Learning Forum

[hamderson]

Hello, 

 

I am currently conducting a modal analysis on a structure that is used in ballistic testing - the structure holds the test specimens in place. I have conducted experimental work and am completing a frequency analysis on my results form the ballistic test. I hope to identify the vibration of the supporting structure in my data and thus have completed a modal analysis in Ansys. I constructed the geometry in solidworks and then imported it as a solid part hoping that this would simplify the procedure.

 

Thankfully, a solution was found first time. However, the solution is not what i expected and i was hoping that someone could help me understand it give advice on how to improve the analysis. I expected there to be a dominant natural frequency and some harmonics of this but the solution has showed that there are many natural frequencies that are very close together. Is this correct? 

 

Any input is appreciated, thanks. 

 

[Wenlong]

Hello Hamderson,

 

According to the company policies I am not allowed to download attachments so if you can share more details about your model through screenshot (and insert as an image, not screenshot) that will be very helpful. 

 

Thank you!

Regards,

Wenlong

[hamderson]

Hi Wenlong, 

This is the geometry, imported as a solidworks part file. 

 

These are the results.

 

Thanks for the reply, I have attached some images.

[Wenlong]

Hi Hamderson,

 

So you don't have a constraint on the model and it is subjected to rigid body motion. You can tell by looking at the mode frequencies, there are 0s. Please apply a boundary condition and see.

 

Regards,

Wenlong

[hamderson]

Wenlong, 

 

I have applied a fixed boundary condition to the bottom surfaces of the model, i thought this would be appropriate as the support is reasonably heavy and rests on the floor. There are now no zero values in the results but i am still receiving many clumped natural frequencies. 

[Wenlong]

Hi Hamderson,

 

It is usually a long and iterative process to match the simulation with the experiment and there is no guarantee. There are several questions I would ask myself when I see a mismatch from the experiment:

 

1. Does my simulation represent the experiment setup well? For example, are there any extra mass I need to add to the model? In the experiment are all the components "perfectly bonded" like the model does? Are my material models

2. I would also plot and play the deformed shape to get an idea of how the equipment deforms under each frequency. That may also provide some insights into what may cause the mismatch. 

 

Regards,

Wenlong