Advanced Computations

General Information about the Research

Advanced-computations research in chemical engineering involves the use of high-performance computing to address problems important to our discipline. The applications are widely varying, but all require advanced computing techniques because of challenges such as solving millions of equations millions of times, performing rapid 3D visualization using hundreds of gigabytes of data, or modeling systems at different levels of detail (electrons, atoms, molecules, meso-scale and continuum). The research is cross disciplinary because of the need to bring together experts in computational modeling, numerical methods, and fundamental science.

At LSU, professors and students involved in computational research benefit from outstanding resources – both facilities and personnel.

LSU is home to four supercomputers with a total of 454 nodes (see LSU High Performance Computing for details). The fastest cluster at HPC-LSU, Tezpur, is a Dell Linux cluster with 360 nodes, each with two dual-core 2.66 GHz Intel Xeon 64-bit processors (a total of 1440 cores), with 4 GB RAM per node, and a peak performance of 15.322 TFlops. In addition, the Louisiana Optical Network Initiative (LONI) provides access to 12 clusters with a total of 1513 nodes. Among these, 1448 nodes belong to Dell Linux clusters with a total of 8512 cores; the rest of the nodes are part of IBM Power5 AIX clusters. The fastest cluster at LONI, Queen Bee, is a Dell Linux cluster with 680 nodes, each with two quad-core 2.33 GHz Intel Xeon 64-bit processors and 8 GB RAM per node. Queen Bee has a peak performance of 50.7 TFlops. On the personnel side, LSU’s Center for Computation and Technology employs a large and diverse staff that ranges from world-class researchers to graphics specialists to hardware technicians. CCT is leading efforts in grid computing, optical networking, algorithm development, and more.

In our Department, advanced computing is being used for computational fluid dynamics, materials characterization, chemical reactor design, process design and optimization, and molecular simulation of nano- and bio-materials. Graduate students are involved with multidisciplinary research through formal programs such as the NSF IGERT on computational fluid dynamics, industrial partnerships such as the PoreSim Consortium, and collaborations that involve research partners from different departments at LSU, as well as from around the nation and the world.
 

Professors Involved

Prof. Karsten E. Thompson

Professor Thompson’s research group focuses on computational modeling of flow and structure in porous materials. Applications include oil & gas production, membrane separations, advanced materials (solid foams and composite fibrous materials), and marine and environmental applications. A partial list of current projects is listed below. See www.poresim.org for more information:

• Multi-scale modeling techniques: sub-pore scale to Darcy scale flows
  
  
• Automatic meshing techniques from 3D computed tomography images
  
  
• Modeling interfacial behavior in porous materials
  
  
• Inertial flows in porous materials
  
  
• The motion and trapping of non-spherical particles in constricted channels

Prof. Martin A. Hjortsø
Simulation of biological systems

Prof. Ralph W. Pike
Simulation of fluid mixing and reaction processes

Prof. José A. Romagnoli
Process Systems Engineering at LSU

Prof. Francisco R. Hung
Nanoporous materials and confined systems, liquid crystals, self- and
directed assembly at the nanoscale, computational modeling and
molecular simulation.

This page was last updated on May 6, 2009