MA10208B – Methods & Applications 1B

Dr Martin Reed

Academic Year 2010/11; Version: January 28, 2011

1 Vectors

This course develops the theory of vectors – a major mathematical tool for modelling and
analysis of various “real world” (physical, mechanical etc.) applications. Vectors have numerous
other applications too, for example, in electromagnetic waves which are important in the design
of mobile phones.
Mechanics deals with the action of forces on bodies and includes both statics and the study of
motion. We shall consider Newtonian Mechanics, including,

• the motion of the planets and satellites,

• projectiles,

• circular orbits.

Before looking at the applications we need to understand the tools, namely vectors, and so the
first few sections are about Vector Algebra and then Vector Calculus.
Text book - Anton “Calculus” 8th Edition This is a very readable book, has lots of
exercises, and I’ll assume everyone has a copy. However, we’ll not follow Anton all the time,
and also we’ll cover certain topics differently. You should read both these notes and the relevant
sections in Anton.
Warning These typed notes do not cover the whole course MA10208B. Some additional material
will be given in the lectures. Also, when the time comes, students may be advised to read parts
of Anton (§12 and §13) and make their own notes.

Acknowledgement These notes have evolved from various units taught in the first year at the
University of Bath over several years. Previous lecturers include: John Willis, Ray Ogden, Keith

1
Walton, Victor Galaktionov, Alastair Spence and Valery Smyshlyaev. These notes may be used
in other courses provided notification is given to Martin Reed (email: mbr20 @ bath.ac.uk).
c The Department of Mathematical Sciences, The University of Bath, February 2011.

1.1 Introduction to Vector Algebra

Some quantities in Mechanics and Physics are characterised by a single real number, for example,
mass, temperature. However other physical quantities require the specification of both a magni-
tude and direction, for example, velocity and force (see Anton p 797). We shall assume that we
live in a 3-Dimensional (3-D) Euclidean Space R3 , and for this course, vectors are 3-Dimensional.
To give a precise mathematical definition we adopt the axioms of Euclidean geometry, namely,
points, straight lines, planes, parallel lines, distance, etc.

Definition 1.1 A vector is a directed line segment (DLS) characterised by two “ordered points”
in 3-D Euclidean space, P and Q, and is visualised as an arrow joining P to Q. P is the initial
−−→
point and Q is the terminal point. The vector is denoted P Q.

Every vector has a “direction” and a “length” associated with it:

−−
→ −−→
Definition 1.2 The length of P Q is the distance between P and Q, and is denoted by kP Qk.
−−→ −−
→
Often we say kP Qk is the norm of P Q.

Definition 1.3 If P and Q coincide (that is P = Q) then the vector is called the zero segment
−
→
(or zero vector), and is denoted 0 , 0 or 0.

From the Euclidean distance axioms, namely, (a) the distance from a point to itself is zero,
and (b) the distance between any two distinct points is strictly positive, we have the following
“obvious” result.

−−→ − → −−
→
Proposition 1.4 P Q = 0 ⇐⇒ kP Qk= 0.
−−→ − → −−→
Proof: P Q = 0 ⇐⇒ P = Q ⇐⇒ kP Qk= 0. 2

2
 −−→ −→
Definition 1.5 Two directed line segments (DLSs) P Q and RS are said to be parallel (or
collinear) if they lie on parallel straight lines L1 and L2. The zero segment is regarded as being
parallel to any vector.

or

L1 L2 L1 L2

An important idea is that two DLSs which have the same direction and magnitude are “equal”
or “equivalent”, and we do not distinguish between them. This leads to the following definition.

−−→ −→ −−
→ −→
Definition 1.6 Two DLSs P Q and RS are said to be equal ( equivalent), that is, P Q = RS, if

1. They are collinear.

−−→ −→
2. They have the same magnitude, that is, kP Qk=kRSk.

3. They point in the same direction.

Note that this definition doesn’t mention the initial point of a DLS, so a DLS can be “freely”
−−→
moved, in that, for any DLS P Q and for any other point R, there exists an equal (and unique)
DLS with initial point at R. Note that all zero DLSs are regarded as being equivalent.
Notation It is convenient to have a single label to describe this, for example,

−−→ −→
a = P Q = RS,

indicating that we don’t distinguish between two equivalent DLSs. Hence, a represents a “class”
−−→
of equivalent DLS’s each with the same magnitude and direction as P Q.
Finally, kak will denote the length of the vector a.

