Strange Mode in Dispersion Curve with HFSS
For wake filed estimation I need to plot dispersion curves for Higher Order Modes (HOM).
90 degree slice with one PEC and one PMC as boundary condition is used to study dipole modes.
First ten Higher Order Modes are studied.
Master Slave Boundary condition is used with step of 30 degree phase change.
I am getting one 8th mode at 9.27466 GHz which remains same on all phase sweep.
Can anybody help what this mode is and why it is not changing with change of phase.
Modes with a constant frequency for all phase points are simple resonances -- like a LC tank.
What do its fields look like?
Thanks for reply I agree that it is resonance frequency but my question is why this resonance frequency is not dependent on phase difference between master and slave boundary. All other modes have proper dispersion curves.
Like HOM1 4.60027, 4.55111,4.48995,4.44566, 4.41721,4.40160,4.39665
While this one remains 9.27466 for all this range
Sorry, a poor choice of wording on my part -- when I said "simple", I should have said "isolated". The particularity of this resonance is (most likely) such that the fields associated with the resonance mechanism are not coupled between adjacent periods.
Like the LC tank, it won't really matter what else is around it -- it's behaviour will pretty much just be determined by L and C. Similarly, your fields here are constrained in some way (most likely highly localized to the ridge I see in the attached image), such that the adjacent periods represented by the M/S boundaries have no discernible impact. You should be able to verify this by looking at the Poynting vector at various phase points -- there shouldn't be much energy coupling through the boundaries.
This obviously isn't the case for regular, propagating modes.
Ok I have multi cell structure unit cell of which I am simulating.
what I get from you answer is following
So these modes are not coupled at all to the neighboring cells?
Resonance frequency is defined by the shape of cell as shape defines L and C of structure ?
These type of modes are also called trapped modes?
Please let me know if all concluded points are correct
They shouldn't be. There should exist some transverse plane within the structure (not necessarily the M/S boundaries you have chosen), through which power does not flow.
The frequency of this mode will be determined by the structure, yes.
Not that I'm aware of. All one can really say is that they're resonant.
Actually my problem is that I have to give some logical explanation for this type of modes as such type of modes are present in my monopole as well as dipole configurations.
If I know what type of mood it is then I will be able to know how it will affect my beam acceleration and beam dynamics. If this mode is trapped in a cell then will it play any role in acceleration? Beam is on Z Axis red in figure?
Unfortuantely, my knowledge of linacs is quite limited. I would think that it could, depending on the field configuration of the mode as compared to those you're using for acceleration.
I found out finally these type of modes are called trapped modes and you very rightly said that they are not influenced by the boundary conditions and the reason is that they have very weak field configuration near boundaries.
How to find if the mode falls in the category of trapped modes is with the K which is called the coupling factor and is defined as the measure of influence of boundary and is given as
K= 2|fm-fe|/fm+fe
where
fm is the frequency of mode with PMC Boundary
fe is the frequency of mode with PEC boundary
If value of K is very small it means the mode is trapped mode and it will not be coupled to the neighboring cell but it will affect the beam dynamics as well as Q of the cavity
Reference Tesla Report 2000-2008 Desy