Discussion Forum

Hi there,

I am using Comsol 4.2's acoustics module to simulate some thermoacoustic instabilities.
The whole process can be described by an inhomogeneous Helmholtz equation.
A simple n-tau-model is used to relate the velocity fluctuation upstream of the heat release zone to the pressure fluctuations within the heat release zone. This should render the involved matrices non-symmetric and result in truly complex eigenfrequencies (i.e., eigenfrequencies with non-vanishing real and imaginary parts). However, the resulting eigenfrequencies are real and identical to the resonance frequencies which are obtained, if no heat release is considered. Moreover, the solver log states that symmetric matrices were found. Even if the solver is forced to use a non-symmetric algorithm, no improvement is gained. Interestingly, the documentation states "Because they are independent of the pressure, the solver ignores any dipole and monopole sources unless you are solving a coupled eigenvalue problem." under the topic "Eigenfrequency Study" . And indeed, I modeled the heat release zone using a monopole source. I implemented the velocity relation with an average operator taken from the Model Couplings bullet in the model's Definitions node. I thought that Comsol automatically recognizes that it should perform a coupled eigenvalue problem analysis. Apparently, this seems not to be the case. So how is this done?

I attached a minimum working example to this request.

Thank you in advance
Georg Mensah