1.2 Basic operations with vectors

In physical problems we often need to evaluate the combined effect of two processes characterised
by two vectors, say a and b, that “follow one after the other” in some sense. This leads us to
the first basic operation, namely, addition of two vectors.

3
 displacement b final position
due to wind in unit of time
a c
position of bird in unit of time
initial position if there were no wind
of bird

Example: Wind is blowing with velocity a and a bird is flying with velocity b with respect to
the air. What is the velocity, say c, of the bird with respect to an observer on the ground. (As
usual the ‘velocity’ is defined to be the change in position (displacement) per unit of time.)
This example motivates the following definition:

Definition 1.7 Given two vectors a and b. Position b so that its initial point coincides with
the terminal point of a. Then c = a + b is defined by the DLS whose initial point coincides with
the initial point of a and whose terminal point coincides with the terminal point of b.

b
a
c=a+b

Exercise: Check that this definition is consistent with the following equivalences. If

−−−→ −−−→ −−−→ −−−→

P1 Q1 = P2 Q2 , Q1 S1 = Q2 S2

then we need to show that

−−−→ −−−→ −−−→ −−−→
P1 Q1 + Q1 S1 = P2 Q2 + Q2 S2 .

Property 1 Commutativity of addition: a + b = b + a

−−→ −→
Proof: Let a = P Q, b = P R. Construct the parallelogram P QSR (see diagram).
Then, from the definition of equivalence of vectors, and the properties of parallelograms

−→ −−→ −→ −→
RS = P Q = a, QS = P R = b.

−−
→ −→ −→
Thus, from the definition of addition, P Q + QS = P S, or

a + b =: c.

On the other hand,

−→ −→ −→
P R + RS = P S = c

4
 Q b S

a a

P b R

Q b S S

a c = a+b c=b+a
a

P P b R

and so
b+a=c

Hence,
a+b=b+a ∀a , b .

Property 2 Associativity of addition: Let a, b and c be three vectors. Then

(a + b) + c = a + (b + c) .

(See Tutorial Sheet 1, Question 1).

−−
→
Property 3 The zero vector: Consider a zero segment, namely, P P . All zero segments are
regarded as equivalent and we write
−
−→ −−→
P P = QQ = 0.

Since by definition
−−→ −−→ −−→ → −−
−− →
P Q + QQ = P Q = P P + P Q,

it follows that
a+0 = a = 0+a .
−−→
Property 4 Negation of vectors: Given a vector a = P Q we denote the negative of a by −a,
and define it by reversing its direction, so that,
−−→
−a = QP .

Then, clearly,
−−→ −−→ −− →
a + (−a) = P Q + QP = P P = 0.

5
and so
a + (−a) = 0 .

Definition 1.8 (Subtraction)

a − b := a + (−b).
−−→ −→
If a = P Q and b = P R, then
T −b Q

a−b a−b
a

P b R

−−
→ −→ −→
a − b = a + (−b) = P Q + QT = P T .
−→ − −→ −
−→
Now, since P RQT is a parallelogram, P T = RQ and hence RQ = a − b.

Next we introduce the concept of multiplication by a scalar (or scalar multiplication). Consider
a + a. It is natural to write a + a = 2a. The vector 2a has the same direction as a but is twice
as long:
k2ak = 2 kak .

a
Q

a
P

−a

S
−−→ −→
P Q = a, P R = 2a

On the other hand, it is natural to write

−→
−a = (−1)a , P S = (−1)a.

Thus, multiplication by −1 results in altering the direction of a vector. This motivates the
following definition of scalar multiplication.

6
Definition 1.9 For any λ ∈ R, the scalar multiplication of a by λ produces the vector b = λa
such that

1. b is collinear to a,

2. kbk= |λ| kak,

3. b and a have the same direction if λ > 0, and opposite if λ < 0. (If λ = 0 then b =
0 × a = 0.)

