Ionization Cooling for a ν-Factory or Collider David Neuffer Fermilab 7/15/06
2 Outline Cooling Requirements ν-Factory Collider Ionization Cooling Cooling description Heating – Longitudinal Cooling Emittance Exchange- Partition Number Helical wiggler-PIC-REMEX Low-Energy “Cooling” Emittance exchange Li lens Solenoid Cooling Scenarios
3 References A. N. Skrinsky and V.V. Parkhomchuk, Sov. J. Nucl. Physics 12, 3(1981). D. Neuffer, Particle Accelerators 14, 75 (1983) D. Neuffer, “ + - Colliders”, CERN report (1999). D. Neuffer, “Introduction to Muon Cooling, NIM A532,p. 26 (2004). C. X. Wang and K. J. Kim, “Linear Theory of 6-D Ionization Cooling”, (PRL) MuCOOL Note 240, April (also COOL03), NIM A532, p. 260 (2004) Y. Derbenev and R. Johnson, Phys. Rev. ST Accel. Beams 8, E (2005); COOL05 proc. Simulation tools R. Fernow, ICOOL T. Roberts, G4BeamLine (Muons, Inc.)
Collider Overview
5 Collider Parameters
6 Overview of -Factory Proton Driver (1-4 MW) – proton bunches on target produce s Front-end: decay + collect and cool s: (phase rotation + ionization cooling) Accelerator - to full energy ( linac + RLAs to 20—50 GeV) - Storage ring Store ’s until decay (~300 B turns) e + e * decays produce neutrino beams directed toward: Long base line neutrino detector (2000—8000 km away …) to ( e, *) /SS/year
7 Producing and Capturing Collaboration baseline: 10GeV p-beam on Target (Hg-jet) immersed in 20 1.75 T solenoid, taking ~300 MeV/c μ -Collider: Rf: ~200 MHz, Capture string of ~20 bunches- Recombine after cooling ν -Factory: Rf: ~200 MHz, 12 MV/m Capture in string Of ~50 bunches
8 Cooling Requirements Beam from target has ,rms 2 × m-rad; ║,rms 1m -Storage Ring -Factory Goal is to collect maximum number of + and/or - that fit within accelerator / storage ring acceptances Transverse cooling by ~10 is sufficient ,rms 0.2 to 0.8×10 -2 m-rad; ║,rms 0.06 m-rad/bunch Collider Goal is maximal cooling of maximum number of both + AND - ; high luminosity needed. Cooling by > ~100 in each of x, y, z is required ,rms 0.5 to 0.025×10 -4 m-rad; ║,rms 0.04 m-rad
9 Cooling Summary
10 Transverse cooling: Particle loses momentum P( and ) in material Particle regains P (only) in RF Multiple Scattering in material increases rms emittance: Muon Cooling-general principle
11 Ionization Cooling Principle Loss of transverse momentum in absorber: Heating by multiple scattering
12 Combining Cooling and Heating: Low-Z absorbers (H 2, Li, Be, …) to reduce multiple scattering High Gradient RF To cool before -decay (2.2 s) To keep beam bunched Strong-Focusing at absorbers To keep multiple scattering less than beam divergence … Quad focusing ? Li lens focusing ? Solenoid focusing?
13 Transverse cooling limits Transverse Cooling – equilibrium emittance equilibrium scattering angle Want materials with small multiple scattering (large L R ), but relatively large dE/ds, density ( ) Want small at absorbers => strong focusing - equilibrium emittances (/ ) smallest for low-Z materials
14 Ionization Cooling problems Must focus to very small β β : 1m → ~1mm Intrinsic scattering of beam is large θ rms > ~0.1 radians Intrinsic momentum spread is large σ P /P > ~0.03 Cooling must occur within muon lifetime = 2.2γ s or L = 660 βγ m pathlength Does not (directly) cool longitudinally
15 Longitudinal “Cooling” Energy cooling occurs if the derivative : (dE/ds)/ E= g L (dp/ds)/p > 0 g L (E) is negative for E < ~ 0.2 GeV and only weakly positive for E > ~ 0.2 GeV Ionization cooling does not effectively cool longitudinally Energy straggling increases energy spread
16 “Emittance exchange” enables longitudinal cooling: Cooling derivative is changed by use of dispersion + wedge (Dependence of energy loss on energy can be increased) (if due to path length)
17 Partition Numbers, δE-δt cooling With emittance exchange the longitudinal partition number g L changes: But the transverse cooling partition number decreases: The sum of the cooling partition numbers (at P = P ) remains constant: Σ g > 0
18 Cooling + Energy straggling... Energy spread E cooling equation: Longitudinal Emittance Cooling equation : Longitudinal Cooling requires: Positive g L (η, “wedge”), Strong bunching (β cτ small) Large V rf, small λ rf Energy loss/recovery Before decay requires: Equilibrium σ p :
