Wednesday 10 July 2013, Strathblane & Cromdale Halls, 16:30-18:30

advertisement
Wednesday 10 July 2013, Strathblane & Cromdale Halls, 16:30-18:30
Poster session B – Software and data analysis
P.221 VATES: Software for advanced visualisation and quantification of neutron scattering data
O Arnold1, T Perring2, M Reuter3, A Buts2, M Gigg1, J Zikovsky3 and N Draper1
1
Tessella, UK 2STFC Rutherford Appleton Laboratory, UK 3Oak Ridge National Laboratory, USA
The VATES project aims to provide advanced tools for visualisation and fitting parameterised models to large
neutron scattering data sets. Increasingly, a full understanding of the materials of interest to solid state physics,
chemistry and materials research requires the complete mapping of data in an n-dimensional manifold.
VATES is an on-going collaboration between ISIS at RAL and the SNS at Oak Ridge. The project has been run in
parallel to Mantid[1], our extensible framework for neutron and muon data reduction and analysis. The core of the
project has been heavily focused on creating and manipulating n-dimensional data, and extending the Mantid suite
of algorithms to operate on this new format.
The major user applications that have been created so far provide full 3D visualisation of manifolds within the ndimensions, and provide new opportunities for data discovery and reduction. By using ParaView[2] as the
visualisation engine, we are able to provide detailed and flexible representations of the data. Two-way interaction
between ParaView and Mantid allows us to view data as it exits Mantid, and also visually drive Mantid. The use of
the visualisation applications will be illustrated using examples of data from ISIS and the SNS.
The project focus is now towards supporting single-crystal elastic diffraction experiments and simulation and fitting
of resolution-broadened parameterised models of n-dimensional data, for which we have drawn on expertise and
proven tools from both ISIS and the SNS. The status of this part of the project will also be illustrated with examples
drawn from both facilities.
[1]
[2]
www.mantidroject.org
www.paraview.org
P.222 vTAS – a virtual assistance for planning neutron experiments
M Boehm, A Filhol, Y Raoul, M Johnson, J Kulda
Institut Laue-Langevin, France
vTAS is a Java based software package developed for the preparation phase of a neutron experiment on a triple-axis
spectrometer [1]. It essentially consists of two interactive graphical displays, one representing the scientific space
(reciprocal lattice, scattering triangle, instrument resolution ellipsoid) and one the corresponding real space
instrument configuration.
Making the best out of a neutron spectrometer in the measurements requires an understanding of the interaction
between the 4D resolution ellipsoid and the expected excitations of the system under investigation. The influence of
the various instrumental configurations on the shape and size of these ellipsoids is nowadays simulated by powerful
ray-tracing algorithms like McStas [2], Restrax [3] or Vitess [4]. Instead of implementing its own resolution
calculations, vTAS profits from these existing packages and establishes a dialogue for exchanging instrumental
parameters and simulation results. We present very recent progress in the synchronization between McStas and
vTAS.
[1]
[2]
[3]
Boehm et al., Nucl. Inst. Meth. A 697 (2013), p. 40.
P. Willendrup et al., Neutron News 350, (2004) e735–e737. http://www.mcstas.org
J. Saroun et al., Physica B 234-236 (1997) 1102-1104. http://neutron.ujf.cas.cz/restrax/ (2007).
ICNS 2013 International Conference on Neutron Scattering
[4]
http://www.helmholtzberlin.de/forschung/grossgeraete/neutronenstreuung/projekte/vitess/index_en.html
P.223 Bayesian model selection and parameter estimation of inelastic neutron scattering spectra
A De Francesco1 and L Scaccia2
1
CNR-IOM-OGG, France 2Università di Macerata, Italy
The extraction of physical quantitative information from inelastic Brillouin neutron scattering data is often a
challenging task, especially when dealing with specific samples or anytime the scattering signal is very weak. These
limiting conditions together with the effect of both instrument resolution and finite statistics make sometimes the
data analysis and the estimation of relevant dynamical parameters similar to the amazing job of a sculptor who
carves a beautiful and well defined figure from an unshaped block of marble. Moreover, when the collected spectra
are poorly structured, classical algorithms struggle to fit the data, easily fall in local minima and results strongly
depend on initial parameter values. Thus several solutions are apparently possible and the discernment of the
researcher plays a heavy role. In addition, measured spectra often comprise an unknown number of structures or
peaks of known parametric family. Choosing the number of such structures is an important issue, which is
commonly dealt with using various criteria, such as the AIC, the BIC or the Schwartz information criteria. These
criteria though fail to provide a probability measure associated with the number of peaks and sometimes different
criteria lead to different conclusions. We suggest a Bayesian approach to tackle these problems. Prior researcher
information on any parameter can be incorporated in a sensible and systematic way that will allow a better fit to the
spectra. Moreover results from this approach are in the form of probabilities, which one can use to compare the
different possibilities for the number of peaks characterizing the spectra.
