-----------------------------------------------------------------------
Minutes of the ABP-RLC section meeting of 18.02.05
present: EB, AG, WH, EM, TP, FR, EV, FZ
-----------------------------------------------------------------------
(1) Minutes of last meeting and pending actions
------------------------------------------------
=> ACTION => compare the size of the longitudinal geometric wake with
RW longitudinal wake from A. Koschik (FR)
STATUS: PENDING.
=> ACTION => EM and BZ will continue the development for waves with beta << 1.
STATUS: EM is developing solution for beta_beam not equal to beta_wave
and will give a presentation on this calculation in two weeks.
The next step will be considering several rings of charge.
=> ACTION => confirm bunch length, intensity, and collimator gaps during
tune-shift measurement (FZ, EM).
STATUS: Discussions and information exchange with G. Arduini, M. Glasior,
and S. Redaelli. EM and FZ obtained raw data of intensity and
bunch length from GA, as well as a first notebook for data analysis.
Further, M. Gasior provided EM with data of the tune spectra.
EM also asked S. Redaelli for the actual gap sizes. Remarkably
nobody else had requested these numbers before.
(2) Coherent Beam-Beam Effects
------------------------------
A) simulation techniques for multi-bunch tune spectra (WH)
-----------------------------------------------------------
WH described the main features of the COMBI program for the simulation of
2-beam multi-bunch interactions. This code can model a variety of phenomena,
like short-range and long-range beam-beam interaction, strong-strong
as well as weak-strong beam-beam effects, and impedance. The user has a
choice between different field computations. WH showed example parts of
input files, where filling patterns, collision schemes, and actions can
be specified. Symmetry or antisymmetry of the interaction points have
profound effects on the simulated beam spectra.
A recent LHC Project Note (no. 356) contains a summary of simulation results
which WH produced for the BDI group. Substantial work is planned for the
coming months. Collaborations with several outside partners (M. Furman,
J. Qiang, F. Jones) have been established.
B) simulation results for rigid and soft-Gaussian bunches (TP)
---------------------------------------------------------------
Simulations were done with rigid-Gaussian and soft-Gaussian models.
A single bunch is kicked and the spectrum computed either for this
bunch or for several (kicked and non-kicked) bunches. The averaging
over several bunches introduces an additional peak in the tune
spectrum, even with only two IPs, no long-range collisions, and
a rigid-Gaussian model. The peak amplitude depends on the kick amplitude.
The separate peak disappears when the beam-beam force is linearized.
Conclusion of this first part of TP's study is that one should kick
and observe a single bunch, avoiding averaging.
The soft-Gaussian model includes aspects like Landau damping, emittance
effects and higher-order modes. The incoherent frequency spectrum is
clearly visible with this model, and consistent with expectation.
Reducing the intensity of 1 beam moves the coherent pi mode into the
continuum (incoherent) part of the spectrum, ensuring Landau damping.
The soft-Gaussian simulations typically consider 10000 macroparticles
per bunch. The simulations are run on LHC@HOME, which is administered
by members of CMS.
In parallel an analytical matrix-based approach is pursued.
Future work includes an extension of this analytical calculation
to account also for the long-range collision, an extension of the soft-
Gaussian simulations, an implementation of the HFMM field calculation,
and parallelization (MPI).
(3) New results of e-cloud simulations for nominal LHC, ultimate LHC,
and two upgrade scenarios (FZ)
----------------------------------------------------------------------
Results of new ECLOUD simulations for the nominal LHC, the ultimate,
and two upgrade scenarios were presented. FZ showed simulated heat load,
central electron density, and electron flux at the wall, averaged over an
LHC arc cell, as a function of the maximum secondary emission yield and
the low-energy electron reflectivity. For the nominal LHC a delta_max
of 1.3-1.5 appears sufficient, and a similar number holds for the ultimate.
