November 17, 2007 1 Chuck AnkenbrandtFermilab Making More Muons mu2e Conversion and higher intensity beams Project X Physics Workshop Chuck Ankenbrandt

November 17, 2007 1Chuck Ankenbrandt Fermilab

Making More Muons

mu2e Conversion and higher intensity beams

Project X Physics Workshop

Chuck AnkenbrandtFermilab

November 17, 2007


Game Plan for this Talk

My title (from the organizers) suggests three topics:

Delivering protons from Project X to mu2e I will discuss this only briefly unless there are

questions.

Providing more/better muons than MECO This is the main topic of my talk. I will present preliminary ideas to induce

brainstorming.

Upgrading the detector for higher rates This is under active consideration by the

collaboration. I will not discuss it here.


Delivering the Protons From the Booster: Cf. Jim Miller’s mu2e presentation At similar intensities (~30 kW), but from Project X:

Cf. Letter to Steering Group: “Using an ILC-Style 8 GeV H- Linac for a Muon to Electron Conversion Experiment” by C. Ankenbrandt, S. Geer, and E. Prebys (also backup slides)

Up to ~ 200 kW (full Project X 8-GeV beam power): Extract one complete Recycler fill (1 to 4 Linac squirts) in a single turn at MI52 and transmit it via P150 line to the Accumulator. Inject the beam into the Accumulator using multi-turn (7-turn) transverse stacking in both transverse planes. That should be feasible because the transverse acceptances in the Accumulator are each about an order of magnitude larger than those of the Recycler.


Muon Beam Introduction Muon Beams would be useful for …

National or International Lepton Colliders (NLC or ILC!) Neutrino Factories Precision Physics beyond the Standard Model

– Muon g-2

– Muon EDM

e– Muon to electron conversion (mu2e): (focus of this talk)

Muons, Inc/Fermilab recently got an SBIR grant to explore whether innovations in muon cooling can be used to improve Stopping Muon Beams, particularly for mu2e.


Intro., continued: Approximate rates An 8-GeV proton beam on an optimized target produces

about one charged pion of each sign per proton.

Approx. Rates mu2e

Muon Collider Nu Factory From Project X From Booster

Pbeam, MW 2 2 0.2 0.027

Np/sec 1.56E+15 1.56E+15 1.56E+14 2.11E+13

Mu/p collected 0.1 0.1 0.0025

fraction cooled 0.07 0.7? 1

Muons/sec 1.09E+13 1.09E+14 5.27E+10


Stopping Muon/proton ratio Reaction of muon collider community: “Why is the ratio

so small?”

Reaction of stopping muon community: “How does MECO get such a large ratio?!”


More intro.: Comparison of Front Ends

Muon Collider mu2e Muon Beam

Cost I s no object Must be reasonableCollect both signs Desirable UndesirableDuty cycle Pulsed CWBeam cooling Many stages Single pass at mostReacceleration Lots of rf No rfBeam contamination Not important I mportant

Those familiar with muon colliders should realize that stopping muon beams have very different requirements.

(If and when a muon collider exists, a stopping muon program could use the cold muon beam when the collider is not operating.)


MECO


Stopping Muon Beams 101 p + A -> + X ~One charged pion of each sign per 8-GeV proton Pion decay length is 7.8 meters for p = mc decay kinematics in lab:

Less than 29 MeV/c of transverse momentum Longitudinal momentum distribution of muons is uniform between

about 60% and about 100% of pion momentum Polarization correlates completely with longitudinal momentum

Passing the muons through material causes the momentum spread to grow (dE/dx causes longitudinal heating)

Where the pions are: cf. next slide

Momentum vs. Cosine of production angle (from C. Yoshikawa)

Momentum vs. Cosine of production angle (from C. Yoshikawa)

MECO

Alt.A

Alt. B


Concepts (a) production

(b) momentum evolution

SBIR Cooling Concept: 6D particle density

increase Absorber density

decrease

MECO acceptance

HCC acceptance

a) b)

(From Mary Anne Cummings)


However, … Above example used collider target solenoid (20 T)

Not a fair comparison with MECO (5 T) Perhaps not practical for mu2e

Also, collaborators dislike proton beam pointed at detector

Ergo, explore other concepts:


Target + Wedge @ Edge of Dipole

Proton beamtarget

BR=1 TmwedgeStart of match to HCC

Followed by low-Z absorber in HCC to cool the beam and reduce its energy.


