Section 1.

1.5 Coordinate Transformation of Vector Components

Very often in practical problems, the components of a vector are known in one coordinate
system but it is necessary to find them in some other coordinate system.

For example, one might know that the force f acting “in the x1 direction” has a certain
value, Fig. 1.5.1 – this is equivalent to knowing the x1 component of the force, in an
x1  x 2 coordinate system. One might then want to know what force is “acting” in some
other direction – for example in the x1 direction shown – this is equivalent to asking what
the x1 component of the force is in a new x1  x 2 coordinate system.

x2 x1 component
of force
x2
f
x1

x1
x1 component
of force

Figure 1.5.1: a vector represented using two different coordinate systems

The relationship between the components in one coordinate system and the components
in a second coordinate system are called the transformation equations. These
transformation equations are derived and discussed in what follows.

1.5.1 Rotations and Translations

Any change of Cartesian coordinate system will be due to a translation of the base
vectors and a rotation of the base vectors. A translation of the base vectors does not
change the components of a vector. Mathematically, this can be expressed by saying that
the components of a vector a are e i  a , and these three quantities do not change under a
translation of base vectors. Rotation of the base vectors is thus what one is concerned
with in what follows.

1.5.2 Components of a Vector in Different Systems

Vectors are mathematical objects which exist independently of any coordinate system.
Introducing a coordinate system for the purpose of analysis, one could choose, for
example, a certain Cartesian coordinate system with base vectors e i and origin o, Fig.

Solid Mechanics Part III 24 Kelly

 Section 1.5

1.5.2. In that case the vector can be written as u  u1e1  u 2 e 2  u 3 e 3 , and u1 , u 2 , u 3 are
its components.

Now a second coordinate system can be introduced (with the same origin), this time with
base vectors ei . In that case, the vector can be written as u  u1e1  u 2 e2  u 3 e3 , where
u1 , u 2 , u 3 are its components in this second coordinate system, as shown in the figure.
Thus the same vector can be written in more than one way:

u  u1e1  u 2 e 2  u 3e 3  u1e1  u 2 e2  u 3 e3

The first coordinate system is often referred to as “the ox1 x 2 x3 system” and the second as
“the ox1 x 2 x3 system”.

x2
x2
u1

u 2

 x1 u2

e2 e1

o x1
u1

Figure 1.5.2: a vector represented using two different coordinate systems

Note that the new coordinate system is obtained from the first one by a rotation of the
base vectors. The figure shows a rotation  about the x3 axis (the sign convention for
rotations is positive counterclockwise).

Two Dimensions

Concentrating for the moment on the two dimensions x1  x 2 , from trigonometry (refer to
Fig. 1.5.3),

u  u1e1  u2e 2
 OB  AB e1   BD  CP e 2
 cos u1  sin  u2 e1  sin  u1  cos u2 e 2

and so

Solid Mechanics Part III 25 Kelly

 Section 1.5

u1  cos u1  sin  u 2

u 2  sin  u1  cos u 2

vector components in vector components in

first coordinate system second coordinate system

In matrix form, these transformation equations can be written as

 u1  cos  sin    u1 

u    sin  cos  u 2 
 2 

x2
x2
u1 P

u2

x1 u2
 C
D

 A B
x1
o
u1

Figure 1.5.3: geometry of the 2D coordinate transformation

The 2  2 matrix is called the transformation or rotation matrix Q . By pre-

multiplying both sides of these equations by the inverse of Q , Q 1 , one obtains the  
transformation equations transforming from u1 u 2  to u1 u 2  :
T T

 u1   cos sin    u1 

u     sin  cos  u 2 
 2 

An important property of the transformation matrix is that it is orthogonal, by which is

meant that

Q   Q 
1 T
Orthogonality of Transformation/Rotation Matrix (1.5.1)

Three Dimensions

It is straight forward to show that, in the full three dimensions, Fig. 1.5.4, the components
in the two coordinate systems are related through

