is approximated as

n
M 1
f(x* k, y* k, z* k) S k

If n →  as Δ𝑆 → 0 , then

78
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

n
M= lim
n→
 1
f(x* k, y* k, z* k ) S k (1)

The limit in (1) can be used to express not only mass but other quantities. We call it the
surface integral of f over  and we denote it by  f ( x, y, z )ds .


n
Hence  f ( x, y, z )ds = lim
n→
 f(x k, y k, z k) S k
 1

NB. If Δ𝑆 → 0 , then 𝑛 →  .But the converse is not necessarily true.

B. Definition of surface Integral of a scalar field

Definition 5: Let  be a finite, smooth and simple surface in xyz space. Let 𝑓 (𝑥, 𝑦, z) be

a continuous scalar field defined at each point of Σ . Let {Σ1, Σ2…, Σ n} be a partition of Σ as
shown in the figure. Then the surface integral of 𝑓 over Σ (with respect to surface area)
is denoted and defined by

 f ( x, y, z )ds = lim
n→
 f(x*k, y*k, z*k ) S k
 1

79
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

(Where S k is the area of Σ𝑘 , (x*k, y*k , z*k) is any point in Σ k, and max {Δ𝑆𝑘 }𝑛𝑘=1 tends to

0 as n →  .)

C. Evaluation: The basic idea is to convert into double integral.

If 𝑟⃗(𝑢, 𝑣)=〈𝑥(𝑢, 𝑣), 𝑦(𝑢, 𝑣), 𝑧(𝑢, 𝑣)〉, (𝑢, 𝑣) 𝜖 𝑅 is a parameterization of Σ and if 𝑓(𝑥, 𝑦, 𝑧)
is a scalar field defined on Σ, then

 f ( x, y, z)ds =  f ( x(u, v), y(u, v), z(u, v)) r u  rv dA

 R

Procedures (To calculate a surface integral):

1. Parameterize the surface. Let 𝑟⃗(𝑢, 𝑣)=〈𝑥(𝑢, 𝑣), 𝑦(𝑢, 𝑣), 𝑧(𝑢, 𝑣)〉; (𝑢, 𝑣) 𝜖 𝑅 be a
parameterization of the surface

2. Compute the differential Area Element ds: It is given by 𝑑𝑠 = ‖𝑟𝑢 × 𝑟𝑣 ‖𝑑𝑢𝑑𝑣 .

3. Convert to double integral over the parameter domain and compute it:

 f ( x, y, z)ds =  f ( x(u, v), y(u, v), z(u, v)) ru  rv dA

 R

Example 7 (surface integral over the graph of z =f(x, y)): Compute the surface integral

∬Σ(𝑥𝑦 + 𝑧)𝑑𝑠, where S is that part of the plane x + y+ z = 2 in the first octant. Ans.: 2√3.

The following properties may help in evaluating surface integrals.

D. Properties

a) ∬S kf = k ∬S f
b) ∬S f + g = ∬S f + ∬S g
c) ∬S f − g = ∬S f − ∬S g

80
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

d) ∬S ∪S f = ∬S f + ∬S f
1 2 1 2

Example 8 (surface integral of a scalar field over a piecewise smooth surface): Evaluate
∬Σ(𝑥 + 𝑦)𝑑𝑠, where ∑ is the cube with vertices A(0,0,0), B(1,0,0), C(1,1,0), D(0,1,0), A’(0,0,1),
B’(1,0,1), C’(1,1,1) and D’(0,1,1). Answer: 6

Example 9: Evaluate ∬Σ 𝑧𝑑𝑠

Where ∑ is the upper half of the sphere cantered at origin and with radius 2.

Solution: A parameterization of the surface is 𝑟⃗ = 2𝑠𝑖𝑛𝑣 𝑐𝑜𝑠𝑢 𝒊 + 2sin 𝑣 sin 𝑢 𝒋 +

𝜋
2cos 𝑣 𝒌) where 0 ≤ 𝑣 ≤ 2
and 0 ≤ 𝑢 ≤ 2𝜋.