P.224 New Rietveld refinement methodology using McStas virtual experiment models
F Emmanuel and R-C Juan
ILL, France
We present a new Rietveld methodology in which the instrument is modeled using a McStas diffractometer
description, rather than using a Caglioti based resolution estimate. The instrument model can be of any complexity.
Currently, the only requirement is to include the sample as an instance of the PowderN or Isotropic_Sqw
component, and a detector matching measured data, that is with a geometry reproducing the real set-up. The initial
sample structure can be set manually, atom per atom, or from a CIF, FullProf CFL/PCR or ShellX INS/RES/SHX file.
This structure is then converted, by making use of the CrysFML, into an intermediate powder or single crystal
reflection list used by the McStas sample component. There is no explicit analytical computation of the instrument
response, as it is implicitly convoluted with the sample as part the virtual experiment. As an example, the refinement
model can be created with the syntax:
>> model = rietveld('cif_file','mcstas_instrument_file');
Constraints can be assigned to structure and instrument parameters (fixed, bounded, crossed, ...), which are then
applied during the refinement procedure. This latter is initiated with the syntax:
>> structure = fits(model,'measurement');
where 'measurement' is a measured data set of any dimensionality matching the virtual experiment detector. The
refinement can be performed using any optimization method and any criteria. This new methodology may be
especially useful as a complement to conventional Rietveld tools in the case of close overlapping Bragg
spots/peaks, stretched and bent resolution functions, or highly asymmetric resolution profiles.
[1]
E. Farhi, et al, J. Neut. Res., 17 (2013) http://ifit.mccode.org
ICNS 2013 International Conference on Neutron Scattering
P.225 Validation of the MCNPX-McStas interface at the BOA beam line
U Filges1, E Rantsiou1, T Panzner1 and E Klinkby2
1
Paul Scherrer Institut, Switzerland, 2Technical University of Denmark, Denmark
Abstract unavailable
P.226 Interactive 3D for scientific applications
A Filhol1 and M Ippersiel2
1
Institut Laue-Langevin, France, 2IPTER, France
Interactively moving parts inside the 3D model of a congested technical environment with potential collision risks
(e.g. neutron strain scanner sample environment), interactively positioning parts inside a complex scientific model
with all the other parts reacting to collisions and readjusting themselves, these are examples of nowadays problems
for which no affordable solution was thought to exist.
In fact, most 3D engines are either limited to volumes into which one can freely navigate or too expensive or specific
(e.g. robotics or haptic interfaces).
We demonstrate the feasibility and usefulness of interactive 3D (3D objects that can be seized) with
Neutrons4Science. This multi-platform application (web, tablets, smartphones) implements a virtual Three Axis
Instrument (ThALES) whose parts can be moved with the fingertip inside a 3D scene that can also be rotated,
zoomed, panned.
This demonstrator could be easily expanded in many creative ways, e.g. with added external control (a) or used to
control the real world (b):
a)A virtual instrument could be controlled by the instrument computer thus providing affordable remote surveillance
and anti-collision monitoring.
b)A real instrument could be safely driven through the 3D interactivity of the virtual model owing to its anti-collision
capabilities.
However, there is more than what is shown by this simple demonstrator since VirtualGrip - the 3D engine we used also natively handles gravity, roughness, collision propagation, etc.
