k-Wave

1. Nikilesh
   

   Member
   Joined: Jan '15
   Posts: 2

   Hi !

   I am working on simulating the propagation of lamb waves in aluminium plates. I've tried using a heterogeneous medium for this application. I cannot use any liquid medium for my plate to be inserted in. However, my simulations seem to be giving me weird results. I am aware that the acoustic impedance of the plate is very high compared to that of air outside the plate. After going through the forum posts, I saw that the new release supported wave propagation in elastic media. Thereby, I tried using the new functions pstdElastic2D and pstdElastic3D. However, I still do not see any difference in results. The source tone burst seems to black out the screen and then slowly starts disappearing. Hence, I wanted to know if the k wave toolbox along with the latest release, can be used for Ultrasonic Fault Detection Applications such as the one mentioned above? If not, can you please suggest any other toolbox available?

   Thanks

   Posted 9 years ago #

2. Bradley Treeby
   

   Developer
   Joined: Mar '10
   Posts: 1,640

   Hi Nikilesh,

   Whether or not k-Wave is the best tool for the job depends quite a lot on the simulation settings. Just to double check, is the blacking-out of the display a plot scale or stability issue? The plot scale defaults to [-1, 1], and can be changed by setting the optional input parameter 'PlotScale'.

   Does your source start in the air and propagate into the plate, or does it start in the plate?

   Brad.

   Posted 9 years ago #

3. Nikilesh
   

   Member
   Joined: Jan '15
   Posts: 2

   Hi !

   Thank you so much for your quick reply!

   As you mentioned, I checked the simulation by changing the 'PlotScale' parameter. It still blacks out the display.

   Also, to answer your second question, I've tried both things. The source is defined as a linear array to give out a toneburst. If the source is inside the plate and the medium is defined as a homogeneous one(only aluminium), then there is no issue with the simulation. However, the waves generated will escape at the boundaries and not reflect as required for lamb waves. Now, if I use a heterogeneous medium comprising of air and aluminium, I can see the tone burst for a fraction of a second and then the black out appears first in the aluminium region and then the entire gridspace. In this case, my source was placed at the boundary between the air and aluminium plate. Next, when I tried to shift my source inside the plate but with a heterogeneous setting outside, the same thing occurred. Thinking in an another direction, I tried to toggle the attenuation coefficients, but still the same results. Moreover, my output signal sensor measures a very huge value of the pressure. As far as the application is concerned, the simulation has to be done in a heterogeneous medium. However, my source will still have to be placed at the boundary between the air and plate layer.

   Thanks

   Posted 9 years ago #

4. Bradley Treeby
   

   Developer
   Joined: Mar '10
   Posts: 1,640

   Hi Nikilesh,

   It sounds like you are seeing an instability, which could be rectified by using a smaller (potentially much smaller) time step. However, for such strong impedance discontinuities and a source at the boundary, the spectral method used in k-Wave is probably not your best choice. You might be better off looking at another type of elastic wave solver. There are some listed here.

   Brad.

   Posted 9 years ago #

5. datbui
   

   Member
   Joined: Nov '22
   Posts: 2

   Hi Dr. Treeby,

   I would like to learn more about the reason why "strong impedance discontinuities" would make the spectral method become unstable. Could you please direct me to some articles as a starting point?

   Thank you.

   Regards,

   Dat

   Posted 10 months ago #

RSS feed for this topic

Reply

You must log in to post.