Latest News On NuMI Beam Simulation Mark Messier Harvard University NBI22, CERN 17 March 22 - Basics of LE beam - Attempt to optimize for flux - Variable target position - Commissioning studies - Hadron production
NuMI Monte Carlos NuMI currently uses three MC's for beam GNUMI - GEANT3/GEANT-FLUKA based - Used for prediction of all neutrino species - No hard coded geometry - All information about decays stored - Inputs from own target simulation or particle list PBEAM - Fast parametric MC for muon neutrino fluxes - Used for quick optimization studies and alignment - Used to generate muon inputs for GNUMI - Approximate treatment of multiple scattering and absorption - Fast! 1x GNUMI MARS - Fermilab FLUKA equivalent - Used for radiation protection and energy deposition calculations - Also used to simulate target cascade as input to GNUMI
Latest NuMI Spectra νµ CC Events/kt/yr/GeV 15 1 5 Medium Energy High Energy Total Evts/kt/yr LE: 44 ME: 14 HE: 217 Low Energy 1 2 3 GNUMI V14 GEANT-FLUKA Target Model Physics results not updated. Still beam changes to put in...
Low Energy Beam Composition Low Energy Beam Low Energy Beam CC Evts/kt/yr/GeV 1 2 1 1 1-1 1-2.4 ν µ ν µ -bar 1 2 3 CC Evts/kt/yr/GeV 1 2 1 1 1-1 1-2.3 ν µ ν e +ν e -bar 1 2 3 ν _ µ /ν µ.3.2.1 (ν e +ν _ e )/ν µ.2.1 1 2 3 1 2 3
LE Beam Kaon Contribution ν µ CC Evts/kt/yr/GeV Fraction 1 2 1 1 1-1 1-2.3.2.1 Low Energy Beam ν µ νµ from kaons 1 2 3 1 2 3 ) CC Evts/kt/yr/GeV (ν e +ν _ e Fraction.4.3.2.1 1.75.5.25 Low Energy Beam ν e νe from kaons 1 2 3 1 2 3
Geometry Update 1 - Summer last year undertook comprehensive audit of beam line geometry to bring 'official' spectra inline with current design - Plot shows effect of each change applied cumulatively ν µ CC events 9 8 7 6 5 4 original new geom. target hall He Air 3.2 mm 6.4 mm target parahorn phorn Horn2 IC 2mm3mm Chase area reduced and distance to Soudan changed (both ~1% effects) Rest is due to increased decay pipe window thickness! 3 2 1 1 2 3 4 5 6 7 8 9 1
Decay Pipe Extension - ME and HE beams are not optimized for current SK and K2K oscillation parameters - Considered giving up optimal ME and HE beams in favor of more LE flux - ME and HE beams still possible via target shifts 15m 28m 45.28m Nominal L ong narrow extention Narrow extention Wide extention 112cm Extension option not taken Flexibility is good (off axis beams), Radiation issues with extension pipe, Major change to nearly complete design
Decay Pipe Extension and Window Options 1% Increase in rate possible with extended pipe and thin window Roughly 1/2 due to extension (35% RL of air) and 1/2 due to window (also about 35% RL of iron) Opted for thinner window w/o extension (M. Kostin)
Semi-Beams - Move Target but leave horn 2 in LE position - If extension option were taken these would be only ME and He options available - Got us to thinking about target motions. LE target already must move 1 m to be inserted into horn. Turns out 2.5 m of motion is not much harder..
Spectra as Function of Target Motion - Up to 2.5 meters possible - Allows experiment to tune beam quickly to place 'sweet spot' of beam in potentially interesting energy bin - Excellent monitoring tool. Muon rates in alcoves go up with energy. Possible to monitor and find error conditions in the beam line. - More handles on beam simulation. Tune MC to data taken during commissioning at various target positions
Other Games to Play During Commissioning Can we convince ourselves that we understand our beam during commissioning? Vary horn currents together (stat. errors for ~1 day of Near Detector) Events /.15 kt / 2.3E+17 POT /.5 GeV 3 25 2 15 1 5 I(1,2) = 2, 2 ka I(1,2) = 15, 15 ka I(1,2) = 1, 1 ka Events /.15 kt / 2.3E+17 POT /.5 GeV 3 2 1 I(1,2) = 2,2, t=inf I(1,2) = 2,2, t= I(1,2) = 15,15, t=inf I(1,2) = 15,15, t= I(1,2) = 1,1, t=inf I(1,2) = 1,1, t= 2 4 6 8 1 12 14 16 18 2 E n 2 4 6 8 1 12 E n Two curves compare current distribution on horn IC at t= and t=inf.