The surface normal vector is ⃗⃗⃗⃗

𝑟𝑢 × ⃗⃗⃗⃗
𝑟𝑣 = −4𝑠𝑖𝑛𝑣(𝑠𝑖𝑛𝑣 𝑐𝑜𝑠𝑢 𝒊 + sin 𝑣 sin 𝑢 𝒋 + cos 𝑣 𝒌)
The differential surface area element is 𝑑𝑠 = ‖𝑟𝑢 × 𝑟𝑣 ‖𝑑𝑢𝑑𝑣 = 4 sin(𝑣).

𝜋
2𝜋
Thus ∬Σ 𝑧𝑑𝑠 = ∫0 ∫02 (2 cos 𝑣)(4 sin 𝑣) 𝑑𝑣 𝑑𝑢 = 8𝜋 //.

Example 10: Evaluate ∬Σ 𝑦𝑑𝑠

Where ∑ is the portion of the cylinder x2 + y2 = 3 that lies between 𝑧 = 0 and 𝑧 = 6. Ans. 0

81
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

Example 11: Evaluate∬Σ(𝑦 + 𝑧)𝑑𝑠, where ∑ is the surface whose side is the cylinder x2 + y2
= 3 and whose bottom is the disk x2 + y2 ≤ 3 in the xy-plane and whose top is the plane 𝑧 =
4 − 𝑦..

𝜋
Answer. [29√3 + 24√2 ]
2

4.4.3 Surface Integral of Vector Fields(Flux Integral)

A. Motivation : Volume of a Fluid flowing through a Surface Per Unit Time

Let Σ be an orientable surface with orientation 𝑛⃗⃗. (a unit and normal vector field which is
continuous at each non-boundary point of the surface). If v is the velocity field of a fluid
flowing through Σ in the direction of 𝑛⃗⃗ , then the volume of the fluid that crosses the surface
per time unit (i.e. per second, per minute, or whatever time unit you are using) is given by:

n
lim
n→
 v( x *
k =1
k , y *k , z *k ).nS k

Where {Σ1, Σ2…, Σ n} is a partition of the surface, ∆𝑠𝑘 ≔ 𝑎𝑟𝑒𝑎(Σ k) and (𝑥𝑘∗ , 𝑦𝑘∗ , 𝑧𝑘∗ ) ∈ Σ𝑘 .
n
In general for any vector field F, the limit lim
n→
 F (x *
k =1
k , y *k , z *k ).nS k

is called the surface integral of F over 𝛴 in the direction of 𝑛⃗⃗. Thus, this limit measures the
flow rate or volume of fluid crossing a given surface (cross-section of a pipe, for instance)
per unit time. The amount of fluid crossing a cross section per unit time depends on
how fast it moves its direction of flow and the area of the cross section.

82
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

Justification: We consider three cases.

Case 1: If n and v have the same direction, then the volume of fluid that passes through Σ k
between a time t and a later time t + ∆𝑡 is 𝑽𝒐𝒍𝒖𝒎𝒆 ≅ (𝒉𝒆𝒊𝒈𝒉𝒕)(𝒃𝒂𝒔𝒆 𝒂𝒓𝒆𝒂) =
‖𝒗(𝒙𝒌 , 𝒚𝒌 , 𝒛𝒌 )‖ ∆𝒕 ∆𝑺𝒌 , which is the volume of the cylinder in the left most figure. Thus the total
volume is approximately ∑𝑛𝑘=1 𝑣([𝑥𝑘 , 𝑦𝑘 , 𝑧𝑘 ). 𝑛]∆𝑡∆𝑠𝑘 . Hence the rate of flow per unit time is
∑𝑛𝑘=1 𝑣([𝑥𝑘 , 𝑦𝑘 , 𝑧𝑘 ). 𝑛∆𝑠𝑘 . The total volume of fluid that crosses the surface per unit time is exactly
equal to the limit of ∑𝑛𝑘=1 𝑣([𝑥𝑘 , 𝑦𝑘 , 𝑧𝑘 ). 𝑛∆𝑠𝑘 . as n approaches to infinity.

The justifications for Case 2 and case 3 are left as an exercise.

B. Definition

83
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

Definition 6: Let Σ be an orientable (hence smooth), finite and simple surface.

