NuFACT06 Summer School-Factory Front End

and Cooling

David Neuffer

fFermilab

2

0utline Lecture I –Front End

General introduction– Study 2A ν-Factory– variations

Capture and decay from target Bunching and φ-E rotation

Lecture II – Cooling Ionization cooling concepts Cooling for a ν-Factory

– MICE experiment µ+-µ- collider cooling

3

References “Cost-effective design for a neutrino factory”, J. Berg et

al., PRSTAB 9, 011001(2006) “Recent progress in neutrino factory …”, M. Alsharo’a et

al., PRSTAB 6, 081001 (2003) “Beams for European Neutrino Experiments (BENE)

CERN-2006-005 S. Ozaki et al., Feasibility Study 2, BNL-52623(2001). N. Holtkamp and D. Finley, eds., Study 1, Fermilab-Pub-

00/108-E (2000). The Study of a European neutrino factory complex,

CERN/PS/2002-080. R. Palmer- NuFACT05 Summer School lecture notes

4

Neutrino Factory - Study 2A Proton driver

Produces proton bunches 8 or 24 GeV, ~1015p/s, ~50Hz

bunches

Target and drift (> 0.2 /p)

Buncher, bunch rotation, cool

Accelerate to 20 GeV or more Linac, RLA and FFAGs

Store at 20 GeV (0.4ms) e ++ e

*

Long baseline Detector >1020 /year

5

Target for π production Typical beam: 10 GeV protons up to 4 MW

1m long bunches up to 4×1013/bunch, 60Hz

Options: Solid targets

– C (graphite targets) (NUMI) – Solid metal (p-source) – rotating Cu-Ni target

Liquid Metal targets– SNS – type (confined flow)– MERIT – Hg jet in free space

– Best for 4MW ??

6

The Fundamental Problem with Solid Targets

What do we need in materials to get us

to multi-MW Power Levels? low elasticity modulus

(limit Stress = EαΔT/(1-2ν) low thermal expansion high heat capacity good diffusivity to move heat away

from hot spots high strength resilience to shock/fracture strength resilience to irradiation damage That’s All !

7

Liquid Targets Contained liquid flow (~SNS)

Damage to containment vessel possible Shock of short pulse

Liquid Jet target Hg jet ~ Jet is disrupted by beam

– δT = 50 s ? Need target material capture and recirculation

system

8

MERIT experiment at CERN

Target date: November 2006! i.e. ready to receive and install the solenoid and Hg-loop

Beam parameters:Beam parameters: Nominal momentum 24 GeV/c Intensity/bunch – baseline: harmonic 16 (i.e. 16 buckets in PS, t=125ns)

2-2.5 1012 protons / bunch total maximum ~30 1012 protons/pulse

Next steps:Next steps: MD time in 2006 assigned

To address the most critical configurations – priorities should be defined Set-up time at the beginning of 2007 may be required to achieve the

highest intensities

Magnet tested at 15THg jets

9

π capture from target Protons on target produce large

number of π’s Broad energy range (0 to 10+GeV) More at lower energies Transverse momentum

(up to ~0.3GeV/c)

Capture beam from target Options:

Li lens

Magnetic horn

Magnetic Solenoid

10

Li Lens properties Current-carrying conducting cylinder Focusing Field:

Fermilab values: R0=1cm, I=0.5MA, L=15cm, B(R0)=10T

Focuses 9GeV/c p with p < 0.45 GeV/c

Problems Pulsed at <1Hz, need liquid for 10+ Hz Absorbs particles (π,p-bar) Forward capture Captures only one sign

θ 0 total 20

rB (r) = μ I

2πR

Focusing angle:Θ=(0.3B(r) L)/P

11

Magnetic Horn after target Baseline capture for superbeams/NUMI Magnetic field from I on wall

Lenses can be tuned to obtain narrow band or broad-band acceptance

Pulsed current, thin conductors Breakage over many pulses Beam lost on material

focuses + or - particles

inside conductorθB (r) = 0

total

θ o

IB (r) = μ

2 r /

θ pathfocus

B LΔθ =

3.33P

NUMI beam line

12

NUMI target, beam

Target segmented graphite, water cooled 954mm long; 47 20mm segments Movable: can be positioned up to

