Presentation is loading. Please wait.

Presentation is loading. Please wait.

IMR-MIP DANSE: Distributed Data Analysis for Neutron Scattering Experiments Brent Fultz, California Institute of Technology, DMR 0520547 Goal: Discovery.

Similar presentations


Presentation on theme: "IMR-MIP DANSE: Distributed Data Analysis for Neutron Scattering Experiments Brent Fultz, California Institute of Technology, DMR 0520547 Goal: Discovery."— Presentation transcript:

1 IMR-MIP DANSE: Distributed Data Analysis for Neutron Scattering Experiments Brent Fultz, California Institute of Technology, DMR Goal: Discovery across the frontier of science and engineering, connected to learning, innovation and service to society. Indicator (I1): Enable people who work at the forefront of discovery to make important and significant contributions to science and engineering knowledge. Data Reduction for Inelastic Neutron Scattering J.Y.Y. Lin, M. McKerns, B. Fultz (California Institute of Technology) "Data reduction" is the essential step of transforming raw data of counts in neutron detectors to intensities in physical units such as barns, electron volts and inverse nanometers. Version 1.1 of the reduction software for inelastic neutron scattering was released in 2006, accommodating requests from users of version 1.0. A hierarchy of menus works well for the user interface of Reduction. These are generated in part with automated help, reducing development time considerably. To date, each instrument for inelastic neutron scattering has used highly customized data reduction software. While the software performs adequately, the variations in style and structure make it challenging for users to work with different instruments. Each software package must be maintained independently, and therefore is rarely updated to

2 IMR-MIP DANSE: Distributed Data Analysis for Neutron Scattering Experiments Brent Fultz, California Institute of Technology, DMR new types of user interfaces or graphics packages, for example. With the design of Reduction v.1.1, approximately 40 of 300 software classes are specific to a particular instrument, and many of these have only minor differences. Those classes specific to instrument configuration and geometry are designed to be used in Monte Carlo simulations of instrument performance, allowing better connection between theory and experiment. Finally, much of the user interface for Reduction was generated automatically, with guided discovery of variables in the code itself. This work is notable because: It provides a generalized software design that is readily extended to other neutron scattering instruments and is readily maintained. It is compatible with simulations used in computational neutron science. Other Indicators: T5: Support research that advances instrument technology and leads to the development of next- generation research and education tools

3 IMR-MIP DANSE: Distributed Data Analysis for Neutron Scattering Experiments Brent Fultz, California Institute of Technology, DMR Goal: Discovery across the frontier of science and engineering, connected to learning, innovation and service to society. Indicator (I1): Enable people who work at the forefront of discovery to make important and significant contributions to science and engineering knowledge. Prototype Science Application for Phonon Scattering O. Delaire, J.Y.Y. Lin, B. Fultz (California Institute of Technology) Neutron scattering was used to discover quantized vibrations in solids called "phonons," and B. Brockhouse won the Nobel Prize in Physics for this discovery in Inelastic neutron scattering remains an excellent way to measure the energies and wavelengths of the different phonons in solids. In recent years it has become possible to calculate phonons reliably using the quantum mechanics of the atoms in a crystal. In a prototype effort, we have used these ab-inito calculations to predict phonon scattering in a Monte-Carlo simulation of a neutron spectrometer. The results of these calculations were in the same format as real experimental data, and were analyzed by Reduction 1.1. In a second type of calculation, we have prototyped a standard calculation of vibrations of atoms on springs, optimizing the spring constants to fit the measured scattering. Successful results from the two methods are shown in the figure. Both methods are moving towards detailed design, allowing improvements in computational speed, flexibility for different crystal structures and greater ease of use.

4 IMR-MIP DANSE: Distributed Data Analysis for Neutron Scattering Experiments Brent Fultz, California Institute of Technology, DMR This work is notable because: It puts the theory of phonons and experimental measurements of them on the same computational footing, far simplifying comparisons. Other Indicators: T5: Support research that advances instrument technology and leads to the development of next-generation research and education tools.

