APSC248 – Engineering Analysis III

Lecture 6 – Vector Functions

𝑃
𝑟⃑(𝑡)

𝑜 𝑦

𝐶
𝑥

Definition of vector function:

Given a vector 𝑟⃑(𝑡) where 𝑡 is a real variable. Suppose that vector 𝑟⃑(𝑡) is defined for each

value of 𝑡 over a given range of 𝑡, i.e., 𝑡 ∈ (𝑎, 𝑏). Then it is said that 𝑟⃑(𝑡) is a “vector

function” of 𝑡 over that interval (𝑎, 𝑏).

Note, 𝑟⃑(𝑡) might represent any physical nature such as position vector a particle or an electric

field vector. Vector 𝑟⃑ can also be a function of arc length 𝑠 along the curve, i.e., 𝑟⃑(𝑠). Or

vector 𝑟⃑ is a function of simply a pure mathematical parameter.

(1) Scalar Function and Vector Function

Scalar functions have forms like 𝑦 = 𝑓(𝑥) or 𝑦 = 𝑓(𝑥! , 𝑥" , … , 𝑥# ). For example,

- 𝑇(𝑥, 𝑦, 𝑧) can be a scalar function for temperature of a location in 3D space;

- 𝑃(𝑥, 𝑦, 𝑧) can be a scalar function for pressure of a location in 3D space;

In 3D space, scalar function tells us how to associate a number (e.g., temperature) with

each point in space.

@Dr. Yang Cao Page 1

 APSC248 – Engineering Analysis III

Vector function is a mathematical function of one or more variables whose range is a

set of multidimensional vectors or infinite-dimensional vectors. Often the input of a

vector-valued function is a scalar, but in general the input can be a vector of either

complex or real variables. Vector functions (in 3d space), is a rule which tells us how

to associate a vector with each point (𝑥, 𝑦, 𝑧) in space. For example,

- Velocity of fluid, 𝑣⃑(𝑥, 𝑦, 𝑧). It specifies the speed of fluid and direction of flow

at a location (𝑥, 𝑦, 𝑧).

- Angular velocity of a rotating body, 𝜔

7⃑(𝑥, 𝑦, 𝑧). It represents angular velocity

of a point on the rotating body.

𝑣⃑(𝑥, 𝑦, 𝑧) means it is a vector function depends on three parameters (or variables). It

can be resolved into components as

𝑣⃑(𝑥, 𝑦, 𝑧) = 𝑣$ (𝑥, 𝑦, 𝑧)𝚤̂ + 𝑣% (𝑥, 𝑦, 𝑧)𝚥̂ + 𝑣& (𝑥, 𝑦, 𝑧)𝑘=

In many cases, parameters 𝑥, 𝑦, and 𝑧 are function of a common variable, such as time

𝑡 . This means vector function 𝑣⃑(𝑥, 𝑦, 𝑧) = 𝑣⃑>𝑥(𝑡), 𝑦(𝑡), 𝑧(𝑡)? = 𝑣⃑(𝑡) becomes a

function of parameter 𝑡. And similarly, it can be resolve into

𝑣⃑(𝑡) = 𝑣$ (𝑡)𝚤̂ + 𝑣% (𝑡)𝚥̂ + 𝑣& (𝑡)𝑘=

@Dr. Yang Cao Page 2

 APSC248 – Engineering Analysis III

In summary, a scalar function is a function 𝑓(𝑥! , 𝑥" , … , 𝑥# ) of one or more variables

whose range is one-dimensional, as compared to a vector function, whose range is

three-dimensional (or, in general, 𝑛-dimensional).

We can relate scalar function and vector function to system concept. The system can

be represented by the function (e.g., 𝑓(∙), 𝑣⃑(∙) etc.). The system can have single input

or multiple inputs. The number of outputs of the system is referred to as “range” of

the function representing the system. If the system has only one output, the function

representing the system is a scalar function since its “range” is one-dimensional. If the

system has multiple outputs, the function representing the system is a vector function.

Example 1: Vector function 𝑟⃑(𝑡) = 〈2 cos 𝑡 , 4 sin 𝑡 , 𝑡〉

The spiral is the path traced by the

tip of the vector as 𝑡 increases from
𝑟⃑(𝑡) zero through 8π.

