DEVELOPMENT OF A BEAM DEVELOPMENT OF A BEAM LOSS DETECTION SYSTEM FOR LOSS DETECTION SYSTEM FOR

THE CLIC TEST FACILITY 3THE CLIC TEST FACILITY 3

T. Lefevre

• Beam loss monitors for the CLIC Test Facility 3

• Preliminary study done in 2003• Geant3 simulations• First experimental data

• Conclusions & Perspectives

BIW 2004, 5 May 2004

T. Lefevre, M. Velasco , M. Wood, Northwestern University

H. Braun, R. Corsini, M. Gasior, F. Tecker, CERN

T. Lefevre BIW 2004, 5 May 2004

CLIC Test Facility 3CLIC Test Facility 3

• The CLIC Test Facility 3 is built to demonstrate the Compact LInear Collider The CLIC Test Facility 3 is built to demonstrate the Compact LInear Collider feasibilityfeasibility

- Drive Beam Generation :

- Efficient way of producing a high current (35A) high frequency (15GHz) 150MeV and 1.6s electron beam

- Done using a fully loaded linear accelerator (94% RF to beam efficiency) and two rings

- Drive Beam Deceleration stability (two beam acceleration section)

- The beam is strongly decelerated in order to provide the 30GHz RF source

- Study the Beam halo & Beam loss mechanisms

- Provide a 30GHz power source to continue the R&D program on high gradient accelerating structures (150MV/m)

- Housed in the LEP injector complex and scheduled for completion before 2010.

- The construction of the linac will be finished by the end of the year

BLM for CTF3 Linac BLM for CTF3 Linac

T. Lefevre

• Northwestern University joined the CTF3 collaboration in 2003 and we are designing and building the beam loss detection system for the CTF3 linac.

BIW 2004, 5 May 2004

• The operation of a fully loaded linear accelerator

• Heavy beam loading in accelerating structures Transient effects where the energy gain per cavity for the beam head can be twice higher than its nominal steady state value

• Beam transient compensation scheme by adjusting the delay between the beam and the RF signal in the cavities

• We designed our system to observe the beam transient loss

• Dangerous beam losses : > 10% of the total beam charge (6C)

• The protection system will rely on wall current monitor

• The BLM system will be a tool for the optimization of the Linac operation

T. Lefevre

Beam position monitor

Acceleratingstructure

Quadrupoles

e-

Beam loss detectors• Fast time response (ns-10ns) for beam transient study• Segmented X-Y beam loss positioning for tuning

Yz

x

Typical Linac section

Design beam optics

High probability that beam loss occurs in the quadrupole region

BIW 2004, 5 May 2004

Goal : To install the BLM at a position where the beam transient would be lost

BLM for CTF3 Linac BLM for CTF3 Linac

T. Lefevre

Beam pipe simulations : Transverse distribution of the e-/e+ shower

Geant3 SimulationsGeant3 Simulations

100MeV

• The total flux of electrons in the shower is proportional to the electron energy• With higher beam energies, the shower asymmetry is more pronounced

BIW 2004, 5 May 2004

Simulations based on a beam loss corresponding to the ‰ of the nominal beam current e-

Position of observation : 1m dowstream Beam loss

at + Y

T. Lefevre

Beam loss in the Central quadrupole

Geant3 SimulationsGeant3 Simulations

BIW 2004, 5 May 2004

Beam loss Positions of observation

e-

Screening effect of the 3rd quadrupole which

reverses the transverse distribution of the e-/e+ Shower

Z=25cm Z=75cm Z=120cm Z=155cm

T. Lefevre

done by Matthew Wood

Positions of the beam loss

ee-- shower efficiency shower efficiency : Number of particles detected / Number of particles lost

Geant3 SimulationsGeant3 Simulations

BIW 2004, 5 May 2004

e-

BLM’s(size and position)

•35MeV, 0mm beam size, 3mrad beam angle• Beam loss at + Y• Ø40mm detector installed at 15cm from the beam axis

Simulations

• The shower transverse distribution is affected by the presence of Quadrupoles• For losses on the beam pipe the asymmetry corresponds to 50%• Beam loss position more than 2 orders of magnitude difference in the shower efficiency • For losses > 1‰ of beam current Detector must be able to measure currents > 100nA