Property 5 Associativity of scalar multiplication: λ(µa) = (λµ)a

Property 6 Distributivity of scalar multiplication with respect to addition of vectors: λ(a + b) = λa + λb
Property 7 Distributivity of scalar multiplication with respect to addition of scalars: (λ + µ)a = λa + µa
Property 8 1 · a = a
(Properties 5 - 8 can be directly derived from the above definitions.)
Exercises:

(a) Draw a figure to show that Property 6 is true. (Hint: Think similar triangles.)

(b) Give arguments to show that Property 5 and Property 7 are true.

Eight Rules of Vector Algebra (or Vector Arithmetic) (see Anton Th 12.2.6, p800)

1. a + b = b + a

2. (a + b) + c = a + (b + c)

3. a + 0 = 0 + a = a

4. a + (−a) = 0

5. λ(µa) = (λµ)a

6. λ(a + b) = λa + λb

7. (λ + µ)a = λa + µa

8. 1 · a = a

7
These 8 properties are satisfied by directed line segments (with addition, negation and scalar
multiplication introduced above). More generally, (1) - (8) may be regarded as “Axioms”, and
quantities that satisfy them are called vectors. The directed line segment is merely one example
of an object satisfying (1) - (8). Other examples are matrices and functions, among others.
Remark Many physical quantities (e.g. velocity, force) are indeed vectors. However not all
quantities having both a magnitude and a direction satisfy these rules. For example consider the
operation of rotation: the axis of rotation gives the direction, the angle is the magnitude. Let
“addition” be superposition of rotations. Then the commutative property 1 (namely a + b =
b + a) does not hold.

1.3 Unit vectors

Definition 1.10 A vector whose length is 1 is called a unit vector. Given a non-zero vector a,
that is a 6= 0, the unit vector in the direction of a is
1
â = a.
kak
1 kak
Check: kâk = kak kak = kak = 1.

1.4 Position Vectors:

A

Let O be the origin (which is a fixed point) in R3 . Any point A can be identified with its
position vector a by
−→
a := OA.

The main concept to remember here is that the initial point of any position vector is always
the origin. It follows that for points A and B
−
−→ −→ −−→ −−→ −→
AB = AO + OB = OB − OA = b − a

(in the obvious notation). Position vectors are often useful for solving problems in geometry
(see §2 of these notes).
Useful “rules”

8
 A B

a b

O
A C B

c
a
b

1. Let the position vectors of A and B be a and b respectively.

−
−→
Then AB = b − a

−−→
2. Let C be the midpoint of AB and so c is the position vector of OC
−→ −→ −→ −−
→
∴ c = OA + AC = OA + 12 AB =
= a + 21 (b − a) = 21 (a + b).

Now let’s consider two examples.

Example 1: Assume a and b are non-zero and non-collinear, and let λ, µ be real numbers.
Then
λa + µb = 0 ⇒ λ = 0, µ = 0.

Proof First, assume λ 6= 0. Then a = − µλ b, implying a and b are collinear: a contradiction.

Thus λ = 0. Hence
µb = 0,

and since b 6= 0, we must have µ = 0.2

Example 2: Prove that the diagonals of a parallelogram bisect each other.

A C
!
QQ

!!



Q !
a
Q P !!

Q!
!

!! Q


!! Q

Q

!!
! -

Q
O b B

9
Diagonals are:

−−→
AB = b − a
−−→
OC = b + a

Let diagonals intersect at P . So

−→ −−
→
AP = λAB = λ(b − a) (for some λ ∈ R)
−−→
and OP = µ(b + a) (for some µ ∈ R).

(We need to show that λ = 21 , µ = 12 .)

Now

−→ −−→ −→
OA = OP + P A

∴ a = µ(b + a) − λ(b − a)

(1 − µ − λ)a + (λ − µ)b = 0

Now use the result in the previous example:


∴ 1−µ−λ=0  1
λ=µ= .
and λ − µ = 0  2

1.5 Vectors in Cartesian coordinates (Reference: Anton §12.1)

Although vectors are entities in their own right, treating them within a coordinate system
is often very useful. To do this in three-dimensional Euclidean space we introduce a right-
handed mutually orthogonal coordinate frame (Oxyz) with origin O, and with coordinate planes
Oxy, Oyz and Ozx. (See diagram following, and Anton §12.1.)
In general, three non-zero vectors a, b, c (which have the same initial point and are not coplanar)
form a right-handed co-ordinate system if a right-threaded screw rotated from a to b through
an angle less than 180o will advance in the direction c. (See page 4. in “Theory and Problems
of Theoretical Mechanics” by Spiegel (in library).)
Let i, j, k be the unit vectors pointing in the positive directions of Ox, Oy, Oz respectively. Let
a be an arbitrary vector. We can always decompose a into a “linear combination” of i, j, k as
follows.