The vector 𝑟⃑(𝑡) has its tail at the

origin and its head at the
coordinates evaluated by the
functions
𝑥 = 2 cos 𝑡 , 𝑦 = 4 sin 𝑡 , 𝑧 = 𝑡.

@Dr. Yang Cao Page 3

 APSC248 – Engineering Analysis III

(2) Derivative of Vector Function

Given a vector function 𝑟⃑(𝑡), the derivative of vector function 𝑟⃑(𝑡) is defined as

𝑑𝑟⃑(𝑡) 𝑟⃑(𝑡 + Δ𝑡) − 𝑟⃑(𝑡)

= 𝑟⃑ ' (𝑡) ≡ lim
𝑑𝑡 ()→+ Δ𝑡

If the limit on the right side exists for a given value of 𝑡 , we say that 𝑟⃑(𝑡) is

-.⃑())
differentiable at that 𝑡, and that 𝑟⃑ ' (𝑡) or -)
is its derivative.

Let 𝑟⃑(𝑡) = 𝑥(𝑡)𝚤̂ + 𝑦(𝑡)𝚥̂ + 𝑧(𝑡)𝑘=, then its derivative is still a vector.

𝑑𝑟⃑(𝑡) 𝑑𝑥 𝑑𝑦 𝑑𝑧
𝑟⃑ ' (𝑡) = = 𝚤̂ + 𝚥̂ + 𝑘=
𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡

Example 2: Position vector function 𝑟⃑(𝑡) = 3 cos 𝑡 𝚤̂ + 3 sin 𝑡 𝚥̂ + 𝑡 ! 𝑘9 (A

spiral)

The velocity vector is

𝑑𝑟⃑(𝑡)
𝑣⃑(𝑡) = = −3 sin 𝑡 𝚤̂ + 3 cos 𝑡 𝚥̂ + 2𝑡𝑘=
𝑑𝑡

The acceleration vector is

𝑑 " 𝑟⃑(𝑡)
𝑎⃑(𝑡) = = −3 cos 𝑡 𝚤̂ − 3 sin 𝑡 𝚥̂ + 2𝑘=
𝑑𝑡 "

The speed is

|𝑣⃑(𝑡)| = P9 cos " 𝑡 + 9 sin" 𝑡 + 4𝑡 " = P9 + 4𝑡 "

To find times when 𝑣⃑ and 𝑎⃑ are orthogonal

𝑣⃑ ∙ 𝑎⃑ = 9 sin 𝑡 cos 𝑡 − 9 cos 𝑡 sin 𝑡 + 4𝑡 = 4𝑡 = 0

Thus, the only time when 𝑣⃑ and 𝑎⃑ are orthogonal is when 𝑡 = 0.

@Dr. Yang Cao Page 4

 APSC248 – Engineering Analysis III

(3) Rules of Vector Differentiation

𝑑
7⃑(𝑡)] = 𝑓 ' (𝑡)𝑢
[𝑓(𝑡)𝑢 7⃑' (𝑡)
7⃑(𝑡) + 𝑓(𝑡)𝑢
𝑑𝑡

𝑑 𝑑 𝑑
[𝑢
7⃑(𝑡) ± 𝑣⃑(𝑡)] = 𝑢 7⃑' (𝑡) ± 𝑣⃑ ' (𝑡)
7⃑(𝑡) ± 𝑣⃑(𝑡) = 𝑢
𝑑𝑡 𝑑𝑡 𝑑𝑡

𝑑
[𝑢 7⃑' (𝑡) ∙ 𝑣⃑(𝑡) + 𝑢
7⃑(𝑡) ∙ 𝑣⃑ (𝑡)] = 𝑢 7⃑(𝑡) ∙ 𝑣⃑ ' (𝑡)
Product rule _ 𝑑𝑡
𝑑
[𝑢7⃑(𝑡) × 𝑣⃑(𝑡)] = 𝑢 7⃑' (𝑡) × 𝑣⃑(𝑡) + 𝑢 7⃑(𝑡) × 𝑣⃑ ' (𝑡)
𝑑𝑡

