Minutes of the ABP-LCE team meeting on 20.02.04
present: EB, WH, AK, EM, FR, EV, FZ
justified absent: DS
----------------------------------------------------------------------
(1) Minutes +ACY- Pending Actions
-----------------------------
ACTIONS (mainly DS):
+AD0APg- clarify geometry of trapped modes
+AD0APg- compute longitudinal impedance
+AD0APg- practical advice for collimator design
+AD0APg- redo estimate for larger gap
+AD0APg- understand stability diagram for m+AD0-4
+AD0APg- what can be measured in the SPS?
DS was not at the meeting. He has written
a draft memo on the collimator impedance
and sent it to his contact in the
collimation team.
+AD0APg- perform full simulation for the SPS (AK)
Not yet done.
(2) Review of resistive wall formulae
and measurements (AK,EM,FZ)
---------------------------------------
EM presented a new general formula
for the transverse resistive impedance.
It is derived from the multi-layer
longitudinal impedance of L. Vos,
which is converted to transverse
impedance following L. Vos'
prescription for the inductive bypass.
This new formula agrees well with
the measurement of Fritz Caspers
and with the Burov-Lebedev formula.
Applying this formula to the LHC
collimators EM found that the
difference to the thick-wall
inductive bypass result was negligible
(1+ACU- or less).
AK evaluated various expressions
for the resistive transverse impedance,
including the new formula from LV-EM.
He confirmed an excellent agreement
between the new L.Vos-EM expression
and the Burov-Lebedev result for the
laboratory measurement of Fritz.
For the collimator, the two
expressions deviate at low frequency.
Why there is a difference in
this case, and which one is
more correct remained unclear for
the moment. It is not relevant
for the LHC. The new expression
approaches the thin-wall inductive
bypass formula at low frequencies,
the Fermilab formula is offset.
The latter was also evaluated for more
than 2 layers, considering copper outside
the graphite. The copper enhances
the impedance at lowest frequencies.
FZ showed similar independent evaluations,
not including the new LV-EM formula. He found
that Burov-Lebedev's result was closest to
both measurements and Tsutsui's simulations,
and that thick-wall inductive bypass
expression deviated considerably. The
discrepancy with the previous two
speakers might hint at an error in
his mathematica notebook or at a unit
problem.
Possible actions:
+AD0APg- compare results and mathematica notebooks
(FZ, EM, AK)
+AD0APg- suggest that Fritz may attempt a
measurement in a parameter region that
is closer to the LHC collimators (FR)
+AD0APg- determine what can be measured in
the SPS (AK, EM, FR, DS, FZ,...)
+AD0APg- check epsilon and mu of collimator
material for impedance calculation
(mu of the graphite was checked by EM
and it is close to 1)
It was suggested that CERN adopts a general
policy of using low-mu stainless steel
for accelerator components near the beam.
(3) Electron Cloud: analytic results and
code comparisons (EB)
--------------------------------------------------------------
HEADTAIL simulations for LHC were compared with
QUICKPIC, considering 1 interaction point in both
codes and a linear rf. Conducting boundaries were
used in both codes. The bunch length in QUICKPIC
extends over 5 sigma, while for HEADTAIL
both 2 sigma and 5 sigma were simulated.
The emittance growth was different from
that in QUICKPIC for either case. QUICKPIC
shows a fast vertical instability, which
is absent in HEADTAIL. Perhaps some other
parameter is not the same, or it could be an
effect of electron space charge. The latter was
not included in HEADTAIL. FZ remarked that GR
had performed simulations with and without
electron space charge in HEADTAIL and that
he saw a negligible effect, in the past.
EB will visit USC for 9 days after ECLOUD'04,
to continue comparison of codes and to learn
the use of QUICKPIC.
FR proposed to perform a series of HEADTAIL
simulations to study the chromaticity needed,
in LHC, for a minimum emittance growth (say
less than 0.1+ACU- over the time of the
simulation) as a function of electron density.
This should use several interaction
points, and perhaps be done for both dipole
fields and field-free regions. Effect of
feedback could be explored in addition.
