collimation systems part 2 · s. redaelli, summer student lectures, 04/08/2016 main points to...

Stefano RedaelliCERN, Beams Department

Accelerator and Beam Physics group

Collimation Systems Part 2

2016 CERN Summer Student Lectures3rd-4th August 2016

CERN, Geneva, Switzerland

S. Redaelli, Summer Student Lectures, 04/08/2016

Outline

2

Recap.: main points of part 1/2

LHC collimation design

Operational performance at the LHC

Collimation simulations

Advanced collimation concepts

Outlook


Main points to retain from part 1 (i)

3

Beam collimation is essential in modern high-power machines to safely dispose of unavoidable beam losses (beam halo cleaning). 

LHC main concerns:  (1) minimize risk of quenches with 360 MJ stored energy, (2) passive machine protection in case of accidental failures. 

Many other important roles, but the 2 above were driving the system design.

Collimation is achieved by constraining the transverse amplitudes of halo particles: collimator jaws are set close to the beam to shield the aperture. Many sources of beam losses (collisions, gas or beam scattering, operational losses,...) are modelled by looking at the time-dependent beam lifetime.  

Required cleaning depends on minimum allowed beam lifetime.

We have see the key parameters involved in the specification of collimation systems (beam intensity and energy, assumed lifetime, quench limits).Single-stage collimation: efficiencies up to ~97-99%. This is not enough: the leakage must be reduced by another factor 100-1000 to avoid quenches. 

Many collimators are needed to catch efficiently high-energy halo particles.


Main points to retain from part 1 (ii)

4

A multi-stage collimation can provide the missing factors and fulfil the cleaning challenge! 

Secondary collimators are placed at optimum locations to catch product of halo  interactions with primaries (secondary halo+shower products). Other collimators are needed to achieve ~1e-5 → complex multi-stage hierarchy.

Dedicated momentum cleaning might be needed if energy losses  are a concern. 

Special optics solutions to protect the off-momentum aperture bottleneck,  otherwise using the same multi-stage approach as for betatron cleaning.

Back-bone of collimation placed in dedicated warm insertions, but some collimators also used for local protection of sensitive magnets.

LHC collimation: unprecedented complexity in particle accelerators!  A total of 44 collimators per beam, ordered in a pre-defined collimation hierarchy: two dedicated warm insertions (2-stage collimation+shower absorbers), local cleaning in experiments, physics debris cleaning and protection collimators.


The aperture sets the scale

5

Cold aperture

Circulating beam

Primary beam halo

Primary collimator

Secondary collimators

Tertiary beam halo  + hadronic showersSecondary beam halo  

+ hadronic showers

Shower  absorbers

Cleaning insertion

Tertiary collimators

Bottleneck

Arc(s) IP

Protection devices

The collimator settings are determined by the machine aperture that we want to protect. LHC design value: Cold aperture = 10 σ.

Typical collimator settings: primaries = 6 σ - secondaries = 7 σ.

10 σ


Outline

6






Outlook


Workflow for collimation design

7

Machine aperture

Quench limitsBeam parameters

Loss assumptions

Collimator designJaw materials

Collimator settings

Cleaning inefficiency

Iterate

Losses on collimatorsIterate


A multi-disciplinary topic...

8

The complete design chain rely on different key ingredients:

Tracking models

Collimation scattering models

Energy deposition simulations

Thermo-mechanical analysis

Operational performance

Standard chain of tools developed and used at CERN:(1) SixTrack with collimation

(2) FLUKA (3) ANSIS / AutoDyn

Important effort worldwide to extend tools:MARS, Geant4, Merlin, BDSIM, ...

Recent workshop within HiLumi-WP5:https://indico.cern.ch/event/275446

https://indico.cern.ch/event/275446


Possible collimator designs

9

x

y

x

y

x

y

Fixed collimators (masks): square, circular, elliptical, ...

x

y

x

y

x

y

Movable collimators: L-shaped, one-sided, two-sided.

LHC choice!


