Tuesday March 27th,
14:30
CCC conference room
874/1-011
Roger explained the reasons for modifying the LHCCWG meeting schedule. In March Steve shifted the LTC by one week so that the synchronization between LTC and LHCCWG was lost. To avoid future timing conflicts, the LHCCWG meetings were relocated to Tuesday afternoon. Roger thanked all team members for agreeing to accommodate this change. One or two requests were made to start the Tuesday meetings at 14:00 rather than 14:30. This time change was approved.
At the newly oriented LTC, presentations are being given on the commissioning procedures developed by the LHCCWG. So far three of these presentations have been delivered: Brennan discussed the injection and first turn. Gianluigi followed with procedures for the circulating beam. And Rhodri reviewed the initial commissioning. At the next meeting, on 11 April, Frank will present optics measurements at 450 GeV. Future presentations will cover increasing the intensity, snapback and ramp, squeeze, and two-beam operation. Two-beam operation will first be looked at in the LHCCWG by Jan and Ralph at the beginning of May. It is foreseen to produce official documents describing each commissioning phase. Magali commented that one possibility to generate this documentation consists in generating pdf files from the corresponding web descriptions. Roger pointed out that producing such official documentation would be important and address a concern which had been raised by Steve.
Ralph
asked about the possibility for a
There were no comments on the minutes of the last LHCCWG meeting.
Frank
discussed the optics measurements to be done at 7 TeV.
After acknowledging the help of several team members, he first laid out a few
basic considerations: The nominal 7-TeV optics is similar, but not identical to the 450 GeV optics. In addition
to different magnet errors, also the nominal 7-TeV optics for IR2 is different
from the injection optics. A number of aspects will facilitate the 7-TeV
commissioning: The BPM offsets are expected to change by less than 50 micron
between 450 GeV and 7 TeV;
the apertures are already checked at 450 GeV, and at
7 TeV the normalized aperture should be 4 times
larger. On the other hand, damage and quench levels are considerably reduced at
7 TeV, and it will take much longer (~1 h) to return
after beam loss. At 7 TeV, the metal damage limit
corresponds to roughly the pilot bunch intensity, whereas the collimators can
withstand the impact of up to 4 nominal bunches. The quench levels set the
tightest constraints: At 7 TeV the local loss of 5e5
protons is expected to quench the
It was asked whether the quench level at injection really corresponds to the local loss of 1e9 protons. In response, Frank pointed to the reference given in his Chamonix 2006 presentation [J.Jeanneret, D. Leroy, L.R. Oberli, T. Trenkler, Quench levels and transient beam losses in LHC magnets, LHC-Project-Report-44, 1996]. According to this source, for fast losses the quench level corresponds to 1e9 locally lost protons or to 1e9 lost protons per meter, respectively (Table 6 of the above reference). In case of striking incidence and distributed losses, the quench level would be higher. Ralph commented that the quoted damage level for collimators applies only to the primary and secondary collimators, while other types of collimators are much more sensitive.
Frank now considered two extreme cases. If the optics hardly changes when going to 7 TeV, there would be no need for detailed measurements, In the other extreme, if the optics strongly varies, optics measurements and corrections would be required already on the ramp, and not only at 7 TeV.
He next addressed the anticipated orbit changes between 450 GeV and 7 TeV. Orbit variation arises (1) from the residual of 50 units change in the transfer function of the arc dipoles, (2) from about 700 units change in the transfer functions for D1, D2, D3 and D4 (in the case of warm magnets, this change is due to saturation, in the case of cold magnets due to very low excitation at injection; assuming ~1% transfer function accuracy, the residual ~7e-4 error in D1 to D4 would correspond to 50 micron change per dipole), (3) from the fact that an exact local correction is not foreseen for the LSS (due to an insufficient number of orbit correctors at Q4, where correction is possible only in one of the two planes), (4) from a reduction of the separation bump between the two beams (the separation distance of 2.5 mm at 450 GeV must be reduced above 5 TeV), and (5) from changes in rf frequency & circumference. The orbit feedback, if active, will correct most of the resulting orbit variation. Frank presented the warm-cold offsets measured for a single cold magnet, D1L103, at BNL. He also illustrated the rf frequency change between injection and top energy for protons and lead ions. For protons the total frequency swing is 10 kHz. A 1% frequency error would correspond to 100 micron orbit change. Extrapolating from a simulation by Ralph S., Frank expected that momentum offsets of 1e-5 can be resolved for pilot bunches.
