m. apollonio

124/3/2010 CM26 - Riverside 1

m. apollonio

(,P) matrix

224/3/2010 CM26 - Riverside 2

- the goal of MICE is demonstrating Ionisation Cooling …

- … for a variety of initial - emittances- momenta

- ideally covering a continuos space (,P)- practically studying some discrete points

- i.e. defining a matrix

- the choice for it is:- = 3 / 6 / 10 mm rad- P= 140 / 200 / 240 MeV/c (at the centre of the H2 absorber)

General Introduction

324/3/2010 CM26 - Riverside 3

3,1403,140 3,2003,200 3,2403,240

6,1406,140 6,2006,200 6,2406,240

10,14010,140 10,20010,200 10,24010,240

P (MeV/c)

eN

(m

m r

ad

)

- finding the element (3,240) means to find the BL optics that matches the MICE optics for a beam of 3 mm rad at a P=240 MeV/c

- the element (10,200) is the BL optics matching a MICE beam with 10 mm rad at P=200 MeV/c

This pair is our goal: how do we get it?

424/3/2010 CM26 - Riverside 4

(*) MICE note 176

BLBL DiffuserDiffuser MICEMICE

- Hyp.: is known (~1 mm rad trace space) - we proceed backward: - fix P/N in the cooling channel- fix the optics in the cooling channel ()- solve the equations giving and t at the US face of the diffuser (*)

t

524/3/2010 CM26 - Riverside 5

- So the question becomes: - how do we “tell” the beamline to be at US_Diff?- solution(s)

- we optimise the BL by varying Q4-Q9 - let us break the BL in two parts: US and DS- in what follows I mean a beamline

Q4Q1

Dip

ole1

DK solenoidQ2 Q3

Dip

ole2

Q5 Q6 Q7 Q8 Q9

- US part: we can optimise the MAX number of pions- but not much magic left …

- DS part: - choose Q4-Q9 - shoot a beam- check at Diffuser vs “target” values- repeat

beam line: typical spectrum at the exit of the DS

- Rationale- select u.s. of DKSol with D1- select d.s. of DKSol with D2

- back scattered muons == purity

24/3/2010 6CM26 - Riverside

24/3/2010 CM26 - Riverside 7

3,1403,140 3,2003,200 3,2403,240

6,1406,140 6,200 6,2406,240

10,14010,140 10,20010,200 10,24010,240

we already have an initial solution: the “central value”

24/3/2010 CM26 - Riverside 8

kinematic limits

250 MeV/c

195 MeV/c

24/3/2010 CM26 - Riverside 9

Will it work? Pdiff = 215

In the original scheme the pi mu beamline is Ppi=444 Pmu=256Best separation PI/MU

acceptance acceptanceNB.: PD2=256 MeV/c becomes Pdif=215 MeV/c

24/3/2010 CM26 - Riverside 10

3,140Pdif=151=0.2

=0.56mt=0.0mm

3,140Pdif=151=0.2

=0.56mt=0.0mm

3,200Pdif=207=0.1

=0.36mt=0.0mm

3,200Pdif=207=0.1

=0.36mt=0.0mm

3,240Pdif=245=0.1

=0.42mt=0.0mm

3,240Pdif=245=0.1

=0.42mt=0.0mm

6,140Pdif=148

=0.3 =1.13mt=5.0

6,140Pdif=148

=0.3 =1.13mt=5.0

6,200Pdif=215=0.2

=0.78mt=7.5mm

6,240Pdif=256=0.2

=0.8mt=7.5mm

6,240Pdif=256=0.2

=0.8mt=7.5mm

10,140Pdif=164=0.6

=1.98mt=10mm??

10,140Pdif=164=0.6

=1.98mt=10mm??

10,200Pdif=229=0.4

b=1.31mt=15.5mm

10,200Pdif=229=0.4

b=1.31mt=15.5mm

10,240Pdif=267=0.3

=1.29mt=15.5mm

10,240Pdif=267=0.3

=1.29mt=15.5mm

24/3/2010 CM26 - Riverside 11 195 350

Pdiff = 148 215 256

Ppi (tgt) = 350

i.o.t. accommodate several mu momenta another “shortcut” scheme was adopted (aug 2009):Define one lower Ppi ~ 350/360 and several different Pmu (we lose in purity …)

acceptance acceptance

Q4Q1

Dip

ole1

DK solenoidQ2 Q3

Dip

ole2

Q5 Q6 Q7 Q8 Q9

d.s. BL tuning: match to diffuser

P=444 MeV/c

fix D1 fix D2

1224/3/2010 CM26 - Riverside 12

24/3/2010 CM26 - Riverside 13

- a first round of the BL optimised (e,P) matrix has been produced in august 2009 (“shortcut”)

- however the few data taken in november reveal a pretty strange look

- one thing I dislike is using only one momentum for the pion (US) component and Select the backward going muons

http://mice.iit.edu/bl/MATRIX/index_mat.html 1424/3/2010 CM26 - Riverside

RUN 1174-1177 – PI- (444MeV/c) MU- (256 MeV/c) at D2

~29.

NB: DTmu(256)= DTmu(300) * beta300/beta256 = 28.55 * .943/.923 = 29.13

0.943269

0.943269

PI- should be here: 30.44

24/3/2010 15CM26 - Riverside

?RUN 1201 – PI- (336.8MeV/c) MU- (256 MeV/c) at D2

PI- should be here: 30.44

MU- should be the same asbefore … what is that?24/3/2010 16CM26 - Riverside

24/3/2010 CM26 - Riverside 17

y

x

x’

y’

COV-MAT

GenerateGaussianBeam with defined COV-MAT(arbitrary statistics)

G4Beamline Generation upTo DS

24/3/2010 CM26 - Riverside 18

a) Consider all 9 cases: one Ppi + one Pmu per case (no “shortcuts”)b) Define initial BL currents (from scaling tables)c) Check tuning with G4Beamlined) use simulation output at DS to infer the COV-MAT of the beame) Generate a Gauss-beam with that CovMat:

a) E.g. MatLab tool, fast + any number of particles …f) Propagate / optimise this beam in the DS section

a) By hand (GUI tool)b) By algorithm (GA)

g) check results versus real data …

wrap-up …

24/3/2010 CM26 - Riverside 19