2.5m upstream of horn to tune beam energy

Parabolic horns Pulsed at up to 200kA; 3T peak field Focus π‘s into decay tunnel

13

Solenoid lens capture Target is immersed in high field solenoid Particles are trapped in Larmor orbits

Produced with p = p‖, p

Spiral with radius r = p/(0.3 Bsol)

Particles with p < 0.3 BsolRsol/2 are trapped

p,max < 0.225 GeV/c for B=20T, Rsol = 0.075m

Focuses both + and - particles

14

Solenoid transport Magnetic field adiabatically

decreases along the transport Transverse momentum

decreases Busch’s theorem: B rorbit

2 is constant B=20T→2T (r=3.75cm →= 12cm) P = 0.225→0.07GeV/c

Emittance = ~σx σpx/105.66

~8cm×25 MeV/c/105.66≅0.02m

P remains constant (P‖ increases)

Transport designed to maximize π→ acceptance

15

Homework problems: targetry How many 24 GeV protons per second are in a 4 MW

beam? With 60Hz bunches, how many protons/bunch? a beam of 24GeV protons produces, on average, one

pion pair per proton with mean momentum of 250 MeV/c (per pion). What percentage of the proton beam kinetic energy is converted to pions?

If the target is surrounded by a 20T solenoid with a 5cm radius, what maximum transverse momentum of pions is accepted?

If B is adiabatically reduced to 1T what is the resulting transverse momentum and beam size?

Estimated normalized emittance?

16

π→μν decay in transport π-lifetime is 2.60×10-8 s

L = 7.8 β m For π μ+ν,

<PT,rms> is 23.4 MeV/c, E=0.6 to 1.0Eπ

Capture relatively low-energy π μ 100 – 300 MeV/c

Beam is initially short in length Bunch on target is 1 to 3 ns rms

length

As Beam drifts down beam transport, energy-position (time) correlation develops:

L=0m

L=36m

0

0.4 GeV

-20m 100m

arrival

Lc

17

Phase-energy rotation To maximize number of

~monoenergetic μ’s, neutrino factory designs use phase-energy rotation

Requires: “short” initial p-bunch Drift space Acceleration (induction linac or rf)

– at least ±100 MV Goal:

Accelerate “low-energy tail” Decelerate “high-energy head: Obtain long bunch

– with smaller energy spread

L=1m

L=112m

rf

0

0.5GeV/c

-50m 50mLL

δL = δβ(p )

18

Phase Energy rotation options Single bunch capture

Low-frequency rf (~30MHz) Best for collider (?) (but only + or -)

Induction Linac Nondistortion capture possible Very expensive technology, low gradient Captures only + or -

“High Frequency” buncher and phase rotation Captures into string of bunches (~200MHz) Captures both + and -

19

Phase/energy rotation Low-frequency rf;

capture into single long bunch 25MHz – 3MV/m +25% 50MHz 10m from target to 50m

But: Low-frequency rf is very

expensive Continuation into cooling

and acceleration a problem (200MHz?)

12 m

Only captures one sign …

20

Induction Linac for φ-E Rotation Induction Linac can provide

long pulse for φ-E rotation Arbitrary voltage waveform

possible

Limited to < ~1MV/m need > ~200MV, > 200m

Very expensive, large power requirements

Only captures one sign

21

Nondistortion φ-E rotation Cancellation with single

induction linac gives distortion (Head has larger δE than tail)

Sequence of 2 (or more) linacs can spread beam out evenly

Goal is to spread beam out evenly mapping Kinetic Energy to length (Δcτ); all at same final energy (0.25 to 0.225MeV) to (0 to 80m)

2 11

0 1final

L LLz

22

Study 2 system Drift to develop Energy- phase correlation

Accelerate tail; decelerate head of beam, non-distortion (280m induction linacs (!))

Bunch at 200 MHz ~0.2 /p

Inject into 200 MHz cooling system Cools transversely (to t= ~0.002m

23

High-frequency Buncher + Rotation

Drift (110m) Allows beam to decay;

beam develops correlation

Buncher (~333230MHz) Bunching rf with E0 = 125 MeV,

1 = 0.01 { L 1 =~1.5m at Ltot

= 150m}

Vrf increases gradually from 0 to ~6 MV/m

Rotation (~233200MHz) Adiabatic rotation Vrf =~10 MV/m

Cooler(~100m long) (~200 MHz) fixed frequency transverse cooling system

Replaces Induction Linacs with medium-frequency rf (~200MHz) !

