The RHIC Collider

The RHIC Collider

Fulvia Pilat

Collider Workshop, JLAB, February 24,

2009

RHIC Collider Complex

100x100

GeV/

u

ions

250x250

GeV polarized

p

Chronology:

1996 commissioning

AtR

1998 sextant test

1999 engineering runs

2000 first collisions

10 years of operations

Outline

Introduction

Collider

design

and

evolution

Optics

,

interaction regions

,

correction systems

Validation of design and correction schemes

Operational use of corrections systems (ex: IR corrections)

Commissioning



operations

: increasing machine performance

example:

lower

beta

*

Recent

developments

Stochastic cooling

of bunched beams

Tune, coupling, orbit and chromaticity

feedback

Lesson learned

from RHIC

Optics for Run-10 Au-Au b*=0.7m

Optics for Run-9 P-P b*=0.7m

Optics: zoom into triplets in IR6 (STAR)

Interaction regions: layout and correction systems

Corrections systems

Orbit correction

BPM + dipole correctors

Coupling correction

3 families of skew quads

120 deg in the

arcs (

2 wired up in software



orthogonal system)

1 skew quad/triplet

for local compensation of the roll misalignment of the triplet quads

(

no experimental solenoid compensators

– all done by the skew quad families)

Chromatic corrections

2 families of sextupoles

in the arcs



linear chromaticity

4 additional sextupole families

in arcs



nonlinear chromaticity (added later)

IR correction packages (each triplet)

1

dipole H

, 1

dipole V

, 1

skew quadrupole

1

normal

and 1

skew sextupole

2

octupoles

1

decapole

2

dodecapoles

(skew octupole and dodecapole layers exists but are not powered)

Validation of design and correction schemes

In the design phase we did extensive modeling and simulations to validate the design and the correction schemesBuilt a offline machine model for extensive DA simulations, including:Optics configurationsMeasured magnetic errors in arc and IR magnetsMeasured misalignments and roll errors in cold masses and cryostatsBeam-beam (weak-strong approximation)Other performance issues dealt with stand-alone codes:Intra-beam scatteringBeam-beam (strong-strong)Electron cloudPolarization trackingSelected capabilities of the offline model are part of the online machine model

But, over

the operational life of the machine we ended relying mostly

on

b

eam based

corrections

(orbit, coupling, IR corrections, nonlinear chromaticity)

The magnetic errors in an accelerator magnet can be described in terms of the multipole errors an and bn defined as: An excursion of the local orbit through a region having non-linear fields generates feed-down effects to lower order field harmonicsThe most useful observable effects come from the feed-down to the beam closed orbit and betatron tunesIt is possible in theory to infer local non-linear effects both from the measure of residual RMS orbit and of tune shifts generated by a local orbit bump in the IR. Given existing limitations on the resolution of the orbit measurement and on the allowable bump amplitude at the triplets, in practice we have used so far almost exclusively the measurement of tune shift as a function of bump amplitude for non-linear correctionThe measured tune shifts arise from either the feed-down to the normal gradient or from the repelling effect of linear couplingThe tune shift (ΔQ) and the linear coupling term (Δc) for different bump planes (H and V) and for different multipole errors (normal, skew, even and odd orders) can be expressed as follows (where cn is either the an or bn and z is x or y): : This table implies that for reasonable measurement of a tune shift due to I.R. magnetic field errors, the following bump types should be used to identify the relevant multipole: horizontal for Sextupole(b2), vertical for Skew Sextupole (a2), horizontal and vertical for octupole (b3) etc. A diagonal bump for skew octupole is necessary In order to simplify the identification of individual multipoles using the observed tune shifts, the conditions should be such that the tune shifts produced by coupling are negligible when compared with the tune shifts from the normal gradient change

Where the functions

g

and

h

are defined as:

IR correction method - theory

Before Correction

After Correction

Example: normal sextupole IR correction

Schematics of IR bumps

Tune shift vs. amplitude

Before correction

Tune shift vs. amplitude

After sextupole correction

Bump power supply



Before Correction

After Correction

Example: skew quadrupole IR correction

Beam decay evolution during the correction

Tune shift vs. amplitude

Before correction

Tune shift vs. amplitude

After skew sextupole correction

Horizontal Tune Shift Before Correction

Vertical Tune Shift Before Correction

Tune Shifts After Correction

Example: octupole IR correction

Before correction H

After correction

Before correction V

“The tune modulation (10 Hz due to triplet vibration via

feed-down

effect; that is, tune modulation due to off-axis beam in sextupoles driven by off-axis beam in triplet

quads) was observed to reduce by a factor of 2-3 after non-linear corrections.