Oliver and Massimo commented that the Q4 correctors can be used for orbit correction. After the meeting, Stephane explained in more detail that those correctors close to Q4 which are needed for the separation bumps are near their maximum excitation level and, therefore, they will likely not be available for correcting the orbit, and that proper orbit correctors are installed for one plane only. Massimo and Ralph added as a further source of orbit variation the change in the magnetic center of the warm magnets between injection and top energy of about 0.2 to 0.3 mm [see Massimo’s presentation at the LTC of 19.03.2003]. Oliver stressed that not only the number of D1 magnets measured cold was limited to a single one, but, even more worrisome, that this measured magnet was also of a different type than the other three D1 magnets installed in the LHC. Massimo later elaborated that the remaining three D1 magnets feature laminations with different magnetic permeability, which renders the measured transfer function of limited use. This problem concerns IP2 and 8. Oliver further commented that tracking errors between sectors are yet another contribution to the orbit change.
Frank turned to the expected optics changes, drawing from his presentation at Chamonix XV [Beam Measurements Required in the First Two Years of LHC Commissioning]. The tune change is of order 0.1; the chromaticity varies by about 320 units. The peak change in the beta beat may reach values up to ~40% with contributions from +/- 2.5 units b2 change in the two beam apertures leading to 7-8% peak beating and from 7 units b3 change resulting via spool-piece misalignments in an additional 6% peak beating, plus several others. A recent WISE calculation by Per Hagen without any geometric errors predicts 13-18% peak beta beat at 7 TeV. Frank emphasized that most of the errors will lead to a beta beating which bears little correlation to the beta beating encountered at 450 GeV, even though the absolute magnitude is similar for both cases.
Other expected optics changes include a variation in the betatron phase advance between IP1 and 5, corresponding to 0.14 units of 2pi, and a coupling change of order ~0.05. In addition, quoting John Jowett, Frank explained that the optics in IR2 may transit from the injection optics to a so-called pre-squeezed optics. Frank added that, based on information from Ralph S., the tune and coupling feedbacks are likely to be active, while an operating chromaticity feedback is less probable. Ralph S. commented after the meeting that the default (safe) fall-back Q' FB solution will be based on tune tracking while slowly modulating the beam momentum through cavity frequency changes. As a further illustration Frank showed the b2 evolution with MB current for the two apertures, as well as the b3 variation with current. He then presented a table summarizing the preliminary WISE predictions for the beta beating caused by magnetic field quality at injection, at 7 TeV with the injection optics, and for the 7-TeV collision optics.
In the following, Frank commented on the IR2 optics change, resorting to a couple of slides on the topic prepared by John Jowett. The reason for the optics change is related to special phase advance conditions that at injection must be fulfilled between MKI and TDI, as well as between the TDI and two auxiliary collimators. These phase advances cannot be maintained up to 7 TeV without exceeding the nominal strength limits for the triplet magnets. The peak beta function values differ by almost a factor 2 between the IR2 injection optics and the IR2 pre-squeezed optics.
Jean-Pierre and others commented that the strong change of coupling could adversely affect the orbit and tune feedbacks. Ralph S. and Rhodri replied that instead of operating with a combined tune & coupling feedback, one could also use a different commissioning working point, if coupling compromises the feedback performance. An option would be the working point proposed by S. Fartoukh at the LCC of 23.10.2002. Jorg assisted by saying that the commissioning will be made much easier with feedbacks. Ralph commented that the tune spectrum may change widely. Verena pointed out that the feedbacks are included in the procedures for injection commissioning. Thijs added that one can always run feedbacks in an open loop for testing purposes. Jean-Pierre and Ralph S. observed that a large change in the coupling would render the control of orbit, tune and chromaticity (& many other higher order beam parameters) impossible.
