------------------------------------------------------------------
Minutes of the ABP-RLC team meeting of 07.04.2006
present: AG, EM, WH, TP, FR, GR, RT, Tom Kroyer, Heiko Damerau
excused: DS and FZ are in Turin for the defense of EB's PhD thesis
web site: http://ab-abp-rlc.web.cern.ch/ab-abp-rlc/
------------------------------------------------------------------
(1) Minutes of last meeting, pending actions: see the following
---------------------------------------------------------------
(2) Progress report on US-LARP Beam-Beam studies (WH)
-----------------------------------------------------
Working with M. Furmann, Ji Qiang (LBL) is using his strong-strong
beam-beam simulation code based on the concept of displaced
Green's function to study the emittance growth in case of
beam-beam interactions with a static offset at the central
interaction point. Collisions in IR 1 and 5 are considered. He
studied 3 cases: head-on collisions with either horizontal or
vertical offset, and a third case with head-on as well as lumped
long range interactions. He evaluated beam-beam offsets between
0.1 and 0.4 sigma. In his preliminary results for the head-on
cases he finds emittance growth in the plane of the offset which
increase with increased separation. Both beams behave similarly.
At present, it is studied whether this growth is real. In case it
is confirmed, it is important to minimize the offset of the two
beams at collision.
The case with long range interactions is less clear and further
studies are presently done to understand the results.
FR: Is this emittance growth an effect of numerical noise? Similar
simulations run by S. Krishnagopal showed no such effect.
WH: The fact that the same behavior is observed for both beams
(which have different random initializations) suggests that the
effect is not purely numerical.
FR: What are the practical implications for the LHC?
WH: The offsets must be compensated as much as possible (tolerable
values could be < 0.2 sigma), because as the separation gets to
values close to 0.4 sigma the emittance increase is significant.
The offset of PACMAN bunches with compensated orbit effects by
alternating H/V crossing planes is about 0.1 sigma, but without
compensation it easily gets to ~0.5 sigma.
FR: Do they correct beam-beam offsets at the TEVATRON?
WH: They don't have PACMAN bunches and have many other problems
before coming to the emittance growth caused by offsets.
(3) Increasing the TMCI threshold in the SPS by linear coupling
---------------------------------------------------------------
EM reported interesting HEADTAIL simulations about a possible
increase of the TMCI intensity threshold at injection in the SPS
by linear coupling. Working near the coupling resonance should
allow an increase of the intensity threshold by 36%. The
simulations assume a broad-band impedance of 20 MOhm/m at 1 GHz
with Q=1, and a single bunch with longitudinal emittance of 0.2 eVs
(instead of 0.3 eVs for the nominal LHC beam).
Using that: 1) thresholds in the horizontal and vertical planes
are equal in a round geometry 2) threshold in the vertical plane
is half of the one in the horizontal plane in a flat geometry (and
it is 12/pi^2 the one of a round geometry), one can think of
introducing linear coupling between the transverse planes in order
to increase the threshold in the vertical plane for a flat
geometry. Simulations have been done with HEADTAIL using the
parameters of a single bunch with low longitudinal emittance and
Qx=26.180, Qy=26.185, and an integrated skew quad strength
K2=0.005 m^-1 (ideally the tunes should be related by
Qx+DeltaQx,imp = Qy+DeltaQy,imp). The simulated threshold in the
vertical plane goes from 3.3x10^10 without coupling to 4.5x10^10
with coupling. The increase is by 36%, that is comparable to the
48% current increase required to go from the LHC nominal intensity
to the ultimate intensity. Linear coupling would also help for the
e-cloud instability because, since the e-cloud is mainly located
in the dipoles, it causes a vertical single-bunch instability.
This could allow running the SPS with lower chromaticity, which is
beneficial for the beam life time. Introducing the ratio between
vertical and horizontal effective impedances lambda=Zy/Zx, the
gain given by linear coupling can be written as:
gain=2 lambda/(lambda+1). This formula gives 39% in the case
simulated with HEADTAIL, which is very close to the 36%
given by simulations.
FR: Do results depend on the tune split?
EM: Yes, because he also tried to improve the vertical threshold
using linear coupling and Qx=26.18 Qy=26.13 and this did not work.
AG: The same effect as linear coupling could be achieved by
introducing in the beam pipe an object with a negative vertical
impedance. This will of course enhance the horizontal impedance
but the overall situation for the beam improves because the total
impedance gets reduced in the vertical plane, which is the
critical plane (it has the lowest TMCI threshold).
FR: Tests in the SPS should be done soon! This will require a
change of working point.
(4) Report about FP420 meeting (EM)
-----------------------------------
FP420 is a new experiment not yet approved, located at 420 m from
IP1 and IP5, which requires Roman Pots or a 8 m long movable pipe,
1-sided, copper, very close to the beam (3 mm). Message given from
the impedance point of view: experience from collimators shows
that one should be very careful introducing such objects, because
they basically contribute to all degrading effects: resistive
wall, geometrical effects, trapped modes, high contributions to
the broad-band impedance. Roughly adding the moving pipe would be
like adding 5-10 collimators, which would put the coherent tunes
far off the stability region.
