Divergence on piezoelectric domain

Please login with a confirmed email address before reporting spam

Hello to all,

I'm trying to create a 2D model of an ultrasonic NDT using Elastic Waves, Time Explicit with Electrostatics. I used the tutorial "Angle Beam Nondestructive Test" (https://www.comsol.com/model/angle-beam-nondestructive-testing-78811) as starting point for my own model, but I'm having some trouble with the Piezoelectric domain.

I assigned an edge as Electric Potential boundary condition and to create a pulse I inserted an Analytic function as voltage input, just like the tutorial. The simulation starts fine, but after a while, the electric potential in the piezoelectric domain grows and I can't figure it out why. I attached a plot of the average electric potential on the edge, in this image is visible the input pulse and how exponentially the voltage grows after some point.

I thought that it could be a mesh problem, so I tried to refine it, but that didn't work. Anyone faced a similar problem?

Thank you in advance,

Rodrigo Pires



5 Replies Last Post 26 giu 2024, 14:16 GMT-4
Mark Cops COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 6 months ago 17 mag 2024, 15:22 GMT-4

The graph is indicating that the solution is diverging. Since there can be various reasons for this, you would need to submit your .mph model file in order to get some specific suggestions for your model.

-Mark

The graph is indicating that the solution is diverging. Since there can be various reasons for this, you would need to submit your .mph model file in order to get some specific suggestions for your model. -Mark

Please login with a confirmed email address before reporting spam

Posted: 6 months ago 21 mag 2024, 08:58 GMT-4

Hello Mark, thanks for your reply!

I’m attaching the .mph file to this topic. I’m sorry, but I had to exclude the mesh and result to get the right file size for uploading.

Thanks in advance, Rodrigo

Hello Mark, thanks for your reply! I’m attaching the .mph file to this topic. I’m sorry, but I had to exclude the mesh and result to get the right file size for uploading. Thanks in advance, Rodrigo


Mark Cops COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 6 months ago 21 mag 2024, 09:20 GMT-4
Updated: 6 months ago 21 mag 2024, 09:29 GMT-4

It looks like there is no damping in any of the solids, even the "Damping material". It could help to add some Rayleigh damping - similar to what was done in the angle_beam_ndt model. Secondly, both piezo domains currently have the same coordinate system used for the poling direction. Is this correct? It would seem that due to the different angle, each piezo would need a separate base vector system to define its poling direction relative to the mount angle.

It looks like there is no damping in any of the solids, even the "Damping material". It could help to add some Rayleigh damping - similar to what was done in the angle_beam_ndt model. Secondly, both piezo domains currently have the same coordinate system used for the poling direction. Is this correct? It would seem that due to the different angle, each piezo would need a separate base vector system to define its poling direction relative to the mount angle.

Please login with a confirmed email address before reporting spam

Posted: 5 months ago 25 giu 2024, 12:19 GMT-4

Hi Mark, sorry for the late reply.

I didn’t notice the damping subsection on the physics model. I tried to include it and worked in that case! But when I tried to use the same configuration but changing the inlet wave shape for a frequency sweep with higher maximum frequencies (about 2.5 MHz) and the divergence appeared again. I used the same damping values than before, is this incorrect? I’m just started learning about acoustics for this project, so I’m a noob haha. I attached the voltage chart of the “receiver” transducer. I also attached a gif of the results, the inlet piezoelectric locks in a weird pressure pattern after the input electric potencial goes to zero.

Hi Mark, sorry for the late reply. I didn’t notice the damping subsection on the physics model. I tried to include it and worked in that case! But when I tried to use the same configuration but changing the inlet wave shape for a frequency sweep with higher maximum frequencies (about 2.5 MHz) and the divergence appeared again. I used the same damping values than before, is this incorrect? I’m just started learning about acoustics for this project, so I’m a noob haha. I attached the voltage chart of the “receiver” transducer. I also attached a gif of the results, the inlet piezoelectric locks in a weird pressure pattern after the input electric potencial goes to zero.


Mark Cops COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 5 months ago 26 giu 2024, 14:16 GMT-4
Updated: 5 months ago 27 giu 2024, 02:09 GMT-4

The latest excitation looks to be higher frequency. It is possible that the mesh somewhere is not fine enough to capture the wave, hence causing the solution to blow up. It is difficult to see where this might be happening from the .gif. I would change the colorbar scale limits back to default and find what time and where in the domain the pressure is first blowing up. Then locally refine the mesh there.

The latest excitation looks to be higher frequency. It is possible that the mesh somewhere is not fine enough to capture the wave, hence causing the solution to blow up. It is difficult to see where this might be happening from the .gif. I would change the colorbar scale limits back to default and find what time and where in the domain the pressure is first blowing up. Then locally refine the mesh there.

Reply

Please read the discussion forum rules before posting.

Please log in to post a reply.

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.