Minutes of the ABP-LCE team meeting on 14.11.03
present: GA, HB, EB, WH, JJ, AK, EM, FR, DS, EV, LV
-------------------------------------------------------------------------
(1) Minutes of Last LCE Meeting and Pending Actions
------------------------------------------------------------------
ACTION -> EM will check possible longitudinal 'Yokoya factor' relating
round and rectangular cases.
Still to be done.
ACTION -> Check discrepancy by a factor ~3 on collimator impedance
(EM+LV) and put numbers into ZBASE (EM).
Solved, no discrepancy in rise time left (see below); numbers still to be put into Zbase.
ACTION -> DS will cross-check these estimates by numerical simulations
and follow-up the impedance of Roman Pots with Roger Perret.
DS got drawings for Roman Pots, and is now performing calculations for Roman pots in OUT position. He has also implemented the collimator taper in GdfidL. Warner has installed a new version of GdfidL, which incorporates a trick from O. Napoly to facilitate the calculation of collimator impedances. DS also asked for the capability to compute the wake of a tilted trajectory as well as for triangular bunch shapes.
ACTION -> AK will repeat simulations with a constant injection offset.
Still to be done. No problem anticipated.
ACTION -> JJ will explain how to produce eps plots with Mathematica.
JJ explains that the mathematica default includes the exportation of fonts. If the default is not properly set, it can be adjusted. The procedure is explained in the mathematica help pages.
ACTION -> FR will read and approve the draft note by AK.
DONE
AK reports that following a suggestion by BZ he changed the sign convention from
A. Chao to the 'normal' convention. This introduced an additional term in the Fourier transform. Therefore, the earlier and the new calculation appear to be inconsistent.
AK is also collaborating with BZ on the impedance of a multi-layered wall. The first example is a think metallic wall surrounded by vacuum. There may also be a sign problem here.
AK reports that he encountered problems installing Mathematica5, in which important bugs have been fixed. FR also had the same problem. On the other hand, JJ has successfully installed Mathematica5 on various computers.
(2) Unified Instability Threshold (EM)
----------------------------------------------
EM explained that various different instability theories can be cast into the same form, which can be derived from the coating beam limit. The SPS calculations with MOSES by E. Shaposhnikova nicely agree with this universal formula, except for a numerical factor of order 3 which could be related to differences in the longitudinal profile and the definition of the bunch length. In EM's formula the instability threshold scales with the longitudinal emittance and not with the synchrotron tune. He recommended that this dependence be checked with the HEADTAIL code.
In addition to the dependence on the longitudinal emittance, the other notable dependence is that on chromaticity. Large chromaticity increases the threshold.
(3) Collimator Impedance Calculation for 9 Options (EM)
---------------------------------------------------------------------
EM presented impedance and stability boundary calculations for various collimator options at injection and top energy, in both transverse planes. Later options improve the stability behavior and move the operating point closer towards the stable region. In practice, one will always operate the LHC close to the limit of stability. Thus it was suggested that collimation efficiency studies be performed for this situation.
EM showed that the collimator impedance also can cause beam instability at injection, when the feedback is turned off and the octupoles are turned on to the maximum value that is compatible with an acceptable dynamic aperture.
Rise times are nearly identical between LV's and Em's calculation, but at injection there is a factor of 2 difference in the tune shift. At top energy there is a nearly perfect agreement, EM's numbers being slightly more conservative.
(4) Space Charge and Octupoles (EM)
EM presented calculations of the tune spread due to space charge for a self-consistent parabolic distribution and an approximation with a linear dependence on Jx and Jy. He showed 3-dim., 2-dim, and 1-dim projections of the tune distributions. In the future, he plans to add the longitudinal degree of freedom.
(5) Electron-Cloud Simulations for Variable Bunch Spacing (DS)
--------------------------------------------------------------------------------
DS finds a factor of 10 difference in the average electron flux for 25-ns and 75-ns spacing over 3 batches. The observed ratio is 2. Since the mean energy is not the same, the residual difference could be due to the bias voltage. DA considered rms bunch lengths of 2 ns and 1 ns. The 2 ns number came from MJ. GA remarked that the bunch length was 1.1 ns for the 75 ns spacing and 0.9 ns for 25 ns. DS reported that the flux ratio is similar for the 2 bunch spacings he simulated. DS will look at the stripe position next.
One surprising result was that without magnetic field the larger bunch spacing was worse.
FR suggested producing a plot of build-up time (without inter-batch gaps) versus bunch length for different values of the yield.
FR also asked whether the ECLOUD code is stable and be frozen for 3 months. DA anwers there are no known serious bugs at the moment.
Another question raised is whether we can extract the low-energy reflectivity by comparing measurements and simulations.
The dependence on the magnetic field is qualitatively consistent with the measurements, but the 'free region' at low field differs by a factor of 2 (DS).
FR recommended a publication of the ECLOUD history.
With the code frozen, a comprehensive comparison of SPS measurements and simulations should be performed. GA will help in collecting relevant beam parameters.
(6) RF Voltage Modulation in the SPS (EV)
Last Monday's MD of amplitude modulation was successful. The beam was stabilized in the same way as it is by the higher harmonic rf in bunch shortening mode. This procedure might alleviate kicker heating problems in the SPS and is of potential interest for the LHC, where neither higher harmonic cavities nor feedback is available.