-----------------------------------------------------------------------
Minutes of the ABP-LCE team meeting of 08.10.04
present: EB, WH, JJ, EM, TP, FR, DS, EV, FZ
-----------------------------------------------------------------------
(1) Beam-Beam Simulation Plans (WH)
----------------------------------------------
WH's presentation addressed tune scans for dynamic aperture
with beam-beam, beam-beam with linear imperfections, and
an assessment of different crossing schemes.
Other studies are underway for coherent beam-beam effects
and 3-dimensional strong-strong simulations, but were not
discussed in this talk.
The tune scans for the dynamic aperture with beam-beam
are done in collaboration with D. Kaltchev, considering
head-on and long-range interactions, without linear or
nonlinear errors, but including collimator apertures.
D. Kaltchev will present results of these studies during
his next visit to CERN.
For beam-beam with linear imperfections, field and
alignment errors are included. Only information available
in the control room is used in the simulation, e.g.,
MAD matching is not employed and the ideal model is
considered as a basis for tuning and corrections.
Only chromaticity and tunes are adjusted with global
knobs, since the result is identical to the knob tuning
of the control room.
Several changes to MAD-X were necessary for these studies,
as, e.g., the installation of BB elements wiped out all
previously defined errors, and a special treatment of
correctors caused additional problems. In consequence,
the treatment of MAD tables was generalized, now allowing
for a fast read-in of errors and for a read-in of orbit
corrector settings.
The status of the linear-imperfection study is as
follows: a suitable "private version" of MAD-X exists,
only a few knobs are missing, and the coupling correction
will be implemented by adapting the module of S. Fartoukh.
The study of crossing schemes compares HH, HV, and
VV crossing, comparing both nominal and PACMAN bunches.
Different integer tunes will be investigated. The study
uses V. 6.4 and the 2 main IPs. Tracking is done
with SIXTRACK.
In the longer term this simulation shall be combined
with that of linear imperfections and corrections.
A draft note with D. Kaltchev exists. Input files
for the imperfection studies are available. The new
MAD-X version will become fully operational with the
help of Eric McIntosh and FS.
(2) Follow-Up on MKE Kicker Impedance with Different
Ferrite (EM)
-----------------------------------------------------
Following up on a comment by F. Caspers that the
ferrite used in the MKE kickers is of type 8C11
and not 4A4 as previously assumed, EM has reviewed
and recalculated the transverse kicker impedance.
In LHC Project Note 234 H. Tsutsui computed the
transverse impedance for 4A4 ferrite. Between 0
and 1.7 GHz the imaginary part decreased from
0.4 MOhm/m to 0.2 MOhm/m, the real part increased
from 0 to 0.2 MOhm/m. The longitudinal impedance
has a different behavior and a much lower
resonance frequency.
In CERN-AB-2003-088 (ABP) L. Vos computed the
effective imaginary impedance of an MKE kicker
as 1.25 MOhm/m, also for MKE. the number seems
about 3 times higher than Tsutsui's. The impedance
of L. Vos is inductive until about 0.6 GHz (much
lower value than in Tsutsui's model).
EM and A. Burov computed the kicker impedance
recently, considering a horizontally infinite width
and also assumed 4A4 ferrite material
They obtained an imaginary effective impedance of
0.6 MOhm/m. The impedance is well represented by
a broadband resonator with a resonance frequency
of 1.7 GHz. For 8C11 ferrite the imaginary part
of the effective impedance is also about 0.6 MOhm/m.
The real component of the magnetic permeability, mu',
is 5 times higher than for the 4A4 kicker. Data for
mu' and mu'' data are available only up to 1.8 GHz.
FR commented that higher frequencies are of somewhat
academic interest, since SPS bunch spectra for mode m=0
do not extend beyond ~0.5 GHz. However higher order mode
spectra extend to higher frequencies and m=2-3 may be
relevant for the transverse mode coupling instability.
ACTION -> EM will check with BT experts and/or with FC
whether 4A4 ferrite data is available at higher frequencies.
Otherwise some analytic model will be used to extrapolate
the available data to higher frequencies.
(3) Emittance Growth by Scattering off the Residual Gas (FZ,JJ)
--------------------------------------------------------------
FZ reviewed the effects of the residual gas on the beam:
beam lifetime and local losses are determined by inelastic
nuclear interactions. The emittance growth is caused by
multiple Coulomb scattering.
