Dynamics of Vibrating Systems

Part 2: Basic Dynamic Models & Vibration

Dr Oksana Trushkevych
 Unit 2 Basic dynamic models
Constructing equations of motion
Solving equation of motion for free vibration of 1DoF system
A glimpse of 2 DoF system

Unit 3 Equivalence
Stiffness
Mass/Inertia
Damping

2
Notation
Scalar variables:
Angles:
Matrices:
some textbooks

Vectors:

Time-derivatives:
Complex numbers:
 Constructing equations of motion
Our tools:
Newton’s 2nd law (consider all forces)
D’Alembert’s principle (dynamic equilibrium) video
𝑑
Conservation of energy 𝑇+𝑉 =0
𝑑𝑡
Virtual displacement principle (in equilibrium for 𝛿 𝑟,
Ԧ δW = 0) video
Lagrange equation (unit 5)

Rao 2.2
4
 Free vibration (1DoF)
one-degree-of-freedom vibrating
system: the spring-mass oscillator

Free
Body
Diagram

Rao 2.2
5
Free vibration (1DoF) – Newton’s 2nd law
Newton’s second law: if a mass is displaced by a distance 𝑥Ԧ 𝑡 when
acted by a resultant force 𝐹Ԧ 𝑡 , then
𝑑 𝑑 𝑥Ԧ 𝑡
Ԧ
𝐹(𝑡) = (𝑚 )
𝑑𝑡 𝑑𝑡

For a constant mass,

𝑑 2 𝑥Ԧ 𝑡 𝑀 𝑡 = 𝐼 𝜃Ԧ
ሷ
𝐹Ԧ 𝑡 = 𝑚 = 𝑚𝑥ሷ
𝑑𝑡 2

For the stated problem, 𝐹 𝑡 = −𝑘𝑥 = 𝑚𝑥ሷ

𝑚𝑥ሷ + 𝑘𝑥 = 0
𝑘 𝑘
𝑥ሷ + 𝑥 = 0 𝑥ሷ + 𝜔𝑛2 𝑥 = 0, 𝜔𝑛 =
𝑚 𝑚

6 Rao 2.2 Where ωn is the angular frequency of the system in rad/s

 Free vibration (1DoF) – energy method
Energy method for a conservative system:
𝑑
𝑇 + 𝑉 = 𝑐𝑜𝑛𝑠𝑡 𝑇+𝑉 =0
𝑑𝑡

𝑚𝑥ሶ 2 𝑘𝑥 2
𝑇= ; 𝑉=
2 2
𝑑 𝑑 𝑚𝑥ሶ 2 𝑘𝑥 2
𝑇+𝑉 = + = 𝑚𝑥ሶ 𝑥ሷ + 𝑘𝑥𝑥ሶ = 0
𝑑𝑡 𝑑𝑡 2 2

𝑥ሶ = 0 is a trivial solution, we also get

𝑚𝑥ሷ + 𝑘𝑥 = 0

Rao 2.2
7
Free vibration – energy method example

How many degrees of freedom?

We know
m – mass
R – radius of the pulley
mp or I – mass or moment of inertia of
the pulley
k – spring constant of the spring

8
Example – energy method
Write down all energies:
1 1 2 1 2
𝑇 = 𝑚𝑥ሶ + 𝐼 𝜃ሶ
2
𝑉 = 𝑘𝑥
2 2 2
Terms are squared so no worries about directions
Use relationship between mass velocity and pulley rotation: 𝑥
2 𝜃=
1 1 𝑥ሶ 𝑅
2 𝑥ሶ
𝑇 = 𝑚𝑥ሶ + 𝐼 ሶ
2 2 𝑅 𝜃=
𝑅
Take derivative in time 𝑑𝑇 𝐼 𝑑𝑉
= 𝑚𝑥ሶ 𝑥ሷ + 2 𝑥ሶ 𝑥ሷ = 𝑘𝑥 𝑥ሶ
𝑑𝑡 𝑅 𝑑𝑡
Add derivatives and equate to zero:
𝐼
𝑚𝑥ሶ 𝑥ሷ + 𝑥ሶ𝑥ሷ + 𝑘𝑥 𝑥ሶ = 0
𝑅2
𝐼
9 𝑚 + 2 𝑥ሷ + 𝑘𝑥 = 0
𝑅
Free vibration (1DoF) – D’Alembert’s principle
The sum of the differences between the forces acting on a system and the time derivatives
of the momenta of the system is zero

