-----------------------------------------------------------
Minutes of the ABP-RLC team meeting of 05.05.2006
present: UD, AG, TP, FR, GR, RT, FZ
excused: EM
web site: http://ab-abp-rlc.web.cern.ch/ab-abp-rlc/
------------------------------------------------------------
(1) Minutes of Last Meeting, Pending Actions
--------------------------------------------
ACTION: AG will write a short note on the TCDS impedance estimates that can be referenced in the Functional Specifications.
Status: AG has just published LHC Project Note 382 on the longitudinal geometrical impedance of the TCDS.
Concerning the collimator crisis, FR corrected his earlier assessment: The crisis is not quite over yet. The first real collimator delivered to CERN has one broken finger, which will be exceptionally accepted, but should not become the rule.
There was some confusion as to whether the broken finger was longitudinal or transverse. After the meeting it was clarified by RA that the broken finger was indeed a less critical transverse one, which is only called longitudinal by the engineers.
FR reported that EM has verified the thickness of Cu layer for FP420. It is only 200 microns instead of 2.5 cm, which were previously assumed. The calculation will be redone by Federico Roncarolo.
(2) Update on Transverse Impedance of SPS MKE Kicker (EM)
---------------------------------------------------------
EM had sent two slides showing the improved agreement between Fritz Caspers' and Tom Kroyer's two-wire measurements and theoretical predictions, after Fritz and Tom re-processed the measured data. The agreement with theory is now nearly perfect for the real part of the impedance.
(3) Coherent Tune Shifts vs Collimator Gap (FZ)
-----------------------------------------------
FZ presented ongoing work on the coherent coupled-bunch head-tail tune shift. The study responds to an urgent request from FR
to clarify the scaling of the tune shift with collimator gap size. After presenting the general formula used and the weight functions
for a Gaussian profile, FZ showed as a first preliminary result the fastest (Q'=0,m=0) CB-mode rise time as a function of half gap size.
The dependence is nearly inversely linear, and does not scale with the inverse third power. FZ will next compute the coherent tune
shifts for various cases.
FR commented that the LHC Design Report contains some remarks about the scaling of the peak impedance with the gap size.
FZ suggested that the Piwinski factor is a small correction for the LHC parameters, and half gaps larger than 4 sigma.
FR expressed the view that maybe the SPS experiment indeed was sensitive to the inductive bypass effect as the theoretical predictions
from Burov-Lebedev and the classical theory differed, and the experiment was much closer to the Burov-Lebedev one (when also
taking into account the Piwinski correction).
(4) Electron-Cloud Instability Threshold vs SPS Injection Energy (GR)
---------------------------------------------------------------------
GR studied in simulations the effect of a higher SPS injection energy of 60 GeV instead of 26 GeV on the single-bunch e-cloud instability.
This talk followed on his earlier RLC presentation on the scaling of the conventional TMCI instability, whose threshold was much improved.
For the moment, he only considered an electron cloud without magnetic field, zero chromaticity, e-cloud densities around 1e12 m^-3, and
the same bunch length (based on input from Elena Shaposhnikova). The grid size awas 30 sigma_x times 10 sigma_y. The beam pipe
dimensions therefore shrink with energy, which is not quite realistic. However, the result does not much depend on the boundary conditions.
10 IPs per turns were selected, and GR confirmed that the result did not significantly change for a larger number of interaction points.
The beam was found to be more unstable at 60 GeV than at 26 GeV in both transverse planes. The horizontal growth rate is higher than the
vertical.
A look at the analytical scaling formula presented by FZ at the GSI-CERN bilateral meeting revealed that a reduction in stability is indeed
expected at 60 GeV without magnetic field. With a dipole field, the pinch effect would be reduced, and the beam is expected to be more
stable at the higher energy. FZ also noticed that a short reduction in bunch length would help suppress the instability.
FZ pointed out that the longitudinal emittance here is 0.35 eVs in the SPS, while it is 1 eVs at injection into the LHC. The team jointly
found that the emittance term was neglected in the GSI presentation. Taking the factor 3 increase into account the analytical scaling
suggests the same e-cloud threshold in the LHC at injection as in the SPS at injection, in perfect agreement with Elena Benedetto's
simulations.
FZ suggested that simulations be performed for 450 GeV, which could be even more unstable than the 60 GeV case, as well as for other
bunch lengths.
=> ACTION => Discussion with Elena Shaposhnikova & rf group on strategy for longitudinal parameters at LHC injection (FR).
After the meeting ES responded that the longitudinal emittance at injection to the LHC will be 0.6 eVs for nominal intensity
(which allows staging the 200 MHz RF system in LHC).
=> ACTION => SPS e-cloud simulations at 26 and 60 GeV for a dipole field (GR).
(5) Measured SPS Correlations of Bunch Intensity/Length (GR)
------------------------------------------------------------
GR presented further analyses of the SPS bunch shortening data, evidencing a general correlation between bunch length and
bunch intensity. He also commented on a possible theoretical explanation.
GR showed examples from several SPS coasts in August 2004. All bunches in a train shrink and lose intensity. The later
bunches lose more than the first, e.g., 75% compared with 25% over about 10 minutes. Plotting bunch length versus intensity
reveals that data for all bunches fall more or less onto the same universal curve. The only difference between bunches
along the train is that later bunches move faster along this correlation curve. In some runs, a peculiar
effect occurred for a bunch length around 0.68 ns, where the length stayed roughly constant for a while before shrinking
again.
Giuliano proposed a theoretical model, based on resonance trapping, which, if shifted but not scaled, well describes the
observed correlation of intensity and bunch length. The time numbers on Giuliano's slide refer to the numbers of SPS supercycles.
GR noticed some puzzling questions: (1) Why is the first bunch in a train affected in the same way? (2) The assumptions in
producing Giuliano's figure are not clear - why should z_t decrease, for a fixed time T, along a bunch train? (3) Why is there no leveling off towards the
end of a store?
In view of (2), FZ added another question: why is z_t decreasing with time for a fixed bunch?
FZ remarked that the high chromaticity, which also leads to trapping, could be responsible for losses of the first bunch.
FR encouraged more systematic measurements in the next scrubbing run. He also suggested contacting the rf group for
eventual higher-quality bunch profiles from August 2004.
=> ACTION => Contact rf group for possibly higher-quality bunch profiles (GR).
Posted on the web: Slides by EM, GR and FZ, Email exchange by RA and AG
Web site: http://ab-abp-rlc.web.cern.ch/ab-abp-rlc/