Wednesday
September 20th, 14:30
CCC
conference room 874/1-011
Roger took up three points discussed in the previous meeting: (1) Everybody agreed with the target values for the ring aperture as laid down by Stephane and summarized in the minutes of the 12th LHCCWG meeting (i.e. n1=7/6.7 at QF/QD with exceptions). (2) Concerning the possibility of parasitic collisions for detector alignment, brought up by some US ATLAS members, Massimiliano reported that there is no official request from the experiments at the moment. The experiments will present a wish list at the next CRAG meeting this fall. Massimiliano asked whether injecting short batches with the 25-ns spacing during the 450-GeV run would be a useful step for accelerator commissioning, especially in view of the 2008 run (in addition to providing the requested parasitic collisions on both sides of the interaction point). Verena and Paul commented that injecting a full batch is not possible in view of machine protection. The acceptable intensity limit for transfer and injection is 2e12 protons until the machine protection systems are fully commissioned, which will not be the case during the 450 GeV run, as, e.g., the transfer-line collimators will not even be fully installed. Frank suggested that even with a third or half the nominal bunch intensity, running with 25-ns bunch spacing could be useful, as one would early on detect regions with strong multipacting and enhanced vacuum activity, if present, which could then be mitigated in the following shutdown (e.g., by additional bake out). Oliver commented that the IR beam pipe will likely not be accessible in the shutdown. However, Paul clarified that the IRs will need to be opened for the phased installation of tertiary collimators. (3) From the discussion at the last meeting, there seems to be a clear request to install auxiliary scintillators for luminosity measurements in the 450-GeV run. Roger asked whether we should become active in this matter. Daniela explained that the installation of these scintillators has not yet been decided. Helmut is following up the issue together with Enrico Bravin. He expressed a strong preference for dedicated detectors installed inside the TAN to be used by the machine. Daniela informed the working group that, next January, a joint workshop on TAN integration and on luminosity measurement will be organized in collaboration with American and Japanese colleagues. The exact dates are yet to be decided.
Yannis
thanked several colleagues for their help in preparing this presentation,
especially
Yannis started out with an LHC overview. In the following he concentrated on the four experimental IRs. IR1 and IR5 are identical except for the beam crossing schemes, the tunnel slope, and the beam screen orientation. He listed all magnetic elements and showed the layout of each IR. IR2 includes injection elements for beam 1, an experimental dipole plus three warm compensator magnets. IR 8 is similar to IR2, except for an IP shift by 11.25 m, which implies a non-symmetric magnet layout in the matching section Yannis also presented the IR1 & 5 injection optics with beta*=17 m.
Roger commented that the beta*=11 m optics had recently been chosen as starting point for the 450 GeV run; see the minutes of the 10th LHCCWG meting on July 12. Yannis replied that he was not aware that this proposal had been endorsed and that the optics presented corresponds to the latest official optics V6.500.
Yannis next explained the crossing scheme in IR1 and IR5. The nominal amplitude of the separation bump is +/- 2.5 mm. The signs of the vertical crossing angle (IR1) and of the vertical separation bump (IR5) are arbitrary, but the horizontal signs are fixed by the ring geometry. Two times 11 steering magnets are used per beam and per IR, for constructing the separation and angle bumps. Both vertical signs should be commissioned. Calibration of all element strengths with beam is needed.
Paul asked whether all 11 correctors are indeed necessary at 450 GeV. Oliver replied yes, since otherwise the bumps may not be closed, and also since, if all are used, we can control the absolute transverse position of the collision point. One alternative possibility would, however, be to employ only the correctors not common to the two beams. Jean-Pierre remarked that not all steering correctors may be needed at 450 GeV, and he advocated a reduced complexity. Jean-Pierre added that the hysteresis of the steering magnets is not known today, but that it will be determined in the “extended measurement programme”. Paul observed that the conditions at 450 GeV will not be the same as later at 7 TeV in any case. Therefore, we do not need to choose the same configuration as at 7 TeV.
Concerning the number of correctors used for the separation and crossing bumps, Werner clarified after the meeting that in order to control position and angle at the IP, 4 correctors are needed per beam and per type of bump, two on each side of the IP. All these 8 correctors per IP are separate for the two beams and are mandatory. Four common correctors (MCBX next to Q1, left and right, horizontal and vertical) are used to simplify the aperture and strength (7 TeV) requirements. Since their strengths must fulfill left-right symmetries, they add only 2 degrees of freedom. Only these correctors are optional at 450 GeV. Not using them may reduce the aperture, which could be an issue especially at 450 GeV. Example bump orbits with and without common correctors can be found in LHC Project Report 315.