P.227 The nDDB project for a neutron dynamics data bank for biological macromolecules
V Garcia Sakai1, L Rusevich2, B Franzetti3, M Johnson4, F Natali4, E Pellegrini4, J Peters4, J Pieper5, M Weik3 and J
Zaccai4
1
ISIS, UK, 2Institute of Physical Energetics, Latvia, 3Institut de Biologie Structurale, France, 4Institute Laue-Langevin,
Grenoble, France, 5University of Tartu, Estonia
Neutron spectroscopy provides experimental data on time dependent trajectories, which can be directly compared
to molecular dynamics simulations. Its importance in helping us to understand biological macromolecules at a
molecular level is demonstrated by the results of a literature survey over the last two to three decades. Around 300
articles in refereed journals relate to neutron scattering studies of biological macromolecular dynamics, and the
results of the survey are presented here. The scope of the publications ranges from the general physics of protein
and solvent dynamics, to the biologically relevant dynamics-function relationships in live cells. As a result of the
survey we are currently setting up a neutron Dynamics Data Bank (nDDB) with the aim to make the neutron data on
biological systems widely available. This will benefit, in particular, the MD simulation community to validate and
ICNS 2013 International Conference on Neutron Scattering
improve their force fields. The aim of the database is to expose and give easy access to a body of experimental data
to the scientific community. The database will be populated with as much of the existing data as possible. In the
future it will give value, as part of a bigger whole, to high throughput data, as well as more detailed studies. A range
and volume of experimental data will be of interest in determining how quantitatively MD simulations can reproduce
trends across a range of systems and to what extent such trends may depend on sample preparation and data
reduction and analysis methods. In this context, we strongly encourage researchers in the field to deposit their data
in the nDDB.
P.228 How to make vortex lattice diffraction patterns prettier and more accurate with Bayes’ theorem, and why
Bayes is the future of neutron data analysis
A Holmes
University of Birmingham, UK
Magnetic flux lines, or vortices, in type-II superconductors form an ordered two dimensional lattice, which has been
studied extensively using small angle neutron scattering (SANS).
The two-dimensional nature of the flux lattice means that all relevant scattering information is concentrated about a
single plane in reciprocal space. Using Bayes’ theorem, we can use this qualitative prior information to significantly
increase the signal to noise ratio, and correct for the Lorentz factor in a natural way.
There is a lot to be gained by using these methods; neutrons are expensive, and we wish to extract the maximum
amount of information from them. In almost all cases significant prior information is available, both about samples
and instrumental characteristics. As computational resources become more powerful these approaches become
more and more feasible. I will argue for a generalised Bayesian framework for data analysis, not just in the area of
SANS.
P.229 Software development for the effective utilization of event-recording data at MLF, J-PARC
Y Inamura1, T Ito2, T Nakatani3 and J Suzuki4
1
J-PARC center, Japan Atomic Energy Agency, Japan, 2Comprehensive Research Organization for Science and
Society, Japan, 3Japan Atomic Energy Agency, Japan, 4High Energy Accelerator Research Organization, Japan
We report the current status of software development for the effective utilization of event-recording data on the
Materials Life Science Facility (MLF), the Japan Proton Accelerator Research Complex (J-PARC).
MLF adopted the event data recording methods on the data acquisition system (DAQ). In this method, the time and
position information of each neutron detection are recorded.Recently we developed successfully a DAQ electric
module TrigNET, which records common electric signals as events synchronized with neutrons. TrigNET will connect
with instrument devices and gather output signals from them, i.e. sample direction from goniometers, thermometers
for sample and phase information from choppers controller. By analyzing events from TrigNET, neutron events can
be filtered for each condition users want to see. This filtering method enables users to measure and analyze the
information of the sample under various outer conditions more easily, for example phenomena induced by magnetic
field or pressure and so on.
In this presentation, we report the current status of software development for the filtering of event data with
instrument and sample conditions.
ICNS 2013 International Conference on Neutron Scattering
P.230 New Rietveld program suited for processing patterns recorded with an inverse- space focusing neutron
diffractometer equipped with position sensitive detector
I Ionita1 and N C Popa2
1
Institute for Nuclear Research Pitesti, 2National Institute for Materials Physics Bucharest, Romania
The efficiency of the neutron spatial focusing high resolution powder diffractometer (SFHRPD) can be significantly
increased by using a linear position sensitive detector (PSD). The principle and the design of the original SFHRPD
described by Ionita et al. (1999) is based on a bent plate single crystal monochromator and a flat sample rotated
according to a certain function, being the Bragg angle, different from the traditional Bragg-Brentano law, imposing
modifications of the Rietveld program. Two of them concern the calculation of the irradiated sample volume and
transmission on the Bragg angle and the parameterization of the dependence of the peak shift on. The third one was
imposed by the inappropriateness for this instrumental geometry of the texture correction models. In place was
used the model of texture correction representation by generalized spherical harmonics as reported by Popa
(1992).
