Wednesday March 22nd, 14:30
CCC conference room 874/1-011
The LHCCWG now has a web site (http://lhccwg.web.cern.ch/lhccwg/), where minutes etc. are posted. A few links of interest, e.g., to the LHC operations project, are included. Comments and suggestions concerning this web site can be sent to Frank. A mailing list consisting of LHCCWG members was also created, with the name ab-project-lhccwg.
At the open session of the LHC experiments Committee (LHCC) on 22.03.2006, there were two talks. The first one by Lyn on the status of the LHC delivered an upbeat message. Only half of his time was spent on the progress of installation, the other half focused on the commissioning with beam. Lyn was optimistic that beam commissioning will start in 2007. The second presentation concerned LHCf (LHCf Design Report), a new experiment related to understanding cosmic-ray measurements, which would be integrated into the TAN on either side of IP1. This experiment might have an impact on the luminosity monitors. The LHCf experiment is also mentioned in Helmut’s presentation from the TAN workshop below, and it should be followed up.
→ future topic: LHCf
Responding to a request by Ralph in the previous LHCCWG meeting, Mike discussed the data format for the sector test. He described that the data are stored in the form of “self-describing data sets” (SDDS), inspired by a development at Argonne (APS SDDS web page), where this scheme is extensively used, including toolkit etc. Greg Kruk has provided a JAVA interface to the SDDS format. This data format was already successfully employed for the TI8/TT40 tests, for recording the shot-by-shot readings of BLMs, BPMs, screens, currents etc. Each SDDS data file contains a time stamp and the cycle number. The SDDS files can be accessed from JAVA, C, and C++ programs, as well as using the Mathematica/JAVA interface. The XPOC (external post operation check) application also uses SDDS data. The SDDS data is saved in binary or ascii format. In addition to SDDS, the LHC will also have a logging database and a measurement database, as for LEP, both based on ORACLE. The TI8 tests utilized both SDDS and the ORACLE databases.
A workshop on TAN integration was organized by D. Macina and E. Tsesmelis on 10.03.2006. Topics covered were the LHC luminosity monitors, the newly proposed LHCf detectors, the ATLAS ZDC and the CMS ZDC. The luminosity monitors are a US contribution, and were originally developed by Bill Turner at LBNL; at the workshop they were presented by Bill Turner and Alessandro Ratti At the TAN workshop, Helmut talked about the roles of the luminosity monitors for LHC operation. He emphasized that a reliable robust signal is needed, and that, based on LEP experience, the luminosity monitors are important early on. The impact of background and sensitivity to the beam orbit angle at the IP needs to be understood. He recommended an automatic background subtraction. The TAN contains several slots in which detectors can be installed, all close to the maximum of the shower. The luminosity monitors are assigned to slot no. 4. The LHCf detectors would be installed in front of these monitors for data taking with luminosities below 1030 and would be removed for higher rates, replaced with copper blocks. The presence or not of these detectors might change the calibration of the luminosity monitors, as might any possible movement of the detectors when in place. By contrast, the ZDCs of ATLAS and CMS, aimed at measuring the forward cross section, are installed behind the luminosity monitors.
Thijs mentioned that, also at the TAN workshop, Nikolai Mokhov reviewed the expected losses in the TAN, in point 7, and in other regions. A load of 220 W, compared with a total of 1.2 kW IP collision debris, ended up in undisclosed locations behind the TAN. Thijs recommended checking the power distribution early on. Paul suggested looking at the BLMs. The expected loss distribution should be discussed with Nikolai. Ralph proposed monitoring the TAN temperature.
→ future topic: loss distribution of collision debris
→ future topic: sensitivity of luminosity signal, e.g., to IP orbit angle or pressure.
Arguments in favour of 75-ns spacing include the reduced effect of long-range beam-beam collisions, a factor 2 smaller transverse emittance from the injectors – easing tolerances in the LHC - , a larger beta* for the same luminosity, higher luminosity for the same beam current, compatibility with partial installation of dilution kickers, no parasitic crossings at the Q1 BPMs, and the absence of electron-cloud effects such as pressure rise or poor lifetime. Paul and Jean-Pierre suggested that the crossing angle could be reduced thanks to the weaker long-range effect, the larger beta*, and the smaller emittance. An estimate after the meeting indicates a minimum crossing angle of about 90 mrad for 75-ns spacing, to be contrasted with 180 mrad minimum crossing angle for 25-ns spacing at a similar luminosity. Disadvantages of 75-ns spacing are a higher number of pile-up events, the possibility of single-bunch instabilities, and, as pointed out by Helmut, a shorter luminosity lifetime. Massimiliano suggested that the number of pile-up events is not taken as a constraint, but that a rapid transition to 25-ns spacing may be desired by the experiments. Paul cautioned that this might imply a temporary drop in luminosity. In any case, it seems that only a comparison of the running conditions for either bunch spacing will allow a decision by the experiments as to which one they prefer. An earlier request by one of the CMS collaborators for a period with 75-ns spacing has not been agreed upon by the rest of the experiment(s). Mike and Paul expressed the view that a step with 75-ns spacing is natural when going from 156 bunches towards higher intensity, and that this will allow for a more adiabatic transition than switching directly to 2808 bunches. As a tentative conclusion, a period of 75-ns bunch spacing is kept.
