Wednesday
July 26th, 14:30
CCC
conference room 874/1-011
According
to information which Massimo received from
Stefano
reviewed aperture measurements at 450 GeV. His talk
was prepared with the help of Gianluigi, Ralph, Stephane, Massimo, and Federico. After recalling the
generic questions from the chairman, Stefano listed the milestones of the
aperture studies, which include the identification of bottlenecks and their
correction, the achievement of the design aperture, checking the IR aperture in
view of later separation, crossing and squeeze, understanding locations that
become critical at 7 TeV, measuring loss locations of
injection failures, determining the momentum aperture, and cross-checking the
tools and assumptions. Tolerances which must be taken into account when
computing the expected aperture include orbit errors, beta beating, spurious
dispersion, mechanical tolerances and alignment errors. The so-called n1
parameter should be larger than 7, according to LHC
Project Note 111 by
Brennan pointed out that the specifications for the injection protection elements assume a minimum aperture of 7.5 sigma and that this assumption is inconsistent with some locations having considerably smaller apertures. He stressed that the warm magnets should be protected against damage as well. Stefano replied that the warm aperture bottlenecks in the cleaning insertions are a known concern that was brought up in various meeting when the aperture issues were discussed. Helmut remarked that the warm magnets at least cannot quench and that the main worry may be the cold aperture. Rudiger suggested lowering the orbit tolerance to 3 mm at the critical locations in order to recover an overall aperture of 7.5 sigma. Yannis added that 4 mm orbit error looks high in the IR, for which, however, according to Stefano and Rudiger, a smaller value of 3 mm is assumed already.
Stefano continued his presentation by mentioning additional constraints from collimation. Simulation studies with errors have identified critical loss locations both at injection and at 7 TeV. These were presented by Guillaume Robert-Demolaize at Chamonix’06.
The optics and beam requirements for all aperture measurements are stringent (stable orbit and beta functions, reproducible emittances, detailed optics knowledge, chromaticity control, single pilot bunch, pencil beam and nominal-emittance beams).
Gianluigi asked whether the BCT can precisely measure the
partial loss of a pilot bunch due to scraping.
Stefano responded that we should be able to detect changes of 1e9
protons with the BCT quoting a discussion at
The beam instrumentation required comprises BPMs (both position signal and sum signal), BLMs (including movable units), Wire scanners/IPM and BCTs, all fully commissioned, and with an acquisition rate higher than 1 Hz. Hardware needed includes kickers, collimators, CODs, protection elements & dump, the injection & dump operation mode, coasting beams, and SPS scrapers. Among the software requirements Stefano mentioned the 3- and 4-bump steering programme (YASP), an on-line optics model, a dedicated application programme for sliding bumps, and an aperture database containing bottlenecks and mitigating bumps.
Rudiger clarified that ‘coast beams’ does not refer to uncaptured beams, but simply to stored beams. Brennan remarked that the aperture kicker should be used for aperture measurements and not the tune kicker.
Stefano next described numerous techniques for measuring the aperture: (1) closed orbit scans with full beam scraping, as proposed by FrankZ in the LHCCWG meeting of April 5th [FZ had there estimated 2-3 h/ring; Stefano considered this estimate as too optimistic], (2) controlled emittance blow-up and wire scans [requires more time and relies entirely on calibration of a single wire scanner per plane per beam – this method would profit from the availability of other instruments, like IPM or SR; an example SPS measurement was performed by inserting all screens in the transfer line], (3) kick combined with BCT measurements [redundant calibrations; needs new injection for each kick; example SPS measurement achieved a resolution of 20% = difference in the result from kick+BCT measurement and from the wire scanner profile], (4) local measurements with sliding bumps [scan vs. amplitude of closed bumps; good for optimizing known bottlenecks; example from SPS], (5) aperture measurements with BLMs [problem of unknown tails can be solved by pre-scraping the beam at a known amplitude; motivated by 2004 SPS measurements; when the collimator is retracted we expect a sharp increase of BLM signal at the new bottleneck and a decrease in signal at the collimator; parasitic 2006 tests supported the basic idea, but suffered from too high a ramp rate].
