Choose timezone
Your profile timezone:
MP3 meeting
Participants
Adriaan Rijllart
Andrea Musso
Antonio Vergara Fernandez
Bernhard Auchmann
Christian Giloux
Giorgio D'Angelo
Howard Pfeffer
Hugues Thiesen
Jim Strait
Karl Hubert Mess
Marco Zanetti
Markus Zerlauth
Marta Bajko
Matteo Solfaroli Camillocci
Mike Koratzinos
Nuria Catalan Lasheras
Reiner Denz
Remo Maccaferri
Robert Henry Flora
Rüdiger Schmidt
Sandrine Le Naour
Walter Venturini Delsolaro
Yves Thurel
MP3 meeting 17 June 2009
Matters arising
Nuria informed that she's now invited to the LMC meetings, and in the TEMB, where she will report on MP3 works
Powering tests have started in Sector 23, in future there will be a report on previous week MP3 activity in the CCC at each meeting
Report from the CCC (S. Le Naour)
Sandrine reported on the issues encountered during the first days of comissioning, in particular on the RCBYHS4.R2B1 circuit which quenches when the crowbar is activated at 72 A. This happens only when starting with a positive polarity. When reducing the current to 70A the problem disappeared, but it was still there when switching back to 72 A.
Giorgio showed the AC transfer function |Z(f)| of the correctors in the same cold mass (see slide); which looks sligthly smaller in the intermediate frequency range for the corrector concerned.
Yves commented that a similar behavior was observed on a bunch (6-7) of correctors of the same type in 2008, and in some cases the magnets would quench on the ramp down instead of upon crowbar activation.
In other words this is not the worst case: indeed the behavior of RCBYHS4.R2B1 could be accepted for beam operation.
Differently from 600A circuits quenching during discharge, in this case the mechanism is not understood.
Yves has already done some analysis work on the subject.
ACTION –> Walter, Yves and Giorgio will look further into the subject and report back to MP3
nQPS test procedure (R. Denz)
Reiner presented the goals and plan for the foreseen test of the new QPS on the weekend 27-28 June (slides)
The objectives of the test include noise measurements, assessement of the reachable thresholds, balacing of busbar inductances, verification of tolerance to high ramp rates, etc. Parameters tuning should be possible from the CCC.
The presentation triggered a few questions:
Rudiger recommended to prioritize the tests as we have only two days available
Mirko asked whether the circuits RB/RQF and RQD must be qualified to 1 kA
Reiner answered that some tests can be done at lower current
Mirko said that the plan is to have the machine closed by midday and use the afternoon to complete the commissioning to 1 kA
Nuria inquired about the access conditions
Rudiger clarified that in case an access is needed during the weekend, the procedure is to lock the circuits, to pass the machine from CLOSED to RESTRICTED mode, then access can be granted and (hopefully) there will be no need to re patroll the machine
Reiner asked to MP3 wheter it is considered necessary to repeat the IST of the quench heater circuits
There was general consensus on not doing it. The ELQA has been extended to the heaters circuits, and anyway the current is low enough for the magnets to be self protected, while the discharges should be kept to a minimum not to stress the fragile wires
Rudiger asked if tests of the coupling between RQF and RQD were foreseen.
Reiner answered that initially they will be tested one by one
Nuria observed that the tests of RQF and RQD at injection could be done in phase one; she asked whether they could not be anticipated
Reiner said that the equipment should be installed at the beginning of next week
Everybody agreed that such an anticipation should be tried if possible.
Status of RQT13 investigation (S. Le Naour)
Sandrine showed (see slides) the findings about the RQT13.L5B1. In syntesis it was made clear that the "quenches on flat top" observed were in fact one true quench and one PC trip (triggered by the QPS). The trip happened while ramping down the main quadrupoles circuit, although apparently not at the maximum didt of the latter. The quench happened following a training quench that had occurred while just approaching 550A.
The PCS tests had not revealed any anomaly
K. H. Mess commented that the course to take was different in case the quench was in the magnet or in the bus bar. In the first case training may help, while in the second case it would be a bad idea.
He added that these are welded joints which are mechanically much more sound.
