From: "Frank Zimmermann"
To: ""Frank Zimmermann"" ; ; ""Francesco Ruggiero"" ; ""Daniel Schulte"" ; ""Elias Metral"" ; ""Frank Schmidt"" ; ""Gilbert Guignard"" ; ""Jacques Gareyte"" ; ""Werner Herr"" ; ""Luc Vos"" ; ""Alex Koschik"" ; ""Bruno Muratori"" ; ""Tommaso D'Amico"" ; ""Walter Wittmer"" ; ""Gregory Penn"" ; ""Lifshitz Ronen"" ; ""Maxim Korostelev"" ;
Cc: ""Jean-Pierre Riunaud"" ; ""Karlheinz Schindl"" ; ""Louis Rinolfi"" ; ""Michel Martini"" ; ""Oliver Bruning"" ; ""Roberto Cappi"" ; ""Charles Hill"" ; ""Gianluigi Arduini"" ; ""Helmut Burkhardt"" ; ""Daniel Brandt"" ; ; ""Ralph Wolfgang Assmann""
Subject: Minutes of LHC Collective Effects meeting 25/04/2003
Date: Saturday, April 26, 2003 12:48 AM
-----------------------------------------------------------------------------------------------------------------------
minutes ABP-LCE meeting 25.04.03
present: EB, WH, JJ, AK, EM, DS, FR, FZ
-----------------------------------------------------------------------------------------------------------------------
(1) Follow Up on Pending Actions
-----------------------------------------------------------------------------------------------------------------------
Instability Simulations:
AK confirmed again that he reproduces G. Dugan's result, but that the
observed effect is not a soliton. Studies are continuing.
Collimator Impedance:
LV has added the different harmonics contributing to the collimator impedance.
His results show a few percent reduction of the effective impedance by isolating
the collimators.
ACTION: LV will report his conclusions in one of the next LCE meetings.
EM and DS discussed with Erk Jensen and Igor Syrachev about the numerical
computation of collimator impedances. MAFIA is not a tool to use. GdFIDL will
be installed at CERN in June. EM has got an account for HFSS. His other
obligations in the next weeks (PS operation) are likely to slow down progress.
FR points out that we cannot wait for months to have an answer on the collimator
impedance, and he encourages to spend free minutes on this topic.
-------------------------------------------------------------------------------------------------
(2) Crossing Angle in IP 2 (WH)
-------------------------------------------------------------------------------------------------
WH described the layout of IP2 and in particular the various definitions of
crossing angle used by the beam-beam team, optics team, Alice, and others.
Alice accommodates a vertical spectrometer (plus three compensators) which
by default introduce a crossing agle of +/- 70 microrad. There is a desire by
the experiment to frequently alter the polarity. For proton operation with 25 ns
spacing, the separation at the long-range collision points is insufficient, and,
in addition, the spectrometer bump only extends over a rather short section.
Therefore long bumps are superimposed, far from the outside, to further separate
the beams, and to vary the crossing angle at the collision point. The outside
bump can either add to or counteract the crossing angle from the spectrometer
bump. This gives net crossing angles of +/- 150 microrad or +/- 100 microrad,
respectively. The former option is preferred. The other option could be used for
ions, where even 0 angle is possible (large bunch spacing).
The situation is similar for LHCb, but in the horizontal plane.
JJ mentioned that a revision of the ion parameters is in progress.
FR suggested that once the parameters are settled updated footprints could be
produced and published. WH only expects a small change.
FR informed the LCE team that S. Myers and L. Evans have decided to publish
an updated LHC design report by September 2003.
-----------------------------------------------------------------------------------------------------
(3) Tune Shifts for Higher Order Head Tail Modes (EM)
-----------------------------------------------------------------------------------------------------
EM calculated the effect of the collimator impedance on higher-order coupled-bunch
head tail modes, using the Sacherer formalism. Modes with m>0 cannot be damped
by feedback, and must be stabilized by Landau damping using octupoles. For each
head-tail mode, EM computes the most critical coupled-bunch mode index, and then he
finds the real and impaginary part of the tune shift for this case. He applied this
calculation to a graphite collimator with and without inductive bypass. The calculation
was done for 2808 bunches. FZ asked whether the gaps between trains are included.
The calculation will be updated to take this into account. FR mentioned that filling all
empty positions with bunches would give the most pessimistic case. [By the afternoon,
EM already had new results. The changes are at the 20% level.]
EM's modelling of the impedance was done in collaboration with LV. For a normal
graphite collimator, he finds that the mode m=2 is at the limit of stability (imaginary
part of tune shift equal to 1e-4). Adding the inductive bypass in the calculation reduces
the imaginary part of the tune shifts by a factor of 100, and there is no longer a problem,
though the real part of the tune shifts is still at the 1e-4 level. How this enormous
reduction comes about was not obvious from the plots of mode spectra and impedances.
It was noted that only every other mode order (0,2,4) is potentialy unsatble, while the
odd modes are suppressed by many orders of magnitude. FR suspected that one can
explain this by a parity argument.
FR concluded that the collimator impedance could make a significant impact on the
higher-order head-tail modes, which cannot be stabilized by the feedback.
With a chromaticity of 0.1 instead of 0.0, the result was essentially unchanged.
