Hi, I would like to setup a study with the HelmholtzBEM solver. The problem would look similar to the AcouSTO Engines' shielding effect tutorial (page 20): I have a source in the middle of a domain and I am interested in the pressure on a sphere around it.
For example, I might have a 2D disk (but I would try to have a 3D source too at some point) and a sphere around it. I looked up the test case, and in that case the domain appears to be a 2D planar domain with line boundaries. In my case, I would have two surfaces: a flat disk and a sphere around it. I would like to apply a flux boundary condition to the disk and solve for the pressure and flux at the sphere. Is that possible?
If so, I tried to setup my geometry with Salome, and I exported the result as a UNV. In ElmerGUI the imported mesh will see the surfaces as bodies and their borders as boundaries. However, I need to apply the boundary condition to the disk. How would I achieve that?
Setting up a HelmholtzBEM Study
-
- Site Admin
- Posts: 4828
- Joined: 22 Aug 2009, 11:57
- Antispam: Yes
- Location: Espoo, Finland
- Contact:
Re: Setting up a HelmholtzBEM Study
Hi
This is an old module which has limited use since the BEM matrix is assembled in full. This means that the memory consumption and speed of solution grow rather unfavorably with problem size. Efficient BEM solvers never create the full matrix but use iterative multilevel methods instead, such as the multipole expansion. Also, the solver only works in serial so parallel computing is not available to resolve the bottleneck.
These are the reasons why in documentation it is under "obsolite and experimental solvers". ElmerGUI knows nothing of this.
If you want to try to use it there is one consistency test "HelmholtzBEM". Start from a very small case so that you can see where you hit the sealing.
-Peter
This is an old module which has limited use since the BEM matrix is assembled in full. This means that the memory consumption and speed of solution grow rather unfavorably with problem size. Efficient BEM solvers never create the full matrix but use iterative multilevel methods instead, such as the multipole expansion. Also, the solver only works in serial so parallel computing is not available to resolve the bottleneck.
These are the reasons why in documentation it is under "obsolite and experimental solvers". ElmerGUI knows nothing of this.
If you want to try to use it there is one consistency test "HelmholtzBEM". Start from a very small case so that you can see where you hit the sealing.
-Peter
Re: Setting up a HelmholtzBEM Study
Makes sense, thanks for the insight!