Skip to content

Minutes of E-Beam Meeting #2

The meeting was held at CERN on 13/03/2019 - See indico

Participants (incomplete list)

A. Saa Hernandez, A. Latina, A. Pikin, A. Rossi, G. Arduini, L. Ponce, E. Metral, N. Biancacci, R. Scrivens, M. Zampetakis, L.V. Joergensen, H. Bartosik, A. Mereghetti, G.A. Tranquille

Actions

  • D. Gamba: Organise doodle to choose date/time compatible with most participant constraints.
  • A. Saa Hernandez and others from LEIR team: Further analyze LEIR data to understand if space-charge or IBS dominated emittance growth.

E-Beam: mandate and organizational matters (A. Rossi)

Adriana presented the scope and mandate of the working group - see presentation.

R. Scrivens asked if the meetings will always be on Wednesday afternoon.

not necessarily. D. Gamba will organize a doodle to check participant availabilities.

E-cooling measurements at LEIR (A. Saa Hernandez)

A. Saa Hernandez gave a comprehensive presentation of various e-cooler-related measurements at LEIR.

Motivation

  • 2018 used as reference for typical use of e-cooler in LEIR
  • Record extracted intensity obtained in 2018
    • but more relative losses between injection and extraction
  • Stressed that the final beam parameters are the result of equilibrium between heating and cooling effects
    • most studies carried out to understand what is the dominating mechanism.
  • Pointed out that it is not possible to measure directly the properties of the e- beam in the e-cooler. Only indirect measurements or based on early stage simulations/measurements.

E. Metral asked which e- beam parameters could be interesting to measure.

It would be good to know the beam position, current (presently measured), transverse profile, temperatures.

Experiment: Ion acceleration

  • Assumption that losses during beam capture at injection plateau are driven mainly by IBS.
  • From simulations, if one would increase the momentum by 40%, the horizontal emittance increase rate would also drop by about 40%.
  • Idea of using the e-cooler to accelerate the ion beam, and so perform capture at higher momentum.
  • Used "Brho_dot" for programming acceleration:
    • this propagates the change of momentum to all magnets
    • also needs to stabilize the tune during acceleration
  • managed to get 6% acceleration:
    • limited by duration of injection plateau
    • no losses during acceleration
    • possible to capture with RF at this higher energy
    • BUT similar loss pattern and emittance increase rate.
      • too small momentum variation.

E. Metral could we imagine to decelerate the beam to enhance the losses effect?

Probably possible, but not tried so far. To be checked.

Experiment: Cooling bunched beams

  • Idea of compensating heating effect during capture with e-cooler
    • Stressed that bunching is increasing longitudinal charge density, which might actually result in enhancing heating effects.
  • By changing the e- beam energy it was possible to generate parabolic, flat and hollow ion beam longitudinal distributions.
  • Measured losses between start RF and start Ramp for different e- beam energies:
    • Found a minima when cooler is set to higher equivalent frequency than the RF.
    • Behavior is similar if one uses two harmonic RF.
  • Measurements performed with low intensity beams compared to NOMINAL.
    • For higher intensities, the overall losses are not different than with standard operation which was proven to be more reliable and so preferable.

Measurement: equilibirium emittance

  • Measured the final emittances as a function of beam intensity injected from LINAC.
  • In theory, for IBS dominated machine one expect a I^n law (with n\approx 0.45), and this should be independent from machine tune
  • In reality, for some tunes and for high-enough intensities one can go in space-charge dominated regime, in which the final emittance is measured to follow still a I^n law but with n\approx1.45.
    • preliminary analytical calculation of tune spread by space charge show that for some tunes the beam might touch resonance lines.

E. Metral pointed out that for some tunes it looks like the behavior goes back to the IBS-dominated regime.

A detailed analysis of the data has still be done, but it might be correct.

Measurement: cooling maps

  • Idea is to measure equilibrium emittances and energy spread as a function of ion-beam orbit inside the e-cooler.
  • Useful as guideline for later cooling set-up.
  • Data collected for different e-beam profiles and currents still to be analyzed.
  • Preliminary observations:
    • Increasing emittance with abs(angle) as expected
    • It seems like we normally run with a 10 mm horizontal offset between e-beam and ion-beam center
    • Vertical emittance is independent from H-angle as expected
    • Vertical emittance increases as e-ion beams overlap decreases
    • There is a momentum dependence with e-ion beams overlap as expected due to e- beam space charge effects
    • Final ion beam momentum spread seems not very sensitive to e-ion beams overlap
    • Horizontal emittance independent of V- angle as expected
    • Vertical emittance not very sensitive to vertical bump, except for very large offsets
  • It seems difficult to reverse engineer the e- beam profile from cooling maps as important ion-beam losses occur when exploring edges of e-beam.

Insight: what can/could we measure from e- beam?

  • During last few days of operation an attempt to measure the e-beam position has been done by looking at the cooler BPMs signals at the moment when e-cooler voltages are short-circuited (i.e. end of cooling).

    • A signal of amplitude is clearly visible.
    • It has been verified by A. Frassier that the signal disappears if e^- gun filament is off.

  • Discussion:

Can such a signal from the BPMs be calibrated?

G. Tranquille: even if possible, the e-beam position during switching off might not be representative of position during steady state cooling.

A. Rossi: can the e^- be modulated in energy and/or intensity?

  • G. Tranquille: this has always been implemented via dedicated GFAs. In the past this was used regularly.
  • D. Gamba: we tried quickly during last few days of operation, but without success. Probably we didn't specify a big enough amplitude modulation.
  • G. Tranquille: note that the modulation on the grid voltage has been taken away to be installed in ELENA.