(5) Follow-up of TCTV and TCLI collimator impedance (EM + AG & FR)
------------------------------------------------------------------
There are 4 TCTH + 4 TCTV (2 of them in a 2-beam region) and 2
TCLI (1 in a 1-beam region and the other in a 2-beam region). The
three devices in 2-beam regions give a quite high contribution to
the broad-band impedance of the machine. Trapped modes are to be
checked. The criterion for these objects is that they should
contribute to the broad band impedance by no more than 1/100 of
the full impedance budget (that means they should not exceed ~0.02
MOhm/m). This translates into a gap of 12 mm for the TCTVs for
nominal beta (~70 m). But squeezing in IP2 or IP8, beta goes to
~660 m, therefore the effective impedance would be multiplied by a
factor 10, which would require a gap of 30 mm to still meet the
requirement on the broad band impedance. Actually both these gap
values are too high because, given their position, with gaps > 8
mm the TCTVs become useless to protect the triplet magnets. FR
points out that one should consider also the beam intensity when
doing these estimations. The TCLIB should also have a gap of 12 mm
following the same criterion as above. This is compatible with the
protection of the arc.
==> ACTION: EM will put all contributions to the broad band
impedance in the stability diagram both at injection and at top
energy to see where we are actually standing with all the objects
so far examined.
Concerning the TCDS functional specifications, AG gave FR a paper
copy of the material shown at the RLC team meeting of 10.06.2005.
Many of the options discussed and investigated are not mentioned
in the functional specification and it would be desirable to add a
single reference to the results of AG.
==> ACTION: AG will write a short note on the TCDS impedance
estimates that can be referenced in the Functional Specifications.
FR informs the team that the impedance of the copper coated
injection septa in IR2 and IR8 (MSIA and MSIB, made out of
mu-metal, 4 m long magnets, in groups of 5) and of the dump septa
in IR6 (15 MSD, 4 m long magnets) should be re-evaluated, even
though it should be negligible according to the LHC Design Report.
After the meeting Stephane Fartoukh informed FR and EM that there
are several transverse steps of the vacuum chamber in
correspondence with these septa.
==> ACTION: EM will follow-up the impedance of LHC septa
FR, AG, and Ralph Assmann discussed the collimators RF fingers.
Apparently many of them come out deformed or broken from
production and should therefore be cut, but a number is needed on
how many could be cut without affecting too much the impedance of
the collimators. AG says it mainly depends on the length of the
fingers, but in principle no more than 1-2 in a row should be
missing to avoid a significant increase of the coupling to the
collimator tank. After the meeting FR sent the following email to
the LHC collimation team in connection with RF fingers with
bad/missing contacts:
Dear Ralph,
I summarize below the conclusions of the LHC impedance team for
the collimator RF fingers (~5 mm long, ~2 mm wide with ~1 mm
spacing):
- all RF fingers are functionally important and should not be
considered as an optional esthetic ornament of the collimators...
- we may exceptionally accept ONE missing finger or maybe TWO
provided they are not consecutive. With two or more missing
fingers in a row, the beam would "see" the outer tank and excite
low-frequency modes very dangerous for the transverse beam
stability. Missing "longitudinal" RF fingers would lead to
potentially large heat deposition in the collimator tank.
- the contact force of all RF fingers (both longitudinal and
transverse) should be around 50 g/finger. Apparently this is not
included in the functional specifications of the collimators, but
is assumed to be the result of a correct mechanical assembly.
- since the collimators will be activated in the LHC, we better
fix this problem now, before installation in the machine. Of
course I hope that the problem can be solved quickly.
(6) Follow-up of CARE-HHH CERN-GSI meeting
------------------------------------------
FR shortly reports about the recent bilateral GSI-CERN meeting of
last week at GSI-Darmstadt (attended by FR, EM, GR, FZ, Gianluigi
Arduini, and Elena Shaposhnikova). Many subjects were discussed
ranging from incoherent effects due to electron cloud (emittance
growth induced by periodic resonance crossing) to cross-check of
different analytical formulae for the resistive wall impedance to
TMCI and benchmarking of different codes for collective effects
including space charge and impedances. All the talks are available
on the web pages
http://care-hhh.web.cern.ch/CARE-HHH/Collective%20Effects-GSI-March-2006/
http://www-linux.gsi.de/~boine/CARE/HHH-meeting.htm
Summaries and conclusions of the two-days discussion will be
available soon.
(7) AOB
-------
FR has created at web page linked to the RLC web site with a list
of references on Accelerator Physics. This list is far from
complete and is intended mainly for future students who may need
basic (online) references to start working in the RLC team. See
http://ab-abp-rlc.web.cern.ch/ab-abp-rlc/ap-literature.htm
Posted on the web: Slides by EM
Web site: http://ab-abp-rlc.web.cern.ch/ab-abp-rlc/