the lhc as a proton-nucleus collider: status and plans john jowett

The LHC as a proton-nucleus collider:

Status and plansJohn Jowett

Special thanks to:R. Alemany-Fernandez, P. Baudrenghien,

C. Carli, E. Carlo-Giraldo, S. Hancock, R. Jones,D. Manglunki, J. Wenninger,

and many others in ABP, BI, OP, RF groups.

1J.M. Jowett, Workshop on Prospects of p-Pb collisions during the 2012 LHC HI run, CERN 17/10/2011

pPb

Plan of talk Brief history The revolution frequency problem– Moving long-range beam-beam effects

Status of preparation of LHC systems Feasibility test in 2011– Goals– Possible bonus: some physics

Possible p-Pb run in 2012

J.M. Jowett, Workshop on Prospects of p-Pb collisions during the 2012 LHC HI run, CERN 17/10/2011 2

3

Before Jan 2011: Vox clamantis in deserto Not part of “LHC Baseline” … no resources. CERN Workshop in 2005 (link)

– Physics case, experiments’ performance, …– Reviewed RHIC experience with d-Au in 2002-3 (T. Satogata)– Schemes for LHC injector operation (C. Carli)– LHC operation, beam dynamics concern identified (JMJ)– Executive summary of accelerator part (in 2011 report)

LHC Project Report 928 (= paper at EPAC 2006) Key systems groups kept aware meanwhile

– Eyes open for any showstoppers Requested by ALICE for 2012

– First Pb-Pb run successful. If not 2012 then would otherwise be much later because of shutdown schedule, energy increase.

– Discussion at Chamonix workshop, Jan 2011. – ATLAS and CMS heavy-ion groups– Some resources now available:

OP, RF, BI, … collaborating on implementation and operation Fellow to work on beam dynamics in ABP arrived … today

J.M. Jowett, Workshop on Prospects of p-Pb collisions during the 2012 LHC HI run, CERN 17/10/2011

http://ph-dep-th.web.cern.ch/ph-dep-th/content2/workshops/pAatLHC/pAworkshop2.html


Relation between Beam Momenta LHC accelerates protons through the momentum

range

– Use this as reference, measure of magnetic field in main bending magnets

The two-in-one magnet design of the LHC (unlike RHIC) fixes the relation between momenta of beams in the two rings

0.45 TeV (injection from SPS) 7 TeV (collision)pp

Pbp

where 82, 208 for fully strippe

d Pb in LHCQ Z

p p

AZ


5

Critical difference between RHIC and LHC

RHIC: Independent bending field for the two beams

LHC: Identical bending field in both apertures of two-in-one dipole


6

RF Frequency for p and Pb

RFRF

p

RF RF

RF frequency , ,

where the harmonic number , 35640 in LHC

hfT p m Q

h h

RF frequencies needed to keep p or Pb on stable central orbit of constant length C are different at low energy.

No problem in terms of hardware as LHC has independent RF systems in each ring.

fRF(Pb)

fRF(p)

Which is Beam 1 and which is Beam 2 ? Initial preference for ALICE spectrometer

asymmetry:

Beam 1=p, Beam 2=Pb

Assume this for definiteness in rest of this talk.

But switching of the beams between the two rings is important and requested– Clearly equally feasible, just some setup time


Distorting the Closed Orbit

2

pp

p

p

2p

2 2

, , 1

( )where is a fractional momentum deviation and

the phase-slip factor , 1 , 55.8 for LHC optics.

Moves beam on

1 1

to off-momentum orb

(

it,

1 )

l

TT

C mcT p m Qc Qp

p QpQp

Qpmc

onger for 0. Horizontal offset given by dispersion: .xx D s


Additional degree of freedom: adjust length of closed orbits to compensate different speeds of species.– Done by adjusting RF frequency


9

Momentum offset required through ramp

22 22Pb

p Pb p2 2p

Minimise aperture needed by .4T mc m

p Z

Limit in normal operation (1 mm in arc QD)

Limit with pilot beams

Revolution frequencies must be equal for collisions at top energy.Lower limit on energy of p-Pb collisions, E=2.7 Z TeV. RF frequencies must be unequal for injection, ramp!Moving long-range beam-beam encounters may be a problem (cf RHIC).

2% - would move beam by 35 mm in QF!!

