MRI/Acq.: Distributed Computing Cluster for Multidisciplinary Research, Rsch Training, and Educ
Project Proposed: This project, acquiring a versatile high-performance computing cluster that will enable multi-disciplinary research, services physics and astronomy, chemistry, computer science and information technology, and mathematics. The requested equipment consists of a computer cluster (16 nodes, each two dual-core processors), three control nodes (1 master and two I/O nodes), a RAID storage subsystem, a tape drive, and necessary compilers and computational software. Research projects involve: . Data mining of large astronomical databases, . Coherent synthesis of high-harmonics, . Understanding of noncovalent assembly of macromolecules. . Application of parallel techniques to traditional applied mathematical methods, and . Tension supported truss systems (tensegrities).
Broader Impacts: The instrument creates new opportunities for undergraduate research, contributing to train undergraduates in the use of high-performance cluster technology. Undergraduate capstone projects will be directed in Computational and Numerical Methods and Computer Science. The facility might enable students to perform graduate studies in applied sciences.
This award has provided the establishment of a shared computational resource, a parallel computing cluster, at Austin Peay State University in Clarksville, Tennessee. This resource is in use by faculty in three different departments (Physics & Astronomy, Chemistry, and Mathematics), and is currently being used to pursue a wide variety of computational topics including inorganic and organic chemistry, atomic physics, x-ray spectroscopy, materials modeling, and numerical algorithm development. These are outlined below. X-ray spectroscopy - A pre-existing set of codes (the Los Alamos Atomic Physics codes) has been deployed on the cluster and is being used to investigate the x-ray spectrum of helium-like neon to determine if resonant processes must be included in an ab initio fashion or if a two-step approximation is sufficient. Atomic physics - A suite of atomic physics modeling codes is being developed that allow the inclusion of energy resonances at the ab initio level. These are based on b-splines. Inorganic chemistry - The electronic structure of a tri-iron carbene is being investigated to better understand its bonding pattern. This molecule has potential as an organometallic catalyst. The electronic structure of a family of nickle oximate complexes is also being investigated as a potential catalyst. Organic chemistry - A study in the use of organometallic catalysts in the polymerization of epoxides with carbon dioxide into a biodegradable polycarbonate plastic has been started. The exact mechanism of the catalysis is not well understood and is being modeled. Computational materials - A study into technologies that could increase the efficiency of solar energy conversion has begun. In particular, the electronic structure of zinc and silicon nanowires are being investigated to attempt to predict their behavior when included in a multi-layered photovoltaic structure (solar cell). In addition, a numerical study of the non-precious metal catalysts for use in fuel cell technology has begun in collaboration with Thrust II of the TNSCORE project (Tennessee Epscore, EPS 1004083). Algorithm development - A study into the implementation of a new numerical linear multistep algorithm developed by an APSU researcher has begun. The method shows a great deal of promise for N-body type problems (as compared to traditional routines such as Runge-Kutta). This award has also had considerable impact beyond the direct research activities outlined above. It has contributed to the development of 10 faculty and to the hiring of two new computational researchers at APSU. The research projects outlined above have involved 15 undergraduate researchers whose work has been presented at a number of different conferences. Furthermore, the cluster was used in conjunction with a required course in the APSU physics major directly affecting 46 students. Many of these undergraduate researchers and students have gone on to graduate programs in STEM related disciplines.