We can rewrite equations of motion from Newton’s second law as:

𝐹Ԧ 𝑡 − 𝑚𝑥Ԧሷ = 0 ሷ
𝑀 𝑡 − 𝐼 𝜃Ԧ = 0
These can be considered equilibrium equations if 𝑚𝑥ሷ is treated as a force
ሷ
(and 𝐼𝜃 as a moment)
Then for the stated problem, −𝑘𝑥 − 𝑚𝑥ሷ = 0

𝑚𝑥ሷ + 𝑘𝑥 = 0
𝑘 𝑘
𝑥ሷ + 𝑥 = 0 𝑥ሷ + 𝜔𝑛2 𝑥 = 0, 𝜔𝑛 =
𝑚 𝑚

10 Rao 2.2 Additional slide to complement videos

Free vibration (1DoF) – virtual displacement principle
In a system in equilibrium the virtual work done by virtual displacement is zero

Ԧ 𝑟Ԧ = 0
𝛿𝑊 = ෍ 𝐹𝛿

• Applies to accelerating systems if we use inertia forces (D’Alembert)

• Very useful in cases when there are various linked motions in different directions
• A virtual displacement 𝛿 𝑟Ԧ is
• infinitesimal variation of the coordinate (in the permitted direction)
• no significant change of geometry
𝑑
• Instantaneous 𝛿 𝑟Ԧ = 0
𝑑𝑡
• Virtual work is the work done by all the active forces in a virtual displacement
• Forces are assumed to remain unchanged while applying 𝛿 𝑟Ԧ

11 Rao 2.2 Additional slide to complement videos

Free vibration (1DoF) – virtual displacement principle
In a system in equilibrium the virtual work done by virtual displacement is zero

Then for the stated problem,

𝛿𝑊𝑠𝑝𝑟𝑖𝑛𝑔 = −(𝑘𝑥)𝛿𝑥
𝛿𝑊𝑖𝑛𝑒𝑟𝑡𝑖𝑎 = −(𝑚𝑥)𝛿𝑥
ሷ

𝛿𝑊𝑖𝑛𝑒𝑟𝑡𝑖𝑎 + 𝛿𝑊𝑖𝑛𝑒𝑟𝑡𝑖𝑎 = −(𝑚𝑥)𝛿𝑥

ሷ − 𝑘𝑥 𝛿𝑥 = 0

𝛿𝑥 ≠ 0, so
−𝑘𝑥 − 𝑚𝑥ሷ = 0

𝑚𝑥ሷ + 𝑘𝑥 = 0

12 Rao 2.2 Additional slide to complement videos

2.2 Solving 2 nd order ODE
Free vibration (1DoF) – solution – periodic functions
𝑘
𝑚𝑥ሷ + 𝑘𝑥 = 0 𝑥ሷ + 𝜔𝑛2 𝑥 = 0, 𝜔𝑛 =
𝑚

𝑥 = 𝑋 cos(𝜔𝑡 + 𝜑)
𝑥ሶ = −𝜔𝑋 𝑠𝑖𝑛(𝜔𝑡 + 𝜑)
𝑥ሷ = −𝜔2 𝑋 cos(𝜔𝑡 + 𝜑)
X
−𝜔2 𝑋 cos 𝜔𝑡 + 𝜑 + 𝜔𝑛2 𝑋 cos(𝜔𝑡 + 𝜑) = 0
−𝜔2 + 𝜔𝑛2 = 0
𝜔 = 𝜔𝑛
𝑥 = 𝑋 cos(𝜔𝑛 𝑡 + 𝜑) or 𝑥 = 𝑋1 cos 𝜔𝑛 𝑡 + 𝑋2 sin(𝜔𝑛 𝑡)

𝑋= 𝑋1 2 + 𝑋2 2
Free vibration (1DoF) – solution – exponential form
𝑚𝑥ሷ + 𝑘𝑥 = 0
ways of solving 2nd order differential equations:

𝑖(𝜔𝑡+𝜑)
𝑥 = 𝐴𝑒 𝑝𝑡 p can be complex 𝑥 = 𝑋𝑒
general form, used for solving cases involving damping
𝑥 = 𝐴𝑒 𝑝𝑡
𝑝2 𝐴𝑒 𝑝𝑡 + 𝜔𝑛2 𝐴𝑒 𝑝𝑡 = 0
𝑥ሶ = 𝑝𝐴𝑒 𝑝𝑡
𝑥ሷ = 𝑝2 𝐴𝑒 𝑝𝑡
𝑝2 = − 𝜔𝑛2 𝑝 = ± 𝑖𝜔𝑛
General solution
𝑥 = 𝐴1 𝑒 𝑖𝜔𝑡 + 𝐴2 𝑒 −𝑖𝜔𝑡
𝑥 = 𝐴1 𝑒 𝑝1 𝑡 + 𝐴2 𝑒 𝑝2 𝑡 𝑥 = 𝐴1 + 𝐴2 𝑐𝑜𝑠 𝜔𝑡 + 𝑖 𝐴1 − 𝐴2 sin(𝜔𝑡)
𝑥 = 𝐶1 𝑐𝑜𝑠 𝜔𝑡 + 𝐶2 sin 𝜔𝑡 = 𝐷𝑐𝑜𝑠 𝜔𝑡 + 𝜑
Constants are defined by initial conditions
15
often specified as displacement and velocity at t=0
Free vibration (1DoF) – solution – other forms
𝑘
𝑚𝑥ሷ + 𝑘𝑥 = 0 𝑥ሷ + 𝜔𝑛2 𝑥
= 0, 𝜔𝑛 =
𝑚
ways of solving 2nd order differential equations:

1) 𝑥 = 𝑋 cos(𝜔𝑡 + 𝜑) 2) 𝑥 = ℜ𝑒(𝑋𝑒 𝑖(𝜔𝑡+𝜑) ) 3) 𝑥 = 𝐴𝑒 𝑝𝑡

𝑥ሶ = −𝜔𝑋 𝑠𝑖𝑛(𝜔𝑡 + 𝜑) 𝑥ሶ = ℜ𝑒(𝑖𝜔𝑋𝑒 𝑖(𝜔𝑡+𝜑) ) p can be complex
𝑥ሷ = −𝜔2 𝑋 cos(𝜔𝑡 + 𝜑) 𝑥ሷ = ℜ𝑒(−𝜔2 𝑋𝑒 𝑖(𝜔𝑡+𝜑) ) More general form,
used for solving
cases involving
−𝜔2 𝑋 cos 𝜔𝑡 + 𝜑 + 𝜔𝑛2 𝑋 cos 𝜔𝑡 + 𝜑 = 0 −𝜔2 𝑋𝑒 𝑖(𝜔𝑡+𝜑) + 𝜔𝑛2 𝑋𝑒 𝑖(𝜔𝑡+𝜑) = 0
damping
𝜔 = 𝜔𝑛 𝜔 = 𝜔𝑛
𝑥 = 𝐴1 𝑒 𝑝1 𝑡 + 𝐴2 𝑒 𝑝2 𝑡
General solution 𝑥 = ℜ𝑒(𝑋𝑒 𝑖(𝜔𝑛 𝑡+𝜑) )
𝑥 = 𝑋 cos(𝜔𝑛 𝑡 + 𝜑) 𝑥 = ℜ𝑒(Xcos 𝜔𝑛 𝑡 + 𝜑 − 𝑖 Xsin 𝜔𝑛 𝑡 + 𝜑 )
Additional slide to complement videos
Amplitude and phase are defined by initial conditions
16
often specified as displacement and velocity at t=0
 Quick example
m=1kg 𝑥 = 𝑋 cos(𝜔𝑛 𝑡 + 𝜑)
Force of 100N to pull 10cm to the right and then let go
Find k: F=kx, k=F/x, k=100N/0.1m=1kN (kg m/s2)
𝑘
Natural frequency is 𝜔𝑛 = =100 rad/s
𝑚
Full equation of motion: 𝑥 = 𝑋 cos(𝜔𝑛 𝑡 + 𝜑)
Initial conditions (t=0), X=10cm;
Velocity at t=0 is zero 𝑥ሶ = −𝜔𝑋 𝑠𝑖𝑛(𝜔𝑡 + 𝜑)
so phase angle is zero 𝑥 = 0.1 cos 100𝑡 [𝑚]

