Minutes from the MCNSI meeting, 21-22/2 2006 Participants: Risoe Kim Lefmann, Peter Willendrup, Peter Christiansen, Kaspar Klenoe HMI: Carlos Fehr NPI: Jan Saroun ILL: Emmanuel Farhi, Klaus Lieutenant ISIS: Dickon Champion PSI: Uwe Filges, Geza Szigmond FZJ: Alexander Ioffe SNS: Mark Hagen A. The group discussed the progress of MCNSI - The intercomparison on FOCUS needs still input from RESTRAX - IDEAS is not in use at SNS -> we will leave this package out of intercomparisons - Flux intercomparisons at IN14: VITESS and McStas are very close - Reflectivity samples are present in VITESS - Monte Carlo workshop will be held at PSI 3-6 Oct 2006 - The next MCNSI meeting must be held before the Sicily October meeting on FP7 application. Perhaps in Berlin June or September B. PKW presented the progress in McStas - Short update news on McStas 1.9: - PC has been employed to implement polarized neutrons - A prototype web-based simulation has been developed - Virtual experiments for planning, start with DMC, PSI - Teaching material for neutron outreach - New components - New tools - The teaching material has been extended - A component manual has been written - The next release of McStas, version 1.9.1. or 1.10, is planned to contain - Bug fixes - New components: Realistic detector, phase space transformer more realistic chopper - VRML - Language extension with new keywords: JUMP, WHEN, LOOP - Concentric component - Polarization components - New sources - 3D data - Frontends to TAS and TOF - Capability for grid simulations C. CF presented the status and future of VITESS - Present status - VITESS runs polarized neutrons - MacIntosh platform - 3D visualization - Simulations of EXCEED (HMI); NEAT (HMI) - Many modules - Future devlopments - VITESS publications will be placed on the home page - Benchmarking will be performed on new instruments and with user exp. - New graphics interface, fitting, manual/tutorial - New modules: Velocity selector, focusing optics - Optimization will be implemented - DIGEST options will input instrument description from file - Virtual experiments at NEAT D. JS presented the developments in RESTRAX - V. Ryukhtin works half time on the project; will finish tasks - Simulations of multichannel focusing benders - Simulations of doubly focusing monochromators - Added flexibility: - Displacement (strain scanner) - velocity (phase space transformer) - rotation - Development directions: - SIMRES documentation - NeXuS data format E. MH presented the general status of SNS and the local simulation efforts SNS is planned for 2 MW and has a construction budget of 1.4 G$. It is a short-pulsed source with a Hg target. At the same time, the ORNL reactor HFIR gets a new cold source. The neutrons programs of the two sources are now merged. The SNS construction project ends in June 2006 when a 20 kW pulse has been produced. The status is that the linac, the ring and the magnet system is OK. Also the target, Hg handling, and the moderators are ready. MH is constructing the single crystal thermal spectrometer HYSPEC. Computational details: - The estimated data generation will be 1 Tbyte per day. - All data will be stored in the NeXuS format - Data are counts plus "metadata": webcam, notebook, environment data... - Webaccess to all data - Software will be community code, e.g. Matlab via SNS license on Teragrid - SNS run McStas virtual experiments on Grid for - Experiment planning / proposals - on-the-fly diagnosis - Data analysis - typically 10 G rays in 5h -> virtual data -> analysis software - DANSE project has recieved 10 M$ funding (Brent Fulz is coordinator) F. KHK presented a simulation of the ISIS TS2 spectrometer LET LET is a cold-neutron high-resolution direct geometry chopper spectrometer, which has the possibility of using a mode with multiple repetitions within the same frame ("multi rep-rate"). The aim of this particular simulation was to investigate the effect of jitter in the fast choppers. It was found that a large amount of jitter (several us) can be tolerated without a loss in energy resolution or flux. G. UF presented the status of the intercomparison