Diaphram displacement on Helmholtz Resonator

Numerical methods and mathematical models of Elmer
Gary R
Posts: 162
Joined: 12 Apr 2012, 07:23
Antispam: Yes
Location: Long Beach CA, USA

Re: Diaphram displacement on Helmholtz Resonator

Post by Gary R »

Hi all;
Boy have you guys run with this one. I have been working on the model and haven't been keeping track of the posts. All are appreciated.
A clarification: The diaphragm is the driven element. I am trying to model the major drawback of a bass reflex enclosure. The driver becomes completely unloaded as the frequency approaches zero. The result is very low frequencies can completely de-cone the driver. Some form of low frequency high pass filter normally needs to be installed to prevent this from happening. If a passive network is used, the components are huge and are really expensive. This is the case where the speaker is a passive device and is fed directly by a power amplifier. A more economical approach is to supply the speaker with its own power amplifier and build in active filtering. Unfortunately for me my amplifier system doesn't have the preamplifier outputs needed for this approach.

The comment about the room side loading of the diaphragm is absolutely true. I have ignored this because in complicates the setup and will probably not have too much effect on the analysis at the low frequencies of interest. If I can get an analysis to run I may go back and retrofit things.

I'm still working through all of your posts and will be back when finished. My model is producing data but the data is total nonsense.

Thanks again.
Gary R.
Gary R
Posts: 162
Joined: 12 Apr 2012, 07:23
Antispam: Yes
Location: Long Beach CA, USA

Re: Diaphram displacement on Helmholtz Resonator

Post by Gary R »

I have tried incorporate your suggestions into my model. I found that the test case "Helmholtz Structure 3 " to be most helpful. But I still get zero output for displacement. The reason became obvious after some thought. The BC for the main body surfaces are set to Flux = 0.0. So, when the diaphragm is placed against the surface, no flux can flow into the cavity. Ergo - no input. I am going to try to create an area on the surface with a different BC from the rest of the surface. Hopefully this will fix things.

Gary R
Rich_B
Posts: 421
Joined: 24 Aug 2009, 20:18

Re: Diaphram displacement on Helmholtz Resonator

Post by Rich_B »

Hello,

Attached is an archive with one gmsh model and three different sif files. The use of Helmholtz and StressSolver for driving, driven, and driver & driven, are demonstrated.

Rich
Speaker6eg.png
(101.58 KiB) Not downloaded yet
Attachments
DoubleDrum.zip
(104.82 KiB) Downloaded 53 times
Gary R
Posts: 162
Joined: 12 Apr 2012, 07:23
Antispam: Yes
Location: Long Beach CA, USA

Re: Diaphram displacement on Helmholtz Resonator

Post by Gary R »

Hi all,
I reworked the project using the test case "Helmholtz Structure 3" and the suggestions here as a guide. It still doesn't work. The coupling between the disk and the resonator seems to be missing. Attached is a tar ball of the whole project. I hope someone can help. The "SaveScalar" solver seems to be working. It records the frequencies but only records zero for the Displacement.
Screenshot.png
Screenshot.png (22.94 KiB) Viewed 768 times
Gary R.
Attachments
SpeakerCabinet.tar.gz
(177.16 KiB) Downloaded 44 times
Rich_B
Posts: 421
Joined: 24 Aug 2009, 20:18

Re: Diaphram displacement on Helmholtz Resonator

Post by Rich_B »

Hello Gary,

A couple of items:

1. Add
Calculate Mesh Pieces = Logical True
to the simulation section of case.sif. This produces this message in the solver log:
CalculateMeshPieces: Number of separate pieces in mesh is 2

Right now the mesh is in two pieces as confirmed in this screenshot, after hiding boundary 11:
TwoMeshPieces.png
TwoMeshPieces.png (256.53 KiB) Viewed 764 times
The diaphragm mesh is separate from the enclosure mesh, and needs to be joined together, as discussed in this post:
viewtopic.php?p=27130#p27130

2. Solver 3 section states:
Equation = Helmholtz Equation
Variable = -dofs 2 Pressure Wave
this states there are two variables for Pressure Wave, which are the real and imaginary components of the wave. In the last boundary conditions, there is this statement
Wave Flux 3 = 1
This BC may or not be needed, but if included it should be Wave Flux 1 or 2 but never 3, since there are only two components.

3. A boundary condition for zero displacement for the rim of the diaphragm is required, specifically:
Displacement 1 = 0.0
Displacement 2 = 0.0
Displacement 3 = 0.0
There may be other comments, but the above items need to be addressed.

Rich.
Gary R
Posts: 162
Joined: 12 Apr 2012, 07:23
Antispam: Yes
Location: Long Beach CA, USA

Re: Diaphram displacement on Helmholtz Resonator

Post by Gary R »

Thanks Rich for the suggestions. The main problem now seems to be the lack of coincidence between the nodes of the face and the nodes of the diaphragm. I have not been able to find a setting in gmsh to correct this nor have I bin able to correct it using -merge 1 with ElmerGrid. I do not have access to Salome or can't afford a copy. So that is out. There seems to be no coupling between the two parts. Any suggestions will be sincerely appreciated.

Gary R.
kevinarden
Posts: 2237
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: Diaphram displacement on Helmholtz Resonator

Post by kevinarden »

Salome is open source and doesn't cost anything.
-merge doesn't work because the nodes between the nodes on the two parts are not close.
You need to boolean the two parts into one part before meshing, either in your cad or in gmsh.
kevinarden
Posts: 2237
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: Diaphram displacement on Helmholtz Resonator

Post by kevinarden »

Here is a geo file that imports the brep and unions the two volumes so that the mesh conforms at the boundary.
3DHelmholtzResonator.geo
(123 Bytes) Downloaded 47 times
Here is a geo that creates the geometry entirely in gmsh with no cad tool involved
speaker2.geo
(367 Bytes) Downloaded 51 times
Gary R
Posts: 162
Joined: 12 Apr 2012, 07:23
Antispam: Yes
Location: Long Beach CA, USA

Re: Diaphram displacement on Helmholtz Resonator

Post by Gary R »

Kevin;
The .geo files definitely solve the edge coincidence problem but end up with only one body. There is no way to assign different materials to the resonator and the diaphragm. Not sure where to go from here.
kevinarden
Posts: 2237
Joined: 25 Jan 2019, 01:28
Antispam: Yes

Re: Diaphram displacement on Helmholtz Resonator

Post by kevinarden »

This geo file will get to 3 bodies with edge coherence.
speaker2.geo
(715 Bytes) Downloaded 50 times
Post Reply