W. Venturini - PowerPoint PPT Presentation

��������������������������� W. Venturini Delsolaro Acknowledgment L. Bottura, A. Butterworth, S. Fartoukh, K. Fuchsberger, M. Giovannozzi, M. Lamont, M. Pereira, S. Redaelli, R. Steinhagen , E. Todesco, J. Wenninger Workshop of the LHC Beam Commissioning Working Group, Evian, 19-20 January 2010

�������� Historic of the energy ramps during the 2009 run Losses and transmission Evolution of orbit, tune and chromaticity Feed forward and feed back Settings incorporation for various systems Software tools Conclusions and outlook

������������������� # Timestamp Pre B1 in, B2 in, Q FB, Q FB, B1 out B2 out cycle protons protons B1 B2 protons protons ≈ 5E7 bad 2.6E9 - no - - (noise 1 2009-11-24 General conditions: level) 00:23:08. ≈ 5E7 No separation, ok 2.5E9 - no - - (noise 2 2009-11-29 No orbit FB, level) 21:47:51.844 No Q’ continuous measurement, ≈ ok 2.2E9 1.25E9 no no 1.2E9 1.9E8 3 2009-11-30 mains 00:33:16.356 “constant” incorporation of injection trims, ok 2 pilots 2 pilots no got to top got to top, yes Ramp with masked BPM interlocks in P6, (no BI (no BI and lost (??) no BI data 4 2009-12-08 logging) logging) 21:32:06.994 Collimators at injection settings, ≈ 8-9E9 ≈ E10 ≈ 5E7 ok yes yes 8.2E9 RF: synch. and phase loops on, constant voltage, (noise 5 2009-12-13 level) 22:41:33.821 no attempt to blow up emittance ≈ 9.6E9 ≈ 1.1E10 ok yes yes 9.6E10 1.1E10 6 2009-12-14 02:31:30.575 ok 1.52E10 1.62E10 yes yes 1.52E10 1.58E10 7 2009-12-15 21:12:33.680 ok 1.15E10 1.9E10 yes yes 1.15E10 1.89E10 8 2009-12-16 00:49:06.019

�������������� Feed forward tool by Mario/Mike Generation (to incorporate injection trims) In general, incorporation was done manually by the specialist. In future could it be sequencer task? Sequencer/Equip State to load functions and drive the systems Tune meter Fixed displays for energy, FBCT, bunch profiles, BLM…

������������������� From old Fidel scaling � undercorrected, ≈ 10 units Q’ left in

������������ Used to tackle the tune evolution. A specific application exists, which computes and applies the trims taking input from logging of previous ramps (Mario) Q ref (LSA) – Q meas (MDB) = Q trim � LSA Q trim knobs for the next ramp When the FB was on, the feedback contribution was isolated: I RQTF-RQTD (MDB) - I RQTF-RQTD (FF) = I RQTF-RQTD (FB) … then translated in an incremental Q trim for the next ramp. A little more cumbersome (working with currents) In one case (4 th ramp) sign was wrong, error was taken care of by FB FF would be possible for orbit and coupling as well Not logged (enough), filtered from MDB to LDB Completely empirical, corrects the resultant: need a mechanism to interact with TF improvements

������ B1 reached 560 GeV, lost on 3 rd order resonance

���������������� �� ����

������ Proper pre cycle, B1 to 1.18 TeV, “golden orbit”, beam finally lost on Qv=0.3333

����������������� �� ���� 1.80E-03 2500 loss TCP.6L3B1 1.60E-03 loss TCP.6R3.B2 loss TCP.A6L7.B1 2000 loss TCP.B6R7.B2 1.40E-03 dipole current (A) 1.20E-03 dipole current (A) 1500 Loss (Gy/s) 1.00E-03 8.00E-04 1000 6.00E-04 4.00E-04 500 2.00E-04 0.00E+00 0 -300 -100 100 300 500 time from start of the ramp (s)

����� 2 beams up to 1.18 TeV, no FB, FF from ramp 2

���������������� �� ���� 1.60E-04 2500 loss TCP.6L3.B1 loss TCP.6R3.B2 1.40E-04 loss TCP.A6L7.B1 loss TCP.B6R7.B2 2000 dipole current 1.20E-04 dipole current (A) 1.00E-04 1500 Loss (Gy/s) 8.00E-05 1000 6.00E-05 4.00E-05 500 2.00E-05 0.00E+00 0 -300 -100 100 300 500 time from start of the ramp (s)

!����������"�����# �� ����$

% �& �����#'�����&���(�'$ Reconstructed bare tunes B2 What we saw in the CCC tune snapback is very small (about 0.005).

