Wednesday
September 6th, 14:30
CCC
conference room 874/1-011
Mike chaired the meeting. He transmitted apologies from Roger who was unable to attend.
In an email exchange after the 11th
LHCCWG meeting, Stephane clarified his comments made
during Stefano’s presentation on aperture measurements. Quoting
Frank reported an interest by
some ATLAS members (Charlie Young at SLAC, and Tobias Golling
at LBNL) in longitudinally displaced
collisions at +/3.8 m on either side of the main IP for aligning the
detectors. Such parasitic collisions could be produced, for example, with 25-ns
bunch spacing or by injecting into non-nominal buckets. He referred them to Massimiliano, in case they want to pursue their
request. After the meeting,
Enrico reviewed the nominal requirements for the four experiments and then recalled the luminosity formula. A 1 sigma separation of the two beams leads to a 20% reduction in luminosity. The LHC collision rate monitors (BRAN), also known as luminosity monitors, detect the flux of forward neutral particles. The initial luminosity originally considered for colliding two pilot bunches with 5x10^9 protons at beta* of 10 or 18 m was a few times 10^26 cm^-2 s^-1 at 7 TeV. The luminosity is proportional to gamma/beta. At 450 GeV with 11 m beta* the luminosity in collisions of two pilot bunches is some 10^25 cm^-2 s^-1. To reach the specification accuracy of 10% requires about 1 minute integration at 10^26 cm^-2 s^-1 luminosity, where 8 pp events are expected per second assuming 80 mbarn total cross section, Only a fraction of these events, approximately a quarter, will be detected in the luminosity monitor. The full specification of the luminosity monitors can be found in LHC-B-ES-0007 rev 1.1.
Different luminosity monitor concepts are being realized for the two high-luminosity experiments at IP1 and IP5 on the one hand, and for IP2 plus IP8 on the other. The monitors for IP1 and IP5 are fast ionization chambers (ICs) built by LBNL, which will be mounted in the 4 TANs. Around IP2 and IP8 CERN will install CdTe detectors produced by CSA-LETI. The main challenge for these monitors is to be sufficiently radiation hard to survive for 20 years of LHC operation. The reason for choosing different solutions was the different radiation levels, as well as limitations of the LBNL budget.
Enrico showed that the IP1 luminosity monitor will be inserted in slot 4 of the TAN. At the start up only LHCf will sit in front of it, since the ATLAS ZDC elements located on either side of the luminosity monitor will be installed at a later date. The situation in CMS is similar, except that there is no LHCf. The ionization chamber is surrounded by a 10-bar pressure vessel. The final IC design is ready. The first chamber is expected to be delivered to CERN at the end of 0206. The remaining chambers are due to arrive in spring 2007. The front-end electronics is well advanced. The predicted bunch-by-bunch resolution just meets the specification. Gas piping is being installed. It is thought that there will be a continuous slow flow of gas (mainly Argon) through the IC. The IC signal scales about linearly with the beam energy. At 7 TeV beam energy, the ratio of signal to electronic noise is about 30. Reducing the beam energy increases the relative level of the noise with respect to the signal. At 450 GeV beam energy, the IC signal may be comparable to the noise: While for 1 TeV neutrons 80% passing the IC of the events are above the noise threshold, for 100 GeV neutrons these are only 6.3% of events. Enrico flashed the installation schedule of all TAN experiments.
ACTION: Enrico or Helmut? – Follow up pp simulations at 7 TeV and 450 GeV with Nikolai or Mika.
In IP2 and IP8 the CdTe semiconductor detectors from LETI will be deployed. At these IPs there is no TAN, but the CdTe detectors will be installed at equivalent locations and placed behind a copper block of 4-20 cm depth. The manufacturing contract was placed with LETI at the end of April 2006. The front-end electronics is produced by FOTEC. It is planned to use the same digital acquisition boards as for the IC, profiting from the LBNL development. The CdTe monitors have a much higher sensitivity than the ICs. Therefore, noise is not an issue for them. However, a F.E. amplifier is needed in order to transport the signal over 50 m to the DAQ electronics.
Position readings of the BPMs near Q1 can be used for a pre-alignment of the two beams at the IP. The remaining uncertainty is about 200 micron, or about 13 transverse sigma. Enrico emphasized that the background conditions at this stage are largely unknown. Background estimates were provided by Bill Turner at LBNL for the nominal LHC with 10^34 cm^-2 s^-1 luminosity at 14 TeV c.m. energy. Extrapolating these estimates to 900 GeV c.m. energy with 10^26 cm^-2 s^-1 luminosity for 10^10 protons per bunch, one arrives at an event rate of 8 Hz (2 Hz counting rate), to be compared with a beam-gas background of about 1 event per second, and halo scraping corresponding to about 3 events. These background numbers assume a vacuum pressure of 0.1 nTorr.
