Minutes of the ICE section
95th meeting on Wednesday 06/11/2013 (08:40-10:30, 6/2-004)
ICE members: Adrian Oeftiger (AO), Andrea Passarelli (AP), Benoit Salvant (BS), Carlo Zannini (CZ), Danilo Banfi (DB), Daria Astapovych (DA), Elias Metral (EM), Elena Wildner (EW), Frank Schmidt (FS), Giovanni Iadarola (GI), Giovanni Rumolo (GR), Javier Barranco (JB), Kevin Shing Bruce Li (KL), Nicolo Biancacci (NB), Nicolas Mounet (NM), Olav Ejner Berrig (OB), Serena Persichelli (SP), Sergio Rioja Fuentelsaz (SRF), Tatiana Pieloni (TP), Werner Herr (WH), Xavier Buffat (XB).
Present/Excused: AO, AP, BS, CZ, DB, DA, EM, EW, FS, GI, GR, JB, KL, NB, NM, OB, SP, SRF, TP, WH, XB, GuidoS, MichaelB, VincenzoF, EiriK, WolfgangH, SimonW, SashaV (FNAL), AlexanderH, HannesB, AndreaP, VittorioV (Naples), ElenaB.
1) Newcomers / visitors
- VittorioV for few days => Impedance studies and impedance workshop for 2014.
2) Comments on the minutes of the previous 94th meeting + Actions
- No comment.
3) General infos
- No particular comment from anyone.
- SL meeting:
- Follow-up about VIAs:
- The DG has issued on 11th September 2013 a memo stating that both VIA and fellows are 2 equivalent programs at CERN, and therefore a VIA should not be awarded a fellowship contract.
- An addendum to this memo was then issued on 9th October 2013 by the DG to specifically address the coming fellowship committee as 18 VIAs had already applied, before the memo was published. This addendum proposes transition measures only for this committee to allow the award of a fellowship contract under specific principles, mainly for applications in different departments/activities from the current VIA, and to avoid a continuation of tasks/activities in the same group.
- These rules are given as strong guidelines, but possible exceptions can be proposed to the DG by the demanding group/department for very specific cases.
- Compulsory membership of the French national social security and health insurance schemes (“frontaliers”). A working group has been set up with a view to determining the impact of this decision. A public announcement will be issued to explain the consequences and who is affected.
- Eric Montesinos is collecting request for special training needs in 2014 (e.g. C++ courses or special software courses) => Everybody sent me his requests?
- RLIUP workshop last week => Summary this Thursday 07/11 afternoon: Main Auditorium (500/1-001) from 14:00 – 17:00 (timetable).
- Impedance web meeting on bench measurements.
- WP2 Task Leader meeting => TP and EM => Decision to go to the metallic collimators as soon as possible (in the plan of GA and StefanoR).
- Possible additional collaboration (Frascati) of ~ 1 man year (for HL-LHC WP2).
- Talk from Vincenzao at the last LIS meeting on space charge simulations for the PSB.
- Talk form NM at the collimation WG on the Collimator impedance: highest contributors and possible gain by relaxing IR3
- LBOC talk yesterday on ecloud scrubbing in the LHC with doublet beams => HannesB.
- PS-LIU meeting on 19/11 to review the PS imped activities => SP and/or NB will give the talk.
4) cobra-HeadTail (Kevin Li): pdf
- Reminder: Ecloud => PyEcloud (Python).
- HeadTail => Cobra-HeadTail (C++ and Python). Cobra = Collective beam dynamics in ring accelerators.
- And Anaconda = Analysis of cobra-HeadTail numerical simulation data
- Why cobra-HeadTail? => Many versions in //, efforts scattered, etc.
- The HeadTail model
- Time evolution of a particle ensemble is generated via a Poisson bracket with the Hamiltonian.
- Several Hamiltonians are formally easily concatenated.
- Symmetrized split operator method immediately yields a 2nd order symplectic tracling algorithm (Leap-frog).
- Design
- Programming in C++.
- Scripting using Python.
- Libraries.
- Documentation for free...
=> Cobra-HeadTail design:
- Extendible: Write a new class respecting the interface conventions and export via boost.python – or – write it all directly as a Python module which naturally merges into cobra-HeadTail.
- Flexible: Write your own personal main function using all the modules provided by cobra-HeadTail and write it in Python!
- Manageable: The code is fully modular, dependencies are minimized. Each module author can be a module keeper of his own part of the code without breaking the global structure of the program or other independent modules.
- Translatable: Python modules can be interfaced from a vast amount of languages → Fortran, C/C++, Matlab.
- Benchmarks against the original HeadTail code
- Reminder: Same algorithm (so we should have the same results) but memory used differently, etc. => Many mistakes possible.
- Tracking in linear and nonlinear buckets, with chromaticity and amplitude detuning.
- Poisson solver => New FFT-type 2D integrated Green's function Possion solver - similar to Ji Qiang's BeamBeam3D or IMPACT codes.
- Same speed as original HeadTail.
- Very nice demo at the end.
- Conclusions
- Developed a modular and scriptable, Python interfaced version of HeadTail – written in C++ and interfaced via Boost.Python.
- Next steps
- Include
- Impedance model.
- MAD-X interface.
- Perhaps some form of beam-beam (Poisson solvers are there).
- Space charge (Poisson solvers are there).
- Open a HeadTail-developers-and-users (HEDEUS) group?
5) Actions to be taken for the next meeting
- Old actions.
6) Miscellaneous
- Deadlines and important dates for ICE.
- The next (96th) meeting will take place on 27/11/2013 => Agenda:
1) Brief reports of the 4 WGs (by the WG leader or replacement) => Plan, activities, deadlines, issues, etc.: BB, SC, EC, IMP.
- See preliminary agendas for the next meetings.
- See List of Actions.
Minutes by E. Metral, 28/11/2013.