Modal Analysis Library - With Triaxial Accelerometer

[VLg]

Hi all,

I have a potential need to develop a modal analysis program, and as a newbie to this field, it is great that the modal analysis library is available for us to build the application easier. After going through the library and its examples, I have some questions which I hope to get some advice from you guys. Thanks in advance 🙂

https://forums.ni.com/t5/Sound-and-Vibration/Getting-Started-with-Modal-Analysis-in-LV/gpm-p/3533797...

In my setup, I would like to obtain the mode shape of a plate. The vibration data will be acquired using an impact hammer with one triaxial accelerometer using the roving accelerometer method. The hammer and accelerometer are hooked up to a NI-9234 in a USB chassis.

1) Version of the SI Estimate FRF VI to use?

After studying the example "Record Impact Testing FRF data.vi" in the library, I observed that this is based on a single-axis accelerometer (1x response channel). To calculate the FRF at each accelerometer position, the "SI Estimate FRF.vi" (SISO) version was used. I would like to check that if I am going to use a triaxial accelerometer, would it be correct to use the MIMO version of the "SI Estimate FRF.vi" as shown below? In my case, there will still be only 1x stimulus channel (hammer), but 3x response signals.

2) Calculation of the Mode

I explored the other example program "Modal Parameter Identification.vi" by using the sample data "Plate FRF measure - 1.dat", which seems to be based on a single-axis accelerometer response channel. In this example, the plate has a total of 27 points where the response is measured, therefore 27 sets of FRF data (27 points x 1 response channel per point) are used to identify the mode, e.g. using "MP_FDPI.vi". From an online tutorial, it is stated for a roving triaxial accelerometer setup, we can calculate the complete mode if we have one column in the FRF matrix as shown below: 

 

If my plate comprises of 10 measurement points, the 30 sets of FRF data (10 points x 3 response channels per point) seems to correspond to this column in the FRF matrix - is it correct if I format these data into a 1D array of FRF and use the Mode extraction VIs just like in the 1-axis accelerometer scenario ? - please see the attached file "Mode Extraction with 3-axis Accelerometer.png" for a better visualization. 

3) Visualizing by animation

Currently, with a plate, it can be represented rather easily with a 3D surface defined by a X and Y array, and correspondingly the animation of each mode shape can be realized without much issue in the example program. I would like to ask, does anyone have any experience or comments with visualizing the mode shapes for more complex objects like a car or an aircraft using LabVIEW's built-in libraries (like the 3D Picture Control functions?). Or would it be more advisable to do this in a 3rd party software (such as ModalVIEW or MEScopeVES)?

Thanks a lot!

Best regards,

Victor

[mich@sg]

Hi Victor,

I am trying to do the same than what you described in your post. Did you find answers to your questions? If yes, I would be very happy if you could share them with the community 🙂 

Thx,

Michael

[VLg]

Hi Michael,

Unfortunately I have not found the answers to these yet. Due to some reasons, this potential project has been set aside until further notice, but I would also be grateful if someone else can shed some light on this matter too. 🙂

Best regards,

Victor

[Ryan4]

Lysandros offered some practical advice to consider when acquiring from passive acoustic sensors.