k-Wave

1. antares
   

   Member
   Joined: Feb '19
   Posts: 5

   Hello,

   recently I started experimenting with k-Wave for ultrasound in air problems. Working with sources, sensors, and reflections on objects with higher densities works great. Now I would like to make a model of a membrane that separates two areas (spaces in 3D) of air, is moved by an incoming wave in one area, and moves the air in the other area. Is that possible?

   If an object of higher density is defined as separation, my current understanding is that it does not move, i.e. the transmission from one area to the other happens by passing of a wave through the "membrane" and two boundaries, not by movement of the "wall". If no direct solution exists, something like taking the results of a first simulation, doing some calculations on it, and feeding it into a second would be OK for me, but I don't know how.

   Regards,
   Heinrich

   Posted 5 years ago #

2. antares
   

   Member
   Joined: Feb '19
   Posts: 5

   push

   Posted 5 years ago #

3. bencox
   

   Developer
   Joined: Mar '10
   Posts: 292

   Hi Heinrich,

   I'm not quite clear what you mean in your second paragraph, so I'll answer the question in the first paragraph.

   You could model wave propagation through a three-layer elastic medium in k-Wave, but perhaps that's not what you want. (The wave speed in the middle medium - the membrane - would depend on the tension and so would, I suspect, be anisotropic.) Alternatively, you could assume the membrane moves with the fluid, ie. does not affect the wave motion at all, and extract the particle velocity from the model to determine the membrane motion. If you really want to couple the equation for the wave in the membrane and the equation for the wave propagation in the surrounding fluid, then that it beyond k-Wave's capabilities.

   Finally, the vibration of a membrane is strongly affected by the boundary conditions at its edge, in that they determine the modes (think of a drumskin, Chladni figures etc). Without explicitly modelling the membrane, it's not clear to me how these conditions would be incorporated.

   Best wishes,
   Ben

   Posted 5 years ago #

4. antares
   

   Member
   Joined: Feb '19
   Posts: 5

   Hi Ben,
   Thank you very much for clarifying this. If the membrane moves with the fluid, this appears to me like an ideal membrane, something that you would like to have in a microphone, right? My case is different: A hollow metal object in air has sections with thin walls, and it is known experimentally that these show resonance, as does the void, and that an acoustic wave from the outside passes to some extent inside the object.

   So it appears to me that I cannot use the 'ideal membrane', since waves would just pass it as if it is not there, and the void of the object could not show resonance?

   The three-layer elastic medium is perhaps what I have tried and tried to explain in my second paragraph: A model with inner space (air)- shell/membrane (aluminum) - outer space (air). My current understanding is that the shell in this case does not move at all, and that penetration of the shell by a wave means passing the solid material and the two surfaces. Can you confirm this? In this case, the membrane resonance shows that the model is not adequate. Since k-Wave is capable of delivering very good results for parts of my problem, I would have to live with partial models in different simulators, one of them for the membrane itself. But knowing this would be a big step for the project.

   Regards,
   Heinrich

   Posted 5 years ago #

5. bencox
   

   Developer
   Joined: Mar '10
   Posts: 292

   Hi Heinrich,

   Yes, your description is right. I don't think k-Wave is the right tool for your problem. There's a discussion of the transmission of an acoustic wave through a thin tensioned membrane in a duct in Morse & Ingard (p543) that might be helpful. I think you need to solve the coupled equations describing the wave motion in the air and the motion of the membrane, with the relevant boundary conditions.

   Kind regards,
   Ben

   Posted 5 years ago #

6. antares
   

   Member
   Joined: Feb '19
   Posts: 5

   Thanks again!

   H.

   Posted 5 years ago #

RSS feed for this topic

Reply

You must log in to post.