Let 𝐹⃗ = 𝑀𝒊 + 𝑁𝒋 + 𝑃𝒌 be a vector field defined at each point of Σ . Let 𝑛⃗⃗ be an
orientation of Σ (i.e., 𝑛⃗⃗ is a continuous unit normal filed to the surface). Let {Σ1, Σ2…, Σ n}
be a partition of Σ as shown in the figure. Then we denote the surface integral of F over 𝛴 in
→
the direction of 𝑛⃗⃗ by  F .nds or  F . ds and define it as
 

 F .nds = lim
n→
 F ( x *k , y *k , z *k ).nS k
 k =1

where

for each 𝑘 = 1, … , 𝑛 , ∆𝑆𝑘 is the surface area of Σ𝑘 ,

(x*k ,y*k, z*k )  Σ𝑘 and

Provide that ‖Δ𝑆‖ ≔ max {∆S1 , ∆S2 , , … , ∆Sn }→0 as n→∞.

* Thus flux is a scalar quantity, defined as the surface integral of the component of a vector
field perpendicular to the surface at each point.

The integral can be interpreted physically as the rate of flow of F through S.

C. Evaluation

Let ∑ be a smooth surface with parameterization

𝑟 (𝑢, 𝑣) = 𝑥(𝑢, 𝑣)𝑖 + 𝑦(𝑢, 𝑣)𝑗 + 𝑧(𝑢, 𝑣)𝑘; (𝑢, 𝑣) ∈ 𝑅

⃗⃗⃗⃗over ∑ in the direction of the unit vector

The surface integral of a vector field 𝐹

𝑛⃗⃗ can be computed using

 F .nds =  F (r (u, v)).( r u  rv )dA

 R

84
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

Where

𝑅 is the projection of ∑ onto uv-plane.

𝑑𝐴 is a differential area element on uv-plane and 𝑑𝑠 is differential surface element

[𝑑𝐴 is the projection of 𝑑𝑠 onto uv-plane].

For clarity, we can apply the following procedures.

Procedures (for Evaluating a Flux Integral):

1. Parameterize the Surface, say by 𝑟(𝑢, 𝑣) ; (𝑢, 𝑣) ∈ 𝑅

𝑟𝑢×𝑟𝑣 𝑟 ×𝑟
2. Orient the surface, i.e, choose 𝑛⃗⃗ = ‖𝑟 or 𝑛⃗⃗ = − ‖𝑟𝑢×𝑟𝑣‖ depending on the
𝑢 ×𝑟𝑣 ‖ 𝑢 𝑣

direction of the flux.

3. Convert the surface integral to double integral over the parameter domain R. To do this,
express 𝐹⃗ as a function of u and v, compute the dot product 𝐹⃗ . 𝑛⃗⃗ and multiply the
result of the dot product by 𝑑𝑠 (that is by ‖𝑟𝑢 × 𝑟𝑣 ‖𝑑𝐴).
Note that 𝑛⃗⃗ ds = 𝑟𝑢 × 𝑟𝑣 and hence, 𝐹⃗ . 𝑛⃗⃗𝑑𝑠 = 𝐹⃗ . 𝑟𝑢 × 𝑟𝑣 .
4. Compute the double integral in step (3).

Example 12: Evaluate  F.nds

S

Where 𝐹⃗ = 〈𝑥 2 , 0, 3𝑦 2 〉

S: portion of the plane 𝑥 + 𝑦 + 𝑧 = 1 in the first octant

𝑛⃗⃗ is chosen so that its k component is positive.

85
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

Figure 9 : Portion of the Plane 𝒙 + 𝒚 + 𝒛 = 𝟏 in the first octant

Solution: A parameterization of the given surface is 𝑟⃗(𝑢, 𝑣) = 〈𝑢, 𝑣, 1 − 𝑢 − 𝑣〉 with

parameter domain : 0 ≤ 𝑢 ≤ 1 & 0 ≤ 𝑣 ≤ −𝑢 + 1 . (R is the triangular region in uv-plane
with vertices (0,0), (1,0) and (0,1). )

𝑟𝑢 = 〈1,0, −1〉 and ⃗𝑟⃗⃗𝑣 = 〈0,1, −1〉 and hence, ⃗⃗⃗⃗

Then ⃗⃗⃗⃗ 𝑟𝑢 × ⃗𝑟⃗⃗𝑣 = 〈1,1,1〉.