The magnet parameters and expected chromaticities depend on whether this part of the commissioning is done on the nominal or on the degauss cycle. It is assumed that the 1st turn was established and the 1st turn orbit corrected. The orbit is closed with two orbit correctors. The integer tune is obtained from a difference orbit. The debunching due to finite momentum spread and momentum compaction will limit the number of turns we can measure to about 140, which could be increased to >200 by using a low-emittance pilot bunch or a lower rf voltage in the SPS. The SPS experience shows that the fractional tune can be measured with a precision of 0.01 from about 16 turns. Jean-Pierre remarked that in LEP a resolution of .001 was reached over 100 turns.
→ future topic: simulation of beam signals and tune resolution in LHC
A large uncorrected chromaticity implies an additional fast decoherence of the signal. For example, with a chromaticity Q’ of 179 units the oscillation decays after 3 turns. If the chromaticity is corrected to 18 units, the decay time is still as small as 30 turns. Conversely, the decoherence signal can be used to correct the chromaticity. Gianluigi suggested that by gating the position reading on the centre of the bunch we could avoid the signal decoherence. Andy next discussed the rf capture and energy matching. The debunching is not critical for the longitudinal pickup. A typical slippage to be expected from initial energy errors is of order 10 rf periods over 0.5 s. In order to perform the rf capture, a circulating beam over >100 turns is required. It is assumed that the 1st turn is already centered. Incoming energy errors can easily be disentangled from betatron oscillations using the arc BPMs.
For energy matching, one has three variables, namely the magnetic field of the SPS, the magnetic field of the LHC, and the LHC rf frequency. One does not really want to change any of these three variables: variation of the LHC bending field could induce snapback effects, but using orbit correctors instead could be a viable option; changes to the SPS field requires retraining of the frequency program; changes to the LHC rf frequency imply retuning of the SPS rf and other side effects. Andy presented two equations which could be used to match in a single iteration by using two of the three variables above.
He emphasized that the rf frequency is normally locked between the two LHC rings. Though the two rf systems can be unlocked, re-synchronization afterwards – and therefore also collision -- is considered difficult. If the 2 rings are locked the beam in the two rings might have different radial positions. A 1 cm length difference for the two rings implies a 1.5-mm radial offset in one of the two rings, and it corresponds to a centre frequency shift of 150 Hz. Andy proposes injecting early on in the second ring to check the difference in circumference.
A discussion ensued on the lack of re-synchronization. RHIC, TEVATRON and HERA were mentioned as examples where the rf systems of two rings, or for two beams, are unlocked, cogged or locked, quickly and without any apparent difficulty. Though, in principle, one could measure chromaticity and dispersion always in both rings at the same time, LHC operation might be restricted if the momentum aperture is asymmetric. Jean-Pierre commented that one does not want to perturb a hadron beam with good lifetime, e.g., by shifting the rf frequency in case the second beam has a bad lifetime.
Elaborating on the frequency locking, Andy added that the rf systems in the two LHC rings use the same frequency reference, but are fully independent. So there is no hardware reason why we should not inject and ramp with different frequencies. However, a hardware limitation is in the SPS, which currently can only deliver LHC beam at one frequency for both rings. A change here would require duplication of hardware. Rephasing in the LHC at top energy is also possible, but the speed of rephasing depends on the aperture one is willing to lose by changing the frequency, e.g. applying a difference of 10 Hz between the 2 rings would allow 10 RF periods per second rephasing, i.e. 1782 seconds if one needs to rotate the relative position over half of the circumference. Of course one could go faster with a bigger frequency offset. Rhodri remarked that ramping the 2 rings with different frequencies implies random parasitic bunch crossings.
→ future topic: rephasing between rf systems of the two rings
→ future topic: energy tracking between sectors
Wednesday April 5th, 14:30
CCC conference room 874/1-011
Provisional agenda
Minutes of previous meeting
Matters arising
450GeV initial commissioning with pilot beam
AOB
Reported by Frank