Summary notes of the forty-fourth meeting of
the LHC Commissioning Working Group
Tuesday April 22nd, 14:00
CCC conference room 874/1-011
Minutes of the Previous Meeting and Matters
Arising
There were no comments on the
minutes of the 43rd meeting. Roger announced that prior to
discussing the three main points on the present meeting’s agenda (5 TeV parameters, cold check out, and dry run), a few open issues from previous meetings would be followed
up, namely the integration of FiDeL into LSA by Mike,
and several collimation questions by Ralph.
FiDeL Data for LSA (Mike)
Mike reported that the FiDeL implementation for LSA is being followed up
with high priority. To illustrate the progress, he showed some pages from
the FiDeL
web site, which features a work plan,
documentation of meetings, and an updated overall work status.
Weekly meetings are organized by Luca Bottura, who will be the point of contact until the end
of May.
Collimation Issues (Ralph)
Ralph presented an update on the
collimators, addressing several questions raised by Roger. He announced that the installation of the collimation is now
complete: 88 are presently installed, after 4 had to be taken out again.
A delay was encountered for the four TCTVBs in IR2 and IR8. These are
special vertical collimators for regions with two beams, with very large
elliptical beam pipes. They were contaminated
during production at CERN, and need to be cut open and cleaned. CERN was
supposed to produce a total of 15 collimators in house, 9 of which were for
protecting warm magnets and are OK. The production of the remaining two
collimators of type TCLIA were stopped, after gas composition measurements had
revealed the contamination of the TCTVBs. Ralph recalled the function of the
TCTVB collimators: vertical triplet protection and vertical background control.
He emphasized that these collimators will be needed only for beta*<6 m in IR2 and 8. Therefore, despite of
this problem, the collimation system is
ready for 450 GeV and 5 TeV
beams, without any limit on the beam intensity, and with no impact on the
optics in IP1 and 5. Thomas Weiler is presently
reviewing the beta*=6 m limit for IR8. Ralph presented the previously
calculated aperture in IR1 as a function of beta*, suggesting that the
dependence in IR2 and 8 would be similar. A proposed crash effort could have
replaced the missing collimators by the fall, but it was incompatible with the
general LHC planning. New collimators will be installed after the 2008 run.
A non-critical non-conformity was encountered in the TCTVA.4L5. Collimator production and installation
was and is affected by non-conformities, that are
solved 1 by 1 and documented in MTF. The primary requirement is that the
collimator must function as specified, e.g. the measured torque needed to move
the jaw must comply with the specifications. If the torque is found to be OK
and if all involved engineers approve, the collimator is accepted. A final
visual inspection of TCTVA.4L5 after transport into the tunnel revealed a
displaced roller. In total each collimator is equipped with 18 rollers per axis
by design. A reduced number of rollers, down to a minimum acceptable number of
14, was approved by TS mechanical engineering based on
the weight supported by each axis. A spare collimator would also be available
if needed. If torque measurements in the tunnel reveal no problem the
collimator will be used as it is. Alick asked whether
the missing roller was located at the end or in the middle of the collimator.
Ralph replied that it was a roller at the very end, but that dedicated shock
monitors had recorded no shock outside the tolerance during transport, so the
origin of the problem is not yet understood (no problem during visual
inspection on the surface).
Ralph mentioned the recently started
urgent study on implementing a temporary
betatron cleaning scheme in IR3 for the 2008 run.
This is motivated by the installation of unsuitable electronics in the
radiation zones of IR7, which was presented by Thijs Wijnands at the last ICC
meeting. Ralph highlighted that the collimators are designed to intercept the
beam and thereby to create radiation. No proper reshuffling was possible to
protect the electronics. The fallback solution therefore is to arrange for betatron cleaning also in IR3. During the run we can then
choose to perform betatron cleaning either in IR7 or
IR3. In IR3 there is a factor 200 less radiation for the same beam loss.
Responding to a question, Ralph clarified that the proposed betatron
cleaning in IR3 would imply a coupled betatron-momentum
cleaning.
Ralph showed recent results from
collimator hardware commissioning, where 14
collimator jaws in IR3 were remotely controlled from the CCC and monitored by 28 independent position
sensors, while exectuting the collimator ramp
functions. The results demonstrate that the specified accuracy of collimator position control is achieved. Observed
differences can mostly be explained by the 20 micron mechanical play that had
been specified, and a change of direction in the collimator-jaw motion. The
error decreases for the tighter settings, which reflects that the calibration
was optimized for high accuracy at small gaps. Jan asked about the interlock
level. Ralph answered that he expects the interlock level at 7 TeV to be set to about +/-50 micron (roughly 1/4 sigma). It
will be less strict at injection energy and for the larger collimator gaps.