Wolfram Fischer

10

Example – beam lifetimes with (Br)d = (Br)Au 2003

beam-beam effect during injection, d and Au with same rigidity, frf = 44 kHz, vertical separation=10mmAside: RHIC is the 1st bunched beam hadron collider exhibiting strong-strong effects [W. Fischer, et al., “Observation of Coherent Beam-Beam Modes in RHIC”, BNL C-A/AP/75 (2002)]

Beam envelopes around ALICE at injection


11

Crossing angle from spectrometer and external bump separates beams vertically everywhere except at IP (also in physics). Parallel separation also separates beams horizontally at the IP during injection, ramp, squeeze.Other experiments vary ..

ALICE – Separation at injection - CMS


12

ATLAS - Separation at injection - LHCb


13

Long-range beam-beam effects


14

,

2 20 0

02 2

For separations , , the (angular) beam-beam kick on a particle of charge , due to an opposing

, where / (4 )

and gives rise to

2 ( , )

beam of total charge

pertu

(

,

i

)

sx y

x yZNr x y r e mc

x yZ

p p

e

x y

e N

ò

2 2, 0

, , , 2 2 2

rbative betatron tune-shi

( , )(

fts

)4 ( )

2x y x y

x y x y x y

x yZNrQ px y

LHC separation configurations were chosen to minimise the tune effects in physics (“footprint”).

Example: beam-beam for Pb around ALICE


15

Dashed lines are “normal” position of encounters which will move 15 cm along IR on next turn …

Overlap knock-out resonances ?


16

,

Bunch harmon

p Pb

1,2,transve

icrse mode

,

s

Encounter points move at speed 1734 m/s 0.15 m/turn2Hamiltonian is no longer periodic in .

Excites modulational resonances

x y

x x y y

m

v vV

s

m Q m Q p k

6

p Pb

3. 10 at injection,decreases in ram

891 or at most 3564

p

; , , ,

Known as "overlap knock-out resonances" at the ISR.

However with LHC tunes, 64.3, 59.3, only ext

2

remely

x y

x x

m m p kv

Q

v

Q

c

high-order resonance conditions can be satisfied. (similar in RHIC, W. Fischer).

We are nevertheless looking at calculations of driving termVery unlikely

s, compensatio

to

n b

be a problem

etween

IRs, etc.

Diffusion models Naively regarding the kicks as purely random

– Works fairly well for RHIC data (W. Fischer)

Better calculate combination of beam-beam kicks on a particle on a given turn as the encounters move– Add them up with proper betatron phases

Partial compensations– Take out static component (closed-orbit) from long-term

averaging and look at fluctuations around it– RMS fluctuation gives emittance growth rate ?– Other resonant effects?

Work ongoing

Assuming pseudo-random Spectral analysis can confirm Implement detailed collision schedule


2, 2

0 , ,1 1 ( ) ( )2

where [..] denotes mean-square deviationgives an emittance doubling time around 40 min

x ynx y x y

df s p s

dt

Build of beam-beam kicks around ring


18

Horizontal and vertical, kicks represented in normalised betatron phase space.On a turn where encounters occur at the usual IpsOn later turns the resultant is different.

Variation over ~100 turns


19


Transverse Feedback 4 independent systems, 1 per plane and per ring High bandwidth to act on individual bunches Located in IP4, so no concerns about timing of p-

Pb bunch passages. Potentially very important for p-Pb: – Damping any coherent oscillations driven by

the coherent dipole kicks from moving beam-beam encounters at injection and during ramp

Outline of p-Pb physics cycle (Pb-p similar) Nominal Pb beam (100 ns basic spacing) Matching proton beam – See injection scheme details (C. Carli)

Inject p beam in Ring 1, fRF for p– Orbit, ramp established in advance

Inject Pb beam in Ring 2, fRF for Pb– Orbit, ramp established in advance

Ramp both beams on central orbits – Orbit feedback decouples RFs

Rephase RF and bring fRF together to lock Squeeze, collide, (almost) as usual Pb-Pb

– Preliminary off-momentum set-up for 3.5 Z TeV?


Review of LHC systems RF (P. Baudrenghien, A. Butterworth)

– Independent for two rings, OK– Decouple radial control

Transverse Damper (ADT) (D. Valuch, W. Hofle)– Independent for two rings, OK– Possible variable Q reference should be OK

Beam instrumentation (R. Jones, E. Giraldo, …)– Common BPMs identified as main concern at LMC50 –

later slide– All other BI independent for two beams

Orbit and tune feedback – very important– Looks OK (J. Wenninger, R. Steinhagen)

Q/Q' systems are independent for both beams Orbit feedback does not use common BPMs Need to decouple radial control Check possibility of variable Q reference ?