Since the k- component of 〈1,1,1〉 is positive, so the unit vector in the direction of 𝑟⃗⃗⃗⃗
𝑢 × ⃗𝑟⃗⃗
𝑣 is
1 1 1
just 𝑛⃗⃗, i.e., 𝑛⃗⃗ = 〈 , , 〉 and = √3 .
√3 √3 √3

1
Thus ∬𝑆〈𝑥 2 , 0,3𝑦 2 〉. 𝑛𝑑𝑠 = ∬𝑅[〈𝑢2 , 0,3𝑣 2 〉. 〈1,1,1〉]dA = 3 //

Example 13: Evaluate  F.nds where 𝐹⃗ ≔ 𝑧𝒌,

S
S is the unit sphere centred at the origin and

4𝜋
n is the outward normal vector field to S. Answer: 3

86
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

4.4.4 Surface Integrals in Differential Forms

Definition 7 (Kreyszig Page 444): Let the vector field 𝑛⃗⃗ be an orientation of a smooth finite
surface S. Let α, β and 𝛾 be the angles between 𝑛⃗⃗ and 𝑖, 𝑗 and 𝑘 respectively. Let 𝑓 be a scalar
field defined at each point of S. We define

a. ∬𝑆 𝑓𝑑𝑥𝑑𝑦 = ∬𝑆 𝑓𝑐𝑜𝑠𝛾𝑑𝑠

b. ∬𝑆 𝑓𝑑𝑦𝑑𝑧 = ∬𝑆 𝑓𝑐𝑜𝑠𝛼𝑑𝑠

c. ∬𝑆 𝑓𝑑𝑧𝑑𝑥 = ∬𝑆 𝑓𝑐𝑜𝑠𝛽𝑑𝑠

Caution: 𝑑𝑥𝑑𝑦 ≠ 𝑑𝑦𝑑𝑥

Note: Note that the element of area 𝑑𝑥 𝑑𝑦 in the xy-plane is the projection |cos 𝛾| 𝑑𝑠 of the
element of area 𝑑𝑠 of 𝑆 onto

⃗⃗ 𝑑𝑠 (because 𝑛⃗⃗ =
Evaluation: ∬𝑆 𝑓𝑑𝑥𝑑𝑦 = ∬𝑆 𝑓(𝑥(𝑢, 𝑣), 𝑦(𝑢, 𝑣), 𝑧(𝑢, 𝑣))𝑛⃗⃗. 𝑘
⃗⃗ = cos 𝛾).
〈cos 𝛼 , cos 𝛽 , cos 𝛾〉 and, so, 𝑛⃗⃗. 𝑘

Definition 8: Let 𝐹⃗ = 𝑀𝒊 + 𝑁𝒋 + 𝑃𝒌 be a vector field defined at each point of a smooth

surface S of finite area. We define:

∬𝑆 𝑀𝑑𝑦𝑑𝑧 + 𝑁𝑑𝑧𝑑𝑥 + 𝑃𝑑𝑥𝑑𝑦 = ∬𝑆 𝑀𝑑𝑥𝑑𝑦 + ∬𝑆 𝑁𝑑𝑦𝑑𝑧+ ∬𝑆 𝑁𝑑𝑧𝑑𝑥

Remark (Alterative way to evaluate flux integral): If 𝐹⃗ = 𝑀𝒊 + 𝑁𝒋 + 𝑃𝒌 is a vector

field defined at each point of a smooth, orientated surface S of finite area, then

∬ 𝐹⃗ . 𝑛⃗⃗ 𝑑𝑠 = ∬ 𝑀𝑑𝑦𝑑𝑧 + 𝑁𝑑𝑧𝑑𝑥 + 𝑃𝑑𝑥𝑑𝑦

𝑆 𝑆

87
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

We can use this formula to evaluate surface integrals by converting them to double integrals
over regions in the coordinate planes of the xyz-coordinate system. But we must carefully
take into account the orientation of S (the choice of n).

