Progress with FCC-ee solenoid modeling and testing the new exact MAD-X solenoid

→ Europe/Zurich
Zoom Meeting ID
61040794000
Host
Rogelio Tomas Garcia
Useful links
Join via phone
Zoom URL

LNO meeting on the on progress with FCC-ee solenoid modeling and testing the new exact MAD-X solenoid and a discussion of computing needs in the technical network (TN)

 

News, Rogelio : The HL-LHC Cost and Schedule Review / Machine Advisory meeting started yesterday and appeared to go well. It will finish this afternoon with the close-out session.

RIccardo mentions a recent MAD-X reported issue and proposed fix on ESAVE after using FILL_KNOB and plans to clarify and discuss this with Tobias.

 

Progress with FCC-ee solenoid modeling, Helmut, slides

The principles of solenoid modeling using slices based on detailed field maps have already been discussed in earlier LNO meetings and were presented at the FCC-week ’22. Helmut now uses the detailed 3D four quadrant field map that he obtained from Mike Koratsinos at the FCC-week. He observes that the left right symmetry is slightly broken and that the orbit bump in the passage through the solenoid described by the field map is not exactly closed. He proposes nested horizontal and vertical orbit correctors in the anti-solenoid (and quadrupoles for IP steering) and shows these can close the bump using small kicks . These ideas were also presented and discussed at the recent FCC-EIC Joint & MDI Workshop ’22. With the help of Michael Hofer this is now implemented in the fcc-ee lattice repository, including testing and sample jobs. A main objective for this first version has been simplicity to make this easily usable and flexible to extend. The bump is generated and closed using orbit correctors slices alone. Helmut says this should be sufficient to already provide quite realistic predictions of the SR power generated and where it will be lost, as well as studies of IR steering. Planned extensions (not yet in the repository, largely tested, more complex to use) are simulation of the overlap with quadrupoles and the addition of thin solenoid slices aligned with the actual orbit.

For this first implementation of nested orbit correctors in quadrupoles Helmut used simply thin orbit correctors at the start and end of the thick quadrupole elements and says this could be improved by upgrading MAKETHIN to generate correctors at each slice. Riccardo and Rogelio propose to follow up on this also in the more general context of allowing for overlapping elements. Rogelio also proposes to follow up more generally on aligning elements around the actual orbit.

Rogelio asks about higher order effects. Helmut believes they can be determined using the field map/tracking approach and would expect them to be small as seen from outside since the anti-solenoids are designed to compensate the solenoid including higher orders.

 

Testing the new exact solenoid in MAD-X, Leon, slides

Leon presented first tests of the new exact solenoid in MAD-X and comments on the use of different frames for the tilted solenoid. He finds that the new MAD-X implementation removes the unaccounted spurious dispersion such that now both orbit and dispersion agree well with SAD. The beta beating also agrees with SAD when removing the extra drift required in the previous solenoid implementation, which was needed to correct for path length changes for the tilted non-exact solenoid, which actually further confirms the better modeling now. Leon says it is also planned to implement the MAD-X solenoid in Xsuite, to follow up on overlapping fields, field expansion around the tilted paths, beam-beam and a coordinate system of both beams in the solenoid.

Helmut reminds that the exact solenoid does not provide information about the extend of fringe fields that is important for synchrotron radiation and proposes to check estimates of the synchrotron radiation effects using EMIT with the orbit-corrector slicing approach.

 

The meeting now turned to a discussion of the computing needs in the technical network : 

The BE-CSS team, in collaboration with the IT department, is defining the future of the CERN Technical Network (used for example in our control rooms).
They are presently collecting input from the users on desired features and requirements focusing on the network isolation (and not infrastructure). We discuss our input:

  • Easier access to Nxcals: This year there were lots of trouble to extract data within the TN possibly due to Kerberos token. Gianni suggests to open a ticket to address this unwanted feature likely in Nxcals.
  • CVMFS access or something similar from the TN. Goal is to have different python releases, maybe as a virtual environment, both in TN and general network, that is self managed by us. Gianni asks what is missing in acc-py environment? It is only accessible from TN, plus needs add-ons. Possibly versioning available. Dedicated meeting possible. 
  • Access to powerful multicore servers already in the general network.
  • EOS access (maybe selected projects) from TN firstly for storage but could be used to solve other software access issues.
  • Keeping afs access!
  • Communicate with us in case there is a possibility to break an existing feature.
     

Two LNO meetings are planned for next week and then meetings once per week, in one case joined with NDC, as circulated by email.

There are minutes attached to this event. Show them.