Frank remarked that the pressure assumption of 0.1 nTorr may be too optimistic. Massimiliano referred to Vadim’s talk on background in the previous meeting. The gas background scales linearly with pressure. Ralph pointed out that there could be other important differences between the 7 TeV and 450 GeV conditions, for example, crossing angles. It was also mentioned, possibly by Masimiliano and Vadim, that the dominant part of the beam-gas scattering background, namely the one coming from the arcs and not from the IR, is missing in Bill Turner’s estimate. Emmanuel suggested that Vadim could check the numbers. Vadim pointed out that knowledge of the exact optics is needed for reliable background estimates. Helmut concluded that background can potentially be a problem.
ACTION: AT/VAC – Produce a realistic estimate of the vacuum conditions expected for the 450 GeV run.
Enrico mentioned that for high beta* values the signal-to-background ratio is worse. One possibility for improving the signal-to-noise ratio might be the use of coincident detection. However, whether the events occur in coincidence is unclear, with about 6 of 10 questioned experimentalists claiming there will be a coincidence, and the other 4 disputing this. Massimiliano remarked that the limited acceptance of the detector is an issue, while the dynamics is intrinsically symmetric. Enrico commented that the detectors are designed to be capable of coincidence detection. He later clarified that a luminosity of 10^26 cm^-2 s^-1 would the minimum measurable value from a purely statistical point of view, but that in practice this could be very difficult to achieve. Oliver asked whether these last numbers would be consistent with a signal-to-noise ratio of 30 for nominal conditions at much higher luminosity. Enrico responded that the ratio of 30 referred to the detection of a single event without averaging. For Mike the key message of Enrico’s presentation was that colliding pilot bunches causes serious problems with the detection and that we should collide beams of higher intensity.
Oliver mentioned that TOTEM could also measure the luminosity. Jean-Pierre recommended looking at alternative ways for steering the beams into collision such as detecting the coupling signal from the beam-beam interaction. He added that Schottky pick ups will definitely be ready and would have a high accuracy. Enrico mentioned yet another possibility, first put forward by Bill Turner, which is to install some a few sensitive luminosity monitors, based on scintillators, in IP1 and 5 for the LHC start up. Concerning the expected initial beam-beam separation due to errors, Oliver commented that the estimate depends on whether correlated or uncorrelated orbit errors determine this offset. Ralph expressed the concern that some errors may cause systematic effects, and lead to spurious responses in the luminosity monitor. Jean-Pierre replied that the expected angle resolution of the luminosity monitor is of order 20 microrad. Helmut added that the monitor acceptance covers the full range of crossing angle. Therefore, the angular acceptance is no problem initially when we operate with head-on collisions. The monitor detects if the beam orbits are not centered. Daniela remarked that there may be no empty place available in the TAN for mounting the supplementary scintillating detectors. Enrico answered that the scintillating component would need only a few centimeters of space. Emmanuel pointed out that scintillating counters are also being set up in CMS, and these could possibly be used for monitoring the luminosity. Enrico ended with the remark that until now, nobody has taken any decision on the proposed back-up solutions.
Helmut reviewed parameters, observables, beam-beam tune shift, and luminosity values, related to bringing the beams into collision. An important question already evident from Enrico’s presentation is at which intensity these collisions should happen. Other questions related to the procedure of beam steering, and to the communication and signal exchange with the experiments. Helmut urged the team to start with simple, flexible, and robust methods, emphasizing the importance of redundancy. Several independent measurements and methods are needed to find errors and determine the uncertainties. The analytical formula for the beam-beam tune shift shows that the latter is independent of energy and beta function. Therefore, the beam-beam tune shift may be a helpful signal at 450 GeV. It scales linearly with intensity, assuming a value 3-4 x10^-3 for a single head-on collision at design beam intensity. One way of utilizing the beam-beam interaction for detection of beam-beam separation is the method of “tune coupling” with excitation, which was employed at HERA.
Next, Helmut commented on some essential measurements. Measurements per bunch are needed, for example a bunch-by-bunch tune measurements, possibly based on a gated BBQ detector. Beam-size measurements together with bunch current and knowledge of head-on collision would allow computation of the absolute luminosity.
Thijs asked about the role of the Schottky monitors for steering beams into collision. Helmut replied that the main system is the luminosity monitors, but that redundancy is clearly useful.
As a reasonable goal for the 1st collisions, Helmut proposed 4e10 protons per bunch with single bunch. Enrico commented that a better signal is obtained by colliding more bunches, and Helmut agreed.
Helmut stressed that smaller emittances, like 2.5 micron, would help in many regards (luminosity, aperture, magnet imperfections). Such emittances may be available from the SPS. Helmut computed event rates for a cross section of sigma=10 mbarn, his estimate for the cross section characterizing the detection of high energy neutrons in a luminosity monitor.
Ralph asked why one does not consider a higher bunch intensity, e.g., 10^11 protons, which would be completely safe at 450 GeV. The signal/noise ratio is better for higher bunch intensity, and this could be a promising way to overcome limitations in the observation.
Quoting discussions with Enrico and Bill Turner, Helmut recalled that the BRAN specification was not meant for 450 GeV, where the monitor performance appears marginal. This motivated the proposal by Bill Turner to add a scintillator on each side of IP1 and IP5. Bill Turner also suggested a second workshop similar to the “TAN workshop”. Background rejection could be improved by coincidence detection.