Advantages of Wedge@edge Hard momentum cutoff at ~ 150 MeV/c Eliminates wrong-sign particles Full width of target magnet is used: magnetic field volume is not wasted Particles heavier than muons stop in degrader, not in stopping target; thus much better hadron background rejection than in MECO design Electrons get eaten by high-Z wedge, thus probably not a problem. Better background suppression means might live with higher intensity Little dispersion in muon arrival times May have better mu/proton ratio: Uses pions at peak of momentum distribution Uses pions produced near zero degrees May be less expensive than MECO (maybe even fly under P5 radar?) Probably can operate with shorter deadtime after proton arrival

=>More usable muons per proton=>Could use titanium or other material having Z greater than

Aluminum Proton beam points away from experiment and is easily dumped. Produces polarized stopping muon beam Polarization can be varied by changing the thickness of the degrader


Cartoon of target toroid (thanks to M. Popovic)

Iron Toroid for Pion Collection


Iron Toroid Muon Beam One Lint (16.7 cm) of iron alloy magnetized to 2.5 T:

Attenuates by 1/e (but interactions in it produce more pions) Bends 400 MeV/c particle by 312 mr Degrades minimum ionizing particles by about 200 MeV/c

So an iron toroid in the forward direction can: Make a parallel pion beam for a selected momentum p* Shift the momentum spectra at each prod. angle to peak at p* Pick p* to maximize muon flux after HCC at stopping target

Works much like a DC horn

Perhaps follow with Dipole+Wedge for muons, then HCC


Summary Project X can supply protons for mu2e

At Booster intensities, and Up to 150-200 kW

Muons, Inc. and Fermilab are exploring alternatives to MECO for pion production, muon collection and delivery. Muon cooling via HCC looks promising. Effective use of cooling requires different production concepts.

The mu2e collaboration is considering detector upgrade possibilities.


Backup slides


Additional Capabilities/Other Compelling Physics

(The MI-based neutrino program) A ->e conversion experiment: mu2e

wants to momentum-stack in the Accumulator, then rebunch and slow-extract from Debuncher. It would benefit from cooling of stopping muons.

A muon cooling experimental program: A prototype front end

– Target station with low beam power

– Muon collection, cooling, and re-acceleration A muon test beam (including slow-spill beam) Test concepts and components

Etc.


AP-4 Line

AP-5 Line

The original SNuMI connections

Euclid: “The shortest distance between two points is a straight line.”


Fundamental Geographyof the Main Injector

Main Injector

MiniBooNE

Pre-Accelerator

Tevatron

P3

P2

RingsPbar

MI-20

MI-10

MI-22

MI-30

MI-32

MI-62

MI-60

MI-52MI-40

MI-50

A1

P1AP-1

DumpAbort

Target

MI-8

Switchyard

Linac

F49

Booster

AP-3AP

-2

MI-8

F17

F0

The proposed path

The cheapest route between two rings is via existing beam lines.


MI and RR time lines for NOA

The Recycler is empty for as many as 8 Booster ticks.

Charge for MI

Charge MI Energy

Time, 15HzMI

RR

<-20 Booster ticks->


Bending the beam out of the Recycler

Q522AQ522B

RECYCLER

MAIN INJECTOR

(The path into the Recyler will be implemented for NOA)


Bending the beam into the P150 Line

~20 mr vertical bend out of Recycler and into P150

P150LINE


Matching Recycler to P150 Line

The MAD output for the P150 line. The red arrows indicate the approximate separation betweenthe two dipole magnets, with the first being in the Recycler and the second in the P150 line.


