Minutes of the ICE section
80th meeting on Wednesday 24/04/2013 (08:40-10:30, 6/2-004)
ICE members: 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), Jean-Luc Nougaret (JLN), 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: BS, CZ, DB, DA, EM, EW, FS, GI, GR, JB, JLN, KL, NB, NM, OB, SP, SRF, TP, WH, XB, Letizia Ventura, Alexey Burov, Uwe Niedermayer, Michael Bodendorfer, Javier Barranco, Danilo Banfi.
1) Newcomers / visitors
- Passarelli Andrea (Technical Student, 01-Jun-13 to 31-May-14) with CZ.
- Bérengère Luthi (summer student) with BS and CZ.
2) Comments on the minutes of the previous 79th meeting + Actions
- Comment from NM about low-frequency impedance (see example data and picture):
- It seems that when you put a strong magnet (I chose the susceptibility chi_m=1000, then 10000, with the same result), at relatively low frequency there is a "plateau" in longitudinal impedance: some induced currents stayed confined in the beam pipe thickness because one finds a longitudinal impedance close to the ohmic resistance of the pipe (here one finds ~5e-3 Ohm, which corresponds to this 1m long stainless steel pipe of 10mm radius and 2 mm thickness - I checked with 20mm radius and the result also scales accordingly).
- But, if one continues to decrease frequency the plateau reaches an end and the longitudinal impedance decreases again to zero. I think that this zero longitudinal impedance at zero frequency is forced in any 2D model, whatever the outside structure, and this has nothing to do with any kind of "hidden bypass". There is a simple symmetry argument for this: at DC, the problem is translationally invariant along the longitudinal direction (again, in a 2D model). At DC we can decouple electric and magnetic fields and solve the electric fields equation individually and write E=-grad(V) with V the potential. Since V must be constant along z (longitudinal coordinate) from the translation invariance, then E_z must be zero. And the longitudinal impedance is proportional to E_z., hence the result.
- Now, this does not mean that the impedance of the real machine must be zero at DC, but simply that the so-called "resistive-wall" impedance (which is defined - I think - by a 2D model neglecting any geometric 3D feature) is zero at DC. Of course then one can imagine that the main part of the impedance of a machine at or close to DC is not the resistive-wall impedance, but is rather geometric, or even could be due to an interplay between resistive and geometric effects (the latest is something that Mauro and Nicolo are checking, if I understood correctly). And a 3D structure is not translationally invariant so the longitudinal impedance has no reason to be zero at DC. But, I maintain that the resistive-wall impedance at least as it is defined now (with a 2D model) is necessarily zero at DC.
3) General infos
- No particular comment from anybody.
- SL meeting:
- I did not attend the last one due to SC2013 workshop.
- The fellow requests should be sent to Delphine and Oliver for 23/04.
- There will be a new training on costing estimate related with WP from projects. This course has been discussed at the BE Workshop. The new training is being organised and EricM proposes to have a mix of newcomers and experienced staff. Proposals should be sent to Eric.
- Meeting of the task force mandated by DG to look at traffic.
- Staff Association:
- Audit report received about the health insurance. SA will then make official comments.
- A creche will be opened in September with 13 places. SA has a moderated Facebook account.
- Invenio (US firm): contacted CERN asking to remove this name from CERN-developed software and to destroy all files.
- Electronic version of the OHS form is now available.
- News from LHC:
- First interconnect about 2 weeks ago and first splice measurements => Start-up phase of this activity and therefore don’t know yet if delays.
- Nice SC workshop last week. This week several experts are around for those who have some questions.
- CZ's defense went very well => Congratulations!
- LIU day on FR 12/04.
- ATS seminar with BB workshop summary by TP.
- WG on Advanced Collimator Materials - Kickoff Meeting => NM will be the contact person for the impedance.
- Nice talk from UweN yesterday about GSI and CERN activities.
- TCTP transverse mode => Hot topic these days => Follow-up.
- Comment from BS about CST licenses: back to 15 licences.
4) Long range beam-beam contribution to stability diagrams (XB): pdf
- Answer to the Action 8 of the List of Actions for 2013 => Clarify the situation with the stability diagram in the presence of both octupoles and LR beam-beam.