Oliver remarked that a 1 mm rms orbit change at the lattice sextupoles would give an additional contribution to the beta beating comparable to the one arising from the b3 change plus spool-piece misalignment. This estimate was confirmed by Stephane [see S. Fartoukh, Revision of the Tolerance Budget for the Beta Beating at Injection, FQWG 08/03/2005].
Ralph asked whether the IR8 optics would also need to be changed in the same way as for IR2. Oliver answered that, yes, if an optics change is needed for IR2 then it would also be required in IR8. However, he then elaborated that these optics changes likely will not be necessary, pointing out that the quoted strength-limit for the quadrupoles includes a 15-20 T/m margin for collision debris, which is not relevant for IR2 and IR8, and especially not at commissioning. Without this margin, the quadrupole strength limit can be raised to 220 T/m. The optics could then stay constant, and an optics change be performed later during the squeeze. That a gradient of 220 T/m can be reached by the triplet magnets has already been verified during hardware commissioning.
In response to a remark by Mike that in IR2 the Q4 magnet is powered at low current, near the edge of power-converter stability, Massimo stated that the strength values in the optics are consistent with the specification that magnets can be excited down to 2% of their nominal value. Oliver commented that one could use a smaller number of modules in the power supply if stability is an issue. Helmut is in contact with Thys on this matter.
Frank now detailed the measurements and corrections at 7 TeV. Tunes can be measured by exciting the beam. If the tune (and coupling) feedback is active, this feedback will directly provide the tune value. It has not been decided whether the collision tunes are chosen at this stage. Chromaticity is measured either in the classical way, using a tune signal in conjunction with radial steering, or by means of the head-tail monitor, or by rf phase modulation. The orbit is measured and corrected, and possibly stabilized with the orbit feedback if active. Coupling is measured and corrected using orthogonal knobs, either by minimizing the closest tune approach or by minimizing the coupling line in the tune spectrum, or, more simply, by the tune & coupling feedback if active (the feedback uses the same algorithms). Concerning the linear optics, beta function and dispersion should be measured as well as the phase advance from IP1 to IP5. ORM measurements on the other hand are not needed for the feedback, since the latter can work even in presence of large optics errors. Also a b3 correction check is likely not necessary at 7 TeV, where the beam is smaller than at injection, and after the corrector circuit powering checks already performed at injection.
Ralph S. commented that based on earlier discussion with Joachim Tückmantel et al. (AB-RF) , fast RF phase modulations (> 20 Hz) seem to be excluded due to the required large RF power that needs to be dissipated and which is limited by the RF cooling system. He remarked it is important to note that the feedback (or basic slow control) minimizes only the "global" coupling parameter that is measured at the location of the tune PLL pickup but does not replace a thorough control of the local coupling that is e.g. based on kicks and FFT spectra of each individual BPM. Oliver commented that one can kick a bunch at least 20 times to determine the tune. Rhodri concurred with him and added that there basically is no limit on the number of times the beam can be kicked, since the excitation levels required for the BBQ are in the range of or less than a micrometre.
Next, discussing the dispersion measurement, Frank mentioned that the machine aperture permits relative momentum offsets of up to +/-3.5e-3. The momentum collimators could, however, be tightened for machine safety to values between 1e-3 and 1.7e-3 (present design setting). The dispersion is measured by recording the orbit shift with varying momentum. The BPM resolution is expected to be better than 20 micron for pilot bunches in orbit mode, which is adequate. Even if the momentum collimators are closed by a factor two with respect to their setting at injection, the transverse normalized aperture remains large.
Jan pointed out that a limit on the maximum acceptable momentum offset equal to +/-2e-3 is set by the energy acceptance of the beam dump system. This is consistent with the present collimator settings at 7 TeV. From Jan’s statement, Ralph inferred the implied need of closing the momentum collimators already during the ramp. Jan mentioned that as an alternative one could rely on an interlock to prevent too large momentum errors. He asked when and how such interlock could be commissioned. Oliver remarked that the momentum aperture could also be limited by b5 errors resulting in a large third order chromaticity, and that as a result an average correction of b5 could prove necessary. Frank replied that only a small b5 effect appears to be expected at 7 TeV prior to the squeeze and that the third-order chromaticity are likely not to exceed the specified tolerances. After the meeting Stephane confirmed this assessment.