+ Y+/- X- Y

T. Lefevre

Test on CTF3 in 2003 Test on CTF3 in 2003

Using the collimator in the cleaning chicane to study the beam transient

BIW 2004, 5 May 2004

Collimator

BPM502

BPM690Accelerating structures

BPM402

Quadrupoles

e -

Beam line layoutBeam line layoutDipoles

Injector

Two Aluminum Cathode Electron MultipliersØ40mm, Sensitivity range [100nA-100mA]

Beam loss

monitors

Cleaning chicane

Steerers

First Linac Section

T. Lefevre

0 200 400 600

-3

-2

-1

0

Slit aperture 14 Slit aperture 19 Slit aperture 23 Slit aperture 27

BPM

502

curr

ent

(A)

t (ns)0 200 400 600

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

Slit aperture 14mm Slit aperture 19mm Slit aperture 23mm Slit aperture 27mm

Beam

loss

cur

rent

(mA)

t (ns)

Observation of the beam transient loss Observation of the beam transient loss

Chicane &Collimator

Case 1 : Slit opened

< 80MeV 35MeV

•The slit is opened so that the full beam enters the next accelerating structure.

•The beam transient is then re-accelerated up to 80MeV and is lost somewhere because the beam optics are not adapted to its energy

BIW 2004, 5 May 2004

Case 2 : Slit Closed

< 35MeV

20MeV

•The slit is closed so that the beam transient is stopped in the collimator.

•The rest of the beam enters the next accelerating structure and is accelerated to 35MeV

0 200 400 600 800-5

-4

-3

-2

-1

0 BLM Right BLM Left

BLM

cur

rent

(mA)

t (ns)

0 200 400 600 800-10

-8

-6

-4

-2

0 BLM Right BLM Left

BLM

cur

rent

(mA)

t (ns)

T. Lefevre

Slit closed : Horizontal and Vertical scansSlit closed : Horizontal and Vertical scans

BLMLeft

BLMRight

e-

0 200 400 600 800-5

-4

-3

-2

-1

0 BLM Right BLM Left

BLM

cur

rent

(mA)

t (ns)

0 200 400 600 800

-6

-3

0 BLM Right BLM Left

BLM

cur

rent

(mA)

t (ns)

Beam goes to the Left Beam goes to the Right

Beam goes Down

Beam goes Up

•In vertical scans the beam loss is equally distributed on the two detectors and their output signals are equivalent (<5% difference)

•In horizontal scans the BLM output signals are different in a ratio of 2 (40-60%)

Localizing the beam loss transversely ?

BIW 2004, 5 May 2004

Geant3 expectationsGeant3 expectations

T. Lefevre

ee-- shower efficiency for different shower efficiency for different beam loss positions and energiesbeam loss positions and energies

Beam loss position 1st Quad 2nd Quad

3rd Quad Pipe 0.9m

Pipe 0.3m

Shower efficiency (%)

5.25 10-

4

1.4 10-3 9.3 10-3 2.7 10-2 0.2

35MeV, 3mrad, 0mm beam size

80MeV, 3mrad, 0mm beam size

Beam loss position 1st Quad 2nd Quad

3rd Quad Pipe 0.9m

Pipe 0.3m

Shower efficiency (%)

4.6 10-3 8.9 10-3 4.5 10-2 9.9 10-2 0.28

BIW 2004, 5 May 2004

-2 -1 0 10.04

0.06

0.08

0.10

-2 -1 0 10.0

0.2

0.4

0.6

0.8

1.0

1.2

Transient

BLM

/BPM

cur

rent

(%)

Steerer current (A)

Steady state

T. Lefevre

Vertical scan with the slit opened

0 200 400 600 800-1.5

-1.0

-0.5

0.0

BLM

cur

rent

(mA)

t (ns)

0 200 400 600 800

-1.2

-0.8

-0.4

0.0

Beam centered I=0.5A I=-1A I=-1.7A

Beam

cur

rent

lost

mea

sure

d us

ing

BPM

(A)

Close to the detector

Detector 3rd Quad

3rd Quad

3rd 2nd Quad

35 < E < 80MeV

E = 35MeV

1. Using BPM data’s to estimate the beam current lost in a linac section2. Using the BLM measurement to estimate the Z position of the beam

loss

BIW 2004, 5 May 2004

T. Lefevre

ConclusionsConclusions

• Beam loss transverse positioning works in agreement with the Geant3 predictions

• During this test, the beam losses were relatively high (beam transient ~ 1A) and they were located near the quad’s region (which was consistent with the design lattice)

• Using the BPM’s data & the energy measurements, the BLM system can be used to localize the losses along the accelerator accurately (< 50cm)

• Without the BPM data, one system per section is not enough to monitor beam loss intensity & position (more complicated for beam losses distributed along the linac)

How can we be quantitative : I & Z ?

