Particle production vs energy M Bonesini Sezione INFN

Particle production vs energy M. Bonesini Sezione INFN Milano Bicocca, Dipartimento di Fisica G. Occhialini

Outline • Targetry for Nufact – HARP • Large Angle Data analysis • Comparison with MC simulations • Targetry for conventional neutrino beams – HARP for K 2 K, MINIBoone – NA 56/SPY for WANF, CNGS, Nu. MI • Targetry for EAS and atmospheric neutrino • Future experiments • Conclusions

Towards a Neutrino Factory: the challenges • Target and collection (HARP/MERIT) – Maximize + and - production – Sustain high power (MW driver) – Optimize pion capture INTENSE PROTON SOURCE (MW); GOOD COLLECTION SCHEME • Muon cooling (MICE) – Reduce +/ - phase space to capture as many muons as possible in an accelerator • Muon acceleration – Has to be fast, because muons are short-lived ! 3

Why dedicated Hadroproduction expts: conventional neutrino beams Ingredients to compute a neutrino flux : (and k) production cross section (use same target and proton energy than proton driver of the experiment) Reinteractions (take data with thin and thick target)) All the rest: Simulation of the neutrino line: An “easy” problem. 4

Simulation of neutrino beams 1. Primary target production 2. re-interactions in target 3. re-interactions in beamline Full Montecarlo simulation (MARS, FLUKA, Geant 3 or 4) Fast simulation (parametrization of hadron production data, re-int models) Good for study of systematics Good for beamline optimization

Available data for simulations of beamlines • Low energy beams (Nu. Fact, K 2 K, Mini. BOONE …); mainly HARP • High energy beams (WANF, CNGS, Nu. MI, …): NA 20, NA 56/SPY and coming soon MIPP, NA 61/SHINE • In addition a lot of old not-dedicated hadron production experiments, mainly with big systematic errors and poor statistics q I will speak mainly of HARP (with an detour on NA 56/SPY): see M. G. Catanesi’s talk for the others

Physics goals of HARP 2000 – 2001 Installation 2001 - 2002 Data taking • Input for prediction of neutrino fluxes for the Mini. Boo. NE and K 2 K experiments • Pion/Kaon yield for the design of the proton driver of neutrino factories and SPL- based super-beams • Input for precise calculation of the atmospheric neutrino flux and EAS Systematic study of hadron production: Beam momentum: 3 -15 Ge. V/c Target: from hydrogen to lead • Acceptance over full solid angle • Final state particle identification • Input for Monte Carlo generators (GEANT 4, e. g. for LHC or space applications)

Targetl Target length ( %) Beam Momentum (Ge. V) 2 (2001) ± 3 ± 5 ± 8 ± 12 ± 15 #events (Mevts) Be C Al Solid targets Cu Sn Ta 5 100 For negative polarity, only 2% and 5% Pb K 2 K Al Mini. Boo. N E Be Cu “button” Cu Cu “skew” Cu 5, 50, 100, replica Harp detector layout and data taken. Barrel spectrometer (TPC) + forward spectrometer (DCs) to cover the full solid angle, complemented by PID detectors +8. 9 15. 27 22. 56 1. 71 2 +12 1. 69 6 cm ± 3 ± 5 ± 8 ± 12 ± 15 58. 43 H 2 18 cm ± 3, ± 8, ± 14. 5 13. 83 H 20 10, 100 +1. 5, +8(10%) 9. 6 08 D 1 H 1 Water +12. 9, +15 N 7 Cryogenic targets 233. 16 8

factory design • • maximize +( -) production yield as a function of: • • proton energy target material geometry collection efficiency (p. L, p. T) but different simulations show large discrepancies for production distributions, both in shape and normalization. Experimental knowledge is rather poor (large errors: poor acceptance, few materials studied) aim: measure p. T distribution with high precision for high Z targets

HARP Large Angle Analysis Beam momenta: 3, 5, 8, 12 Ge. V/c Data: 5% I targets Be, C, Al, Cu, Sn, Ta, Pb TPC tracks: >11 points and momentum measured and track originating in target PID selection Corrections: Efficiency, absorption, PID, momentum and angle smearing by unfolding method Backgrounds: secondary interactions (simulated) low energy electrons and positrons (all from 0) predicted from + and - spectra (iterative) and normalized to identified e+-. Full statistics analysed (“full spill data” with dynamic distortion corrections) although no significant difference is observed with the first analysis of the partial data (first 100 -150 events in the spill). 10

