FRF using laser vibrometer 1

[xydrex]

I have been measuring FRF of a structure using a laser vibrometer and electromagnetic shaker. The system has been set up to automatically give the FRF as output variable per unit input of voltage sent to the shaker from signal generator. E.g. Mobility in [(mm/s)/V]. Traditional techniques using accelerometers that I have come across always use a force transducer at the tip of the shaker so that FRF's are in output variable per unit of force input, so Mobility in [(mm/s)/N]. To leave the input in terms of volts implies that the shaker is consistent enough that a certain voltage will always give the same level of force. Is this a reasonable assumption? Do the dynamics of the shaker itself have something to do with this? Why wasnt voltage used as the input in the past with accelerometers?

Many Thanks......

 

[GregLocock]

It is not a reasonable assumption, unless you know quite a bit about the system.

The danger is that the electrical side of your shaker (and the coil and stinger) are now coupled into the system you are measuring, so your resonances are now of the test specimen+the stinger and shaker coil+the output stage of the power amp.

However, I have seen this done, it used to be reasonably common practice - I guess before they realised that compromising 100 man hours of work for the sake of a thousand dollar piece of instrumentation made little sense.






Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[xydrex]

Thanks for the previous post, I have now mounted the force transducer, the shape of FRF's are very similar, peaks are at the same freq's as expected but there are a few concerns:
*Firstly, the laser vibrometer software automatically creates animations of the deformed shapes. When I select shaker voltage as the reference, the animations are what you would expect & I get the same ones via FEA at similar freq's. But when I select force transducer as the reference, even though freq's match up good, the animations look very different. There are many more 'bulges' through the middle of the object (car bonnet) as if it was a much higher mode.
*Secondly coherence between force transducer & vibrometer is very low - roughly constant and below 0.3. The coherence between vibrometer and shaker voltage range between 0.5&0.9, peaking at the resonances.
*Finally there is a huge difference between FRF estimators H1 & H2 when Force Transd. is the reference. Although the shape of the two curves are very similar, H2 is nearly ten times the magnitude. When shaker voltage is the reference, H2 is only about 10% larger. From definition of H1 & H2, I wouldnt expect them to be too different.

Could the issue possibly be the signal amplifier b/w the force transducer and analyzer?

Many thanks.....

 

[GregLocock]

You have some huge instrumentation issue. Is the coherence low throughout the frequency range, or just at the resonances, or just at the antiresonances?

Until you get the coherence (vibrometer/force) sorted out don't worry too much about the mode shapes. it needs to be >.9, and I've had customers who would sniff at 0.95

Sounds to me like your force transducer is screwy.





Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[SrinivasAluri]

You could have speckle noise or "drop-outs" in your laser signal. Are you using any reflective tape/paint on your test specimen? One way to check for it is to input a pure sine wave and measure the output from laser in time domain, if you get a good clean signal then there are no drop-outs in your laser signal, if the signal is not clean then applying a reflective paint ot tape could solve your problem.