Using a PSD in place of a simple detector a supplementary problem appears for the Rietveld program because of a
very peculiar dependence of the instrumental line breadth on the scattering angle. The whole pattern is divided in
segments of lengths equal to length of PSD and for every segment the sample orientation is set to an angle such
that the ideal focusing is realized for the channel sitting the middle of the segment. The FWHM for every segment
was fitted by least square with a three degree polynomial as function of.
P.231 Teaching with the Virtual Neutrons for Teaching project
P Jensen1, L Udby1, J Bruun1, P K Willendrup2, H Schober3, J Neuhaus4, J SB Nielsen2, J Pulz4 and K Lefmann1
1
University of Copenhagen, Denmark, 2Technical University of Copenhagen, Denmark 3Institut Laue Langevin and
Université Joseph Fourier, France 4Technische Universitet München, Germany
Experimental neutron scattering is conventionally restricted to large-scale facilities. It is imperative to educate future
users of these facilities in order to get a better scientific output and to stimulate growth of the community. However,
students and future users alike can find it hard to locate material to prepare them on the subject of neutron
scattering, since very few of their home institutions teach on the topic.
The Virtual Neutrons for Teaching (VNT) project [1] aims to develop an online and freely accessible teaching system
to educate both students and scientists interested in neutron scattering and the use of the facilities. The neutron
scattering course at the University of Copenhagen has been running for several years, teaching master and PhD
students both theory, experimental methods, and simulation of neutron experiments. The last years, the course has
been using the newly developed VNT platform in the teaching, as a first test of the project. The platform consist of a
wiki e-book containing theory, a site with interactive problems with hints and solutions, and finally a web front-end
to the Monte Carlo simulation package McStas [2], with which the users can simulate typical neutron instruments.
This poster will give an overview of our initial results and experiences with using the platform, as well as give
information to initiate discussion on what could improve the VNT project.
[1]
[2]
Udby et al, Neutron News 24 (2013)
Lefmann et al, Neutron News 10 (1999)
ICNS 2013 International Conference on Neutron Scattering
P.232 Evaluation of preferred orientation using MTEX from multi-detector TOF neutron diffraction data collected on
GEM instrument at ISIS
W Kockelmann2, M Xie1, N Baimpas1 and A Korsunsky1
1
University of Oxford, 2Rutherford Appleton Laboratory, ISIS Facility, UK
The time-of-flight (TOF) neutron diffraction multi-detector instrument GEM at ISIS covers a large fraction of the 4π
solid angle, thus offering the possibility of collecting sufficient data for extracting the entire orientation distribution
function (ODF) from textured samples with a few sample rotations, or even from one exposure. Recently, two major
improvements have been made on GEM. Firstly, a four axis (three translations and one rotation) sample stage was
installed on GEM to make scanning measurements possible. Secondly, a novel data analysis procedure was
developed and commissioned. The well-established method of Rietveld refinement was used to analyse GEM data
and to extract texture information. However, this approach involves multi-parameter fitting of hundreds of spectra.
This makes the data processing slow and cumbersome for new users. In the present study, a different approach
based on a Matlab toolbox (MTEX) was employed. The method is based on the harmonics method and allows one to
estimate the ODF from raw pole figures. The data processing speed is massively improved, and the routine can also
be readily implemented for iterative improvement. A copper standard sample with known preferred orientation was
used to demonstrate the power of the approach. The comparison between MAUD Rietveld refinement and MTEX
harmonics method is discussed.