Collimation – Settings according to beam, mostly separate

– IR2 vertical tertiary collimators (TCTVs) require special consideration ALICE zero-degree calorimeter (ZDC) shadowing with

present location, we have to minimise real crossing angle

IR2 modification (see LMC 79) to move (new) TCTVs behind ZDC should be done for 2012 p-Pb run, otherwise may have to request open TCTs L2 (for incoming proton beam, as for Pb beam in 2010-11) – machine protection ?


BPMs (from Eva Calvo Giraldo, Rhodri Jones) BPM sensitivities for B1 and B2 are completely

independent (no limit of p intensity from this) Strip-line monitors can have two types of issues: – Both beams cross exactly at the same time

Located so cannot happen with 100 ns once the frequencies are cogged.

While frequencies separate, both beams will sometimes cross BPM at same time. – Then orbit can have errors 5.5 mm (10% bunch

intensity ratio, unlikely), or ~1 mm (50% bunch intensity ratio, more likely).

– False triggers: see next slide …


BPMs (continued)– False triggers: – possible for bunch intensities close to the high end of

each sensitivity mode range (B2 port of monitor sees B1-induced signals), can trigger acquisitions in the wrong channel, degrading orbit precision.

– Nevertheless with bunch intensities of 1e10 p/bunch and ~6e9 charge/bunch, there will be no false triggers.

Common BPMs: disabled on the orbit feedback system most of the time. – Only used in squeeze.

Ghost bunches: – Simulations showed that they could lead to an

error of up to ~120 μm. (More detail in notes from E. Giraldo)


Machine Protection (preliminary!) Most protection issues dealt with independently in

the two rings, either for p or Pb Eg, either beam becomes unstable, losses

Injection: must avoid trying to inject wrong beam in wrong ring– Since SPS is phase-locked, impossible to pass

momentum acceptance of transfer lines– Protect SPS and transfer lines with software

interlocks – detailed scheme being implemented


Injection Nominal Pb beam has 592 bunches, 100 ns

minimum bunch spacing p-Pb collisions in LHC will be done with a special

proton filling scheme, designed to match the Pb beam – Other p-p schemes are different– Schemes worked out in 2005 by K. Schindl and

C. Carli– This beam is now ready in the injectors (S.

Hancock et al in PS) Useful flexibility in bunch intensity available


System Status Date for test Possible issuesBPM - BPM sensitivities for B1 and

B2 are completely independent.- Two independent RF systems used to trigger BPMs

1 . Both beams cross exactly at the same time in these monitors Issue only until cogging is doneSolution: disable common BPMs until cogging is done 2. False triggers: NO with bunch intensities of 1e10 p/bunch (p+)and ~6e9p/bunch (Pb)

RF RF settings completely independent for each ring

RF Rephasing 28.08.2011 (MD3) Last test in Oct 2010 showed high beam losses during rephasing

Feedback Fully decoupled radial loop - p+p+: 05.09.2011 (Start up after TS)- p+Pb: MD4 Nov

Interlocks for MP Protect TL&LHC from wrong particle injection

- 28.08.2011 (MD3)- SIS interlock by Jorg in place- Injection line TT10 settings consistent after TS

Timing: Particle type & Accelerator Mode

-Two telegram words to define the circulating particle type: Proton, Pb82, Ar18, D, Xe54 - Accelerator Mode: PROTON ION PHYSICS

- In place after the September TS


From R. Alemany

Proton-lead feasibility study in 2011 Remarks:– This is a new way to operate LHC, very little

previous experience in other hadron colliders– We do not know if it will work!

Could be strongly limited in intensity/luminosity No resources for study (will change on 15/10/2011) Discussed at CERN Machine Advisory Committee

https://indico.cern.ch/contributionDisplay.py?contribId=12&confId=149070

Aims: – Inject and ramp with unequal RF frequencies

Possible bonus, if all goes extremely well:– Re-phase RF to collide beams at proper IPs and

make a few hours collisions


https://indico.cern.ch/contributionDisplay.py?contribId=12&confId=149070

p-Pb filling Beam 1: 100 ns proton beam, ~1010 p/bunch– Should be something close to Nominal Pb

scheme Beam 2: start with a few (probably 2) Pb bunches

for MP reasons– If we succeed in ramping and manually re-

phasing the RF, this could give 1 (or 0) collision/turn in each experiment

– More than this is unlikely but not impossible– Need to clarify conditions for declaring Stable

Beams


Potential luminosity In 2011 feasibility test


20 0 Pb

2Pb

26 -2 -1 8 10

44 *10 cm s (1 colliding bunch, ~10 Pb/bunch, ~10 p/bunch

nominal emittance 1.5 m)but some emittance blow-up, intensity loss likely,

c Pb p b b

n

k N N f k N fL

5

combined factor 3 ?1.8 barn

10 events/h ?

