1. Marina Golubeva, Alexander Ivashkin Institute for Nuclear Research RAS, Moscow B+C@20 AGeV simulations with Geant4 and Shield Geant4 with Dpmjet-2.5 interface Comparison of Geant4+Dpmjet-2.5 and Shield results for Be7+Be9@80 AGeV Simulation of Be7+Be9@80 AGeV for thin (1.5%) and thick (10%) targets with Geant4+Dpmjet-2.5 NA61 collaboration meeting, CERN, June 6 - 10, 2011 Simulation of PSD response to light ion collisions at NA61 beam

2. B+C@20 AGeV simulations with Geant4 and Shield. Vacuum Carbon Target 5 cm PSD B ions The transport code in both simulations was Geant4 No magnetic field was applied Statistics for both simulations – 10000 events 17 m 2 Vacuum It was found that Geant4 do not produce events with low energy deposition in PSD (central events)

3. Energy deposition spectra. Geant4 and Shield comparison Geant4 without cut on PSD acceptance in PSD acceptance Shield 3 The shapes of the distributions are very different Geant4 does not produce central events with low energy deposition in PSD

4. Geant4 and DPMJET-2.5 interface http://www.astro.isas.jaxa.jp/conference/g4space5/slide/14/IonIonTokyo.pdf http://www.inta.es/g4suw2009/docs/Presentations/22_Fri_May_2009/G4SUWS_2009_Truscott_DPMJET.pdf DPMJET is a Monte Carlo model for sampling of hadron-hadron, hadron-nucleus, nucleus-nucleus and neutrino-nucleus interactions in the framework of the Dual Parton Model with emphasis as well in the central as in the fragmentation region in wide energy range from 5 GeV/n up to about 1000 TeV Versions of DPMJET are available in FLUKA Implementation of DPMJET-2.5 model in Geant4 DPMJET-2.5 FORTRAN code requires Glauber profile data: integral probability function used to sample impact parameter Current version of database of Glauber profile data available applicable to projectiles from AP = 2 to AP = 58 on target nuclei with AT = 2 to AT = 58 Total inelastic cross-section class G4DPMJET-2_5CrossSection created to do this, and also covers projectiles from A=2 to A=58 on targets from A=2 to A=58 (by-product of DPMJET-2.5) This has recently been extended to AP=210 and AT=210, as well as for AP=1 and AT=1 (several CPU years of simulation!!) and the simulation results are being processed Na61 contacted with CERN/PH-SFT Geant4 group and got the example how to use DPMJET-2.5 with Geant4 Now we also have possibility to use Geant4+Dpmjet-2.5 in VMC framework (thanks to Ivana Hrivnakova) We should ask about additional precomputed data files with AP and AT more than 58 4 Geant4 native nucleus-nucleus codes do not currently provide detailed spectra for projectile energy above ~5 GeV/nucleon C++ interface to DPMJET-2.5 (G4DPMJET-2_5Model class)

5. Deposited energy distributions. Comparison of Geant4, Shield and Geant4+Dpmjet-2.5 results Geant4 Shield Geant4 + Dpmjet-2.5 in PSD acceptance 5 Good agreement between Shield and Geant4+Dpmjet-2.5 results linear scale logarithmic scale

6. Be7+Be9@80 AGeV simulations with Geant4+Dpmjet-2.5 and Shield (vacuum in beam pipe) 6

7. Vacuum Be9 Target Be7 ions PSD 23 m Target thickness - 0.1956 cm (1.5% probability of inelastic interactions) The transport code in both simulations was Geant4 No magnetic field was applied Statistics for both simulations – 100000 events Be7+Be9@80 AGeV simulations with Geant4+Dpmjet-2.5 and Shield The comparison was done for PSD energy deposition distributions, for multiplicities and for p T and p Z distributions of protons, neutrons and different projectile fragments in PSD acceptance 7

8. Shield Geant4+Dpmjet-2.5 Multiplicity Energy deposition and multiplicity distributions Energy deposition Fragments with Z>2 Multiplicity alpha p n td He3 8 w/o cuts on energy E part > 50 GeV/n Cut E part > 50 GeV/nucl is used to select spectators

9. p T distributions Shield Geant4+Dpmjet-2.5 Fragments with Z>2 pnt dHe3alpha 9 Shield produces more nucleons, while Dpmjet-2.5 produces more fragments

10. p Z distributions Shield Geant4+Dpmjet-2.5 pnt d He3alpha Fragments with Z>2 10 The shapes are similar but the numbers are different We see that there is definite difference in number of produced nucleons and fragments between Geant4+Dpmjet-2.5 and Shield. Nevertheless, one can not declare that some model is better or worse a priory. But the difference is not crucial. Now it is the only tool we have for simulating ion-ion collisions at NA61 energies. (Dpmjet is one of the Fluka models)