The upgrade with 12.5-ns spacing requires a delta_max of 1.1 or lower,
while for the 75-ns spacing upgrade no strong electron-cloud effects are
predicted even for a delta_max of 2.
One reason why these simulation results are more favorable than previous
ones is thought to be a recent correction to the dependence of the
energy at which the secondary yield is maximum, epsilon_max, on the
impact angle. The erroneous formula was discovered by G. Bellodi (RAL).
FZ mentioned that V. Baglin has asked for the updated prediction of
the heat load, which he plans to present at a CARE Meeting on 3/4 March.
=> ACTION => crosscheck the new results for the nominal LHC (FZ and DS)
(4) Damping of injection oscillations at the PS (EV)
----------------------------------------------------
EV discussed simulations of the CERN CPS transverse damper performance
for beam injection. This study was inspired by A. Blas. First the general
feedback theory was presented. Then EV elaborated the specific problems
of the CPS, e.g., that the new CPS feedback amplifier cannot damp the
lowest betatron sideband. The questions asked by A. Blas are:
Is the frequency response sufficient? Can we operate without damping the
lowest betatron line? EV pointed out that the damper samples the
centroid beam position along the bunch about 13 times (73 MHz sample
frequency). The pick up signal scales with intensity, which has
the (undesirable) consequence of damping the center of the bunch
more than the head and the tails. Taking synchrotron oscillations
into account, this may have the effect that the transverse damper
and the incoherent collective beam dynamics interact.
Simulations are needed to fully understand this interplay.
(5) Follow-up of space-charge decoherence (EM, EB)
--------------------------------------------------
EM discussed decoherence due to chromaticity, nonlinear detuning with
amplitude, and space charge. For the former two effects, a complete
analytical formula exists, e.g., one by Minty, Chao and Spence.
Not only the centroid oscillation, but also the analytical emittance
growth is in perfect agreement with the HEADTAIL simulations.
For the space charge decoherence, he referred to a paper by L. Vos.
Experimental CPS data are available from A. Blas and M. Glasior.
Marek's data show a 1.5 ms decoherence time, Alfred's 2 ms with
feedback off and much less than 1 ms with feedback on. All these
data were taken at 1.4 GeV.
The beam pararameters obtained from A. Blas correspond to a s.c.
tune shift of 0.35 vertically and 0.30 horizontally, consistent with
HEADTAIL. EM did not think it was possible to obtain such beam
in the CPS, and he suspected that the emittance was underestimated.
The inverse Fourier transform of the frequency distribution is
the Green function (the response to a kick excitation). For the
effect of the chromaticity, the calculated Green function agrees
with the analytical formula for the centroid response.
L. Vos applied the same calculation to the space charge effect.
This may not be fully justified as the space charge force moves
with the beam. Nevertheless, following this scheme, EM could
reproduce curves by L. Vos. The decoherence times obtained
for different distributions are always much shorter than the
observed ones.
EM also estimated the head-tail damping from broadband impedance
and resistive wall. It is about 5 ms for the m=0 mode.
EM concluded that the HEADTAIL code shows a beneficial effect of
space charge, but the decoherence time is larger than measured.
One reason could be the assumed tune shift which may be almost a
factor 2 too high. An analytical model of decoherence with
space charge will be worked on by EM, EB and FZ.
FZ commented that the HEADTAIL code considers only the
linear space-charge force. Including the nonlinear components
may also improve the agreement of simulations and measurements.
EB presented an update on the HEADTAIL simulations.
Extending the simulations to a longer time does show the
expected recoherence. The latter is not perfect, which could
be due to the sinusoidal shape of the rf voltage.
HEADTAIL computes a horizontal space charge tune shift of 0.24
with dispersion taken into account, slightly different from EM's
number.
EB's near-term plan is to scan the bunch intensity or the
emittance (recommended by EM) to see at which value the measured
decoherence can be reproduced. A second plan is to include the
nonlinear force.
Posted on the web: Slides by WH, TP, FZ, EV, EM, EB