Tuesday May 22nd,
14:00
CCC conference room
874/1-011
Roger excused Jan who was away this week and, therefore, needed to cancel his report from the sub-working group.
He next announced that there are three talks from LHCCWG members at the next LHC MAC (namely by Jan on machine protection, Mike on the sector test, and Ralph on the collimation system update).
Concerning the presentations at the LTC, the following day Massimo would report on squeeze, two weeks later Frank on optics measurements at 7 TeV, and another two weeks afterwards Helmut on top energy collisions. A comment was made that we should also cover the 450 GeV run at the LTC.
The first EDMS documentation on commissioning procedures is ready to go. Target days for the following EDMS documents need to be set. The whole enterprise should be completed in the coming months.
On the repair of the triplet, Jean-Bernard Jeanneret alerted us of a potential problem related to the fiducialization of the triplets which will be de-cryostated during repair. Massimo described the details. The proposal had been made not to redo the surveying.
The survey measurements are normally performed without the beam screen. A new tool is required to measure with beam screen present. Massimo stressed that we do need a clear measurement of geometry and alignment. It was decided that the LHCCWG and LTC should push for such measurement. Oliver had already asked Luca Bottura to give a pertinent presentation at the LTC. Massimo emphasized that quick action is required. The production of a new mole would take at least a couple of weeks. He stressed that our goal is to know the magnetic and mechanic center of all triplet quadrupoles. Ralph suggested that we could apply “K modulation” to determine the magnetic center. Massimo replied that we already know that the differences with respect to the mechanical center are large, of the order of 0.5 mm. The topic would also come up at the MARIC the following morning, as well as at the ICC on Friday. In addition Luca would also present the issues to the MEB.
Roger introduced Alick MacPherson – our seventh EIC. The LHCCWG welcomed Alick on board.
Helmut described the two motivations for this presentation, namely to have a first discussion on the topic of experimental conditions and background at the LHCCWG and to serve as a dry-run for his talk at the LHC MAC on 14 June. At the LHC MAC he will also address optics and beam control, and the procedure for bringing the beams into collision. The latter will also be discussed at the LTC on June 20. Helmut next reminded the team that the aim of LHC is to deliver usable luminosity to the experiments, where “usable” refers to stable conditions and low background. He considered possibilities of active background control such as moving collimators for background reasons if needed. Helmut encouraged any further input and suggestions for the LHC MAC presentation.
Machine-detector interface issues and background calculations were already presented by Emmanuel Tsesmelis at the LHCMAC17 on 9 September 2005. This presentation was complete, and Helmut plans no repetition. In addition, there is no need to discuss collimation in his presentation, since this topic had been thoroughly covered already. For these reasons, Helmut’s idea is to focus his talk on commissioning, addressing the question of how to deal with backgrounds empirically, using existing hardware and instruments. The top priority is of course safe operation. Helmut cautioned that LHC would quickly enter into a new parameter regime. He stressed that we should be in a position to quickly diagnose background problems and allow for some flexibility.
Following
a discussion at and after
The
framework of background studies at the LHC is similar to the one at LEP. Reyes
and possibly other EICs will be involved, as well as
the LHCCWG, the run coordinators, the collimation team and the physics
coordinator.
Returning to the ultimate objective of low and stable backgrounds, a naïve expectation would be that since the luminosity increases with the square of the beam current, the signal to background-noise ratio should improve when the luminosity is increased. However, this behavior is not always observed in actual machines. The contrary dependence is more common, that is the backgrounds often rise dramatically as a function of the beam current. The LHC background may benefit from two factors, however: The LHC machine is designed for very high luminosity and intensity, and it requires extremely clean operation in order not to induce quenches. In addition a staged commissioning is foreseen, in which currents are raised only gradually.
Now discussing background sources and their optimization, Helmut explained that background primarily arises from beam-gas interactions, from halo, and from the collisions themselves. Mitigations include a vacuum pressure as low as possible, particularly around the experiments, the minimization of halo production and maximization of the cleaning efficiency, and an adequate control of the beam-beam collisions. He recalled that under good conditions the ISR featured a lifetime of several months, while a beam lifetime of a few weeks already resulted in severe background problems. He quoted a note by himself and Rudiger on beam scraping at injection into the LHC (CERN-AB-2004-032). At the LHC, also heating and vibrations must be minimized.
As for collimation and background, Helmut emphasized that the collimation systems in other machines are mainly used to reduce the background. By contrast the LHC collimation is designed for a high cleaning efficiency.
Ralph explained the reason for this difference. Namely for the LHC no request had been made to install any collimators for background control, and as a consequence no collimators for background control are included in the LHC design.
