-----------------------------------------------------------
Minutes of the ABP-RLC team meeting of 02.12.2005
present: EB, FC, UD, AG, WH, EM, GR, TP, FR, RT, FZ, Mario Deile
web site: http://ab-abp-rlc.web.cern.ch/ab-abp-rlc/
------------------------------------------------------------
(1) Minutes of Last Meeting, List of Actions (FR)
------------------------------------------------------------
=> ACTION => Check exact time distance of bunches (RA)
No news.
=> ACTION => literature search for 2-beam wake field effects (EM, FZ)
No literature was found so far. FZ contacted W. Chou, who responded that he
did not study this question.
=> ACTION => HFSS simulation (EM, AG), may require 3D drawings, not yet existing
Pending.
=> AG will discuss with RA.
Pending?
=> ACTION => Attend dedicated collimation meeting on Thursday morning (AG)
Done.
=> ACTION => compute stability limit for copper secondary collimators (EM)
Done. See EM's report below.
=> ACTION => A one page recommendation is needed by Wednesday afternoon (FR)
Done. One page was sent to SLAC. SLAC's response was received (see later).
=> ACTION => Draw tentative conclusion for two limiting cases (EM)
Pending. EM will report to the APC next week.
=> ACTION => Check adequacy of conducting boundaries (GR)
Partly done (see below).
(2) Measuring negative transverse impedances with a single
(displaced) wire: theoretical predictions (EM)
--------------------------------------------------------------
EM reviewed a paper by Danilov and Burov (PRL 82, 11, 1999), which
calculates driving and detuning wakes for three different geometries.
Drawing the analogy with the wire measurement, EM concludes that a
negative impedance is no surprise, since it is the sum of driving and
detuning impedance which is measured by a single wire. The detuning wake
can be of opposite sign and larger than the driving wake.
FC pointed out the special conditions of the PS kicker, where the
boundary is half electric (left and right) and half magnet (top and bottom),
and a displacement of the wire towards the metal wall leads to a decrease
in the impedance.
FZ suspected that a linear term was missing in the expansion of the
longitudinal impedance in powers of the transverse displacement
for the asymmetric case.
The Danilov-Burov paper proposed a cylinder with negative detuning wake
as a means of increasing the TMCI threshold.
EM also derived the following relations between the impedance of a
1-sided vertical boundary with that of a 2-sided boundary or a
circular boundary:
Z_x,1-sided = 0.6 Z_x,2-sided,
Z_y,1-sided = 0.3 Z_y,2-sided, and
Z_x,1-sided=Z_y,1-sided=Z_circular/4
(3) LHC stability limit for Cu secondary collimators (EM)
---------------------------------------------------------
The total number of collimators considered is 42. EM replaced
the 15 marked as secondaries by Copper collimators and re-computed
the stability diagram. The beam is stabler than for the phase-1
collimator, and stability is lost at about 85% of the nominal
intensity (instead of 50% for phase 1).
The calculation needs to be further refined:
- modify TCDQ contribution using correct impedance for asymmetric devices
- update collimator list (from Guillaume)
- take into account heating of Cu, which could increase
resistivity by factor 3 (comment by FC)
Replacing graphite collimators by copper ones reduces the impedance
at high frequencies, but it increases the impedance at 10 kHz.
(4) Comparing resistive-wall impedance for Cu and stainless steel (EM)
----------------------------------------------------------------------
EM showed a plot of the real and imaginary impedance for a copper
and a stainless steel pipe of identical thickness. The curves are shifted
up by about a factor of 50 in frequency for stainless steel.
(5) Report from Engineering Design Review of TOTEM Roman Pots (EM)
------------------------------------------------------------------
EM gave a summary of the review meeting. The impedance of the
Roman pots could be significant. It was measured by M. Deile with
help of FC to be about 18 nOhm at 740 MHz, with Q=114, and power < 200 W,
for a single Roman pot. In total there are about 12 Roman pots per
beam in CMS, and possibly 4 in ATLAS.
To reduce the impedance, ferrite can be installed for damping.
