-----------------------------------------------------------------------
Minutes of the ABP-RLC section meeting of 11.02.05
present: EB, AG, WH, EM, FR, DS, EV, FZ
special guest: Bruno Zotter
-----------------------------------------------------------------------
(1) Minutes of last meeting, pending actions, future presentations
-------------------------------------------------------------------
Due to lack of time, 4 presentations are moved to the next meeting on February 18, namely
- Damping of injection oscillations at the PS (EV)
- New results of e-cloud simulations for nominal LHC, ultimate LHC, and 2 upgrade scenarios (FZ)
- Beam-Beam Studies (WH and TP)
FR reported on a meeting with JM Jimenez and FZ on the LHC flanges/gaskets.
Their total number is about 1000 per ring, representing a broadband impedance of about 10%
of the total, longitudinally and transversely. It was agreed that the gaskets will be modified
for most of these flanges, so as to fill the deep cavities that otherwise existed and to reduce
the impedance. JMJ will issue an ECR for this change. Exempt from the modification are elliptical
chambers (risk of accidental aperture limitation),
which constitute about 10% of the total and flanges with welds on the side of the beam pipe,
which are less than 10%. With these modifications, the total broadband impedance from the flanges
will be about 1-2% of the total, which is acceptable.
=> ACTION => compare the size of the longitudinal geometric wake with RW longitudinal
wake from A. Koschik (FR)
PENDING.
(2) LHC collimator impedance:
-----------------------------
A) resistive-wall impedance from field matching (BZ)
------------------------------------------------------
BZ first reviewed the history of resistive-wall wake field calculations, the earliest
one being the paper by Laslett, Neill and Sessler in 1965 and a 1976 Erice lecture by
Sacherer, where an effect of the finite wall thickness was introduced. E. Keil and
B. Zotter developed the LAWAT model, based on field matching, which initially considered
surface charges and currents as source, and was applied to the ISR and many other machines.
The matrix formalism of LAWAT takes into account the redistribution and inductive bypass
effects for homogeneous structures of infinite length.
More recently the formalism has been changed starting from a different source term, as used
by A. Chao and R. Gluckstern. This has given rise to the LAWAT2000 version, which exists in
mathematica format. BZ commented on developments by Burov and Lebedev, which starts with some
approximations (no scalar potential, vector potential purely longitudinal) by which some
physics can be lost, and by L. Vos (where the inductance of free-space chamber of mu_0/(4 pi)
is introduced without a clear justification). BZ described in detail all derivations, based
on the general solution of the vector Helmholtz equation and field matching, leading to the
final expression for the impedance, which is valid for all beam energies (even in the
non-relativistic case). As mentioned by EM in the last meeting, for the collimator, the
result agrees with Burov-Lebedev's, for the kicker there is a difference of ~20%. BZ pointed
out that it is better to derive the general case first, and then make approximations, instead
of starting with approximations.
Relevant for LHC are slow waves with velocities much less than the speed of light.
In the present LAWAT2000 formalism it is assumed that source and wave propagate at the
same speed. An interesting case for LHC would be a wave at beta~0.04 and a relativistic
bunched beam, worked on by BZ and Bob Gluckstern (and independently also by Stephane Fartoukh).
Namely, the formalism can be extended to different betas of wave and source, e.g., using several
charge rings, as it was the case in the original LAWAT. The finite-length effect is another question,
which is being worked on by BZ and Bob Gluckstern.
=> ACTION: EM and BZ will continue the development for waves with beta << 1.
FR suggested a more detailed comparison between measured and expected SPS collimator-induced
tune shift, by gathering all pertinent information from the participating groups. It was proposed
that changes in beam intensity or bunch length could explain the difference with regard to the
theoretical expectation. FZ reported a comment by Gianluigi Arduini that the tune spread appeared
to increase when the collimator jaws were closed. This was presented by Marek Gasior in the last
meeting of the Collimation Working Group.
=> ACTION: confirm bunch length, intensity, and collimator gaps during tune-shift measurement (FZ).
(B) trapped modes and multi-bunch instabilities (EM)
------------------------------------------------------
EM reviewed the effects of trapped modes for the SPS & LHC collimators. All previous results
had to be divided by a factor of 2, due to a difference in the definition of shunt impedances
for linacs and rings. The expected most unstable mode wass calculated using AG's parameters
for the 13 strongest trapped modes. For a single collimator at 400 m beta function, the rise
time is 155 s (m=4) with a Gaussian profile and 110 s with a parabolic profile (m=2) in LHC
at top energy. If contributions from the other 18 collimators are added, one may expect growth
rates between 20 and 100 s.
(C) impedance of trapped modes for smaller collimator gap and single bunch
tune shift in the SPS (AG)
-------------------------------------------------------------------------------------------------------
AG explained why a confusion had arisen on the definition of the shunt impedance. The linac shunt
impedance is 2 times the storage-ring shunt impedance. He computes the kick factor in two different
ways, which offers a numerical check of the precision. Following a previous suggestion by DS he has
repeated the calculation of trapped transverse modes for a smaller full gap (2.5 mm instead of 5 mm).