The LHC design vacuum pressure is specified to correspond
to a beam lifetime of 100 h. One can thus calculate, for
different gas species, the corresponding molecule density
and pressure, and, from this, the related emittance doubling
time, which depending on the species amounts to a value of
a few hours to a few tens of hours at injection energy.
These numbers were published in the LHC design report,
and recently verified independently by B. Jeanneret.
The emittance growth time scales with the square of the energy
and the inverse beta function. The effect should be much
larger at injection into the SPS. The expected emittance
growth time can be estimated from the measured gas pressure
and gas composition. The plans for the future include
an improved understanding of the limitations imposed by
the cryogenic system, an estimate of the effect of
elastic nuclear interactions on the beam emittance,
and a calculation of beam-halo formation in the LHC based
on the formalism of Lebedev and Nagaitsev. Intermediate
results may be presented at the HHH-2004 workshop.
ACTION: Estimate emittance growth time and beam lifetime
from gas scattering in the SPS (FZ?)
JJ presented some chapters of the Mathematica notebook
computing luminosity lifetimes, which was developed
by him in collaboration with Amy Nicholson. This notebook
also computes emittance doubling times and beam lifetimes
due to gas scattering. JJ uses the formulae by Barashenkov
for the nuclear cross section. Assuming the gas densities
calculated by A. Rossi and N. Hilleret for the LHC
interaction regions, the emittance growth time for
protons is much longer than 100 hours.
FR commented that the values for the interaction region
are necessarily optimistic, and much lower than the densities
promised for the arcs. Assuming the same densities as FZ,
the beam lifetime and emittance growth times calculated
by JJ appeared comparable.
(4) Effect of CNGS Extraction Kicker (EV)
-----------------------------------------
EV presented results from an SPS MD of 23.09. where he studied
the effect of feedback on the 2nd CNGS batch after extracting
the first batch. The second batch is excited by the rising and
falling edge as well as the flat-top ripple of the extraction
kicker pulse. Without feedback orbit oscillations above 0.5 mm
result. With feedback the oscillations are damped within about
150 turns. The second batch is extracted about 2000 turns after
the first. The remaining problem are proton losses for the first
and last bunches. Regardless, the MD result was successful and
showed that the kicker ripple has been improved.
(5) Electron-Cloud Simulations with HEADTAIL (EB)
-------------------------------------------------
Following up on an earlier suggestion by EM to study the
possibility of a regular head-tail instability, EB performed
HEADTAIL simulations with Q'=2 and Q'=0. Both above the TMCI
threshold and below it, the emittance growth is almost unchanged
for the two chromaticity settings. This suggests that the
regular head-tail instability is not responsible for this
emittance growth. EM suggested to perform the same simulations
for a BB resonator model.
EB showed simulations below TMCI threshold for different
numbers of interaction points per turn, up to 100. There is
no clear convergence of the number. The result in the
horizontal plane seems to oscillate between two levels.
FR, DS, and FZ noted that changing the number of interaction
points is equivalent to changing the effective tune plus the
electron density.
ACTION (EB): Plot results as a function of effective phase advance
Perform tune scan with a constant electron density
Simulate the situation of the recent SPS experiments
(6) Future LCE Work on LHC and SPS Impedance Database (FR)
----------------------------------------------------------
EV has offered his help and exchanged emails with FR.
However, the man power remains critical. DS is 90% working
for CLIC. A. Grudiev will help us to a certain extent.
Until the arrival of a new staff, a reorganization of the work
is required for the next several months.
FR asked whether the impedance of the instrumentation is
correctly represented by a canonical number Z/n~1mOhm from
L. Vos and whether all instruments are taken into account.
FZ suggested that with DS a new user-friendly database program
suiting our needs might quickly be developed. JJ suggested
that Mathematica would be the ideal tool for such a database.
DS and FR brainstormed that a statistical calculation may
be needed in many cases, e.g., for varying collimator
positions or other uncertainties, and that the future database
should facilitate such descriptions.
FR also stressed that the database is not our primary goal,
but a main objective of the team is to collect all hardware
data relevant to the LHC (and SPS) impedance, and then to construct
impedance models, and make beam stability estimates.
Attached: Slides by EB, WH, JJ, EM, FZ