Elaborating on the measurements of beta function and linear optics, Frank recalled that the standard method for measuring beta functions in the LHC is to extract the phase beating from turn-by-turn BPM data (see the presentation by Rogelio in LHCCWG#8). This method faces two obstacles at 7 TeV. One is the limited amplitude of beam excitation: The maximum strength of the aperture kicker corresponds to maximum oscillation amplitudes of 1.6 sigma, or about 300 micron (see Jan’s presentation at the LTC on 15 June 2005); the tune kicker can apply a maximum kick of 0.7 sigma. The only device which could conceivable produce larger beam oscillations is the ac dipole. It has to be clarified whether the ac dipole will be available at this stage of commissioning (this question will be addressed in the following LHCCWG meeting). Machine protection issues are a related concern. Repeated large kicks of the same bunch may be problematic, as the bunch emittance will blow up. An option could be injecting several pilot bunches simultaneously and then kicking only individual bunches. The beta function at the critical collimator locations can also be verified by collimator scans and/or by K modulation in the collimator straights.
The second obstacle is the rms BPM resolution of ~200 micron ~ 1sigma for pilot bunches in trajectory mode. Rogelio and Rama have shown that the phase-measurement error scales with the ratio of the BPM noise over the oscillation amplitude (about 1) and decreases with the inverse square root of the number of turns. It is expected that oscillations at 7 TeV persist for a few 100 turns, before decaying due to decoherence. By studying many error seeds in simulations, Rama and Rogelio have also demonstrated that the optics correction works reliably only for peak phase-measurement errors of less than 1 degree. As an example, considering data taken over 400 turns with 1.6 sigma kick, a much larger peak phase error close to 8 degrees is estimated. Frank presented a couple of simulation examples. In summary the limited kicker strength and the large BPM noise for pilot bunches conspire to render optics measurements with pilot bunches extremely difficult.
A discussion ensued on the strength of the various kickers. Frank S. quoted Gene Vossenberg as a further reference for a maximum kick of about 1.4 sigma available from the aperture kicker, corresponding to 6-8 sigma at injection. By contrast the ac dipole could easily provide 4 sigma oscillations. Detailed requirements will need to be specified, but Frank S. has checked with Gene that such amplitudes are feasible. In particular, there is no danger for the machine, as the oscillation is not excited instantaneously by the ac dipole. Ralph nevertheless expressed some concern that the ac dipole might be dangerous in view of machine protection. Jean-Pierre replied that the typical rise time for an ac dipole is 100-200 ms and that its value can be chosen.
After the meeting, Jean-Pierre expressed his surprise at the “poor” accuracy announced for the measurement of the phase advance. He pointed out that for LEP, with 12 “turns” only, a similar BPM resolution but a much larger, 15-mm, oscillation amplitude, the phase advance was measured with 0.2 degrees accuracy (0.05 predicted). This high accuracy was obtained by a method of “quadratic detection” implemented by British radar engineers during the second world war. Both method and LEP results were discussed in a paper presented at EPAC’90 [A.M. Fauchet, J.-P. Koutchouk, Betatron Phase Advance Measurement in LEP, EPAC1990].
=> ACTION: Finalize procedure for beta function measurements at 7 TeV (Rogelio, Jean-Pierre?, Frank)
=> ACTION: Agree on kicker amplitudes which are safe and/or useful
Frank now proposed a measurement sequence, developed together with Ralph, which gives priority to machine protection. The first step is to close the collimators by a factor 2, already during the ramp. Then, at 7 TeV, the local beam sizes at all collimators are measured with collimator scans to roughly check the optics and to re-adjust the jaw positions, so that secondary collimators indeed act as secondaries. If the inferred beta beating lies within a safe range (<100%), the bunch intensity can be raised to about 3e10 for beta measurements with either aperture kicker or ac dipole. At this intensity the BPM noise is only 50 micron, promising a peak phase error of less than 2 degrees using the aperture kicker. If this resolution should still turn out to be insufficient, the measurement could be repeated with the ac dipole.