• Adding detectors every 50cm to get the Z beam loss position:

• Longitudinal positioning using a Cherenkov fiber and a time of flight measurement (already developed at SLAC and TTF)

BIW 2004, 5 May 2004

T. Lefevre

Perspectives Perspectives

BIW 2004, 5 May 2004

• The system to be installed in the next months :

• The detectors are developed at Northwestern University (M. Velasco and A. Dabrowski) in conjunction with Fermilab (G. Tassotto)

• 12 sets of 4 detectors (SEM/SIC) located near the quadrupoles region

• Special set-up to study the losses one a single linac section using 12 detectors

• The signals are then amplified and acquired using 100MHz ADC’s

1mm gap chamber

• Can be operated with gas (ionization) or vacuum (SEM) • Radiation hard

T. Lefevre

CTF3 little shop of horrors CTF3 little shop of horrors

BIW 2004, 5 May 2004

Damage on a Vacuum valve

Spectrometer line

Respect

the steering limitation

T. Lefevre

Suggestion !Suggestion !

BIW 2004, 5 May 2004

The CLIC Test Facility 3The CLIC Test Facility 3

T. Lefevre

Drive Beam generation : Efficient way of producing a 35A beam bunched at 15GHz• Acceleration of a high current beam in a 3GHz fully loaded Linac (95% RF to beam efficiency)

• Production of a high frequency (15GHz) bunched beam using a delay loop and a combiner ring

CLEX : CLic EXperimental area: • Provide a 30GHz power source for the development of high gradient (150MV/m) accelerating structures

• Test the Drive beam stability in the Drive Beam Decelerator (beam loss rate)

• Housed in the LEPPre-injector complex

• Scheduled for completion before 2010

BIW 2004, 5 May 2004

30 GHz Power Source and distribution line

3TeV Compact LInear 3TeV Compact LInear ColliderCollider

T. Lefevre

e- Main Linac e+ Main LinacBDS

Damping rings• 4ps, >10me-/e+ Source

• 42ns, 2,424GeV• 157 bunches (400pC)• 4ps, >50m

Main Linac• 9 1500GeV• 100fs, >1m

CombinerRing 1

Delay loop

CombinerRing 2

Source

25 Drive Beams Decelerators per linac• 1.79GeV, 144A over 56ns• 1ps, >50m

Drive Beam Generator •4.5A over 92s with an final energy of 1.79GeV: 43000 bunches (9.6nC each)• 10ps, >50m

BIW 2004, 5 May 2004

What have been achieved up to What have been achieved up to nownow

CTF3 – Preliminary phase - 2002Low-charge demonstration of electron pulse combination and bunch frequency multiplication by up to a factor 5

Streak camera image of the beam

time structure evolution

LEP injector EPA ring

Operation of a fully loaded linacOperation of a fully loaded linac

T. Lefevre

Drive beam acceleration in 2003

Beam current – BPM 402

4 A

1.5 s

Beam current 4 ABeam pulse length 1.5 sPower input/structure 35 MWOhmic losses (beam on) 1.6 MWRF power to load (beam on) 0.4 MWRF-to-beam efficiency ~ 94%Phase variation along pulse ±4º

RF signals / output coupler of an accelerating cavity

RF phase

RFpower Power to load (beam

off)

phase

Power to load (beam on)

±4º

1.5 s

Heavy beam loading in the accelerating structure

Beam head sees a much higher accelerating field

Strong transient effects so that in the first 50ns of the pulse the beam energy can be twice higher than the energy of the rest of

the beam

Beam transient compensation by adjusting the delaying the RF pulse in the

accelerating structure to suppress the transient effect

BIW 2004, 5 May 2004

BLM detectorsBLM detectors

T. Lefevre

Two different types of detectors (ns time response) have been tested in parallel

- A 4mm thick plastic scintillator (Ø40mm) coupled to XP2020 photomultiplier tube

e-/e+VisiblePhotons

1.75%

e-

Photocathode25% QE • e- / e+ :