19 μ Cooling Regimes Efficient cooling requires: Frictional Cooling (<1MeV/c) Σ g =~3 Ionization Cooling (~0.3GeV/c) Σ g =~2 Radiative Cooling (>1TeV/c) Σ g =~4 Low-ε t cooling Σ g =~2β 2 (longitudinal heating)
20 Focusing for Cooling Strong focussing needed – magnetic quads, solenoids, Li lens ? Solenoids have been used in most (recent) studies Focus horizontally and vertically Focus both + and - Strong focussing possible: β = 0.13m for B=10T, p = 200 MeV/c β = m for B=50T, p = 20 MeV/c But: Solenoid introduces angular motion L damped by cooling + field flips No chromatic correction (yet) B within rf cavities ? ( )
21 Solenoidal Focusing and Angular Momentum Angular motion with focusing complicates cooling Energy loss in absorbers reduces P , including P Orbits cool to Larmor centers, not r = 0 Solution: Flip magnetic fields; new Larmor center is near r=0
22 More complete coupled cooling equations: θ D, θ W are dispersion, wedge angles Scattering terms Wang and Kim, (MuCOOL 240) have developed coupled cooling equations including dispersion, wedges, solenoids, and symmetric focussing (β x = β y = β T )
23 Cooling with exchange and solenoids (Wang and Kim) Example: rms Cooling equations with dispersion and wedges (at = = ) in x-plane: The additional correlation and heating terms are “small” in “well-designed” systems.
24 Study 2 Cooling Channel (for MICE) Cell contains Rf for acceleration/bunching H 2 absorbers Solenoidal magnets sFOFO 2.75m cells 108 m cooling channel consists of: m cells m cells Focusing increases along channel: B max increases from 3 T to 5.5 T Simulation Results
25 Study2 cooling channel Focusing function at absorbers: 0.5m→0.2m Total length of channel 100+m Cools to ~0.002m
26 Study 2A cooling channel Lattice is weak-focusing B max = 2.5T, solenoidal β ≅ 0.8m Cools transversely from ~0.018 to ~0.007m in ~70m Before After cooling -0.4m+0.4m
27 RFOFO ring cooler performance Example cools longitudinally more than transversely Can be adjusted for more transverse cooling E-ct before and after Transverse before and after
28 RF Problem: cavity gradient in magnetic field is limited? Vacuum Cavities 800 MHz results: 40MV/m →13MV/m Muons, Inc. results: 50+ MV/m no change with B Rf breakdown field decreases in magnetic fields? Solenoidal focussing gives large B at cavities But gas in cavity suppresses breakdown 10% of liquid H 2
29 Helical 6-D Cooler (Derbenev) Magnetic field is solenoid B 0 + dipole + quad + … System is filled with H 2 gas, includes rf cavities Cools 6-D (large E means longer path length) Key parameters: a, k=2π/λ, solenoid field B 0, transverse field b(a)
30 Comments on Helical Wiggler parameters 1/ T 2 ≅ 0.67 for equal cooling at ∑ g =2 Energy loss at liquid H 2 density is ~30MV/m (800atm-e gas) Typical simulations have used ~ 15MV/m energy loss Need more rf gradient: ~22MV/m (could use less if needed…) PP 200 MeV/c (0.67T-m) λ1.0m a0.16m =ka= P /P z 1 B5.5T (Bρ/B=0.12m) b(a)1.28T (Bρ/b=0.52m) b ' (a)-0.46T/m D(a)- dispersion 0.29m Δ g =1/ T Typical case
31 Helical Wiggler 3-D Cooling (P µ =250MeV/c) l=1.0 l=0.8 l=0.6 l=0.4 Cooling factor ~ 50,000 × Yonehara, et al.
32 Helical wiggler R&D Need Magnet design Solenoid, dipole + quad Displaced solenoid coils can provide needed field Matching in/out
33 + - Collider Cooling Scenarios + - Collider requires energy cooling and emittance exchange (and anti-exchange) to obtain small L, ε x, ε y emittances required for high-luminosity Start with large beam from target, compress and cool, going to stronger focussing and bunching as the beam gets smaller …
34 Updated Scenario (Palmer ) PIC? Low Emttance Muon Collider REMEX? “Guggenheim” 6D cooler 800 MHz 6D cooler
35 Palmer scenario to do: Matching from section to section Buncher/Wiggler – 3D dynamics Reoptimization Phase rotation buncher is ν-Factory case (not Collider optimum) Tapered Guggenheims Final cooler Matching in/out Rf match in/out “realistic” field models Reacceleration scenario from low-emittance Lower frequency rf + buncher Can PIC/REMEX or … get us to smaller emittance?
