Participants:
R. Bailey (chairman), G. Arduini, R. Assmann (secretary),
O. Brüning, H. Burkhardt,
R. Jones, V. Kain, J.P. Koutchouk,
M. Lamont, L. Ponce, R. Steinhagen, W. Venturini Delsolaro, T. Wijnands
Excused: M. Giovannozzi,
B. Goddard, S. Redaelli, R. Schmidt, J. Uythoven, F. Zimmermann, M. Ferro-Luzzi
None.
LHC planning is being finalized with several meetings involving the DG. The new schedule should be internally officialised in the TCC at June 16th.
LHC aperture restrictions have been found in the installed dipole magnets. Some were found by the reflectometer, others not. This is a potentially serious problem and is treated at future MARIC and LTC meetings.
Mike presented the plans for operational handling of the snapback and the ramp with a single beam. His slides are available for detailed information. Mike explains the nominal cycle and then discusses the important topics:
DC Settings: Needed
are transfer functions, DC errors, MAD magnet strengths. FIDEL determines
harmonics to get DC calibration curve: B vs I. From
this we get settings for harmonic correctors. Power converter test in IR8 was
successful: an irregular squeeze was tested due to out of date data č
redo with recent squeeze settings. The ramp settings are determined from Fidel
and
Action: Provide recent
squeeze data to operation and prepare for including them into future power
test.
Decay and snapback: Decay effect: It was discussed whether it should be included in Fidel or in LSA? Fidel should have trimming capabilities. We must be careful with implementing trims into the model. Feedbacks should be switched off ideally, including all trims into the model. Fidel could use data for all single magnets but at the moment uses a statistical average per sector. General goal in AT: provide 20 parameters per magnet. Mike is using data per sector. The discussion concluded that it is initially OK to use data per sector but later we should use it per magnet. There can be a choice for each ramp whether to re-calculate settings (~5min) or use functions of previous fill. Injection plateau: correct dynamic effects will be implemented in some automatic mode. Snapback model (L. Bottura) will predict changes in chromaticity. Incorporate injection trims ~10s before start of ramp.
Action: Decide
where to handle decay effects: Fidel or LSA?
Action: Prepare
to later use all data available from AT (per magnet).
Action: Specify
in detail incorporation of trims before ramping. Ensure and test compatibility
with feedbacks and make sure that machine safety cannot be compromised.
Power converters: Continuity of functions within 0.01A is automatically ensured when functions are loaded. Functions are checked against limits. There can be no trims after loading the ramp. Changes can still be put through the real time channel. However, these real time trims are not ramped (č 0.5 A stays 0.5 A).
Comment by Ralph S after the meeting: The 'not ramping' is a design choice. The inputs could technically be ramped but at the moment its not clear whether “all” the effects corrected during the injection scale with energy during the ramp. E.g the correction of misaligned quadrupoles does whereas the decay correction do not scale. Hence, it was decided - as safest assumption (design choice) - to let these correction "die out" while ramping.
Action: Make sure that
there will be no problem with the 0.01 A check, for example for correctors with
low currents.
Functions and timing: No stop of ramp during execution of functions. Set up ramp for beam dump: how and how many stop points? Several participants felt that well-defined stops during the ramp would be very beneficial for commissioning.
Action: Specify the
number of stops required for the set-up of the beam dump, before we can ramp to
7 TeV.
Action: Present
detailed procedure for steps in commissioning the energy ramp.
RF: Ramp with the synchro loop. Accurate frequency functions are required. Radial loop can work with one or two pickups.
Action: Clarify number
of pickups available for the radial loop during commissioning.
Comment by Ralph S. after the meeting: Using only one (or two) pickups may lead to a compensation of the dispersion orbit (desired) “and” betatron oscillation and orbit respectively (not desired) using the RF frequency - this may be dangerous because one may automatically steer the beam against the aperture this way if the orbit is not stabilized using the orbit feedback.
Beam tolerances: 450 GeV tolerances should also apply for the ramp as the available beam aperture stays constant. The discussion pointed out the need to allocate budgets for static and dynamic tolerances. Beta beat values are under preparation in AT. The collimator settings for the ramp are required for Mike. No easy beam loss free-steering will be possible in the LHC: For example minimize BLM readings by putting local orbit bumps. This can be disastrous, as the LHC beam halo can be fully asymmetric.
Action: Allocate tolerance
budgets for static and dynamic errors.
Action: Provide
collimator settings during the ramp.
Measurements: Feedbacks should be commissioned early on.
Ralph S commented after the meeting: Presently, it is not planned to compensate/re-calibrate the BPMs with respect to changing bunch intensities or varying bunch length dependence. The expected effects are small and well below 2% with respect to the BPM half-aperture.
Operation: A strategy must be decided for incorporation of trims into ramp: constant or ramped. Should there be freedom of choice or should a fixed strategy be imposed? Are there implications on machine safety?
Action: Clarify the
strategy for incorporation of trims before the ramp and evaluate consequences.
General discussion:
RB asked about possible stops during snapback: could they be required for any reason? First ramps could be done with slower speed to allow accurate snapback measurements. The discussion concluded that this might be useful, also for other issues like beam loss studies at the start of the ramp.
RB asked about the baseline target energy for the first ramp? The discussion concluded that this will likely be given by the beam dump requirements (see action above).
GA asked if real time trims are possible for operators or only for feedbacks? The discussion concluded that they will be possible for operators but are first implemented with the feedbacks. There was a concern about initial operation when the feedbacks might not yet work. It was concluded that real time trims must be available early on.
Action: Test real-time
channel for trims before commissioning.
RB asked about the time required for ramp commissioning. Mike stated that 7 days are required for both rings. The milestone is to have at least 2e9 p in both beams at top energy.
RJ asked how long it will take to have the next fill? Mike stated that it will take 90 min to 120 min to recycle the machine.
Next meetings will be on June 14th (450 GeV optics, circumference issues), June 28th, (tracking between sectors), July 12th (IR aperture and IR bumps) and July 26th (450 GeV aperture).