modal analysis of an oscillator with non uniform stiffness in tension/compression

[BeFEA]

I am trying to find the fundamental period of a special oscillator which consists of a mass attached to a spring that has stiffness k1 in compression but stiffness k2 in tension.

Is there any way to model this behavior in Finite Elements Method?

I can not even think of a hand solution. Can anyone help me with an idea how to calculate manually or through FEM?

 

[Denial]

Your mass as described, in the absence of damping or any other forces, will move in an alternation between two different simple harmonic motions:[ ] one for the compression part and one for the tension part.[ ] The resulting motion will have a period of
pi * [sqrt(m/k₁) + sqrt(m/k₂)]

Or have I missed something?

 

[GregLocock]

The easy way to do it physically would be to use a swept sine signal to excite it, you would probably use the frequency at which the peak to peak displacement divided by the applied force is a maximum as the 'resonance'.

There are many programs that could be used to model this, but linear FEA is not one of them. You need a time domain simulation. I'd suggest Matlab would be most instructive, but you could use simulink, adams, motionview, dymola, excel etc etc.



Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[BeFEA]

Thank you very much for your ideas. I think solved the problem. It was quite easy though.

 

[bkal]

GregLocok,

I am not disagreeing with you, but my first instinct would be to give the mass an initial displacment and let it go. Do you think there is an issues with this approach?

 

[GregLocock]

That is a very good suggestion.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[BeFEA]

I think it is a very good and theoretically correct approach. Thanks everyone for the suggestions.