PIC-Parametric-resonance Ionization Cooling (JLab, Y. Derbenev) (also Balbekov, 1997) Excite ½ integer parametric resonance (in Linac or ring) Similar to vertical rigid pendulum or ½-integer extraction Elliptical phase space motion becomes hyperbolic Use xx’=const to reduce x, increase x' Use Ionization Cooling to reduce x' Detuning issues being addressed (chromatic and spherical aberrations, space-charge tune spread). Simulations underway. First: Then: IC reduces x’
37 PIC/REMEX cooling (Derbenev) PIC ,eff : 0.6cm 0.1cm Transverse + longitudinal cooling Reverse emittance exchange to reduce transverse emittance (“REMEX”) Chromaticity correction a problem Depth of focus a problem L absorber < ~ β No “realistic” simulations
38 PIC/REMEX examples (Bogacz, Beard, Newsham, Derbenev) Example: Solenoids + quads + dipoles+rf 2m cells β = 1.4cm, η x = 0.0m Problems: Large σ p /p (~3%) Large σ θ (>0.1) Short absorber –1cm Be = 3MeV Solution approach: Use simulations to tune this as a resonant beam line
39 Cooling scenario (~Muons, Inc.)
40 Low-Energy “Cooling”=REMEX without wedges At P μ = 10 to 200 MeV/c, energy loss heats the beam longitudinally Transverse cooling can occur emittance exchange Equilibrium transverse emittance decreases dE/ds scales as 1/β 2 β t scales as β –Solenoid β t p/B ε N,rms P μ 2 ??? Decrease ε N,transverse while ε long increases “wedgeless” emittance exchange ε N,rms × 1/30, ε long ×300 ???
41 Low-Energy “cooling”-emittance exchange dP μ /ds varies as ~1/β 3 “Cooling” distance becomes very short: for H at P μ = 10MeV/c Focusing can get quite strong: Solenoid: β =0.002m at 30T, 10MeV/c ε N,eq = 1.5×10 -4 cm at 10MeV/c Small enough for “low-emittance” collider PµPµ L cool 200 MeV/c cm 100 cm
42 Emittance exchange: solenoid focusing Solenoid focusing(30T) 0.002m Momentum (30→10 MeV/c) L ≈ 5cm R < 1cm Liquid Hydrogen (or gas) PµPµ L cool 200 MeV/c cm Use gas H 2 if cooling length too short ε N,eq = 1.5×10 -4 cm at 10MeV/c -Will need rf to change p to z
43 Li-lens cooling Lithium Lens provides strong- focusing and low-Z absorber in same device Liquid Li-lens may be needed for highest-field, high rep. rate lens BINP (Silvestrov) was testing prototype liquid Li lens for FNAL But FNAL support was stopped - and prototypes were not successful... β =0.026m (200 MeV/c, 1000 T/m) β =0.004m (40 MeV/c, 8000 T/m)
44 Summary Cooling for neutrino factory is practical Collider cooling scenario needs considerable development: Longitudinal cooling by large factors … Transverse cooling by very large factors Final beam compression with reverse emittance exchange Reacceleration and bunching from low energy
45 Linac-area MuCool Test Area Test area for bench test and beam-test of Liquid H 2 absorbers Enclosure complete in ~October 2003 Can test 200 and 805 MHz rf for MuCOOL and also for Fermilab Assemble and beam test cooling modules (absorber + rf cavity + solenoid)
46 MTA experimental program Rf: 805, 201 MHz, gas-filled 201MHz just reached 16 MV/m 805 MHz 3T, gas-cavity test H 2 absorbers
47 MICE beam line (Drumm, ISS) MICE (International Muon Ionization Cooling Experiment) To verify ionization cooling (for a neutrino factory) with a test of a complete cooling module in a muon beam Muon beam line and test area in RAL-ISIS (Oxford) Installation Jan. – Oct Experiment occurs in ~ time frame MICE beam line and experimental area (RAL) TTargetSheffield APion CaptureISIS BDecay SolenoidEID CMuon Transport Channel Liverpool ? DDiffuserOxford
48 Incoming muon beam Diffusers 1&2 Beam PID TOF 0 Cherenkov TOF 1 Trackers 1 & 2 measurement of emittance in and out Liquid Hydrogen absorbers 1,2,3 Downstream particle ID: TOF 2 Cherenkov Calorimeter RF cavities 1RF cavities 2 Spectrometer solenoid 1 Matching coils 1&2 Focus coils 1 Spectrometer solenoid 2 Coupling Coils 1&2 Focus coils 2Focus coils 3 Matching coils 1&2 10% cooling of 200 MeV muons requires ~ 20 MV of RF single particle measurements => out / in ) = 10 -3
49 MICE Experiment
50 Last Slide
51 Postdoc availability – Front end Optimization