1June 2 nd 2009MICE CM24 - RAL1 m. apollonio Beamline+( ,P) matrix
-demonstrate ionization cooling in a variety of initial emittances and momenta ( ,p) emittance due to decaying in the PSI solenoid (~1-2 mmrad geometric) - use a diffuser MICE goal MICE experiment MICE beamline beamline/experiment interface: the MICE diffuser 2June 2 nd 2009MICE CM24 - RAL2
Q4 Q1 Dipole1 DK solenoid Q2Q3 Dipole2 Q5Q6Q7Q8Q9 1) beamline momentum tuning P =208 MeV/c P =444 MeV/c P =214 MeV/c fix D1 fix D2 2) beamline quadrupole tuning P =255 MeV/c 3June 2 nd 2009MICE CM24 - RAL3
BeamLine Completion: tuning for all ( ,P) Also known as... the MATRIX search for Q4-5-6 & Q7-8-9 currents to match optics at the u.s. face of the diffuser fine tuning for every chosen [3,6,10] mm rad (norm.) at every defined P = [140,200,240] (MeV/c) 4June 2 nd 2009MICE CM24 - RAL4
Q4 Q1 Dipole1 DK solenoid Q2Q3 Dipole2 Q5Q6Q7Q8Q9 INPUT: beamline US section >>>>>>>>> GA procedure - Define 6-genes genotypes: Q currents - Express it as a phenotype: the Twiss parameters at some Z’s - Evaluate phenotype’s fitness - Mate genotypes and produce new individuals - Choose the best & repeat for several cycles GA+Turtle Optimiser OUTPUT: Twiss diffuser US face (+TOF1) 5June 2 nd 2009MICE CM24 - RAL5
6June 2 nd 2009MICE CM24 - RAL6
=78 cm =132 cm =0.4 cm =0.2 cm 7June 2 nd 2009MICE CM24 - RAL7
Pro’s and con’s of the method - running is fast (2.5 h for initial 70K muons after DKsolenoid) - preparing it a bit cumbersome (TTL…) - 3 decks (US / MID / DS) - I think of a change: - generate (G4BL) and record after DKsolenoid - use THIS output as input for GA+TTL optimisation NB In principle the GA algorithm should be able to run G4BL directly - It calls a script via a system call - possibility of tuning the solenoid - CAVEAT : a) need to insert the spectrometer solenoid in the simulation (fringe field) b) could take some time to have a working version 8June 2 nd 2009MICE CM24 - RAL8
are we ready? (shutdown 17 Aug/1 Sept) 2Sept – 25 Oct STEP I depends on what we want to reach - beam optics for a beam of 214 MeV/c before diffuser IS ready - it can be easily scaled for other momenta - requires more work IF we need tuning (e.g. define some specific Twiss parameters at Z=fixed) STEP II a bit more laborious - as before, optics exists for =6,10 and P=214 MeV/c - matrix has to be completed - this requires values for B(tracker) at 140/240 MeV/c [ B(140)=140/200 * B(200) and B(240)=B(200) ] 9June 2 nd 2009MICE CM24 - RAL9
are we ready? XY (a) well defined problem know in solenoid know at US-diffuser -line tuned ( =6,10 mm rad, P=214 MeV/c) as case (a) but beam diverges due to absence of solenoid SUGGESTION: use case (a) tuning STEP II Q7Q8Q9 Q7Q8Q9 STEP I 10June 2 nd 2009MICE CM24 - RAL10
P TARGET (MeV/c)P (MeV/c) P TARGET (MeV/c)P (MeV/c) 484~250 P TARGET (MeV/c)P (MeV/c) 390~160 - a level-0 optics for any P can be obtained by momentum rescaling - then a Q tuning is required to match the line P after DKsolenoid 11June 2 nd 2009MICE CM24 - RAL11
recalculated =6mm rad, P=207 new tuning =10 mm rad, P=207 investigate case =3 mm rad P=207 (does not converge) PLAN: fix P=140/240 and cover the remaining empty elements - either changing D1 (and DK) - or letting it fix and using D2 only (to select ) ??? June 2 nd 2009MICE CM24 - RAL12
- a central optics exists for ( ) beamline with P ~200 MeV/c and =6 mm rad, 10 mm rad - level-0 optics for P =140/240 can be generated by momentum rescaling - a fine tuning of Q[4-9] can be done to match the optics to US-diffuser values [STEPII] - optics for STEPII can be used for STEPI - so far DKsol is kept fixed (not tuned) - could be inserted in tuning if TTL G4BL (but requires work) STEPI: emittance calculation via TOFs - STEPII: via TOFs + Tracker Summary 13June 2 nd 2009MICE CM24 - RAL13