Data libraries as a collaborative tool across Monte Carlo codes.ppt

1、Data libraries as a collaborative tool across Monte Carlo codes,Maria Grazia Pia INFN Genova, Italy,Mauro Augelli1, Marcia Begalli2, Mincheol Han3, Steffen Hauf4, Chan-Hyeung Kim3, Markus Kuster4, Maria Grazia Pia5, Lina Quintieri6, Paolo Saracco5, Hee Seo3, Manju Sudhakar5, Georg Weidenspointner7,

2、Andreas Zoglauer81 CNES,France - 2 State University Rio de Janeiro, Brazil 3 Hanyang University, Korea 4 Technische Universitt Darmstadt, Germany 5 INFN Sezione di Genova, Italy 6 INFN Laboratori Nazionali di Frascati, Italy 7 MPE and MPI Halbleiterlabor,Germany 8 University of California at Berkele

3、y, USA,SNA + MC 2010 Joint International Conference on Supercomputing in Nuclear Applications + Monte Carlo 2010,Data libraries,Compilations of evaluated theoretical or experimental dataEssential tool for Monte Carlo simulation Simulation results are as good as the data on which they are based well,

4、 sometimes they can be worseLargely shared across different Monte Carlo codesCan be a powerful instrument for the community to study modeling systematic to exchange physics knowledge across Monte Carlo codesRelatively limited documentation of their validation in the literature Sometimes even of thei

5、r content,Ongoing activities,Validation of current data libraries EADL (Evaluated Atomic Data Library) EEDL (Evaluated Electron Data Library) ENSDF (Radioactive decay data) Evaluation of improvements to current data libraries Evaluation of atomic parameter collections Available in the literature New

6、 data libraries Proton and a ionisation Electron ionisationSoftware tools for data library management G3: Information Technology and its Applications II, Data management tools: computational evolutions and benchmarks,EADL Evaluated Atomic Data Library,For Z = 1-100, all subshells: Subshell data Numb

7、er of electrons Binding and kinetic energies Average radius Level widths Average number and energies of released electrons and X-rays Average energy left to residual atom Radiative and non-radiative transition probabilities,To date, the most complete collection of atomic parameters relevant to Monte

8、 Carlo transport,S. T. Perkin, et al.,Tables and Graphs of Atomic Subshell and Relaxation Data Derived from the LLNL Evaluated Atomic Data Library (EADL), Z = 1-100, UCRL-50400, Vol. 30, LLNL (1991),Atomic binding energies,Inner shells Evaluated through experimental measurements of characteristic X-

9、ray energies E = subshell binding energy difference In progress: evaluation through direct experimental dataOuter shells Ionisation energy (AKA ionisation potential): direct experimental measurements Evaluated indirectly through their effects on electron impact cross section calculations compared to

10、 experimental data,Sources,Data libraries/collections,EADL Geant4 Lotz Carlson MCNP, Penelope 2008 Geant4 Table of Isotopes 1978 EGS4, EGSnrc, Geant4 Table of Isotopes 1996 EGS5 X-ray Data Booklet,Experimental data,Review Des Lattes et al., X-ray transition energies: new approach to a comprehensive

11、evaluation, Rev. Mod. Phys. 75, 3599 (2003) NIST Physics Reference Data See references therein High precision evaluated experimental data Bibliography in forthcoming publication,KL3 transition,Full set of results in a forthcoming publication,Lotz,KN3 transition,EADL,Ionisation energies,Difference w.

12、r.t. NIST Reference (experimental) Data,Full set of results in a forthcoming publication,Effect of atomic parameters on ionisation cross sections,Comparison with experimental data,BEB cross section with EADL binding energies * with Lotz binding energies with EADL (inner shells) b.e. and NIST ionisat

13、ion energy,Significant effect of outer shell binding energies on ionisation cross section valueLimited effect of different electron kinetic energies,Full set of results and references to experimental data in a forthcoming publication,See I2: Low Energy Electrons and Photons, Design, development and

14、validation of electron ionisation models for nano-scale simulation,Statistical analysis to distill sound conclusions,EADL radiative transition probabilities,EADL Calculations based on Hartree-Slater method by Scofield GUPIX database (commercial) Calculations based on Hartree-Fock method by Scofield