Captures both μ+ and μ- !!

24

Adiabatic Buncher overview

Want rf phase to be zero for reference energies as beam travels down buncher

Spacing must be N rf

rf increases (rf frequency decreases)

Match to rf= ~1.5m at end:

Gradually increase rf gradient (linear or quadratic ramp): 2

D Drf 2

B B

z - z z - zE (z) = B + C

L L

Example: rf : 0.901.5m

25

Adiabatic Buncher example Adiabatic buncher (z=90→150m)

Set T0, : 125 MeV/c, 0.01

In buncher:

Match to rf=1.5m at end:

zero-phase with 1/ at integer intervals of :

Adiabatically increase rf gradient:

m5.1L11

L rf1

tot01

tot

2

2( ) 2 6 /D Drf

tot D tot D

z z z zE z MV m

L z L z

1

0n

n11

rf : 0.901.5m

)(z)z( 1rf

26

Rotation At end of buncher, change rf to

decelerate high-energy bunches, accelerate low energy bunches

Reference bunch at zero phase, set rf less than bunch spacing

(increase rf frequency)

Place low/high energy bunches at accelerating/decelerating phases

Can use fixed frequency (requires fast rotation) or

Change frequency along channel to maintain phasing “Vernier” rotation –A. Van Ginneken

27

Fast “Vernier” Rotation At end of buncher, choose:

Fixed-energy particle T0

Second reference bunch TN

Vernier offset Example:

T0 = 125 MeV Choose N= 10, =0.1

– T10 starts at 77.28 MeV Along rotator, keep

reference particles at (N + ) rf spacing 10 = 36° at =0.1 Bunch centroids change:

Use Erf = 10MV/m; LRt=8.74m High gradient not needed … Bunches rotate to ~equal

energies.

R10rf10R10 z)sin(Ee)0(T)z(T rf : 1.4851.517m in rotation;rf = ct/10 at end

(rf 1.532m)

Nonlinearities cancel:T(1/) ; Sin()

28

“ICOOL” Adiabatic Rotation

At end of buncher, choose: reference particle T0

Reference bunch TN (N bunches from 0)

V’ rf gradient, offset Example:

T0 = 125 MeV Choose N= 10, =0.1

– T10 starts at 77.28 MeV

In ICOOL T0 = T0

+ E0'z…

TN =TN + EN' z

Rotate until T0 ≅ TN

Along rotator, keep reference particles at (N + ) rf spacing EN' ≅eV' sin(2π)

λrf ~1.4 to 1.5 m over buncher

•~Adiabatic•Particles remain in bunches as bunch centroids align•Match into 201.25 MHz Cooling System

29

Initial Study 2A(12/03) Drift (110.7m) Buncher (51m)

P1=280, P2=154 MeV/c, NB=18

Vrf= 3 L/51 + 9 (L/51)2 MV/m

Vernier Phase Rotator (54m) NV = 18.05, Vrf=12 MV/m

Cooler (up to 100m) Alternating solenoid 2.7T, 0.75m cells 2cm LiH/cell 16MV/m rf (30°)

5000 particle simulation

30

ICOOL results-Study 2A (12/03)

0

500

1000

1500

2000

2500

3000

3500

0 50 100 150 200 250 300

accepted / 5000

Transverse emittance

0.00E+00

2.00E-03

4.00E-03

6.00E-03

8.00E-03

1.00E-02

1.20E-02

1.40E-02

1.60E-02

1.80E-02

2.00E-02

0 50 100 150 200 250 300 m

~0.23 μ/p within reference acceptance at end of 80m cooling channel (ε⊥<0.03m)

~0.11 μ/p within restricted acceptance (ε⊥<0.015m)

At end of φ-E Rotator:

A~0.10 μ/p and 0.05 μ/p

Rms emittance cooled from ε⊥= 0.0185 to ε⊥= ~0.008m

Longitudinal rms emittance ≅0.070m (per bunch)