The Target/TPC Region target MWPCs TPC readout connectors beam HALO veto RPC modules 11

Spectrometer performance momentum resolution momentum calibration: cosmic rays elastic scattering: absolute calibration of efficiency momentum angle (two spectrometers!) PID: d. E/dx used for analysis TOF used to determine efficiency -p PID with d. E/dx -e PID with d. E/dx

The two spectrometers match each other 13

9 angular bins: p-Ta + Pion production yields p forward 0. 35 < q < 1. 55 backward 1. 55 < q < 2. 15

p-Ta Pion production yields forward 0. 35 < q < 1. 55 backward 1. 55 < q < 2. 15

Neutrino factory study + p+ yield/Ekin Cross-sections to be fed into neutrino factory studies to find optimum design: Ta and Pb x-sections at large angle (see Eur. J. Phys C 51 (2007) 787)

Comparisons with MC Many comparisons with models from GEANT 4 and MARS are being done, starting with C and Ta Some examples will be shown for C and Ta Binary cascade Bertini cascade Quark-Gluon string models (QGSP) Frittiof (FTFP) LHEP MARS Some models do a good job in some regions, but there is no model that describes all aspects of the data

3 Ge. V/c p-Ta +/- 18

8 Ge. V/c p-Ta +/5% target MODELS

8 Ge. V/c p-C +/5% target MODELS

Comparison with MC at Large Angle 1. Data available on many thin (5%) targets from light nuclei (Be) to heavy ones (Ta) 2. Comparisons with GEANT 4 and MARS 15 Monte. Carlo show large discrepancies both in normalization and shape – Backward or central region production seems described better than more forward production – In general + production is better described than production – At higher energies FTP models (from GEANT 4) and MARS look better, at lower energies this is true for Bertini and binary cascade models (from GEANT 4) – Parametrized models (such as LHEP) have big discrepancies – CONCLUSIONS: MCs need tuning with HARP data for pinc<15 Ge. V/c

beams flux prediction • Energy, composition, geometry of a neutrino beam is determined by the development of the hadron interaction and cascade needs to know spectra, K/ ratios • K 2 K : Al target, 12. 9 Ge. V/c Al targets 5%, 50%, 100% (all pbeam), K 2 K target replica (12. 9 Ge. V/c) Oscill. MAX special program with K 2 K replica target M. G. Catanesi et al. , HARP, Nucl. Phys. B 732 (2006)1 M. H. Ahn et al. , K 2 K, Phys. Rev. D 74 (2006) 072003. • Mini. Boo. NE: Be target 8. 9 Ge. V/c M. G. Catanesi et al. , HARP, Eur. Phys. J. C 52(2007) 29 Be targets: 5%, 50%, 100% , Mini. Boone target replica 0 0. 5 1. 0 Beam MC 1. 5 2. 0 2. 5 E (Ge. V Beam MC confirmed by Pion Monitor Precise p. T and p. Lspectra for extrapolation to far detectors and comparison between near and far detectors 22

HARP forward p K TOF for p=2+-0. 25 hadrons electrons Ncherenkov for p below pion threshold Calorimeter E/p and E(1 st layer)/E for p above pion threshold 23

PID performance 0 e 1 2 3 4 5 6 7 CERENKOV CALORIMETER p TO F k TOF CERENKO V TOF pions CERENKOV protons: 1 -2% 24

HARP Be 5% 8. 9 Ge. V/c Results 0. 75<p<5 Ge. V/c 30<theta<210 mrad relevance for Mini. Boo. NE HARP results (data points), Sanford-Wang parametrization of HARP results (histogram) 25