After the meeting Jan commented that the aperture kicker can only be used with “safe beam”, which at 7 TeV corresponds to a pilot bunch. Therefore, kicking a higher-intensity bunch is not an option for improving this type of optics measurements. The only remaining possibility appears to be the ac dipole.
Taking a look at the synchrotron-radiation monitors, which will need to be switched from undulator radiation to dipole radiation, Frank reported that according to Rhodri no change in the beam size is expected during this transition. Calibration of the monitor requires time with “stable” circulating beam. Most synchrotron-light studies can be done parasitically if the emittance blow-up is limited. A cross calibration with wire scanners will be required at both beam energies.
Ralph asked whether we can measure the emittance, which, he stressed, should be the first quantity to measure. Rhodri answered that measuring the beam size is easy, but that for the emittance a calibration would be needed. Replying to question by Stefano on the maximum intensity at which wire scanners can operate, Bernd D. explained that the wire-scan limit arises from secondary particles hitting cold magnets downstream. The exact limit is presently under study, and it seems to be lower than previously thought (which was 1/10th of nominal total intensity).
Mike asked whether we should stop on the ramp. Ralph replied that we can always close collimators, for which no stop is needed. Although stops are not explicitly required for the beam dump and machine protection (Brennan), Mike proposed that stops in the ramp be planned as a way of carefully staging energy increase. Basic optics checks could be performed at these intermediate points, along with machine protection checks and the requisite beam dumps.
Towards the end of his presentation, Frank returned to the issue of the D1/D2 transfer function, which for 450 GeV had already been addressed at LHCCWG#10 and earlier by Oliver at Chamonix XII. Frank reiterated that the D1-D2 transfer function errors can have a significant effect on the closed orbit during the squeeze (10 units error causing a 3 sigma orbit change at triplet). Local correction of this effect requires a careful analysis and distinction between D1/D2 transfer function errors, triplet alignment errors, and triplet gradient errors with crossing-angle bump offsets. On either side of the IP three common BPMs are located within the first 60 m. At 150 m, close to D2, the first separate BPMs for the two rings are found. Frank proposed (1) to assume that the alignment was done at injection (i.e. that BPM readings for straight reference line were identified, and quadrupole offsets determined); (2) to consider the main error after the ramp as being caused by the D1 & D2 transfer function uncertainty (if necessary one could perform Q1/Q2/Q3 K modulation to verify the quadrupole misalignments), and (3) to correct the orbit of the incoming beams for ~0 orbit and ~0 slope upstream of D2, and finally use the D1 and D2 dipole strengths to steer both beams onto the same orbit at 7 TeV. The open questions related to the D1/D2 transfer function problem remain the same as in the previous LHCCWG discussion on this topic: (i) the misalignment of mechanical and magnetic axes of the low-beta quadrupoles by 0.1-0.2 mm in x, and 0.5 mm in y (the beam needs to be steered through the mechanical center); (ii) the BPM offsets causing an error of up to 5 mrad, compared with a total dipole deflection angle of 1.5 mrad, resulting in a relative error larger than the desired precision of 3e-4, and (iii) the difference in the BPM offsets for beam1 and beam2. The D1/D2 issue may need to be settled prior to the squeeze.
=> ACTION: Finalize procedure for D1/D2 transfer function measurement & two-beam orbit steering.
Frank recalled that a longitudinal blow up is required for nominal LHC operation and he suggested that such blow up may need to be put in operation during commissioning if the beam is found to be longitudinally unstable. He then presented a draft flow chart for the complete 7-TeV measurement procedure.