10
 y
y
j

O i
x x
z

k
z

Proposition 1.11 For any a, there exists a1 , a2 , a3 ∈ R such that a = a1 i + a2 j + a3 k. The

a1 , a2 , a3 are found uniquely.

Definition 1.12 The a1 , a2 , a3 are called components of a, and we write a = (a1 , a2 , a3 ).

Proof: In part (1) we prove existence, and then in part (2) show uniqueness.

1. Existence (geometrical):
−→
Let A be such that a = OA. Construct a “box” (rectangular parallelepiped) with faces
parallel to the coordinate planes
y

j A
O x
i B
k

−→ −→ −−→ −−→ −→ −−→

Let a = OA. Now a = OA = OB + BC + CA, and since OB is parallel to i, there exists
−−→ −−→ −→
a1 such that OB = a1 i. Similarly BC = a2 j, and CA = a3 k. Therefore

a = a1 i + a2 j + a3 k.

2. Uniqueness (algebraic):

11
 Assume there exists another representation

a = a′1 i + a′2 j + a′3 k,

for some a′1 , a′2 , a′3 ∈ R. Subtracting gives

(a1 − a′1 )i + (a2 − a′2 )j + (a3 − a′3 )k = 0.

Assume, without loss of generality, a1 6= a′1 . Then

a2 − a′2 a3 − a′3
i=− j − k.
a1 − a′1 a1 − a′1

Therefore i is “co-planar” to j and k (i.e. i lies in same plane as j & k). But i is orthogonal
to the Oyz plane, a contradiction. 2

Examples of the use of components:

1. Length:
1
kak = (a21 + a22 + a23 ) 2 by Pythagoras’ theorem (twice) :
−→ −−→ −→  −−→ −−→  −→
kak2 = kOAk2 = kOCk2 + kCAk2 = kOBk2 + kBCk2 + kCAk2 = a21 + a22 + a23 .

2. Addition:
a + b = (a1 i + a2 j + a3 k) + (b1 i + b2 j + b3 k)

= (a1 + b1 )i + (a2 + b2 )j + (a3 + b3 )k

= (a1 + b1 , a2 + b2 , a3 + c3 ) (∗)

3. Multiplication by scalar:

λa = λ(a1 i + a2 j + a3 k) = (λa1 )i + (λa2 )j + (λa3 )k = (λa1 , λa2 , λa3 ) (∗∗)

Exercise: Which of the laws of vector algebra have been used in (*) and (**) above?
Remark Anton starts by discussing coordinates in §12.1 and then discusses vectors in §12.2.
We talk about vectors first and then go on to talk about coordinates.

12
1.6 The dot product (scalar product)

Definition 1.13 Given two vectors a and b, their dot (or scalar) product (denoted a.b) is a
scalar defined by


 kakkbk cos θ, if a 6= 0 and b 6= 0
a.b =
 0 , if a = 0 or b = 0,
where θ is the angle between a and b (0 ≤ θ ≤ π).

For a 6= 0, b 6= 0, we have
a.b
cos θ = ,
||a|| ||b||
and so the dot product tells us about the angle between two vectors. In fact, if a.b = 0, then a
is orthogonal (perpendicular) to b. (See also Anton pp.809-811.) The converse is also true, so
we have the result that “two non-zero vectors a and b are orthogonal if and only if a.b = 0”.
The dot product will be used extensively as a mathematical tool in the manipulation of vectors.
However it also has meaning in many physical situations, for example, if a represents a force
and b represents a displacement, then W = a.b is the work done. (See the discussion in Anton
on work on pp813-814.)
The component (or “algebraic”) form of the dot product
From the cosine rule

ka − bk2 = kak2 + kbk2 − 2 kak kbk cos θ

| {z }
:= a.b
Therefore,

1

a.b = kak2 + kbk2 − ka − bk2
2
1 2 
= a1 + a22 + a23 + b21 + b22 + b23 − (a1 − b1 )2 − (a2 − b2 )2 − (a3 − b3 )2 = a1 b1 + a2 b2 + a3 b3
2

