National Science Foundation (NSF) Scientific Software Innovation Institute for Advanced Analysis of X-Ray and Neutron Scattering Data (SIXNS)


Workshop

Theoretical software and analysis tools and software integration for scattering science 




Schedule

Friday, 1-17-2014

 Time  Topic  Location  Speaker
 7:45-8:30  breakfast  Pacific Dining Room  
 8:30-8:50  Opening remarks; Scientific Software Innovation Institutes as part of NSF's SI2 program  Cedar Room  J. Rehr; D. Katz
 8:50-9:05  Outcomes of SIXNS-I  Cedar Room  Brent Fultz
 9:05-9:20  Outcomes of SIXNS-II  Cedar Room  Simon Billinge
 9:20-9:30  Goals for SIXNS-III  Cedar Room  John Rehr
 9:30-9:50  XAS analysis and modeling of x-ray absorption experiments  Cedar Room  Matt Newville
 9:50-10:10  Analysis and modeling of electron scattering experiments  Cedar Room  Cecile Hebert
 10:10-10:30  break    
       
 10:30-10:50  The Materials Project  Cedar Room  Wei Chen
 10:50-11:10  The NWChem Platform  Cedar Room  Eric Bylaska
 11:10-11:30  Experiences w/ the UNEDF and PSI DOE SciDAC Computational Science Projects  Cedar Room  Kenneth Roche
 11:30-12:00  Advanced theory codes  Cedar Room  S. Louie, C. Draxl
       
 12:00-13:30  lunch  Pacific Dining Room  
       
 13:30-15:00  Breakout Sessions 1:  Breakout rooms  
   1.A Theory codes for Ground-State properties    
   1.B Hardware, infrastructure, and data    
   1.C Theory codes for neutron spectroscopy    
   1.D Theory codes for excited states, x-ray, and electron spectroscopy    
 15:00-15:30  break    
 15:30-17:00  Breakout Sessions 2:  Breakout rooms  
   2.A Analysis tools for x-ray and electron experiments    
   2.B Analysis tools for neutron experiments    
   2.C Integration of software workflows    
   2.D SIXNS institute and proposal    
       
 19:00-21:00  conference dinner  Ivar's Salmon House  



Saturday, 1-18-2014

 Time  Topic  Location  Speaker
 7:45-8:30  breakfast  Pacific Dining Room  
 8:30-10:00  Prepare reports for breakout sessions  Breakout rooms  
 10:00-10:30  break    
 10:30-11:20  Presentation and discussion of breakout reports 1.A-D (4x12 mins.)  Cedar Room  
 11:20-12:10  Presentation and discussion of breakout reports 2.A-D (4x12 mins.)  Cedar Room  
 12:10-12:20  Brief participant contributions (10x1 min.)  Cedar Room  
       
 12:20-13:30  lunch  Pacific Dining Room  
       
 13:30-14:50  Conclusions and SIXNS proposal  Cedar Room  
 14:50-15:00  Closing remarks  Cedar Room  John Rehr



Breakout topic descriptions:

SESSION 1

  • 1.A Theory codes for ground state properties. It is usually necessary to refine a model by optimizing atom positions, lattice constants, and defect structures. Ground-state densities and potentials are usually a necessary first step for calculating excited-states properties. The SIXNS institute therefore needs a repository of robust ground-state engines that together are capable of describing all relevant material classes.
  • 1.B Hardware, infrastructure, and data. What will be required of the SIXNS code in terms of handling data and properly utilizing the institute's hardware and infrastructure?
  • 1.C Theory codes for neutron spectroscopy. Inelastic neutron scattering is used to measure the dynamics of nuclei or spins. Phonon studies can benefit today from DFT methods because the electrons remain in the ground state throughout the scattering process. Here the question is why computational scattering science is not further along. For magnetic scattering studies, calculations on systems with spin disorder presents a challenge. A second issue is that much of the scattering research is on quantum magnetism in low dimensional materials, where ab initio methods are not available.
  • 1.D Theory tools for excited states, x-ray and electron spectroscopy. Generally, modeling of spectra requires theoretical tools that go beyond the approaches used for calculating ground state properties. The response of a material to a perturbation must be modeled carefully to the specifics of the excitation responsible for a given x-ray or electron spectroscopy. As no "perfect codes" exist that can reliably simulate all spectra of all materials, the SIXNS institute requires a portfolio of complementary codes that cover as much as possible of the range of spectroscopy experiments of interest.

SESSION 2

  • 2.A Analysis tools for x-ray and electron experiments. Experimentally measured data generally needs to be treated to remove undesirable contributions. Analysis requires comparison to theoretical reference data while fitting or extracting quantities of interest, while accounting for experimental conditions. The software tools for such analysis must be versatile yet user-friendly.
  • 2.B Analysis tools for neutron experiments. Analysis tools for neutron experiments. We will assume that the data are already "reduced," and corrected for most characteristics of the instrument. Further analysis today is usually done without reference to materials theory. How can our present analysis methods be adapted to better include the capabilities of computational materials science? What new workflows can be built? How should these be prioritized?
  • 2.C Integration of software workflows. The plethora of software elements must be combined into a versatile and intuitive workflow. This involves program control, data management, and user interfacing. Design decisions must be taken on interfacing codes and streamlining I/O.
  • 2.D SIXNS institute and proposal. The goal of the SIXNS workshops is a blueprint of a theoretical institute, and a matching proposal for this institute to be funded. In this workgroup we will bring together the work done at the three workshops and begin to integrate it into a blueprint and proposal.






For more information on the SIXNS workshop please contact:
Dr. Kevin Jorissen
University of Washington, 
jorissen@uw.edu



Acknowledgements:
- Thanks to NSF for funding to run this workshop. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.
- Thanks to University of Washington for hosting the workshop.