Determining natural frequency

[CoastalMaineBird]

It was a standard sledge hammer to which someone had attached a set of electrical contacts. 

Well, with the impact hammers, you set up the DAQ to trigger when the voltage rises above X volts.  The sensor is built into the head.

If you use pre-trigger recording, you can get the entire impact event, although you may not need to. 

[g0dam0ng1n53ct5]

Is it correct that the stimulus hammer is significant in that it supplies a signal similar to the accelerometer, which is used to determing the natural frequency?

It's not just used to trigger recording, is it? Our test was intended to be performed while dropping a weight into the structure, so as to determine the natural frequency during the drop. We can record the force at time at the point where the load from the dropped weight is applied, but no vibration signal is output by it.

Would this be enough to find the tower's natural frequency?

[preston_johnson]

We have a nice impact hammer example that will ship with Sound and Vibration Measurement Suite 2010.  All the calculations for frequency response already exist in our Sound and Vibration Measurement Suite.  With these, you will not need to test and verify the calculations, that becomes NI's responsiblity. 

Are you interested in learning more about the impact test and two channel frequency response examples we have in the sound and vibration measurement suite?

for example:

http://zone.ni.com/reference/en-XX/help/372416C-01/sndvibtk/freqncy_response_mag_phase/

You can evaluate these VIs at:

https://lumen.ni.com/nicif/us/evalsvmeassuite/content.xhtml

Let us know how we can help. 

[g0dam0ng1n53ct5]

I'm definitely interested in learning more about the sound and vibration suite. However, I'm not sure the stimulus-response route is the way we want to go. It was recommended that we could perform any impact, without gathering its signal, and then simply gather the accelerometer's output and find the resonant frequency of the structure using that.

Is it necessary to use a known stimulus signal to do this?

Is there a way in the vibration suite to do what we need?

[CoastalMaineBird]

Is it correct that the stimulus hammer is significant in that it supplies a signal similar to the accelerometer, which is used to determing the natural frequency?

--- Yes, it measures the force applied to the structure.

It's not just used to trigger recording, is it?

--- No - it's main purpose was to establish how much force was applied, so that the true transfer function was obtained.  Since it was doing that already, it has handy to use it as the trigger.

Our test was intended to be performed while dropping a weight into the structure, so as to determine the natural frequency during the drop.

--- If all you want is the frequency, then this is probably sufficient.

We can record the force at time at the point where the load from the dropped weight is applied, but no vibration signal is output by it.

--- I don't know what that means. 

Would this be enough to find the tower's natural frequency?

---Well, I'm not a mechanical engineer, so take this with a grain of salt, but if the weight STAYS on the structure, then it will affect the resonant frequency, I think.

It needs to be a glancing blow, such that the structure is free to do whatever it will, without interference from the weight.

In any case, the code shows how to process the signal: integrate it (if you need to convert acceleration to displacement), window it, transform it, then pick out the peak frequency.

Hope that helps. 

[CoastalMaineBird]

 It was recommended that we could perform any impact, without gathering its signal, and then simply gather the accelerometer's output and find the resonant frequency of the structure using that.

--- Thinking more clearly about this, if all you need is the frequency, then yes, you'll be fine doing that.

Is it necessary to use a known stimulus signal to do this?

--- The only thing necessary is that the stimulus cover the true resonant frequency.  In other words, if the resonance is at 100 Hz, hitting it with nothing but 13 Hz force will not tell you anything.

But some sort of strike is a wide-band stimulus, so I think you're OK. 

Is there a way in the vibration suite to do what we need?

I'm sure there is, but it's going to boil down to the same DC removal - integrate - window - FFT - peak-pick as shown above.

If you feel comfortable with the processing that's going on in my code, then adapt it to fit.

If not, then the SVT may make it easier. 

[dsb@NI]

If any are still interested in impact test example for LabVIEW, please refer to related post:

http://forums.ni.com/ni/board/message?board.id=231&message.id=6798&jump=true

[dsb@NI]

The posted example includes features such as interactive UI, data acquisition, software triggering, rejection criteria, averaged FRF, and save to UFF 58b. With that feature complexity, some of the elegance was lost. Please weigh in with your opinion on where you see value in (advanced) examples provided by NI for impact test and other specific applications.