-----------------------------------------------------------------------
Minutes of the ABP-LCE team meeting of 01.10.04
present: EB, EM, WH, TP, FR, EV, FZ
-----------------------------------------------------------------------
(1) Follow Up on Pending Actions and News (FR)
----------------------------------------------
FR informed the LCE team that the ABP group will be
restructured in 2005.
The number of news posts in ABP may be much less
than previously expected. Existing senior staff is
assumed to contribute to CARE and EUROTeV.
PAST ACTION: EM will check the effect of the lattice
on the Montague resonance.
EM contacted I. Hofmann, who shared EM's opinion
that the PS lattice is sufficiently smooth that
a constant focusing approximation is valid for
space-charge effects. Nevertheless, the GSI group
is performing new simulations which will test this
assumption.
Concerning the dynamical effect of the resonance
I. Hofmann believes that longitudinal mixing alone
does not suffice to reproduce the PS measurements,
and that an additional ingredient, perhaps intrabeam
scattering, may be needed to achieve agreement
with observations.
FR mentiond that during Massimo Giovannozzi's AB
seminar presentation, Steve Myers asked whether
we can simulate space charge effects on the island
shape for a multi-turn injection scheme.
Presently no good simulation tool exists at CERN.
FZ recommended exploring this question with the
ORBIT code, which has been benchmarked on the LANL
PSR and on the PS and which can read a MAD input
file. EM suggested collaborating with GSI.
EM announced that a new intensity record was
reached in the SPS with 5.6e13. FR asked for
the reason behind this intensity record, e.g.,
whether it is related to the increase of linac
intensity, and what the present limitations are.
One reason is that the injection energy of the
PS has been increased from 1 GeV to 1.4 GeV.
The SPS intensity may be limited longitudinally.
EM informed the team that studies with EB et al.
on the TMCI threshold with space charge make
good progress.
(2) MKE kicker impedance and heating (EM)
-----------------------------------------
EM explained that earlier calculations of the kicker
impedance performed with A. Burov were invalid, since
both the ferrite material and the geometry were incorrect.
A revised calculation is being worked on.
FR emphasized that earlier computations of the longitudinal
kicker impedance likely also were wrong. He asked
whether there exists a simple relation between longitudinal
and transverse impedance. EM answered that the present
scheme of computing exhibits no obvious relation.
FR stressed that coating of the kicker is not possible
due to high voltages which could lead to sparking.
FZ mentioned that all SNS kickers are coated and suspected
that the SNS extraction kickers operate with similar voltage.
EM discussed the amplitude and power spectra for various
distributions following a request by J. Uythoven. The power
spectrum shows little qualitative difference between a
Gaussian and other distributions. If the full length of
other distributions is chosen as 4 sigma (where sigma
is the standard deviation of the Gaussian), the heating
of the kicker varies by up to 50%. FZ pointed out
that if the width of the distribution function would
be adjusted to match the shape of the beam core in each
case, the variation of the power would be much lower.
An alternative approach would be to interpolate the measured
beam power frequency distribution by splines and integrate
numerically over the product of beam power and impedance.
EV asked whether the measured bunch length is representative
for all bunches and over a longer time; this may not
be the case.
(3) LCE future work on LHC and SPS impedance database (FR et al)
----------------------------------------------------------------
FR's presentation to the LTC was postponed due to a computer
problem and the news on the MKE kicker impedance. Another
complication is the lack of a complete updated impedance model
for the LHC and the SPS. Recent systematic work on the SPS
resistive-wall impedance was done only by A. Koschik who compiled
apertures all around the machine. The work on LHC and SPS
could go in parallel. The LCE team should make a new effort
to proceed with this task.
This work requires a proper interface with several other
sections and groups, in particular the rf group, which
contributes both impedance measurements and some simulations.
(4) Torino School on Plasma Physics and e-cloud simulations (EB)
----------------------------------------------------------------
EB reported highlights from a school on NON EQUILIBRIUM SYSTEMS:
TURBULENCE IN FLUIDS AND PLASMAS (see web site
http://www.isi.it/42/events_detail.html )
Among other items, the plasma dynamics in a Tokamak was discussed,
with currents of the order of MA and plasma densities of
1e20 m^-3. FR asked whether part of the electron cloud dynamics
can be described by a fluid approach.