P.233 Highlighting the advantages of the 2D analysis in neutron scattering
A Len1, G Pépy1 and L Noirez2
1
Wigner Research Center for Physics, Hungary 2Laboratoire Léon Brillouin, CEA Saclay, France
While analyzing isotropic SANS patterns is quite common, the non centrosymmetric 2D scattering maps are often
treated as if they contained information about randomly oriented samples or by selecting one or more preferred
directions, separating one direction from the rest of the pattern. The oriented scattering objects appear in most of
the cases interesting to perform on them 2D data treatment. It is obvious that it is preferable fitting together data
lying in different directions, because of their different behavior. Eg. in case of the determination of the polymer form
factor, a series of difficulties would be caused in determining the cross sections at zero scattering vector, which is
linked to the polymer molecular mass independently of the orientation angle. Another reason is that if only 1D
fitting is used restriction to few pixels, where the intensity does not change versus the angle is necessary; therefore
most of the data cannot be used. Finally if a real 2D data treatment is available it is easy to expand its capacity to
treat several files together or to enhance and extend the advantages of a 2D treatment from small to large scattering
vectors (typically from 10-3 Å-1 up to 4Å-1), from conventional gas detectors to high resolution 2D scintillator neutron
devices (up to 1024 x 1024 cells). We shall show examples for "embedded" scattering intensities of a sample
obtained at different scattering vector ranges, fitted with one single model in 2D, examples of scattering patterns
with anisotropy changing with temperature, treated with the same model changing only one parameter, 2D data
treatment of a complex sample (rhombohedric nanochannels) observed at orthogonal angles, but treated with the
same model.
ICNS 2013 International Conference on Neutron Scattering
P.234 The improvement of the instrument control system toward 1MW in J-PARC/MLF
T Nakatani1, Y Inamura1 and T Ito2
1
J-PARC, 2 Comprehensive Research Organization for Science and Society, Japan
The proton beam power of J-PARC has been increased more and more in spite of the Great Earthquake. Now we can
experiment using the 300kW proton beam, during the several years, we will be able to use the 1MW beam. At that
time, our experiment method will have to change high quality as well as high quantity. We have been developed the
universal event recording system and the dynamic device control software. We have introduced the new DAQ board
called "TrigNET" into our DAQ system for the universal event recording. TrigNET can output the event data which
consist of the elapsed time from the trigger, the event source and the signal values. By using TrigNET, we can
measure the device signals, for example, the rotating pulse of the choppers, the phase of the neutron polarizer,
temperature, electric and magnetic fields, etc. Because TrigNET has the SiTCP interface, we can simultaneously
collect the measured data as well as the neutron event data by DAQ-Middleware which is the standard DAQ
software in J-PARC/MLF. The dynamic device control software can concurrently control the neutron optics and
sample environment devises as well as the DAQ in the instrument. The instrument control software framework called
"IROHA" integrates the DAQ and device control, but its architecture is sequential control. So, we improve IROHA to
be available dynamic device control. In this presentation, we will show the details and applications of this system.
P.235 A method to evaluate secondary extinction of pole figures measured by neutron diffraction
J Palacios Gómez, R Sebastián Salat Figols and T Kryshtab
Instituto Politécnico Nacional, Mexico
Secondary extinction could be present in measurements of pole figures of textured polycrystals. In order to evaluate
its influence, a method has been developed, to determine the main losses of intensity due to a second diffraction
process: The beam diffracted from the sample, is approximated as the primary beam incident on the same sample
for a new diffraction process. The sample has the orientation for measuring one point of the pole figure; secondary
extinction arises then from the scattering of this beam, which happens mainly as the Debye-Scherrer diffraction
rings. This beam does not reach the detector, and therefore is a measure of secondary extinction. The integrated
intensity of the Debye-Scherrer rings of the most intense possible reflections of this beam is then calculated. Each
point of a Debye-Scherrer ring, characterized by an angle omega, receives the intensity of crystallites oriented along
a curve in Eulerian space, determined through the Laue-conditions. The number of poles for each curve for every
value of omega, summed over the whole ring, is obtained from the CODF calculated by the method of three
measured pole figures, as an approximation to the true CODF. The integrated intensity of all Debye-Scherrer rings is
then calculated. The same procedure is applied to a powder, and the quotient of these quantities is evaluated,
corresponding to the measured pole density for the point under analysis of the pole figure. From this expression a
relation between measured and true pole densities arise. In this way, with the said approximations, a pole figure
can be corrected for secondary extinction. This method has been applied to verify that secondary extinction in
recrystallized copper is negligible.