Exercise 4.4:

1. Find a parameterization of the surface given by the paraboloid 𝑧 = 𝑥 2 + 𝑦 2 and

below the plane 𝑧 = 4.
2. Find a unit vector with positive z component which is normal to the surface 𝑧 = 𝑥 4 𝑦 +
𝑥𝑦 2 at the point (1,1,2) on the surface.
3. Find the surface area of the surface given by the parameterization 𝑟⃗(𝑠, 𝑡) ≔
〈−𝑠 − 𝑡, 𝑠 + 2𝑡, 2𝑠 + 𝑡〉 where 0 ≤ 𝑠 ≤ 1 and 0 ≤ 𝑡 ≤ 1.
4. Evaluate the surface integral ∬𝑆 𝑓𝑑𝑠 where 𝑓(𝑥, 𝑦, 𝑧) = 𝑥 2 over the unit sphere 𝑥 2 +
𝑦 2 + 𝑧 2 = 1.
⃗⃗⃗⃗⃗ where 𝐹⃗ (𝑥, 𝑦, 𝑧) = 〈𝑧 2 , 𝑥, −3𝑧〉 and S is the
5. Evaluate the surface integral ∬𝑆 𝐹⃗ . 𝑑𝑠
surface cut from the parabolic cylinder 𝑦 = 𝑥 2 with −1 ≤ 𝑥 ≤ 1 by the planes 𝑧 =
0 and 𝑧 = 2.
6. Parameterize the hyperboloid 𝑥 2 − 4𝑦 2 + 𝑧 2 = 1.
7. Find a parametric representation for the part of the plane z = x + 3 that lies inside the
cylinder x2 + y2 = 1.
8. Parameterize the upper hemisphere of the sphere 𝑥 2 + 𝑦 2 + 𝑧 2 = 9.
9. Find a parameterization of the portion of the sphere 𝑥 2 + 𝑦 2 + 𝑧 2 = 1.between the
1 1
planes z=− 2 &𝑧 = 2.

10. Find the equation of the tangent plane to the surface given by :
𝑟⃗(𝑢, 𝑣) ≔ 〈𝑢, 2𝑣 2 , 𝑢2 + 𝑣〉at the point (2,2,3).

11. For the cylinder of radius 3 and height 5 given by x2 + y2 = 32 and 0  z  5, let the
charge density be proportional to the distance from the xy-plane. Find the total charge
on the cylinder. Ans. 75𝜋

88
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

12. [Mass problem]: Find the mass of the lamina that is the portion of 𝑥 2 + 𝑦 2 − 3𝑧 −
9
1 = 0 inside 𝑥 2 + 𝑦 2 = 9/4. The density is δ(x, y, z) = ( + 𝑥 2 + 𝑦 2 ).
4

89
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

4.5 Divergence Theorem

Divergence theorem relates triple integral over a solid with surface integral over the
bounding surface of the solid. It was stated first by Gauss; hence it is also called Gauss’s
Theorem. It can be viewed as a generalization of Green’s theorem; it is analogous to the
divergence form of Green’s Theorem.
Divergence theorem is useful for evaluating triple integrals via surface integrals or
vice versa. It is also useful in the study of physical problems like heat flow, fluid flow,
potential theory and electro-magnetic theory.

Definition 1: The flux of a three-dimensional vector field F across an oriented surface S in

the direction of unit normal vector n is

𝑭𝒍𝒖𝒙 = ∬ 𝑭. 𝒏𝒅𝒔
𝑺

The definition of the flux of a two dimensional field across a curve is given in section 4.3.

Theorem 1 [Divergence Theorem]: Let

i) 𝐷 be a closed and bounded three dimensional region in xyz-space with piecewise smooth
and orientable boundary ∑ that is oriented outward.

ii) 𝐹⃗ (𝑥, 𝑦, 𝑧) = 𝑀(𝑥, 𝑦, 𝑧)𝑖 + 𝑁(𝑥, 𝑦, 𝑧)𝑗 + 𝑃(𝑥, 𝑦, 𝑧)𝑘 be a vector field for which M,N and
P are continuous and have continuous first order partial derivatives in a three dimensional
region containing D.