𝑑 𝑑𝑢
7⃑ 𝑑𝑥
Chain rule: e𝑢
7⃑>𝑥(𝑡)?f =
𝑑𝑡 𝑑𝑥 𝑑𝑡

𝑑
[𝑢 7⃑' (𝑡) ∙ 𝑣⃑(𝑡) × 𝑤
77⃑(𝑡)] = 𝑢
7⃑(𝑡) ∙ 𝑣⃑(𝑡) × 𝑤 7⃑(𝑡) ∙ 𝑣⃑ ' (𝑡) × 𝑤
77⃑(𝑡) + 𝑢 77⃑(𝑡) + 𝑢 77⃑ ' (𝑡)
7⃑(𝑡) ∙ 𝑣⃑(𝑡) × 𝑤
𝑑𝑡

;⃑(𝑡) = 𝑡𝚤̂ + 𝑡 ! 𝚥̂ − 3𝑘9 ; 𝑣⃑ (𝑡) = 𝑡𝚤̂ + 𝑡𝚥̂ + 𝑡𝑘9

Example 3: 𝑢

𝑑
[𝑢 7⃑' (𝑡) ∙ 𝑣⃑(𝑡) + 𝑢
7⃑(𝑡) ∙ 𝑣⃑(𝑡)] = 𝑢 7⃑(𝑡) ∙ 𝑣⃑ ' (𝑡)
𝑑𝑡

i ? + e𝑡𝑖h + 𝑡2 =𝑗 − 3𝑘
= >𝑖h + 2𝑡𝑗=? ∙ >𝑡𝑖h + 𝑡𝑗= + 𝑡𝑘 i f ∙ >𝑖h + 𝑗= + 𝑘
i?

= (𝑡 + 2𝑡 " ) + (𝑡 + 𝑡 " − 3)

= 3𝑡 " + 2𝑡 − 3

@Dr. Yang Cao Page 5

 APSC248 – Engineering Analysis III

Example 4: vector function of constant length, |𝑟⃑(𝑡)| = 𝑐

Position vector representing a particle moving on the surface of a sphere will always have

a constant length equal to the radius of the sphere.

𝑟⃑(𝑡) ∙ 𝑟⃑(𝑡) = |𝑟⃑(𝑡)|" = 𝑐 "

𝑑
[𝑟⃑(𝑡) ∙ 𝑟⃑(𝑡)] = 0 ⟹ 𝑟⃑ ' (𝑡) ∙ 𝑟⃑(𝑡) + 𝑟⃑(𝑡) ∙ 𝑟⃑ ' (𝑡) = 0
𝑑𝑡

𝑑𝑟⃑
2𝑟⃑ ' (𝑡) ∙ 𝑟⃑(𝑡) = 0 ⇒ 𝑟⃑ ∙ =0
𝑑𝑡

The vectors 𝑟⃑ ' (𝑡) and 𝑟⃑(𝑡) are orthogonal, because their dot product is 0. In summary, if

-.⃑
𝑟⃑(𝑡) is a differentiable vector function of 𝑡 of constant length, then 𝑟⃑ ∙ -) = 0. Similarly, if

-.⃑ -.⃑ -.⃑

𝑟⃑ is a function of the arc length, we have 𝑟⃑(𝑠) ∙ -2 = 0. -2
and -)
point to the same

#⃑
!" #⃑
!"
-.⃑ -.⃑
direction, but have different length. Note that -2 is a unit vector since -2 = !%
!& = !%
#⃑
!"
.
!%
3 3
!%

(4) Integrals of Vector Function

Let 𝑟⃑(𝑡) = 𝑥(𝑡)𝚤̂ + 𝑦(𝑡)𝚥̂ + 𝑧(𝑡)𝑘=, then

m 𝑟⃑(𝑡)𝑑𝑡 = m 𝑥(𝑡)𝑑𝑡 𝚤̂ + m 𝑦(𝑡)𝑑𝑡 𝚥̂ + m 𝑧(𝑡)𝑑𝑡 𝑘=

Or if 𝑅7⃑ is any antiderivative of 𝑟⃑(𝑡), then

m 𝑟⃑(𝑡)𝑑𝑡 = 𝑅7⃑ (𝑡) + 𝐶⃑

@Dr. Yang Cao Page 6

 APSC248 – Engineering Analysis III

Example 5: 𝑟⃑(𝑡) = 2𝑡𝚤̂ − 𝚥̂ + 3𝑡 ! 𝑘9 , ∫ 𝑟⃑(𝑡) 𝑑𝑡 =?

