04/10/2013

New injection system - concept

B.Goddard, O.Aberle, B.Balhan, J.Borburgh, E.Carlier, K.Cornelis, L.Ducimetière, L.K.Jensen, T.Kramer,

D.Manglunki, A.Mereghetti, V.Mertens, D.Nisbet, B.Salvant, L.Sermeus, J.Uythoven, F.Velotti

LIU-SPS 50 ns Pb ion injection system review

Motivation

• Strong request from LIU-ION to study improving injection rise time, to 50 ns– For Pb ions only– Keeping existing injection energy for ions (rigidity of 60 Tm)– Not impacting negatively on injection for p+ or other ions– Flat top to allow injection of 4 bunches at 50 ns spacing (i.e. ~160

ns)– “Acceptable” emittance growth (can assume damper) – so 2-98%

rise-time to respect in B field

SPS injection

MKPAC MKP TBSJ

MSI

• LHC p+ injection (26 GeV/c) uses all kickers, at 52.0 kV (225 ns)• CNGS p+ injection (14 GeV/c) uses MKPA/C, at 49.2 kV (< 1000 ns)• Pb82+ ion injection (17 GeV/c per charge)

– used MKPA/C, at 52 kV for Q26 (200 ns)– uses all kickers at ~40 kV for Q20 (225 ns)

• TBSJ is injection dump (use MDSH119 to improve centering)

Present ion injection

• Present “ion” kickers type MKP-S– Were designed to reach 115 ns (with PFLs, never installed…)

• Measured at <150 ns (2-98%) with present PFNs– Need also to use “LHC p+” MKP-L kickers for Q20

• Imposes 225 ns rise-time• Look at split tunes, or back to Q26 for ions, to reach 150 ns

with present layout

• Addition of PFLs to the present MKP-S magnets is possible (already foreseen in BA layout)– Will give 125 ns for sure, maybe 100 ns– No additional tunnel equipment– Kick strength then only sufficient for ≤13 GeV/c depending on

optics

MKP usage for ions(from discussion with Django)

• 2011– Q26 optics, use first three tanks only (MKP-S)– Pushed voltage to 52 kV and tweaked trajectories and orbits to

make a 17 GeV/c be injected in a system which was designed for 13 GeV/c

– Rise time used was 200 ns• The MKP-S measured rise time 2 % - 98 % is 148 ns. See later slide from

2000.

• 2012– Q20 optics – less help from quad in front of MKP– Used the fourth tank (MKP-L) as well, which was not modified 13

years ago– Rise time used, as for LHC protons: 225 ns, due to the slower

MKP-L

Concepts studied

• Rebuild entire injection kicker system• Add thin pulsed septum and rebuild entire

injection kicker system• Add separate new ion kicker and new thin

pulsed septum, for ion injection only• Add PFLs to existing kickers, with new thin

pulsed septum• Do nothing to existing kickers, either with or

without new thin pulsed septum

Option (i)

• Rebuild existing magnets completely– Higher impedance (16.6 -> 25 Ohm)– Much shorter magnets (0.6-0.7 -> 0.28 m)– Many more individual magnets (12 -> 48)– Much longer magnetic length needed (8 -> 12 m)– No aperture for dumping p+ beam– Extremely expensive even if possible

• Discarded: does not look feasible– Many integration issues– Many technical issues– No aperture for dumped beam– Huge cost and investment even if solutions could be found

Option (ii)

• Add thin in-vacuum septum and rebuild existing MKP, with rearrangement of layout– Only need to rebuild a subset of the MKP– Same issues with very short 28 cm magnets– More issues with magnet aperture (vertical gap not enough)– No solution found to dump p+ beam cleanly

• Discarded: does not look feasible– Some integration issues– Some technical issues– Aperture limits in vertical plane– No way to dump p+ beam– Large cost and investment

Option iii: the minimalist

• Don’t change anything and we can probably go down to 150 ns rise time (2 – 98 %)– If we want to have more margin / less tweaked orbits and

trajectories one can install the MSI-V, but might not be strictly necessary

• In this case the installation of an ion injection dump might be nice to keep it ‘clean’

• Without MSI-V Q20 clearly makes things more difficult – try split tunes?

– More work on the MKP-S front cells in the PFN might make it possible to go down to 125 ns rise time

– Would be available for tests after LS1

Option iv: the intermediate

• To go below 125 ns kicker rise time, aiming for 100 ns kicker rise time (2 – 98 %): install a PFL in parallel to the present PFNs on the MKP-S system. – No kicker magnets to be installed in the tunnel– This reduces the maximum voltage on the PFL to 40 kV and the

installation of the MSI-V becomes obligatory– With MSI-V no problems with Q20 optics– The switching between PFN and PFL requires development and

lab tests– Injection dump for ions would be nice, to control losses

Option v: the full monty

• Install the new MKP-I magnets & MSI-V– If all works well this would allow for a rise time of 75 ns (2 – 98

%). But not easy and a lot of work– Adds impedance to the machine – beam dynamics, and risk of

kicker heating– Not sure that impedance shielding is possible with this rise time– New kickers and cables needed in highly radioactive zone

downstream of TIDVG– Ion injection dump again needed, to stay clean

Option v

Layout in SPS

MKP-I

MSI-V

Option v

17 GeV/c/u Pb82+ ionswithout injection bump

Option v

14 GeV p+ (as done today)

Option v

• Could conceivably have a dump block in a new location just upstream QD.121

• Could misalign MBB.12070/90 by -10 mm in H

~1.4 m available

Option v

Dumping of injected ion beam(continue to use TBSJ for p+ as at present)

Dump blockQD.12110 chamberMBB.12070/90 chamber,with 10 mm offset

30 mm

Considerations

• Performance reach is one important aspect• Others are

– Required effort level (resources, cost) vs. benefit– Radiation doses for installation, operation and

maintenance– Technical risk– Compatibility with injection of p+ beams (aperture)– Compatibility with HL-LHC beams (impedance)

TIDVG

LSS1 is a very active area…

MSI-V MKP-I TBS-I

Summary of performance/costs

Option iii) - optimised present Option iv) - intermediate Option v) - full Monty

Description Trims of existing PFNs/timings, +MSI-V? New MSI-V, PFLs on MKP, new dump New MSI-V, new MKI-P, new dump

batch spacing (ns) 150 (125?) 100 75

New MSI-V no/yes yes yes

New MKP-I no no yes

Cost (kCHF) 100/1300 5500 9000

Manpower (MY) 1.0/3.3 13.1 19.4

CommentAvailable before LS2 w/o MSI-V125 ns needs studies

Attainable rise time to demonstrate. Can already work on existing MKP-S Feasibility of shielding?

Radiation impact None/medium medium high

Impedance impact None/low low potentially high

Compatible with Q20 No/yes with MSI-V (Q26 or split tunes)? yes yes

fin