������)�����*

������+�����,

����������������) �& ���� loss TCP.6L3.B1 1.80E-05 2500 loss TCP.6R3.B2 loss TCP.A6L7.B1 1.60E-05 loss TCP.B6R7.B2 dipole current 2000 1.40E-05 dipole current (A) 1.20E-05 Loss (Gy/s) 1500 1.00E-05 8.00E-06 1000 6.00E-06 4.00E-06 500 2.00E-06 0.00E+00 0 -300 -100 100 300 500 time from start of the ramp

����������������* �& ���� loss TCP.6L3.B1 1.80E-06 2500 loss TCP.6R3.B2 loss TCP.A6L7.B1 1.60E-06 loss TCP.B6R7.B2 2000 dipole current 1.40E-06 1.20E-06 dipole current (A) Loss (Gy/s) 1500 1.00E-06 8.00E-07 1000 6.00E-07 4.00E-07 500 2.00E-07 0.00E+00 0 -300.00 -100.00 100.00 300.00 500.00 time from start of the ramp (s)

����������������+ �& ���� 1.80E-05 2500 loss TCP.6L3.B1 loss TCP.6R3.B2 loss TCP.A6L7.B1 1.60E-05 loss TCP.B6R7.B2 2000 dipole current 1.40E-05 dipole current (A) 1.20E-05 Loss (Gy/s) 1500 1.00E-05 8.00E-06 1000 6.00E-06 4.00E-06 500 2.00E-06 0.00E+00 0 -200 0 200 400 600 time from start of the ramp (s)

����������������#, �& ����$ loss TCP.6L3.B1 2.50E-06 2500 loss TCP.6R3.B2 loss TCP.A6L7.B1 loss TCP.B6R7.B2 dipole current 2.00E-06 2000 dipole current (A) Loss (Gy/s) 1.50E-06 1500 1.00E-06 1000 5.00E-07 500 0.00E+00 0 -300 -100 100 300 500 700 900 Time from start of the ramp (s)

ORBIT EVOLUTION �������-������� rms growth if extrapolated at higher energy will trigger a beam dump from interlocked BPM in P6 � 3.5 TeV needs orbit feedback

.������/��-�������

.&������������-������� Not measured continuously (problem of tune “noise”) Q’ measurements before and after the last ramps: Ramp #4 → beam 1 ∆ Q’ H ≈ -6.3, ∆ Q’ V ≈ -14.7 Ramp #5 → beam 1 ∆ Q’ H ≈ -2.7, ∆ Q’ V ≈ -13.2 Ramp #6 → beam 1 ∆ Q’ H ≈ -3.0, ∆ Q’ V ≈ -10.8 Ramp #6 → beam 2 ∆ Q’ H ≈ -9.2, ∆ Q’ V ≈ -8.1 Possible sources: imperfect b 3 corrections for snapback and magnetization components, hysteresis of MS, …

���������� Possible sources of Q evolution during the ramp: B 2 /B 1 tracking error (visible during decay and snapback) Feed down from CO in the main sextupoles (was checked with measured orbit and betas, effect is small, ≈ 10 -4 ). Feed down from CO in random (uncorrected) b 3 of dipoles (also checked with measured orbit and found small ≈ 10 -3 ) MCS misalignments and powering. From preliminary calculations this appears to be a good candidate to explain the tune drift. Tracking error MCS-b 3 also gives a contribution The effect is bigger for beam 2, as it is the case for orbit and coupling (…?)

.����������������������� Ramping was easier than anticipated, however several issues are still on the table: Beam parameters evolution not fully understood (in particular the differences between beams) Fidel corrections to be updated with best estimate for snapback correction Disentangle feed forward and TF updates Orbit feedback at least in the dump and collimation regions Chromaticity measurement on line Incorporate incorporation (in the sequence) RF: commissioning of emittance blow up, global orbit FB Link logging of BI to beam operation through mode Ramp with separation bumps