Rudiger mentioned that R. Wolf had proposed of doing some calorimetry to identify the location of problems, such as heating of bus-bars leading to a quench
Mike cautioned the cryogenics of line N was not enough undestood
Andrea Musso mentioned that line N interconnect could be opened to look at the weld quality.
Gianluigi stated that a minimum of 390 A is needed for beam operation, without any margin.
It was agreed to reduce I_PNO to 400A, in case more will be needed we will come back on this circuit. At 400A an attempt to measure the resistance of the cold part
Calorimetry tests for Hardware Commissioning (J. Strait)
Jim presented the available tools that could be used to analyse the calorimetric tests during the next hardware commissioning (see slides).
The calorimetry tests are meant as a cross check, given the fact that the new QPS, which is potentially the most adequate tool to measure resistances, will not have been fully qualified at the time of the powering.
Ruediger asked whether the new QPS will cover all the splices of the machine with the same resolution
Mike answered that this true within a factor 2
Tolerances on dT/dt (displayed in the Cryogenics PVSS application) are used to assess the measurement conditions. They are missing for the inner triplets (where the He mass is smaller but presumably the J-T valves are dimensioned for theheat load from high luminosity collisions)
The JAVA tool prepared by BE-OP gives estimates of the resistances using a quadratic fit of the power vs current data.
The helium properties are approximated linearly to translate temperature increases into dissipated power and the cooling power is assumed to be constant, i.e. it is assumed that the J-T valves do not move.
Tolerances for some of the collected parameters are missing.
The algorithm used by L.Tavian is published in his latest PAC paper.
It uses non linear approximations for the helium properties and corrects for the J-T valves changes.
It does not systematically fit parabolas to the power vs current data to get the resistance.
It was asked (Nuria, Walter ) what is the agreement bewteen the two tools in the case of the 50 Ohm resistance found in sector 67.
Jim answered the agreement was within few nOhm.
He added that the best would be to combine the merits of both tools , using quadratic fits on L. Tavian data.
L. Tavian could provide tolerances on the valve opening changes, tables with his analysis of the old calorimetric tests to allow a more extensive comparison of the two tools, and eventually his algorithm could be implemented in the JAVA application, which could become then the baseline, as it is easy to use and readily available.
Rudiger said that we must make sure that Java application will be further supported.
AOB
topics for the next meeting:
Interleaving HC and beam commissioning (G. Arduini)
DFB commissioning procedures (A. Perin)
Visibility of NC in the layout DB
Matters arising
Nuria informed that she's now invited to the LMC meetings, and in the TEMB, where she will report on MP3 worksPowering tests have started in Sector 23, in future there will be a report on previous week MP3 activity in the CCC at each meeting
Report from the CCC (S. Le Naour)
Sandrine reported on the issues encountered during the first days of comissioning, in particular on the RCBYHS4.R2B1 circuit which quenches when the crowbar is activated at 72 A. This happens only when starting with a positive polarity. When reducing the current to 70A the problem disappeared, but it was still there when switching back to 72 A. Giorgio showed the AC transfer function |Z(f)| of the correctors in the same cold mass (see slide); which looks sligthly smaller in the intermediate frequency range for the corrector concerned.
Yves commented that a similar behavior was observed on a bunch (6-7) of correctors of the same type in 2008, and in some cases the magnets would quench on the ramp down instead of upon crowbar activation.
In other words this is not the worst case: indeed the behavior of RCBYHS4.R2B1 could be accepted for beam operation.
Differently from 600A circuits quenching during discharge, in this case the mechanism is not understood.
Yves has already done some analysis work on the subject.
ACTION –> Walter, Yves and Giorgio will look further into the subject and report back to MP3
nQPS test procedure (R. Denz)
Reiner presented the goals and plan for the foreseen test of the new QPS on the weekend 27-28 June (slides) The objectives of the test include noise measurements, assessement of the reachable thresholds, balacing of busbar inductances, verification of tolerance to high ramp rates, etc. Parameters tuning should be possible from the CCC.
