The project will perform fundamental research in developing computational methods and tools for dynamic Monte Carlo (MC) neutron transport simulations on exascale class computer architectures. The NC State research team will perform studies on hybrid deterministic-MC and multi-level computational methods. The expected results of this research will form the background for development of efficient numerical methods and tools for time-dependent neutron transport problems and will be used to determine the next step in this direction. The results will be delivered as technical reports and papers published in technical journals and proceedings of major technical conferences. The project will educate two Ph.D. students

This project will develop and implement new algorithms in the LANL thermal radiation Sn transport code Capsaicin. This radiation transport code will be interfaced with the Eulerian Applications Project to obtain a radiation-hydrodynamics capability with a multigroup Sn radiation treatment. The new algorithms will be based on the second-moment method. The goal is to develop new computational methods with improved efficiency and reduced memory requirements relative to existing capabilities.

This research project will develop approaches and methodologies for formulating efficient reduced-order models for radiation hydrodynamics of physical systems and their optical emission properties.

Reduced-order-models of the radiation hydrodynamics equations in the early-time strong shock stage after the detonation of a nuclear weapon and yield is achieved have the potential to provide accurate and computationally-efficient initial conditions for subsequent simulations of the next stages and resulting nuclear weapons effects. We propose to investigate two separate but synergistic approaches to model-order-reduction for the coupled thermal radiation/shock hydrodynamics equations in comparison with high-fidelity rad/hydro simulations. Our goals are to evaluate the utility of each of the approaches separately, and in combination, by applying them to a) reduced spatial dimension problems with sophisticated physics models, and b) to reduced physics problems involving all three spatial dimensions. Investigating these parallel tracks will guide national laboratory code developers in bridging the gap from cheap empirical approaches to the full-physics, high-fidelity simulations currently available but reserved for hero/benchmark simulations. These reduced-order-models allow analysts to more thoroughly explore solution spaces for a wide variety of possible scenarios, naturally including uncertainty and/or parameter variation.

The Consortium for Advanced Simulation of Light Water Reactors, CASL, supports the broad national missions of enabling energy independence; supporting economic growth through the offering of superior technology ; and being good stewards of the environment, buy enabling predictive simulation of nuclear power plants. Such capability will make possible power uprates, lifetime extension and higher fuel burnups for currently operating and new Generation III+ nuclear power plants.

To perform analysis of characteristic-based radiation transport methods in 2D r-z geometry to assess their accuracy and efficiency.

Project Summary/Abstract Applicant: North Carolina State University Principal Investigator: Paul Turinsky* Co-Principal Investigators: Hany Abdel-Khalik* Dmitriy Anistratov* Michael J. Doster* *Members of NC State?s Department of Nuclear Engineering Computer Committee. Title: Enhancement of Computational Facilities in Support of GNEP Research and Training Abstract: This proposal requests support to enhance computational facilities that are utilized in GNEP related research and training. Specifically, 19 computer nodes (76 processors) and 10.5TB of NFS attached storage with tape backup are requested to be added to the NC State University IBM Blade Linux Cluster. Each computer node will support two 15130 Xeon processors with 8GB of memory. These are dual core processors, so each node has 4 cores and 2GB of memory per core. This addition of 19 nodes will increase the IBM Blade Linux Cluster to have a total of 194 computer nodes, a modest increase from the current number of computer nodes. However, with the current priority system that is used on the cluster, jobs that require more than a few nodes and significant CPU cycles are given very low priority access, implying a day or more turnaround time for job completion. When additional computer nodes are provided by a faculty member, they are given highest priority access to their nodes, which substantially decreases turnaround time, hence improving research productivity. For the 19 nodes to be added, highest priority access will be given to faculty, staff and students who are engaged in research supportive of GNEP goals. This will allow software development, which requires frequent submission of jobs, to progress rapidly on up to 76 processors, later accessing the fuller capabilities of the cluster as needed for production jobs and finer parallel algorithm tuning. Given the current level of GNEP related computational based research being conducted at NC State, and the anticipated even higher level of GNEP related computationally based research based upon pending proposals, GNEP research advancements will benefit from the enhancement of computational resources dedicated to supporting such research. Training of students, particularly graduate students who are our future researchers, will also benefit by providing better access to computational facilities that lead to greater efficiency in conducting research and learning.

To accelerate multigroup diffusion calculations for neutron transport problems with upscattering and fission processes, it is proposed to develop a methodology based on nonlinear projection and prolongation (restriction) operators.

We propose to investigate the properties of a variety of characteristic methods in 1D spherical and cylindrical geometry. We will analyze the behavior of the discretizations in optically-thick and diffusive regions, using the asymptotic diffusion limit analysis, the infinite-medium low-order polynomial solutions, as well as a variety of special test problems. The expected results of this research will form the background for development of accurate and robust transport methods for r-z geometry and will be used to determine the next step in this direction.