Minutes of the ABP-LCE team meeting on 03.10.03
present: EB, TdA, WH, AK, EM, FR, DS, EV, FZ
------------------------------------------------------------
(1) Actions and Comments on Minutes
-----------------------------------
EV had some comments on the last minutes, which
may already have been partly included in the meeting
announcement sent out by Francesco. He pointed out
that there is no strong connection between his approach
and the feedback at the B factories.
old ACTION: EM will further check the space charge detuning.
The previous solution was not self-consistent, the
stability diagram was computed using a quasi-parabolic
distribution and the incoherent space charge tune shift
assuming a parabolic distribution in the transverse
spatial coordinates. Following a recommendation by
FR, EM has redone the space-charge derivation by computing
the projection of the quasi-parabolic distribution onto
the two spatial coordinates. From this, he then derived the
incoherent tune shift and tune spread. There will be some
slight, but not decisive difference to the previous result.
10 double integrals are still left to compute, which
can possibly be done in one afternoon.
ACTION -> compute double integrals (EM)
It was debated how to compute the coherent tune shift
vector, which can be superimposed on the stability
diagram with space charge. EM suggested that as a
first guess one could take the tune shift
at the center. FR recalled that the standard way
to compute the tune shift is to solve the Vlasov equation,
in this case with space charge, similar to what Y. Alexahin
did for the beam-beam interaction, and to derive the
relevant 'Yokoya factors'.
EM emphasizes that space charge is not always stabilizing,
even if the good sign of the octupoles is chosen.
ACTION -> Automatic Excel spreadsheet for collimator
impedance (EM+LV)
Still to be done.
(2) Early status report on LHC resonator impedance,
kickers, BPM's, and cold-warm transitions (FZ)
---------------------------------------------------
The status and literature for the various components
were reviewed:
200-MHz cavities - no data on damped dipole modes
400-MHz cavities - R and R/Q values in D. Angal's paper
without source; otherwise data
for damped and undamped monopole and
dipole modes
transverse damper - dipole mode data missing
experimental chambers - MAFIA calculations by Y. Luo to
be recovered, data only for CMS;
trapped mode at Y chamber
injection kickers - formulae by G. Lambertson;
ceramic-coating theory & experiments?
dump kickers - no information found
BPM - paper by L. Vos and A. Wagner, complete!
cold-warm transitions - resistive part with coating & inductive
bypass, requires to know Luc's theory;
also geometric impedances
(tapers, rf junction - references?)
quadrupolar impedance - geometric wake from Stupakov, resistive
wake from Yokoya
FZ concluded that for roughly 50% of the elements a complete
description of the impedance is available; for some others he
has found little information so far.
FR suggested to contact B. Spataro and Y. Luo for their MAFIA
calculations and related input/output files. Also H. Tsutsui,
B. Zotter, and R. Gluckstern may have computed kicker impedances.
(3) Tensor Transformation (FR)
-------------------------------
FR derived the tensor transformation needed, e.g., to
compute the impedance for a flat collimator rotated
by 45 degree. The impedance must be weighted with the
correct product of square roots of beta functions and
multiplied with coefficients 3/4 (same plane) or
1/4 (orthogonal plane).
(4) Collimator Wake Calculations using GdfidL (DS)
---------------------------------------------------
DS reports that computations with resistivity
do not give the expected equality of real and
imaginary impedance, the imaginary impedance
decreasing much more steeply for increasing
frequency.
The geometric wake was computed for two example
structures. The first case was for a round
geometry with entrance and exit taper. The
longitudinal wake is the same as that obtained
by MAFIA.
The second structure was a 'brick' collimator
inside a round chamber, with a tapered transition
at 10 degree. The wake looks like the shape of
the bunch, and not inductive. At longer times,
behind the exciting bunch the computed wake grows
continuously. Is this the sign of a trapped mode,
or an artifact of the code?
The real impedance for this case changes sign
at some frequency, which does not look physical
and could be an edge effect.
ACTION -> check sign of real part of impedance (DS)
FR suggested to define a criterion for concluding
this study. FZ proposed that quadrupole wakes
also be calculated for the collimators (with
offset test particle).