Discussion Forum

Hello Daniel Atehortua

Your Discussion has gone 30 days without a reply. If you still need help with COMSOL and have an on-subscription license, please visit our Support Center for help.

If you do not hold an on-subscription license, you may find an answer in another Discussion or in the Knowledge Base.

Any idea?? Thanks.

Hi Dan

In the Pressure Acoustics, Transient interface you can change the Fluid model from Linear Elastic to Viscous and/or Thermally conducting. This will add bulk attenuation due to viscosity and thermal conduction.

If you want to add your own user defined loss term you need to enter it as a domain Dipole Source. Here you can define a user defined term that is proportional to the time derivative or the gradient of the pressure.

More detailed thermal and viscous losses in the acoustic boundary layer can actually be modeled using the Linearized Naveri-Stokes, Transient interface. This is like Thermoacoustics in the time domain.

It all depends on which type of attenuation you need to model.

Best regards
Mads

Hi everyone,

The transient acoustic pressure study does not take into account wave attenuation when traveling accross a material (i.e. there is no lineal elastic with attenuation study option as the pressure acoustic module does). So, If I want to include attenuation losses in the model, what would be the best way to do it.

Thanks in advance.

Dan

Hi Mads,

Thank you very much for your reply. Actually, I have not been able to do it yet, because I do not want a dipole source, nor a monopole one, not even I want to switch to a viscous model. I want just to model the acoustic field as a function of time when an incident ultrasonic wave is propagating through a solid media (i. e. steel). So I just want to add the pressure time derivate times lossy factor (from the media i. e. absortion and scatter coeficient) to the transient pressure acoustic equation. But I have not been able to do it yet. Anytime I try it, there is a mistake. Shall I add it as a weak expression? Global equation? or how?

Thanks in advance.

Dan

Hi Dan

The monopole and dipole sources defined on the domains are just names we have given these two types of domain terms you can add. They are not dipole/monopole sources as you classically think of them in points. Use the monopole domain source to add a right hand side to your equation. It can be anything you like. See for example the nonlinear_acoustics_westervelt_1D model in the model library. Here we have added a non-linear term to the wave equation.

Th other way is to select linear elastic with attenuation and define your own attenuation expression. This would result in the same. Look at where the terms enter the equations.

Best regards
Mads

Dear Mads,

Thank you very much for your time and the help!! Thanks to your words, I have been able to do it!!!!!

Best Regards,

Dan

Hi, Daniel,

I would like to do some simulation of elastic waves passing through solid with large attenuation. How did you accomplish it? can you please provide some guidance?

Regards,
Jing

Hi Jing,

It is something that can be as complicated as you desire. For example, attenuation in lossy media may be frequency dependent, which means additional terms in weak formulations. Another thing is that when handling elastic waves, you have to account for longitudinal components, but also for transversal ones (and conversion modes when the wave front hits an interface), which might become the problem more complicated. And if your problem is time-dependent, things might get even worse!!!!

The final answer depends on the problem you have and the purpose of the analysis. Is your analysis frequency based? Time dependent? What attenuation model can describe the phenomenon? And, not less important, what comsol version do you have available?

Best Regards,

Daniel

Daniel,

Thank you very much for your prompt and detailed reply.
I want to simulate a xdcr with piezoelectric disk as active material, and a composite (rubber mixed with tungsten power) as backing material. The backing material provide extremely high damping to the piezo excitation, and high absorbing of elastic wave.
The problem troubles me is that it seems none of the damping mechanism in simulation is able to provide enough damping and absorbing comparable to what I obtain experimentally.
The case is frequency dependent, I am working around 100kHz to 600 kHz frequency range. I prefer a time dependent analysis to see waveform generated in water. A frequency domain analysis is also valuable.
My comsol version is 5.2

Best regards,
Jing

xdcr is transducer in short form

Ok,

You have the latest comsol version, so I think it can be very useful as it might already include the attenuation model you would need into transient analysis, which means that you would not need to implement them from zero in PDE nor weak forms. So, you would save lots of time!!!! I have not worked with latest version, only up to 5.0, so I cannot tell you how it should look like.

There are some things to have in mind: from your problem physics, what attenuaton model should you have to use? If you are using the right-for-the-purpose model, then you can always adjust constant parameters of the model in order to fit the model results with experiments. Of course, you cannot "go crazy" here, but it is acceptable if you change the parameters within some expected ranges. There is no exact answer, you just have to play a little bit with it and to analyse what is going on.

The results depend on other model features too, as it should be. Do you have the right boundary conditions? Are you modelling in 2D? have you checked mesh and time step dependence? Have you checked absolute and relative error dependence? Are material properties checked too?

An advise: I would look for making the model as simple as possible by discarding what you do not really need to model or by simplifying the phenomenon (making rigth assumptions). The simulation can be very time consuming and you might need enough computational resources available. You are dealing with and ultrasonic wave, several equations to solve at the same time (¿Can you switch to iterative solvers or seggregate algorithms?), and the problem is transient. You need to have a well enough conditioned problem. If so, you can optimize your comsol model.

Best Regards,

Daniel

Thank you, Daniel.
It's very kind of you to provide all those thoughtful suggestions. I will try and see how.

Hi Daniel

I have similar problem. I want to simulate soil as an elastic medium and add wave attenuation on time domain in the model. I have Comsol 5.2 but I don't know where I can add this feature.
Can you help me by detail because I am a beginner?

I have similar problem. I want to simulate soil as an elastic medium and add wave attenuation on time domain in the model. I have Comsol 5.2 but I don't know where I can add this feature.
Can you help me by detail because I am a beginner?

Hi,

If you right-click on Linear Elastic, you can add a Damping subnode.

Regards,
Henrik

I have similar problem. I want to simulate soil as an elastic medium and add wave attenuation on time domain in the model. I have Comsol 5.2 but I don't know where I can add this feature.
Can you help me by detail because I am a beginner?

Hi,

If you right-click on Linear Elastic, you can add a Damping subnode.

Regards,
Henrik

Hi,

I have a similar problem. I want to do a transient acoustic pressure study trough a solid. I want to take into account attenuation but I do not see the attenuation model.

You mention that I can right click on Linear Elastic to add a Damping subnode. I can't right-click on the linear elastic option inside the "Transient Pressure Acoustics Model" section, inside the settings window for "Transient Pressure Acoustics Model". So where do I find the "Linear Elasic' that you mention?

Is there also another way to add the values of attenuation?

I am new in Comsol

Regards,
Guillermo

Hi

On the elastic domains (solids) where you use Solid Mechanics you have to right click the Linear Elastic node.

In the fluid domains where you use Pressure Acoustics, Transient you change the fluid model to include viscous and/or thermal damping (this is in the settings of the Transient Pressure Acoustics Model node). If you want to add a more complex damping mechanism you need to add a Domain Monopole Source and write a user defined expression.

Best regards
Mads

I can see the options.

Thanks.