The presentation triggered a few questions:
Rudiger recommended to prioritize the tests as we have only two days available
Mirko asked whether the circuits RB/RQF and RQD must be qualified to 1 kA
Reiner answered that some tests can be done at lower current
Mirko said that the plan is to have the machine closed by midday and use the afternoon to complete the commissioning to 1 kA
Nuria inquired about the access conditions
Rudiger clarified that in case an access is needed during the weekend, the procedure is to lock the circuits, to pass the machine from CLOSED to RESTRICTED mode, then access can be granted and (hopefully) there will be no need to re patroll the machine
Reiner asked to MP3 wheter it is considered necessary to repeat the IST of the quench heater circuits
There was general consensus on not doing it. The ELQA has been extended to the heaters circuits, and anyway the current is low enough for the magnets to be self protected, while the discharges should be kept to a minimum not to stress the fragile wires
Rudiger asked if tests of the coupling between RQF and RQD were foreseen.
Reiner answered that initially they will be tested one by one
Nuria observed that the tests of RQF and RQD at injection could be done in phase one; she asked whether they could not be anticipated
Reiner said that the equipment should be installed at the beginning of next week
Everybody agreed that such an anticipation should be tried if possible.
Status of RQT13 investigation (S. Le Naour)
Sandrine showed (see slides) the findings about the RQT13.L5B1. In syntesis it was made clear that the "quenches on flat top" observed were in fact one true quench and one PC trip (triggered by the QPS). The trip happened while ramping down the main quadrupoles circuit, although apparently not at the maximum didt of the latter. The quench happened following a training quench that had occurred while just approaching 550A. The PCS tests had not revealed any anomaly
K. H. Mess commented that the course to take was different in case the quench was in the magnet or in the bus bar. In the first case training may help, while in the second case it would be a bad idea.
He added that these are welded joints which are mechanically much more sound.
Rudiger mentioned that R. Wolf had proposed of doing some calorimetry to identify the location of problems, such as heating of bus-bars leading to a quench
Mike cautioned the cryogenics of line N was not enough undestood
Andrea Musso mentioned that line N interconnect could be opened to look at the weld quality.
Gianluigi stated that a minimum of 390 A is needed for beam operation, without any margin.
It was agreed to reduce I_PNO to 400A, in case more will be needed we will come back on this circuit. At 400A an attempt to measure the resistance of the cold part
Calorimetry tests for Hardware Commissioning (J. Strait)
Jim presented the available tools that could be used to analyse the calorimetric tests during the next hardware commissioning (see slides). The calorimetry tests are meant as a cross check, given the fact that the new QPS, which is potentially the most adequate tool to measure resistances, will not have been fully qualified at the time of the powering.
Ruediger asked whether the new QPS will cover all the splices of the machine with the same resolution
Mike answered that this true within a factor 2
Tolerances on dT/dt (displayed in the Cryogenics PVSS application) are used to assess the measurement conditions. They are missing for the inner triplets (where the He mass is smaller but presumably the J-T valves are dimensioned for theheat load from high luminosity collisions)
The JAVA tool prepared by BE-OP gives estimates of the resistances using a quadratic fit of the power vs current data.
The helium properties are approximated linearly to translate temperature increases into dissipated power and the cooling power is assumed to be constant, i.e. it is assumed that the J-T valves do not move.
Tolerances for some of the collected parameters are missing.
The algorithm used by L.Tavian is published in his latest PAC paper.
It uses non linear approximations for the helium properties and corrects for the J-T valves changes.
It does not systematically fit parabolas to the power vs current data to get the resistance.
It was asked (Nuria, Walter ) what is the agreement bewteen the two tools in the case of the 50 Ohm resistance found in sector 67.
Jim answered the agreement was within few nOhm.
He added that the best would be to combine the merits of both tools , using quadratic fits on L. Tavian data.
L. Tavian could provide tolerances on the valve opening changes, tables with his analysis of the old calorimetric tests to allow a more extensive comparison of the two tools, and eventually his algorithm could be implemented in the JAVA application, which could become then the baseline, as it is easy to use and readily available.
Rudiger said that we must make sure that Java application will be further supported.
AOB
topics for the next meeting: Interleaving HC and beam commissioning (G. Arduini)
DFB commissioning procedures (A. Perin)
Visibility of NC in the layout DB
There are minutes attached to this event.
Show them.
The agenda of this meeting is empty