m 𝑟⃑(𝑡)𝑑𝑡 = m>2𝑡𝚤̂ − 𝚥̂ + 3𝑡 " 𝑘= ?𝑑𝑡

= 𝑡 " 𝚤̂ − 𝑡𝚥̂ + 𝑡 4 𝑘= + 𝐶⃑

where 𝐶⃑ = 𝑐! 𝚤̂ + 𝑐" 𝚥̂ + 𝑐4 𝑘= is a constant vector.

(5) Tangent Vector, Unit Tangent Vector, and Unit Normal Vector

A tangent vector is a vector that follows the direction of a curve at a given point.

𝑧
7⃑(𝑡)
𝑇

𝑃
𝑟⃑(𝑡)

𝑜 𝑦

𝑥 𝐶

Suppose that curve 𝐶 is represented by vector function 𝑟⃑(𝑡) = 𝑥(𝑡)𝚤̂ + 𝑦(𝑡)𝚥̂ + 𝑧(𝑡)𝑘=.

Then the derivative of 𝑟⃑(𝑡) is the tangent vector of curve 𝐶 at point 𝑃 (in the direction of

increasing 𝑡). Note that if you think of 𝑟⃑(𝑡) as a position vector, derivative of 𝑟⃑(𝑡) gives the

velocity vector and velocity vector is always tangent to the trajectory.

7⃑(𝑡), then
Let tangent vector be denoted as 𝑇

𝑑𝑟⃑(𝑡) 𝑑𝑥 𝑑𝑦 𝑑𝑧
7⃑ (𝑡) =
𝑇 = 𝚤̂ + 𝚥̂ + 𝑘=
𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡

@Dr. Yang Cao Page 7

 APSC248 – Engineering Analysis III

Unit tangent vector is then

7⃑(𝑡)
𝑇 𝑟⃑ ' (𝑡)
𝑇=(𝑡) = = '
7⃑(𝑡)o |𝑟⃑ (𝑡)|
o𝑇

-.⃑
We can also show that 𝑇=(𝑡) = -2
. Proof is given below

𝑑𝑟⃑ 𝑑𝑟⃑ 𝑑𝑡 1
= = 𝑣⃑ = 𝑇=
𝑑𝑠 𝑑𝑡 𝑑𝑠 |𝑣⃑|

Principal unit normal vector

There is one vector of particular significant among the vectors orthogonal to the unit

tangent vector (as shown in the diagram below). This vector points in the direction in

which the curve is turning. Since 𝑇= has constant length, from Example 4, we have

𝑑𝑇=(𝑡)
⋅ 𝑇=(𝑡) = 0
𝑑𝑡
-56
7⃑(𝑡) =
We obtain a vector 𝑁 which orthogonal to 𝑇=(𝑡). Its unit vector is
-)

7⃑
𝑁 𝑑𝑇=⁄𝑑𝑡
i=
𝑁 =
7⃑ o o𝑑𝑇=⁄𝑑𝑡o
o𝑁

𝑁 is called principal unit normal vector for the smooth curve in plane. The following

diagram illustrates the relationship between the curve, the unit tangent vector and the unit

normal vector.

𝑑𝑇=⁄𝑑𝑡 𝑟⃑(𝑡)
i=
𝑁
o𝑑𝑇=⁄𝑑𝑡o

𝑇=
𝑇=

𝑑𝑇=⁄𝑑𝑡
i=
𝑁
o𝑑𝑇=⁄𝑑𝑡o

@Dr. Yang Cao Page 8

 APSC248 – Engineering Analysis III

Binormal vector

The binormal vector is defined as

𝐵= = 𝑇= × 𝑁
i

𝐵= is a unit vector.

Normal Plane of curve 𝐶

The plane through a given point on curve 𝐶 and normal to vector 𝑇=, is called the normal

plane of curve 𝐶.

