r.banner
Newbie level 1
Hello,
I have successfully done cloaking simulations with COMSOL's RF module. Trying to do the same for the Helmholtz equation, I ran into some problems. I think this has to do with incorrect boundary conditions, but I can't see the solution right now. Maybe someone can help me!?
The simulation results are shown in the following pic:
It looks like scattering from the boundaries.
I've used the following setup in COMSOL:
2D > Mathematics > Classical PDEs > Helmholtz equation
Global Definitions > Parameters: k (wavenumber)
Equations: grad(-c grad(u)) + a u = f
Outside the ring: c=1, a = -k^2, f=0
Inside the ring: c = usual cloaking parameters, a = -k^2, f=0
Zero flux: top and bottom boundary
Flux/Source: left and right boundary g = -k*sin(k*x)
For the RF wave equation I used PMLs and scattering boundary conditions. That's not possible for the Helmholtz equation and I thought that it wouldn't be necessary. Can someone tell me what's going wrong there!?
Regards,
RB
I have successfully done cloaking simulations with COMSOL's RF module. Trying to do the same for the Helmholtz equation, I ran into some problems. I think this has to do with incorrect boundary conditions, but I can't see the solution right now. Maybe someone can help me!?
The simulation results are shown in the following pic:
It looks like scattering from the boundaries.
I've used the following setup in COMSOL:
2D > Mathematics > Classical PDEs > Helmholtz equation
Global Definitions > Parameters: k (wavenumber)
Equations: grad(-c grad(u)) + a u = f
Outside the ring: c=1, a = -k^2, f=0
Inside the ring: c = usual cloaking parameters, a = -k^2, f=0
Zero flux: top and bottom boundary
Flux/Source: left and right boundary g = -k*sin(k*x)
For the RF wave equation I used PMLs and scattering boundary conditions. That's not possible for the Helmholtz equation and I thought that it wouldn't be necessary. Can someone tell me what's going wrong there!?
Regards,
RB
