Part I

Position, Velocity and Acceleration

An extensive overview of this section on Kinematics can be found in Chapter 3 of

"Fundamental of Applied Dynamics", JH Williams Jr, 1996, Wiley.

Practically all kinematic analyses follow the same recipe;

find the position vector with respect to a reference frame,

differentiate to find the velocity,

differentiate again to find the acceleration.

This is explained in a video on LMS.

Let's consider a student rolling around in a thin circular hoop, where we want to study
the kinematics of two points on the students body. This problem is Example 3-1 from
Chapter 3-2 (p71-75).
We can use software such as Mathematica to perform many of the tasks required to ensure we
do not make mistakes. The Mathematica file for this problem is available on LMS.
Part II

Time derivatives in rotating reference frames

What you have learned up to this point has dealt with rotation about a fixed point. But what if
the body you are studying is rotating while also rotating about another coordinate system?

Now we will introduce you to situations where there is a time rate of change of a vector in a
rotating frame.

This is a Threshold Concept.

You are very familiar with the fact that velocity is the time derivative of position, however you
are about to see that in 3D rotation, angular velocity is not a derivative of anything!
A video on LMS explains the content in Chapter 3-3 (p77-85) in "Fundamentals of Applied
Dynamics".
3-3.4 A CauUonal'y Note about Flnite Rot.itlons
The discussion of angular displacements, velocities, and accelerations of a rigid
body has so far parallelled the discussion of ti11,8(lr displacements, velocities, and
accelerations of a point. In both cases it i5 possible to represent displacements,
velocities, and accelerations by direct�d line segments. There is, however, a
significant difference between the charac­ter of the directed line segments that
represent aiigular displacements and the directed line segments that represent
linear displacements. This difference emerges when nonplanar motions of finite
bodies are analyzed, or equivalently, when we must consider comb ining a number
of angular displacements. Such considerations are usually explored under the
cat­egory ofji.ni.te rot.ationS.
I! a polnl P undergoes several linear displacements in succeii&ion, the
resulting total linear displacement can be represented by a directed line segment
that is the vector sum of the directed line segments that represent the individual
displacements. However, when a finite body underg0es several finite aiigular
displacements about different axes in succession, the result depends on the order .in
which tJte i.n<lividual angular displacements are taken. So, even though finite angular
displacements can be represented by directed line segments, in
Luckily, the derivative of angular velocity is angular acceleration; it is the link between
angular position and angular velocity that is the complicated part.
Now we will see how to take a derivative in a rotating coordinate system.
To explain this concept, consider a rock rotating at ⃗Ω ⃗ about a fixed coordinate system, and
there is an ant on the rock. So, the both the rock and the ant are moving.

We want to find the total change in position and velocity of the ant, over a small time
increment.

In order to find this information, we will consider two situations.

Situation 1

First consider the situation as if the rock is not rotating, but the ant is walking on the rock.

Here, Δ𝑟 is a small change in position of the ant w.r.t. the rock.

So, this change in

position can be
expressed as;

𝜕𝑟 ∆𝑟
| =
𝜕𝑡 𝑎𝑠 𝑖𝑓 𝑟𝑜𝑐𝑘 ∆𝑡
𝑛𝑜𝑡 𝑟𝑜𝑡𝑎𝑡𝑖𝑛𝑔

𝜕
Here, we use to
𝜕𝑡
signify a small change
of a vector w.r.t. time.

69
Situation 2

We will now consider the situation where the rock is rotating and the ant is not moving.

Here, we use 𝑅⃗ to show that the

position is w.r.t. to the fixed
coordinate system, and ∆𝑅
⃗⃗⃗⃗⃗ is due
to the rotation of the rock only.

If we consider the position vectors, we can see that, ∆𝑅 ≈ 𝜃𝑅, as due to the small angle 𝜃, the
length ∆𝑅 is a straight line.

And that in a in a simple 2D situation, ∆𝑅 ≈ Ω. ∆𝑡. 𝑅

Δ𝑅
So the time derivative of the position vector, = Ω. 𝑅 = 𝑉, which is the velocity.
Δ𝑡

Therefore, ∆𝑅⃗ ≈ Ω
⃗⃗ 𝑥 𝑅⃗ . ∆𝑡

And;

𝑑𝑅⃗ 𝜕𝑅⃗
= | ⃗⃗ 𝑥 𝑅⃗
+Ω
𝑑𝑡 𝜕𝑡 𝑎𝑠 𝑖𝑓 𝑟𝑜𝑐𝑘
𝑛𝑜𝑡 𝑟𝑜𝑡𝑎𝑡𝑖𝑛𝑔

Where the ⃗Ω
⃗ 𝑥 𝑅⃗ term is due to the axis rotating.