IR7 collimator settings at 450 GeV

10

19.8 19.9 20 20.1 20.2 20.3-20

-15

-10

-5

0

5

10

15

20C

ollim

ator

gap

s [ m

m ]

Longitudinal coordinate, [ km ]

IP7

Gapmin = ± 4.6 mm

ATCP = 5.7 σ ATCS = 6.7 σ ATCLA = 10 σ

3σy 3σx

� =�

��

S. Redaelli, Summer Student Lectures, 04/08/2016 11

19.8 19.9 20 20.1 20.2 20.3-20

-15

-10

-5

0

5

10

15

20C

ollim

ator

gap

s [ m

m ]

Longitudinal coordinate, [ km ]

IP7

ATCP = 6 σ ATCS = 7 σ ATCLA = 10 σ

Gapmin = ± 1.1 mm

3σy

3σx

Optimum settings can only be guaranteed  with high-precision movable collimators jaws!

10% of sigma = ~ 20 micrometers!

IR7 collimator settings at 7 TeV� =

��


LHC collimator design

12

Beam

Main design features:• Two jaws (position and angle)

• Concept of spare surface

• Different angles (H,V,S)

• External reference of jaw position

• Auto-retraction• RF fingers • Jaw cooling

A. Bertarelli et al.


LHC collimator “jaw”

13

Beam

Special “sandwich” design to minimize the thermal deformations: Steady (~5 kW) ➙ < 30 μm Transient (~30 kW) ➙ ~ 110 μmMaterials: Graphite, Carbon fibre composites, Copper, Tungsten.

Collimating Jaw (C/C composite)

Main support beam (Glidcop)

Cooling-circuit (Cu-Ni pipes)

Counter-plates (Stainless steel)

Preloaded springs (Stainless steel)

Clamping plates (Glidcop)

Carbon jaw  (10cm tapering for RF contact)


RF contact Longitudinal strip (Cu-Be)

Beam

Vacu

um ta

nkJa

w (C

arbo

n)

A look inside the vacuum tank

What the beam sees!

A. Bertarelli, A. Dallocchio


Complete collimator assembly

15

Beam

Motors   position survey systemBellows

Support

Quick plug-in system

Vacuum tank


Complete collimator assembly

16

Beam

Motors   position survey systemBellows

Support

Quick plug-in system

Vacuum tank


Tunnel layout:Tertiary collimators in IR1

Beam


Outline

18






Conclusions


Collimator gaps and betatron cuts

19

3 primary collimators are needed to protect the

machine against transverse betatron losses.

Only horizontal collimation for momentum losses.

Horizontal Vertical Skew

-4 -3 -2 -1 0 1 2 3 4-4

-3

-2

-1

0

1

2

3

4

y [ m

m ]

x [ mm ]

Front view (x,y)

: Normalized gap  (beam size units)N� =

g

21�z

xc ± N� · �z : Collimator jaw   positions

: RMS betatron  beam size�z =

��z

�z�

: normalized emittance✏z/�

: beta functions�z

: collimator gap in millimetersg

z � (x, y) : Hor. and Ver. planes

Settings: units of local beam size


Smallest collimator gaps in operation

20

3σ beam envelope

Transverse cuts from H, V and S primary collimators in IR7 2€ coin

The LHC beam carrying ~250MJ passes more than 11000 per second

in such small collimator gaps!


Fixed display in the LHC control room showing the IR7 collimator gaps.

Side view of the vertical TCP

21

60 cm flat active length, gap = ± 1.05 mmBeam: RMS beam size σv = 250 microns!

Beam

2€ coin

L. Gentini


6.5TeV collimation gaps

22

How do we set the collimators so tightly around

the circulating beams?

https://op-webtools.web.cern.ch/vistar/vistars.php?usr=LHCCOLLB1

https://op-webtools.web.cern.ch/vistar/vistars.php?usr=LHCCOLLB1


Collimator beam-based alignment

23

Closed orbit

β-beat!

Due to the very small gaps involved, collimators cannot be set deterministically using nominal parameters: alignment errors, orbit

imperfections and optics errors cause uncertainties large compared to gaps.

Beam orbit and beam size at each collimator is measured with beam-based alignment techniques.

Normalized collimator settings must be converted to positions in [mm]:• Center the two collimator jaws ➙ Need local orbit!• Adjust the gap to the correct setting ➙ Need local beam size!


LHC alignment technique

24

(1) Reference halo generated with primary collimators (TCPs) close to 3-5 sigmas.(2) “Touch” the halo with the other collimators around the ring (both sides) → local beam position. (3) Re-iterate on the reference collimator to determine the relative aperture → local beam size.(4) Retract the collimator to the correct settings.Tedious procedure that is repeated for each machine configuration.