Thus a.b = a1 b1 + a2 b2 + a3 b3 . (†)

(Note that we define the dot product in a geometric manner and then deduce the component
form. Anton does the reverse - the dot product is defined in Def 12.3.1 and then the geometric
form is deduced in Theorem 12.3.3.)
Properties of the dot product: Given the vectors a, b, c, and for any scalar λ:

1. a.b = b.a

13
 2. a.(b + c) = a.b + a.c

3. λ(a.b) = (λa).b = a.(λb)

4. a.a = kak2

5. a.0 = 0

Remark These properties can also be used as axioms in an abstract theory of vectors. If we
proceed along this line, then we first define a “Linear Space” using the addition and scalar
multiplication properties we’ve seen before, and then introduce an “inner product” based on the
rules for the dot product. You’ll meet all this in your later courses.)
Exercise:

(a) Derive (1) and (4) from the definition of a.b

(b) Derive (2) and (3) from the component form of the dot product.

The dot product is used frequently when solving problems. Here are some examples.
Example 1: (Similar to the example considered in lectures) Let a = 2i−j+k, and b = i+j+2k.
Find the angle θ between a and b.
From(†), a.b = 2 × 1 + (−1) × 1 + 1 × 2 = 3;
1 √
kak = (22 + (−1)2 + 12 ) 2 = 6;
1 √
kbk = (12 + 12 + 22 ) 2 = 6;
a.b √ 3√ 1
∴ cos θ = kak·kbk = 6· 6
= 2

∴ θ = π3 .
Example 2: The dot product provides an easy way of finding components of vectors. For
example, if
a = (a1 , a2 , a3 ) = a1 i + a2 j + a3 k

then
a.i = a1 , the component of a in the direction of i.

For the “basis” vectors i, j, k we have

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

14
 y

O θ B
x

Example 3: Find the angle between a diagonal of a cube and one of its edges.
−→ −−→
Solution Consider a cube of size d, let a = OA be the diagonal and let b = OB.

−→
a = OA

= di + dj + dk

= (d, d, d)
−−→
b = OB = di = (d, 0, 0)
a.b d2 1
cos θ = = √ =√
kakkbk 3d.d 3
 
1
∴ θ = cos−1 √
3

2
Example 4: Find the acute angle between the diagonals of a quadrilateral having vertices
(0, 0, 0), (1, 0, 0), (2, 1, 0) and (0, 1, 0).

j 6
C  B(2, 1, 0)
@
@ 
@ 
@ 
@ 

@ Θ
 @
 @ -
O A i
NB: No k direction needed.

15
 −→ −−→
OA = i, OB = 2i + j,
−−→
OC = j,
−→
CA = i − j.

−−→ −→
OB.CA 2.1 + 1.(−1) 1
cos Θ = −−→ −→ = √ √ =√
kOBk.kCAk 5. 2 10

 
1
Θ= cos−1 √ .
10
Example 5:
Beware A common error is to say:
“If x.a = x.b then a = b”
This is clearly incorrect. The correct reasoning goes as follows:
“If x.a = x.b then x.(a − b) = 0. Thus x is orthogonal to (a − b) .”
However, we have a useful Lemma and Corollary, where the key phrase in both is “for all x”.

Lemma 1.14 If x.c = 0 for all x, then c = 0.

Proof Take x = c. Then c.c = 0, so kck2 = 0 ⇒ c = 0.2 (Can you think of different proof,
where you prove all the components of a are zero?)

Corollary 1.15 If x.a = x.b for all x, then a = b.

Proof Rearranging gives x.(a − b) = 0 for all x, and using Lemma 1.14, we have a − b = 0 ⇒
a = b.2

(You should make sure you realise the difference between the “beware” example and Corollary
1.15.)
Orthogonal projection Consider a 6= 0 and b. Assume we wish to represent b in the form:

b = b|| + b⊥ ,

where b|| is parallel to a, and b⊥ is orthogonal to a, so that a.b⊥ = 0. (See diagram.)

