Hi,
I have been trying to calculate the propagation matrix that is found in Rayleigh II integral, using simulated data from COMSOL. I want to use such matrix in pressure-to-pressure calculations with simulated data. I have seen that COMSOL's 'Far-field calculation' solves Kirchhoff-Helmholtz integral, however it seems not to solve for arbitrary pressure values in space, i.e. computes far-field plots. Neither I have found if it is somehow possible to impose Dirichlet's boundary condition in COMSOL's 'Far-field calculation', which could lead to Rayleigh II integral from K-H integral.
First attached you can find the model 'PMatrixRayleighII.mph'. It consists of a 2D model with a compact dipole (two point sources at the origin, opposite in phase and closely positioned), outputting a pressure vector from a dataset at fixed y-coordinate ∆y. As a starting point I'm interested in plan-parallel planes, where the propagation matrix is Toeplitz and can be filled with the pressure vector obtained from just one dipole. In order to validate the calculated matrix, I made an additional model 'PMatrixValidation.mph'. In this case I used 3 point sources and obtained pressure vectors from datasets at fixed y0 and y1, such that y1 - y0 = ∆y. Then in MATLAB, I convolved the pressure vector at y0 with the propagation matrix, and compared it with COMSOL's pressure vector at y1; which yielded no similarity between them. I tried other source configurations too but no success.
I couldn't find then a source-independent way in COMSOL to solve pressure-to-pressure convolutions (which Rayleigh II integral solves); but I'm unsure if the computation and/or the validation model I have done are wrong for this particular purpose. I would like to know if you have any suggestions, corrections or ideas about this, I have been stuck for about two months with no luck. If you need me to clarify something else please let me know. Many thanks in advance for your kind attention and support, I look forward to hearing from you.
Best regards,
/E
I have been trying to calculate the propagation matrix that is found in Rayleigh II integral, using simulated data from COMSOL. I want to use such matrix in pressure-to-pressure calculations with simulated data. I have seen that COMSOL's 'Far-field calculation' solves Kirchhoff-Helmholtz integral, however it seems not to solve for arbitrary pressure values in space, i.e. computes far-field plots. Neither I have found if it is somehow possible to impose Dirichlet's boundary condition in COMSOL's 'Far-field calculation', which could lead to Rayleigh II integral from K-H integral.
First attached you can find the model 'PMatrixRayleighII.mph'. It consists of a 2D model with a compact dipole (two point sources at the origin, opposite in phase and closely positioned), outputting a pressure vector from a dataset at fixed y-coordinate ∆y. As a starting point I'm interested in plan-parallel planes, where the propagation matrix is Toeplitz and can be filled with the pressure vector obtained from just one dipole. In order to validate the calculated matrix, I made an additional model 'PMatrixValidation.mph'. In this case I used 3 point sources and obtained pressure vectors from datasets at fixed y0 and y1, such that y1 - y0 = ∆y. Then in MATLAB, I convolved the pressure vector at y0 with the propagation matrix, and compared it with COMSOL's pressure vector at y1; which yielded no similarity between them. I tried other source configurations too but no success.
I couldn't find then a source-independent way in COMSOL to solve pressure-to-pressure convolutions (which Rayleigh II integral solves); but I'm unsure if the computation and/or the validation model I have done are wrong for this particular purpose. I would like to know if you have any suggestions, corrections or ideas about this, I have been stuck for about two months with no luck. If you need me to clarify something else please let me know. Many thanks in advance for your kind attention and support, I look forward to hearing from you.
Best regards,
/E