damping determination at 20kHz 1

[shardin2014]

Hello All,

I am doing one experiment to find the damping of component made of polycarbonate material at 20 kHz frequency (the first natural frequency of the part is in the range 80 Hz). I want to use this damping in the FEA to simulate the response of the component.
Source that I am having can produce vibrations at fixed 20 kHz. Instrumentation I have are accelerometers and CRO.

Is there anyway to find the damping.

Regards

 

[GregLocock]

YUp, vibrate the part at 20 kHz, then disconnect the excitation. the decay in the amplitude of the vibration is directly related to the damping by a well known equation.

Cheers

Greg Locock


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

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

 

[shardin2014]

Thanks.

 

[jimkinney]

Method suggested by Greg at much lower frequency.

https://www.youtube.com/watch?v=PnsFFhKqXXo

Jim

Jim Kinney
Kennedy Space Center, FL

 

[hacksaw]

I am puzzled by your question and the lack of specificity.

The stated 80 Hz natural frequency suggests that the part is rather large and yet you want the damping at 20 KHz.

As far as modeling goes, you assume a damping parameter.

 

[FeX32]

The decay method works well with linear systems, and those with small number of DOF's.
You state that you need the damping at 20kHz, however the decay method will yield a damping coefficient at its resonance frequency only.
Many materials (likely polycarbonate too) have damping that is dependent on deflection amplitude, frequency, heat, ect.

 

[GregLocock]

No, if the system is being driven at 20 khz the decay will be at 20 khz, absent non linearities. try it!

Cheers

Greg Locock


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

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

 

[FeX32]

That's interesting Greg. I have never tried it at different frequencies, just nat. freq.
Do you have an analytical reference for this?

 

[shardin2014]

Thanks Greg,

I will find the damping by decay method as you suggested.
The lower frequency measurement issue will occur in hammer test, but if I am operating at 20 kHz, then decay method should also work.

Regards

 

[GregLocock]

A hammer test will mostly excite resonances, if you need to know at a specific non resonant frequency it won't help.

Cheers

Greg Locock


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

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

 

[FeX32]

http://www.ijesi.org/papers/AMAAS/Q1028.pdf

 

[GregLocock]

That's good, if the OP's structure obeys the same sort of trend then the damping seems to be approaching some asymtotic value as the modal frequencies increase, so the chances are pretty good that the modal damping of a (say) 21 kHz mode is representative for 20 kHz, so you can use an impact hammer, so you get a good damping estimate. Using a shaker for high frequency damping is problematic, but I'd assumed that was a given from the way the question was asked.

Cheers

Greg Locock


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

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

 

[Strong]

What kind/make/model impulse hammer has flat frequency response to 20-kHz? The presumption would be that the impact tip would be metal, yet the structure being impacted is polycarbonate plastic! Does the OP accelerometers have flat frequency response to 20-kHz? What about mass loading of accelerometer on polycarbonate plastic structure? I am trying to understand the OP and the responses for making practical measurements at 20-kHz on a plastic material.

Walt

 

[GregLocock]

For a damping determination at a resonance the flatness of the frequency response is somewhat unimportant, and yes I used to have some tiny impact hammers, about the size of a teaspoon, but lighter, like a cocktail stick, for HF work. Typical upper frequency limit for an impacthammer's force gage is way more than 20 kHz.

In fact if he flicks the part with his fingernail or a small screwdriver he'll generate a spike, so long as his accelerometers work that'd do. It's the decay characteristic that matters.

Good point though.

Cheers

Greg Locock


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

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

 

[FeX32]

Attached is a decent comprehensive reference for anyone reading and wants more.
Damping measurement in complex systems is always an interesting topic.

Unrelated to the attachment;
Seems first try is: 1) Setup experiment and locate (ideally a few) sensors onto the structure. Keep in mind the locations in relation to the theoretical mode shapes.
2) Use hammer test method to excite as many modes as possible. A harder hit will likely excite more of the high frequency modes.
3) Collect data of several trials.
4) Break data up into frequency components with FFT. Then packet it into the frequency you are interested in.
5) IFFT and perform 1dof decay method calculations.

Any comments?

Cheers,

 

[GregLocock]

"A harder hit will likely excite more of the high frequency modes. "

That's a testable hypothesis. Of course if the system is linear it'll make no odds.

Cheers

Greg Locock


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

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

 

[FeX32]

Or excite the higher frequency modes in the power spectrum with a higher amplitude so that they are easier to detect by experiment.

 

[shardin2014]

Higher modes will need high energy for excitation.

 

[GregLocock]

Pourquoi?

Cheers

Greg Locock


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

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