ACTION +AD0APg- perform comprehensive
HEADTAIL simulation study for LHC (EB)
EB then reported on progress with the analytical
treatment of the electron pinch. For a Gaussian
longitudinal profile and a linear force
she solved the electron motion using the WKB
approximation. Inverting the result, the
density of electrons is obtained as a function
of radius and time during the bunch passage.
This can be compared with simulations.
Except for the WKB approximation the agreement
should be exact, but some differences were
visible, perhaps due to small differences
in the parameters. If nonlinear forces
are added to the simulation, the electron
density starts to resemble a step function
which assumes a nearly constant value after
the first quarter oscillation in phase space.
The solution using the WKB approximation
preserved the symplecticity. It is not
clear whether this is generally the case.
(4) Progress on Feedback Simulations etc. (EV)
-----------------------------------------------
Work in progress:
1. Create common program for beam-beam simulations
and feedback. This could be done in SIMULINK.
The option +ACI-accelerator+ACI- compiles the code and
gains a factor 10 in speed.
2. Paper on simulations of CNGS extraction with
feedback.
3. In contact with J. Tuckmantel regarding
technical possibilities of rf amplitude modulation
in the LHC.
FR recommended to look at the kickers for extraction
and injection into the LHC. EV remarked that plans exist
to cancel the ripple of LHC injection and SPS extraction
kickers against each other.
(5) Reports from meetings:
LTC and follow-up of Chamonix workshop (FR)
---------------------------------------------
In the LTC, S. Myers reported open questions
from Chamonix. FR listed topics relevant to us
and people who will report back:
Pulse-to-pulse emittance variation up to 20+ACU-
due to kicker rise time (V. Mertens)
Satellite bunch intensities - satellites are
created in the PS
MKE kicker heating problem.
Impedance of all kickers in the LHC and in SPS+ADs-
estimates, measurements, how to improve shielding
of MKE kickers in the SPS (EM, FR, FZ)
Scraper in the SPS+ADs- collimation in transfer
line (HBu, P. Collier)
75-ns beam and short bunches in PS booster and PS
(needed by end 2004)
156-bunch TOTEM beam in the SPS, 1 test in SPS
with equidistant bunches (GA)
Beam-loss monitors
Beam quench limits of LHC magnets, sector test
or slow extraction from SPS?
JPK remarked that beam-loss monitors are not
sufficient to adjust the collimators. Alternative
is to equip collimator jaws with bent crystals.
FZ suggested to use the wake field kick and
closed-orbit change for determining the
distance between collimator jaw and beam and
to center the two jaws. EV suggested to
look at the background in the experiments.
EM mentiond that the collimator distance from
the beam is 1.2 mm to 3.8 mm in physics.
FR remarked that this distance is significantly
larger at injection (4.7 to 11.1 mm) and the
corresponding wake field kicks become very small.
Beam efficiency for each system on a single
paper.
Injector chain emittance blow up.
Ion operation, lifetime of some magnets operated
close to the quench limit.
(6) Other News: Computing Power (FR)
-------------------------------------
L. Evans agreed to the purchase of 50 new
CPUs+ADs- for the next months LCE studies gain
priority in the queues.
ACTION +AD0APg- FR will investigate whether there
exists a special priority queue or whether
priorities are automatic.
(7) Geometric collimator impedance (EM)
---------------------------------------
EM applied L. Vos formulae to estimate
the geometric impedance of the LHC collimators,
both longitudinal and transverse.
The real part is zero until the pipe
cut off and then assumes a finite constant
value. The imaginary part peaks at the
cutoff. The longitudinal impedance is
about 0.1 mOhm per collimator for a
distance of 2 mm and 15 degrees taper.
The transverse impedance is not negligible.
It is estimated at 0.2 MOhm/m.
This could be compared with DS' Gdfidl
results.
(8) Stability diagram for m+AD4-0.
-------------------------------
EM remarked that the stability
diagram should not change with head-tail
mode number m. FZ suggested that there
could be an effect if the synchrotron
tune spread contributes to the
Landau damping. Since often the
bucket is filled, this could
be the case, at least at injection.
FR mentioned that the contribution
from synchrotron tune spread
had been studied by Y.H. Chin
(see also LHC-Project-Report-121,
enclosed).
Enclosed: Slides by EM, AK, FZ.