Optimization of phase rotation parameters for the n -factory.ppt

Optimization of phase rotation parameters for the n -factory M. Apollonio – University of Oxford Leading idea: 5 figures used to parametrize the F-rotation Goal: Trying to reach the “best” combination of these values to produce an “ideal” muon beam Instead of doing it by hand (lengthy/tedious) use a minimization procedure ? MINUIT Wrap ICOOL+endof9+ecalc9 in a MINUIT main routine Discuss results ... Parameters initial (“typical”) values meaning/function ref1 ---------------------- -.48 reduce energy (negative ramp) ref2 ---------------------- 2.64 increase energy (pos. ramp) delta --------------------- 18.04 # of w.l. separating the ref. particles grad ---------------------- 12 RF gradient ref1(2)=ref2(2) -------- .214 energy level in the cool. section ref part 1 ref part 2 F -rotation final energy level t or Z E Which minimization? Probably a matter of taste, but I ended up with a SIMPLEX minimization which looks faster than usual MINImize and reliable Structure of the program minuit routine: change parameters modify ICOOL for001.dat file (new input) run ICOOL+ENDOF9+ECALC9 compute sigma(E) at the END of the COOLING channel END verify minimization convergence Technical note: To speed the algorithm up I used a smart set of 40 muons (only) provided by Bob Energy distribution at the beginning of the channel (MeV) Muons are injected at the beginning of the drift channel Function to minimize: Sigma(E)/3 MeV (at z=216, 226, 232, 239.6, 247 m) 1st attempt: leave 1 or 2 running while fixing the other par’s Check shape of minimized function Subsequently release 1-4 par’s (p5 fixed) and run minimization Results F -rotation final energy Z E sE(initial) sE(final) REF2 2.85 2.59 3.20 2.45 initial values -0.4 -0.5 initial values Fix all but P2=REF2 Fix all but P1=REFP NB: MINImize routine used here ~2.67 ~-0.446 FCN vs P1 P2 vs P1 Fix 3,4,5 and minimize over P2 scanning on