Correction benefits: reduction of tune modulation

Operational correction for IR decapole and dodecapole

15

Generic scanner

scans magnet strength (for list of magnets)

observes beam loss rate

minimizes beam loss rate with strength

Decapole/Dodecapole correction result

16

Tested effect of 10- and

12-pole correctors on beam loss rate by switching off all correctors

10998

Estimate of luminosity gain

17

Nonlinear chromaticity – Run 10 experience

Momentum apertur

e essential for

re-bucketing

at store (turn on

196 MHz

RF system at store – on top of the accelerating

28 MHz

RF system to get more beam in the experiment acceptance)

Nonlinear chromaticity

reduces the available momentum aperture

2

nd

order chromaticity

minimized for phase advance

of (2n+1)*

p

/2

between 2 equal IP’s

Running now with increased arc phase advance from 86 to 93deg/cell (IBS reduction lattice, lower dispersion in the arcs)

Also lower beta* (0.6m instead of 1m) reduced aperture in the triplets

Insufficient momentum aperture for re-bucketing

Tried nonlinear chromaticity corrections but

measurements not reliable at small radial offsets

Had to step back beta* from 0.6m to 0.7m and shift the tunes for momentum aperture

Outline

Introduction

Collider design and evolution

Optics

,

interaction regions

,

correction systems

Validation of design and correction schemes

Operational use of corrections systems (ex: IR corrections)

Commissioning



operations

: increasing machine performance

example:

lower

beta

*

Recent developments

Stochastic cooling

of bunched beams

Tune, coupling, orbit and chromaticity

feedback

Lesson learned

from RHIC

Performance increase

Heavy ion runs

Polarized proton runs

Integrated luminosity L [pb

-1

]

Integrated nucleon-pair luminosity LNN [pb-1]

Increase:Bunch intensity (limits: instabilities)Number of bunches (limits: electron cloud)Total intensity (limits: losses, beam-beam)

Decrease:

Beta star

(limits: aperture, lifetime)

Emittance

(via stochastic cooling and

electron cooling at low energies)

Beta* squeeze at RHIC

GOALS:Increase of luminosityPreparation for dynamic beta* squeeze with transverse stochastic coolingHISTORY:

Beta* squeeze: methodology

Before beam

the optics matching to lower

b

*

in IP6 and IP8 is turned into a

ramp

with ramp application software The ramp, typically 300 s, is first tested without

beam for

power supply

limits

and the

quench protection system

.

Ramp development

Ramp development follows with 6-12 bunches/ ring. Care is taken to avoid transverse emittance growth to avoid losses in the aperture limiting triplets.

The ramps are done with

tune & coupling feedback

.

Orbits are corrected to to 0.1-0.2 mm rms

Store set-up

We tune for

lifetime

at store (orbit, tunes, coupling, and chromaticity), then steer for

collisions

, compare rates

and test

collimation

.

Optics

measurements with the AC dipole follow.

Measured

b

*

are typically in within 10-15% from nominal, and

b

*

is also verified with

Vernier scans

in operation.

Test of physics ramp and store

We test the new configuration with a

physics store

(56-109 bunches/ring for ramp transmission, collimation, experimental backgrounds. If successful we can use the lower

b

*

in operations. We then readjust

non-linear corrections

for the new configuration, namely local IR triplet correction and possibly non-linear chromaticity corrections.

Example results: d-Au Run-8

We first reduced b* in the yellow ring (gold), where we ran a lattice with higher phase advance per arc cell to minimize intra beam scattering effects. After 2 attempts, the 3rd ramp with tune & coupling feedback brought the beam to store with good transmission A 56x56 physics ramp allowed us to establish that the normalized collision rates ratios between the baseline (yellow at 1m) and the one with squeezed optics (yellow at 0.70m) yielded the expected 15% luminosity increase. Once we established the feasibility of operations with yellow at b*=0.7m, we repeated the development for the blue ring, running deuterons. The entire development took an integrated beam time of ~24h, over a few days. We ran the reminder of the d-Au run with b*=0.7m in both rings, gaining ~30% in integrated luminosity increase for the run.

