| research | data | how to reach us | in the news | pointers |

from the MSEL Annual Report 1997:  

THE CENTER FOR THEORETICAL AND COMPUTATIONAL MATERIALS SCIENCE

The interdisciplinary field of materials science presents researchers with rich, complex problems often of a universal nature. The theoretical underpinnings are relatively undeveloped, partly because it is a new field of research, and partly because the problems are quite complex. Indeed, these problems are often far too complicated to solve with pencil and paper. However, the revolution of fast information access, high speed computation, and widespread availability of computational resources has made possible new approaches to research in materials science and engineering. For example, previously intractable problems have become manageable through numerical simulation. Computer experiments allow fast, inexpensive, and nondestructive prediction of materials properties and behavior. It is now possible to study the physics of materials on all length scales using computer simulation, allowing researchers to verify and improve theories and propose more realistic models. Visualization of microscopic and mesoscopic phenomena give researchers new insights not previously possible through traditional experiments. These insights often lead to the discovery of new phenomena, and can serve as guides to experiments. With the advent of these new techniques NIST is in a unique position to bring together industry, academia, and government labs to form research collaborations. Thus, in 1994 MSEL created the Center for Theoretical and Computational Materials Science (CTCMS) to further computational materials science and promote its application to industrial problems.

The CTCMS is now a recognized national center of expertise in computational materials science. The Center's mission is to:

• investigate important problems in materials theory and modeling using new computational approaches;

• create innovative and productive opportunities for collaboration in materials theory and modeling;

• develop powerful new tools for materials theory and modeling and accelerate their integration into industrial research.

FY 1997 ACCOMPLISHMENTS

INFRASTRUCTURE

CTCMS members include researchers from industry, universities, national labs, and the NIST labs. Within MSEL, each CTCMS researcher is a member of one of the five MSEL divisions. Consequently, the research agenda of the CTCMS is coupled to division agendas. The scientific diversity of CTCMS researchers encourages communication between mathematicians, physicists, chemists, and materials scientists and engineers, and facilitates interactions and collaborations on a wide range of problems in materials and materials processing.

The CTCMS consists of a central hub in MSEL, and partnerships with other NIST labs, industrial labs, universities, and national labs. In FY 1997, the CTCMS expanded its network of powerful workstations and multi processor computers with advanced graphics, computing power, and data storage. The facility provides computational support, audio/visual capabilities, and teleconferencing for CTCMS working groups and workshops. The CTCMS maintains a World Wide Web server (http://www.ctcms.nist.gov) that provides information on Center activities, research results, and software. Each CTCMS working group has a Web page which describes its research agenda, publications, meetings, and advances. In some cases, the CTCMS server has become a tool for confidential information exchange among the participants of a working group. New features, including a web-based calendar and search engine, were implemented.

WORKING GROUPS

The CTCMS forms short-term multidisciplinary and multi-institutional research teams to attack key problems in materials and materials processing. As a first step, CTCMS holds workshops to (1) define technical research areas with significant technological impact, (2) identify team members, and (3) design and build the infrastructure for collaborative research.

This year CTCMS working groups added new members, held group meetings and new workshops, and made numerous, significant advances:

MICROMAGNETIC MODELING

In the area of micromagnetics, modeling efforts are hampered by the fact that different computational methods give different solutions to elementary micromagnetic domain-structure problems. Today's fast workstations now make standardization of methods and solutions possible, and easily accessible to the community. The CTCMS Micromagnetic Modeling group was formed in FY 1996 to address the need for accurate, standardized micromagnetic modeling tools.

In FY 1997, CTCMS co-sponsored researchers from MSEL's Metallurgy Division, NIST's Information Technology Laboratory, IBM, Kodak, and a number of universities continued their efforts to develop interactive computational tools for micromagnetic modeling, provide solutions to standard problems, and conduct benchmark verification. This past year, seven submitted solutions to the first mMAG standard problem were collected and displayed for comparison, both on the CTCMS web page, and at a major international magnetism conference. This was the first time that the solutions of independently developed micromagnetic codes had ever been directly compared, and the results were alarmingly different. As a result, preparations are underway for a second standard problem geared toward finding the borderline conditions where solutions start to deviate.

Significant progress was made in the design and implementation of micromagnetic code for public distribution. The software has been designed to allow the maximum amount of flexibility and modularity for comparing and updating computational techniques. The initial release of mMAG software, "mmdisp," displays and manipulates micromagnetic spin configurations. In January 1998 the group plans to release a full micromagnetic code, "Object Oriented MicroMagnetic computing Framework"(OOMMF). These public domain software tools are available through the CTCMS web server.

GREEN'S FUNCTION LIBRARY

Industrial designers and computational engineers need to reduce the time-cycle of the component design process to remain competitive in today's marketplace. An electronic database of region-dependent Green's functions would allow rapid design modifications of complicated material geometries and reduce the time cycle by weeks. Data storage and input/output capabilities of current computers now make such modern computational tools possible.

