Issue in Bloch Analysis

[njain094]

Hello,

I am trying to find the critical periodicity of a discrete structure using Bloch Analysis. From Bloch Analysis, I get the critical periodicity as 1x2. Now, when I compare the first frequency obtained from Bloch Analysis for the critical periodicity (i.e. 1x2) to the enlarged 1x2 unit cell model with conventional periodic boundary conditions, they do not match. I get a lower frequency from Bloch Analysis (67) as compared to the enlarged model (91). Note that, all other higher mode frequencies match for the two cases. Ideally, these two (first) frequencies should match as both the models represent the same periodicity. (They do match in a continuum case).

Please let me know if anyone has faced this issue.

Regards

 

[FEA way]

Can you attach some pictures showing your models and say more about their settings (element types, materials and so on) ?

 

[njain094]

Hello,

I apologize for getting back to you so late.

I have attached a pdf containing the models used in enlarged 1x2 unit-cell periodic analysis (page 1) and Bloch Analysis (page 2). I am using a simple beam element (B21) with just one element between the joints as I am not creating a fine mesh. The mesh doesn't matter though when we compare results from the enlarged unit cell model and Bloch analysis. Moreover, the material is elastic.

https://files.engineering.com/getfi...1-afce-41b4-bed1-95d0400352b7&file=Models.pdf

Regards,
Nidhish

 

[FEA way]

It seems that the pdf file upload failed for some reason. Please try attaching it again.

 

[njain094]

Yes, I was struggling to do that. The link was not appearing. I had to manually do that.

 

[njain094]

Hello FEA way,

Did you get a chance to look at the figures?

 

[FEA way]

If the mesh refinement doesn't help (it often does) then, due to the complexity of the problem itself, I would try with even simpler model. What I mean is a structure as simple as possible for these analyses (something like a single element test). It may help you verify the current approach.

 

[njain094]

Hello,

I did with a simpler model and it works.