HARP 12. 9 Ge. V/c p+Al Results HARP in black, Sanford-Wang parametrization in red Sanford-Wang parametrization HARP data used to: in K 2 K and Mini. Boo. NE beam MC Translate HARP pion production uncertainties into flux uncertainties Compare HARP results with previous results 26

p+Al versus GEANT 4 27

p+Be versus GEANT 4 28

A small detour: the NA 56/SPY experiment at SPS Ø Measure p, kaon fluxes by 450 Ge. V/c p on Be ( 5% precision) ->knowledge of neutrino spectra Ø Measure k/p ratio (3% precision) -> knowledge ne/nm ratio v Equipped H 6 beam from NA 52 experiment in North Area Ø Available results were parametrized v Primary p flux measured by (BMPT parametrization) or used to tune SEM available MC (such as Fluka). Used for v Different Be targets (shapes, L) the study of available high-energy v PID by TOF counters (low neutrino beamlines: WANF at SPS, momenta) and Cerenkov (high CNGS, Nu. Mi momenta) 29

An application to NUMI (from M. Messier et al. ) ØComparison BMPT, Mars, GFLUKA in Minos near/far detecor 30

Atmospheric flux • Primary flux (70% p, 20% He, 10% heavier nuclei) is now considered to be known to better than 15% (AMS, Bess p spectra agree at 5% up to 100 Ge. V, worse for He) • Most of the uncertainty comes from the lack of data to construct and calibrate a reliable hadron interaction model. Model-dependent extrapolations from the limited set of data leads to about 30% uncertainty in atmospheric fluxes cryogenic targets (or at least nearby C target data) primary flux N 2, O 2 hadron production decay chains • • 31

Extended Air Showers Primary particle p + p 0 p e+ p p+C 0 n e- + e+ e target incoming protons and pions spectra: and - + Several targets + Forward direction + Relevant energy range: 10 -400 Ge. V

Daughter energy Hadron production experiments 10 Population of hadronproduction phase-space for p. A → πX interactions. NA 56/SPY Atherton et. al. Barton et. al. 1 Te. V Serpukov νμ flux (represented by boxes) as a function of the parent and daughter energies. Allaby et. al. 100 Eichten et. al. Measurements. Cho et. al. Abbott et. al. 1 -2 p. T points 3 -5 p. T points >5 p. T points But with different targets (mainly Be) 10 1 Ge. V 10 100 1 Te. V 10 Parent energy HARP 33

Model comparison: HARP p+C→ ++X 34

Model comparison: p+C→ +X 35

+C @ 12 Ge. V/c (lower statistics) • stat error ~ 30 -40 % • syst error ~ 10% 36

+C @ 12 Ge. V/c (high statistics) Stat error ~ 10% Syst error ~ 10% 37

Measurements with N 2, O 2 cryogenic targets Shape looks similar =>may use simpler C target data (solid, not cryogenic target) 38

Comparison with GEANT 4 preliminary 39

Covered phase space region • New data sets (p+C, +C and +C, p. O 2, p. N 2 at 12 Ge. V/c) • Important phase space region covered • Data available for model tuning and simulations • Results on N 2 and O 2 data are preliminary [Barton 83] Phys. Rev. D 27 (1983) 2580 [NA 49_06] Eur. J. Phys. , hep-ex/0606028 HARP (Fermilab) (SPS) (PS) 40

Data with incident Just an example for FW production HARP paper in preparation • All thin target data taken in pion beams also available • Interesting to tune models for re-interactions (and shower calculations in calorimeters etc. ) 41

Next measurements/analyses Energy range and phase space of interest Ebeam 8 -1000 Ge. V p 0. 5 -11. 0 Ge. V/c q 0 -300 mrad p+C, +C and K+C @ 20, 60, 120 Ge. V/c p+C and +C @ 30, 40, 50, 158 Ge. V/c p+C and +-C @ 3 -15 Ge. V/c N 2, O 2 targets 42

Summary • HARP has provided results useful for conventional n beams study, n factory design, EAS, atmospheric n studies and in addition for general MC tuning (Geant 4, FLUKA …) with full solid angle coverage, good PID identification on targets from Be to Pb at low energies (< 15 Ge. V) with small total errors (syst+stat < 15 %). About 10 physics paper published or submitted • More HARP results coming : forward production with incident pions, protons on Be to Ta targets; production with long targets, … • Comparison with available MC show some problems 43