Welcome
The scope of computational fluid dynamics continues to expand beyond its original applications in high-technology, aero-mechanical engineering industries and meteorology. Yet, CFD has not yet reached its potential as an engineering tool - much work remains to improve the quantitative accuracy of various techniques used for flow prediction; to establish greater confidence in numerical simulations over more flow regimes; and to make CFD cost-effective for a much broader range of projects and industries.
The Cluster Fluid Dynamics group at the University of Kentucky is addressing these challenges, laying the groundwork for the next century of computational fluid dynamics. This website is intended as an overview of these activities, from active flow control for unmanned aerial vehicles to planetary atmospheres. Further, as these efforts are of limited use when confined to an academic laboratory, underlying all our research is the goal of establishing the quality and trust of CFD prediction tools and the transfer of CFD technology to our scientific and industrial partners.
One of the great barriers to CFD usage is the cost of the necessary computational hardware. A critical part of our research is to increase access to computation fluid dynamics by combining highly optimized, parallel codes with low cost commodity clusters. The research includes constructing relatively inexpensive clusters from commodity PC components and performance analysis of these systems on computational fluid dynamics codes. Our vision is one of CFD tools with collectively optimized hardware and software, incorporating the latest technologies and techniques in a ready-to-use, cost-effective package that could be accessed by many research endeavors beyond the traditional realms of CFD, thereby creating a new paradigm of CFD technology transfer in the 21st century.