Beam

Reference collimator

Collimator i

BLM

1

Beam


BLM

3

Collimator i

Beam

Reference collimator Collimator i

BLM

2

Beam


Collimator i

BLM

4


How did we make it faster

25

MAY 2010 MAR 2011 MAR 2012 MAY 2012 MD OCT 2012 MD0

5

10

15

20

Setu

p Ti

me

per C

ollim

ator

[min

]

Collimator Alignments

Setup time per collimator (2010-2012)

12.5 Hz (until 2015)100Hz (2016) 1.0 Hz

Movements 8.0 Hz ↦ 50Hz

PhD thesis work G. Valentino

1) 2010: fully manual procedure > 15 min/device  Limitation of operational efficiency

2) 2011: automated procedure based on feedback  loop between BLM and motors

3) Since 2012: further improved algorithms, faster rates of BLM acquisition and settings trims

4) Since 2015: BPM collimators


BPM collimator design

26

Courtesy of A. Dallocchio for the MME team

18 new BPM collimators installed in experiment and dump regions.

Aims: faster precise alignment; continuous orbit monitoring.

New software implementation to align these collimator through a BPM-based feedback loop.

Phys. Rev. ST Accel. Beams 17, 021005 (2014)��


BPM alignment performance

27

1 hour vs 20 seconds

Time [s]0 500 1000 1500 2000 2500 3000 3500

Jaw

pos

ition

s [m

m]

-10

-5

0

5

10

Time [s]0 2 4 6 8 10 12 14 16 18 20 22

Jaw

pos

ition

s [m

m]

-10

-5

0

5

10BPM alignment: done at large gaps, several collimators in parallel!Improved safety: jaws far from circulating beams. Can be done with any beam intensity.

G. Valentino

Standard method: collimators closed until each jaw touches the beam halo.One collimator per beam at a time.


Beam validation through “loss maps”

28

After having setup the collimators, we need a direct measurement of what the beams “will see” and of how the collimation system will behave in presence of high beam losses! 

Can we exclude setting errors? Is the setting hierarchy respected?  Is the local cleaning in cold magnets as expected for a given hierarchy? Does the system - and the machine - provide stable performance in time?

Each set of settings of the collimation system is validated through loss maps with low-intensity beams (few bunches)Beam loss rates are abnormally increased in a controlled way to simulated large beam losses that might occur during nominal high-intensity operation. 


Collimation cleaning

29

0 5 10 15 20 250

0.05

0.1

0.15

0.2

0.25

Longitudinal position [ km ]

Beam

loss

es [

Gy/

s ]

IP2 IP3 IP4 IP5 IP6 IP7 IP8 IP1IP7:Betatroncleaning

What is going on there?

Beam 1

3600 beam loss

monitors (BLMs)

along the  27 km

during a loss map

MEASUREMENTS


Off-momentumDump

TCTs

TCTsTCTs

TCTs

Betatron

1/10000

Collimation cleaning: 4.0 TeV, β*=0.6 mLo

cal c

lean

ing

inef

ficie

ncy

0.00001

0.000001

Beam 1

Highest COLD loss location: efficiency of > 99.99% ! Most of the ring actually > 99.999%

B. Salvachua

MEASUREMENTS


Zoom in IR7

31

1/10000

Critical location (both beams): losses in the “dispersion suppressor”.With “squeezed” beams: tertiary collimators (TCTs) protect locally the triplets.

B. Salvachua

MEASUREMENTS


Outline

32






Outlook


Do we understand the observed collimation

losses? Can we predict them accurately for new

machines/configurations?


Simulation tools

34

Accurate tracking of halo particles  6D dynamics, chromatic effects, δp/p,  high order field errors, ...

SixTrack†

Detailed collimator geometry  Implement all collimators and protection devices, treat any azimuthal angle, tilt/flatness errors

Scattering routine  Track protons inside collimator materials K2

Detailed aperture model Precisely find the locations of losses BeamLossPattern

All combined in a simulation package for collimation cleaning studies:

G. Robert-Demolaize,  R. Assmann, S. Redaelli,  

F. Schmidt, A new version of SixTrack with collimation and aperture interface, PAC2005

Collimator jawIncoming

halo particle

An illustrative scheme

† See also talk by F. Schmidt .