Possibly higher if we have time to try some bunch trains.

Potential luminosity (2) In possible 2012 physics run at 3.5 Z TeV

But we should really wait until after the feasibility test to say anything meaningful.– (With extreme optimism, using full proton

intensity we can dream of factors ~15 more …).


20 0 Pb

2Pb

28 -2 -1 8 10

44 *3 10 cm s (~300 colliding bunch, ~10 Pb/bunch, ~10 p/bunch

nominal emittance 1.5 m)but reduction factors almost certainly apply

c Pb p b b

n

k N N f k N fL

Accessible energies and CM rapidities


33

p-p Pb-Pb p-Pb / TeVE 0.45-7 287-574 (2.7-7,287-574)

/ TeVNE 0.45-7 1.38-2.76 (2.7-7, 1.38-2.76)

/ TeVs 7-14 73.8-1148 48.9-126.8

NN / TeVs 7-14 0.355-5.52 3.39-8.79

CMy 0 0 -2.20

NNy 0 0 +0.46

p p

p P b

P b P b

0 1 2 3 4 5 6 70

2

4

6

8

1 0

1 2

1 4

p p T eVc s N

NTeV1 2

1 2

1 2

1 2

1 2

Charges , in rings with magnetic field set for protons of momentum

2 ,

1log2

p

NN p

NN

Z Zp

Z Zs c pA A

Z AyAZ

Possible range of collision energiesMinimum p-Pb energy for equal revolution frequency.Relations between these numbers are a simple consequence of the two-in-one magnet design

Commissioning and run schedule Different from 2010, now available and will be

kept updated:– http://lhc-commissioning.web.cern.ch/lhc-commissioning

/commissioning-ions.htm– Must commission ALICE squeeze– Interruption by Technical Stop

Delays not impossible …– Commissioning and p-Pb feasibility test are

interleaved of necessity (beam availability)


http://lhc-commissioning.web.cern.ch/lhc-commissioning/commissioning-ions.htm

http://lhc-commissioning.web.cern.ch/lhc-commissioning/commissioning-ions.htm

Schedule in 2011


35

Set up p beam, Pb injection, test injection of Pb on p (2 shifts designated MD)

Time to think … review p-Pb

Set up ALICE squeeze with protons, aperture measurement

Test ramp of p-Pb, while p still available from injectors, possible collisions

Use of physics time for MD will of course be minimised, but this is a very tight schedule.

Pb beam, ramp, squeeze, crossing angles, collimation in two instalments

Conclusions (1) Operational cycle for physics taking shape Matching p beam ready in injectors– Flexibility on intensity available

Check-out of all LHC systems complete– No showstoppers for p-Pb

Machine protection, ~OK Main uncertainty remains beam dynamics– Overlap knock-out resonances very unlikely to

be a problem– Possible diffusive emittance growth ?

Studies advancing …


Conclusions (2) Feasibility test this year is crucial– Plan carefully, interleaved with Pb-Pb

commissioning– New, more complicated mode of LHC operation– Risks of being cut short by limited time …

We must be very cautious about projections for 2012– Today’s discussion could usefully define

minimum performance acceptable for physics Go-ahead from Chamonix workshop ?– 6-10 February 2012


A second opinion: CERN Machine Advisory Committee, August 2011


38

Possible Proton Lead running in 2012 Findings:Asymmetric proton-lead operation is considered for 2012 pending a successful test at the end of 2011. In the test it is planned to demonstrate injection and acceleration of the proton and lead beams with unequal revolution frequencies, a condition necessitated by the equal magnetic fields in both rings due to the 2-in-1-magnet design. The checkout of all system (rf, instrumentation, collimation, machine protection) is almost complete, and no showstoppers were found. Perhaps most importantly the test will also allow evaluating the beam dynamics issues stemming from unequal revolution frequencies, which have been found detrimental in RHIC when operating under such conditions.Comments:The dominating beam dynamics effect from unequal revolution frequencies is likely emittance growth due to the modulated long-range beam-beam interactions. Estimates for RHIC for d-Au operation with unequal revolution frequencies give a time to double the emittance of order minutes. Applying the same analysis to the LHC case one would expect an emittance growth effect that is an order of magnitude smaller with the proposed bunch intensities.With the detailed LHC calculations to estimate the emittance growth completed, comparison with the measured emittance growth in the planned test will be possible. Since an order of magnitude more protons are available than requested for the physics program, the detrimental effects on the lead beam can be enhanced and evaluated over a wide intensity range.

the lhc as a proton-nucleus collider: status and plans john jowett

Documents