16
 b
b⊥

a b||

Now,
b|| = λa,

for some λ ∈ R. Therefore,

b.a = (b|| + b⊥ ).a

= (b|| ).a

= λa.a

= λkak2 .

Hence, since a 6= 0,
b.a
λ=
kak2
and so we can determine b|| . Next, b⊥ can be found from the equation b⊥ = b − b|| . Thus,

(b.a)
b|| = a
kak2
(b.a)
b⊥ = b− a
kak2

Definition 1.16 b|| is called the orthogonal projection of b on a: it can also be written as
(b.â)â, where â is the unit vector along a. b⊥ is the vector component of b orthogonal to a.

(See also Anton, bottom of p812.)

Later we shall use the fact (easily deduced from the previous diagram) that ||b⊥ || = ||b|| sin θ,
where θ (0 ≤ θ ≤ π/2) is the angle between a and b.
−→
Direction Cosines (Anton p810) Let a = OA be a vector from O to A, and denote by
−→
α, β, γ the angles that OA makes with Ox, Oy, Oz, respectively. We define the direction
cosines of a to be cos α, cos β, cos γ. Now

a1 = i.a = kik kak cos α

17
and so,
a1
cos α = .
kak
a2 a3
Similarly, cos β = kak , cos γ = kak . It is easily shown that

cos2 α + cos2 β + cos2 γ = 1.

A = (a1 , a2 , a3 )

γ β

α
x
O

Exercises

1. What are the direction cosines of i?

2. A line makes angles of 60o with both the x−axis and the y−axis and is inclined at an
1 1 √1
obtuse angle to the z−axis. Show that its direction cosines are 2, 2, − 2 and find the
angle it makes with the z−axis.

Addition of Orthogonal projections and Orthogonal Components: Given a 6= 0, b, c.

We can write b = b|| + b⊥ , and c = c|| + c⊥ . Similarly, b + c = (b + c)|| + (b + c)⊥ . We have
the following result:
(b + c)|| = b|| + c|| , (b + c)⊥ = b⊥ + c⊥ .
(b.a)
Proof Now b|| = a, and there is a similar expression for c|| . Adding gives:
kak2
(b + c).a

b|| + c|| = 2
a = (b + c)|| . Then b⊥ + c⊥ = b + c − b|| + c|| = (b + c)⊥ 2
kak

1.7 The vector (or cross) product

Often in applications one needs to find a vector orthogonal to two other vectors. This is accom-
plished using the vector product as we now define.

18
 c=a∧b

Definition 1.17 Given two vectors a and b, their vector (or cross) product c, denoted by
c = a ∧ b (or c = a × b) is a vector

(i) orthogonal to both a and b

(ii) kck = kak kbk sin θ, (0 ≤ θ ≤ π)

(iii) If a k b, then a ∧ b = 0. Otherwise, the direction of c is such that a, b, c form a right-

handed system (that is, all the angles are less than π).

In fact, it is easy to see that “a k b if and only if a ∧ b = 0”.

There is a simple geometrical interpretation of the length of a vector product. First, note that
ka ∧ bk = ||a|| ||b|| sin θ = kak kb⊥ k, since kb⊥ k = kbk sin θ. Thus the length of ka ∧ bk equals
the area of the parallelogram generated by a and b.

b
b⊥

θ
a

Properties of the vector product:

(1) a ∧ b = −b ∧ a (anti-commutativity)

(2) a ∧ (b + c) = (a ∧ b) + (a ∧ c)

(3) λ(a ∧ b) = (λa) ∧ b = a ∧ (λb)

(4) a ∧ a = 0, ∀a (since sin θ = 0).

19
Properties (1), (3), (4) follow from the definition.
Proof of (1): (Note: to prove two vectors are equal we need to check the conditions of Definition
1.6.)
b
b
d =b∧a

a a

c=a∧b

First, a ∧ b and b ∧ a are both orthogonal to a and b: so they are parallel. Next, ka ∧ bk =
kb ∧ ak = kak kbk sin θ: so a ∧ b and b ∧ a have the same magnitude. Finally, in the figure,
a, b, a ∧ b form a right handed system. Also, b, a, b ∧ a form a right handed system. Thus d is
in the opposite direction to c and so,

a ∧ b = −b ∧ a .