5 IMR-MIP DANSE: Distributed Data Analysis for Neutron Scattering Experiments Brent Fultz, California Institute of Technology, DMR Goal: Discovery across the frontier of science and engineering, connected to learning, innovation and service to society. Indicator (I1): Enable people who work at the forefront of discovery to make important and significant contributions to science and engineering knowledge. Interaction of Alzheimer’s peptide amyloid b with lipid bilayer membranes M. Loesche (Carnegie Mellon University), M. Doucet, P. A. Kienzle (University of Maryland) Figure 1: Neutron reflectivity curves measured on a tethered lipid membrane with and without amyloid β. The results clearly show a fully reversible membrane thinning in the presence of amyloid β.

6 IMR-MIP DANSE: Distributed Data Analysis for Neutron Scattering Experiments Brent Fultz, California Institute of Technology, DMR Amyloid plaques associated with dead or damaged neurons are a characteristic property of patients suffering from Alzheimer's disease. A major component of these plaques is the amyloid β peptide (1-42) which plays a key role in the progress of the disease. Two leading hypotheses exist: either amyloid β forms ion channels in the membrane or amyloid β leads to membrane thinning and disruption. Neutron reflectivity, which is able to test both hypotheses by measuring membrane thickness and completeness in presence of amyloid β, favors the second hypothesis. The layer densities and thicknesses were obtained by simultaneously fitting of all datasets using DANSE reflectometry software. The DANSE project aims to exploit the natural parallelism of the genetic algorithm search strategy and improve the user interface so that results can be obtained more quickly and easily. This work is notable because: Improved fitting software allows us to more easily understand the processes involved in a disease important to an aging population. Other Indicators: T5: Support research that advances instrument technology and leads to the development of next-generation research and education tools.

7 IMR-MIP DANSE: Distributed Data Analysis for Neutron Scattering Experiments Brent Fultz, California Institute of Technology, DMR Goal: Discovery across the frontier of science and engineering, connected to learning, innovation and service to society. Indicator (I1): Enable people who work at the forefront of discovery to make important and significant contributions to science and engineering knowledge. Examining Complex Magnetic Structure using Polarized Neutron Reflectometry P. Kienzle, Z. Fu, W. Chen (University of Maryland) Figure 2: Spin valve Ta/Cu/IrMn/CoFe/Cu/CoFe/NiFe/Ta/Si reflectometry measurements and strength and angle of magnetization determined from fits. To appear in IEEE Transactions on Magnetics. Modern storage media rely on spintronics, in which the manipulation of electron spin controls the magnetic properties of materials. A typical device consists of layers of magnetic and nonmagnetic metals deposited to form a spin valve similar to those in the magnetic readers on computer harddrives. Understanding how the magnetic properties change within the device is important to developing readers for higher density storage materials.

8 IMR-MIP DANSE: Distributed Data Analysis for Neutron Scattering Experiments Brent Fultz, California Institute of Technology, DMR These systems are particularly complex, in our case involving eight layers of a few nanometers each with the exact thickness unknown. Polarized Neutron Reflectometry, which is sensitive to both structure and magnetism, indicates that the direction of the magnetic moment is changing within the sample. The layer thickness and magnetic properties were obtained by fitting using the DANSE reflectometry software. The DANSE project aims to provide models for patterned arrays, such as those being developed for magnetic random access memory devices, in addition to the uniform layers currently modeled. This work is notable because: We provide a better understanding of the details of the magnetic structure within electronic devices, which will help industry to increase the storage density for next generation harddrives.

9 IMR-MIP DANSE: Distributed Data Analysis for Neutron Scattering Experiments Brent Fultz, California Institute of Technology, DMR This work is notable because: We provide a better understanding of the details of the magnetic structure within electronic devices, which will help industry to increase the storage density for next generation harddrives. Other Indicators: T5: Support research that advances instrument technology and leads to the development of next-generation research and education tools.


Download ppt "IMR-MIP DANSE: Distributed Data Analysis for Neutron Scattering Experiments Brent Fultz, California Institute of Technology, DMR 0520547 Goal: Discovery."

Similar presentations


Ads by Google