Remarks:
6
-5 6
-5⁄-)
• 𝑇= is an unit vector, but one can’t conclude that -) is a unit vector. Of course, |-56⁄-)|

guarantees to be a unit vector. You can use 𝑇= = cos 2𝑡 𝚤̂ + sin 2𝑡 𝚥̂ to verify.

• To relate to a physical meaning, we can also say tangent vector gives us the velocity

7⃑ = -.⃑()) and acceleration would be 𝑎⃑ = -9:⃑ = -5:⃑.

𝑣⃑ = 𝑇 -) -) -)

:⃑
-5 -56
7⃑ (𝑡) ≠
Note that normal vector 𝑁 7⃑(𝑡) =
, instead, 𝑁 .
-) -)

As pointed out in Textbook on page 735, we can decompose acceleration vector as

𝑎⃑ = 𝑎 5 𝑇= + 𝑎; 𝑁
i

@Dr. Yang Cao Page 9

 APSC248 – Engineering Analysis III

Example 6: A roller coaster based on the parametric equations of a 3D

Lissajous curve
𝑟⃑(𝑡) = sinE𝑠(𝑡)G 𝚤̂ + sinE2𝑠(𝑡)G 𝚥̂ + cosE𝑠(𝑡)G 𝑘9
𝑠(𝑡) represents the distance along the track. Obtain the velocity of the
roller coaster.

Solution:

𝑥(𝑡) = sin>𝑠(𝑡)? ; 𝑦(𝑡) = sin>2𝑠(𝑡)? ; 𝑧(𝑡) = cos>𝑠(𝑡)?

𝑑𝑥(𝑡) 𝑑𝑠
= cos>𝑠(𝑡)? = cos>𝑠(𝑡)? 𝑠̇
𝑑𝑡 𝑑𝑡

𝑑𝑦(𝑡) 𝑑𝑠
= 2 cos>2𝑠(𝑡)? = 2 cos>2𝑠(𝑡)? 𝑠̇
𝑑𝑡 𝑑𝑡

𝑑𝑧(𝑡) 𝑑𝑠
= − sin>𝑠(𝑡)? = − sin>𝑠(𝑡)? 𝑠̇
𝑑𝑡 𝑑𝑡

The velocity is then

𝑑𝑟⃑(𝑡) i ?𝑠̇
𝑣⃑(𝑡) = = >cos>𝑠(𝑡)? 𝑖h + 2 cos>2𝑠(𝑡)? 𝑗= − sin>𝑠(𝑡)? 𝑘
𝑑𝑡

@Dr. Yang Cao Page 10

 APSC248 – Engineering Analysis III

Example 7: 𝑟⃑(𝑡 ) = 𝑡 ! 𝚤̂ + 3𝑡𝚥̂ + 5𝑘9 , Obtain the unit tangent vector and
unit normal vector at 𝑡 = 0.

Solution:
𝑑𝑟⃑(𝑡)
7⃑(𝑡) =
𝑇 = 2𝑡𝚤̂ + 3𝚥̂
𝑑𝑡
7⃑(𝑡)
𝑇 1
𝑇=(𝑡) = = (2𝑡𝚤̂ + 3𝚥̂)
7⃑(𝑡)o √4𝑡 " + 9
o𝑇

𝑑𝑇=(𝑡) 2𝑡 '
3 '
=y z 𝚤̂ + y z 𝚥̂
𝑑𝑡 √4𝑡 " + 9 √4𝑡 " + 9

< ' <' 9=9 ' <

Recall the quotient rule, e9 f = 9(

1 !
2𝑡 ' 2√4𝑡 " + 9 − y2 (4𝑡 " + 9)=" 8𝑡z (2𝑡) 2(4𝑡 " + 9) − 8𝑡 " 18
y z = = 4 = 4
√4𝑡 " + 9 4𝑡 " +9 (4𝑡 " + (4𝑡 " + 9)"
9)"
'
3 𝑑 ! 1 4 −12𝑡
y z =3 y(4𝑡 " + 9 )=" z = 3 y− z (4𝑡 " + 9 )=" (8𝑡) = 4
√4𝑡 " + 9 𝑑𝑡 2
(4𝑡 " + 9)"