15、Validation w.r.t. experimental data collected by Salem K and L transitions Experimental data span several decades Data quality is largely variable One can draw sound conclusions only based on rigorous statistical analysis,Hartree-Fock method produces significantly more accurate results,Further detai

16、ls in I2: Low Energy Electrons and Photons Conceptual challenges and computational progress in x-ray simulation,Inconsistencies in EADL,L2M1,L2M4,L3M1,L3M5,L3M4,EADL, Scofield Hartree-Slater,Discrepancies of EADL values w.r.t. experimental data were documented in R. D. Bonetto, A. C. Carreras, J. Tr

17、incavelli, G. Castellano,“L-shell radiative transition rates by selective synchrotron ionization”, J. Phys. B, vol. 37, pp. 14771488, 2004,EADL improvement,The current version of EADL dates back to 1991 It would profit from some rejuvenationState-of-the-art binding energies Other data collections ex

18、hibit better accuracy w.r.t. experimental data Are they complete? What effect on other electromagnetic calculations? State-of-the-art radiative transition probabilities Based on Hartree-Fock calculations We could retrieve Hartree-Fock calculations in the literature only for a small set of transition

19、s Are further theoretical calculations needed? V&V prior to release,EEDL Evaluated Electron Data Library,Elastic scattering Ionisation Bremsstrahlung Excitation,Cross sections Spectra Angular distributions Average energy to the residual atom,Z = 1-100 Electron energy: 10 eV 100 GeVData from theoreti

20、cal calculations S. T. Perkins, D. E. Cullen and S. M. Seltzer, Tables and Graphs of Electron-Interaction Cross Sections from 10 eV to 100 GeV Derived from the LLNL Evaluated Electron Data Library (EEDL), Z = 1-100, UCRL-50400, Vol. 31, LLNL(1991),Validation of EEDL ionisation cross sections,Limited

21、 validation of EEDL cross sections in the literature “Comparing our K-shell ionization cross section to the compiled experimental data of Tawara and Kato and Long et al., we find that the uncertainties in our data appear to be 30 to 40% at low energies. The error might be as large as a factor of 2 a

22、bove several hundred MeV for intermediate-Z elements because of the uncertainty in the onset in energy of the density effect. Errors in the outer subshells are probably at least as much.”Systematic validation of EEDL ionisation cross sections w.r.t. a large collection of experimental data 181 experi

23、mental data sets for 57 atomsEvaluation of possible alternatives in the low energy range (below 1 keV),Further details in I2: Low Energy Electrons and Photons Design, development and validation of electron ionisation models for nano-scale simulation,Results,% elements for which EEDL is compatible at

24、 95% CL with at least one data set,SystematicsDependence of the results on the type of measurementSingle ionisationAbsolute measurement,First validation of EEDL ionisation cross sections below 1 keV,New data library for electron ionisation cross sections,New data library in progress Electron impact

25、ionisation cross sections at low energies (1 keV) Binary-Encounter-Bethe model Deutsch-Mrk modelSupported by extensive experimental validation Further details in I2 Low Energy Electrons and Photons: Design, development and validation of electron ionisation models for nano-scale simulation Paper in p

26、reparation,Example of data library values compared to EEDL and Penelope electron impact cross sections,New data library for proton/a ionisation cross sections,K and L shell ionisation Theoretical models ECPSSR model and variants Hartree-Slater corrections “united atom” approximation specialized for

27、high energies PWBA Empirical models Paul and Sacher Paul and Bolik Kahoul et al. Miyagawa et al. Orlic et al. Sow et al.,Example: K shell ionization by proton impact on Cu,RSICCs collaboration is kindly acknowledged,Review of radioactive decay data,ENSDF Evaluated Nuclear Structure Data Files half l

28、ives decay branches energy levels and level intensities of the decaying nucleus The current database used by Geant4 does not include references to the origin of the individual datasets or their actuality Work in progress to compare Geant4 radioactive decay datasets and the current version of ENSDF F

29、urther experimental activity to validate the simulation,Further details in F4: Monte Carlo Applications III (Others) Radioactive decay simulation with Geant4: experimental benchmarks and developments for X-ray astronomy applications,Conclusions,Significant investment in validating improving extending physics data libraries for Monte Carlo simulationThe Monte Carlo community would profit from a collaborative common effort to review existing data libraries to create new data libraries, to facilitate exchanges of physics models across Monte Carlo codes,