This can reduce the impedance by a factor of 5. FC pointed out
that the ferrite should not be seen by the beam. The drawing,
and perhaps also the experimental setup, should be modified accordingly.
AG asked for the distance of the Roman pot edge from the beam.
MD answered that for the TOTEM optics this is 0.8 mm, for the nominal
LHC conditions it may be 5 mm.
(6) Ecloud: code benchmarking for long-term emittance growth (EB)
-----------------------------------------------------------------
EB reported the outcome of recent code benchmarking with
G. Franchetti, who visited CERN the week before. The hypothetical
mechanism of the slow emittance growth due to electron cloud
is resonance crossing and trapping as for the effect of
space charge. HEADTAIL code employs a numerical field calculation,
which could contain noise. Giuliano's code, originally developed
for studying space-charge phenomena, is based on the analytical
field, typically for a Gaussian distribution.
benchmarking was done for parameters approximating LHC at injection.
For simplicity a Gaussian electron distribution was considered,
for which the analytical force is known, and it was further assumed
that the electron density increases linearly along the bunch.
Comparisons were made for cloud rms sizes equal to various fractions
(0.25, 0.5 and 1) of the beam rms size, and for different maximum
tune shift values. In genberal there is an astonishingly good
agreement between Giuliano's code and HEADTAIL, which indicates that
the HEADTAIL Results are physical and not determined by noise.
Sometimes, for large tune shifts, the HEADTAIL result shows a fast
initial rise, which is not seen in the other code and needs to still
be understood. Both codes exhibit the same dependence on the number
of cloud-beam interaction points.
As a next step of benchmarking a more realistic model is being used,
where the maximum tune shift along the bunch is taken from HEADTAIL
simulations of the pinch. Early results also here suggest reasonable
agreement. G. Franchetti computes the rms size of the electrons
dynamically assuming charge conservation.
FZ pointed out that this ignores electrons streaming in from
large amplitudes, which are included in the HEADTAIL simulation
and may account for a somewhat reduced growth for Giuliano,
with this model.
EB will still investigate a potential problem due to the grid size
when the pinched electron distribution is narrower than the grid size.
As a next step, it is planned that Giuliano's fast code is used
to simulate a quasi-continuous interaction for a realistic model
of the SPS or LHC.
RT suggested to track backwards in order ti understand the size of
the noise.
FC asked for the time scale over which the incoherent tune spread
should be measured by the Schottky monitor. FZ thought the measurement
may need to be faster than the synchrotron period.
(7) SLAC collimator design for Phase II rotating collimators (FR)
-----------------------------------------------------------------
SLAC has responded to our recommendation, asking two or three questions.
One is whether a taper angle of 5 degrees or 15 degrees should be used.
5-degrees is better in all likelihood. There is some concern about
the proposed tapering of the grooves. Tapering is necessary to reduce
the impedance, which otherwise, with 250 kOhm per collimator, becomes
larger than acceptable.
The effect of the tapered grooves on the impedance has not yet been
simulated.
(8) Meeting on TCLI collimators (AG)
------------------------------------
AG reported from a meeting on the design of TCLI. There are two types
of this secondary collimator, one with tungsten the other with graphite.
To assess the impact of these objects we need to understand the
wake field for two beams. For trapped modes, this may be fairly
straightforward. A theoretical treatment appears needed for the
resistive wall.
(9) Conducting boundary conditions in HEADTAIL (GR)
---------------------------------------------------
GR checked the boundary conditions as implemented in HEADTAIL,
looking at 3 different cases. No problem was discovered.
In the case of two stripes, the boundary really changes
the field at the center of the chamber significantly,
which probably explains the observed strong effect on the
instability growth time.
(10) AOB
-----------------------------------------------------------
FR mentioned an idea by SF to use directional impedance
for compensating the effect of long-range collisions in the
flat-beam optics.
FR also pointed to a recent note by Theo Demma, summarizing
first comprehensive e-cloud map results for the LHC.
This may offer the potential for many additional and fast studies.
A paper by W. Fischer and U. Irizo was published recently as well.
Posted on the web: Slides by EB, EM, and GR.
Web site: http://ab-abp-rlc.web.cern.ch/ab-abp-rlc/