Then

∬ ⃗𝑭⃗. 𝑛⃗⃗𝑑𝑠 = ∭ 𝑑𝑖𝑣(𝐹⃗ )𝑑𝑣

Σ 𝐷

90
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

(In words: The flux of a vector field 𝐹⃗ across a closed oriented surface ∑ in the direction of
the surface outward unit normal field 𝑛⃗⃗ equals the triple integral of div (𝐹⃗ ) over the region D
enclosed by the surface).

Partial Proof:

Case 1: If T is a simple region (i.e., the intersection of T with every line parallel to the
coordinate planes is a point or a single segment)

Case 2: if T is divisible to finite number of simple regions

Case 3: The general case

Example 1[using triple integral to evaluate surface integral]: Use Divergence theorem to
evaluate

∬ 𝐹⃗ ∙ ⃗⃗⃗⃗𝑑𝑠
𝑛
Σ

where 𝐹⃗ (𝑥, 𝑦, 𝑧) = 𝑥𝒊 − 2𝑦 2 𝒋 + 𝑧 2 𝒌 , Σ is the boundary surface of the solid bounded by the

cylinder 𝑥 2 + 𝑦 2 = 4 and the planes 𝑧 = 0 and 𝑧 = 3 and ⃗⃗⃗⃗
𝑛 is the unit surface normal vector
field of Σ that is directed outwards.

Solution: ∬Σ 𝐹⃗ ∙ 𝑛
⃗⃗⃗⃗𝑑𝑠 = ∭𝐷 𝑑𝑖𝑣(𝐹⃗ )𝑑𝑣 = ∭𝐷(1 − 4𝑦 + 2𝑧)𝑑𝑣 = 48𝜋

Example 2 (Verification of the Divergence Theorem): Verify divergence theorem for the
problem of Example 1.

Solution: It follows from the solution of Example 47 that ∭𝐷 𝑑𝑖𝑣(𝐹⃗ )𝑑𝑣 = 48𝜋. It remains

to show that ∬Σ 𝐹⃗ ∙ ⃗⃗⃗⃗𝑑𝑠

𝑛 = 48𝜋, without using divergence theorem. Observe that Σ is a
piecewise smooth surface bounding a cylinder. It is composed of two circular regions and a
circular cylindrical surface. The circular faces form the upper and lower faces of the cylinder.

91
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

Let Σ1, Σ2 & Σ3 be respectively the upper face, the lateral face and the lower face of Σ. Each
of these surfaces can be parameterized as follows:

𝛴 1: x = u, y = v, z = 3; (u, v)𝜖 R where R is the disk in uv-plane with center (0,0) and
radius 2

𝛴 2: x= 2cos u, y = 2 sin u , z = v ; (u, v)𝜖 R where R is the rectangle in uv-plane bounded

by u = 0, u=2𝜋, v = 0 and v=2𝜋

𝛴 3: x = u, y = v, z = 0; (u, v)𝜖 R where R is the disk in uv-plane with center (0,0) and
radius 2

The outward unit surface normal vector to 𝛴 1 and 𝛴 3 is the standard unit vector k and a
𝑟 ×𝑟 〈−2𝑠𝑖𝑛𝑢,2𝑐𝑜𝑠𝑢,0〉×〈0,0,1〉
set of unit normal vectors to 𝛴 2: is given by n =± ‖𝑟𝑢 ×𝑟𝑣‖ =± ‖〈−2𝑠𝑖𝑛𝑢,2𝑐𝑜𝑠𝑢,0〉×〈0,0,1〉‖ = ±
𝑢 𝑣

〈−𝑐𝑜𝑠𝑢, −𝑠𝑖𝑛𝑢, 0〉. For u =0 , the unit vector n has to be the standard unit vector i ; in
view of the assumption that the unit vector has to be outward. Thus n = 〈𝑐𝑜𝑠𝑢, 𝑠𝑖𝑛𝑢, 0〉.

