Tuesday January 29th,
14:00
CCC conference room
874/1-011
Concerning
the minutes of the 38th meeting and with reference to a message from
Massimiliano, Roger clarified that
Referring to a comment made at the previous LHCCWG meeting that Paul wished to render the access system operational as quickly as possible, Roger mentioned that some experience during the hardware commissioning was adding an extra push to rapidly go ahead with the full commissioning and activation of the access system.
Roger turned to the central question of how we will transit from the hardware commissioning to the first beam. In the preceding week a small team of people had started to look at the coordination of all pertinent issues (dry runs, machine check out, etc.), in particular at how the relevant information can be collected and made visible. Related to this topic, the consequences of non-conformities would be addressed in one of the following LHCCWG meetings. The pertinent information from the hardware commissioning must be available to the beam commissioners.
Oliver recalled that this should be taken care of by the “as-built database”. However, Jan reported his impression that the development of the latter was not advancing very rapidly. Massimo and others commented that a more centralized coordination could be useful. Oliver remembered a presentation at the 65th LTC by T. Petterson, at the end of which a decision may have been taken that the operation and equipment groups should provide lists of which additional databases or information would be of interest.
Mike presented the schedule for dry runs which is accessible from the LHC beam commissioning web site. Dry runs on the injection system had already been performed in mid-December 2007 and a presentation had been given at the 37th meeting of the LHCCWG. Dry runs for the beam dump are scheduled on 14 and 15 February. The first cold sector dry run, for sector 5L, could take place as early as the end of week 5 (the present week) depending on the hardware commissioning progress.
Jean-Jacques commented that BI would prefer an improved integration and coordination of dry runs in order to conduct various checks together in a single week, including the pertinent checks for BI instruments. He stressed that there was a heavy involvement of BI in every dry run, and he recommended that an attempt be made to minimize the total number of dry runs and to cover various aspects together. Verena remarked that there always was the same group of people involved for all dry runs, and that this group could not do everything at the same time. Gianluigi mentioned that an overview of equipment availability is being considered, which would help to optimize the dry runs and the machine check out. Roger commented that the Extended LTC Days will address these sorts of issues, but that the discussion could also be resumed earlier, at the next LHCCWG meeting.
At the end of this meeting, Jean-Jacques explained that for BI there is large difference between dry runs (tests of operational performance, synchronization etc) and the cold checkout (specific tests to check exact settings). He highlighted that different people are involved in these two activities.
Massimo followed up on his presentation at the 27th LHCCWG meeting in June 2007, after which it had been decided to investigate the possibility of relaxing tolerances for the commissioning, in collaboration with the Machine Protection System Commissioning Working Group (MPSCWG). In dedicated discussions with the MPSCWG a number of comments and conclusions were drawn.
The proposed approach was to accept a reduction of the cold aperture below 7.5 sigma (corresponding to n1=7). Massimo presented a table with relaxed tolerances suggested for the various commissioning phases (43 bunches, 156 bunches, 75-ns spacing). He pointed out that the so-called “nominal” reference case in this table is not the “nominal LHC”, but the beam of commissioning stage III, with 25-ns bunch spacing and 5e10 protons per bunch. By accepting the proposed aperture reduction one gains an additional tolerance budget that can be redistributed to orbit errors, beta beating or dispersion beating. Example allocations were shown.
Massimo reported that the proposals of relaxed apertures for stages Ib and II (156 bunches and 75 ns spacing) were rejected quickly since the pertinent beams are unsafe. Frank asked whether the collimators could not alternatively be closed to 4 or 5 sigma which should allow a safe machine while providing margin for larger optics errors. He recalled that the HERA proton ring had had a physical or dynamic aperture of less than 4 sigma in the first years of operation. Massimo replied that closing the collimators had not been considered as an option.
The discussions with the MPSCWG concluded that the tolerances can be relaxed via the proposed approach only for stage Ia with 43x43 bunches, and even for this stage the relaxation of the required cold aperture is possible solely at injection through commissioning phase A.4, which would not be very helpful in practice since the beta beating would first be corrected in phase A.4 anyhow.
Two alternative paths were, therefore, identified:
(1) Reducing the transverse emittance (there is a natural link between intensity and emittance, which should facilitate this scheme). Natural limits are set by the requirement that the transverse energy density must not exceed the damage threshold, and by possible difficulties of emittance preservation, since the emittance growth due to steering and dispersion errors is rather insensitive to the value of the injected emittance.
(2) At 7 TeV an increased beta beating may be acceptable if beta* is not less than 2 m for all three stages considered. This second proposal is still controversial, however. In any case the beta beating and dispersion beating will have to be under control and nominal at injection and on the (first part of the) ramp.
Roger summarized the basic message of Massimo’s presentation which is that we must correct the beta beating down to the level of the specification. Massimo confirmed, yes, indeed this is the case, adding that we must start by having a better-than-specified closed orbit, as this is certainly much easier to control than beta-beating and it would gain some margin in aperture. Nevertheless, the beta-beating should be also correct according to specification.
Frank pointed out once more that we could also close the collimators down to e.g. 4 or 5 sigma in case we have difficulties in achieving or stably maintaining the required beta-beating level and find no other fast possibility to proceed. Ralph signaled his full agreement, and he emphasized that we should always close protection elements so as to ensure a safe machine. Assuming the numbers from Massimo’s stage-Ia example, Ralph considered a closing of the primary collimators from 5.7 sigma to 4.2 sigma in the ring in case we need to give up 1.5 sigma aperture. He commented that in such scenario indeed smaller beam emittances may help, and they would also be OK for low beam intensities.
