Stephen Brooks / RAL / March
20051 of 13
Muon Front Ends
Providing High-Intensity, Low-Emittance Muon Beams for the Neutrino Factory
and Muon Collider
Stephen Brooks / RAL / March
20052 of 13
Contents
• Future Accelerator Projects Requiring Muon Front Ends– Neutrino Factory– Muon Collider
• Choice of Particle – why Muons?• Front End Components and Options• Context: National R&D Programme
– UK Neutrino Factory (UKNF) group
Stephen Brooks / RAL / March
20053 of 13
The Neutrino Factory
• Goal: To fire a focussed beam of neutrinos through the interior of the Earth– What’s the point?
• Constrains post-Standard Model physics– But why does this involve muons?
• Neutrinos appear only as decay products• Decaying an intense, high-speed beam of
muons produces collimated neutrinos
Stephen Brooks / RAL / March
20054 of 13
The Neutrino Factory
• p+ + + e+e
• Uses 4-5MW proton driver– Could be based on ISIS
Stephen Brooks / RAL / March
20055 of 13
The Neutrino Factory
• “Front end” is the muon capture system
Stephen Brooks / RAL / March
20056 of 13
The Muon Collider
• Goal: to push the energy frontier in the lepton sector after the linear collider
• p+ +,− +,−
+
-
3+3TeV MuonCollider Ring
Stephen Brooks / RAL / March
20057 of 13
Why Collide Muons?Particle Proton Electron MuonMass 938 MeV 0.511 MeV 106 MeVSynchrotron radiation limit (LEP-II RF)
28.5 TeV 0.102 TeV 5.55 TeV
Machine issues
B-field limit at 7 TeV (LHC)
Linear 1 TeV collider more cost-effective Half-life of
2.2 sPhysics problems
Messy collisions None
Stephen Brooks / RAL / March
20058 of 13
Design Challenges
• Must accelerate muons quickly, before they decay– Conventional synchrotrons cycle too slow– Once is high, you have a little more time
• High emittance of pions from the target– Use an accelerator with a really big aperture?– Or try beam cooling (emittance reduction)– In reality, do some of both
Stephen Brooks / RAL / March
20059 of 13
Muon Front End Components
• Targetry, produces pions (±)• Pion to muon decay channel
– Uses a series of wide-bore solenoids• “Phase rotation” systems
– Outside scope of this talk• Muon ionisation cooling (as in “MICE”)
– Expensive components, re-use in cooling ring• Muon acceleration (RLAs vs. FFAGs)
Stephen Brooks / RAL / March
200510 of 13
The Decay Channel
• Has to deal with the “beam” coming from the pion source
Evolution of pions from 2.2GeV proton beam on tantalum rod target
Stephen Brooks / RAL / March
200511 of 13
The Decay Channel
• Has to deal with the “beam” coming from the pion source
• Pion half-life is 18ns or 12m at 200MeV– So make the decay channel about 30m long
• Grahame Rees designed an initial version– Used S/C solenoids to get a large aperture
and high field (3T mostly, 20T around target)• Needed a better tracking code…
Stephen Brooks / RAL / March
200512 of 13
The Decay Channel (ctd.)
• Developed a more accurate code, Muon1• Used it to validate Grahame’s design…
– 3.1% of the pions/muons were captured• …and parameter search for the optimum
– Within constraints: <4T field, >0.5m drifts, etc. – Increased transmission to 9.6%
• Increased in the older code (PARMILA) too– Fixed a problem in the original design!
Stephen Brooks / RAL / March
200513 of 13
UKNF Research Efforts
• MICE at RAL (phase I~2007; II~2009-10)• FFAG electron model at Daresbury
– Under definition!• Target shock studies program• Beamline design and optimisation work
– Myself, Grahame (+ new recruit soon)– Network with European “BENE” collaboration
• http://hepunx.rl.ac.uk/uknf/
Stephen Brooks / RAL / March
200514 of 13
BACKUP!
