Hello,
I am trying to model acoustics in a simple geometry. Later on I want to use a modified Helmholtz equation, which is why I'm trying to mimic the pressure acoustics boundaries in the Helmholtz equation module.
The problem is that these do not match.
I've started by modelling the acoustics in the pressure acoustics module.
I've used a 2D axisymmetrical geometry existing of a simple rectangular with the pressure at a part of the bottom. Sound hard boundaries on the outside and a sound soft boundary on the top. Water is the medium.
I've used the same 2D axisymmetrical geometry for the Helmholtz equation module. I've made the boundaries with dirichlet boundary conditions.
The results from these 2 models are quite different for "high" frequencies. 20kHz is comparable, 200kHz and 570 kHz is very different.
My first idea was of course that my model conversion was wrong. But after checking, double checking and triple checking, I don't see an error. So I tried to model both in a 3D geometry with the same boundaries (actually an eight of the cyllinder to lower the necessary memory for solving this.) and at that point I get the same result for the helmholtz module as I have for the pressure acoustics module (which is the same for the 2D axisym and the 3D geometry).
This lead me to believe the boundaries are correct and the problem lies somewhere else.
So, now to the question: Does anyone have an idea why the helmholtz equation does not work correctly in the 2D axisym geometry? (Do I perhaps have to add some special boundary to define the axisymmetry?)
It's also not about the mesh sizes, I've tested this aswell.
Some explanation or thought would really be appreciated as I'm stumped.
Thanks in advance for taking your time.
~Niels
I am trying to model acoustics in a simple geometry. Later on I want to use a modified Helmholtz equation, which is why I'm trying to mimic the pressure acoustics boundaries in the Helmholtz equation module.
The problem is that these do not match.
I've started by modelling the acoustics in the pressure acoustics module.
I've used a 2D axisymmetrical geometry existing of a simple rectangular with the pressure at a part of the bottom. Sound hard boundaries on the outside and a sound soft boundary on the top. Water is the medium.
I've used the same 2D axisymmetrical geometry for the Helmholtz equation module. I've made the boundaries with dirichlet boundary conditions.
The results from these 2 models are quite different for "high" frequencies. 20kHz is comparable, 200kHz and 570 kHz is very different.
My first idea was of course that my model conversion was wrong. But after checking, double checking and triple checking, I don't see an error. So I tried to model both in a 3D geometry with the same boundaries (actually an eight of the cyllinder to lower the necessary memory for solving this.) and at that point I get the same result for the helmholtz module as I have for the pressure acoustics module (which is the same for the 2D axisym and the 3D geometry).
This lead me to believe the boundaries are correct and the problem lies somewhere else.
So, now to the question: Does anyone have an idea why the helmholtz equation does not work correctly in the 2D axisym geometry? (Do I perhaps have to add some special boundary to define the axisymmetry?)
It's also not about the mesh sizes, I've tested this aswell.
Some explanation or thought would really be appreciated as I'm stumped.
Thanks in advance for taking your time.
~Niels