2
Proof of (2): We’ll delay the proof of this till after we introduce the scalar triple product.
Important examples:

i ∧ i = j ∧ j = k ∧ k = 0.

i ∧ j = k, j ∧ k = i, k ∧ i = j.

j ∧ i = −k, k ∧ j = −i, i ∧ k = −j.

(Proofs from the definition.)

The component form of the vector product: Given

a = a1 i + a2 j + a3 k

b = b1 i + b2 j + b 3 k

20
then

a ∧ b = (a1 i + a2 j + a3 k) ∧ (b1 i + b2 j + b3 k)

= a1 b1 i ∧ i + a1 b2 i ∧ j + a1 b3 i ∧ k +

a2 b1 j ∧ i + a2 b2 j ∧ j + a2 b3 j ∧ k +

a3 b1 k ∧ i + a3 b2 k ∧ j + a3 b3 k ∧ k

= a1 b2 k − a1 b3 j − a2 b1 k + a2 b3 i + a3 b1 j − a3 b2 i

= (a2 b3 − a3 b2 )i + (a3 b1 − a1 b3 )j + (a1 b2 − a2 b1 )k.

There is a convenient way of calculating the vector product using 3 × 3 determinants (see Anton,
p817–819, and MA10209/210 Algebra).

i j k
a2 b3 a3 a1 a1 a2
a∧b= a1 a2 a3 =i +j +k
b2 b3 b3 b1 b1 b2
b1 b2 b3

“The rule of Sarrus”

For 3 × 3 determinants, one can use the “Rule of Sarrus” which goes as follows:
To evaluate
a1 a2 a3
b1 b2 b3
c1 c2 c3
the first step is to repeat the first two columns

a1 a2 a3 a1 a2
@ @ @

b1 @
Rb @
Rb @
Rb b2
2 3 1
@ @ @

c1 c2 @
Rc @
Rc @
Rc
3 1 2

Now the determinant equals

(a b c + a2 b3 c1 + a3 b1 c2 ) − (a3 b2 c1 + a1 b3 c2 + a2 b1 c3 )
| 1 2 3 {z }
@
”the terms” − ”the terms”
@
R

21
You can easily check this equals

b2 b3 b3 b1 b1 b2
a1 + a2 + a3 .
c2 c3 c3 c1 c1 c2

Remark In these notes we define the vector product geometrically and go on to deduce the
component form. Anton first defines the vector product using the determinant (Def 12.4.2) and
then deduces the geometric properties (Th 12.4.5).

1.8 Scalar triple product

Definition 1.18 Given three vectors a, b, c, their scalar triple product, denoted by [a, b, c], is
the scalar
[a, b, c] := a.(b ∧ c)

Component form:
[a, b, c] = a1 (b ∧ c)1 + a2 (b ∧ c)2 + a3 (b ∧ c)3 =
= a1 (b2 c3 − b3 c2 ) + a2 (b3 c1 − b1 c3 ) + a3 (b1 c2 − b2 c1 )

a1 a2 a3
= b1 b2 b3 - the 3 × 3 determinant again!
c1 c2 c3

Geometric interpretation:

b∧c

a
c
θ

Consider the parallelepiped P (spanned by a, b, c). The area of the base is kb ∧ ck. The height
π π
is kak cos θ if (0 ≤ θ ≤ ), or −kak cos θ if ( ≤ θ ≤ π). (Here θ is the angle between a and
2 2
b ∧ c.) If the volume of P is denoted by V , then

 V > 0, if a, b, c is right handed
[a, b, c] =
 −V < 0, if a, b, c is left handed.

22
So [a, b, c] is a “signed” volume of a parallelepiped.
Since the volume of the parallelepiped as well as the right-handedness property are unchanged
if the vectors a, b and c are interchanged cyclically we see immediately that

a.(b ∧ c) = b.(c ∧ a) = c.(a ∧ b).

Here we have kept the . and the ∧ fixed and cyclically interchanged a, b and c. We give a full
list of the properties of the scalar triple product on the next page.
Co-planar vectors: Three vectors are co-planar if there exist scalars λ, µ, ν, with λ2 + µ2 +
ν 2 > 0, such that λa + µb + νc = 0. (We say that a, b and c are “linearly dependent” vectors.)