Helmut continued his presentation, pointing out that it is not evident what to do with the LHC collimation system in regard to backgrounds. Setting up 100 collimators empirically for optimum background conditions is not realistic. Some flexibility will be needed. Helmut recommended (1) allowing for some cleaning or scraping with the primary collimators; (2) opening the tertiary collimators to investigate their effect on the experiment; and (3) preparing collimator settings for both nominal and reduced emittances. The motivations for the various recommendations are that we would like to check occasionally that there are no significant tails at 4-5 sigma, that tertiary collimators could also be a source of background, and that settings for smaller emittance should help for background control.
Mike asked whether any simulations were done to study these effects. Helmut replied no, not by him personally. However, the collimation working group has done general simulations. Ralph commented that smaller gaps are difficult to realize, since the jaws are already nearly “colliding” in their nominal positions. The impedance also increases as the third power of the inverse gap size. All tolerances will become tighter. Tightening the settings could therefore perhaps be attempted after a few years of operation when many aspects are understood, but not in the commissioning. While Ralph agreed that this proposal could be tried at least in an MD, he stressed that for operation with tighter settings similar arguments hold as for the Roman pot experiments. Helmut argued that flexibility will be desirable. Ralph shared this view. And he announced that the LHC will have no collimation police, and that the collimation system is intrinsically very flexible. Nevertheless he advised against too daring experiments.
Helmut highlighted that the LHC is a complex machine with complicated insertions, two beams, alternating crossing angles, separation bumps and asymmetric apertures. A step-wise approach will facilitate the understanding of the background.
Ralph pointed out that asymmetry between experiments indeed is a very important point. As the phase advances from the background sources to the various IPs differ, the background can be different. Also with respect to the collimation system there is no symmetry.
The experiments in IR2 & IR8 will only see an increase in background and no increase in luminosity, as the luminosity is raised in IP1 and 5. Spectrometers lead to asymmetric crossing angles and may require switching polarities once per week, further complicating operation and possibly enhancing sensitivity to off-momentum background. The shift in the longitudinal position of IR8 adds yet another asymmetry.
Quoting the report CERN 77-15, “ISR Performance for Pedestrians,” by K. Hubner, Helmut illustrated that the background conditions can worsen with increasing beam current, as it was the case at the ISR. LEP1 operation observed background spikes. For LEP2 synchrotron radiation was a main concern. Continuous background monitoring and follow-up was needed. The study of beam-induced backgrounds at LEP was organized by Georg von Holtey. Helmut plans to play a similar role at the LHC.
Helmut now reported on the need for robust, normalized figure of merit signals, which were discussed in the various machine-detector interface working groups (LEMIC, LEADE, and LHC Background Working Group). A number of questions by Helmut have been circulated to the experiments, including the question: which detector components are planned to be used for the normalized BKG1.2 signals?
Roger asked who runs the LHC machine-interface etc. working groups. Helmut replied that these working groups are led by Emmanuel Tsesmelis.
A normalized background signal of “1” would indicate good conditions; a signal larger than “5” would imply very bad background.
Massimiliano asked why the signals are integers and not floating numbers, and also how these signals will be defined, e.g. if their definitions can be changed at any time. Helmut answered that these signals may indeed be presented on a "continuous" scale and that the background signals fall under the responsibility of each experiment. Ralph added that the properties of the background signals were specified at LEADE. They should be provided continually at a rate of 1 Hz. These signals are included in the collimation application software, as they will provide useful information for understanding the sources of background.
Alick remarked that the signal should not stop at a value of 5. Thijs commented that more details could be useful rather than a single number. Helmut responded that only a minimum has been specified for easing the accelerator operation. Witold made some comments based on his experience with PEP-II background. He stressed that normalized quantities are extremely useful. Their introduction represented a “quantum leap” in PEP-II operation and communication (where floats are used, not integers). At least 4 or 5 signals from different parts of the Babar detector were needed. He also recommended that the LHC infrastructure should allow for the combination of several signals. Helmut reassured the team that this full infrastructure is available. The few signals proposed here provide an additional guidance.
Ralph asked whether the experimental background signals are already defined and if we know how to access them. Mike and Massimiliano replied to the second question that, yes, we know how to get and read the signals at 1 Hz. They stressed however that the signals themselves must still be defined by the experiments. For illustrating possible difficulties, Massimiliano mentioned the example of a few bunches with very high background, but far from any trip level. Helmut replied that the background signals are normalized to the beam current. Massimiliano remarked that complicated signal processing may render the situation not easily understandable.
Roger recommended that Helmut and Massimiliano follow up the open questions on the background signals.
