Two degree-of-freedom systems

Find the natural frequencies and describe the modes for the linear vibration system below.

Printable View

- August 25th 2006, 10:48 AMbobby77mechanics:please help
Two degree-of-freedom systems

Find the natural frequencies and describe the modes for the linear vibration system below. - August 25th 2006, 12:13 PMCaptainBlackQuote:

Originally Posted by**bobby77**

1. Decide on your variables (use the deflection of the first trolley from its

equilibrium position as x, and the second trolley from its equilibrium position

as y, positive to the right).

2. Write out the equations for the accelerations of the trolleys in terms of

x and y.

3. Rewrite as a vector equation X''=AX', where A is now a 2x2 matrix, and

X=[x,y]'.

4. Now the normal modes correspond to the eigen vectors of A.

5. If we now rewrite the equation in 3 in terms of the normal modes the

new A will be diagonal (with the eigen values of the original A on the

diagonal), and the solution to the DE is:

Y=exp(sqrt(A)t)

which will give the natural frequencies.

RonL - August 25th 2006, 01:06 PMCaptainBlackQuote:

Originally Posted by**CaptainBlack**

,

and on trolly 2:

,

where the notation is as in the attached figure.

So:

Now substitute in the given values:

RonL - August 25th 2006, 01:39 PMCaptainBlackQuote:

Originally Posted by**CaptainBlack**

which may be rewritten as:

.

So writing , this becomes:

,

where:

Now we may find the eigen values and vectors of by hand, but I will resort to numerical packages at this point:

Code:

>A=[-73.333,6.667;6,-80]

-73.333 6.667

6 -80

>help eigen

function eigen (A)

## eigen(A) returns the eigenvectors and a basis of eigenvectors of A.

## {l,x,ll}=eigen(A), where l is a vector of eigenvalues,

## x is a basis of eigenvectors,

## and ll is a vector of distinct eigenvalues.

>

>{aa,bb,cc}=eigen(A)

-69.5171+0i -83.8159+0i

>aa

-69.5171+0i -83.8159+0i

>bb

0.867895+0i -0.536649+0i

0.496748+0i 0.843805+0i

>cc

-69.5171+0i -83.8159+0i

>

and .

RonL

I will complete the last part of this tommorow, its midnight here now and I

am going to bed. - August 25th 2006, 09:58 PMCaptainBlackQuote:

Originally Posted by**CaptainBlack**

in terms of the eigen vectors, so we put:

where:

,

and:

Also:

.

Now we rewrite the differential equation in terms of :

.

Hence:

,

and:

.

So the differential equations for , and

are uncoupled, and of the form which give simple harmonic motion, so the

usual methods can now be applied to obtain the natural frequencies.

RonL - August 25th 2006, 10:49 PMCaptainBlackQuote:

Originally Posted by**CaptainBlack**

,

then:

,

where what this means when is diagonal is obvious (and I don't propose

to go into what conditions we need on for this to work in general

or what it means for non-diagonal matrices).

RonL - August 27th 2006, 12:55 PMbobby77
according to the file attached..

I get characterstic equation as:

7500s^4+1150s^2+43=0

where s=natural frequency

solving this equation gives s=+/- 0.3i and s= +/-0.25i

is it correct natural frequency..? - August 28th 2006, 05:06 AMCaptainBlackQuote:

Originally Posted by**bobby77**

equation, and how you found it.

RonL - August 28th 2006, 08:59 AMbobby77
here is the characteristic equation i got..

- August 28th 2006, 09:03 AMCaptainBlackQuote:

Originally Posted by**bobby77**

kN/m rather than converted them to N/m.

RonL