17
 Free vibration (1DoF) summary:
tools for constructing equations of motion

𝑑 𝑑 𝑥Ԧ 𝑡
Newton’s 2nd law 𝐹Ԧ 𝑡 = 𝑚 = 𝑚𝑎
𝑑𝑡 𝑑𝑡
𝑑
Conservation of energy 𝑇+𝑉 =0
𝑑𝑡
D’Alembert’s principle, dynamic equilibrium, ma treated as a force
Virtual displacement principle (in equilibrium for 𝛿 𝑟,
Ԧ δW = 0)
Lagrange equation (Unit 5)

Rao 2.2
18
 Free vibration (1 DoF) summary
Harmonic equation:
2nd order ODE of the form 𝑥ሷ + 𝜔2 𝑥 = 0
Oscillation with angular frequency 𝜔.
Solution in the form 𝑥 = 𝑋 cos(𝜔𝑛 𝑡 + 𝜑) or 𝑥 = 𝑋1 cos 𝜔𝑛 𝑡 + 𝑋2 sin(𝜔𝑛 𝑡)
Two parameters (e.g. amplitude and phase or X1 and X2)
Can also be done using exponentials and complex
number notation
𝑥 = 𝐴1 𝑒 𝑖𝜔𝑡 + 𝐴2 𝑒 −𝑖𝜔𝑡 𝑥 = 𝐴1 𝑒 𝑝1 𝑡 + 𝐴2 𝑒 𝑝2 𝑡

19
2.3 2DoF equations
Undamped Free Vibration – 2-DoF
k1 k2

x1 x2
Build free body diagrams:
k1x1 m1 m2
k2(x2-x1) k2(x2-x1)

Apply Newton’s second law to each mass:

Take care
with signs!
This can be written in matrix form:

𝑚1 0 𝑥1ሷ 𝑘 +𝑘2 −𝑘2 𝑥1 0

+ 1 =
0 𝑚2 𝑥2ሷ −𝑘2 𝑘2 𝑥2 0

Rao 6.1
Undamped Free Vibration – 2-DoF– solution
𝑴xሷ + 𝑲x = 𝟎

The general solution can be written as

𝑥1 = 𝐴1 cos(𝜔𝑡 + 𝜑)
𝑥2 = 𝐴2 cos(𝜔𝑡 + 𝜑)

Then, 𝑥1ሷ = −𝜔2 𝐴1 cos(𝜔𝑡 + 𝜑)

𝑥2ሷ = −𝜔2 𝐴2 cos(𝜔𝑡 + 𝜑)

−𝜔2 𝑴x + 𝑲x = 𝟎

Similar for torsional vibration – see videos

Rao 6.9
Undamped Free Vibration – 2-DoF– solution in more detail
𝑚1 0 𝑥1ሷ 𝑘1 +𝑘2 −𝑘2 𝑥1 0
+ =
0 𝑚2 𝑥2ሷ −𝑘2 𝑘2 𝑥2 0
𝑴𝑿ሷ + 𝑲𝑿 = 𝟎 Let’s rename k1 and k2
combinations and
𝑀11 0 𝑥1ሷ 𝐾 𝐾12 𝑥1 0 solve in the most
+ 11 =
0 𝑀22 𝑥2ሷ 𝐾21 𝐾22 𝑥2 0 general form

The general solution can be written as

𝑥1 = 𝐴1 cos(𝜔𝑡 + 𝜑) This implies all coordinates can
𝑥2 = 𝐴2 cos(𝜔𝑡 + 𝜑) vibrate with the same frequency
and phase angle, but only at the
Then,
𝑥1ሷ = −𝜔2 𝐴1 cos(𝜔𝑡 + 𝜑) frequency that satisfies the
𝑥2ሷ = −𝜔2 𝐴2 cos(𝜔𝑡 + 𝜑) equations

Rao 6.9
Additional slide to complement videos
Undamped Free Vibration – 2-DoF– solution in more detail
𝑴𝑿ሷ + 𝑲𝑿 = 0
Substitute into the differential equations:

