240 likes | 350 Vues
Explore the latest features in McStas simulation software presented at the Berlin MC School 2013, including modernized build systems, metapackages for easy installation, new ESS sources, brilliance monitoring, interface with iFit, Python-based tools, Rietveld refinement methodology, and more. Collaborative developments on McStas-MCNPX interfaces and proof of concept solutions are highlighted, offering improved accuracy and efficiency in neutron scattering experiments.
E N D
New features in McStas • Peter Willendrup1,5, Emmanuel Farhi2, Erik Knudsen1,5, Emmanouela Rantsiou3,6, Kim Lefmann4,5 • 1Physics Department, Technical University of Denmark, Denmark2Calcul Scientifique, Institut Laue-Langevin, France3Laboratory for Developments and Methods, Paul Scherrer Institute, Switzerland4Niels Bohr Institute, University of Copenhagen, Denmark5ESS design update programme, Denmark6ESS design update programme, Switzerland New features in McStas - Berlin MC school 2013
Agenda • New features in McStas 2.0 • McStas 2.1 and beyond... • New project areas New features in McStas - Berlin MC school 2013
Modernized build/package system - McStas 2.0 New features in McStas - Berlin MC school 2013
Modernized build/package system - McStas 2.1 Metapackages!!! Easy to install on Linux systems via dependency-requirements, i.e. apt-get install mcstas-suite apt-get install mcstas-suite-perl apt-get install mcstas-suite-python Meta-package now defined on the Mac Will come for 2.1 on windows also... - will allow install without internet connection New features in McStas - Berlin MC school 2013
The 1.12c ESS source (released) • Either cold or thermal • Flat, rectangular source • 2001 cold and thermal brightness New features in McStas - Berlin MC school 2013 5
The 2.0 ESS source (released) • Cold and thermal moderators • Cylindrical cold source • Flat thermal wing • 2012 cold and 2001 thermal brightness • Hard to use / align for thermal instruments New features in McStas - Berlin MC school 2013 6
The 2.0a ESS source • Cold and thermal moderators • Cylindrical cold source • Flat thermal wings • Geometry close to that from MCNPX • 2012 cold and 2013 thermal brightness (thermal was just released by target??) • Easier to align for both cold, thermal and bispectral instruments • Later: space-dependent brilliance from the moderators New features in McStas - Berlin MC school 2013
The 2.0a source (in the pipe) New features in McStas - Berlin MC school 2013
Overview of full monolith, including in-pile optics, shielding etc. New features in McStas - Berlin MC school 2013
Brilliance_monitor output New features in McStas - Berlin MC school 2013
Easy way to get brilliance curves New features in McStas - Berlin MC school 2013
Interface to iFit New features in McStas - Berlin MC school 2013
McStas-MCNPX interfaces for beam losses New developments 1 - work on McStas-MCNPX interfaces Collaboration DTU Nutech DTU Fysik ESS New features in McStas - Berlin MC school 2013 16
Proof of concept solution: One detector geometry so far, yet general: Allows to define an IDF XML instrument description using a standard McStas. NeXus needs to be installed for producing NeXus output. New features in McStas - Berlin MC school 2013
New Python-based tools plus web frontend New features in McStas - Berlin MC school 2013 18
New Rietveld refinement methodology using McStas virtual experiment models E. Farhi and J. Rodriguez-Carvajal, ILL New features in McStas - Berlin MC school 2013
Science path: conventional Raw data Conventional Rietveld New features in McStas - Berlin MC school 2013 Picture from B. Fultz, DANSE
Science path: direct Raw data McStas with CrysFML calls Optimisation with iFit New features in McStas - Berlin MC school 2013 Picture from B. Fultz, DANSE
Implementation overview • We set-up a diffractometer model from its geometry. No 'effective' parameters. The model should contain a realistic detector geometry (incl. 3He gas). • We define the sample as part of the diffractometer description, with structural parameters as input (CIF/CFL/ShelX). • We get some measured data. • We refine the structural parameters on the measurement, using an ad-hocoptimizer. New features in McStas - Berlin MC school 2013
Some selected peaks That's not perfect, but a reasonable fit. Much slower than traditional Rietveld, but no a-priori assumptions about peak shape or resolution. New features in McStas - Berlin MC school 2013
People New features in McStas - Berlin MC school 2013