Fermilab, Proton Driver, PRISM/PRIME

David Neuffer

Fermilab

FFAG05

2

Intro–Fermilab Future Plans (~Oddone 9/05)

(Tevatron shuts down ~2009) First Priority – ILC

Global Study ~1 year –present to DOE (end of 2006)

IF favorable, push for near-term construction at Fermilab

Second Priority – protons Becomes high priority if ILC near-term not encouraged

PROTON DRIVER: 8 GeV, MW superconducting linac

If ILC near-term, continue facility incremental upgrades Priorities are NUMI, NoVA, … Some excess proton capacity for other experiments

3

From Oddone P5 Talk (9/2005)

0

20,000

40,000

60,000

80,000

100,000

120,000

140,000

160,000

180,000

FY05 FY06 FY07 FY08 FY09

DELAYED ILC DECISION -WITH PROTON DRIVER

EARLY ILC DECISION - NOPROTON DRIVER

EARLY ILC DECISION -WITH PROTON DRIVER

2005 2006 2007 2008 2009

RF module assembly and testing

CDR effort

Industrialization

Not affordable

Much longer R&D needed

Looks good! Press for early decision

Nothing yet

LHC discovery: GO

This induces other program choices

ILC

Proton Driver

BudgetRequired…

4

Proton Driver and Muon beams

8GeV Linac can produce streams of 1.5×1014 8GeV protons at 10Hz > 1022 protons/year

Neutrino Physics: main goal But: Only 1/15 of these needed for

Main Injector/NUMI

Are there muon beam experiments that could use this intensity ??

Tertiary muon beams: P + X → π π → μ + ν

10-2 μ/p → 1020 μ/year or more

(other experiments will also be possible)

~ 700m Active Length

8 GeV Linac

X-RAY FEL LAB

Slow-Pulse Spallation Source& Neutrino Target

Neutrino“Super-

Beams”

MainInjector@2 MW

8 GeVBooNe

NUMI

Anti-Proton

SY-120Fixed-Target

Off-Axis

Neutrinos to Homestake…

~ 700m Active Length

8 GeV Linac

X-RAY FEL LAB

Slow-Pulse Spallation Source& Neutrino Target

Neutrino“Super-

Beams”

MainInjector@2 MW

8 GeVBooNe

NUMI

Anti-Proton

SY-120Fixed-Target

Off-Axis

Neutrinos to Homestake…

Main Injector: 120 GeV, 0.67 Hz Cycle, 2.0 MW Beam PowerLinac Protons: 8 GeV, 4.67 Hz Cycle, 0.93 MW Beam Power Linac Electrons: 8 GeV, 4.67 Hz Cycle, 0.93 MW Beam Power

8 GeV Linac Cycles 1.5E14 per Pulse at 10Hz

Main Injector Energy

H-Injection

8 GeVProtons

8 GeVElectrons

0

20

40

60

80

100

120

140

0 0.5 1 1.5 2 2.5 3

Time (sec)

MI Energy

H- Injection

8 GeV Protons

Electrons

5

Primary Parameter List (Foster, March 2005 reference)

PRIMARY PARAMETERS 8 GeV Initial 0.5 MW {Ultimate 2MW in Brackets}Linac beam kinetic energy 8 GeVLinac Particle Types Baseline Mission

via foil stripping in transfer linePossible w/upgrade of Phase Shifters & Injector

Linac Stand-Alone Beam power 0.5 {2.0} MW 8 GeV beam power available directly from linacLinac Pulse repetition rate 2.5 {10} HzLinac macropulse width 3.0 {1.0} msLinac current (avg. in macropulse) 8.7 {26} mALinac current (peak in macropulse) 9.3 {28} mALinac Beam Chopping factor in macropulse 94 % For adiabatic capture with 700ns abort gap.Linac Particles per macropulse 1.56E+14Linac Charge per macropulse 26 uCLinac Energy per macropulse 208 kJLinac average beam current 0.07 {0.26} mALinac beam macropulse duty factor 0.75 {1.0} %Linac RF duty factor 1.00 {1.3} %Linac Active Length including Front End 614 m Excludes possible expansion lengthLinac Beam-floor distance 0.69 m =27 in. same as Fermilab Main InjectorLinac Depth Below Grade 9 m same as Fermilab Main InjectorTransfer Line Length to Ring 972 m for MI-10 Injection pointTransfer Line Total Bend 40 deg two 20-degree collimation arcsRing circumference 3319.4 m Fermilab Main InjectorRing Beam Energy 8-120 GeV MI cycle time varies with energyRing Beam Power on Target 2 MW ~ independent of MI Beam EnergyRing Circulating Current 2.3 ARing cycle time 0.2-1.5 sec depends on MI beam energy & flat-topRing Protons per Pulse on Target 1.50E+14 protonsRing Charge per pulse on target 25 uCRing Energy per pulse on target 200-3000 kJ at 8-120 GeVRing Proton pulse length on target 10 us 1 turn, or longer with resonant extractionLinac Wall Power 5.5 {12.5} MW approx 3 MW Standby + 1MW / Hz