Comparisons of Near Spectra at Different Horn Currents Sensitivity to IC Current Distribution Sensitivity to IC Current Distribution I(1,2)=15,15 ka, t=inf I(1,2)=15,15 ka, t= I(1,2)=2, ka, t=inf I(1,2)=2, ka, t= I(1,2)=1,1 ka, t=inf I(1,2)=1,1 ka, t= I(1,2)=,2 ka, t=inf I(1,2)=,2 ka, t= Varied Currents/Nominal Currents 1.5 1.5 Varied Currents/Nominal Currents 1.5 1.5 2 4 6 8 1 2 4 6 8 1 2 4 6 8 1 2 4 6 8 1 (swapping in dummy load during commisioning is too hard so these are not possible)
Near - Far Spectrum Comparison MINOS Near and Far detectors are built to be a similar as possible - iron and scintillator thickness and spacing are same - average B field Neutrino flux as two sites is different π,k Decay Volume.75 km long ν Near Detector z=1.4 km Sees extended neutrino source Far Detector z=735 km Sees point neutrino source Predict far flux by extrapolating high statistics measurement at near detector predicted N(E) = FAR N(E) measured NEAR R(E) predicted FAR/NEAR point source: line source: R(E) predicted FAR/NEAR = Z 2 /Z 2 = 1.4 2 / 735 2-6 = 2x1 NEAR FAR predicted R(E) = FAR/NEAR exp(-z/γcτ)/(1/(z-z exp(-z/γcτ)/(1/(z-z far near ) 2)dz ) 2 )dz Ultimately need simulation of beam line to account for - production of particles in target - horn acceptances - beam line acceptances
Uncertainties Due To Hadron Production 1 6 CC Events/kt/year Difference from Ave. % 1 8 6 4 2 5 1 15 2 25 3 4 2-2 -4 GFLUKA BMPT MARS MALENSEK Absolute Rate GFLUKA BMPT MARS MALENSEK 5 1 15 2 25 3 1 6 Far/Near Difference from Ave. (%) 2 1.5 1.5 5 1 15 2 25 3 2-2 Far to Near Comparison GFLUKA BMPT MARS MALENSEK GFLUKA BMPT MARS MALENSEK 5 1 15 2 25 3 1 to 3% uncertainties in absolute rate 2-1% uncertainties in far to near comparison
ν µ CC Events/kt/yr/.5 GeV Components of LE Beam 1 7 1 6 1 5 1 4 Low Energy Beam, Near Detector OverFocus-Focus Focus-Neck Focus-Focus Neck-Neck Neck-Focus All Kaons 5 1 15 2
Far/Near By Track Type x 1-5.3 Low Energy Beam x 1-5.3 OverFocus-Focus Focus-Neck Focus-Focus Neck-Neck Neck-Focus All Kaons Near/Far.2.1 Near/Far.2.1 π lifetime 5 1 15 2 5 1 15 2
Target Model How many neutrinos come from primary interaction in target? Secondary? 1 9 Low Ener gy Beam - Far Det ec t or For LE beam full cascade is important Hadron production measurements beyond primary interaction needed ν µ CC- event s / kt / yr / GeV 8 7 6 5 4 3 2 t ot al s ec ondar y t er t i ar y >t er t i ar y 1 5 1 15 2 25 3 ( GeV)
Target Model - Currently different models of target are produced using weighting functions of p and p_t. Effects of target length, phi, etc. are ignored - Official neutrino spectra based on GEANT-FLUKA. Would like to update to better model. Candidates are: MARS FLUKA 'standard' at FNAL. Access to source code is difficult Widely used. Good agreement with data. Access to source code is not permitted DMPJET3 Widely considered best model at NuMI energies. Source freely available. But interface to tracking code (GEANT3) does not exist. Leaning towards MARS Have detailed simulation of target using MARS and have used it to produce results with GNUMI
Comparison of MARS Target Model with GFLUKA 1 8 x 1-5 Near/Far.2.1 CC events/kt/yr/gev 6 4 2 GFLUKA Target MARS Target (N/F MARS)/(N/F GFLUKA) 1.2 1.1 1.9.8 1 2 3 E n 1 2 3 E n 5 1 15 2 25 3 35 4 E (statistics were better with reweighting!)
8 7 ν µ CC E vents/kt/yr/gev 6 5 4 3 On axis 5 km off axis 1 km off axis 2 km off axis 2 1 2 4 6 8 1 12 14 16 18 2 (GeV ) ~1% measurements should be possible Detailed studies of detector technologies underway