---------------------------------------------------------------------------------------------------------------------
(4) Coupled-Bunch Wakes for Electron Cloud (DS)
----------------------------------------------------------------------------------------------------------------------
DS presented a new routine for calculating multi-bunch wake fields and presented
first intriguing results for the SPS. The old routine was based on momentum balance,
a questionable approach if image charges are present. The results were also doubtful.
the new routine computes the field on axis. A large offset, a large number of particles,
and several seeds are needed to reduce the noise to an acceptable magnitude.
FR suggested to not only look at the center of the beam but also at the transverse
variation, which was found to be a large effect in single-bunch simulations.
DS showed that without offset there is a large random wake during the build-up phase
of the electron cloud, but it decays later towards zero. After this stabilization
he displaces a single bunch to compute the wake on subsequent bunches. The horizontal wake
lasts for more than 1 bunch, with a decay length of about 3 bunches, that would fit experimental
observations in the SPS by Karel Cornelis. Most interestingly, the wake field oscillates inside
the bunch. The vertical wake shows a fast decay after1 bunch.
FR recommended to plot the hidden part of the wake, between bunches.
JJ cautioned to be careful how to apply this type of 'wake field', which depends not only
on the offset bunch. FZ reminded the team that E. Perevedentsev had generalized
the wake and impedance concept by Fourier transforming in two dimensions.
DS also showed that the x-y interference in the simulation can be significant. The linearity of the
wake is good up to offsets of about 5 mm.
------------------------------------------------------------------------------------------------------------------------------
(5) Information on Visitors (FR)
--------------------------------------------------------------------------------------------------------------------------------
FR informed the LCE team about a new procedure for visitors who come to CERN for
one day or more. The purpose, objectives and goals of the visit should
be declared in advance. After the visit the results should be summarized by the host
CERN person in a short note and sent to Jean-Pierre Riunaud. This 'mini-Maps' scheme
should allow an assessment of the benefit we gain from our visitors.
------------------------------------------------------------------------------------------------------------------------------------
(6) Summary of Oxford Workshop on Space Charge Simulations (EM)
--------------------------------------------------------------------------------------------------------------------------------------
EM summarized the highlights of the Space Charge Workshop organized by Chris Prior.
There were 38 participants and 28 presentations, including 3 presentations from CERN
by Roberto Cappi and Elias. The CERN talks were on longitudinal microwave instability
with space charge (EM reported his revised theory, which now agrees perfectly
with simulations. Beams more stable (~ factor 30) than thought originally; stability is
better below transition; FZ commented that this might be an option
for the CLIC damping ring), on transverse space charge for the PS, and on emittance
sharing and Montague resonance.
The experimental stop band in the PS is 3 times higher than simulated by I. Hofmann.
A large number of codes were discussed, including ORBIT, ACCSIM, ESME, SYNERGIA,
GPT, BEST, GenTrackE. GenTrackE is a 3D self-consistent electron-cloud program
presented by Andreas Adelmann. BEST was advertised as the only code that can model
two-stream instabilities, by H. Qin of Princeton. The code VADOR is a Vlasov solver written by
Eric Sonnendrucker of Univ. of Strasbourg. It was suggested that we could invite him to give a
presentation at CERN and to install this code.
W. Chou reported about the situation at Fermilab. The booster is the bottleneck. 30% of
the beam are lost in 3 ms. There are two suspected reasons: aperture limits by doglegs,
and space charge. Removal of one dogleg improved the losses (champagne celebration).
But the codes ORBIT and SYNERGIA make different predictions for the space-charge
effect.
A set of benchmark tests were proposed by H. Qin. The cases to benchmark are two
measurements from the PS: emittance exchange on Montague resonance, and emittance
blow up when crossing integer and half-integer resonances. Some workshop participants
will be invited to CERN for these benchmark discussions and experiments.
A question was asked about the space-charge counterpart at CERN writing the mini-Maps
summary for F. Jones' visit. The answer was not clear, since H. Schonauer has retired.
-------------------------------------------------------------------------------------------------------------------------
(7) Update on Emittance-Growth Simulations with HEADTAIL (EB)
-------------------------------------------------------------------------------------------------------------------------
In recent simulations a random phase advance was selected between consecutive
beam-electron interactions. The emittance increases much more than in the previous
simulations with a fixed phase advance. But for the random phase advance there is a
monotonic decrease in emittance growth as the number of interaction points is
increased. Plotting the growth rate versus the number of IPs shows that it seems
to decrease as 1/N, approaching zero (N ist the number of interaction points
per turn). The physics of the emittance growth looks, however, different from
the case of fixed phase advance. Here it appears to be an incoherent blow up,
perhaps due to continous filamentation, in a randomly mismatched optics, while
with fixed phase advances it mostly appeared to be a dynamic instability.
The lack of convergence without randomization is still not completely clear.
But whether the results with randomization describe the correct physics also needs
to be further explored. FR cautioned that we have to be careful when presenting
these results to the outside.
EB has received a computer account on NERSC, where she can run the continuous
plasma code QUICKPIC. Unfortunately she will leave at the end of next week, and return
only in September. Continuation of these studies to be discussed.
Slides from Werner Herr, Daniel Schulte, Elias Metral, and Elena Benedetto are attached.
So is an EXCEL file shown by Elias, which summarizes features of space-charge simulation
codes presented at the Oxford workshop.