ProtonsH - ions

Electrons

6

Review of the PD/Linac Cost

(All costs in 2004 K$) Linac

Project M&S* 229,779

Project SWF 89,118

Project Subtotal 318,897

G&A# (16.05% M&S & 30.35% Labor)

63,927

Project (incl. G&A) 382,824

Contingency (30%) 114,847

Total Cost 497,670

* Davis Bacon labor shows up as M&S# G&A rate will be lower on large POs

7

Proton Driver Parameters8 GeV Superconducting LINACEnergy GeV 8Particle Type H- Ions, Protons, or ElectronsRep. Rate Hz 10Active Length m 671Beam Current mA 25Pulse Length msec 1Beam Intensity P / pulse 1.5E+14 (can also be H-, P, or e-)

P/s 1.5E+15Linac Beam Power MW avg. 2

MW peak 200

MAIN INJECTOR WITH 8 GeV LINACMI Beam Energy GeV 120MI Beam Power MW 2.0MI Cycle Time sec 1.5 filling time = 1msecMI Protons/cycle 1.5E+14 5x designMI Protons/hr P / hr 3.6E+17H-minus Injection turns 90 MI Beam Current mA 2250

8

LFV: A eA can use high intensity

Experiment I0/Im T

[ns]

T

[s]

p

[MeV]

p/p

A eA

e eee

e e

1021

1017

1017

1016

< 10-10

n/a

n/a

< 10-4

< 100

n/a

n/a

< 1000

> 1

n/a

n/a

> 20

< 80

< 30

< 30

< 30

< 5

< 10

< 10

1…2

1014 < 10-4 100 > 20 30 < 10

g-2

EDM

1015

1016

< 10-7

< 10-6

< 50

< 50

> 103

> 103

3100

<1000

< 2

< 2

dtI

Desirable Beam Characteristics

But bunched beam is needed

9

A eA experiments

Next generation of A eA experiments has been proposed MECO – was to be based at

BNL (DOE and NSF rejected)

PRISM-PRIME – based at KEK/JHF

Could be hosted at FNAL proton driver ?

e Conversion workshop Jan. 11-12 at BNL

A eA produces monoenergetic e- (~105MeV)

SUSY SUSY predictions predictions ofof AA e e--A A

0 0

MECO single event MECO single event sensitivitysensitivity

10 -11

10 -13

10 -15

10 -19

10 -17

10 -21

PRIME single event sensitivityPRIME single event sensitivity

Rem (GeV)

From Barbieri,Hall, Hisano …

10

MECO expt. (BNL proposal rejected)

Superconducting Production Solenoid

(5.0 T – 2.5 T)

Muon Stopping Target

Muon Beam Stop

Crystal Calorimeter

Superconducting Transport Solenoid

(2.5 T – 2.1 T)

Superconducting Detector Solenoid

(2.0 T – 1.0 T)

Collimators

Tracker

Time structure

11

PRISM-PRIME (Y. Kuno et al.)

High intensity pulsed proton beam (bunch length <10ns)•100-1000Hz bunches producing π —> μ bunches

•Phase rotation with rf field: Δp/p : ±20% ± 2%

P = 68 MeV/c ±20%

12

PRISM rf-rotation

P = 68 MeV/c ±20%t = ±12ns (5ns rms)

Pµ

p=±1.9%p=±3.4%

5-turns, 38% beam decay 6-turns, 44% decay

13

Proton Beam requirements MECO experiment

Requires pulses of ~8 GeV protons (<30ns long) every ~1μs (1.4μs) – Obtained by slow extraction of short bunches (in AGS)

Design requires 41013 p/s, 1.5 10-3 captured μ’s/proton 105 μ/pulse ~61017 μ/year from ~4 1020 p/year

PRISM-PRIME experiment Requires proton pulses (<10ns long) at 103/s (~1ms)

– 4 1014 p/s (50GeV) 10-2 to 10-3 μ’s/proton– Up to 1022 p/year, > 1019 μ/year

Single-turn extraction of short bunches (<10ns) Up to 4109 μ/pulse

Both require pulsed beams, proton linac beam must be repackaged in an accumulator ring

14

Recycler as accumulator ring ? 8GeV Linac produces

1ms pulses at 10 Hz

H- injection into Recycler 1ms fills circumference

– (100 turns) Bunch beam into pattern

required for expt.