𝑀11 0 −𝜔2 𝐴1 𝐾11 𝐾12 𝐴1 0

+ = −𝜔2 𝑴𝑿 + 𝑲𝑿 = 𝟎
0 𝑀22 −𝜔2 𝐴2 𝐾21 𝐾22 𝐴2 0

𝐾11 − 𝜔2 𝑀11 𝐾12 𝐴1 0

=
𝐾21 𝐾22 − 𝜔2 𝑀22 𝐴2 0

For non-trivial solutions, the determinant of the matrix should be equal to zero

𝐾11 − 𝜔2 𝑀11 𝐾12

2 =0
𝐾21 𝐾22 − 𝜔 𝑀22

Rao 6.9
Additional slide to complement videos
Undamped Free Vibration – 2-DoF– solution in more detail
From which a b c

𝜔4 𝑀11 𝑀22 − 𝑀11 𝐾22 + 𝑀22 𝐾11 𝜔2 + 𝐾11 𝐾22 - 𝐾12 𝐾21

2
−𝑏 ± 𝑏 2 − 4𝑎𝑐
𝜔1,2 =
2𝑎
𝑎 = 𝑀11 𝑀22 ; 𝑏 = − 𝑀11 𝐾22 + 𝑀22 𝐾11 ; 𝑐 = 𝐾11 𝐾22 − 𝐾12 𝐾21

Once we know frequencies, can substitute into the original matrix equation and extract A1, A2 for each frequency
2
𝑀11 0 −𝜔1,2 𝐴1 𝐾11 𝐾12 𝐴1 0
+ =
0 𝑀22 −𝜔1,2 2
𝐴2 𝐾21 𝐾22 𝐴2 0
The result – two modes of vibration, with different amplitudes for each mass. More about this in Unit 6
Additional slide to complement videos
Can be shown that all roots are real and positive when K and M are symmetric
and positive definite (xTKx >0 and xTMx >0)
 Another way of expressing it – eigenvalue
(characteristic value) problem
ሷ 𝑴𝑿 + 𝑲𝑿 = 𝟎
−𝜔2 𝑴𝑿 + 𝑲𝑿 = 𝟎
2
−𝜔 𝑿 + 𝑴 −𝟏
𝑲𝑿 = 𝟎 𝑲𝑿 − 𝜔2 𝑴𝑿 = 0
1
2
(−𝜔 𝑰 + 𝑴 −𝟏
𝑲)𝑿 = 𝟎 ( 𝑲 − 𝑴)𝑿 = 0
𝜔2
1
𝜆 = 𝜔2 ; 𝑺 = 𝑴−𝟏 𝑲 𝜇 = 2 ; 𝑫 = 𝑲−𝟏 𝑴
𝜔
(𝑺 − 𝜆𝑰)𝑿 = 𝟎 (𝜇𝑰 − 𝑫)𝑿 = 𝟎 D - dynamical matrix

This has solution if the determinant is equal to zero:

𝑺 − 𝜆𝑰 = 𝟎 characteristic equation 𝜇𝑰 − 𝑫 = 𝟎
This will lead to a similar quadratic equation. µ and λ are called eigenvalues. Once these are
known, can find frequencies and amplitudes (eigenvectors)
We can do it all in Matlab! Additional slide to complement videos
Rao 6.10 The code is [eigvector,eigvalue] = eig (K, M)
[eigvector,eigvalue] = eig (S, I) or eig (D,I) – remember to get 𝜔 back from substitution
The best bit? We can do it in Matlab!
k1 k2
m1 m2
−𝜔2 𝑴x + 𝑲x = 𝟎
x1 x2

Matlab can solve equations of the type 𝑲𝑿 − 𝜆𝑴𝑿 = 𝟎

The code is [eigvector,eigvalue] = eig (K, M) – remember to get 𝜔 back from substitution

For m1 = 1kg, m2 = 2kg, all k = 1N/m:

eigvalues = eigvectors =
𝜔2 0.2192 0 𝐴1 0.5615 1.0000
0 2.2808 𝐴2 1.0000 -0.2808
𝜔1= 0.47 rad/s and 𝜔2 = 1.51 rad/s
Torsional Vibration (2-DoF)
θ1
θ2
k2, G2, Ip2 k1, G1, Ip1

