Minutes of the ICE section
43rd meeting on Wednesday 26/10/2011 (08:40-10:30, 6/2-004)
ICE members: Benoit Salvant (BS), Christian Hansen (CH), Carlo Zannini (CZ), Hugo Alistair Day (HD), Elena Benedetto (EB), Elias Metral (EM), Elena Wildner (EW), Frank Schmidt (FS), Giovanni Iadarola (GI), Giovanni Rumolo (GR), Jean-Luc Nougaret (JLN), Kevin Shing Bruce Li (KL), Nicolo Biancacci (NB), Nicolas Mounet (NM), Olav Ejner Berrig (OB), Tatiana Pieloni (TP), Werner Herr (WH), Xavier Buffat (XB).
Present/Excused: BS, CH, CZ, HD, EB, EM, EW, FS, GI, GR, JLN, KL, NB, NM, OB, TP, WH, XB, Alexey Burov, Elena Shaposhnikova, Georges Trad, Hannes Bartosik, RamaCalaga, Theodoros Argyropoulos.
1) Newcomers / visitors
- None.
2) Comments on the minutes of the previous 42nd meeting + Actions
- None.
3) General infos
- SL meeting:
- New associated (or requestors) CERN members: Israel, Serbia, Slovenia, Turkey, Cyprus, India, China, Ukraine.
- The 2011 ABP Christmas Party will be organized by the ICE section. Current proposed date (as I sent you by email): TH 15/12/2011 (9/12 as a backup) => OK?
- News on the LHC:
- Neon injection in LHCb was endorsed at the LMC.
- MD planning: 25 ns got strong support: 12 h from physics and 12 from MD time => Took place on Monday => EM to report this afternoon at the LMC as GA is in vacation.
- TDI vacuum pressure disappeared with the large gap (+- 55 mm) BUT the heating remains! => Slides from VincentB. TDI MD during the next 4th LHC MD block (to be followed up by BS, GiuliaP and other volunteers).
- Action for ABP to be followed up (and explained): abort gap re-population.
- Octupoles in the HEADTAIL code => NM found an error by a factor 2 (the octupoles were too powerful in HEADTAIL) and therefore the stability diagrams from theory and HEADTAIL simulations are finally in very good agreement => See work from Raymond Wasef, where a factor 2 was reported: the theoretical (Gaussian) stability diagram for - 10 A was reproduced with HEADTAIL with only - 5 A. With this error by a factor 2 found, - 10 A is now also found as predicted by theory. The next step for this work would then be to introduce other distributions in HEADTAIL (quasi-parabolic, 6-sigmas, etc.) and compare to theory to check the huge changes in beam stability.
- Concerning the comparison between measurements of the single-bunch instability and HEADTAIL simulations (IPAC11 paper): It was found from measurements that he beam was stabilized for an octupole current between - 20 A and - 10 A (with some uncertainty on the transverse emittance), whereas HEADTAIL simulations predicted a beam stability for ~ - 10 A. With this bug fixed, it means that HEADTAIL simulations predicts a beam stability for ~ - 20 A (for the nominal transverse emittance considered). Therefore, the agreement between the 2 remains the same...
- Discussions ongoing with impedance team (Fritz etc.) to have an impedance room/lab where we could store some materials (from FritzC et al.) and prepare the measurements to be done in clean rooms where the time is limited.
- There should be a new working group to look at the transverse emittance preservation etc. along the injector chain.
- Discussion with RalphS about the BBC = BBLR compensator => We will have to estimate the impedance of this equipment.
- Following the LHC instability observed last Monday 17/10/2011 => Questions from GianluigiA and answers by NM:
- What is the required octupoles' strength for the 50 ns beam with ~ 1.5E11 p/b within ~ 2 microm, Qprime ~ 2, and at 7 TeV (collimators settings taken in sigmas at 3.5TeV and converted to mm with the 7 TeV nominal emittance)? Answer from NM shows that we need 770 A in H and 670 A in V to reach stability (to stabilize the TCBI mode 1, assuming the mode 0 can still be stabilized by the ADT), PictureH and PictureV, whereas the maximum is 550 A => Should not be possible! Furthermore, in this case the TMCI would also become a problem.
- Same question for the nominal collimators settings (i.e. a bit less tight): in this case, we have more margin and the beam is stable at 360 A in V and 400 A in H (and well inside the diagram for 500 A), PictureHnom and PictureVnom .
- Plot of the longitudinal impedance (of our current impedance model) sent by NM:
- Injection: Contributors for Re and contributors for Im.
- 3.5 TeV: Contributors for Re and contributors for Im.
- Longitudinal profile deduced from luminous region => JoachimT's note. The real distribution is not necessary Gaussian!
- Montague resonance: follow-up by Ji Qiang: MontagueSC2.
- Transverse Gaussian distribution found by Helmut Burkhardt => To be followed with him (and with the Massi).
- Concerning the beam instability observed last Monday 17/10/11 (see Slides of last ICE meeting + BeamIntensitiesDuringTheInstability): After discussion with StephaneF, we agree that we need to continue our investigations and try and explain why for instance this instability is not observed already at the end of the ramp where the impedance should be more or less the same (to be confirmed quantitatively) and (linear) chromaticity the same (is it really the case?). We said that we see at least these 3 possible explanations:
- 1) The complex tune shifts are the same (as the impedances are the same) but the transverse distribution is more critical at the end of the squeeze, maybe due to beam-beam or other effects => To be discussed also with WH and beam-beam team.
- 2) Detrimental effect from the beam-beam long range (in particular also to the transverse distribution, cutting the tails and then ending up with a distribution close to the one we used for our estimate?). In this case we should see first some (incoherent) losses. Was this the case? It seems it was the case during the first test with beta*= 1 m... To be followed up.