The derivative of any vector is composed of two parts:

(a) The change in magnitude as if the axis is not rotating.

(b) The change due to the axis rotating.

70
 But of course we are not only concerned with position, we want to be able to take a time
derivative of any vector in a rotating coordinate system – for example, we want to be able to
differentiate velocity to get acceleration.

Let’s consider a vector 𝐴 that can be described as;

⃗
𝐴 = 𝐴𝑥 . 𝑖 + 𝐴𝑦 . 𝑗 + 𝐴𝑧 . 𝑘

where 𝑖, 𝑗, 𝑘
⃗ are unit vectors in the rotating coordinate system. Taking a derivative of the
scalars 𝐴𝑥 , 𝐴𝑦 , 𝐴𝑧 is simple; it is only the derivatives of the versors* we need to consider.

These versors are special cases of position vectors that follow the specific axis for one unit.
Then the time derivative of these constant versors in a rotating system with angular rotation
⃗⃗ , is;
Ω
𝑑 𝑑 𝑑
𝑖 = ⃗Ω
⃗ 𝑥𝑖 ; 𝑑𝑡 𝑗 = ⃗Ω
⃗ 𝑥𝑗 ; 𝑑𝑡 𝑘
⃗ = ⃗Ω ⃗
⃗ 𝑥𝑘
𝑑𝑡

This worked earlier for the position vector 𝑅⃗ so it must work for these versors too as they are
special cases of position vectors.

So;

𝑑𝐴 𝑑 𝑑 𝑑
⃗)
= (𝐴𝑥 . 𝑖) + (𝐴𝑦 . 𝑗) + (𝐴𝑧 . 𝑘
𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡

𝑑𝐴 𝑑𝐴𝑥 𝑑𝑖 𝑑𝐴𝑦 𝑑𝑗 𝑑𝐴𝑧 ⃗

𝑑𝑘
= 𝑖 + 𝐴𝑥 + 𝑗 + 𝐴𝑦 + ⃗ + 𝐴𝑧
𝑘
𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡
𝑑𝐴𝑥 𝑑𝑖
Where 𝑖 is a change in magnitude of 𝐴𝑥 in the 𝑖 direction, and 𝐴𝑥 𝑑𝑡 is due to the axis
𝑑𝑡
rotating where the magnitude does not change, but the direction does.

By substituting in what we found earlier, we get;

𝑑𝐴 𝑑𝐴𝑥 𝑑𝐴 𝑑𝐴
= ⃗ 𝑥 𝑖 + 𝑦 𝑗 + 𝐴𝑦 . ⃗Ω
𝑖 + 𝐴𝑥 . ⃗Ω ⃗ 𝑥𝑗+ 𝑧𝑘 ⃗ + 𝐴𝑧 . ⃗Ω
⃗ 𝑥𝑖
𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡

𝑑𝐴 𝑑𝐴𝑥 𝑑𝐴 𝑑𝐴
= ⃗ 𝑥 𝑖 + 𝑦 𝑗 + 𝐴𝑦 . ⃗Ω
𝑖 + 𝐴𝑥 . ⃗Ω ⃗ 𝑥𝑗+ 𝑧𝑘 ⃗ + 𝐴𝑧 . ⃗Ω
⃗ 𝑥𝑖
𝑑𝑡 𝑑𝑡 𝑑𝑡 𝑑𝑡
which can be summarised by;

𝑑𝐴 𝜕𝐴
= + ⃗Ω
⃗ 𝑥𝐴
𝑑𝑡 𝜕𝑡

And we can see that this is the same as before.

*As used here, versors simply means the unit vectors [i, j, k] but in the mathematics of quaternions, the
orthoganal unit vectors are only a subset of versors.
71
Now, let’s try perform a kinematic analysis of the rock and ant using intermediate frames.
The fixed coordinate system is [𝑋, 𝑌, 𝑍] and the local coordinate system attached to the rock
is [𝑥, 𝑦, 𝑧] and is rotating at ⃗Ω
⃗.