Solid Mechanics Part III 26 Kelly

 Section 1.5

 u1   cos( x1 , x1 ) cos( x1 , x 2 ) cos( x1 , x3 )   u1 

u   cos( x , x  ) cos( x , x  ) cos( x , x  ) u  
 2  2 1 2 2 2 3  2 

u 3  cos( x3 , x1 ) cos( x3 , x 2 ) cos( x3 , x3 )  u 3 

where cos( xi , x j ) is the cosine of the angle between the xi and x j axes. These nine
quantities are called the direction cosines of the coordinate transformation. Again
denoting these by the letter Q, Q11  cos( x1 , x1 ), Q12  cos( x1 , x2 ) , etc., so that

Qij  cos( xi , x j ) , (1.5.2)

one has the matrix equations

 u1  Q11 Q12 Q13   u1 

u   Q Q23  u 2 
 2   21 Q22
u 3  Q31 Q32 Q33  u 3 

or, in element form and short-hand matrix notation,

u i  Qij u j  u  Qu (1.5.3)

x2
x2
u
x1

x1
x3
x3
Figure 1.5.4: two different coordinate systems in a 3D space

Note: some authors define the matrix of direction cosines to consist of the components
Qij  cos( xi , x j ) , so that the subscript i refers to the new coordinate system and the j to
the old coordinate system, rather than the other way around as used here.

Transformation of Cartesian Base Vectors

The direction cosines introduced above also relate the base vectors in any two Cartesian
coordinate systems. It can be seen that

e i  e j  Qij (1.5.4)

Solid Mechanics Part III 27 Kelly

 Section 1.5

This relationship is illustrated in Fig. 1.5.5 for i  1 .

e2
cos( x1 , x2 )  e1  e2

e1 cos( x1 , x1 )  e1  e1

e1

e3
cos( x1 , x3 )  e1  e3

Figure 1.5.5: direction cosines

Formal Derivation of the Transformation Equations

In the above, the transformation equations u i  Qij u j were derived geometrically. They
can also be derived algebraically using the index notation as follows: start with the
relations u  u k e k  u j ej and post-multiply both sides by e i to get (the corresponding
matrix representation is to the right (also, see Problem 3 in §1.4.3)):

u k e k  e i  u j ej  e i
 u k  ki  u j Qij
 u i  u j Qij  u   u Q 
T T T

 u i  Qij u j  u  Qu

The inverse equations are {▲Problem 3}

u i  Q ji u j  u   Q T u (1.5.5)

Orthogonality of the Transformation Matrix Q 

As in the two dimensional case, the transformation matrix is orthogonal, Q T  Q 1 .    

This follows from 1.5.3, 1.5.5.

Example

Consider a Cartesian coordinate system with base vectors e i . A coordinate

transformation is carried out with the new basis given by

Solid Mechanics Part III 28 Kelly

 Section 1.5

e1  n1(1) e1  n2(1) e 2  n3(1) e 3

e2  n1( 2) e1  n2( 2 ) e 2  n3( 2 ) e 3
e3  n1( 3) e1  n2(3) e 2  n3( 3) e 3

What is the transformation matrix?

Solution

The transformation matrix consists of the direction cosines Qij  cos( xi , x j )  e i  ej , so

 u1  n1 n1(3)   u1 

(1)
n1( 2 )
u   n (1) n2( 2 )

n2(3)  u 2 
 2  2
u 3  n3(1) n3( 2 ) n3(3)  u 3 

1.5.3 Problems
1. The angles between the axes in two coordinate systems are given in the table below.
x1 x2 x3
x1 135 o
60 o
120 o
x 2 90 o 45 o 45 o
x3 45 o 60 o 120 o
Construct the corresponding transformation matrix Q  and verify that it is
orthogonal.
2. The ox1 x 2 x3 coordinate system is obtained from the ox1 x 2 x3 coordinate system by a
positive (counterclockwise) rotation of  about the x3 axis. Find the (full three
dimensional) transformation matrix Q  . A further positive rotation  about the
x 2 axis is then made to give the ox1x 2 x3 coordinate system. Find the
corresponding transformation matrix P  . Then construct the transformation matrix
R  for the complete transformation from the ox1 x2 x3 to the ox1x2 x3 coordinate
system.
3. Beginning with the expression u j e j  ei  u k ek  ei , formally derive the relation
 
u i  Q ji u j ( u   Q T u ).

Solid Mechanics Part III 29 Kelly