⃗⃗⃗⃗𝑑𝑠 = ∬𝑅〈𝑢, −2𝑣 2 , 9〉 ∙ 〈0,0,1〉 𝑑𝐴 = 36𝜋. Likewise we obtain ∬Σ 𝐹⃗ ∙ ⃗⃗⃗⃗𝑑𝑠

∬Σ1 𝐹⃗ ∙ 𝑛 2
𝑛 =0

and ∬Σ 𝐹⃗ ∙ 𝑛
⃗⃗⃗⃗𝑑𝑠 = 12𝜋.
3

Thus ∬Σ 𝐹⃗ ∙ 𝑛
⃗⃗⃗⃗𝑑𝑠 = ∬Σ 𝐹⃗ ∙ 𝑛
⃗⃗⃗⃗𝑑𝑠 + ∬Σ 𝐹⃗ ∙ 𝑛
⃗⃗⃗⃗𝑑𝑠 ∬Σ 𝐹⃗ ∙ 𝑛
⃗⃗⃗⃗𝑑𝑠 = 36𝜋+0+ 12𝜋= 48𝜋.
1 2 3

The divergence theorem can also be used to evaluate triple integrals by turning them into
surface integrals. This depends on finding a vector field whose divergence is equal to the
given function.

Example 3[using surface integral to compute triple integral]: Let E be the region
defined by 𝑥 2 + 𝑦 2 + 𝑧 2 ≤ 1. Use the divergence theorem to evaluate

∭ 𝑧 2 𝑑𝑣
𝐸

4𝜋
Answer: 15

92
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

Example 4 [Geometrical Application of Divergence Theorem]: Kreyszig Page 463, Q7

Example 5][Physical application of Divergence Theorem]:

Exercise 4.5:

1. Verify the divergence theorem for the vector field 𝐹 = 𝑥𝑖 + 𝑦𝑗 + 𝑧𝑘 over the
sphere S of radius a centred at the origin .
2. Verify the divergence theorem for the vector field 𝐹 = 𝑥𝑖 + 𝑦𝑗 + 𝑧𝑘 S where S is
the boundary of the solid bounded by the coordinate planes and the plane 2𝑥 + 2𝑦 +
2𝑧 = 6 .
3. Verify the divergence theorem for the vector field 𝐹 = 𝑥𝑖 + 𝑦𝑗 + 𝑧𝑘 over the
sphere S of radius a centered at the origin .
4. Verify divergence theorem for the vector field 𝐹 = (4𝑥)𝑖 − (2𝑦 2 )𝑗 + (𝑧 2 )𝑘 taken
over the region bounded by 𝑥 2 + 𝑦 2 = 4, 𝑧 = 0 and 𝑧 = 3.
5. Use Divergence theorem to evaluate ∬𝑆 𝐹⃗ . 𝑑𝑆⃗ where 𝐹 = 3𝑥𝑖 ⃗ + 2𝑦𝑗 ⃗ and S is the
sphere 𝑥 2 + 𝑦 2 + 𝑧 2 = 9.
6. Prove that the volume of a solid T with boundary surface S is equal to ∬𝑆 zdxdy.

93
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

4.6 Stokes’s Theorem

We have examined several versions of the Fundamental Theorem of Calculus in two and
three dimensions that relate the integral around an oriented boundary of a domain to a
“derivative” of that entity on the oriented domain. We now see the final theorem of this sort.
Stokes’s Theorem relates surface integral over a surface with line integral over the
boundary curve of the surface (assuming that the surface is not a closed surface).It was
appeared first in the form of an examination question at Cambridge University. It can be
viewed as a generalization of Green’s theorem; it is analogous to the curl form of Green’s
Theorem.
Stokes’s theorem is useful in computing line integrals using surface integrals or vice
versa. It can also be used in solving certain physical problems like heat conduction.
Since we will be working in three dimensions, we need to discuss what it means for a
boundary curve of a surface in space to be oriented positively.
Definition 1: Let S be an oriented surface with unit normal vector n and let C be the
boundary of S. Then, we say C is positively oriented if its orientation follows the right
hand rule, that is if your right hand curls around n in the direction of C's orientation, then
your thumb will be pointing in the direction of n.