A vector 𝐴 is described relative to

the rotating coordinate system, but
the time derivative must be
expressed relative to a global or
fixed coordinate system. We have
shown already that the time
derivative of 𝐴 is;

𝑑𝐴 𝜕𝐴
= | + ⃗Ω
⃗ 𝑥𝐴
𝑑𝑡 𝜕𝑡 𝑙𝑜𝑐𝑎𝑙

𝑑𝐴
where is the total change w.r.t.
𝑑𝑡
𝜕𝐴
the fixed system [𝑋, 𝑌, 𝑍]; |
𝜕𝑡 𝑙𝑜𝑐𝑎𝑙
is the velocity as if the local
coordinate system was not moving;
and ⃗Ω
⃗ 𝑥 𝐴 is the velocity due to the
rotation of the system.

The position vectors relative to the two

coordinate systems [𝑥, 𝑦, 𝑧] and [𝑋, 𝑌, 𝑍]
are;

⃗
𝑟 = 𝑥𝑖 + 𝑦𝑗 + 𝑧𝑘

𝑅⃗ = 𝑋𝐼 + 𝑌𝐽 + 𝑍𝐾
⃗

In the local coordinate system [𝑥, 𝑦, 𝑧]

(subscript ‘relative’), we know that the
velocity 𝑉 and acceleration 𝛼 is;

𝜕𝑟
⃗⃗⃗⃗⃗⃗⃗ ⃗
𝑉𝑟𝑒𝑙 = ( ) = 𝑥̇ 𝑖 + 𝑦̇ 𝑗 + 𝑧̇ 𝑘
𝜕𝑡 𝑟𝑒𝑙
𝜕𝑟
⃗
𝛼𝑟𝑒𝑙 = ( ) = 𝑥̈ 𝑖 + 𝑦̈ 𝑗 + 𝑧̈ 𝑘
⃗⃗⃗⃗⃗⃗⃗
𝜕𝑡 𝑟𝑒𝑙

And in the fixed coordinate system [𝑋, 𝑌, 𝑍] the velocity is;

 𝑑𝑅⃗ ⃗⃗⃗⃗𝑜 𝑑𝑟
𝑑𝑅
⃗ =
𝑉 = +
𝑑𝑡 𝑑𝑡 𝑑𝑡

Because we can see from vector addition that; 𝑅⃗ = ⃗⃗⃗⃗

𝑅𝑜 + 𝑟
⃗⃗⃗⃗⃗𝑜
𝑑𝑅
Here, is the velocity of the centre of the rotating coordinate system w.r.t. the fixed system
𝑑𝑡
𝑑𝑟
and 𝑑𝑡 is the velocity as seen in the rotating system.

⃗⃗⃗⃗⃗𝑜
𝑑𝑅 𝑑𝑟
Because is already given w.r.t to the fixed system, we only need to consider further .
𝑑𝑡 𝑑𝑡
𝑑𝑟
So, we apply our equation from earlier to , that is;
𝑑𝑡

𝑑𝑟 𝜕𝑟
= | + ⃗Ω
⃗ 𝑥𝑟
𝑑𝑡 𝜕𝑡 𝑙𝑜𝑐𝑎𝑙

So now the velocity is;

⃗⃗⃗⃗𝑜
𝑑𝑅 𝜕𝑟
⃗ =
𝑉 +( ) +Ω ⃗⃗ 𝑥 𝑟
𝑑𝑡 𝜕𝑡 𝑟𝑒𝑙

which is the same as;

⃗⃗⃗⃗𝑜
𝑑𝑅
⃗ =
𝑉 + ⃗⃗⃗⃗⃗⃗⃗
𝑉𝑟𝑒𝑙 + ⃗Ω
⃗ 𝑥𝑟
𝑑𝑡

Of course, we also want the acceleration which we know is the derivative of the velocity.

⃗
𝑑𝑉 ⃗⃗⃗⃗𝑜
𝑑 𝑑𝑅
𝛼= = [ + ⃗⃗⃗⃗⃗⃗⃗
𝑉𝑟𝑒𝑙 + ⃗Ω
⃗ 𝑥 𝑟]
𝑑𝑡 𝑑𝑡 𝑑𝑡

𝑑2 ⃗⃗⃗⃗
𝑅𝑜 𝑑𝑉 ⃗⃗⃗⃗⃗⃗⃗
𝑟𝑒𝑙 𝑑
𝛼= + + ⃗⃗ 𝑥 𝑟)
(Ω
𝑑𝑡 2 𝑑𝑡 𝑑𝑡

𝑑2 ⃗⃗⃗⃗
𝑅𝑜 𝑑𝑉 ⃗⃗⃗⃗⃗⃗⃗
𝑟𝑒𝑙
⃗⃗
𝑑Ω 𝑑r
𝛼= + + 𝑥𝑟+ 𝑥Ω⃗⃗
𝑑𝑡 2 𝑑𝑡 𝑑𝑡 𝑑𝑡
⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝑑𝑉 𝑑r⃗
We can see that 𝑟𝑒𝑙
and are relative to the local rotating coordinate system and so must
𝑑𝑡 𝑑𝑡
be considered further using our expression.