L2 L1
J2 J1
• Two disks on a shaft
• Angular displacements θ1 and θ2
• The mass moment of inertia of each disk is J1 and J2
• The two segments of the shaft have stiffness k1 and k2 (calculated from shear moduli G1 and G2,
polar moment of inertia IP1 and IP2) and given by
𝐺1 𝐼𝑃1 𝐺2 𝐼𝑃2
𝑘1 = and 𝑘2 =
𝐿1 𝐿2

Additional slide to complement videos

Torsional Vibration (2-DoF) – matrix
θ2 θ1 θ1
θ2
k2(θ2- θ1)
k2, G2, Ip2 k1, G1, Ip1

L2 L1
J2 k2(θ2- θ1) k1 θ1
J1

−𝑘1 𝜃1 + 𝑘2 𝜃2 − 𝜃1 = 𝐽1 𝜃ሷ1
−𝑘2 𝜃2 − 𝜃1 = 𝐽2 𝜃ሷ 2
𝐽1 𝜃ሷ1 + 𝑘1 + 𝑘2 𝜃1 − 𝑘2 𝜃2 = 0
𝐽2 𝜃ሷ 2 − 𝑘2 𝜃1 + 𝑘2 𝜃2 = 0
𝐽1 0 𝜃ሷ1 𝑘 + 𝑘2 −𝑘2 𝜃1 0
+ 1 =
0 𝐽2 𝜃ሷ2 −𝑘2 𝑘2 𝜃2 0

Additional slide to complement videos

Summary for 1 and 2 DoF
We can use different methods to construct equations of motion
Get (simultaneous) 2nd order differential equation(s)
For more than 1 DoF express equations in the matrix form
We expect oscillatory behaviour and so choose either a cosine function or complex exponential, Aeiωt
This substitution reduces the set on ODEs into a set of simultaneous algebraic equations, yielding one
value of ω for each degree of freedom
Then calculate 2 parameters (amplitudes or amplitude and phase angle) per degree of freedom from
initial conditions
Exactly the same for torsional vibration
Dynamics of Vibrating Systems
Part 3: Equivalence
Dr Oksana Trushkevych
 Introduction to Equivalence
Back to 1 DoF – what if the system is more complex than just a
mass and a spring?

Often, we can combine (lump) a number of elements

equivalent to another element,
– e.g. bundling a group of springs into one

Need rigorous criteria for bundling elements

– e.g. match kinetic or potential energy

JRK and OT:32

3.1 Equivalent Mass/Inertia
Elements
Back to the Example
Write down all energies:
1 1 2 1 2
𝑇 = 𝑚𝑥ሶ + 𝐼 𝜃ሶ
2
𝑉 = 𝑘𝑥
2 2 2
Terms are squared so no worries about directions
Use relationship between mass velocity and pulley rotation: 𝑥
𝜃=
2
𝑅
1 1 𝑥ሶ 𝑥ሶ
2
𝑇 = 𝑚𝑥ሶ + 𝐼 ሶ
𝜃=
2 2 𝑅 𝑅

𝐼
𝑚 + 2 𝑥ሷ + 𝑘𝑥 = 0
𝑅

34
 Equivalent mass or inertia elements
For a single-DOF system, we can model several mass or inertial elements as a single
lumped mass
Kinetic energy must be unchanged between original system and model

To do:
1. Write kinetic energy for mass/inertia elements
2. Write kinetic energy for an equivalent component, using same coordinates
3. Equate two energies
4. Solve for equivalent mass/inertia

Rao, section 1.8

Back to the Example
Write down all energies:
1 1 2 1
𝑇 = 𝑚𝑥ሶ + 𝐼 𝜃ሶ
2 𝑇 = 𝑚𝑒𝑞 𝑥ሶ 2
2 2 2
Terms are squared so no worries about directions
Use relationship between mass velocity and pulley rotation: 𝑥
𝜃=
𝑅
𝑥ሶ
2 ሶ
𝜃=
1 2
1 𝑥ሶ 𝑅
𝑇 = 𝑚𝑥ሶ + 𝐼
2 2 𝑅

𝐼
𝑚𝑒𝑞 = 𝑚+ 2
𝑅

36
 Mass/Inertia elements – rack & pinion

How many degrees

of freedom?

