-----------------------------------------------------------------------
Minutes of the ABP-LCE team meeting of 23.07.04
present: EB, WH, AK, EM, TP, FR, Toshio Suzuki, FZ
excused: DS, EV
-----------------------------------------------------------------------
(1) Previous Minutes & Open Actions
-----------------------------------
Check how contact resistance of collimator jaws
relates to the usual beam impedance and budget
==> ACTION EM & FR?
STATUS: pending
The algorithm by Luc Vos could be used to get an estimate.
==> ACTION FZ, further RHIC simulations for larger
delta_max and a larger number of turns
STATUS: DONE - see presentation below
==> Compare simulation for quadrupole
(FZ and DS)
STATUS: quadrupole simulations done by FZ;
discussion with DS still needed
==> ACTION: Compute energy loss in the LHC kickers
due to higher-order modes
STATUS: pending (action for DS?)
FZ mentioned that several versions of the ELCOUD code
presently co-exist. The code should be re-unified again.
==> ACTION (DS & FZ): create new standard version of ECLOUD
==> ACTION (DS & FZ): create sample input files for
debugging of new code versions
(2) Stability Diagrams using L.Vos & Burov-Lebedev's Formulae (EM)
------------------------------------------------------------------
EM showed new calculations of complex tune shifts using either
theory for all collimators with and without 5-micron Cu coating.
The difference between LV and B-L is about 20%. The Cu coating
moves the working point from an unstable situation to a marginally
stable position. This calculation did not include the other
impedances in the machine. If all impedances are added, the beam
loses stability at about half the nominal design bunch intensity.
FR emphasized the advantages of a Cu coating, which can
only improve the impedance and reduce the ohmic heat load.
A possible negative recommendation from the LHC Collimation Review
should therefore be reconsidered.
There persists uncertainty regarding the results of BZ, which
differ by a factor of 100 from all other estimates.
==> ACTION (BZ) presentation in an LCE meeting
AK reported that he discussed this problem with Eberhard Keil
at EPAC, and that EK performed a LAWAT [=LAminated WAll Transverse
(impedance)] calculation, essentially confirming the results of
LV and B-L. A word file and mathematica notebook by Eberhard
Keil will be linked to the code repository web page.
FR suggested that EK publishes a note on this calculation.
(3) Beam-Beam Simulations (WH & TP)
-----------------------------------
Work is ongoing in collaboration with Dobrin Kaltchev
to model the effect of linear imperfections on the beam-beam
interactions. Linear optics errors will perturb the distance
between the two beams, phase advances, and the beta functions
at the long range collision points. The weak-strong simulation
first models correction procedures, e.g., for orbit errors, as
would be applied in the control room. An rms orbit of
0.3-0.5 mm can have a significant effect on the tune footprint.
These simulations will allow the definition of the required
correction levels.
DK is presently performing tune scans.
FR asked whether the tune footprints can be measured in the
LHC. This is thought to be difficult, but perhaps not impossible
using a weak pencil probe beam.
TP is developing a tracking program for coherent beam-beam
interaction with multiple bunches. She will later also
participate in experiments on coherent beam-beam effects
at RHIC.
(4) Simulations of Electron-Cloud Emittance Growth (EB)
--------------------------------------------------------
From a large number of simulations, EB found that the
simulated emittance growth (below TMCI threshold) scales as
the 1.7 th power of the electron density, different from
earlier analytical predictions that suggested a quasi-exponential
dependence. EB also studied the effect of varying the
number of longitudinal rms beam sizes included in the
PIC calculation of beam-electron interaction. The number
of sigmas retained has a strong effect and even changes
the qualitative behavior. Changing in addition the number
of longitudinal slices or the number of macroparticles
also influences the result.
FR brainstormed that the disruption of a few trailing
macro-particles could be responsible for the large
sensitivity.
AK and FR proposed to consider distributions other than
a Gaussian, e.g., cosine, parabolic, or quasi-parabolic.
FR asked EB whether there was any feedback from the LHC
MAC. This was not the case. The conclusions from the
MAC have not yet been published.
(5) Electron-Cloud Build-Up Simulations (FZ)
--------------------------------------------
FZ showed some new simulation results for RHIC.
At RHIC injection, the threshold of multipacting
in the cold arcs is found for a delta_max between
2.1 and 2.2. The heat load and electron density
increase by orders of magnitude in this region,
and saturate for larger values of delta_max.
The simulated heat load at delta_max=2.2 and 2.3
is 60-70 mW/m, while the sensitivity at RHIC is 5 W
over 100 m (which amounts to 50 mW/m if the electron
cloud exists over the full length of 100 m).
So the simulated number is at the resolution
limit.
Simulations of electron-cloud decay in the SPS
for the detector of Jean-Michel Laurent show that
as expected the decay time is shorter for lower
reflectivity. But for all cases the cloud has decayed
after 1 or 2 microseconds, confirming that a
low-energy reflectivity of 1 does not mean that the
cloud density stays constant.
Simulated heat load for an LHC quadrupole at delta_max=1.2
shows a maximum near 8e10 protons per bunch, and an
overall heat load that is lower than for drift with
the same delta_max.
Simulations have started of the electron-cloud
build up in the positron linacs of future linear
colliders. The model for the field is primitive
at the moment. It is taken to be the sum of
a free-space beam field and the rf field of the
fundamental mode in a pillbox cavity.
The simulation suggests that moderate rf fields
enhance multipacting, whereas very high
fields suppress the multipacting. In particular,
preliminary results indicate that TESLA and CLIC
operate in the suppressed region, while
NLC/GLC is in the enhanced parameter range.
Field ionization can be important in the GLC/NLC
at top energy and over most of the length of
CLIC. The azimuthal distribution of electrons
at the wall is peaked in the horizontal plane,
reflecting the smaller disruption here.
(6) AOB
-------
FR reminded the team of the set up of a code repository,
which should include electron-cloud, beam-beam, and
other accelerator-physics codes.
FR also mentioned the EPAC paper by P. Tenenbaum et al. on
the measured resistive wake of tapered collimators, where
the measured wake for Copper is a factor 3 larger
than the prediction, and the paper by J. Jowett
on limits of the ion-beam performance in the LHC,
which however does not much discuss lifetimes and
emittance growth due to beam-gas interactions.
FR recommended a fresh look at the pressure limits in LHC.
FZ questioned the statement that LHC could operate
at 100 times the nominal vacuum pressure.
EM reported that he received the MOSES code from ES.
First simulation runs agree with previous results by
the HEADTAIL code.
FR suggested to compute the longitudinal loss factor
for LHC including the collimator impedance. This number
should already be available since the heating of the
collimator was computed. Coating should play a role.
Vladimir Shiltsev developed a CPT theorem for
accelerators, which predicts 4.5-6 years
commissioning time for the LHC (time until design
luminosity is reached).
Attached: Slides by EB, EM, FZ
CPT theorem by VS