Gianluigi inquired about the danger of a smaller beam in terms of its local energy deposition, and he specifically asked whether the granularity of the energy deposition was really at the mm level. Brennan remarked that this was the general criterion for defining the safe beam limit. Ralph explained that the usual assumption for the local energy deposition is its scaling with the inverse square of the beam size. He highlighted that, however, the variation of the beta functions around the ring is not taken into account normally. Verena remarked that a single number was a convenient criterion. Ralph commented that the limit should correctly be specified in terms of beam energy per mm^2. He quoted the example of SNS, where quadrupole errors can lead to reduced beam sizes at the location of the beam loss. For the LHC a similar situation could arise with a triplet error at top energy. Verena suggested that we need to be more flexible when defining the safe beam.
Ralph pointed out that 43 bunches at 5e10 intensity as foreseen for stage Ia were not safe in his understanding. Stefano and Mike commented that in case there were a beam-loss problem in stage Ia, this problem would be much bigger for the nominal beam if the same failure scenario was applied to that beam. Brennan stated that the real question was if and where we can relax the tolerances. Stefano asked whether the underlying assumption was that with relaxed tolerances we could lose the entire beam in a single spot, or, in other words, whether the 43-bunches scenario with relaxed tolerances was a case with no protection at all or with a relaxed protection. Reformulating Stefano’s question, Mike inquired which failure scenario would make the 43 bunches with relaxed tolerances dangerous. He suggested that energy tracking errors might be such type of failure. However, Massimo reminded the working group that the relaxed tolerances are considered only at injection and not for the ramp.
The general conclusion of this presentation and discussion was that the correction of beta beating and dispersion beating must work with the specified precision for all LHC commissioning phases.
Jorg reported the outcome of the Joint LHC Machine-Experiments Workshop on Machine Protection, held in June 2007. At that workshop all LHC experiments expressed their concern about the maximum intensity that may be injected into an EMPTY LHC ring. The main message was that the intensity limit for injecting beam into an empty LHC ring should be as low as possible. In particular, the experiments felt quite uncomfortable with our previously assumed limit of 1e12 protons.
Following the workshop, a proposal how to address the experiments’ concern had been presented by Jorg at the 85th LTC meeting in December. More specifically, their concern was beam injection into an experiment, due to wrong settings of certain critical magnets like D1 and MCBX. Two remedies against such failures were introduced: the software interlock system – SIS (exploiting the fact that the relevant failure scenarios imply enormous setting errors which can easily be detected); and a new interlock flag in the SPS, offering greater flexibility.
The extraction/injection interlock ingredients comprise four safe beam flags: the SPS Safe Beam Flag, which is TRUE for proton intensities below 1e12; the SPS Probe Bam Flag for injection into an empty LHC ring, which is TRUE for intensities of “fx1e11” protons where “f” is a factor smaller than 1 and equal to 0.1 by default, the LHC Safe Beam Flag, which is TRUE for up to 1e12 protons at injection, and up to 1e10 protons at 7 TeV; the LHC Beam Presence Flag, which is TRUE for LHC intensities >2e9 protons.
These flags exist to be able to mask certain interlocks, e.g. not to dump the beam immediately in IR6 as the orbit moves around during commissioning.
Ralph asked whether the limit of 1e10 for the LHC Safe Beam Flag at 7 TeV had been agreed upon, stressing that even the pilot bunch may not be safe at top energy. Replying to Ralph’s subsequent question if one could move out all collimators and protection devices with a “safe beam”, Jorg responded, yes, this was the case for the time being, but, he added, that there was an ongoing effort to construct a system which would prevent any masking of interlocks at 7 TeV.
Jorg now explained the logics behind the flags. If BFP_SPS = FALSE, one can inject into the LHC only if BPF=TRUE. For MD purposes the injection of up to 1e11 protons into an empty machine will be possible (exceptional limit). The detailed conditions still need to be defined. If the LHC ring is filled there is no limitation on the injection, but the LHC Safe Beam Flag must be forced to the value FALSE in order to inject an unsafe beam.
Frank asked for the reason why the experiments would accept larger injected intensities for MDs, while the latter might be inherently more unsafe. Jorg replied that the detector electronics would be completely switched off for the MDs, while this was not the case for regular operation (“standby mode”), so that the experiments were much more sensitive in the latter case.
Jorg reviewed the SPS extraction ingredients. The extraction from the SPS in LSS4 and LSS6 is controlled by individually designed special MASTER BIC modules which take into account the type of beam, the dump states, the LHC injection BIS state, and all transfer line BIC inputs. He illustrated the LSS4 and LSS6 master inputs.
The commissioning of the system is supported by a special monitoring program, which will ease the diagnostics for extraction with a large number of rapidly varying interlock signals. A special panel will help in the commissioning of the MASTER BIC.
The full commissioning of the BIC requires a lot of ingredients: Super cycles with CNGS and LHC, LHC beams of varying intensity, all BICs with at least one input that may be toggled between TRUE and FALSE, and the possibility to move all TEDs which relies on an operating LHC access system. The commissioning of the LSS4 MASTER BIC must be done for mid-May for CNGS operation and the TI8 extraction. The LSS6 MASTER BIC is needed for the TI2 extraction.
Roger asked whether the factor “f” was under the strict control of a few people. Jorg replied, yes. He added that the value of “f” will be sent to the experiments. Roger recalled that ATLAS was the experiment which had been most worried. Rossano asked who would generate or receive the safe beam flags. Jorg answered that a “magic parameter box” would produce and distribute the flags based on numerous input signals.
Tuesday February 12th, 14:00
CCC conference room 874/1-011
Provisional agenda
Minutes of previous meeting
Matters arising
Filling schemes for zero crossing angle
(Massimiliano and Werner)
AOB
Reported by Frank