Stephane commented that 8.5 sigma is the design goal for the aperture and not 7.5 sigma. Stefano agreed. However, he replied that, since on paper the collimation, dump and protection systems have been designed to work with an injection aperture of 7.5 sigma, he suggests that the bare minimum aperture optimization at injection be targeted at 7.5. A larger aperture is obviously welcome.
The time estimate for the aperture measurement scales linearly with the number of bottlenecks. If there are many, as may happen at 450 GeV, the time needed could also be many hours. The aperture is scanned in three directions (x, y and 45 degrees). Time required further increases if local bumps cannot improve the aperture and an optics change may be needed instead. BLM coverage, fast acquisition, automatic procedures and online data analysis (e.g., plotting BLM or BCT vs. bump amplitude) are essential for rapid measurements and their interpretation. The accuracy of the measurement and correction should be 0.2 sigma, which might be relaxed at the startup. A good knowledge of the optics, and calibrations of kickers, BPMs, and BCTs are the key for success. The error on the measured mechanical aperture has contributions both from the beam-loss measurements and from the error in the rms beam size at the loss location. The latter includes optics errors and an insufficient knowledge of the emittance. Sliding bumps should be applied at a prioritized list of critical locations, including elements which become critical at 7 TeV (change of beta and orbit during the squeeze), IR commissioning with crossing and separation, critical beam loss locations, at dedicated systems [collimation, dump, injection – done by system commissioners], and other locations [suspected regions, alignment errors]. The aperture scans must be repeated off energy (+/-1.5e-3)
Stephane pointed out that the aperture measurement at a momentum offset +/-1.5e-3 is not only a desirable option, but that the aperture measurements must be done off energy. The n1=7 number refers to the aperture including the effect of a momentum offset. On energy a much larger aperture is expected, and on-energy measurements would indeed not provide the relevant information.
The measurement of the momentum aperture is performed with radial steering. A 1% energy change corresponds to a frequency change of about 1.3 kHz. Interlocks must be disabled prior to such a measurement. For increased resolution the beam can be pre-scraped in momentum and in the horizontal betatron beam size. The measurement may need to be repeated after vertical crossing has been set up. The loss locations must also be determined for various injection failure scenarios (SPS extraction kicker, LHC injection kicker, SPS energy), where again the use of a scraped beam is proposed. Stefano also commented on the 450 GeV collision run, for which the same methods can be applied.
Many groups and people would be interested in contributing to the aperture studies. Open issues concern the software for speeding up the data-taking procedure, as well as software for on-line analysis. Off-line understanding of the available information may be required in addition.
Brennan underlined that software for on-line analysis is most crucial. Stefano agreed with him, but he pointed out that if for some unexpected reasons we were to encounter many aperture bottlenecks distributed all around the machine also a fast measurement/correction application could prove essential.
The
aperture measurements have been included in a new set of web procedures. All
LHCCWG members and friends are encouraged to check out these
draft web pages. Comments can be sent to Verena
and Stefano. After the meeting Stefano clarified that the procedure which he
has prepared for the web page is targeted to the required measurements and it
is not coupled to specific techniques applied for performing these
measurements. The description of the proposed techniques is instead given in
separate htm files. The corresponding links can be
changed if we later agree on a different better method. Detailed rules for
dealing with the interlock system (which was pointed out as an important issue
by Ruediger at the end of Stefano’s presentation –
see below) are not yet included in the web procedures. This topic will require
input from the machine-protection experts. ACTION:
Rudiger et al?