Definition 2(circulation): The circulation of a three-dimensional vector field F around an

oriented closed path C in the direction of unit tangent vector T is

Circulation =

94
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

In short, circulation is the line integral of a vector field around a closed path. The definition
of circulation of a two dimensional vector filed is given in section 4.3.

Theorem 1(Stokes’s Theorem):

Let

i) S be an oriented surface with unit normal vector n and C be the positively oriented
boundary of S.

ii) F be a vector field defined with continuous first order partial derivatives in a three
dimensional region containing S.

Then

∫ 𝑭. 𝒅𝒓 = ∬ 𝑐𝑢𝑟𝑙 (𝑭). 𝒏𝑑𝑠

𝐶 𝑆

(In words: the line integral of the tangential component of F along C equals the surface
integral of the normal component of curl (F) over S.)

Note: The surface integral in Stokes’s theorem is taken over a closed surface hence it is
called a closed surface integral. Closed surface integral is sometimes denoted using the
integral symbol

Example 1 (Using Surface Integral to Compute Line Integral): Let S be the part of the
plane 𝑧 = 4 − 𝑥 − 2𝑦 in first octant and with upward pointing unit normal vector. Use
Stokes' theorem to find ∫𝐶 𝐹. 𝑑𝑟 Where 𝐹 = 𝑦𝑖 + 𝑧𝑗 – 𝑥𝑦𝑘 and C is the boundary of S
with positive orientation.

95
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

Solution: First notice that without Stokes' theorem, we would have to parameterize three
different line segments. Instead we can find this with just one double integral. We have

and 𝒏 𝑑𝑆 = 𝒊 + 2𝒋 + 𝒌

So that 𝐶𝑢𝑟𝑙 𝑭. 𝒏 𝑑𝑆 = 1 + 𝑥 + 2𝑦 − 1 = 𝑥 2𝑦

We integrate

Example 2(Using Line Integral to Compute Surface Integral): Use Stokes’ Theorem to

evaluate ∬S curl(𝐅) . 𝐧ds where F = x 2 𝐢 − 3xy𝐣 + x 3 y 3 𝐤 and S is the part z = 5-x2 –y2 above

the plane z = 1. Assume that S is oriented upwards.

Answer: 0

Example 3 (Verification of Stokes’s Thoerem) Verify Stokes's Theorem for F = y2i - xj +

5zk if S is the paraboloid z = x2 + y2 with the circle x2 + y2 = 1 as its boundary.

96
Applied mathematics III Unit 4. Line Integrals and Surface Integrals Compiled by: Tadesse Bekeshie(PhD)

Answer: ∬S curl(𝐅) . 𝐧ds = −π and ∫C F. dr = −π.

Example 4[Geometrical Application of Stokes’s Theorem]:

Example 5 [Physical Application of Stokes’s Theorem]:

Exercises 4.6:

1. Let S be a solid sphere. Show that ∬𝑆 ∇ × 𝐹. 𝑛⃗⃗𝑑𝑠 = 0 by using Stokes’s theorem

2. Suppose 𝑆1 and 𝑆2 are two oriented surfaces that share C as boundary. What can you
say about ∬𝑆1 𝑐𝑢𝑟𝑙(𝐹). 𝑛𝑑𝑠 and ∬𝑆2 𝑐𝑢𝑟𝑙(𝐹). 𝑛𝑑𝑠 ?
3. Let S be a solid sphere. Show that ∫∫ (∇XF)·n dS = 0 by using Stokes’s theorem
4. Verify Stokes’ theorem for the given vector field 𝐹⃗ and surface Σ where 𝐹⃗ (𝑥, 𝑦, 𝑧): =
⃗⃗ ; 𝛴: 𝑥 2 + 𝑦 2 + 𝑧 2 = 1, 𝑧 ≥ 0.
2𝑦𝑖⃗ − 𝑥𝑗⃗ + 𝑧𝑘
5. Use Stokes theorem to compute the flux of 𝐹⃗ through the upper half ellipsoid 𝑥 2 +
𝑦2 𝑧2
+ = 1, 𝑧 ≥ 0 , oriented by its outward normal.
9 4

97