- The stability diagrams can be computed theoretically in some cases. In the general case, a numerical evaluation of the dispersion integral is needed => A code has been written for this, called PySSD = Python Solver for Stability Diagrams. The tracking is made with MAD-X and we can have therefore the full LHC complexity => Possible issues: large computing time and noise.
- An analytical model has been used last year by AlexeyB for the LR BB tune spread and a detailed comparison has been made between this model and the numerical code.
- The analytical model from AlexeyB should be valid under the assumptions below
- Linear detuning (r >> 1),
- Constant separation,
- No // separation,
- IP1&5 only, with aletrnating Xing angle.
- Comparing the 2 models first in the case without // separation already leads to quite a large difference, due to the fact that the linear detuning is over-estimated and LR is not an octupole (higher oder multipoles have an impact on the stability diagrams).
- A comparison has then been made with // separation => // separation does also play an important role. It was also found/confirmed that the effect of IP2&8 is indeed marginal in the current LHC configuration.
- In summary, the estimation of the LRBB tune spread in AlexeyB's recent paper in term of "octupole equivalent" is a factor ~ 550 / 200 = 2.75 too large: instead of ~ an equivalent octupole current of ~ 550 A it should be ~ 200 A. All the consequences of this modification on beam stability should be reviewed.
- Question from XB about a curve from AB which seems to be the opposite. Why? => AlexeyB to check.
5) Future accelerator neutrino activities: EUROnu conclusion, and then? (EW): ppt
- What is the world planning / thinking about?
- Japan => Conventional beams.
- US => Conventional beams, Neutrino Factory.
- China? => A medium-baseline neutrino facility based on 15-MW CW superconducting linac.
- Europe => Conventional beams, Super Beam, Beta Beam, Neutrino Factory.
- CERN LBNO (LAGUNA) 2300 km => Will see enssentially Mass Hierarchy.
- CERN LBNO & short baseline
- NA could host LBL and SBL neutrino beams.
- With upgrades LBNO aims at 2 MW, SBL is a few hundred kW.
- Outcome of EUROnu:
Several "conventional" beams are planned, not too expensive, not unrealistic
1) But for better results for neutrino research one needs “next generation”, in general expensive, facilities studied in EUROnu.
2) NF is the “facility of choice”, very expensive.
3) The Super Beam is also good, less expensive.
4) The beta Beam less good and expensive.
5) Build a test facility: nuSTORM!
- Reminder: In all these projects what is important is intensity and not brightness, which means that emittances are not important.
- nuSTORM
- A revived, now mature, option (started in FNAL) => Workshop at CERN on 26-27/03/2013 (EW and K. Long).
- We inject pions which decay in muons (very rapidly) and then muons which decay on a much longer time into neurtinos.
- Motivation at CERN:
1) Testbed for a future next generation neutrino facility.
2) X-sections.
3) New neutrinos.
=> 100 GeV at CERN (SBL), whereas 60 GeV considered at FNAL.
- nuSTORM physics reach at CERN:
- 4.5E19 POT / year may reasonably be expected.
- 5 years with 100 GeV p, 5 × 4.5E19 = 2.3E20 POT.
- FNAL MI at 60 GeV over five years is 1E21 POT. With linearly (energy) increased yield 60->100 GeV, we still lack some factor 2…
- Timeline
- EOI sent to SPSC in June. Aim: get a mandate and manpower. In 2 years, TDI and LOI.
- New project: EUROSB at ESS:
- In contrast to EUROnu Super Beam, the second oscillation maximum is used => Longer baseline.
- 2.5 GeV at 5 MW => Lower energy => Higher SC effects etc.
6) LTEX report: Status of impedance studies of the LHC forward detectors planned to be upgraded during LS1 (in particular ALFA and TOTEM) (BS)
- BS made a reahearsal of the talk he will give this afternoon at the LMC.
7) Actions to be taken for the next meeting
- Old actions.
8) Miscellaneous
- Deadlines and important dates for ICE.
- The next (81th) meeting will take place on 15/05/2013 => Agenda:
1) Discussion about the review being organised for ~ mid September on the LHC performance limitations during run I (EM).
- See preliminary agendas for the next meetings.
- See List of Actions.
Minutes by E. Metral, 14/05/2013.