5-TeV Stage-A Parameter List (Massimo)
Massimo reviewed the parameter list
for 5 TeV, following up his previous presentation on
“Beam
Parameters for 5-TeV Operation in 2008” from the 43th
LHCCWG. There a proposal had been to redefine
the target beta* for 5 TeV to 3 m, in order to
leave the same aperture margin as had previously been foreseen for 7 TeV, which has now been implemented. For IR3 and 8 the
minimum beta* limit is taken to be 6 m
imposed by the missing TCTVBs, which is also taken as the target value for IP8. IP2 should stay at 10 m.
Longitudinal beam parameters still had to be reviewed, and were presented this
time. Massimo recalled that the blow-up hardware would likely not be available
from day 1. Three longitudinal emittance values were
therefore considered: 0.5 eVs (SPS natural value at
low intensity), 1 eVs (with blow up in the SPS and
injection mismatch), and 2.5 eVs (nominal, with blow
up in the LHC). Massimo presented the momentum spread and bunch length as a
function of energy for a constant rf
voltage of 16 MV. The nominal rf
voltage at injection is 8 MV.
Now turning to a question that had
been posed by Oliver, Massimo showed that the synchrotron frequency crosses 50
Hz at about 800 GeV and 1.6 TeV
for 8 and 16 MV rf voltage,
respectively. No harmful effects from such crossing are expected for the
initial run at 5 TeV.
Massimo reminded the team that the
longitudinal damping will be lost if the longitudinal emittance
is too small. Resulting instability thresholds were provided by Elias Metral, who considered the latest estimate of the expected
value of the LHC impedance (previously, the longitudinal beam parameters were
studied and reported by E. Shaposhnikova, see LHC
PN 242). For the 0.5 eVs emittance
the intensity threshold is 2.2e10 per bunch, which is too low to be acceptable
even for phase A. For an emittance of 1 eVs the threshold is already 1.3e11, higher than the
nominal LHC intensity, which would be perfectly acceptable. Accordingly,
Massimo proposed the 1 eVs value as candidate target
number for the 5 TeV run. This value also implies a
luminous region smaller than nominal, and a lower momentum spread which is good
in view of the mechanical aperture. No issue for machine protection is
expected, since the geometric emittances are larger
at 5 TeV than at 7 TeV.
Massimo also commented that the reason why we do not want to use 1 eVs as target emittance for 7 TeV is that the intensity threshold depends on the bunch
length, and at 7 TeV the threshold for 1 eVs would correspond exactly to an intensity of 9e10
protons per bunch, as foreseen for commissioning phase A, and less than the
nominal one.
Synchrotron radiation effects were
also reevaluated for the lower energy. A table compiled by John Jowett
summarized the situation. The transverse and longitudinal emittance
damping times at 7 TeV are 26 h and 13 h, respectively.
At 5 TeV these values increase significantly.
Radiation damping rates were compared with IBS growth rates for an intensity of
9e10 particles per bunch, as a function of longitudinal emittance,
and in particular for 1 eVs and 2.5 eVs. In the transverse plane the radiation damping always
balances or surpasses the growth rates. However, at 5 TeV with 1 eVs
longitudinal emittance, in the longitudinal plane the
IBS growth time (20 h) is about half the damping time (35 h).
Massimo next compared the integrated
luminosity distribution for 5 TeV with 1 eVs with the nominal case. The luminous region is roughly
30% shorter than nominal.
Finally, Massimo presented the
overall target parameters for 43 bunches, 156 bunches, 75 ns spacing, 50 ns
spacing, and 25 ns spacing. He mentioned that a note summarizing the various
filling options is in preparation by W. Herr, M. Ferro-Luzzi
and T. Pieloni. Different filling patterns allow
varying the numbers of bunches colliding at the various IPs. For 43 bunches,
the idea was to maximize IP8 luminosity and to minimize the IP2 luminosity. The
50 ns scheme also included in the table appears attractive. An extended phase-A
parameter table now features a 5-TeV central column. Crossing scheme details
were added, listing both internal and external angles (the latter being 0 in
phase A). Several alternative collision schedules are indicated as well. The
luminosity at each IP and the number of events per crossing are shown as well.
