This project provides funding for the purchase of a 32-processor, 192-core computer cluster for research in molecular and materials modeling in the Chemical Engineering Department at the University of Massachusett, Amherst.
Intellectual Merit: This computing facility allows the study of large model systems across multiple length and time scales, as well as systematic parametric analyses using a range of modeling techniques including molecular computer simulation and quantum chemistry. The equipment will serve the computation needs of at least seven research projects in the groups of five faculty members. David Ford's group will use the cluster for classical density functional theory calculations on solid-fluid equilibrium and also for stochastic modeling of colloidal self-assembly processes. Peter Monson's group will use the new facilities for research projects on the dynamics of fluids confined in porous materials. Dimitrios Maroudas' group will use the equipment for their research on multiscale modeling related to the surface engineering of metals and semiconductors and plasma processing of carbon nanostructures. Furthermore, several collaborative projects across the groups will be supported. T.J. (Lakis) Mountziaris works with Maroudas on modeling the doping and synthesis of core/shell semiconductor nanocrystals, while Scott Auerbach works with Monson to model the self-assembly of ordered nanoporous materials, specifically zeolites.
Broader Impact: A commonality among the research projects to be supported by the new computing facilities is fundamental research in molecular and materials modeling in areas where there is a close connection with practical application. As an example, the development of new types of porous materials with properties tailored for specific applications is a major area of research throughout the world. Understanding of how the collective behavior of adsorbed molecules is influenced by the microstructure of the porous material can contribute significantly in this effort. The investigators have reached the point where adsorption experiments can be accompanied by a much more sophisticated understanding of the structure. Monson's projects in this area could provide a foundation for new approaches to the characterization of porous materials. The range of potential impact extends across the range of applications of porous materials. Similar connections to application exist for all of the fundamental research projects to be impacted by the new facilities.
Research in the department has consistently had a strong educational component through the involvement of graduate students, postdoctoral scholars and undergraduates. The equipment requested will be used by 20 graduate students and 6 postdoctoral researchers, as well as undergraduates engaged in independent study projects. The junior researchers involved will learn important techniques in parallel computation for engineering applications. The faculty involved have an established record of bringing molecular and materials modeling into the Chemical Engineering curriculum and these activities will be supported by the new facilities.
This grant was for the purchase of an advanced scientific computing cluster to be used by multiple investigators in the Departments of Chemistry and Chemical Engineering at the University of Massachusetts Amherst. In April 2011 we completed installation of a 10-node cluster with four processors per node and 12 cores per processor, for a total of 480 cores. The main storage is provided by two, 2 TB hard drives in a RAID configuration and communication between nodes is via a 24-port Ethernet switch. Graduate students and postdoctoral researchers in five different research groups are currently running their codes on this new cluster. David Ford’s group is using the cluster for classical density functional theory calculations on solid-fluid equilibrium and also for stochastic modeling of colloidal self-assembly processes. Peter Monson’s group uses the new facilities for research projects on the dynamics of fluids confined in porous materials. Dimitrios Maroudas’ group is using the equipment for their research on multiscale modeling related to the surface engineering of metals and semiconductors and plasma processing of carbon nanostructures. Furthermore, several collaborative projects across the groups are supported. T.J. (Lakis) Mountziaris works with Maroudas on modeling the doping and synthesis of core/shell semiconductor nanocrystals, while Scott Auerbach works with Monson to model the self-assembly of ordered nanoporous materials, specifically zeolites. The key theme in all of this computational research is to accelerate the development of new functional materials, which will benefit the advanced materials industry in the United States. The new cluster provides a state-of-the-art platform for accomplishing our goals.
