Skip to content

Minutes of #11 Remote E-Beam Meeting

The meeting was held on Zoom on 12/05/2021 - See indico

Participants

Agnieszka Borucka, Jean Cenede, Roberto Corsini, Alexander Frassier, Bruno Galante, Davide Gamba, Lars Joergensen, Gerard Tranquille, Alexander Pikin, Laurette Ponce, Volodymyr Rodin, Adriana Rossi, James Storey


New AD e-cooler project status (Gerard Tranquille)

Gerard started with an overview of the three operational coolers at CERN: from the lowest energy one in ELENA (down to 55eV) to LEIR (2.4 keV) to AD (up to 27 keV in operation). The AD one is 40 years old, designed in 1978 for ICE experiments and is now time to replace it mainly du to reliability considerations. Presently, the strategy is to have a fully new cooler ready before LS3 (ideally, to be installed in 2024, i.e. one year before LS3, scheduled for beginning of 2025 to mid 2027), but have parts ready much earlier and compatible with the present e-cooler, e.g. the collector which was cause of major downtime in 2018.

The main changes will be:

  • Be able to reach electron energies up to 80 keV with the idea of using it to cool up to 500 MeV/c pbars (68.125 keV electrons).
    • The actual gain in performance and/or the actual pbar energy to use is still to be quantified and/or learned with experience.
  • Introduce the possibility of having an expansion factor 2 (2400 Gauss in the Gun, 600 Gauss in the drift) (now using 590 Gauss everywhere)
    • Possibility to vary the expansion factor during cooling. (Davide: Note that this is in principle possible already in ELENA, and could be tested!)
    • This will allow to further reduce the electron beam transverse temperatures
  • Improve the field quality by a factor 10:
    • Measurements from 1997 (PS/CA/Note 99-06 (local copy)) showed that the straightness of present B field is of the order of 2e-3 to 3e-3. (Also see measurements from 1984: CERN-EP-INT-84-01)
  • Install the cooler horizontally instead of vertically
    • Note that this will make pbar orbit perturbations in vertical instead of horizontal as now
    • Davide: TODO: Orbit compensation scheme to be checked!
  • Add NEG coating where possible and add NEG strips in the gun and collector sections to obtain 4e-12 torr vacuum (compared to present 2e-11)
    • Additionally, pumping ports will be added below the toroids.
  • Increase x4 the electron current (at 100 MeV/c plateau) using a new design (Alexander Pikin) gun with low transverse temperatures (~2 eV)

Whit those changes, there is the hope that we will be able to have faster cooling especially on the tails. (Davide: Actual gain on is all to be quantified. Complex interplay between many different effects).

Discussions are ongoing (mainly between EN-MME, TE-MSC, TE-VSC, SY-BI) on the mechanical design and integration of different parts:

  • Magnetic system:

    • After detailed analysis, it was decided to go for a pancake structure (like LEIR cooler) which allows to obtain a better field quality and avoids the need to have local compensating coils, as each coil can be aligned independently.
    • Return flux will be shielded with bars for the drift solenoid, while full shield will be deployed on the gun/collector to avoid perturbation to/from nearby ring magnets.
    • Toroids will bend the electron beam by 36 degrees (as present cooler)
    • The force between coils have been computed and found to be negligible, i.e. no complex mechanical structure needed to hold the coils.
    • Special care is being taken to optimize the assembly procedure, allowing all necessary radial space to install baking jackets and corrector coils.
      • Still, in case of need to replace a broken coil, one will have to disassemble most of the cooler. Discussions are ongoing to allow for position tracking of each single coil, which would allow to exchange one coil without the need to re-measure (and re-optimize) the magnetic field straightness.
  • Gun:

    • Design with flat cathode and transverse temperature of order of 2 eV.
      • Such a gun will provide, by design, a flat electron beam profile, slightly more intense at the periphery. This profile could be modified acting on control electrode voltage, but at the expense of transverse temperature of the electrons.
    • Care is being taken on HV design in order to avoid issues presently experienced in the new collector.
    • Investigation on using standard 100 keV feedthroughs t simplify the design.
    • Depending on final gun design, design of expansion solenoid will follow.
  • Collector:

    • Main idea to keep cooling water circuits outside the vacuum and use standard material/pieces.
    • Only one squeeze coil instead of two to simplify the design.
    • First version already built as urgent item for the present cooler.
      • Unfortunately, HV problems occurred which are being investigated.
      • Hope to make an operational version to be installed in AD during next year winter shutdown.
  • HV Platform:

    • Most parts have been already purchased and are being used at the cooler test stand.
    • Discussions ongoing on where to install the new platform, aiming to have it closer to the cooler instead upstairs near the AD control room.
      • Discussion to be followed up with Francois.
    • Contacts with EPC have started to discuss about power supplies
  • Beam Instrumentation:

    • Two BPMs will be installed in the drift section as in the present cooler.
    • Design still to be started by O. Marqversen and M. Wendt.
      • Critical item to continue in the design of e- beam steering coil design.

Electron beam generation/dump is being simulated with COMSOL by Gerard.

In conclusion, design is well advanced but still many items to be followed up.

Discussion

Francois asked if the frequency of NEG coating activation/lifetime was evaluated taking into account that the nearby sector might have a worst than the new cooler one. Gerard replied that VSC is confident the proposed design will allow to keep good vacuum with junction in the ring. There were no detail discussion about NEG re-activation, but in principle this should be possible to do with no need to dismount anything. Presently, the vacuum pressure in the nearby sector is already of the order of 10^-11/10^-12 without electron, and NEG is expected to simply allow to better preserve those level when e- are on. Adriana commented that efficiency of NEG will depend on gas composition. Davide asked if a better vacuum is needed mainly to preserve e-cooling efficiency or for pbar lifetime. Gerard replied vacuum has an impact on both. From experience with the IPM, vacuum levels of the order of 10^-9 are a serious threat for pbar lifetime.

Alexander asked if the collector will be at ground potential, and which power will be dumped on it. Gerard replied that no: it will be at a few kV with respect to the gun potential, and it will typically need to dissipate about 10 kW of power. Alexander commented that with those power level, vacuum will certainly be worst than 10^-12 in the collector region. He also asked if there are plans to add a gate valve between gun/collector and ring. Gerard replied that the idea of a gate valve was evaluated, but excluded by VSC group as it would be too expensive and difficult to operate/integrate.

Adriana asked if coils alignment will be done during commissioning or during construction. Gerard replied that in principle this will only be done once at construction, and that each single coil will be fiducialized/referenced to allow to maintain a proper alignment even in the case of replacement of a single coil without the need of new magnetic measurements.

Adriana asked if there will be coils to steer the electron beam. Gerard replied that there will be 2 steerer in the drift solenoid and one in each toroid.

Francois asked to obtain a new spending profile for this year. This will be followed up by Gerard.

Davide asked if one should keep some margin on the maximum magnetic field of the drift solenoid in case higher field will be necessary due to the possibly higher energy of the electrons. Gerard confirmed that this will be possible.

Davide commented that with an e-cooler installed horizontally, the circulating beam will experience a vertical kicks due to toroid fields. The configuration of the nearby orbit corrector should then be considered in any case, irrespectively of other ideas proposed by Pavel in the past (and which were discussed in the next presentation). Action needed!


On possible AD optics improvements (Davide Gamba)

Davide started giving an historical overview from the expected AD performance at the time of its design to the latest status of the machine in 2018. In 2018 the cycle length was about 110 seconds, with about 20 seconds taken by e-cooling for each 300 MeV/c and 100 MeV/c plateaus. Poor e-cooling performance, compared to expectations, have affected the AD cycle since the very first years of operation. According to the documentation found, a few attempts to improve the situation were made, but it is not always clear how effective those tests were. Among several observations, the following are the most recurrent/noticeable:

  • In several occasions it was mentioned that orbit corrector strength around the e-cooler is not enough for correcting fully the orbit perturbations induced on the circulating beam due to the magnetic system of the e-cooler.
  • Already in 2002 tests with a "better" optics with non-zero dispersion in the cooler was successfully tested, but no further attempts in using such an optics are documented (at least to Davide's knowledge).
  • The installation of compensating solenoids far from the e-cooler was imposted by the tight space constraint and by the available hardware at the AD conception time.