Example: trajectory of a halo particle

35

Interpolation: ∆s=10cm(270000 points!)

0.02

0.03

0.0420 20.5 21 21.5 22 22.5 23 23.5 24

-0.2-0.16-0.12-0.08-0.04

00.040.080.120.16

0.2

s [ km ]

Aper

ture

/ be

am p

ositi

on [

m ]

23.05 23.1 23.15 23.2 23.25 23.3 23.35-0.06

-0.04

-0.02

0

0.02

0.04

0.06

s [ km ]

Aper

ture

/ be

am p

ositi

on [

m ]

IR8

Magnet locations : ∆s ≤ 100mTrajectory of a halo particle

A dedicated aperture program checks each halo particle’s

trajectory to find the loss locations.

∆s=10cm


Example of simulated “loss map”

36

0 5 10 15 20 25 30

106

107

108

109

1010

1011

Longitudinal coordinate, s [ m ]

Loss

rate

per

uni

t len

gth

[ p/m

/s ]

IP2 IP3 IP4 IP5 IP6 IP7 IP8 IP1

Beam 1

Nominal 7 TeV case, perfect

machine



37

0 5 10 15 20 25 30

106

107

108

109

1010

1011


Loss

rate

per

uni

t len

gth

[ p/m

/s ]


Beam 1


machine



38

0 5 10 15 20 25 30

106

107

108

109

1010

1011


Loss

rate

per

uni

t len

gth

[ p/m

/s ]


Beam 1

-20 -10 0 10 20

-20

-10

0

10

20

Y [ m

m ]

X [ mm ]

Sloss: s1=20290m; s2=20300m.Nloss = 304 Statistics for a typical case:

20-60 million protons, 200 turns.Up to [5.4x106m] x [60x106p] =

3.24 x1014 m = 0.034 lightyears for one high-statistics simulation case!


machine

This simulation results are used for detailed energy deposition studies!At CERN, this is done with program FLUKA. Output provided to magnets

and collimator design teams.


Comparison with measurements

39

Simulations

Measurements

R. Bruce


R. Bruce

Comparison in the betatron cleaning

40

3000 3200 3400 3600 3800 400010!7

10!6

10!5

10!4

10!3

10!2

10!1

100

s!m "

loss#!loss

atTC

P"

Measured

Collim atorWarmCold

3000 3200 3400 3600 3800 400010!7

10!6

10!5

10!4

10!3

10!2

10!1

100

s!m "

loss#!loss

atTC

P"Simulated

IR7Collim atorWarmColdSimulations

Measurements

Losses in dispersion suppressor: limiting location

Note the log scale!

Cross-talk on BLM signal from upstream losses

We are comparing measured BLM signals against losses in the collimators or protons touching the aperture! Proton tracking alone

is not sufficient to reproduce the deposited energy profile!


Integrated simulations

4121

E. Skordis for the CERN FLUKA team

>700m if IR7 in FLUKA geometry

Tracking results input in FLUKA: energy deposition simulations compared to measured BLM signal.


Outline

42






Conclusions


Why studying new collimation concepts

43

LHC (Design) HL-LHC FCC-hh (Baseline)

Beam energy 7 TeV 7 TeV 50 TeV

Beam intensity 3.2 x 1014 6 x 1014 10 x 1014

Stored energy 360 MJ 690 MJ 8500 MJ

Power load ( τ=0.2h) ~500 kW ~960 kW ~11800 kW

Energy density ~1 GJ/mm2 ~1.5 GJ/mm2 ~200 GJ/mm2

The approved High-Luminosity upgrade of the LHC and future colliders like FCC call for an upgrade of the

collimation system and for new techniques.See talks by M. Lamont and D. Schulte


Dispersion suppressor cleaning

44

L. Gentini, D. Ramos


New collimator materials

45

Test%1%(1%LHC%bunch%@%7TeV)%

Test%2%(Onset%of%Damage)% Test%3%

(72%SPS%bunches)%

Tertiary collimator that protects the inner triplet

Excellent results: full MoGR jaw survived as well as CFC to impact of 288b of 1.3x1011p

with σ=350μm (density beyond LIU)A. Bertarelli, F. Carra

Copper Diamond: candidate tertiary collimator material, 10-15 times more robust.