EB reported new HEADTAIL e-cloud instability simulations for LHC
where she compared with nominal PIC simulations with a hybrid method
where the field of the electrons is computed by the PIC module
but the field of a beam slice is deduced in a soft-Gaussian
approximation. The emittance growth is not the same. The PIC code
shows initially larger growth horizontally. At later times, the
trend reverses and the soft-Gaussian scheme leads to a fast
instability. EM asked whether the emittance growth seen below
TMCI threshold could be explained by a regular head-tail effect.
FZ answered that regular head-tail instabilities had been seen
in earlier microbunch simulations with K. Ohmi, but were never
seen with the codes HEADTAIL, PEHTS and ECI. For confirmation,
EB will re-run one of the cases considered with zero chromaticity
instead of the chromaticity Q'=2.
EB also performs simulations with a frozen cloud, where electrons
do not move. The emittance growth for this case can be compared
with one using an analytical expression for the electron field.
Any additional emittance growth in the former case must be due
to the PIC calculation. Appropriate techniques for suppressing
PIC nose will be explored.
(5) Highlights from SNS Review (instability diagrams,
kicker impedance, resistive wall wake, and e-cloud) (FZ)
----------------------------------------------------------
FZ presented accelerator-physics highlights from the
SNS ASAC review 27-29 September, in particular slides and
studies by S. Henderson, J. Wei, D. Raparia and V. Danilov.
The topics addressed were stability diagrams with space
charge and momentum spread for bunched and coasting beams,
impedance of the coated ferrite kickers, low-velocity
resistive wall wakes, and electron cloud.
Following an earlier recommendation, V. Danilov, together
with A. Burov, studied the EPAC 96 paper by Karliner et al.,
which described a novel resistive-wall wake effect for the LHC.
VD and AB found the results in this paper to be invalid.
Being asked, the authors of the 1996 paper have distanced
themselves from their earlier result. This effect may thus
not occur in the LHC.
The SNS injection kicker consists of a ceramic coated with
copper and TiN on the inner side of a ferrite-filled tank.
The longitudinal impedance is given solely by the thickness
and resistivity of the coating and the transverse impedance
follows from the usual scaling, suggested by the Panofsky-Wenzel
theorem. This impedance is valid for a kicker of finite
length and much higher than what would be expected from
the Piwinski solution for an infinitely long object.
The extraction kickers are placed inside the vacuum
chamber, and the ferrite is directly coated with TiN.
The impedance of these devices was computed (Davino et al.)
and measured at BNL (Hahn et al), including the full
assembly with PFN and the coating. The coating made
no difference to the impedance (this is very different
from the injection kicker where only the coating determines
the impedance). The coating of the extraction kicker is
applied in strips.
The SNS ring instability threshold is at 2-3e14 with a
nominal intensity of 1.4e14 protons. The threshold depends
on energy spread, chromaticity, beta functions and tunes,
transverse distribution, etc. A first attempt was made
to compute a stability diagram, for a coasting beam using the
simulated energy distribution due to painting and space charge.
Without space charge the result agrees with ORBIT simulations.
With space charge the stability diagram suggests that the
beam would always be unstable by a large factor, while ORBIT
simulations for a bunched beam predict stability. It is now
planned to compute stability diagrams for the bunched-beam
case, and to start from the Landau solution to the initial
conditions instead of the van-Kampen approach.
L. Wang simulates the 3D e- evolution at the stripper
foil with the CLOUDLAND code. One input is the backscattering
probability of electrons with energies between 10 keV and several
MeV, which was estimated by EGS. It is of the order several
percent and depends on the material. The stripped electrons
are guided by a tapered magnetic field towards a carbon
catcher with grooves which retain secondary electrons.
The crossing of the Montague resonance was simulated by
S. Coursineau and J. Holmes with ORBIT. The result is close
to the experimental observation and may give a better agreement
than other simulations using the IMPACT code.
The cost of the last SNS magnets has dramatically increased
due to high steel prizes driven by China boom and use of
steel for humvee armor plates.
Attached: Slides by EM, FZ