The proposals from Pavel, presented at the AD Electron Cooler Consolidation Review in 2019 (see indico), were meant to 'solve' those problems by:

  1. Minimize coupling by moving compensation solenoids next to the e-cooler.
  2. Turn away an horizontal orbit corrector from next to the e-cooler, i.e. to increase the lever arm and therefore allow for better pbar orbit control in the e-cooler.
  3. Try again to implement a new optics with lower beta-function and non-zero dispersion in the e-cooler, i.e. similar to what done in 2002.

All those ideas, based on the long experience of Pavel, are surely of interest. On the other hand their likely positive impact on AD performance should be analysed in simulations and priced with the possible cost of hardware modifications and/or MD time needed to test the new optics. For example, recent simulations with RF-Track and Betacool also showed that the gain on cooling time going to lower beta functions might not be so evident. Additionally, the present idea of installing the new AD e-cooler horizontally instead of vertically, should be taken into account when evaluating the impact on pbar orbit and its correction/optimization.

Davide also stressed that AD deceleration efficiency is mainly linked to the ability of stochastic and electron cooling to catch the tails of the beam. Back in 2018, at least for the last month of operation, data shows that most losses were linked to the first 2 deceleration steps, i.e. more likely to be linked to stochastic cooling performance rather than e-cooling. Additionally, with the introduction of ELENA, tails might become even more important because of the necessity of a loss-less transport down to ELENA's 35 MeV/c plateau. This might also change the actual requirements and mode of operation of AD itself.

The next a few months of operation will therefore be crucial to re-assess the performance and needs of the antimatter factory as a whole, and this will serve as input to better evaluate Pavel's proposal. In parallel, work still needs to be done to recover more knowledge about the present AD optics and the peculiar hardware limitations, to ensure that realistic ideas/optics can be evaluated/explored, as well as to refine tools and techniques to perform optics measurements in AD.

Davide concluded mentioning that an updated webpage and git repository of the AD optics is now available.

Discussion

Gerard suggested to compare the simulation of cooling as a function of beta function with data from LEIR (see CERN-PS-99-033-DI, and also LHC-PROJECT-Report-1111 and CERN-PS-2001-056-BD) where indeed for low beta function one sees poorer e-cooler performance.

Lars commented that the maximum intensity of pbars was limited in the past also by Radiation Protection constraints at the level of the target area. This limitation might have been partially overcome after LS2.

Gerard mentioned that after LS2 the Cryogenic Current Comparator (CCC) will be fully operational to measure beam intensity all along the cycle. In principle, the system is already operational, but another leak on the liquid helium circuit was found recently, and it is being investigated. Additionally, the Ionization Profile Monitor (IPM) should also be usable now in AD: the hardware was already upgraded in 2017, but there was no time for testing it in 2018. Alexander F. commented that from a control point of view everything is read, but there is still a problem on the MCP driver that, apparently, pulls too much current and breaks the front-end electronics: this should be looked at during this year run.

Adriana asked about a possible timeline when to perform the magnets reshuffling. Gerard replied that the e-cooler could go in the machine at the earliest one year before LS3 (expected to start as of Dec 2024). This would allow to gain operation experience before going into LS3. In this case, there would be not enough time to also change the magnets position in the nearby sections, therefore this will only be possible during LS3 at the earliest. After the meeting, Francois commented that one year is a reasonable notice for planning piping, supporting, cabling etc., i.e. we should aim for knowing what to do by end of 2023. On the other hand, budget wise such an intervention should be asked as an independent project which should be asked through an ECR as soon as possible, and no later than mid 2022.