sps scrubbing runs 2014 week 45 and week 50 (part i) h. bartosik, g. iadarola, k. li, l. mether, g....

SPS Scrubbing Runs 2014week 45 and week 50 (part I)

H. Bartosik, G. Iadarola, K. Li, L. Mether, G. Rumolo, M. Schenk

Thanks to: SPS OP crew, G. Arduini, T. Argyropoulos, : T. Bohl, E. Carlier, K. Cornelis, H. Damerau, J. Esteban Muller, A. Findlay, S. Gilardoni, B. Goddard, A. Guerrero,

S. Hancock, W. Höfle, V. Kain, G. Kotzian, A. Lasheen, M. Meddahi, E. Métral, B. Mikulec, G. Papotti, E. Piselli, A. Romano, R. Salemme, B. Salvant, E. Shaposhnikova, M.Taborelli

… and thanks to the PS complex for the preparation of the challenging beams!!

LIU SPS intermediate scrubbing review, 23 February, 2015

Outline

• Introduction

• Overview Scrubbing Run Week 45

• MKP, TIDVG temperatures and pressures


• Vacuum pressure evolution (arcs vs a-C coated cells)

• Scrubbing observations in the e-cloud monitors

• Main observations with different beam configuration

• 25 ns beam – standard scheme

• 25 ns beam – BCMS

• 8b+4e – low e-cloud pattern

• High intensity 25 ns beam

• Doublet beam

• Some comparison against simulations

• Doublets: MBA vs MBB

• Doublets: Dipoles vs field free

Scru

bbin

g ru

n

Scru

bbin

gru

n

Injector schedule 2014

• Originally 2 weeks foreseen for SPS scrubbing, then reduced to two blocks of 7 and 2.5 days in Week 45 and 50, respectively

• Some goals (or desired outcome) of SPS scrubbing in 2014 Recover 2012 performance with LHC 25 ns beams and qualify machine behavior with LHC beams

after long shutdown Test doublet beams for SPS scrubbing and in preparation for LHC scrubbing in 2015 Study high intensity 25 ns beams at 26 GeV/c to gain information to steer LIU decision coating vs.

scrubbing in 2015 Test new LHC-type beam variants

Outline

• Introduction











• Doublet beam




Monday

Tuesday Wednesday Thursday Friday Saturday Sunday

Scrubbing week 45: overview

daysMD3: 26.4 s, long flat bottom cycle 26 GeV/c

MD1: 7.2 s, short flat bottom cycle 26 GeV/c

LHC25NS: 22.8 s, nominal LHC filling cycle to 450 GeV/c

Monday



daysMD3: 26.4 s, 26 GeV/c

Monday

• Prepared machine for scrubbing (Fast vac. valves blocked open, extraction elements off, ZSs at -30 kV)

• In agreement with TE-ABT increased MKP4 SW pressure interlock to 4e-7 mbar

• Setup of scrubbing cycles (early dump to control pressure in MKP and TIDVG)

• COLDEX moved into IN position (observed no impact on beam lifetime)

• Started maximizing integrated beam intensity on scrubbing cycle

− limited by TIDVG vacuum interlock in agreement with TE-ABT and EN-STI interlock level raised from 4e-7 mbar to 6e-7 mbar (while the pressure in MKP4, initially lower, was slowly creeping up….)

Monday



daysMD3: 26.4 s, 26 GeV/c

Tuesday

• Continued scrubbing at 26 GeV (MD3 cycle, inject more, move the dump later in the cycle)

• In parallel, setup of 200 MHz LL RF on acceleration cycle

− 6h in the afternoon, mostly without beam, reduced the duty cycle

• Reached 4 batches on FB with dump at 13s

• Limited by MKP4 vacuum (mainly related to increase of static pressure due to heating)

− Had to go back to 3 batches during the night for cool down

Wednesday

• Dedicated MD on new 800 MHz LLRF (from 10:00 to 00:45, fully dedicated in first part, then in parallel with setting up of doublet)