In case the time is longer than my slides.
Web report
Stephen Brooks / RAL / March
200515 of 13
Muon Acceleration Options
• Accelerators must have a large aperture• Few turns (or linear) in low energy part, so
muons don’t decay• Recirculating Linacs (RLAs, studied first)• FFAGs (cyclotron-like devices)
– Grahame is playing with isochronous ones
Stephen Brooks / RAL / March
200516 of 13
NuFact Intensity Goals
• “Success” is 1021 /yr in the storage ring
Proton Energy/GeV Intensity/MW Target eff (pi/p) MuEnd eff (mu/pi) Operational mu/year in storage ring Current/uA
8 4 20% 1.0% 30% 5.90497E+19 500 "Not great" scenario
8 1 60% 2.0% 35% 1.03337E+20 125 ISIS MW only to reach 10^20
8 5 60% 3.5% 40% 1.03337E+21 625 "Quite good" 5MW scenario (gets 10^21)
8 5 1.75 8.5% 55% 1.00646E+22 625 Required to reach 10^22
1.75 = PtO2 target inclined at 200mrad, see Mokhov FNAL PiTargets paper 20% = 2.2GeV dataset from Paul Drumm
Stephen Brooks / RAL / March
200517 of 13
Tracking & Optimisation System
• Distributed Computing– ~450GHz of processing power– Can test millions of designs
• Genetic Algorithms– Optimisation good up to 137 parameters…
• Accelerator design-range specification language– Includes “C” interpreter
Stephen Brooks / RAL / March
200518 of 13
Decay Channel Lattice
Drifts Length (m)
D1 0.5718 [0.5,1]
D2+ 0.5 [0.5,1]
Solenoids Field (T) Radius (m) Length (m)
S120[0,20]
0.1 [fixed] 0.4066 [0.2,0.45]
S2-4−3.3, 4, −3.3[-5,5] 0.3
[0.1,0.4]0.4[0.2,0.6]
S5-S24±3.3 (alternating)[-4,4]
S25+ 0.15 [0.1,0.4]
Final (S34) 0.15 [fixed]• 12 parameters– Solenoids alternated in field strength
and narrowed according to a pattern• 137 parameters
– Varied everything individually
Tantalum RodLength (m) 0.2 [fixed]
Radius (m) 0.01 [fixed]
Angle (radians) 0.1 [0,0.5]Z displacement (m) from S1 start
0.2033 (S1 centred) [0,0.45]
Original parameters / Optimisation ranges
Stephen Brooks / RAL / March
200519 of 13
Improved Transmission• Decay channel:
– Original design: 3.1% + out per + from rod– 12-parameter optimisation 6.5% +/+
• 1.88% through chicane– 137 parameters 9.6% +/+
• 2.24% through chicane
• Re-optimised for chicane transmission:– Original design got 1.13%– 12 parameters 1.93%– 137 parameters 2.41%
3`700`000 runs so far
1`900`000 runs
330`000 runs
Stephen Brooks / RAL / March
200520 of 13
Optimised Design for the Decay Channel (137 parameters)
0
5
10
15
20
25
Fiel
d (T
esla
)
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Size
(met
res)
Solenoid Field Solenoid Radius Solenoid Length Drift Length
•Maximum Length
•Minimum Drift
•Maximum Aperture
•Maximum Field
(not before S6)
(mostly)
(except near ends)
(except S4, S6)
Stephen Brooks / RAL / March
200521 of 13
Why did it make all the solenoid fields have the same sign?
• Original design had alternating (FODO) solenoids• Optimiser independently chose a FOFO lattice• Has to do with the stability of off-energy particles
FODO lattice
FOFO lattice
Stephen Brooks / RAL / March
200522 of 13
Design Optimised for Transmission Through Chicane
• Nontrivial optimum found
• Preferred length?
• Narrowing can only be due to nonlinear end-fields
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Length
Radius
0.463 m
0.402−0.003n m