Stephane reiterated that off-momentum aperture measurements
are fundamental, and that an aperture of 10-12 sigma is expected on-energy. Off-energy
the dispersion reduces the aperture by about 4 mm. The simulations and calculations done in the
past together with
Rudiger suggested that aperture measurements with blow up by noise not be discarded. The noise blow-up method was extremely useful at the SPS (to be reported in a future presentation by Werner). Rudiger emphasized that the protection aspects need to be included and that this is likely to add significant overhead. In particular he asked how the exclusive use of a pilot bunch would be ensured, when some interlocks are switched off, and that the machine is still alive after the measurements. Brennan proposed that the draft procedures be circulated, stressing the need of making sure that the machine safety is not sacrificed. Rudiger added that employing noise for aperture measurements is better than applying a kick, as the noise can be used for any beam intensity. Stephane noted that the ac dipole may be an interesting alternative, as it can slowly increase the oscillation amplitude.
Roger
recommended discussing the AC dipole proposal together with Werner’s
presentation. ACTION: Stephane?
Set
up an aperture database? ACTION:
Stefano?
Helmut asked for the real goals of the injection failure studies, and whether these measurements would address optics as well as aperture.
After the meeting Stefano emphasized once more that the basic idea described – i.e., global measurement of aperture + detailed local bumps to correct the aperture; procedure to be repeated until we achieve the design machine aperture, – is not necessarily coupled to the two different methods which he has proposed and prefers, namely kick+BCT for quick global aperture measurement, followed by a local scan with scraped beams for more detailed measurements and corrections.
Vadim reviewed possible constraints from experimental background during commissioning. Several teams are working on the various aspects. He distinguished three types of detector background: beam-gas losses in the long straight sections (LSS’s), beam-gas losses in the cold sectors, and tertiary halo losses in the interaction region (IR). The particle density in the cold sectors is 100-1000 times higher than that in the LSS. It has to be understood that: a) this is true for the average gas pressure levels in the LSS's (which were given in the talk) in comparison with the 10^15 (H2 mol/m^3) estimate for the cold arcs, while b) the profile of the gas pressure in the LSS's varies between a few x 10^12 (warm sections) and several x 10^13 (cold magnets) H2 equivalent mol/m^3, so that the ratio quoted above is different for different parts of the LSS's and particular scenarios of the machine start-up.
The background related to the cold arcs is sensitive to the value of beta*, while the one in the LSS’s is not. In the LSS’s, the pressure with 156 bunches is higher than that predicted for the nominal LHC. The tertiary losses finally depend on the loss rate at the primary collimators. The radial distribution of the tertiary halo is different from that of the beam-gas background. At large radial positions, the tertiary background dominates. The calculated rates assume a 30 h beam lifetime.
Rudiger pointed out the large difference in gas density between cold and warm regions. After the meeting, Vadim elaborated on two reasons for this (as were discussed, for example, by A.Rossi in her talk at Machine Background WG), namely the photon flux to the wall of the vacuum chamber in the cold arcs is 10 higher than in the LSS's and the photon mean energy is higher by a factor ~3; together this yields a gas pressure level ~30 times higher than e.g. in Q6 (several times 10^13 H2 mol/m^3) and results in the quoted value of 10^15 H2 mol/m^3.
Stefano explained that the TCTV/H collimators are set at 8.3 sigma. Without such tertiary collimators, the losses would occur in the triplets. Rudiger suggested generating a plot of charged hadron flux versus radial position in the absence of tertiary collimators for comparison. Vadim stressed that his team needs a scenario of collimator operation at the 450-GeV start up, and, in particular, loss maps from the collimation project. ACTION: Stefano & Ralph
Estimates for IR8 are based on 2001 pressure data and suggest 6% of level 0 muon trigger bandwidth at 200 kHz. In IR5 the nominal background amounts to 1% of the pp-related events. Only occupancies were quoted, not the trigger rates. In nominal conditions, LHCB is more sensitive to background since it has a much lower luminosity, while the machine background is about the same. A new study considering the use of background for detector alignment was performed for ATLAS, with 43 bunches at 7 TeV. Losses in the cold sectors were here found to add 55% and 374% to the hadron or muon background from LSS losses. The different IRs receive different amounts of background from the cold sectors depending on their distance to the cleaning insertions.
