1. WP10.5: HOM Distribution
Task 2 – Presentation 2

2. Cascaded 3cell benchmark Dipole 2nd band example
4.429
4.767
Band 2
5.3599
5.471
Here are the quarter cell calculations with E and H symmetry planes with Nb 20°C are in good agreement with the full HFSS simulation
Driven modal (S matrix calculations) are in good agreement with eigen mode calculations

3. The 3.9GHz cavity couplers
Here are the HOM couplers derived from the SAT files that I was given.
HOM2
HOM1

4. HOM1 coupler S21 Geometry issues….
The design of HOM coupler 1 from the SAT file contains an error……. The beam pipe is 18mm – it should be 20mm. For initial cascading purposes I have edited the drawing to increase the pipe radius to 20mm.
Coupler dimensions still need to be verified….

5. Other coupler issues Other design issues
I do not have any detailed schematics showing the exact orientation of the upstream and down stream couplers with respect to each other.
I do not have any schematics on the orientation of the couplers between cavities (the cavities are rotated with respect to each other for each additional alternative cavity)
I do not have any drawings of the power couplers
The HOM import SAT files that I have been sent, may not be the correct designs that were actually installed in the cryomodule….. There are five different designs for each coupler…… this needs verifying
Other design issues
The simulations made by Dr N.Solyak of a two cavity string do not exhibit the band structure that is experimentally seen – is this a result of scattering from the bellows (are the bellow shielded?)
As a point of interest the warm cavity data taken by Dr N.Solyak resembles the cavity string data taken as part of the collaborative participation last year

8. Cascaded coupler benchmark Dipole 2nd band example HOM2, Left Endcell and Mid cell – meshing problems in HFSS……..
Cascade calculation of HOM1 Left endcell with a Middle cell for the 3.9GHz structure i.e. 2 section cascade calculation.
Finite conductivity of room temperature NB used on conductive walls, Cu used on interior of coupler
Discrepancies between cascaded and full simulation are yet to be resolved…… meshing issues
Dominant coaxial mode from HOM1 coupler into regular cavity modes

9. Possible causes of the HFSS/Cascading issues – reasons for discrepancies within full 3D simulations
Modes with distinct polarisations flip as a function of mesh for 3D calculations – this can cause problems with the cascading calculation (i.e. wrong polarisations cascaded together). This issue could be resolved by defining a distinct electric polarisation direction within the unit cell calculations
Meshing issues exist within HFSS, i.e. tests with 3D calculations (no symmetry planes) should be independent of axes rotation – rotating the geometry often produces different results. This suggests geometric drawing issues within HFSS.
As a result of the previous point the results of cascading tend to be more accurate – tests using various symmetry planes against full 3D benchmark simulations tend to confirm this.
These geometric meshing issues need further investigating and testing to resolve them; i.e. meshing methods, order of mesh, surface approximations and ways of drawing the structure etc….

10. 9cell Cascaded coupler Dipole 1st and 2nd band using HOM 1 and HOM2 couplers at 90 degrees – Nb 20°C
Cascade calculation of full 9cell 3.9GHz cavity with HOM couplers 1 and 2 i.e. an 8 section cascade calculation.
As the angle between the couplers is unknown (and whether or not these are the couplers that are in the cryomodule) this simulation is but a demonstration of the method and as such the couplers were set at 90°
Finite conductivity of room temperature NB used on conductive walls, Cu used on interior of coupler
24 resonances identified from cascade simulation across two dipole bands
Somewhat resembles Dr N.Solyak’s calculation

11. S21 warm 3.9GHz comparison between cascade calculation and cavity 1 data (taken across the single cavity) – Nb 20°C
Full Cavity simulations for S21 for cascade calculations for first mode
Comparison between cascaded and experimental data shows similarities but is not in complete agreement
Cause of discrepancies could be due to:
a) incorrect coupler angle in simulation
b) incorrect couplers in simulation
c) meshing issues within HFSS
Issues yet to be resolved and remain the subject of current work.

12. WP10.5: HOM Distribution task 2 work package, Future work
Resolve meshing issues in HFSS for the full structure simulations
Find better more robust peak fitting method
Continue dipole circuit model for this structure
Sensitivity studies of the Dipole and HOM’s in this structure
Full analysis of both the “warm un-tuned” and “cold tuned” DESY data and comparing it to the theoretical results – including phase calculations by considering a complex lorentzian fits to the data
Wakefield analysis