The CTCMS Green's Function Group formed following a CTCMS-sponsored workshop in late 1994 on Green's Functions and Boundary Element Analysis for Mechanical Properties of Advanced Materials. Co-funded by CTCMS and the National Science Foundation, the group formed partnerships between researchers at several universities, NIST, PDA Engineering, Gates Rubber Company, Ford, and Motorola. Gates Rubber Company dedicated an employee full-time to this project. Publications from the group can be found on the CTCMS Web Server. The workshop report is available to the public as a NIST special publication.

The group refocused its goals for FY 1997 to develop an interactive, electronic library tool consisting of Green's function and boundary element solutions for standard materials geometries arising in elastostatics, elastodynamics, acoustics and ultrasonics. Several Greens functions were constructed this past year. This tool will be available on the CTCMS web server in FY 1998.

PATTERN FORMATION IN MULTIPHASE LIQUID-CRYSTAL/POLYMER MATERIALS

The performance and properties of devices made from multiphase polymer/liquid crystal materials depend crucially on the microstructure that emerges during processing. Present understanding of the nonequilibrium physics of morphogenesis in these materials is weak. Numerical simulations that implement newly-developed theories of these materials and compare morphologies with experimental morphologies can improve processing and design.

In FY 1997, the CTCMS continued its collaboration with the NSF Science and Technology Center for Advanced Liquid Crystalline Optical Materials (ALCOM) in Ohio to model the kinetic and morphological phenomena arising during processing in these materials. NIST principal investigators include members of the Polymers Division and NIST's Information Technology Laboratory. General Motors, Raychem, IBM, DPix, and Phillips are regular workshop participants.

With sponsorship of the CTCMS, ALCOM and the Department of Library Sciences and Media Services at Kent State University continued development of an interactive web tool for communication and collaboration between participants. This tool now allows researchers to easily and confidentially post experimental and simulation data, images, and comments to the rest of the group. The tool is accessible from the LCP group's home page, and password authentication is required for viewing or entry.

 GLASSES AND GLASS-FORMING MATERIALS

Improved understanding of the microscopic physics of disordered and partially-ordered materials is required to make better glass. Recent advances in parallel computing now make previously intractable simulations of these materials possible. Following a workshop organized and sponsored by the CTCMS on Glasses and Glass-Forming Materials: Challenges in Materials Theory and Simulation, a group of researchers from various universities, national laboratories, and NIST formed the CTCMS Glasses group to measure and characterize the microstructure and dynamics of heterogeneities arising in liquids as they form glasses. CTCMS and the Parallel Computation Division at Sandia National Laboratories are co-supporting a postdoctoral researcher on the project. In FY 1997, the Glasses group conducted large-scale massively parallel simulations of standard models of glass-forming systems, and reported a number of startling results that may contribute to our understanding of anomalous structural relaxation and physical aging in glasses. Proceedings from the FY 1996 workshop were published in the journal Computational Materials Science.

Inspection and repair of solder defects in electrical and optical interconnects is expensive and labor-intensive. As circuits become smaller, the importance of computational tools in integrating design and performance with manufacturing and reliability increases. The electronics industry wants to optimize solder interconnects but lacks tools to predict the performance and reliability of the wide range of geometries used.

In FY 1997 the CTCMS Solder Interconnect Design Team (SIDT) continued to develop and evaluate computational methods and software tools for modeling geometries that arise in solder interconnects. Problems identified by group members that are under current consideration include tombstoning (lifting of a small component off the circuit board), forces on the gull wing lead, solidification of the solder interconnect, reactive wetting (dissolution and the formation of intermetallics), and optoelectronic interconnects.

The group held workshops at the CTCMS attended by team members from the Edison Welding Institute, DEC, Motorola, BOC Gases, Ford Motor Co., Lucent Technology, AMP, Rockwell, Delco, Texas Instruments, Susquehanna University, University of Colorado, University of Massachusetts, University of Wisconsin, University of Loughborough, Lehigh University, University of Greenwich, Marquette University, RPI, University of Minnesota, Sandia, and, of course, NIST.

At the June 1997 workshop, work on a variety of topics was presented spanning the industrial concerns of AMP, TI, Ford, Motorola, and Lucent, as well as exciting new scientific work from many of the team's academic and government partners. A new modeling technology is currently being developed by at U. Greenwich to describe the stresses which develop during the solidification of solder interconnects. This effort also should improve the ease of use of the Surface Evolver, an NSF Geometry Center software tool that underlies the solder tools, and which currently requires substantial specialized expertise. A new model describing the dissolution and spreading of a solder drop over a reactive substrate was developed. CTCMS-supported researchers at the University of Greenwich are examining the buildup of stress during the solidification of optical interconnects, as well as the behavior of the solder-flux paste under reflow. A CRADA with Boeing was established in FY 1997 in conjunction with the ATP program in optoelectronics. It is expected that the research fostered by the SIDT will play an important role in modeling solder interconnects in optoelectronic devices, where solder is used solely as an adhesive.