=> ACTION: Follow-up on background signals (Helmut, Massimiliano)
Helmut next presented the proposed LHC status page. The preparation of operational procedures for the status page is being done by the EICs and OP. Summing up his presentation, Helmut concluded that background and experimental conditions are expected to be a central issue and that their continuous follow up in operation is being prepared.
Massimiliano commented on the historical teletext pages, asking whether there will also be a web page providing additional information including graphs, e.g., showing the recent luminosity history. Oliver commented that the displacement of IR8 does not amount to 3 rf wavelengths. After the meeting Mike clarified that IR8 is displaced by 15 rf wavelengths.
Mike asked why only the background at LEP and ISR is considered as reference, but not e.g. the background at HERA and the Tevatron. Helmut replied that he had less experience with these other machines, but that he was going to follow up this excellent suggestion. Oliver remarked that HERA as electron-proton collider might be special. Ralph and Jorg commented that the Tevatron performs a scraping at the beginning of every fill, where each collimator is set up individually. Ralph reiterated that for LHC this type of adjustment was never requested. Such schemes would become important if we had serious background problems without quenching. A main argument for not implementing collimator-based background control at the LHC is that the background from IP debris is expected to be enormous in comparison with all other types of beam-induced background.
Offline feedback to Helmut is strongly encouraged.
This talk on cryogenics was prepared in collaboration with Stefano. Important help came from Luigi Serio and Adriaan Rijllart. Some other LHCCWG members also contributed with comments. The talk was organized as follows: overview, instrumentation and signals, cryo-organization during commissioning, application software, cryogenics & powering, cryogenics & beam commissioning, items which could go wrong, tools needed, and summary.
Gianluigi first reviewed the layout of the cryogenic system, showing two images including magnets, distribution line, refrigerators, interconnection boxes, and storage units. The refrigeration units are located at 5 points in total. The separation of point 2 from the additional odd point 1.8 is an exception resulting in no redundancy for sector 2-3. No fast cooldown is therefore possible in this sector. A diagram of pressure vs. temperatures visualized the states of the helium in the different parts of the cryogenic system.
The available instrumentation comprises pressure gauges, temperature sensors, level gauges, valve opening, and “virtual” flow meters. It includes about 8000 devices in total. The uncertainty of the temperature measurement is a few hundredths of a K.
Gianluigi now turned to the cryogenics organization during beam commissioning. At the moment there are insufficient resources for the full (24-h) on-line support which would be desirable. Continuous presence of cryo-operators could limit the recovery time, may avoid beam dumps, and could even prove essential in case of “teething” problems. Off-line support will be available in form of the cryogenics performance panel (CPP), chaired by L. Serio, which will provide crucial feedback for the beam commissioning.
Concerning applications software, a high level of detail is available in the CCC. In particular, the possibility exists of navigating through the cryogenic system. There are four access levels. The monitoring level is recommended for LHCCWG members. Presently under deployment are role-based access rights.
Starting from the navigation bar of Sector 7-8, Gianluigi demonstrated how to retrieve information on the structure and status of the cryogenic system in the tunnel. As an example he looked at the inner triplet. Summary information, such as the temperature over a sector, can be displayed as bar charts. Other signals are the cold mass temperatures, pressures, line C temperatures, and helium levels. Different colors are used to highlight if thresholds are exceeded. Another important display option is trends.
Ralph asked whether trends can also be used to predict the future, e.g. allowing for an extrapolation to the quench recovery. Luigi replied that such extrapolation of trends is indeed foreseen, but not yet available. So far the signals are only used for cryogenics control. Reyes asked whether only monitoring is possible at the moment. Gianluigi answered that, no, also expert control is available, but he did not dare to use it.
Another source of information on the cryogenics is provided by the post-mortem (PM) system. Cryogenics PM analysis is done based on check functions defined by experts. There are four PM event triggers (CRYO_START, CRYO_MAINTAIN, ALARM, and QUENCH). Triggering can be fired on request. Therefore, the PM can also be used for analysis purposes. Gianluigi highlighted that the tools available look extremely flexible. In particular, it should be possible to add “beam-oriented” logics to the PM analysis. He also mentioned that the PM application at present retrieves its data from the logging database. A snapshot of the cryogenics post-mortem application was shown. The software offers the possibility of saving to, and loading from, a file. This feature can be used for keeping data over more than a few weeks.
Different cryogenics conditions are being distinguished for powering. Three logic states are used to indicate the status for each powering subsector. CRYO-START and CRYO_MAINTAIN can equal true or false. If CRYO_MAINTAIN is false, the request for a slow current discharge of the magnets is issued. Gianluigi explained that no direct connection exists between the cryogenics and the BIC but only with the PIC. The insulation vacuum is directly interlocked to the cryogenics, but not the beam vacuum.