Matched Transfer Line from Recycler to P150

15.000 mm X 1.000 mrad C:\wint\charsets\

20.000 Deg X 1000.00 keV C:\wint\charsets\

15.000 mm X 1.000 mrad

20.000 Deg X 1000.00 keV NP2= 14

10.00 mm (Horiz) 0.0 Deg (Long.)

10.00 mm (Vert)

NP1= 1 C:\wint\charsets\

Length= 40164.98mm

1 C:\wint\charsets\ 2 C:\wint\charsets\ E 3 C:\wint\charsets\ B(v) 4 C:\wint\charsets\ E

5 C:\wint

6 Q

7

8 Q

9

10 E 11 B(v)

12 E 13

H A= 2.4000 B= 53.100 C:\wint\charsets\ V A=-0.80000 B= 14.100 C:\wint\charsets\

Z A=-5.00000E-02 B= 1.00000E-02 C:\wint\charsets\

BEAM AT NEL1= 1 C:\wint\charsets\ H A= 0.62846 B= 11.324 V A= -2.9004 B= 67.022

Z A= 5.07258E-02 B= 1.00000E-02

BEAM AT NEL2= 14 I= 500.0mA W=8000.0000 8000.0000 MeV

FREQ= 53.00MHz WL=5656.46mm EMITI= 0.900 0.900 3700.00 EMITO= 0.900 0.900 3700.00

N1= 1 N2= 14 PRINTOUT VALUES PP PE VALUEMATCHING TYPE = 8

DESIRED VALUES (BEAMF) alpha beta x 0.6000 11.1000 y -2.9000 66.3000 MATCH VARIABLES (NC=4) MPP MPE VALUE 1 6 -4.49451 1 8 4.76013 1 5 8893.92450 1 7 17000.00000

CODE: Trace 3-D v69LY FILE: rrp150.t3d DATE: 02/18/2007 TIME: 14:50:32


Synergies with SNuMI

B->R->AS will: eliminate the need for the AP-4 line, provide opportunity for early commissioning of momentum-

stacking in the Accumulator, and motivate mitigation of radiation issues in the enclosure.

Modifications for SNuMI era: AP-5 line or equivalent will still be needed. There will still be a slot in the Recycler to transport the beam. We’ll need a fast extraction system for Recycler to P150 line. Only two “free” Booster batches will remain for other programs.


B->R->AS Requirements

Two dipole magnets (Cooling Ring Dipoles) Two Transrex Power supplies (available) Two quads such as existing Recycler quads Two trim dipoles and two trim quads All these components are available


Recent documentation (note contributors)

Delivering Protons to the Antiproton Source after the Tevatron Collider Era

Charles Ankenbrandt, David Harding, David Johnson, David McGinnis, and Milorad Popovic

Fermilab

March 6, 2007

…

Communications with Dixon Bogert, Bruce Brown, Steve Geer, Nancy Grossman, Dave Neuffer, and Eric Prebys contributed to the development of the ideas presented here.


Abstract=Summary

“A way to deliver 8-GeV protons from the Booster to the Antiproton Source via existing enclosures and beam lines is described. By using the existing 8-GeV line and most of the Recycler as parts of the beam path, the scheme avoids the need for new civil construction and new beam transport lines. In this way, as soon as the Tevatron Collider era is over, the Antiproton Source can be rapidly transformed into a very useful pair of proton storage rings for various applications.”


Momentum-stacking in the Accumulator

“Self-Bunching of a Coasting Beam in the Accumulator” by Dave McGinnis

Conclusion of that paper: “For the projected SNUMI intensity of 14.1x1012 particles in

the Accumulator at a 95% momentum spread of 15.9MeV, and an RF feedback gain of 14dB, the beam should be stable with a factor of two margin. This result can be tested by cooling 100mA of antiproton beam on the stacking lattice (h=0.10) to a frequency width of 7.9 Hz.”