• Agreed with TE-ABT and EN/STI about the MKP4 and TIDVG thresholds

− Raise MKP4 SW interlock to 5e-7 mbar

− Raise TIDVG interlock to 1.2e-6 mbar

• Started setting up the doublet beam on MD1

− 1 batch of 72 bunches from PS split into 72 doublets, low intensity (~1.2e11 p/doublet)

− Observed strong pressure rise in the arcs (but none in a-C coated cells and in sensitive elements)

• Mains tripped during the night (4h no beam)

Monday



days

MD3: 26.4 s, 26 GeV/c MD1: 7.2 s, 26 GeV /c

LHC25NS: 22.8s, 450 GeV/c

Monday



days

Thursday

• Continued scrubbing with standard 25 ns beam (injected up to 5 batches of 72 bunches)

• In parallel, LHCPILOT cycle for LHC mastership and RF synchronization tests, and TI2 polarity checks

• Continued transverse damper setup for doublet beams

• Tests with multiple batches on cycle with acceleration (longitudinal instabilities observed, since longitudinal damper was still not functional on the ramp)

• In the night, scrubbing exclusively with doublet beam to allow for MKP4 cool down

MD3: 26.4 s, 26 GeV/c MD1: 7.2 s, 26 GeV/c


Monday



days

Friday

• No beam in the morning due to PS problems (TT2 quad and then RF)

• Continued setup of the 200 MHz longitudinal dampers

• In parallel scrubbing with doublet beam on MD1 cycle

− Further transverse damper setup

− Observed transverse instabilities on the trailing bunches of the train (no vertical instability observed on standard 25 ns beam, even with low chromaticity)

• In the night acceleration of 4 batches of 72 bunches up to 450 GeV/c

• Scrubbing with 4 batches accelerated (interleaved with doublet scrubbing when pressure in TIDVG became too high)

MD1: 7.2 s, 26 GeV/c LHC25NS: 22.8s, 450 GeV/c

Monday



days

Saturday

• Scrubbing with standard 25 ns at 26 GeV/c and with acceleration

• Optimized closed orbit on the energy ramp to reduce losses

• Tested multiple injections (up to 4) of doublet beam on the long 26 GeV/c cycle

− Beam quality initially quite poor, could be later improved by fine tuning

− RF voltage ramped to 7 MV after the last injection to increase scrubbing

− Left in for scrubbing over night

MD3: 26.4 s, 26 GeV/c MD1: 7.2 s, 26 GeV/c


Monday



days

Sunday

• Characterization of the 25 ns beam (4x72b.) at 26 GeV (to be compared against 2012 meas):

− Emittance measurements, bunch-by-bunch tune measurements

− Lifetime as a function of vertical chromaticity

• Characterization of the doublet beam (4x72doublets) at 26 GeV:

− Emittance measurements, bunch-by-bunch tune measurements

− Lifetime as a function of vertical chromaticity

• Few tests with orbit bumps to localize the scrubbed region

• Monday 5.00: beam stopped for cool down and access to move COLDEX out

MD3: 26.4 s, 26 GeV/c

MKP pressure and temperature evolution

MKP pressure and temperature evolution

• Increase of MKP4 SW pressure interlock to 4e-7 mbar, then 5e-7 mbar necessary to advance with scrubbing

• Relatively low dynamic pressure rise, but static vacuum following the temperature profile

• MKP heating due to impedance with nominal 25 ns beams, but cool down with low intensity doublets (consistent with the impedance model)

TIDVG pressure and temperature evolution

TIDVG pressure and temperature evolution

• Increase interlock level from 4e-7 mbar to 6e-7 mbar was necessary to advance with scrubbing. Increase to 1.2e-6 mbar needed when dumping beam on TIDVG

• Conditioning visible over scrubbing run

• Temperature rise by more than 4 degrees when dumping consecutive shots on TIDVG

Important facts of Scrubbing Run I (1/2)