Ralph asked how much temperature margin exists at the moment the beam will be aborted. Luigi replied that a temperature of 2.05 K is the threshold for the CRYO_MAINTAIN. At 2.17 K the system would no longer be functioning. Ralph now inquired whether the numbers for cooling capacities with regard to electron-cloud heat–load presented by Frank at various occasions were too pessimistic. Oliver answered that the numbers shown by Frank referred to the heat load which can be coped with, and not to the temperature.
Gianluigi underlined the assumption that the cryogenics systems will be fully commissioned during hardware commissioning. The main remaining unknown then is the interplay of the beam with the cryo-system.
The beam will give rise to an additional heat load on the beam screen, due to resistive-wall heating, synchrotron radiation, and electron cloud, as well as to an additional heat load on the cold bore, due to inelastic nuclear beam-gas scattering, and other types of beam losses. He summarized the values of different contributions for the nominal beam in a table with numbers taken from Laurent Tavian’s presentation on LHC cryogenics at the LTC of 2 June 2004, and from Frank’s update on LHC electron-cloud simulations at the LTC of 6 April 2005.
Parasitic follow up of the cryogenics system during beam commissioning is expected. The CPP will steer the commissioning progress.
Ralph commented that the number for the halo particle losses must be updated. For special magnets this number can be of the order W/m, e.g. for a special cell in a dispersion suppressor. Luigi and Gianluigi replied that the numbers listed referred to a standard arc cell.
Critical
elements include Q6 in IR1 and IR5, standalone magnets
at 4.5 K, the MQTLs,
In principle the cryo-system is characterized by a high reliability. But sub-system failures cannot entirely be excluded. Gianluigi next addressed the question what could go wrong during beam commissioning. Quenches may be the routine. Typical recovery times vary between one and a few hours.
Oliver asked if the cryosystem needed to be changed for the quench recovery. Luigi Serio replied that, no, the system recovers by itself for most types of quenches. Ralph commented that for the quench of a full sector, the quench diode could be damaged. Luigi answered that he considered here only the cryogenics. Walter added that for diode survival the quench of half a sector would be worse than the quench of a full sector.
Gianluigi pointed out that a strong correlation is expected between the cryogenics and the vacuum, in particular in regard to vacuum transients.
Oliver asked at which level the beam screen temperature is controlled. Sergio replied that the control happens with a precision of +/-2 K.
Possible problems during commissioning include a heat load above specification. The resolution in the heat load deposited on the beam screen is 0.5 W/cell compared with an expected total beam-induced heat load of 280 W/cell. Gianluigi suggested that this heat load could be used to detect pressure bumps. Local heating of the cold masses can be measured with one cell resolution. Another problem that might occur is the electromagnetic interference of the beam field with sensors. Past SPS experience has shown that temperature sensors can be affected by high beam intensity.
Tuning now to the tools needed, Gianluigi commented that certainly a summary of the CRYO_START/MAINTAIN conditions for all sectors would be required. Following up trends after changing the mode of operation will be essential. In addition, we could specify analysis types relevant to the LHC operation, e.g., defining virtual heat loads on beam screens and cold masses, monitoring heat load and temperature and correlating these signals with vacuum, beam parameters, etc, and finally adding temperature/flow trends to identify critical behavior. At a later stage fixed displays could provide information needed recurrently.
In summary, some unexpected effects might occur with beam operation, 24-h cryogenics support would be desirable, and the interaction of LHCCWG with CPP and MPP should be strengthened so as to focus on the crucial elements and to refine the analysis tools for beam commissioning. For beam operation, heat loads appear to be the most meaningful parameters. Non-expert tools need to be “enhanced”, e.g., by allowing the PM analysis to access the measurement database.
Adriaan commented that the response of the cryo-system is slow. And that therefore there should be no reason not to take the data from the logging database in view of either speed or time. Gianluigi replied that apparently some filtering is applied on the data in the logging database which can change the trends by up to 50%. The difference between measurement and logging database is not entirely understood. Luigi added that another reason for using the measurement database is to enable faster decisions on how to react.
=> ACTION: Discussion on LHCCWG interaction with CPP and MPP
Tuesday June 5th, 14:00
CCC conference room 874/1-011
Provisional agenda
Minutes of previous meeting
Matters arising
Parameters and tolerances (Massimo)
RF commissioning with first beam
(Philippe Baudrenghien)
Sector Test (Mike)
AOB
Reported by Frank