Outline

Introduction

Collider design and evolution

Optics

,

interaction regions

,

correction systems

Validation of design and correction schemes

Operational use of corrections systems (ex: IR corrections)

Commissioning



operations: increasing machine performance

example:

lower

beta

*

Recent

developments

Stochastic cooling

of bunched beams

Tune, coupling, orbit and chromaticity

feedback

Lesson learned

from RHIC

Dynamic beta* squeeze– Motivation

Run10: longitudinal and vertical Stochastic Cooling (SC) are operational => potential for luminosity increase improve luminosity by a ~factor 2 The goal is to have an application similar to the one used for orbit correction at store: β* as a function of time should follow the change in emittance as achieved by Stochastic Cooling.

To help reaching higher peak luminosity, an application is being developed using the RHIC online model to further push the squeeze of β* in the experimental insertions IR6 and IR8.

2006: first test of longitudinal cooling2007: longitudinal cooling operational 2009: first transverse testsystem installed and tested

M. Blaskiewicz, J.M. Brennan

“signal suppression” demonstrated:

feedback on

feedback off

2010: first test of transverse cooling of ion beams

2012: 200 GeV,

Au+Au

, full stochastic cooling

RHIC luminosity upgrade

(for ions):

Au+Au, 200 GeV: 40 × 1026 cm-2s-1 (×4)(with * = 0.5 m, 56 MHz rf cavity)

Stochastic cooling system

New pickup, Blue longitudinalNew pickup, Yellow longitudinalMicrowave link, upgraded kicker (9 GHz), new low-level enclosureNew pickup, Blue vertical (from 1.)New kicker, Blue vertical

System

Schematics

Mike Blaskiewicz C-AD

28

Predictions for longitudinal cooling Run-10Current profile at 0, 2.5 and 5 hours without burn-off. 4 MV on storage system, IBS suppression lattice(Vertical cooling only dQmin=0.01, dQbare=0)

Blue longitudinal cooling – measurements Run-10

System status:

Yellow transverse operational

Yellow longitudinal being repaired

Blue transverse operational

Blue longitudinal operational

(surprise: ring cross talk)

orbit

tune

coupling

chromaticity

“10 Hz”

dynamic reference orbit control andfeed-forward demonstrated (02/04/10)

extensively improved in

Run-9fully operational in Run-10

ready for test

To counteract 10 Hz orbit jitter from triplet vibrationsUnder development

RHIC Weekly Meeting, February 8, 2010

successful ramp to store (02/04/10)

Orbit, Tune & Coupling, Chromaticity, 10Hz feedbacks

orbit feedback ON

orbit rms in the blue ring

orbit rms in the yellow ring

orbit feedback OFF

APEX Meeting, February 5, 2010

N

b

6Nppb ~ 1E9 (Au)SVD tolerance 100FB gain 10% / 10%

Ramp development with continuous orbit and tune/coupling feedback

RHIC Weekly Meeting, February 8, 2010

01/08/10

01/08/10

Chromaticity measurement algorithm improved: extracts chromaticity from ‘wiggled’ tunes

blue ring, ramp chromaticity measurement

yellow ring, ramp chromaticity measurement

ready for chromaticity feedback test

x

x

x

y

xx

xy

Qx

Qy

Qx

Qy

Chromaticity measurements (prep for feedback)

conclusions

Disclaimer: RHIC in operations for 10 years – thousands of design and operational issues not covered in this talk

Lessons learned:

Flexibility in the design pays off in operations and performance

(example beta*)

Beam-based corrections play a critical role in a SC

hadron

collider

Stochastic cooling of bunched beams is a reality – although a very specialized one. Could/Should become part of the design of new

hadron

colliders at high energies.

Feedback systems enhance performance and operability of a

hadron

collider

(Available to discuss topics not covered here)

APEX

STAR

Polarity