Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
Modeling Radiation Losses From Pinhole in RF Cavity
Posted 24 juil. 2017, 17:37 UTC−4 2 Replies
Please login with a confirmed email address before reporting spam
Hello,
I am trying to model radiation losses for a microwave cavity resonator with a small optical port (pinhole smaller than the microwave wavelength). I am interested in evaluating the effect of pinhole diameter and location on the quality factor and the shape of the magnetic field eigenmode.
My cavity is a volume of air with a Material Link specifying the exterior boundaries as copper, and an Impedance Boundary Condition to account for resistive losses in the cavity walls.
The way I have tried to model the radiation losses is by adding a cylinder of air (as long as the cavity wall is thick) with the same Impedance Boundary Condition on the sides, and a Perfectly Matched Layer on the end, to simulate radiation into the surrounding environment. I have joined a toy model of this technique with a very simple rectangular waveguide cavity as an example of what I am talking about.
The problem I have is that the model is very slow to converge (>100 iterations of the Eigenvalue Solver in some cases) and for certain pinhole diameters or depths does not converge at all. In addition, the presence of a pinhole does not seem to affect the eigenmode at all : the quality factor does not decrease as expected, and the mode shape does not seem to be perturbed.
For example, in the model that I have joined, the cavity with a pinhole has a slightly higher (10721 vs 10717) Q-factor than the reference cavity; however the difference is so small that I think it's safe to interpret it as numerical artifacts arising from the differences in meshing between the two models.
Is there a better way to model these radiation losses? The PML method seems very inefficient since the majority of the elements is outside the cavity, and I am not interested in the properties of the field outside the cavity.
Thanks,
-- Louis Haeberle
MSc Student (Physics)
Silicon Spintronics Research Group (Pr. Pioro-Ladrière)
Université de Sherbrooke
P.S. I get a "File extension error" when I try to attach the MPH files to this forum post, which has never happened before... Here are links to download them:
www.filehosting.org/file/details/681887/ReferenceCavity.mph
www.filehosting.org/file/details/681888/PinholeCavity.mph
I am trying to model radiation losses for a microwave cavity resonator with a small optical port (pinhole smaller than the microwave wavelength). I am interested in evaluating the effect of pinhole diameter and location on the quality factor and the shape of the magnetic field eigenmode.
My cavity is a volume of air with a Material Link specifying the exterior boundaries as copper, and an Impedance Boundary Condition to account for resistive losses in the cavity walls.
The way I have tried to model the radiation losses is by adding a cylinder of air (as long as the cavity wall is thick) with the same Impedance Boundary Condition on the sides, and a Perfectly Matched Layer on the end, to simulate radiation into the surrounding environment. I have joined a toy model of this technique with a very simple rectangular waveguide cavity as an example of what I am talking about.
The problem I have is that the model is very slow to converge (>100 iterations of the Eigenvalue Solver in some cases) and for certain pinhole diameters or depths does not converge at all. In addition, the presence of a pinhole does not seem to affect the eigenmode at all : the quality factor does not decrease as expected, and the mode shape does not seem to be perturbed.
For example, in the model that I have joined, the cavity with a pinhole has a slightly higher (10721 vs 10717) Q-factor than the reference cavity; however the difference is so small that I think it's safe to interpret it as numerical artifacts arising from the differences in meshing between the two models.
Is there a better way to model these radiation losses? The PML method seems very inefficient since the majority of the elements is outside the cavity, and I am not interested in the properties of the field outside the cavity.
Thanks,
-- Louis Haeberle
MSc Student (Physics)
Silicon Spintronics Research Group (Pr. Pioro-Ladrière)
Université de Sherbrooke
P.S. I get a "File extension error" when I try to attach the MPH files to this forum post, which has never happened before... Here are links to download them:
www.filehosting.org/file/details/681887/ReferenceCavity.mph
www.filehosting.org/file/details/681888/PinholeCavity.mph
2 Replies Last Post 28 juil. 2017, 11:07 UTC−4
Please login with a confirmed email address before reporting spam
Posted:
7 years ago
I would try putting a half sphere on the outside, centered on the pinhole, and use the scattering boundary condition.
If you are seeing no change in Q then probably you need to make the pinhole bigger. For pinholes much smaller than a wavelength the radiation will be really small and is apparently overwhelmed by ohmic losses.
D.W. Greve
DWGreve Consulting
If you are seeing no change in Q then probably you need to make the pinhole bigger. For pinholes much smaller than a wavelength the radiation will be really small and is apparently overwhelmed by ohmic losses.
D.W. Greve
DWGreve Consulting
Please login with a confirmed email address before reporting spam
Posted:
7 years ago
Updated:
7 years ago
Hello Dr. Greve,
thank you for your suggestion! Using the Scattering boundary condition has cut the running time from over 10 hours to 30 minutes, and the Q-factor decreases monotonically as the pinhole diameter increases, as expected.
-- Louis Haeberle
MSc Student (Physics)
Silicon Spintronics Research Group (Pr. Pioro-Ladrière)
Université de Sherbrooke
thank you for your suggestion! Using the Scattering boundary condition has cut the running time from over 10 hours to 30 minutes, and the Q-factor decreases monotonically as the pinhole diameter increases, as expected.
-- Louis Haeberle
MSc Student (Physics)
Silicon Spintronics Research Group (Pr. Pioro-Ladrière)
Université de Sherbrooke