Finally, Frank drew the following conclusions: The optics errors at 7 TeV can be significant and are different from those at 450 GeV/c. Both orbit and optics will differ. We should close the collimators during the ramp and perform collimator scans prior to any beam excitation at 7 TeV. Feedbacks, if available and active, may take care of (most of) the orbit, tune and coupling correction. The limited strength of the aperture kicker together with a 200-micron BPM noise does not allow for clean optics measurements with pilot bunches. Either higher bunch intensity or the ac dipole will be needed. The dispersion and chromatic properties can be measured by radial steering. Other issues include the D1/D2 (D3/D4) transfer functions, the feedback calibration, and the choice of beam parameters.
Massimiliano asked to which level the optics must be known, referring to the TOTEM experiment which assumes that the optics is known to within 3%. Jorg and John commented that the optics between two points, which is the only relevant information for TOTEM, can be known, in principle, to a much higher accuracy.
Helmut’s presentation covered the following topics: IP parameters at 7 TeV both for the squeezed and un-squeezed optics; getting the beams into collision; bunch-by-bunch variation; beam-beam effect; separation scans; absolute luminosity from machine parameters, and the communication between machine and experiments. He presented a list of nine related earlier presentations, where more detailed information can be found. The rms IP beam size at 7 TeV varies between roughly 100 micron for beta*=11 m, and 30 micron for beta*=2 m. From the formula describing the dependence of the luminosity on the transverse beam-beam offset, Helmut concluded that to “see collisions” in the luminosity signal the radial distance between the two beams should be less than 2 sigma (this offset translates into 37% of the peak luminosity). For the un-squeezed optics, this normalized distance corresponds to less than 133 micron absolute distance in both transverse planes.
The pick ups around the IP can be used to bring the two beams onto the same orbit. The expected resolution for the relative transverse position of the two beams at the IP is 100-200 micron (after K modulation this number may decrease to ~50 micron). The offset resolution is mainly limited by the separate BPM electronics for beam 1 and beam 2. The estimated 100-200 micron for the pilot-bunch BPM resolution should be (just about) sufficient to get the two beams close enough to observe collisions in some of the luminosity-related signals. Helmut recalled that a request has been made for an improved BPM system at the IP, which is in any case needed for high-beta operation of TOTEM and ATLAS (assuming 5-micron and 10 micron resolution in their respective TDRs).
Helmut now explained that for operation with zero crossing angle and a small number of bunches, it should be possible to use non-directional button pick ups with identical electronics for beam 1 and beam 2, aiming for 10 micron resolution in the IP offset. Jean-Pierre remarked that the sensitive button BPMs will easily be perturbed by collision debris. Helmut replied that these buttons would only be used to bring the beams into collision, when collision debris is not yet generated. Massimiliano asked whether ATLAS requires the high resolution only for high beta running at low luminosity. Helmut answered that this indeed was the case. He reiterated his recommendation of an early approval of the button pick ups, since these combined BPMs would prove extremely useful for early physics running. Also the installation of such pick-ups would be easier as long as the IR zones are not highly irradiated. Roger suggested that this be followed up at the LTC.
=> ACTION: Follow-up of two-beam IR button BPMs at the LTC (Oliver)
Ralph S. suggested that a van-der-Meer scan could provide the relative offset for the two beams. Helmut agreed, saying that this and other schemes would be mentioned in the following parts of his presentation.
Helmut next looked at the longitudinal position. The latter appears not to be too critical for the commissioning due to the large beta*. LEP experience indicates that it may be important to understand how to adjust the longitudinal position and how to detect offsets occurring later. From a discussion with Rhodri the adjustment and detection look possible by adding electronics. Namely, using a new electronics card for the BPM system, the relative beam arrival times could be determined with sub-ns resolution. Next Helmut commented on the beta* and waist measurements. The tune shift resulting from a 1e-5 change in the integrated K value of Q1 is 0.0016 for 11 m beta* and 0.008 for 2 m beta*. Such tune changes can easily be detected by a PLL. They can be used for inferring the waist location. One also needs to make sure that the minimum of beta* for each beam coincides with the collision point. These were critical issues at LEP, where all four experiments wanted to get equal luminosity. However, they are unlikely to be an issue for the LHC in view of the much larger bunch length and much larger beta* value. In particular, this question will not be relevant during commissioning.