Harmonic 10 buncher for MECO, slow extraction

Harmonic 100 buncher for PRIME, single bunch extraction

Circumference C=2πRave 3320m

Momentum P 8.89 GeV/c

Rev. frequency,

Period

f0

T0

89.8 kHz

11 μs

Slip factor η=1/γ2- 1/γt2 0.0085

Tunes νx, νy 25.4,24.4

But:

Recycler circumference is large

100ms may be too short a time for bunching

15

Space Charge Difficulty Space Charge tune shift:

Parameters: Ntot=1.51014,εN =20π mm-mrad

MECO: 30ns/1μs : BF= 0.03 →δν = 4 : too large

Reduce N to 1.51013 →δν = 0.4 Reduce N to 0.41013 →δν = 0.1

PRISM/PRIME 10ns bunches, 100/ring BF= 0.1 →δν =1.2: too large (but closer)

Larger εN, smaller Ntot,

Smaller circumference proton ring could be better

p tot

2F N

3r N

B

F

0.12

B

16

Recycler – Bunching (~for PRISM) Harmonic 100 buncher (9MHz)

Bunch for 0.1s (Vrf ramps to 140kV)

Bunch lengths reduced to

~5ns rms(Prism wants < 10ns full width.)

Could then extract bunches one at a time over ~0.1s

Uses 1/2 the possible linac pulses (500 bunches/s for PRISM) (100 at 5Hz)

17

Other potential proton storage schemes Accumulator or Debuncher (C= ~454m) after

2010… Large aperture machines t 5

Difficult to inject H- (must bend beam from Linac) (B < 0.05T at 8 GeV, ρ > ~600m)

Could take debunched protons from Recycler or Main Injector(in ~450m chunks) Bunch into pattern needed for experiments Bunching easier than Recycler

New 8 GeV Storage Ring ??

18

New 8 GeV Accumulator/buncher/stretcher

FFAG Type: FODO racetrack,

Superferric arcs nonscaling

H- injection into NewRing (10Hz) 700 turns

δν = 0.4 at BF=0.15 (σ=1.5ns)

Harmonic 42 buncher for PRISM, single bunch extraction (40ns spacing)

Slow extraction, single bunch extraction modes

Circumference C=2πRave ~454m

Momentum P 8.89 GeV/c

rf frequency,

Voltage

H=42

V0

26 MHz

1MV

Slip factor η=1/γ2- 1/γt2 -0.02

Tunes νx, νy 6, 8

aperture a, b ~3,2cm

19

Proton Linac (H-)

NewRing (P)

20

Fermilab – w/o SRF linac proton driver

NOW: Linac8 GeV Booster (C=454m) Currently Limited by losses in Booster 5Hz ~ 3 1012/ pulse 15Hz, 5 1012/pulse possible (0.1MW) (61013 protons/s)

NUMI, MiniBOONE, Tevatron p-source MiniBOONE will terminate soon ??

Single turn extraction is 1.5s For pulsed experiments, need a storage ring Can use Recycler, Accumulator, Debuncher? Future: New 8 GeV Booster

With 1 GeV linac ?? larger apertures, larger injection energies, deeper tunnel (51013/ pulse ??) (1MW ??) With new storage ring

21

• Detailed source design does not exist Straw man design worked out for the front end of a factory supported by MARS simulations (Ray et al.)

•Target + capture solenoid + drift (forward capture)

• 1.4 x 1022 protons/year at 8 GeV yields ~3 x 1021 muons/year.

Charged particlespectra at end ofdecay channel

Generic High intensity muon beam

22

Summary Muon Beams from the Proton Driver could be

enabled Potential muon beam facilities could be

developed:

PRISM, etc. … could be hosted More Detailed design needed

Proton Collection– Recycler ….– New Stretcher/Buncher ring ??

Beam line(s) Experimental area(s)

23

References W. Foster et al., Proton Driver http://tdserver1.fnal.gov/project/8GeVLinac/DesignStudy/

W. Molson, “The MECO Experiment to Search for -Ne-N with 10-17 Sensitivity”, U. Va. Seminar, June 2004

MECO ‘RSVP’ Rare Symmetry Violating Processes (MECO-

KOPIO) NSF proposal, October 1999. PRISM Working group “An Experimental Search for the μ−−e− Conversion

Process at an Ultimate Sensitivity of the Order of 10−18 with PRISM”, The Prime Working Group, Jan. 1, 2003.

R. Ray & D. Roberts, Proton Driver physics study

24

Recycler – Bunching for ~MECO Harmonic 10 buncher

(0.9MHz) Barrierbucket rf

Bunch for ~1s (Vrf ramps to ~30kV)

Bunch lengths reduced to

~50ns rms(MECO wants ~30ns full width.)

Could then extract bunches in slow extraction over ~1s