New collimator materials

46

Test%1%(1%LHC%bunch%@%7TeV)%

Test%2%(Onset%of%Damage)% Test%3%

(72%SPS%bunches)%

Tertiary collimator that protects the inner triplet

Excellent results: full MoGR jaw survived as well as CFC to impact of 288b of 1.3x1011p

with σ=350μm (density beyond LIU)A. Bertarelli, F. Carra

Copper Diamond: candidate tertiary collimator material, 10-15 times more robust.

Advanced material composites bases on carbon and on various metals promise a significant

improvements of the present collimator!


Crystal collimation

47

MCS ~ 3.4 μrad (7 TeV)

Amorphous (0.6 m of C)

Crystal  (Channeling) 

(3 mm Si) ~ 40-50 μrad (7 TeV)

Bent crystals allow bending high-energy particles trapped between lattice planes.

Crystal collimation setup installed in the LHC-IR7 for beam tests.

Application for hadron beam collimation: Exploit large angle (~50μrad) and the

reduced diffractive events!

Equivalent B field = 300T (l = 4mm)  Challenges: small angular acceptance,

localization of large losses (0.5-1.0MJ) in one single collimator.

Beam

Absorber

Crystal  ???

Crystal-based collimation 

Primary  Secondaries  Absorbers 

Beam

Standard multi-stage collimation 

Promises of crystal collimation at the LHC: 1. Improved collimation cleaning; 2. Reduce electro-magnetic perturbations  of collimators to the beams (impedance):  less secondary collimators at larger gaps; 3. Much improved cleaning for ion beams.


Do crystals really work?

48

Beam tests 2015: First demonstrations of channeling of proton beams at 6.5TeV and of

Pb ion beams at 450GeV

Beam core

Collimator

CrystalHalo

(1)

(1) Angular scan: strong reduction of local losses in channeling compare to amorphous.

~1/30

Beam

loss

es a

t cry

stal

[ a.

u. ]

Crystal angle [ μrad ]

Loss rates in amorphous

Reduced losses in channeling

Example: scan at 450GeV

Secondary collimator position [ mm ]

Loss

es a

t col

limat

or [

a.u.

]Example: 

scan at 6.5TeV

(2) Linear collimator scan: measures the profile of the channeled halo.

(2)

Channeled halo

Beam core

Offset at collimator

Fresh results of collimation cleaning with crystals from LHC bema tests last Friday


PRELIMINARY cleaning with crystals

49

Beam

Longitudinal position [ km ]19.8 19.9 20 20.1 20.2 20.3 20.4

Beam

loss

es [

Gy/

s ]

10 -8

10 -7

10 -6

10 -5

10 -4

10 -3

10 -2

10 -1

10 0IP7

Longitudinal position [ km ]19.8 19.9 20 20.1 20.2 20.3 20.4

Beam

loss

es [

Gy/

s ]

10 -8

10 -7

10 -6

10 -5

10 -4

10 -3

10 -2

10 -1

10 0IP7

“Crystal”

Note: data not yet properly normalized.


Collimation techniques discussed so far are ‘passively’ catching losses determined by

the machine itself.

Can we actively control halo losses and diffusion

rates on collimators?


Hollow electron beams

51

Hollow electron lenses, running co-axial to the LHC beam over a short distance, generate an electro-

magnetic field that affects only tails and not the core.

Can enhance the performance of the a collimation system by actively

controlling the time profile of losses!

G. Stancari, FNAL


LHC hollow e-lens design

52

D. Perini, MME

First design study for installation in the LHC point 4 (RF)

LHC beam


LHC hollow e-lens design

53

D. Perini, MME

First design study for installation in the LHC point 4 (RF)

LHC beam


Outlook

54

Beam cleaning and collimation becomes increasingly important for large circular accelerators (“supercolliders”).The basic design strategy for multi-stage collimation in high-energy hadron accelerators was presented. 

Key design parameters reviewed, collimation settings worked   out from aperture. Seen how this defines the collimator design.The present LHC collimation system was presented  

Detailed look at collimation operation and performance.

Simulation of collimation cleaning were discussed.Selected examples of advance concepts were presented.Advertisement: Looking for motivated students to work on the present and upgraded systems. Contact me if interested!

collimation systems part 2 · s. redaelli, summer student lectures, 04/08/2016 main points to...

Documents