project with FOCUS at PSI - The guide system was simulated by RESTRAX, VITESS, McStas - In September 2005, the deviations were 20%, now they are below 2%, due to bug corrections and more accurate instrument descriptions. - The PSI cold moderator has been corrected in McStas - Comparison with gold foils are good: - at 54 m simulations are 4% too high - at 62 m simulations are 11% too high - McStas and VITESS are compared to the sample position. Presently, the deviations are 7% at the Fermi chopper position. H. GS presented the simulations for the UCN project. The main idea is to convert ultra-cold neutrons (UCN) to cold neutrons with a very narrow energy spread through phase space transformation. The UCN, 400-1000 A, will gather at the bottom of a vessel, trapped by gravity. - UCN facility at PSI: Neutron moderation with solid D2 at 6 K - VITESS simulations: multiple reflections, gravity, lifetime - Flux: 1.8 E6 n/sec - PST is a moving crystal - PG: resolution 0.3 meV, flux as FOCUS; will not be built - UCN for reflectivity spectroscopy, flux only 6 E3 n/s/cm2, res. 20 ueV - UCN for spin-echo spectroscopy could go to Fourier times of 50 us (!) I: AI presented the new development at FZJ/FRM2 and the simulation efforts. 8 FZJ instruments will be placed at FRM2 until 2008: - Backscattering spectrometer - Spin-echo spectrometer - Reflectometer - Diffuse scattering diffractometer - 3 SANS - Thermal time-of-flight spectrometer In the build-up there is a need to simulate focusing optics for the 4 FZJ simulators (using both McStas and VITESS): - sphericial/non-spherical surfaces - roughness - parabolic/elliptic guides - pulsed fields - He-3 spin filter - MgF2 neutron lenses S. Manoshin works for FZJ to write VITESS modules for this purpose. J. KLi presented the ongoing effort with validating McStas components Validation methods: - Internal (check code vs. analytical equations) - External (Consider component as black box) - Integrated (Insert component into instrument) Tested components: - V-selector (error found in one component) - Sources (no errors found) - Guide (gravity errors; corrected) - IN14 guide system: 1% difference McStas-VITESS Test plan: - Monochromators - Choppers - Momnitor_nD AI mentioned that the new FRM2 instruments could be used for comparisons, especially SANS and the double focusing diffuse scattering instrument. K. EF presented the inelastic multiple-scattering sample component for McStas. The sample component Isotropic_Sqw is capable of - Coherent and incoherent - Elastic-inelastic - Absorption - Multiple scattering - Input from Fullprof file, like PowderN - S(q,w) file for input (4 files produced; isotropic scatterers) - Concentric behaviour - Comparison virtual exp. with experiments (FOCUS) is almost OK: - contributions from coherent, incoherent, and multiples are separated - background is quantified - Can be used to test optimal sample size from multiple scattering - Can be used to test empty-cell background subtraction L. Discussion of milestones (KL) Instruments for code intercomparison and/or virtual experiments - Reflectometer and quasielastic (OSIRIS) has been done with VITESS - TOF seems OK (Exceed, LET); we may need NISP on board M. Discussion on simulations in FP7 - Case for funding - (EF) Guide systems are standard, well known. We need to cross the border to science and go harder into sample simulation. - (MH) This may end package development in 2-4 years?! - (JS) Development IS needed, users will have particular demands. - (EF) Yes, but that goes under "support for scientists" - (MH) We could have commitment from facilities - (??) We will coordinate existing activities (PSI, ILL, ISIS) which will come with matching funds (personnel/money/beamtime) - New ideas - We could go for a larger degree of support, less development. This would be useful for the JRAs - Instrument scientist training effort: "Road show of neutron simulation" - Simulations for teaching/outreach: NSF has given money, EU may be willing to do the same - Idea: www virtual neutron school under FP7 MCNSI ! TODO items: - KL will contact Phil Seeger on NISP code intercomparison