The tables will be stored on the LHCCWG web site.
Massimo’s main conclusions were: for 5 TeV
operation a longitudinal emittance of 1 eVs is OK and a longitudinal blow up in the LHC is not
needed.
John commented that for an emittance of 0.5 eVs the IBS
effects would be much stronger. Gianluigi suggested that
the IBS blow up could be self-stabilizing. Massimiliano
asked whether it was indeed possible to have 1 eVs in
the LHC for 2008. Massimo replied that, yes, 0.8 eVs
had already been demonstrated from the SPS. Massimiliano
remarked that perhaps some experiments might prefer to have a larger luminous
region. He recommended to preferably choose the
nominal bunch lengths for a later 7 TeV run, if there
is no reason from the machine side calling for shorter bunches. John pointed
out that the 1 eVs emittance
could be better for limiting debunching. Massimiliano proposed to discuss these parameters at the
LPC. A remaining open issue is the number of bunches for ALICE and how they
will be distributed. Massimo remarked that the detailed collision schedule is
included in the tables.
Preparation for the Machine Check-Out - Status (Gianluigi)
Gianluigi represented a team of four people, which was looking at the
preparation of the machine check-out. The other three members were Reyes, Rossano, and Mike. The presentations at the Extended LTC were used as a
starting point. The team already identified the machine check-out contacts,
i.e. the relevant people “in the field”.
A first round of discussions collected pertinent information on the
planned HW/SW availability, status, requirements for
the commissioning time, status of CCC tools, possible dates for preparation
tests, interface between systems, and the scheme for standard cold check out.
The information flow is still slow, since many people are busy finalizing their
hardware. Gianluigi presented the systems table with
contact names (slides 3 and 4). He next reported on the various systems one by
one.
General issues: Verification of
the UPS connections is part of HWC, but it was not done for sectors 4-5 and 5-6 so
far. The tests would have an impact on the cryogenics and the organization.
Ralph wanted to know how these tests are done. Gianluigi
replied that emergency stops are to be pushed. Roger questioned whether these
tests had been done anywhere. Gianluigi answered that
they had partially been performed in Sector 7-8. Responding to a question by
Ralph, Rossano clarified that several problems were
found during these tests in 7-8, e.g. devices that were not connected. Roger
assumed that this type of tests had been skipped in the following sectors
because of their impact on the commissioning programme.
Jean-Jacques remarked that it was known that several components were not
connected to the UPS. Ralph cautioned that we should not compromise the safety of the machine without any official statement
or clear decision. Paul commented that this issue was on the list of follow ups
from the e-LTC to be discussed at the LTC the following day. Other general
issues are the status of the compensators and behavior at full load, especially
in points 2 and 4, the post mortem, which was not really addressed yet, the
logging and alarms, and a few more.
Verena asked how the philosophy differed
from that of the dry runs. Gianluigi imagined
verifying and including all the other still missing systems like vacuum etc.
Magnet training was done anyhow only in the evening. Special items would be
looked at, like the cryo system. One goal was to
arrive at a fixed display presenting the full status of the machine. Verena commented that many items should be looked at system
wise instead of sector wise. Gianluigi remarked that
the proposed approach would teach us how to prepare for more sectors. Roger
added that all the discussion had focused on what we could do in parallel with
ongoing HWC, but another question was
what needed to be done at the very end. Verena
commented that there should be a distinction between the dry runs and the
machine checkout. Gianluigi agreed, also mentioning
the need to test the machine protection system. Ralph announced that he was
drafting a document for the collimator cold check out. Jan pointed out that the
cold check out tests should be done at the end, when nobody was touching the
machine anymore. Ralph said while the cold check out of the collimation was
being prepared, he was concerned by the complexity
of the whole machine protection system. Gianluigi
summarized that we need to define which
tests are required after the individual systems tests, and in particular
look at correlations between systems.
Dry Runs – presentation
postponed (Mike)
Mike’s talk was postponed due to
lack of time. The slides were nevertheless available in his public directory.
Mike announced the dates of the next dry
run, which is planned for week 19,
starting on Monday and continuing for 4 days. It will comprise injection tests,
beam dump tests, and include whatever is available in 5-6 (squeeze, pre-cycle,
etc., expanding on what had been done in 4-5).
Demo of vacuum
displays & PVSS tools (Isabelle Laugier)
Non-conformities and their
consequences (Massimo)