Transcript SJ_AWAKE_UK_Spectrometer_21-03-14

AWAKE Electron
Spectrometer
Simon Jolly
21st March 2014
Spectrometer Specifications
• Wakefield accelerated electrons ejected collinear with
proton beam: need to separate the 2 and measure
energy of electron beam only.
• Must be able to resolve energy spread as well as energy:
spectrometer must accept a range of energies, probably
0-5 GeV.
• Current conceptual layout:
– Dipole mounted ~2 m downstream of plasma exit induces
dispersion in electron beam.
– Scintillator screen 1 m downstream of dipole intercepts
electron beam ONLY.
– Dispersion gives energy-dependent position spread on
screen.
– Scintillator imaged by intensified CCD camera viewing
upstream face of scintillator screen.
21/03/14
Simon Jolly, UCL, AWAKE-UK Meeting
2
2 GeV Beam, 1.86 T Field
21/03/14
Simon Jolly, UCL, AWAKE-UK Meeting
3
Spectrometer Status
• Previously, UCL focus has been on “optical” parts of system:
scintillator, camera etc.
• CERN much more concerned about experimental integration:
– What infrastructure is required for spectrometer?
– How much space do we need to “land grab” to fit in complete
system?
– How do we get electrons out of vacuum?
• Scintillator light output:
– Scintillator performance crucial for this design.
– After hiatus, giving it both barrels:
• Lawrence Deacon (postdoc).
• James Goodhand (MSci).
– Geant4/BDSIM simulations: start with better-known
scintillators…
• Full system simulations:
– Beam-gas interactions (vacuum vessel constraints).
– Efficiency of optical system.
– Energy reconstruction for various input distributions.
21/03/14
Simon Jolly, UCL, AWAKE-UK Meeting
4
Experimental Issues
• CERN MBPS dipole selected:
– It’s free (cheap is good).
– It’s at CERN (local is good).
– Field uniformity only good in 300 mm wide central region (narrow is
bad).
– Magnet quality may not be good enough (who foots bill for new
magnet(s)…?).
• Need to make sure we have enough space for spectrometer
downstream of plasma cell:
– Camera needs to be well-shielded from backgrounds.
– Far away better for backgrounds, close better for photon collection.
– Needs light tight path: run light path along floor with pair of mirrors?
• Lawrence’s Geant4 results show proton beam MUST stay in vacuum:
– Need transition between protons+electrons in vacuum and scintillator
outside vacuum.
– Vacuum window to let electrons pass out of vacuum with minimal
absorption/scattering.
– Either use beampipe with side window or larger vacuum vessel with
“parallel” exit window.
21/03/14
Simon Jolly, UCL, AWAKE-UK Meeting
5
MBPS Magnet
21/03/14
Simon Jolly, UCL, AWAKE-UK Meeting
6
MBPS Magnet: Good Field Region?
300 mm
21/03/14
Simon Jolly, UCL, AWAKE-UK Meeting
1000 mm
7
AWAKE Spectrometer Area (1)
21/03/14
Simon Jolly, UCL, AWAKE-UK Meeting
8
AWAKE Spectrometer Area (2)
21/03/14
Simon Jolly, UCL, AWAKE-UK Meeting
9
Reverse Geometry
21/03/14
Simon Jolly, UCL, AWAKE-UK Meeting
10
Reverse Geometry: Camera Position
21/03/14
Simon Jolly, UCL, AWAKE-UK Meeting
11
Reverse Geometry: Vacuum Vessel
21/03/14
Simon Jolly, UCL, AWAKE-UK Meeting
12
Light Tight Vessel
21/03/14
Simon Jolly, UCL, AWAKE-UK Meeting
13
Off-Axis Camera + Focussing Mirror
21/03/14
Simon Jolly, UCL, AWAKE-UK Meeting
14
Geant4 Simulations (Lawrence)
• Have set up simulation
using GEANT4/BDSIM
quadrupole
• Includes upstream
doublet, dipole field
Coils (yellow)
map (MBPS
measurements),
Vacuum
scattering, all optical
Chamber
processes from
(grey)
Yoke
screen, capture
Camera
(green)
plane
efficiency of camera
Scint. screen
Poles (blue)
system.
• Camera can measure
single photons, but
need to optimise
21/03/14
Simon Jolly, UCL, AWAKE-UK Meeting
15
Scintillator Screen
• Default scintillator choice is Lanex:
•
•
•
•
•
Manufactured by Kodak.
Used in Medical Physics as X-ray phosphor for imaging.
Gd2O2S:Tb – Gadolinium sensitiser, Terbium dopant
activator/wavelength shifter.
Phosphor grains on reflective backing.
Properties don’t seem to be well documented/studied…
Aerial & Industrial Mark
• Need to simulate light production
(photons
per MeV conversion
Screen
construction
efficiency) to ensure we have enough photons emitted in
direction of camera.
• Procuring GadOx samples from
Applied Scintillation
Technologies:
•
•
Variants of Medex screens (AST
trade name for GadOx medical xray phosphor screens).
More phosphor = higher light
output = lower resolution.
Curlcontrol
Backing
Polyester
support
178 microns
Phosphor
Layer
50 to 305 microns
• Need to compare results from
Geant4 simulations with
Clear
overcoat
experimental tests using known
source.
21/03/14
Simon Jolly, UCL, AWAKE-UK Meeting
©
7.6 microns
16
GadOx: Geant4 Photons
(J. Goodhand)
500
0
•
•
500
500
500
0
0
500
21/03/14
Photons mostly emitted
normal to screen.
Optimum camera position
normal to screen: 0.7 m
for 1 photon per electron
(all electrons
reconstructed).
500
Simon Jolly, UCL, AWAKE-UK Meeting
17
Reconstruction (Alexey Side
Injection)
Spectrum – measured (red), actual
ctrum – measured (red), actual
ogram) – scaled to match the(histogram)
peaks
21/03/14
Simon Jolly, UCL, AWAKE-UK Meeting
18
Conclusions
• Simulation results for scintillator light output reasonably
promising:
– Can collect light from all electron hits with camera at 0.7 m.
– Can still reconstruct energy spectrum with camera further
away without focussing lens.
• Need to refine vacuum chamber dimensions: where is
transition between vacuum and air for electrons?
• Magnet still a concern:
– Is MBPS going to give us large enough field region to
measure sufficient energy range?
– Would we be better designing our own magnet (£100k per
magnet at least…)?
– Would 2 magnet “chicane” spectrometer give better
performance? Would certainly reduce energy
acceptance…
• Difficult to give precision measurement of 0.01-5 GeV
beam…
21/03/14
Simon Jolly, UCL, AWAKE-UK Meeting
19