B->R->AS would enable commissioning of momentum stacking in the Accumulator

http://beamdocs.fnal.gov/SNuMI-public/DocDB/ShowDocument?docid=198

http://beamdocs.fnal.gov/SNuMI-public/DocDB/ShowDocument?docid=198


8 GeV Linac (Project X) and mu2e

Chuck AnkenbrandtFermilab

Mu2e Collaboration MeetingAugust 1, 2007

My Preferred title:Linac & Recycler & Accumulator & Debuncher: Consider the Possibilities

(Like 1969 movie)


Considering the possibilities


Outline 1) Introduction/Overview

2) Linac & Recycler & Accumulator & Debuncher: Consider the Possibilities a) Review/discuss Fermilab BEAMS-DOC-2812-V:

Using an ILC-Style 8 GeV H- Linac for a Muon to Electron Conversion Experiment by Ankenbrandt, Geer, and Prebys

b) Update of that note

3) Summary/Conclusions


Proton Linac (H-) 8 GeV?

H-

t

8 GeVProton sources


1) Introduction/Overview The question that is addressed is: How will mu2e get proton beam in the Project X era? In a note to Young-Kee Kim’s Steering Group, we (Geer, Prebys, and I) discussed that question. In this talk, I will briefly review and update that note. For completeness, a copy of that note is included in this document. Assumptions that went into that note: 1) By the time that Project X is commissioned, mu2e will already be running (concurrent with NOnA or SNuMI)

using beam from Booster to Recycler to Accumulator to Debuncher. 2) Lab management will want to decommission the present Linac and Booster once Project X is commissioned. What’s new since the note was written: 1) Rumor has it that experts are pessimistic about the feasibility of slow extraction from the Recycler. If verified, that

would preclude running directly off the Recycler. 2) Ways have been conceived for using more beam from Project X; variations of option 3a. (Thanks to interesting

conversations with Tom Roberts and Dave Johnson) 3) Hence I’ve been thinking more recently in terms of an intensity upgrade rather than a mere continuation of mu2e at

similar intensities.


Title page of the note Using an ILC-Style 8 GeV H- Linac for a Muon to Electron Conversion

Experiment C. Ankenbrandt, S. Geer, and E. Prebys

Fermi National Accelerator Laboratory, PO Box 500, Batavia, IL 60510 Abstract

We describe how the H- beam from an ILC-Style 8 GeV H- Linac can be collected, rebunched, and slowly extracted to provide a beam suitable for a muon to electron conversion experiment (mu2e). The scheme would permit simultaneous operation of the muon program with the future NuMI program, delivering O(1020) protons per year at 8 GeV for the mu2e experiment.


INTRODUCTION

In a communication to the Fermilab long-range steering committee, D. McGinnis [1] has proposed to upgrade the Fermilab proton source by replacing the existing Linac/Booster with an ILC-Style 8 GeV H- Linac delivering 9 mA in 1 ms long pulses at 5Hz. There would be 7 Linac pulses per 1.4 sec Main Injector cycle, each pulse containing 5.7 1013 H-. The first 3 pulses would be injected via H- injection into the Recycler. The circulating 1.7 1014 protons would then be extracted in a single turn and transferred to the Main Injector to be used for the NuMI program. This would leave up to 4 Linac pulses, which could also be injected into the Recycler, available for an 8 GeV physics program. Previous notes have described how 8 GeV protons from the Booster can be transferred to the Fermilab Antiproton source [2], then rebunched and slow extracted [3] to produce a primary beam suitable for a muon to electron conversion experiment [4]. In this note we describe how protons from the McGinnis scheme [1] can be used for a muon to electron conversion experiment. We assume that the scheme described in References [2] and [3] to deliver protons from the Booster to the Accumulator for the experiment will have already been implemented by the time the new Linac is commissioned. It is also assumed that the present Linac and Booster will be decommissioned shortly after the new Linac is commissioned, so beam for the experiment would have to come from the Recycler. One way to accomplish this would be to run the experiment directly off the Recycler; the other way would be to transfer beam to the Accumulator. The two alternatives are discussed in the following paragraphs.


