Hello all,
Good day. I just started a new project on metamaterials using a mechanical beam made of 3D printed PETG. My unit cell representing the repeated structure is about 50mm in length. I need to apply Floquet's periodic condition on the both ends of this beam on Comsol, But I don't know how to go about it.
The Comsol simulation is in 3D. I'm trying to figure out how to apply this periodic boundary condition and define my k values. If anyone can help urgently, I will really appreciate it. I want to get the eigen frequencies of the unit cell, representing other repeated patterns. Thanks a lot.
For applying Floquet's periodic conditions, you should study the following examples in the library of the Comsol:
1. RF_Module/EMI_EMC_Applications/frequency_selective_surface_csrr 2. RF_Module/Verification_Examples/fresnel_equations
3. Wave_Optics_Module/Gratings_and_Metamaterials/bandgap_photonic_crystal
4. Wave_Optics_Module/Gratings_and_Metamaterials/negative_refractive_index 5. Wave_Optics_Module/Optical_Scattering/scatterer_on_substrate
Nagi did a blog post for 2D dispersion in Comsol: https://www.comsol.com/blogs/modeling-phononic-band-gap-materials-and-structures/
Thanks so much, guys. I will go through it all. I deeply appreciate.
Just one more thing to reminder, in electromagnetic simulations of Floquet boundary conditions, the opposite faces should be the same meshes. Normally, one can do that by simply copying the meshes from the opposite face.
I've been able to implement the periodic boundary condition, the challenge now is that the bandgap diagram I'm getting is just straight horizontal lines. I know something is basically wrong somewhere, I'm still struggling to sort that out.
If you get a straight line on your band diagram that means you are sweeping your wavevector with a very small value (long wavelength). Try using a wavevector that will give you a wavelength of the order of your lengthscale.
1) Is it possible to reduce your geometry to 2D? From the image it looks like your geometry has a small out-of-plane thickness so I'm wondering if a 2D analysis would be more appropriate.
2) How are you defining/sweeping your wavevectors?
3) Are you normalizing your wave vectors by the lattice constants k*pi/L?
Your problem is a combination of 2 and 3 (which are related).
I believe this is what Ganesh was trying to convey in his response as well.
As I said in my previous post, using the model I linked in my first reply (simple 2D phononic crystal) you can see the difference in the band diagrams when the wavevectors (kx,ky) are not normalized with respect to the lattice constant (based on the models geometry). See the image attached -- in the geometry the lattice constants in x and y are the same (width = height = L).
Thanks so much Ganesh Patil and Miles Vincent Barnhart . I found your inputs extremely helpful. I appreciate
- Badges
- Science topic