--

Would be possible to have a screen somewhere on the e- path to measure e- beam profile?

A. Pikin commented that there is normally no space for such a device and it would be destroyed by the e-beam. In the past, for other cooler/gun, such measurements have been done, but probably only for pulsed e-beam operation.

Would it be possible to use a wire-scanner-like device?

G. Tranquille: Probably possible, but design and integration might be difficult.

A. Rossi asked if one could measure something at the level of the collector

G. Tranquille replied that the only feasible measurement there (already available) is the e- beam current.

R. Scrivens: Would it be possible to use the ion-beam to probe the e-beam?

G. Tranquille commented that this was done in the past and measurements should be available in literature. A. Saa Hernandez on the other hand commented in her presentation that major losses occur when the ion beam approaches the edges of the e- beam, making such a measurement considerably error prone.

G. Arduini suggested to verify if it is consistent that the losses are driven by emittance blowup and therefore by the beam touching the geometric aperture.

H. Bartosik replied that in 2016 measurements were performed and they were consistent with vertical emittance/beam size touching geometric aperture.

G. Arduini: would one expect more blowup in the H plane via IBS? Is this what is seen in the machine?

H. Bartosik: yes.

G. Tranquille pointed out that equilibrium emittances strongly depend to vacuum levels, and they should be checked while doing those experiments

H. Bartosik replied that all measurements where performed in parallel to NOMINAL operation and no vacuum degradation of lifetime was observed.

H. Bartosik: in previous presentation it was possible to see simulated and measured profiles of the e-beam. How those were performed?

G. Tranquille: those were done in Novosibirsk with a wire on a test bench.

G. Arduini: would it be possible to use phase-space displacement and/or Barrier buckets acceleration?

H. Bartosik: this was investigated in the past, but also very slow process.

Simulation and code development plans for LS2 (A. Latina)

A. Latina presented the latest status of RF-Track

Introduction

  • Theoretically the cooling performance depends on several electron and ion parameters (transverse and longitudinal e- temperatures, e- distribution, etc.), but only the total e- current is normally measurable.
  • use of complex simulation codes (e.g. BETACOOL, RF-Track, others) is necessary to study the cooling effects.

Theory and RF-Track implementation

  • A summary about theory is available in earlier presentations of A. Latina - e.g. link
  • There are two crucial aspects to be considered:
    • single ion interacting with many e-
    • effect of e- temperature
      • normally derived with complex analytical expression
      • in RF-Track, numerically integrated via montecarlo integration.
  • RF-track implementation based on hybrid-kinetic model:
    • ion beam treated as particle distribution
    • e- plasma on a 3D mesh
      • easy to define different e- distribution
      • For LEIR simulations, A. Latina used distributions simulated/measured at the time of LEIR e-cooler construction.
  • RF-track has been benchmarked against
    • literature - for Xe54+ measurements and betacool simulations at ESR Ring (GSI)
      • had to fit on angular parameter between ion and electron
    • measurements of cooling force at LEIR
      • had to fit on e- beam temperatures and density.

Cooling force measurement at LEIR

  • Simulations and measurements showed some differences, even after fitting e- beam density and temperatures.
  • Measurements performed suddenly changing the requested e- velocity by acting on the e- gun voltage.
    • it was observed that the gun voltage response is not immediate, but takes approximately 100 ms to reach a steady state
    • impact on cooling force calculation
  • A first attempt to simulate the whole measurement process has been made
    • the gun voltage response was proven to have a significant impact on the cooling force measurement, therefore new detailed simulations are needed to better compare measurements and simulations.

On-going and new studies

  • Detailed analysis of collected data at LEIR.
  • Possibly, optimization of parameters/performance of AD/ELENA cooling.
  • Development of a software interface between RFTrack and pyHeadTail and BLonD for impedance calculation.
  • Review of IBS model in MAD-X and Sire.
  • Implementation of IBS model in RF-Track.
  • Implementation of e-lenses in RF-Track:
    • this might require considerable coding work.

A. Latina would need additional resources (e.g. Technical or PhD student(s)), especially for software development and simulations.

  • Discussion:

G. Arduini: could the LEIR e- beam temperatures and densities be fit using all data collected?

probably possible, but a careful and detailed data analysis has still to be done.

--

G. Arduini: would it be possible to measure the e- gun parameters, such as e- temperatures and distributions?

  • G. Tranquille: it is basically impossible to measure e- temperatures. From experience a T_{\parallel} = 0.001 eV seems reasonable, while T_{\perp} should probably be about 0.05 eV.
  • A. Pikin: the transverse distribution of e- beam depends on several parameters. Measurements were indeed done in the past on test bench, but they could also be simulated with dedicated simulation codes.

G. Arduini: will it be possible to measure the e- beam distribution for e-lenses?

A. Rossi: yes. It will therefore be possible to crosscheck simulations and measurements, and give more realistic estimate of e-cooler beams as well.

--

G. Arduini: Do we have a spare of the LEIR Gun that could be used for new measurements/studies?

R. Scrivens: we should have it, but to be checked.

As final remarks, G. Arduini suggested to make sure to use as much as possible the existing codes and conventions, not to invest too much effort on a code that then cannot be used by other codes and/or for other studies. He also suggested to find more synergies with A. Rossi for simulation and measurements.