09-09-2022

Orthogonality and normal modes

Orthogonality and normal modes

1
 09-09-2022

Orthogonality and normal modes

Forced response

2
 09-09-2022

Forced response

Forced response

Animation related to (posted) problem #2 on string vibrations

https://www.acs.psu.edu/drussell/Demos/Pluck-Fourier/Pluck-
Fourier.html

3
09-09-2022

4
 09-09-2022

Introduction to nonlinear
vibrations
Course: ME 603
Instructor: Sripriya Ramamoorthy

Nonlinear vibrations
➢ Real systems are often nonlinear
➢ Linear analysis is insufficient to even qualitatively predict
some nonlinear behavior. Examples:
❑ Free response oscillations dependent on amplitude
❑ Existence of multiple solutions (and jump)
❑ Chaos
➢ Uniqueness of solution may not exist, and the response could,
in general, depend on initial conditions.
➢ Superposition principle does not hold.
➢ Analysis can be very hard except for simplified nonlinear
systems.

5
 09-09-2022

Nonlinear vibrations: Examples

https://www.acs.psu.edu/drussell/Demos/Pendulum/Pendulum.html

https://www.acs.psu.edu/drussell/Demos/Pendulum/Pendulum.html

6
 09-09-2022

Nonlinear vibrations: Examples

http://scienceworld.wolfram.com/physics/DoublePendulum.html

A double pendulum is a linear multi-DOF system for small angles.

But for large angles, it displays chaotic behaviour.

Chaotic system => Deterministic system, but extremely sensitive to

initial conditions. So, it is very difficult to predict its response.
Beyond a time-span, it starts to “appear random”.

Nonlinear pendulum
Newton’s second law of motion
F = ma
θ
a = L𝜃ሷ
F = -mg sinθ
T

m
Therefore, the equation of motion is:
-mg sinθ = mL𝜃ሷ
mg sinθ
This equation can be simplified as
mg cosθ 𝑔
𝜃ሷ + sin 𝜃 = 0
𝐿
𝑔
For small θ: 𝜃ሷ + 𝜃 = 0
𝐿 3
𝑔 𝜃 𝜃5
For large θ: ሷ𝜃 + 𝜃− + +⋯ =0
𝐿 6 120

7
 09-09-2022

Nonlinear pendulum

For large θ:
𝑔 𝜃3 𝜃5
θ 𝜃ሷ + 𝜃− + +⋯ =0
𝐿 6 120
T For not too large angles, this can be reduced to:
m 𝑔 𝜃3
𝜃ሷ + 𝜃− =0
mg sinθ 𝐿 6
mg cosθ
This equation is nonlinear due to the cubic term.

The type of nonlinearity analyzed here is called as

“Geometric Nonlinearity” because it arises from
geometry rather than material.

Nonlinear Spring

A spring-mass system with a nonlinear spring can be represented as:

𝑚𝑥ሷ = 𝑓 𝑥

Here 𝑓(𝑥) is the restoring force, 𝑥 is the deformation of the spring.

𝜕𝑓
If 𝜕𝑥 = −𝑘 is constant, then the spring is linear.
This is what we have considered earlier.

Now, consider the case where 𝑘 is not constant with 𝑥.

The type of nonlinearity analyzed here is called as “Material

Nonlinearity” because it arises from material property rather than
from geometry.

8
 09-09-2022

Nonlinear Spring
Here 𝑓(𝑥) is the restoring force, 𝑥 is the deformation of the spring.

Restoring force should act in the direction opposing the motion.

So, we require that 𝑓 𝑥 > 0 when 𝑥 < 0 and vice versa.

In other words, 𝑓(𝑥) should be an odd function of 𝑥

𝑓 𝑥 = −(𝑘𝑥 + 𝛾𝑥 3 )

𝛾 > 0 represents hard spring, whereas 𝛾 < 0 represents a soft spring.

Nonlinear Spring
In other words, 𝑓(𝑥) should be an odd function of 𝑥
𝑓 𝑥 = −(𝑘𝑥 + 𝛾𝑥 3 )

𝛾 > 0 represents hard spring, whereas 𝛾 < 0 represents a soft spring.

−𝑓 𝑥 −𝑓 𝑥

𝑘 increases with 𝑥:
𝑘 decreases with 𝑥:
Hard spring
Soft spring

9
 09-09-2022

Duffing’s oscillator

For cubic nonlinear spring, 𝑓 𝑥 = −(𝑘𝑥 + 𝛾𝑥 3 ):

This leads to the governing equation:

𝑚𝑥ሷ + 𝑘𝑥 + 𝛾𝑥 3 = 0 DUFFING’s OSCILLATOR

When 𝛾 ≪ 𝑘, we have weak nonlinearity. 𝛾

We can therefore write: =𝛼≪1
𝑘

This is of the same form as:

𝜃3
𝜃ሷ + 𝜔0 2 𝜃 − =0
6
1
𝑥ሷ + 𝑥 + 𝛼𝑥 3 = 0
Both equations are of the form: 𝜔0 2

10