• 1 full week (week 45) high efficiency scrubbing with very good beam availability (only minor hiccups) and main limitations coming from vacuum in MKP4 and TIDVG (not the arcs)

• Large use of long flat bottom (26 GeV/c) cycle, gradually increasing number of batches injected and moving the dump later in the cycle (as vacuum in TIDVG or MKP4 permitted)

• Successfully deployed low intensity doublet beam (up to four batches at 26 GeV/c with no more than 1.4 x 1011 p/doublet)

Enhanced electron cloud in high-field non a-C coated regions and none in a-C coated regions and field free regions Experimental evidence of electron cloud suppression with a-C coating

• No MKE heating, but significant heating on MKP4 leading to outgassing and pressure above interlock value

Lower MKP heating (as expected from C. Zannini’s kicker impedance model) with doublet beam, which allowed scrubbing while recovering MKP4 vacuum


• Recovered nominal LHC beam (4 batches 72b) at 450 GeV/c Commissioned the new TIDVG with standard 25 ns beam at 450 GeV/c, data

taken for thermal response Measured transverse emittances below 3 mm Minor fine tuning yet to be done

εv ≈ 2.4 μm


• Recovered nominal LHC beam (4 batches 72b) at 450 GeV/c Commissioned the new TIDVG with standard 25 ns beam at 450 GeV/c, data

taken for thermal response Measured transverse emittances below 3 mm Minor fine tuning yet to be done

• Vertical electron cloud instabilities hard to reproduce with Q20 optics! Observed only with doublets due to high electron cloud operation Standard 25 ns beam was stable even with low vertical chromaticity setting Further studies including comparison with Q26 optics planned

• Important conditioning of both newly installed elements (e.g., MKP4, TIDVG) and pressure in arcs

→ Scrubbing overall successful so far, re-conditioning is always easier than first conditioning…

Outline

• Introduction











• Doublet beam




Scrubbing run in week 50: overview

MD3: 26.4 s, long flat bottom cycle 26 GeV/c

LHC25NS: 22.8 s, nominal LHC filling cycle to 450 GeV/c

LHCMD4: 31.2 s, long cycle with 3x slower ramp to 450 GeV/c

days


MD3: 26.4 s, 26 GeV/c:High intensity 25 ns beam

Monday (08/12/2014)

• High intensity 25 ns beam (up to 1.9e11 p/b injected) on flat bottom cycle

− Lifetime quite poor

− Horizontal instability after injection of 2nd batch

− Emittance blow-up in last part of trains (starting from 2nd batch), especially horizontal

• Intensity scan during the night (but problems in PSB RF)

→ Back to strong and visible electron cloud effects:

− Stronger electron cloud in drift regions due to higher intensity, in dipoles due to different stripe position

− Beam more sensitive to collective effects due to high intensity

days


Tuesday (09/12/2014)• Slow acceleration cycle (initially, cycle setup with 12 bunches)

− Doublet beam (up to 1.5e11 p/b injected)

− Instabilities and losses on the ramp

− High losses on flat bottom and beginning of ramp, increasing significantly with the amount of beam injected

• Doublets also on cycle with nominal LHC filling ramp in parallel

− Maximum intensity on flat top of about 1.4e11 p/doublet, large emittance growth at the tails of the doublet trains

• TT10 quad trip + RF problems in PS during the night

→ Many issues due to enhanced electron cloud. Improvement needed with scrubbing, but also with systematic work on set up (working point, transverse damper, LLRF, longitudinal settings)

− 2015 doublet cycle from MD to ‘operational’ to prepare beam for LHC

MD3: 26.4 s, 26 GeV/c LHC25NS: 22.8 s, 450 GeV/c: Doublet LHCMD4: 31.2 s, slow ramp to 450 GeV/c: Doublet

days


Wednesday (10/12/2014)• Two parallel cycles

− Long flat bottom cycle for scrubbing with high intensity standard beam

− Nominal LHC filling cycle (with acceleration to 450 GeV/c) to test LHC beam variants