[a, b, c] = 0 ⇐⇒ a, b, c are co-planar .

(The proof follows from the geometric interpretation.)

Properties of the scalar triple product

(1) [a, b, c] = [b, c, a] = [c, a, b] =

= −[a, c, b] = −[b, a, c] = −[c, b, a]
(Setting c = a gives [a, a, b] = [a.b.a] = [b, a, a] = 0, ∀a, b)

(2) [a, b, c + d] = [a, b, c] + [a, b, d]

(3) λ[a, b, c] = [λa, b, c]

Properties (1) - (3) follow from the properties of the scalar and the vector products.
Example:
[i, j, k] = i.(j ∧ k) = i.i = 1.

N.B. [a, b, c] = a.(b ∧ c) = [c, a, b] = c.(a ∧ b) = (a ∧ b).c. Therefore the order of the . and ∧
doesn’t matter. Also, the brackets can be dropped in a.(b ∧ c), since (a.b) ∧ c makes no sense
(a scalar vector product a vector).
Proof of the distributive law, property (2), for the vector product: We wish to prove:

a ∧ (b + c) = a ∧ b + a ∧ c.

23
Proof We use the cyclic property of the scalar triple product (twice) and the distributive
property of the scalar product. Consider

x.a ∧ (b + c) = (b + c).x ∧ a (cyclic STP)

= b.x ∧ a + c.x ∧ a (distributive property of scalar product)

= x.a ∧ b + x.a ∧ c (cyclic STP)

= x.(a ∧ b + a ∧ c).

Therefore
x.(a ∧ (b + c)) = x.(a ∧ b + a ∧ c)),

for all x. Corollary 1.15 gives

a ∧ (b + c) = a ∧ b + a ∧ c,

as required.2

1.9 Vector identities

Two very useful identities are the vector triple products:

a ∧ (b ∧ c) = (a.c)b − (a.b)c (∗)

and
(a ∧ b) ∧ c = (a.c)b − (b.c)a (∗∗)

To remember these, note that each of the vectors inside the brackets on the left appears once
outside the brackets on the right and the “middle” vector b appears first. Each term contains
a, b, c once only.
Proof of a ∧ (b ∧ c) = (a.c)b − (a.b)c.
We use components. First, if a = 0 then the result is true. Assume a 6= 0, and select the axes
such that a = ai.

LHS = ai ∧ {i(b2 c3 − b3 c2 ) + j(b3 c1 − b1 c3 ) + k(b1 c2 − b2 c1 )}

= j(−a(b1 c2 − b2 c1 )) + k(a(b3 c1 − b1 c3 )),
RHS = (ai.c)b − (ai.b)c
= (ac1 )b − (ab1 )c

24
 = (ac1 )(b1 i + b2 j + b3 k) − (ab1 )(c1 i + c2 j + c3 k)
= ja(c1 b2 − c2 b1 ) + ka(c1 b3 − c3 b1 ) = LHS.2

Identity (**) immediately follows from (*):

(a ∧ b) ∧ c = −c ∧ (a ∧ b) = − {(c.b)a − (c.a)b} = (a.c)b − (b.c)a.

Remark: The position of the brackets in these identities is important, since, in general,
a ∧ (b ∧ c) is in the (b, c) plane, whereas (a ∧ b) ∧ c is in the (a, b) plane. (See also an example
in Lectures.)
Exercise: Prove the Jacobi identity:

a ∧ (b ∧ c) + b ∧ (c ∧ a) + c ∧ (a ∧ b) = 0.

Hint: Expand out the 3 vector triple products.

1.10 Summary

In this section the main tools for manipulating vectors have been defined and discussed. These
tools are, the dot product, the vector product, the scalar triple product and the vector triple
product. We shall see how these are used in the following sections.
As we have discussed in the text, Anton takes the opposite approach to the one we take. We work
from a “geometric” viewpoint and then deduce various “algebraic” (or “component”) properties,
whereas Anton works from “algebraic” to “geometric”.
You are recommended to read Anton carefully to see the alternative derivations, to see the
worked examples and to read the the physical examples (see, for example, the paragraphs on p.
824 about “Moments and Rotation in 3-Space”.)

25