Ralph asked why both signs of the vertical bumps should be tested. He pointed out that commissioning the second sign requires additional beam time, and that, especially for the crossing angle bumps, commissioning both signs may not be useful. Yannis replied that at least for IR2 and 8 the commissioning of both signs is unavoidable, since the experimental magnets will change their polarity frequently.
Frank asked whether we plan to commission the crossing angles at 450 GeV. Ralph recalled that the installation of the collimation system is phased. He stressed that the triplet should never become the aperture bottleneck. An aperture analysis by Stephane had shown that with zero crossing angle, IR1 and 5 are no bottlenecks for beta*=17 m, though this may not longer be true for 11 m beta*. On the other hand, the IR2 and 8 apertures definitely are bottlenecks. Indeed, to increase the aperture in these IRs the IP needs to be displaced, e.g., by 1.5 mm in IP2.
Yannis
now turned to the IR2 injection optics, which superimposes a +/- 80 microrad external angle on the internal angle of +/- 70 microrad generated when compensating the orbit distortion
induced by the spectrometer. The situation in IR8 is similar. In IR2 and IR8,
some n1 values are below 7 sigma. Magnets are shifted
transversely in order to increase the acceptance. Alignment data should be
included in the optics and aperture models. Yannis
showed aperture plots for the 4 IRs created by
Oliver noticed that the beam emittance at 4e10 protons per bunch will be smaller than nominal. Ralph replied that, regardless of the emittance value, the triplet would still remain the bottleneck, even if the aperture in units of sigma increases. In addition, a lower-emittance beam has a larger damage potential. Responding to a question by Frank, Yannis explained that off-momentum orbit and spurious dispersion and beating are included in the calculated apertures. Oliver pointed out that at 7 TeV with squeeze we assume a peak orbit error of 3 mm rms in the triplet rather than the canonical value of 4 mm. Similar assumptions might be made for the 450 GeV conditions.
Yannis next discussed beam instrumentation needs, first addressing beam position monitors and beam-loss monitors (BLMs). Among the latter are ionization chambers, two of which are located close to Roman pots, and SEM monitors at special locations. Detecting losses from pilot bunches requires integration times above 1 s. For fast BLM response, <0.25 ms, bunch intensities above 6e10 protorns are needed. DC BCTs and fast BCTs are installed in IR4. The ideal machine conditions for IR aperture studies comprise a well corrected and stable closed orbit, a measured optics model, stable beams with reproducible emittance, cross-calibrated BCT and DCCT, and application software for sliding bumps. Yannis considered a bunch intensity of 1e10 as safe in view of quenches, but this number may still have to be reviewed.
Helmut commented on the absence of calibrated BLMs. A beam-based BLM calibration had been foreseen for the sector test, which was cancelled. Ralph mentioned that all BLMs are calibrated in the tunnel by a radioactive source. Verena and Helmut pointed out that this only reveals the functionality of the BLMs, but does not provide a calibration of the BLM readings in terms of quench level. Frank asked which crossing angles, if any, should be chosen in IP2 and 8 for the 450 GeV run.
The two principal aperture measurement methods are two-corrector orbit oscillations, yielding the global aperture, and local or sliding bumps across individual IRs. Yannis commented that the data from the magnet evaluation activity should be transferred to the machine operation database, as the quality of comparison with the measured available aperture will depend on the knowledge of the magnetic field quality and alignment. Knowledge of the triplet aperture can be used for the set up of the tertiary collimators. He illustrated the orbit-oscillation measurement scheme and presented an example from TI8 commissioning. The sliding bumps are greatly facilitated by the YASP software. Some cross calibration is possible in the process. A complete aperture map may ultimately be obtained.
ACTION: Jean-Pierre, Mike or Massimo? – Transfer data from magnet evaluation activity to machine operation database (Massimo mentioned that a presentation is foreseen at a future LTC on the data needed for operation)
Separation bumps are a special type of IR bumps. Yannis recommended that they be implemented after the IR aperture measurements, in a beam-by-beam, pilot bunch, step-by-step approach, and then repeated with opposite sign. The optics and aperture will be re-measured with separation bump active. The validity of the bump should then be checked with the other beam, and finally both beams be injected for final optimization.