Ralph wanted to know whether we can observe the collisions with the beam-loss monitors around the IR. Helmut: replied that this had not been looked at. Helmut now addressed the question how well we can get the beams into collision by steering based on BPM readings, which had been answered in an LTC presentation by Stephane in March 2001. Oliver asked why the formula for the expected beam-beam offset contains a multiplication by the factor sqrt(2). He pointed out that there should be some cancellation of errors when using difference readings. Frank asked whether indeed the same electronic offsets are expected for the two beams. Jean-Pierre commented that the cross talk between the BPM readings for the two beams is important at higher intensity. Helmut granted that his estimate of an initial separation 300 micron or 1 sigma may be conservative. Oliver and Jean-Pierre both pointed out some additional redundancy, offered by the additional two BPMs in each triplet, which could be used for improved interpolation and for identifying bad BPM readings.
ACTION: Bernd or Rudiger? - Collision detection with IR beam loss monitors?
Helmut now surveyed separation scans (“van der Meer scans”, pioneered at the ISR). He presented a few real scans from LEP based on detecting the luminosity as a function of beam separation. For the LHC, he showed a hypothetical scan measuring the changes in the tune-signal width. For the start up, he recommended simple orthogonal x and y scans.
Oliver suggested that there might be a hysteresis problem for the closed-orbit correctors in separation scans at 450 GeV, which could be significant. Walter agreed that this should be checked.
ACTION: Walter? - Hysteresis of orbit correctors used in separation scans at 450 GeV?
Lastly, Helmut addressed the status and communication with experiments. He showed an example of a real LEP TV status display, as well as an artist’s view of an LHC status display. The latter contains information on two types of normalized background, the precise definition of which will need to be discussed in.the LHC Machine Induced Background Working Group.
He ended his presentation on a friendly note. Since the BPM resolution is comparable to the beam size, steering the beams into collision should not be difficult. Separation scans relying on BRAN detectors are marginal at 450 GeV. Therefore a request has been made for additional scintillators and for left/right coincidences. Redundancy is important, and collision effects in tune signal are one example. The absolute luminosity for head-on collisions can be predicted from currents and beam sizes. The experiments are also expected to determine the absolute luminosity by themselves.
Massimiliano remarked that he once made a presentation describing an alternative way to deduce the absolute luminosity using the distribution of events in the detector. This method requires as additional input only the bunch charges with a good accuracy. Paul commented that the absolute calibration of the BCTs is not obvious.
ACTION: Rhodri? - Absolute calibration of BCTs?
Jean-Pierre drew attention to the method of excitation coupling and he explained that this method, which he had used at the ISR, detected the amplitude of the beam response and not the tune shift. He conceded that usually such tools become operational too late to be helpful for the commissioning. He added that for him this remark is a motivation to implement it in the LHC from the start to be helpful. Frank noticed that the tune coupling was used very early on in HERA, and Oliver added that the hardware was simple. Jean-Pierre suggested that the resolution be checked.
ACTION: Jean-Pierre and/or volunteer? – Check resolution of beam-beam coupling?
After the meeting, Jean-Pierre remarked that answering this question may be challenging. He is not aware that the beam-beam transfer function for bunched beams had ever been mathematically expressed; He received the same question from P. Cameron a few months ago, checked with A. Hofmann and they could not find any material on it. So this deviation may need to be done. If anyone else is interested, Jean-Pierre would be happy to help. Otherwise he would try to do it over the coming months (probably there is no urgency as the basic hardware, BBQ, is with us).
Mike asked about the coordination of this overall effort. Daniela responded that a workshop is foreseen either in November or January, addressing questions such as “can we use the detectors inside the TAN for bringing beams into collision?”.
Responding to a question of Frank, Enrico explained that if the luminosity signal and the neutron energies are too low, the coincidence measurements do not help but become a problem.
Concerning refined simulations of background conditions for the 11-m optics at 450 GeV, Vadim needs estimated loss maps for 450 GeV and the planned configuration of the collimation system. Producing such loss maps had been identified as an action item at the previous LHCCWG meeting, which had been forwarded to Ralph. Ralph wondered how much effort should be put into preparing the 450 GeV run. Vadim elaborated that if the tertiary collimators (TCT) are present, the local losses will be at these locations, but if not they will be at other places. Ralph proposed a pragmatic approach, i.e., considering no collimators at all, and not to spend any time on the collimation optimization. Massimiliano remarked that the experiments are interested in knowing the background conditions beforehand. Ralph agreed that the collimation team will produce loss maps for the case of using primary collimators only, i.e., for a single-stage collimation system; the secondary collimators will be inserted only if the background conditions are unacceptable.
ACTION: Ralph et al - Loss maps for 450 GeV
ACTION: Vadim? - Simulation of 450-GeV background conditions.
Paul brought up a topic which is
on the agenda for a future LTC on 26th of October, and which is also
a
ACTION: Paul and Oliver? - Specifications for as-built database. Recruitment of help.
Wednesday September 20th, 14:30
CCC conference room 874/1-011
Provisional agenda
Minutes of previous meeting
Matters arising
450GeV optics – IR aperture and IR bumps (Yannis Papaphilippou)
AOB
Reported
by Frank