Therefore;
 𝑑2 ⃗⃗⃗⃗
𝑅𝑜 𝜕𝑉 ⃗⃗⃗⃗⃗⃗⃗ 𝜕𝑟
𝛼= +
𝑟𝑒𝑙
| ⃗⃗ 𝑥 𝑉
+Ω ⃗⃗⃗⃗⃗⃗⃗ ⃗⃗ ̇ ⃗⃗
𝑟𝑒𝑙 + Ω 𝑥 𝑟 + Ω 𝑥 [ | ⃗⃗ 𝑥 𝑟]
+Ω
𝑑𝑡 2 𝜕𝑡 𝑙𝑜𝑐𝑎𝑙 𝜕𝑡 𝑙𝑜𝑐𝑎𝑙

⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝜕𝑉 𝜕𝑟
We can see that 𝑟𝑒𝑙
| is the local acceleration, ⃗⃗⃗⃗⃗⃗⃗
𝛼𝑟𝑒𝑙 , and 𝜕𝑡 | is the local velocity ⃗⃗⃗⃗⃗⃗⃗
𝑉𝑟𝑒𝑙 .
𝜕𝑡 𝑙𝑜𝑐𝑎𝑙 𝑙𝑜𝑐𝑎𝑙

Therefore;

𝑑2 ⃗⃗⃗⃗
𝑅𝑜
𝛼= 𝛼𝑟𝑒𝑙 + ⃗Ω
+ ⃗⃗⃗⃗⃗⃗⃗ ⃗ 𝑥 ⃗⃗⃗⃗⃗⃗⃗ ⃗ ̇ 𝑥 𝑟 + ⃗Ω
𝑉𝑟𝑒𝑙 + ⃗Ω ⃗ 𝑥 ⃗⃗⃗⃗⃗⃗⃗
𝑉𝑟𝑒𝑙 + ⃗Ω
⃗ 𝑥 (Ω
⃗⃗ 𝑥 𝑟)
𝑑𝑡 2

And finally, as we have two (Ω 𝑉𝑟𝑒𝑙 ) terms;

⃗⃗ 𝑥 ⃗⃗⃗⃗⃗⃗⃗

𝑑2 ⃗⃗⃗⃗
𝑅𝑜
𝛼= + ⃗⃗⃗⃗⃗⃗⃗ ⃗⃗ 𝑥 ⃗⃗⃗⃗⃗⃗⃗
𝛼𝑟𝑒𝑙 + 2Ω ⃗ ̇ 𝑥 𝑟 + ⃗Ω
𝑉𝑟𝑒𝑙 + ⃗Ω ⃗ 𝑥 (Ω
⃗⃗ 𝑥 𝑟)
𝑑𝑡 2

This is our final expression for acceleration and in this expression we find that;

𝑑2 𝑅
⃗⃗⃗⃗⃗𝑜
is the acceleration of the centre of the local coordinate system
𝑑𝑡 2

𝛼𝑟𝑒𝑙
⃗⃗⃗⃗⃗⃗⃗ is the acceleration of the point of interest in the local coordinate system

⃗⃗ 𝑥 ⃗⃗⃗⃗⃗⃗⃗
2Ω 𝑉𝑟𝑒𝑙 is the Coriolis acceleration

⃗⃗ ̇ 𝑥 𝑟
Ω is the Euler acceleration

⃗⃗ 𝑥 (Ω
Ω ⃗⃗ 𝑥 𝑟) is the centripetal acceleration.

Using this new knowledge, we can solve Examples 3-7 and 3-8 from Chapter 3-5
(p96-101). A video solving Example 3-7 is on LMS. There is no video for 3-8, please
read the solution carefully.
More practice with taking time derivatives of vectors; look at Example 3-5 from Chapter 3-4 (p89).
We finish this topic with Example 3-6 (p93), the Anatomy of Centripetal and Coriolis accelerations.
The table below (adapted from Chapter 3-5 (p96) summarises the major kinematic results for
points. In this table, OXYZ refers to the fixed coordinate system and oxzy is the local
coordinate system.