MDOF Flashcards
What is the damped and undamped MDOF system matrix equation of motion?
Damped: [M]{x..(t)} + [C]{x.(t)} + [K]{x(t)} = {f(t)}
Undamped: [M]{x..(t)} + [K]*{x(t)} = {f(t)}
[M] = Mass matrix
{x..(t)} = Acceleration
[C] = Damping Matrix
{x.(t)} = Velocity
[K] = Stiffness matrix
{x(t)} = Dispalcement
{f(t)} = Prescribed time-variting Loads
Descibe the equation below in words:
[M]{x..(t)} + [C]{x.(t)} + [K]*{x(t)} = {f(t)}
Equilibrium of inertia, damping, elastic (or stiffness, or
restoring) forces and external forces, acting in the direction
of each DOF, is satisfied at each instant of time.
Why can’t the matrix equation for a damped MDOF system be solved with normal partial differential equations?
It can’t be solved by normal PDEs because there are coupling terms. This means that there are terms for each floor/node in both equations needing to be solved. So the movement at each node effects the other nodes.
What are the two groups of methods that find general solutions to the equation of motion and which is better for linear dynamic problems?
1) Mode superposition methods
2) Direct time integration methods
Mode superposition methods are better for linear dynamic problems as:
- it provides a physical insight into the vibration behaviour of the structure modelled.
- it is considerably more efficient computationally.
what does the so called expansion theorem state?
Any N-dimensional vector can be presented as a linear combination of N-dimensional vectors which are orthogonal and, therefore, linearly independent (i.e. none of them can be expressed as a linear combination of others).
If all nodes vibrate with the same frequency how would you find the displacement and amplitude at each level?
{x(t)} = {X} * sin(wt + θ)
{x..(t)} = -w²{X} * sin(wt + θ)
{X} = {Φr} * Ar
{Φr} = Mode shape r
Ar = Amplitude factor of mode r
How would you find the angular frequency of the system (w) if it was an undamped system?
det[ -w² *[M] + [K] ] = 0
(Once expanded this is called the characteristic equation or frequency equation)
det[ a,b ; c,d ] = ad - bc
What is the standard eigenvalue problem for an undamped system?
wr² [M]{Φr} = [K]{Φr}
or
( [K] - wr^2 * [M]) * {Φr} = 0
r = mode shape number
wr = angular frequency of mode r
what is the generalised coordinate?
It is the response of the equivalent SDOF system corresponding to mode r.
qr(0) = initial displacement of mode r
q.r(0) = Initial velocity of mode r
How do you calculate the mass and stiffness of each mode?
Modal mass = mr = {Φr} T * [M] * {Φr}
Modal stiffness = kr = {Φr} T * [K] * {Φr}
How can you find the lowest natural frequency of a system?
You find it by minimising the strain energy and maximising the kinetic energy. I.e. the easiest movement mode 1
How do you get the final solution for the undamped free vibration of a MDOF system?
x(t) = (N, r=1) Sum( {phir} * qr(t) )
Use the equation for qr transient but make wrd = wr and all damping terms = 0.
What is the spectral matrix and where is it used?
The spectral matrix [Ω^2] is defined as an NxN diagonal matrix containing squared natural frequencies on its diagonal.
[K] * [φ] = [Μ] * [φ] * [Ω^2]
φ = Node Shape
What is Classical, Proportional damping or Rayleigh damping and when is it possible?
Systems where diagonalisation of the damping matrix is possible are called Classical, Proportional damping or Rayleigh damping.
It is achieved when the matrix is proportional to the mass and stiffness matrices in the form:
[C] = α[M] + β[K]
Where α & β are constant numbers
For a MDOF system under harmonic load, what affects the amplitude and response?
Amplitude is frequency-dependent and increases around all resonant frequencies.
The response will be at the same frequency as the excitation.
The transient response is a superposition of a lower frequency decaying sinusoid having frequency which is lower than the frequency of the constant steady-state sinusoid superimposed on it. This lower frequency corresponds to the natural frequency of the system