P.236 Scatter: a neutron scattering simulation and analysis suite for GULP
D Roach1, M T Garba2, H Gonzalez-Velez3, J D Gale4 and D K Ross1
1
University of Salford, UK 2Robert Gordon University, UK 3NCI Cloud Competency Centre, 4Curtin University, Australia
A methodology has been developed for the direct interpretation and analysis of coherent inelastic scattering from
polycrystals; data that contains Q dependent information scattering analogous to dispersion curves in single
crystals. As such data has to be averaged over all possible orientations of the crystalline structure, it is difficult to
ICNS 2013 International Conference on Neutron Scattering
interpret directly and hence requires a combined modelling-experimental approach. Presented here is an overview
of the software suite developed to identify and interpret such data sets; the ‘Scatter’ code (a Fortran code) has been
built into the GULP software package that simulated one-phonon S(Q, ω) for single and polycrystalline ‘samples’
using the full semi-empirical modelling capabilities of GULP. This code has been fully parallelised and GPU
optimised and tested on architectures ranging from laptop and desktop workstations through cluster and large-scale
computing resources and is now cloud-capable, with new cost-function analysis that dynamically determines the
optimal distribution of calculations for cloud computing. Finally, the visualisation of data sets (and their essential
contribution to the analysis of such data) is explored, along with a python based ‘toolbox’ developed to provide a
user-friendly analysis environment compatible with the MANTID framework.
P.237 SpinW: Matlab library for spin wave calculations and more
S Tóth
Paul Scherrer Institut, Switzerland
SpinW is a set of Matlab functions that can perform classical Monte Carlo and spin wave calculations on interacting
magnetic moments. It is aimed at neutron scatterers and experimental scientists, who need an easy to use
modeling tool for magnetic properties. It allows a general spin Hamiltonian to be defined with arbitrary 3x3
interaction matrices between magnetic moments, single ion anisotropy and external magnetic field. Equivalent
interactions between pairs of magnetic atoms are identified based on interatomic distances, thus making the
definition of extended unit cells extremely simple. SpinW can calculate spin wave dispersions for arbitrary
commensurate magnetic structures [1,2]. It can also handle incommensurate structures using a rotating coordinate
system [3,4], with the only restriction that the moments have to be coplanar. Furthermore it can perform classical
Monte Carlo simulations and parallel tempering on large unit cells up to 105 spins on a desktop PC, these
calculations can be used to determine the magnetic ground state properties, e.g. magnetization, susceptibility,
specific heat and structure factor. The SpinW package is available under the GNU General Public License [5].
[1]
[2]
[3]
[4]
[5]
S. Petit, École Thématique De La Société Française De La Neutronique 12, 105 (2011).
S. R. White, M. Sparks, and I. Ortenburger, Physical Review 139, A450 (1965).
A. Chernyshev and M. E. Zhitomirsky, Physical Review B 79, 144416 (2009).
S. Toth, B. Lake, K. Hradil, T. Guidi, K. Rule, M. Stone, and A. Islam, Physical Review Letters 109, 127203
(2012).
https://wiki.helmholtz-berlin.de/spinw
P.238 E-learning large scale facility users of tomorrow
L Udby1, P Jensen1, J Bruun1, P K Willendrup2, H Schober3, J Neuhaus4, J S B Nielsen2, J Pulz4 and K Lefmann1
1
University of Copenhagen, Denmark, 2Technical University of Denmark, Denmark 3Institut Laue-Langevin &
Université Joseph Fourier, France 4Technische Universität München, Germany
Since experimental neutron scattering is mostly restricted to large-scale facilities, not all students have access to
learning the technique at their home institution.
Providing a freely accessible e-learning portal for neutron scattering is therefore an important outreach task in order
to secure and educate the future users and scientists at neutron scattering facilities. This task and challenge has
been taken up by the Integrated Infrastructure Initiative for Neutron Scattering and Muon Spectroscopy (NMI3)
which are financing the construction of the Virtual Neutrons for Teaching (VNT) web-portal [1].
The VNT portal is based on a new e-learning model which we are developing through co-financing from the
University of Copenhagen. The key point of the model is balancing the didactical triangle between topic, teacher
and students with careful consideration on the interactive web-tools to be used for the interaction between the three
ICNS 2013 International Conference on Neutron Scattering
parts. Particularly we are developing tailored interactive tools such as live-simulation, quizzes and exercises with
interactive hints and solutions for the students to interact directly with the topic of neutron scattering in the VNT
portal [2].
[1]
[2]
http://vnt.nmi3.org
L. Udby et al., Neutron News 24, pp 18- (2013)
ICNS 2013 International Conference on Neutron Scattering
Download