[1, 20]MeV 500 - 1000 photo-e-

• & x rays : [10keV,20MeV] 4 - 100 photo-e-

e- current amplification<106- 107

HV

Scintillator

Signal

x &

rays 50mV/50Ω A-pA

- An Aluminum Cathode Electron Multiplier (ACEM) (Ø38mm)

e-/e+

x &

rays

Aluminum cathode100nm thick • e- / e+ :

[1, 20]MeV 1 - 5% SEM e-

• & x rays : [10keV,20MeV] 4.10-6- 2.10-9 SEM e-

e- current amplification<105- 106

e- HV

Signal50mV/50Ω 100mA-100nA

BIW 2004, 5 May 2004

1.5 1.6 1.7 1.8 1.9 2.01E-4

1E-3

0.01

0.1

1

0.32V

Sin

gle

phot

on s

igna

l (m

V)

PMT High Voltage (kV)

For 1.5kV High voltage

20keV energy deposition 0.32±0.03V

1mV 6.2±0.6MeV

(≈ the calibration using radioactive sources)

Calibration is done at ESRF using a 20keV X-ray beam

BLM detectors : BLM detectors : CalibrationCalibration

0.4 0.6 0.8 1.0

1E-12

1E-11

1E-10

1E-9

1E-8

1E-7

5.8pA

14pA40pA

Calibration using a 330MBq 137Cs Source

I (A

)

High voltage (kV)

Calibration using a very intense Cesium source

( - emitter: 53pA)

High Voltage (V) 500 400 300

ACEM current (pA) 40 14 5.8

Efficiency (%) 75 26 11

Output voltage on 50Ω (nV)

2 0.7 0.3

Calibration : 1mV xx A

26.5 75.7 177

ACEMScintillator + PMT

T. Lefevre BIW 2004, 5 May 2004

Lattice design in a linac sectionLattice design in a linac section

Example in the central Quad:

≈ 10m ≈ 100 .mm.mrad≈ 80 (40MeV)

≈ 6mm

)().(

)(mmmradmm

mm

Beam size

Beam emittance Relativistic factorBPM790

BPM890

Quadrupoles

Steerer

Beam lossdetectors

T. Lefevre BIW 2004, 5 May 2004

Beam optics reconstructed from experimental dataBeam optics reconstructed from experimental data

T. Lefevre

BLM’s

• High probability to have beam losses in the quadrupoles region

BIW 2004, 5 May 2004

35MeV, 0mm beam size

T. Lefevre

done by Matthew Wood

• Beam loss position more than 2 orders of magnitude difference in the shower efficiency

• Can the longitudinal beam loss position be determined by the ‘angular shower shape’ ?

80MeV, 0mm beam size

Positions of the beam lossShower asymmetry as a function of the beam loss position

Geant3 SimulationsGeant3 Simulations

BIW 2004, 5 May 2004

T. Lefevre

done by Matthew Wood

Shower versus beam angle

• During this test we were using very small steering forces (I<1.5A) so that 5mrad can be considered as a maximum deviation angle

• Beam loss angle effects are small compared to the effect due to the beam loss position

35MeV

Positions of the beam loss

Geant3 SimulationsGeant3 Simulations

BIW 2004, 5 May 2004

T. Lefevre

done by Matthew Wood

~ e-/ e+ shower of 0.3nA, not seen by the ACEM with a 400volts bias

Shower generated by the beam losses in the collimator

Geant3 SimulationsGeant3 Simulations

BIW 2004, 5 May 2004

T. Lefevre

Example 1Example 1 : Observation of the beam transient loss : Observation of the beam transient loss

• Time – Energy correlation in the beam transient (Each slit aperture selected a given beam energy range between 35-70MeV)

• You normalize the BLM signals to the beam current loss seen by the BPM502 (≈ BPM690)

16 18 20 22 24 26 28

0.05

0.10

0.15

0.20

0.25

BLM

/BPM

cur

rent

(%)

Slit aperture (mm)

‘BLM signals depend on beam energy, position and current’

Possibility to estimate where the beam transient is lost

• The different energies are not lost at the same position

• Beam loss distributed between the detector and the 3rd quadrupole

Low - - - - - - - - - high energy

BIW 2004, 5 May 2004