Beowulf Cluster for Computational Corrosion and Catalysis Studies
Abstract
This DURIP-funded Beowulf cluster will enhance our AFOSR-funded research in several areas. Large-scale modeling of complex systems in the areas of corrosion and catalysis is a vital Air Force need. Our previous computational resources alone would not allow us to pursue the most important and complex portions of our current contact with AFOSR. The DURIP funding has now provided our personnel with the computational capability to examine the corrosion of aluminum alloys, to understand the role of stress fields in materials and in the multi-scale modeling of fatigue and fracture, and to study reactivity on patterned surfaces including superlattice, nanoparticles, and piezoelectric oxide supported metal. To perform this research, we will use ab-initio density functional theory (DFT), which gives quantitative results by detailed modeling of the valence electrons in materials. The DFT has reportedly proven itself as a highly accurate and highly efficient first-principles computational tool. However, since DFT is a quantum-mechanical method, high accuracy on real-world systems requires significant computational cost. This grant, by funding a Beowulf cluster supercomputing environment, has provided a novel computer architecture which will enable AFOSR-funded researchers to gain new insights into materials of Air Force needs.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 11, 2002
- Accession Number
- ADA405604
Entities
People
- Andrew M Rappe
Organizations
- University of Pennsylvania