- 3) Effect of Q'' which is much bigger at the end of the squeeze => For this Stephane could provide us with some estimates and we could try and run the HEADTAIL code with these Q''.
- After discussion with MikeL and WalterV we could also try and check the evolution of the chromaticities from the end of the ramp till the end of the squeeze.
- Info from StephaneF => First order of magnitude for the Q" induced by the triplet and the octupoles.
- 1) Contribution of two low-beta insertions with beta*=1 m. Q" depends on the betatron phase advances between IR1 and IR5, therefore is not the same in the H and V planes. It is also not the same for beam1 and beam2 and vary with energy due to the systematic quadrupole error of main dipoles (which is energy dependant). A range of Q" can nevertheless be given: -1000 < Q" < +5000. The ref is LPR308 using Fig. 1 (Fig.1 was produced with 2 IRs at beta*=25 cm, and knowing that Q" goes with 1/beta*^2, it is easy to rescale it).
- 2) Contribution of the Landau octupoles (LPR501, Tab. 20) The rules are: In H plane in units of 1000: Q" = 1.55*KOF-0.30*KOD In V plane in units of 1000: Q" =0.07*KOF-0.39*KOD. Therefore neglecting the cross-talk and taking a K of +/- 9 as I found in LSA at the end of ramp we get Q"_x =+/- 14000 and Q"_y=+/-3500. I conclude from above that the contributions of the Landau octupoles is still dominant in the H plane at beta*=1m, but Q" can easily be doubled due to the IT contribution in the V plane. However even if so, the Q"_y will remain below the Q"_x at the end of ramp.
- PS LIU meeting => Talk by RolandG on the production of the same luminosity as now with the 25 ns beam => Production of very small transverse emittances and preservation until LHC collisions => Should be certainly followed-up by the new working group.
- High-pile MD yesterday => 11 bunches of ~ 2.5E11 p/b within transverse emittance of 2.5 microm in collision => Means also that a single bunch of ~ 2.5E11 p/b is stable until collision, which is a good news (as predicted...).
- 2 impedance meetings: internal + webmeeting.
- What is the status of the beam-beam tune spread etc. to be available at the CCC for each fill? => Will be available for next year (XB): 2 versions discussed:
- 1 tune footprint computed (with tracking) before and then just scaled and moved depending on the beam parameters (tunes, emittances etc.).
- 1 version in the on-line model.
4) Can circular optics be useful for upgrades of the LHC complex (Alexey Burov) - Part II: pdf
- AlexeyB summarized first the results of last week.
- Planar-circular mode transformation => Showing what the beam adapter is doing.
- Red arrows show the direction of velocities.
- There is no big difference between beam and light.
- Perturbation (not necessary symplectic) theory:
- Can be solved in the same way as quantum mechanics.
=> Tune shift is found to be proportional to the diagonal elements of the perturbation matrix.
- Rate-sum theorem => Sum of the 2 growth rates is independent on the eigenvectors and depends only on the Trace of the perturbation matrix.
- Coherent motion: If the tune shift is smaller than the difference between the coupled tunes then a transformation rule is given to be able to apply the uncoupled results.
- 2 possible applications for the LHC upgrade?:
1) Space charge suppression => The important transverse emittance in this case is not the smallest emittance but the largest!
- Could be maybe important when injecting from the linac to the PSB.
- Need a transverse emittance (much) bigger than the other one to be something useful. For equal emittances, there is no gain.
2) For LHC IP: When IP optics is circular, the canonical angular momentum is preserved in the beam-beam interaction, thus higher beam-beam parameters are available => Some resonances will not be excited.
- Example of VEPP-2000 e+e- ring (BINP, Novosibirsk, designed from the very beginning with circular optics) => HO DeltaQbb ~ 0.08 per IP and there are 2 IPS => Total HO DeltaQbb ~ 0.16. This is the highest tune spread achieved. Note: this machine is operating in the HO regime, there are no LR effects. Comment from RamaC: KEK-B achieved ~ 0.1 with 1 IP.
- Reminder: the HO limit in the LHC seems to be very high for the moment, meaning that this scheme is maybe not so important for the moment but could be in the future...
=> To be discussed in more detail in the beam-beam working group and the Linac4-PSB and PS people fighting space charge.
5) Anomalous Diffusion for Phase Space Density Modification (Alexey Burov): pdf
- Modulation of the RF phase near the synchrotron frequency (in a single RF system).
- Initial conditions:
- Hamiltonian = 0.6 => Close trajectories remain close.
- Hamiltonian = 0.5 or 0.48 => Trajectories diverge => Before a certain Hamiltonian value there is a region of chaotic region.
- This is a technique (without any noise!) to flatten the distribution in the centre, preserving the tails (as we don't use noise!).
- No coherent motion is excited (if we switch it off correctly, i.e. adiabatically).
- Reference to the paper from H. Huang et al., PRE 48, 4678 (1993) => 3 regions: outer (same as without shaking) + 2 inner circles.
- This anomalous diffusion appears because particles are crossing the separatrix (but there is no noise injected!).
- For this method, a good precision for the detuning is needed. In practice, due to beam loading etc. it might face some difficulties => To be checked in detail with RF colleagues.
- Possible applications: Loss of longitudinal Landau damping, ecloud?.
- Next step: Plan an MD for next year discussing with ElenaS and GR.
6) Actions to be taken for the next meeting
- Old actions.
7) Miscellaneous
- The next (44th) meeting will take place on 23/11/2011 => Agenda:
1) 2011 ICE goals: where do we stand? (all the ICE members).
- See preliminary agendas for the next meetings.
Minutes by E. Metral, 26/10/2011.