Following this point, Helmut addressed the effect of bunch-to-bunch variation. He specified that our aim is achieving less than 10% variation in intensity and less than 20% in emittance. Data for individual bunches could prove valuable for a detailed analysis and the optimization of lifetime, background and stability, whereas for optimizing the collisions and the total luminosity the average measurements over all bunches are sufficient. Gianluigi commented that at the moment no bunch-by-bunch emittance measurement is available in the SPS, but that the emittance average over about 10 bunches is measured instead. Helmut considered that scrapers could be used to equalize bunches. Unfortunately, the SPS scrapers are not really designed for bunch equalization.
Now Helmut turned to separation scans, also called van-der-Meer scans (which were first introduced at the ISR). At LEP a scan took about 5 min per IP. The scans should be faster in the LHC, but are here needed in the two planes, x and y. The commissioning of the scans is simple; they are orthogonal in x and y. Replying to a question by Thijs, Helmut detailed that the scans in LEP were done in a stepwise manner, as they will also be at the LHC. He assumed a duration of about one minute per step, conceding that this is not a fast procedure, but could imply some luminosity loss and additional background. He emphasized once more that higher-resolution BPMs would help to speed up the procedure.
An alternative method of bringing the beam into collision utilizes the beam-beam effect. An advantage of this method is its independence of energy and beta*. The method can already work for pilot bunch intensity, with a beam-beam tune shift of 1.6e-4. Helmut referred to former applications of beam-beam signals for collision optimization at the ISR and at HERA. Frank commented that this scheme had first been invented and used at the initial double-ring version of DORIS-1 by Anton Piwinski [DESY H2-75/3 (1975)]. Jean-Pierre stressed that the beam-beam method is much more sensitive than scans based on luminosity. A clear signal is seen even at 4-5 sigma separation, and the scans are much faster. Helmut cautioned, however, that this method would require some preparation, such as setting up a PLL.
The next item addressed by Helmut is the calculation of the luminosity from the machine parameters, which has been requested by the experiments. The machine can provide an estimate of the absolute luminosity. The crossing angle is not an issue in commissioning, and achieving an accuracy of 20-30% should be easy. A PhD thesis project on the LHC machine luminosity determination will start this autumn. Massimiliano asked whether this luminosity estimate will be based on the van-der-Meer scans. Helmut replied that it will be based either on the van-der-Meer technique or on beam-beam scans, or on both. Ralph S. commented that three measurement points may be sufficient for maintaining maximum luminosity.
Adding a few words on the experimental conditions, Helmut pointed out the need for good, or at least acceptable, background in the experiments. An efficient communication between accelerator and experiments is essential in this respect. The technical preparation of the signals to be communicated was done in the frame of the LHC Experiment-Accelerator Data Exchange working group (LEADE) and it is now followed up within the LHC Machine Induced Background working group and the LHC Experiment Machine Interface Committee (LEMIC). Normalized figures of merit for the background from both beams are already included on the proposed LHC status page.
Helmut concluded that getting collisions and optimizing luminosity at 7 TeV is the main commissioning objective. We could profit from an improved BPM resolution, which anyway is necessary later on for high-beta operation. The luminosity monitors BRAN can be used for collision centering. An alternative method is based on the beam-beam effect. Data from either method can be used to compute the absolute luminosity. Equal sharing of luminosity will need to be ensured. The importance of low or at least tolerable backgrounds was stressed.
Massimilano asked whether the injection into the right bucket is guaranteed. Helmut answered that in LEP the injection was stable for long periods of time, but once in a while bunches ended up in the wrong bucket, which was very difficult to detect.
=> ACTION: Follow up identification of wrong bucket injection (Helmut)
Thijs discussed the possibility of using various radiation and beam-loss monitors around the IP for steering the beams into collision at 450 GeV. The four types of monitors he considered are BLMs, RadMons, BSCs, and BCMs.