31

Study2A June 2004 scenario Drift –110.7m Bunch -51m

V(1/) =0.0079 12 rf freq., 110MV 330 MHz 230MHz

-E Rotate – 54m – (416MV total) 15 rf freq. 230 202 MHz P1=280 , P2=154 NV = 18.032

Match and cool (80m) 0.75 m cells, 0.02m LiH

“Realistic” fields, components Captures both μ+ and μ-

32

Features/Flaws of Study 2A Front End

Fairly long section – ~300m long Study 2 was induction linac 1MV/m, 450m long

Produces long bunch trains of ~200 MHz bunches ~80m long (~50 bunches)

Transverse cooling is ~2½ in x and y No cooling or more cooling ?

Method works better than it should …

Requires rf within magnetic fields 12 MV/m at B = 1.75T

33

Another example: ~88 MHz Drift – 90m Buncher-60m

Rf gradient 0 to 4 MV/m Rf frequency: 166→100 MHz Total rf voltage 120MV

Rotator-60m Rf gradient 7 MV/m – 100→87 MHz 420MV total

Acceptance ~ study 2A (but no cooling yet) Less adiabatic

0 GeV/c

0.5 GeV/c

34

rf in Rotation/Cooling Channels: Can cavities hold rf gradient

in magnetic fields??

MUCOOL 800 MHz result : V' goes from 45MV/m to

12MV/m (as B -> 4T) Vacuum rf cavity

Worse at 200MHz ??

35

Use gas-filled rf cavities? Muons, Inc. tests:

Higher gas density permits higher gradient

Magnetic field does not decrease maximum allowable gradient

Gas filled cavities may be needed for cooling with focusing magnetic fields

Density > ~60 atm H2 (7.5% liq.)

Energy loss for µ’s is >~ 2MV/m

Can use energy loss for cooling

Mo electrode, B=3T, E=66 MV/mMo B=0 E=64MV/mCu E=52MV/mBe E= 50MV/m 800 MHz rf tests

36

Gas-filled rf cavites (Muons, Inc.)

Cool here

Add gas + higher gradient to obtain cooling within rotator

~300MeV energy loss in cooling region

Rotator is 54m; Need ~4.5MeV/m H2 Energy 133atm equivalent 295ºK gas ~250 MeV energy loss

Alternating Solenoid lattice in rotator

20MV/m rf cavities Gas-filled cavities may enable

higher gradient (Muons, Inc.)

37

High-gradient rf with gas-filled cavities

0

0.005

0.01

0.015

0.02

0.025

100 120 140 160 180 200 220

c

Transverse emittance

Acceptance (per 24GeV p)

Pressure at 150Atm Rf voltage at 24 MV/m

Transverse rms emittance cools 0.019 to ~0.008m

Acceptance ~0.22/p at εT < 0.03m

~0.12/p at εT < 0.015m About equal to Study 2A

0

0.1

0.2

0.3

0.4

0.5

0.6

160 170 180 190 200 210 220

n0

e < 0.015

e < 0.030

38

Simulation results

50m

-50m

0

0.5 GeV/c

39

Cost impact of Gas cavities Removes 80m cooling section (-185 M$)

Increase Vrf' from 12.5 to 20 or 24 MV/m Power supply cost V'2 (?) 44 M$ 107M$ or 155M$

Magnets: 2T 2.5T Alternating Solenoids 23 M$ 26.2 M$

Costs due to vacuum gas-filled cavities (??)

Total change: Cost decreases by 110 M$ to 62 M$ (???)

40

Cost estimates: Costs of a neutrino factory (MuCOOL-322, Palmer and Zisman):

Study 2 Study 2A

“Study 2A front end reduces cost by ~ 350MP$

41

Summary Buncher and E Rotator (ν-Factory) Variations

Gas-filled rf cavities can be used in Buncher-Rotator Gas cavities can have high gradient in large B (3T or more?)

Variations that meet Study 2A performance can be found Shorter systems – possibly much cheaper??

Gas-filled rf cavities

To do: Optimizations, Best Scenario, cost/performance … More realistic systems

42

Postdoc availability – Front end

SBIR with Muons, Inc. – capture, - rotation and cooling with gas-filled rf cavities