Frank commented that the estimates consider a 100 beam-gas lifetime in the cold arcs and 30 h total lifetime for tertiary losses. The real beam lifetime during commissioning might be worse and the detector increase correspondingly.
In IR8, the total muon background is about 24% of pp related events for nominal conditions without shielding. For IR8 the situation at 450 GeV in commissioning is better than that at top energy, if a significant number of collisions will be provided for LHCB. The beam-gas background may then be less than 1% (not including any tertiary background). For IPs 1 and 5, with 43 bunches in commissioning, one expects about 50% machine induced background at beta*=18 m. The percentage number goes down as the IP beta function is reduced. With 156 bunches the machine background will be completely dominant.
Summarizing all the studies performed, during commissioning in IP1/5 the relative size of machine background with respect to the pp related events increases at least by a factor of 3 with respect to the nominal LHC, and it may even reach full dominance over the pp related events. IP8 has an advantage in the commissioning as compared to the nominal LHC, provided the number of colliding bunches amounts to a sizable fraction. None of these estimates includes any tertiary halo, since the rate on the tertiary collimators is not yet known.
Background becomes less important if we squeeze beta* as soon and as much as possible. An estimate is needed of the collimation inefficiency and the loss rate in commissioning. ACTION: Collimation Working Group. The gas pressure in both warm and cold regions needs to be monitored. Without safety factor, the machine background is already at the limit of acceptability. It needs to be measured as soon as possible, with the first beams stored in the machine.
Rudiger remarked that a more precise value for the real gas pressure should be available soon from the hardware commissioning. Frank suggested that the fraction of colliding bunches should not matter for the background-over-event ratio at LHCB, if only the colliding bunches are measured by gating. The background should arrive synchronously with the collision events, and the bunches are widely separated, so that it should be easy to separate the background generated by successive bunches. After the meeting, Vadim informed the LHCCWG that Gloria Corti has already been contacted for a direct response to this question from LHCb, and that she will be back at CERN by mid-August.
Frank further commented that a comparison between 450 GeV and 7 TeV backgrounds would be of interest, since the situation may conceivably deteriorate at 450 GeV. Helmut emphasized that the question of tolerable background is important. He recommended that all numbers be quoted also as absolute values and not only as fractions. Vadim replied that the experiments could be encouraged to measure the absolute value of the background with a dedicated detector. Related to this point, it was suggested that the experiments be asked to identify a signal from their sub-detectors which can be used to monitor the machine background. Helmut added that single-beam studies are a key for understanding the background.
Responding to a question by Emmanuel concerning the efficiency of the staged shielding, Vadim elaborated that the full shielding in LHCB will reduce the background to one third of its present value from beam-gas alone, that the partial shielding installed during commissioning reduces the beam-gas background by about a factor of two, and that the full shielding will decrease the total background also by about a factor of two. This estimate refers only to the total particle flux, not to the LHCb trigger rates (the analysis of the latter is still undergoing in LHCb). Thijs asked why the curve of background versus radial position shows an increase at r=1.5 m. Vadim explained that the 1.5-m value equals the distance between the beam and the tunnel wall on one side of the beam. Also, since a) the beam position is shifted with respect to the center of the LSS tunnel, b) the tunnel cross-section is cut at the bottom by the tunnel floor - the radial (centered at the beam axis) distributions of the background flux density at the radius above ~1.5 m actually represent the superposition of several geometrical factors. Details of the background distributions in this region can be much better seen from the X-Y plots, for example, from Vadim’s talk at the LHCb Plenary meeting in June 2006.
After the meeting Helmut recalled that it was decided in
Roger announced that there will be no LHCCWG meetings in August.
Wednesday September 6th, 14:30
CCC conference room 874/1-011
Provisional
agenda
Minutes of
previous meeting
Matters arising
AOB
Reported
by Frank