11. Be7+Be9@80 AGeV (VACUUM) Geant4+Dpmjet-2.5 simulations for 1.5% and 10% targets 11

12. Vacuum Be9 Target Be7 ions PSD 23 m Target thicknesses - 0.1956 cm (1.5%) and 1.304 cm (10%) No magnetic field was applied Statistics – 100000 events Simulations with Geant4+Dpmjet-2.5 for 1.5% and 10% targets Energy deposition distributions and ratios of energy depositions in 10% and 1.5% targets at event-by-event basis were compared in PSD acceptance 12 The aim is to see distortion of PSD energy spectra due to more thick target and, respectively, to multiple interactions in target

13. Energy deposition spectra 1.5% and 10% targets 10% and 50% targets Ratio blue/red spectra 1.5% target 10% target 50% target w/o cuts on energy E part > 50 GeV/nucl 13 We do not see essential distortion of PSD energy spectra due to thick (10%) target E part > 50 GeV/nucl w/o cuts on energy

14. Ratios of deposited energy in 10% and 1.5% targets at event-by-event basis 14 The ratios confirm that basically we have one interaction in 10% target E tot / event < 260 GeV w/o cuts on energy E part > 50 GeV/nucl The ratios are done only for events that had the first interaction in 1.5% target (the worse case) Then all particles are transported through the rest part of 10% target E part > 50 GeV/nucl && E tot / event < 260 GeV

15. Be7+Be9@80 AGeV (HELIUM) Geant4+Dpmjet-2.5 simulations for 1.5% and 10% targets 15

16. Energy distributions (vacuum and helium) Simulations for vacuum and helium between target and PSD are done in the same conditions 16 10% target + vacuum 10% target + helium 1.5% target + vacuum 1.5% target + helium In comparison with thin target the effect of helium (interactions) is important. For thick target the effect is less visible w/o cuts on energy

17. Conclusions 17 A few simulations of PSD response to light ion collisions were performed Geant4 native models do not simulate properly nuclei interactions at high energies ( > 5 GeV/nucleon) Using of Geant4 with Dpmjet-2.5 interface improves the situation essentially The results from Shield and Geant4+Dpmjet-2.5 are in satisfactory agreement that allows the realistic simulation of the ion collisions The reaction Be7+Be9@80 AGeV was studied for a few cases: thin(1.5%) and thick (10%) targets, transportation in beam pipes with vacuum and helium No serious degradation in PSD energy spectra was found for thick target comparing with thin target The interactions in helium beam pipe are not essential for thick target These preliminary results allow to consider the use of thick target in the experiment Finally, we have the simulation tool Geant4+Dpmjet-2.5 (generator and transport) which describes realistically ion interactions (extension of Glauber profile database to more heavy nuclei is needed)

18. Thank you

19. Backup slides

20. The target thicknesses calculations: For Be nuclear interaction length is 42.1 cm, so for p+Be9 the 1.5% target will be 0.6315 cm. Inelastic cross-sections for p+Be9 (153.978 mb) and Be7+Be9 (497.06 mb) were calculated acording to formula: sigma^inel = 10 * pi * r0^2 * ( A^(1/3) + B^(1/3) - b * ( A^(-1/3) + B^(-1/3) ) )^2 with r0 = 1.3 fm, b = 0.93 The target 1.5% thickness for Be7+Be9 is equal then to 0.1956 cm. and 10% 1.30416 cm

21. Geant4+Dpmjet-2.5 simulations:  Geant4.9.4.p01  VMC 2.11  QGSP_BIC_HP (below 5 GeV/nucleon)  Dpmjet-2.5 (above 5 GeV/nucleon) QGSP_BIC and QGSP_BIC_HP QGSP is the basic physics list applying the quark gluon string model for high energy interactions of protons, neutrons, pions, and Kaons and nuclei. The high energy interaction creates an exited nucleus, which is passed to the precompound model modeling the nuclear de-excitation. QGSP_BIC and QGSP_BIC_HP - like QGSP, but using Geant4 Binary cascade for primary protons and neutrons with energies below ~10GeV, thus replacing the use of the LEP model for protons and neutrons In comparison to teh LEP model, Binary cascade better describes production of secondary particles produced in interactions of protons and neutrons with nuclei. Both lists, QGSP_BIC and QGSP_BIC_HP, also use the binary light ion cascade to model inelastic interaction of ions up to few GeV/nucleon with matter. The list QGSP_BIC_HP is like QGSP_BIC with the addition to use the data driven high precision neutron package (NeutronHP) to transport neutrons below 20 MeV down to thermal energies.

22. http://www.inta.es/g4suw2009/docs/Presentations/22_Fri_May_2009/G4SUWS_2009_Truscott_DPMJET.pdf Comparison for C+C @ 10 and 200 AGeV The same spectra calculated with Geant4+Dpmjet2.5 in VMC framework

23. Generated with Shield, 20000 events, without normalization Pb+C @ 158 AGeV Z distributions for Pb+C @ 158 AGeV Provided by Herbert