𝑑𝑇=(𝑡) 1 6
= 4
(18𝚤̂ − 12𝑡𝚥̂) = 4
(3𝚤̂ − 2𝑡𝚥̂)
𝑑𝑡 (4𝑡 + 9)
" " (4𝑡 + 9)
" "

𝑑𝑇=(𝑡) 6
} }= P
4 9 + 4𝑡
"
𝑑𝑡 (4𝑡 " + 9)"

𝑑𝑇=⁄𝑑𝑡 1
i=
𝑁 = (3𝚤̂ − 2𝑡𝚥̂)
o𝑑𝑇=⁄𝑑𝑡 o √9 + 4𝑡 "

At 𝑡 = 0, we have

1 1
𝑇=(𝑡) = (2𝑡𝚤̂ + 3𝚥̂)~ = (0𝚤̂ + 3𝚥̂) = 𝚥̂
√4𝑡 " + 9 )>+
3

1
i=
𝑁 (18𝚤̂ − 12𝑡𝚥̂)~ = 𝚤̂
√324 + 144𝑡 " )>+

@Dr. Yang Cao Page 11

 APSC248 – Engineering Analysis III

(6) Curvature of a Curve (optional)

Consider the unit tangent vector 𝑇=(𝑡) of a curve at point 𝑃 which moves on a smooth

curve. Since 𝑇=(𝑡) is unit vector, its length remains constant and only its direction changes

as the point 𝑃 is moving along the curve. The rate at which 𝑇=(𝑡) turns per unit of length

along the curve is call the curvature.

Definition:

If 𝑇=(𝑡) is the unit vector of a smooth curve, the curvature function of the curve is

𝑑𝑇= 𝑑𝑇= 𝑑𝑡 1 𝑑𝑇= 1 𝑑𝑇=

𝜅=} }=} }= } }= } }
𝑑𝑠 𝑑𝑡 𝑑𝑠 |𝑑𝑠/𝑑𝑡| 𝑑𝑡 |𝑣⃑| 𝑑𝑡

𝜅 is a Greek letter pronounced “kappa”.

Comments:
6
-5
• If • -2 • is large, 𝑇= turns sharply around point 𝑃 location, and the curvature at 𝑃 is

large.

-56
• If • -2 • is close to zero, 𝑇= turns slowly and the curvature at 𝑃 is small.

Example: Curvature of a circle, 𝑟⃑(𝑡 ) = (𝑎 cos 𝑡 )𝚤̂ + (𝑎 sin 𝑡)𝚥̂

𝑑𝑟⃑
𝑣⃑ = = −(𝑎 sin 𝑡)𝚤̂ + (𝑎 cos 𝑡)𝚥̂
𝑑𝑡

𝑣⃑
𝑇= = = −(sin 𝑡)𝚤̂ + (cos 𝑡)𝚥̂
|𝑣⃑|

𝑑𝑇=
= −(cos 𝑡)𝚤̂ − (sin 𝑡)𝚥̂
𝑑𝑡

1 𝑑𝑇= 1 1
𝜅= } }= =
|𝑣⃑| 𝑑𝑡 𝑎 radius

@Dr. Yang Cao Page 12

 APSC248 – Engineering Analysis III

i in terms of radius of curvature 𝜅.

We can also obtain the principal unit normal vector 𝑁

Consider the unit tangent vector 𝑇= as a function of arc length 𝑠. Since 𝑇=(𝑠) has constant

-56
length, then the derivative -2
is orthogonal to 𝑇= . The principal unit normal vector 𝑁
i is

-56
obtained by normalizing -2
, i.e.,

𝑑𝑇=⁄𝑑𝑠 1 𝑑𝑇=
i=
𝑁 =
o𝑑𝑇=⁄𝑑𝑠o 𝜅 𝑑𝑠

The vector 𝑑𝑇= ⁄𝑑𝑠 points in the direction in which 𝑇= turns as the curve bends. Therefore, if

we face in the direction of increasing arc length, the vector 𝑑𝑇=⁄𝑑𝑠 points toward the right if

𝑇= turns clockwise and toward the left if 𝑇= turns counterclockwise. In other words, the

i will point toward the concave side of the curve.

principal normal vector 𝑁

@Dr. Yang Cao Page 13