The BLM ionization chambers measure the energy deposition by radiation in air. Dose rates and absolute doses are available on-line. There are 4000 BLM devices distributed around the ring and their signal strength varies between pA and mA. The Radmon detectors measure the energy deposition by radiation in Si. About 300 RadMon devices are installed. On-line signals available are dose, dose rate, hadron flux and the equivalent neutron fluence. Thijs presented photos illustrating the BLMs and Radmons located near quadrupole Q1 in the tunnel. The BLMs are attached to a quadrupole. The RadMons are movable units, close to the floor. They can be relocated closer to the IP if needed. The third type of monitor is the beam condition monitor (BCM), which measures the energy deposition by radiation in C (diamond produced by Polycrystalline Chemical Vapor Deposition (PCVD)). The BCMs are characterized by a very good time resolution, in the ns to micro-second range, and they are located closest to the IP at a distance of 1.8 m. The last type is the beam scintillator counter (BSC), consisting of paddles and disks, and installed about 10 m from the IP, in front of hadron forward calorimeter. With 3-5 ns its temporal resolution is also superb.
At 450 GeV with 11 m beta* and smaller than nominal emittance (2.5 micron), the single-bunch luminosity is 2.5e27 /cm^2/s luminosity. The nominal dose rates and particle fluxes in the TAS-Q1 region of IP1 and 5 were calculated by N. Mokhov et al [N.V. Mokhov, I.L. Rakhno, J.S. Kerby, J.B. Strait, Protecting LHC IP1/IP5 Components Against Radiation Resulting from Colliding Beam Interactions, CERN LHC Project Report 633]. The ATLAS particle flux for a luminosity of 2.5e27 /cm^2/s has been computed by M.Schupe, and the corresponding CMS particle flux by M. Huhtinen. These simulation data provide an estimate of the expected fluxes for commissioning.
Thijs next presented the measured monitor response rates,
which include proton and neutron data. A calibration of BLM and RadMons with 180 MeV neutrons was
performed at TSL in
Thijs now proposed a scenario of using the BSC monitors for the 450 GeV commissioning. He presented example data from ZEUS, pointing out that a higher number of bunches improves the time resolution. He estimated that it may be possible to resolve changes in the beam-beam distance at the 2 sigma level, but that a 1 sigma resolution in relative transverse position can only be expected at higher luminosity.
Thijs concluded that the radiation levels at 2.5e27/cm^2/s will be extremely low. If the background is sufficiently benign, collisions with more than two bunches will increase the total luminosity and the monitor signal. If the background is high, the bunch current should be raised first to increase the difference in signal from the BSC between incoming and outgoing beam. The latter is a direct measurement of the luminosity. Open issues to be addressed include the resolution and response of the BSCs, for which more information is needed from CMS and ATLAS, the BSC data exchange rate (5 Hz had been proposed), and the optimized locations of BLMs and RadMons in the LSS (downstream of Q1).
Replying to a question by Roger, Thijs confirmed that the monitor commissioning is done parasitically. Emmanuel commented that at the moment no connection exists between the radiation monitors and the accelerator control system. Helmut remarked that these monitor signals could be important for the engineering run. Frank commented that single-bunch luminosities of 1.1e30/cm^2/s, where a signal is expected, cannot be reached at 450 GeV, the maximum with nominal intensity being 1.4e28/cm^2/s. To reach luminosities of 1.1e30/cm^2/s the number of bunches will need to be increased, given that that the maximum bunch intensity is limited. Massimiliano mentioned that the message was passed to the experiments that the machine luminosity monitors are marginal at 450 GeV. Roger announced that the 2008 commissioning strategy will be decided in May and presented to the Council in June, recalling that two possible approaches are on the table.
=> ACTION: Further discuss data exchange with experiments (Thijs, Helmut)
Tuesday April 10th, 14:30
CCC conference room 874/1-011
Provisional agenda
Minutes of previous meeting
Matters arising
Beam excitation:
- Using the AC Dipole (Javier Serrano)
- AC dipole as multi-purpose instrument
(Stephane)
- Aperture kickers (Frank S.)
Commissioning of accelerator system -
vacuum (Frank)
Documentation and procedures
AOB
Reported by Frank