Sound and Vibration Toolkit

[chericks1]

Hi WantToGolf,

When you post, you are given the option below the text area to include an attachment. Include all your screenshots, VIs, and other documents you want to send into a single zip file, browse to the file using this feature, and attach it along with the text you type.

Sorry for the confusion! Let me know if you have any problems.

Chad AE
Applications Engineer

[WantToGolf]

Ok, the files must be in zip format. No problem.

The VIs I used are from the SVT with no modifications so your code is the same as mine.

Procedure:
1. I clamp a long and thin piece of steel in a vice at one end. The steel piece was about 1/8" x 3" x 30".
2. I placed the accelerometer (channel 1) on the unsupported end and used the Level Meter(DAQmx) vi.
3. I bumped the piece of steel opposite the acc with my hand and recorded around a 25Hz spike as you see in the first screen shot. So now I know from this example that one of the natural frequencies of this piece of steel clamped as I have it is 25Hz.
4. Changed to Baseband Frequency Response vi and connected impulse hammer to channel 0.
5. Bumped the steel with the hammer and recorded 5 averages on the Bode Plot. Captured image.
6. Changed to Coherence Plot and bumped with the hammer, recorded 5 averages. Captured imgae.

[chericks1]

Hi WantToGolf,

What is the status on this issue? How's it going?

Please let me know,

Chad AE
Applications Engineer

[WantToGolf]

On 4/27 you mention forwarding this issue to a product support engineer I believe. So far, no one has contacted me so I'm kinda stuck right now.

[WantToGolf]

Chad,

Just want to update you. After some conversations with my FSE, Andrew Akers, and more reading I managed to modify an example application from the SVT. In short, I acquire a fixed amount of samples from the accelerometer and hammer using the hammer to trigger the acquisition. The output of the hammer is force(N) and the accelerometer of course is acceleration (g). I first convert the acceleration values to distance by double integration then feed both force and distance waveforms into the FRF. The magnitude output has the following units:

distance / force

Thanks to a paper by Brian J. Schwarz & Mark H. Richardson entitled "Experimental Modal Analysis" I discover that distance/force = Compliance and Dynamic Stiffness = 1/ Compliance. So, the reciprocal of the magnitude values gave me Dynamic Stiffness units of force/distance ie. N/micron. Making this adjustment resulted in what looked to be the correct plot.

Still concerned that I was getting the correct data I searched for another method to compare my results. More reading and trying to understand lead me to a method described by many authors as the method used in portable vibration analyzers. This method uses both Cross-Power Spectrum and Auto Power Spectrum to calculated dynamic stiffness. I added this calculation to my vi and plotted the resulting graph. I was surprised, and I might add very happy, to see virtually identical graphs. The FRF vs. the Cross-Power/Auto Power Spectrum calculations produced the same results.

Many thanks to you, and Andrew for your assistance in the project.

regards,

Glen

I will cc: Andrew Akers

Message Edited by WantToGolf on 01-05-2006 10:15 AM

[qBot]

Hello !

Does this toolkit support 24-bit soundcards (thus it's possible to use both 24-bit input and output configuratio)? (If it matters, I've got both LV 7.1 and 8.0)

Thanks in advacne 🙂

[Alfeos]

Is any possibility to find a vi that uses Cross-Power Spectrum and Auto Power Spectru. I need it to test my own vi that uses frf.

Please I am in really need of that

Or maybe to explain me how I can do that.

Message Edited by Alfeos on 11-24-2008 07:55 AM