For many gap modes the shunt impedance increased by a factor 3-4, while the Q Value decreased by a
factor 2, yielding an overall factor 2 increase in the net effect. As shown by EM, the most dangerous
modes seem to be transition modes at about 1.2 GHz, with the gap modes at 1.5-2.0 GHz being slightly
less important. AG also computed tune shifts for higher-order head-tail modes. Perhaps a factor 1/(m+1)
was missing where m is the mode order. An impedance comparison of GdfidL and HFSS over the range
0 to 2 GHz shows excellent agreement. It was decided that a detailed study of modes above 2 GHz by HFSS
is not needed, as these modes will certainly be less important than the modes already included.
FR recommended that future plots should carry a horizontal axis in units of Hz.
Next week, AG will report to the Collimation Project on the topic of rf fingers.
(3) Decoherence with and without space charge in the PS (EB)
------------------------------------------------------------
Alfred Blas has observed that several-ms decoherence times observed in the PS are 10-100 times longer
than predicted by simple formulae for the effect of chromaticity or space-charge tune spread. The long
decoherence time would relax requirements for the transverse damper system, for which the TOTEM beam is
the biggest challenge. Collaborating with A. Blas, EM and FZ, EB is presently performing simulations
with the HEADTAIL code. First simulations include the large chromaticity xi=-1 (first and optionally
second order), but not yet the broadband impedance. Some simulations were done with and some without space
charge, in order to observe its effect. Without space charge the simulated bunch centroid oscillation, after
a 15-mm kick, decays within less than 200 microseconds. With space charge, there is no decay over 1 ms.
Emittance growth his also greatly reduced over the same time scale. This strong effect of space charge
is qualitatively consistent with the experimental observations. Snap shots of phase space distributions
with and without space charge reveal that in the former case the distribution quickly filaments into a
'donut' shape, while the space-charge force has the effect of keeping the bunch together and preventing
filamentation. A next set of simulations will include the machine impedance as well.
There was a concern about the space-charge tune-shift value used in the HEADTAIL simulation. It is 0.35
for a Gaussian bunch (the 0.7 quoted by the code and mentioned in the meeting was not the actual tune shift).
According to EM it should be about 0.2 for a parabolic bunch. The two numbers are approximately consistent.
EV asked whether the feedback can be included in the HEADTAIL code. At the moment HEADTAIL contains
a feedback model with an exponential damping plus noise. DS remarked that this model could be extended
to take into account reduced gain along the bunch, if the time-response function of the feedback system is
known.
EM offered presenting an analytical calculation of the space-charge decoherence, based on a paper
by L. Vos, possibly in the next meeting.
(4) First discussion on beam physics tools, instrumentation, and software requirements
for the LHC commissioning (FR+ALL)
-------------------------------------------------------------------------------------------------------------------------
FR asked all members of the RLC team to draft descriptions of beam-physics tools, instruments and
procedures which are needed for LHC beam commissioning. As an example he mentioned the modulation of
long-range beam-beam collisions at injection by a betatron mismatch. The question each team member should
try to answer is how to go from a single stored proton over half the nominal intensity, and the nominal
to the ultimate, without destroying the machine. A catalogue of topics/scenarios should be identified, ranging
from injection to collision. Specifications for most of the instrumentation are available on the web. The
goal is to issue a common document which summarizes the requirements for each stage of the commissioning.
In particular, it should be spelt out which software is needed.
FR recommended to contact outside groups for input. In more detail, emphasis should be on the procedures required
to measure and correct suitable observables related to high intensity LHC operation without damaging the machine.
The main contribution of the RLC team could be to identify integrated scenarios such as beam scrubbing or
luminosity tuning, and describe in a logical manner the required set of measurements, data flow, and correction
strategies leading to LHC-specific algorithms and possibly to the associated software specs/implementation.
For example, another such scenario could be:
"Switching off the damper during the ramp and use of the Landau octupoles to stabilize the beam".
The logical steps for this may include:
- at what energy can/should we switch off the damper?
- damper noise -> emittance growth -> when and how do we measure it?
- what is the strategy for Q and Q' control during the ramp?
- can we measure Q and Q-spread with damper on? -> check with BDI+RF group
- how do we measure tune spread and decoherence during the ramp?
- can we excite a single or a few bunches?
- what is the excitation level compatible with machine protection?
- at what energy do we need to switch on the Landau octupoles?
- the space charge tune spread decreases as 1/gamma^2
- do we close the collimators during the ramp? -> check with Ralph/Rudiger
- if yes -> the impedance increases -> we need more Landau damping
- can we use the octupoles to stabilize the beam in case of slow beam losses?
- software specs for BLM readings/archiving -> post-mortem analysis
- how do we disentangle losses from beta-beating or slow instabilities
- do we need to monitor beam losses on a bunch-by-bunch basis?
Related issues are:
- what beam intensity/luminosity can we reach without using the octupoles
- at what beam intensity do we need a PLL for Q control?
- at what beam intensity do we need a closed loop for Q-spread control
- do we need a closed loop for Q' control?
- can we give arguments/constraints/specs for a combined ramp&squeeze?
BZ commented that the RLC team might want to limit its proposals to the study of beam oscillations, bunch length,
and emittance. FR responded that many issues are intertwined with optics, for example the detuning with amplitude.
FR also mentioned that the idea of inter-group mini-teams has been revived by H. Schmickler. Progress on this issue
will be discussed in two or three weeks.
Attached: Slides by BZ, EM, AG, EB