Additional slide to complement videos

Rack & Pinion example
Equivalent translational mass
– Kinetic energy of system
1 1
𝑇 = 𝑚𝑥ሶ + 𝐽0 𝜃ሶ 2
2
2 2

– Kinetic energy of equivalent system

𝑥 = 𝑅𝜃
1 𝑥
𝑇 = 𝑚𝑒𝑞 𝑥ሶ 𝑒𝑞 2 𝜃=
2 𝑅

Additional slide to complement videos

Rack & Pinion example
Relationships 𝑥ሶ 𝑒𝑞 = 𝑥ሶ
𝑥ሶ
𝜃ሶ =
𝑅
𝑇 = 𝑇𝑒𝑞

2
Substituting 1 1 1 𝑥ሶ
𝑚𝑒𝑞 𝑥ሶ 2 = 𝑚𝑥ሶ 2 + 𝐽0
2 2 2 𝑅
𝐽0
𝑚𝑒𝑞 = 𝑚 + 2
𝑅

Over to you! – what is equivalent inertia Jeq?

Additional slide to complement videos

3.2 Equivalent Stiffness Elements
 Springs in Parallel and Series
In parallel

𝐾𝑒𝑞 = 𝑘1 + 𝑘2

1 1 1
In series = +
K eq k1 k2

JRK and OT:42

 Modelling beams as springs

Assume m>>mbeam

Static deflection 𝑊𝑙 3
𝛿=
3𝐸𝐼

3𝐸𝐼
E- Young’s modulus 𝑊= 3 𝛿
I – second moment of area of cross- 𝑙
section of the beam keq
JRK and OT:43
Orange book, p131
 Equivalent stiffness elements
We can model several stiffness elements as a single stiffness
Potential energy must be unchanged between original system and
model

1. Write expression for potential energy of system

2. Write expression for potential energy of equivalent component
3. Equate the two
4. Solve

JRK and OT:44

Rao, section 1.7
 Real Springs

JRK and OT:45

Additional slide to complement videos
3.3 Introducing Damping
(more on damping in Unit 4)
 Damping
Damping is present in all oscillatory systems. It dissipates energy into either
heat or sound waves.
Examples:
– bending metal back and forth generates heat
– sound emitted from an object given a sharp blow

In vibration systems, interested in effect of damping:

– Decaying response in free vibration
– In forced vibration, this contribution to energy has to be provided by the
excitation

JRK and OT:47

 Types of damping -1
Coulomb: dry friction, sliding, relative vibratory motion transmitted across two
dry surfaces
Damping force is proportional to normal force (no dependence on velocity
magnitude but depends on its direction – acts in the opposite direction)

Linear viscous damping works well for slow fluid motion.

For fast fluid motion, modelling works better if damping varies with square of
velocity
a is a constant, parameterising strength of damping

JRK and OT:48

Rao, section 1.9
 Types of Damping -2
Hysteretic: internal losses in the material
• Area of the loop corresponds to energy absorbed per cycle
• Depends on temperature, independent of frequency, proportional to amplitude
or amplitude2

Displacement-squared: d = constant

Solid structural: b = constant

Structural damping often represented by
complex stiffness:
(solution of this eqn. is non-examinable)
JRK and OT:49
 Modelling of Damping
• Choosing correct form of damping is hard:
• Difficult to measure
• Difficult to model mathematically

• Linear viscous damping – a term that is proportional to velocity

𝑥ሷ + 𝑏𝑥ሶ + 𝜔2 𝑥 = 0
• Mathematically convenient
• Gives reasonable and often sufficiently accurate prediction of real damping

JRK and OT:50

3.4 Equivalent Damping Elements

Rao, section 1.9

 Effect on System Behaviour

JRK and OT:52

Additional slide to complement videos
 Example: Coulomb Damping
Material Kinetic Static
friction friction
Metal on metal 0.07 0.1
(lubricated)

Wood on wood 0.2 0.25

Steel on steel 0.3 0.75

(unlubricated)

Rubber on steel 1 1.2

JRK and OT:53

Additional slide to complement videos
 Example: Coulomb Damping
Because the sign of the damping force is always opposite to that of
velocity, the differential equation of motion for each sign is valid for half
cycle intervals:

To determine decay of amplitude, use work-energy approach:

Half cycle X1 → X-1:

k ( X12 - X-1)
1 2
- Fd ( X1 + X-1 ) = 0
2
1
Þ k ( X1 - X-1 ) = Fd
2
Half cycle X-1 → X2:
2Fd
X-1 - X2 =
k
JRK and OT:54
Additional slide to complement videos
 Example: Coulomb Damping
4Fd
For full cycle, the decay will be: X1 - X2 =
k
General solution for the motion:
𝜇𝑁
First half cycle, from R to L 𝑥 = 𝐴1 cos 𝜔 𝑡 + 𝐵1 sin 𝜔 𝑡 +
𝑘

𝜇𝑁
Second half cycle, from L to R 𝑥 = 𝐴2 cos 𝜔 𝑡 + 𝐵2 sin 𝜔 𝑡 −
𝑘

Special case: B1 = B 2 = 0

JRK and OT:55

Additional slide to complement videos
 Example: Coulomb Damping

JRK and OT:56

Additional slide to complement videos
 Example: Coulomb Damping

JRK and OT:57

Additional slide to complement videos
 Example: Coulomb Damping
Comparisons with viscous damping:

1. amplitude decays linearly rather than exponentially

2. frequency of oscillation is the same as that for undamped

oscillation

3. motion comes to a complete stop at a (potentially) different

equilibrium position (spring force < static friction)

JRK and OT:58

Additional slide to complement videos
 Equivalent Damping
Treat non-linear damping by modelling as an equivalent linear
system
Damping gives energy loss in the system
Equivalence condition: equal energy dissipation per cycle

∆𝐸 = ර 𝐹𝑑 𝑑𝑥
Note: equivalent damping is an approximation

JRK and OT:59

Rao, section 1.9
 Equivalent Damping
For linear viscous damping:

Where we’ve used: Move ω inside dt

From data book integral
of cos2x over 0- pi is pi/2

JRK and OT:60

Additional slide to complement videos
 Equivalent Damping
For Coulomb damping:
– Damping force Fd = mg is constant
– Work done per cycle = 4XFd
So viscous equivalent of Coulomb damping satisfies:
𝜋𝑏𝑒𝑞 𝜔𝑋 2 = 4𝑋𝐹𝑑
4𝐹𝑑 4𝜇𝑚𝑔
𝑏𝑒𝑞 = =
𝜋𝜔𝑋 𝜋𝜔𝑋

Can do the same for other types (e.g. air)

Equivalent damping will not give the same behaviour but may be good enough

JRK and OT:61

Additional slide to complement videos
 Example: Air Damping
𝐹𝑑 = ±𝑎𝑥ሶ 2

Determine the equivalent damping.

Assume SHM: 𝑥 = −𝑋 cos 𝜔 𝑡

8
p beqw X 2 = aw 2 X 3
3
8aw X
beq =
3p
JRK and OT:62
Additional slide to complement videos
 Damping equivalence
𝑥ሷ + 𝑏𝑥ሶ + 𝜔2 𝑥 = 0
Damping model application Damping force Equivalent viscous
damping beq
Linear viscous Slow fluid 𝑏𝑥ሶ 𝑏
Air damping Fast fluid ሶ 𝑥ሶ 2
𝑎 𝑠𝑔𝑛(𝑥) 8𝑎𝜔𝑋
3𝜋
Coulomb damping Sliding friction 𝛽 sgn 𝑥ሶ 4𝛽
𝛽 = 𝜇𝑚𝑔 𝜋𝜔𝑋
Displacement squared Material damping ሶ 2
d sgn(𝑥)𝑥 4𝑑𝑋
damping 3𝜋𝜔
Solid or structural damping Internal damping 𝑐 sgn 𝑥ሶ |𝑥| 2𝑐
𝜋𝜔

JRK and OT:63

 Conclusions Unit 03
Looks at equivalent models for mass, inertia, spring
constant and damping
The basic approach to lumping elements into equivalent
system is equating energy:
– Potential for springs
– Kinetic for masses/inertias
– Loss of energy per cycle for dampers
Considered different types of damping

JRK and OT:64

 For the next session
Examples – work through examples sheet
In a video James also asks you to think about derivation of
equivalent spring constant for springs in series
Do the quiz
On Friday we will cover more complex examples

Next lecture – damped systems, solving equation of motion with

damping

JRK and OT:65