Directly off the Recycler (probably not feasible)

Once per Main Injector cycle, one Recycler fill (5.7 1013 protons) would be used for the muon to electron conversion experiment. The protons would be rebunched into about 7 equally-spaced bunches, then slowly extracted for about 0.7 seconds from the Recycler near MI52 and transferred to the Antiproton Source enclosure via the P150 line. The beam would then be transported directly to the experiment, bypassing the Accumulator. That would require a new beamline in the Antiproton Source enclosure connecting the P150 line to the transfer line to the experiment. The other three available Linac cycles would not be used. This scenario would provide about 50 kW of beam power at 8 GeV with a duty cycle of about 50% for the muon conversion experiment. It would preclude other uses of 8 GeV beam which require the Recycler.


Recycler to Accumulator Various ways to take beam from the Recycler to the Accumulator for the experiment can be conceived. Subsequent beam

processing in the Accumulator and Debuncher would then be similar to the plan that uses beam from the Booster. a) In one scheme, one complete Recycler fill is extracted in a single turn at MI52 and transmitted via P150 to the Accumulator. The beam is injected into the Accumulator using multi-turn (7-turn) transverse stacking in both transverse planes. That should be feasible because the transverse acceptances in the Accumulator are each about an order of magnitude larger than those of the Recycler. That scheme would provide about 50 kW of beam power to the experiment with a duty cycle of about 90%. The three other available Linac cycles could be used for other 8-GeV physics. b) Another method would “steal” about 1/7th of the beam destined for the Main Injector. The beam occupying one seventh of the circumference of the Recycler would be kicked out and transmitted to the Accumulator via the path described previously. No stacking would then be required in the Accumulator. That would provide about 21 kW of beam power for the experiment with a duty cycle of about 90%. It would also create a useful abort gap in the Main Injector beam. The four other cycles available from the Linac could all be used for other 8-GeV physics. c) The third method would “steal” Accumulator-length batches from each of the four Recycler fills not destined for the Main Injector. Each of those could then be separately rebunched in the Accumulator, transferred to the Debuncher, and slowly extracted over 200 msec. That would provide about 28 kW of beam power for the experiment with a duty cycle of about 50%. The rest of these four cycles could be used for other 8-GeV physics.


Update: intriguing options for more proton beam power.

1) Full 200-kW capability using transverse stacking into the Accumulator: a) Take two linac squirts at a time into the Recycler; b) Single-turn extract the whole circumference at once ; c) Transverse stack into the Accumulator; d) Form a single bunch in the Accumulator; e) Transfer to the Debuncher; f) Slow spill from the Debuncher; g) Repeat this process twice per Main Injector cycle. 2) As above, but using momentum stacking: a) Take two linac squirts at a time into the Recycler; b) Extract Accumulator-length batches at 33-ms intervals; c) Momentum stack into the Accumulator; d) Form a single bunch in the Accumulator; e) Transfer to the Debuncher; f) Slow spill from the Debuncher; g) Repeat this process twice per Main Injector cycle.


Email from Nagaslaev From vnagasl <[email protected]> Sent Monday, July 30, 2007 8:40 pm To [email protected] Subject Accumulator apertures Chuck, I think principal limitation of the Accumulator acceptance after necessary rearrangements would be around 20 pimm unnormalized. However, one of the horizontal limitations currently is Extraction Lambertson (15 pi), and one would need to increase substantially the EKIK power in order to bring it to 20 pi. These numbers are not official. Vladimir


Summary from beam note If the proton source is upgraded with an ILC-style 8 GeV

Linac, it appears there are several options that would enable an intense beam of 8 GeV protons to be provided, with the appropriate bunch structure, for a muon to electron conversion experiment.

Documents

November 17, 2007 1 Chuck AnkenbrandtFermilab Making More Muons mu2e Conversion and higher intensity beams Project X Physics Workshop Chuck Ankenbrandt