• 8b+4e beam on LHC cycle (1.7e11 p/b injected, 1.4e11 p/b at flat top)

− Well behaved electron cloud-wise, i.e. no signal on e-cloud monitors (as predicted)

− Pressure spike on ZS (impedance?), however quickly conditioning while running with this beam

− Large losses at beginning of ramp

• BCMS in rather good shape (1.4e11 p/b injected, 1.15e11 p/b at flat top), but still large losses at beginning of ramp

LHC25NS: 22.8 s, 450 GeV/c:8b+4eBCMS MD3: 26.4 s, 26 GeV/c:High intensity 25 ns beam

days

Important facts of Scrubbing Run II

• High intensity LHC beam (4 batches 72b) at 26 GeV/c Strong electron cloud effects

o Poor lifetime, strong emittance blow up at the tails of the batcheso Clear coherent instabilities contributing to the blow up (in spite of the transverse

damper and/or high chromaticity)

• Doublet beam accelerated to 450 GeV/c Many transverse instabilities observed on the slow ramp Maximum intensity at flat top limited to 1.4e11 p/doublet on nominal LHC filling ramp,

lower on slow cycle (but with 3.6 s injection plateau). Strong emittance blow up. In general, beam quality strongly and quickly degraded by electron cloud, but room for

improvement if taken on ‘operational’ basis (scrubbing, better set up)

• Other LHC beam variants tested 8b+4e well behaved in terms of electron cloud (as expected), but sparking and

outgassing on ZS (however quick conditioning was observed), up to three batches accelerated to 450 GeV/c

BCMS in rather good shape, less electron cloud thanks to shorter batches, up to three batches accelerated to 450 GeV/c

For both cases, no clear beam degradation due to electron cloud, but quite large losses at start of the ramp

Comparison of integrated BCT

• 2012 scrubbing run limited by MKE kicker heating

• 2014 scrubbing run W45 and W50 limited by MKP4 pressure (hopefully only temporary problem since kicker was newly installed)

Outline

• Introduction











• Doublet beam




Pressure evolution (normalized): Arc 5

DoubletStd. 25 ns Doublet Std. 25 ns 8b+4e / BCMS

Scrubbing W45(~1.2 x 1011 ppb)

Scrubbing W50(~1.5-2.0 x 1011 ppb)

Std 25 ns Doublet 25 ns (450 GeV) Doublet Final tests

clear conditioning

clear conditioning

clear conditioning and very high efficiency with doublet

Pressure evolution: Arc 5





conditioning conditioning with doublet similar pressure at beginning

conditioning high intensity

Pressure evolution: a-C coating





No pressure rise with doublet beam

No pressure rise with doublet beam

residual pressure rise from neighboring

elements

Outline

• Introduction











• Doublet beam




ECM 1a (MBA StSt) at 1 s





Nov. 2012(Nominal beam)

Stronger E-cloud with doublet

less e-cloud with high intensity

no e-cloud with other beams

ECM 1a (MBA StSt): maximum signal





e-cloud not completely suppressed when many batches in the machine

(for standard and BCMS)

ECM 1b (MBB StSt): at 1 s






Some conditioning, but e-cloud never completely

suppressed even at 1 s

ECM 1b (MBB StSt): maximum signal





Conditioning is less obvious

No e-cloud with 8b+4e

ECM 2a (MBB Cu): at 1 s






Strong deconditioning since 2012 but also quick re-conditioning

ECM 2a (MBB Cu): maximum signal





(Re)-conditioning

Conditioning with high intensity

e-cloud with short bunches

No e-cloud with 8b+4e

Thank you for your attention!

sps scrubbing runs 2014 week 45 and week 50 (part i) h. bartosik, g. iadarola, k. li, l. mether, g....

Documents

ns beams

beam lifetimestarted

mba vs mbbdoublets

lhc beams

tidvg temperatures

challenging beams

sps op crew

desired outcome of sps