Paul emphasized the need to
correct in both planes at the same time, as there may be important cross talk.
The crossing angle plane in IP1 and 2 is vertical and cannot be changed. A
great concern is the effect of the
experimental magnets in IP2 and IP8 at 450 GeV.
The
Ralph wanted to know the reason why we cannot inject at full strength, as 19-20 sigma clearance is acceptable. Replying to Ralph’s question Oliver mentioned failure modes as one concern. The time constants and margins of the compensator magnets may be a critical input to this discussion, and they should be checked. Frank asked whether we should aim for zero crossing angle at IP2 and 8 in the 450 GeV run. The answer was that this may depend on the experiments. In particular, Massimiliano added the related important question if one can close the VELO at point 8 with the experimental field fully on. If this is not the case, the LHCb experiment might want to start with field off and close the VELO, then later try with field on and VELO open (or half closed). Massimiliano reminded the team that the experiments presently do wish to have the possibility to take data at 450 GeV with the dipole magnets at full '7 TeV' field strength.
ACTION: Yannis? - Check time constants and margins of warm compensator magnets
ACTION: Yannis with input from Massimiliano? – Can VELO be closed with experimental field at full strength?
Ralph who had replaced Roger at an LHCB meeting last week, suggested that we may obtain important information from studying the effect of the experimental solenoids on the beam. He specifically recommended exploring solenoid effects rather than implementing crossing-angle bumps, as the latter constitute a potential threat to machine protection and collimation. Related to this point, Massimiliano asked about the effect of the solenoids on the beams. The answer by Oliver and others was that the XY coupling would be tiny at 450 GeV, and, that, therefore it would not matter whether the experimental solenoids are ON or OFF.
Despite of his above concerns, Ralph pointed out that at least IR8, where two collimators will have been installed already, should be safe with regard to angle bumps, provided there is no bottleneck in the triplets. Ralph raised the question how we can make sure that we do not violate this condition. Massimo explained that the general idea underlying Yannis’ presentation was to commission the “nominal” scheme. The internal angles in IP2 and 8 do not cause any aperture problems by themselves.
Frank asked for the time required to ramp the magnets. Paul replied seconds, whereas Massimiliano thought it was on the order of minutes. Ralph suggested that attempts be made to slow down the ramp for safety reasons. After the meeting Massimiliano confirmed that the LHCb magnet is currently ramped to full strength in about 15 minutes. A further reduced ramp rate may ease the accompanying thermal shock. On the other hand, much faster ramp is not possible, A "few minutes" (exact number yet unclear) is the absolute minimum, though an acceleration of ramping below the present rate is not desired by some of the experimental magnet experts.
Yannis listed a number of other possible measurements with bumps, such as identification of triplet errors, studies of long-range beam-beam effects, and impedance characterization. He stressed that the time needed for IR aperture studies may strongly depend on the machine conditions, in particular on the number of aperture bottlenecks and on machine protection issues. He indicated candidate members for aperture measurement teams and bump commissioning teams.
Paul advocated to start only with the internal angles in IP2 and 8, and not to commission the external crossing scheme at the start. The latter could be commissioned together with all the other separation and angle bumps. He recommended that all studies be done with both beams present in the machine, phased such that they collide in a separated region of the ring. He emphasized the advantages of working with both beams at the same time, such as gaining additional information, avoiding any negative effects of bumps on the second beam, and saving time. Oliver recalled a possible interference of the two beams in the Q1 BPMs. Paul assumed that the BPM readings should be fine with a few 100 m longitudinal separation.
Jean-Pierre commented that the priorities need to be defined. Roger explained that the commissioning of the full machine is different from the 450 GeV run. Ralph requested that all studies be restricted to safe beam.
From Yannis’ numbers, Massimiliano roughly estimated the effect of the full field in LHCb on the beam position at the exit of the VELO detector, the latter extending from -0.35 to +0.75 m around IP8. According to this first estimate, one would lose about ~2 mm in aperture. Therefore, compatibility of full-field operation with VELO is not excluded at first sight. Yannis will further look into this.
There was no particular point raised.
Wednesday October 4th, 14:30
CCC conference room 874/1-011
Provisional agenda
Minutes of previous meeting
Matters arising
Fill-to-fill reproducibility and differences between the two apertures
(Luca Bottura)
AOB
Reported
by Frank