Over the past 30 years, the Department of Energy’s (DOE) supercomputing program has played an increasingly important role in scientific research by allowing scientists to create more accurate models of complex processes, simulate problems once thought to be impossible, and to analyze the increasing amount of data generated by experiments. To help the research communities fully tap into the capabilities of current and future supercomputers, Under Secretary for Science Raymond Orbach launched the Innovative and Novel Computational Impact on Theory and Experiment (INCITE) program in 2003. The INCITE program was conceived specifically to seek out computationally intensive, large-scale research projects with the potential to significantly advance key areas in science and engineering. The program encourages proposals from universities, other research institutions and industry.
Since 1974, DOE’s Office of Science, the nation’s single largest supporter of basic research in the physical sciences, has provided supercomputing resources for unclassified research through the National Energy Research Scientific Computing Center (NERSC) at Lawrence Berkeley National Laboratory. During the first two years of the INCITE program, 10 percent of the resources at NERSC were allocated to INCITE awardees. However, demand for supercomputing resources far exceeded available systems and in 2003, the Office of Science identified increasing computing capability by a factor of 100 as the second priority on its Facilities of the Future list. The goal was to establish Leadership Class Computing resources to support open science. As a result of a peer-reviewed competition, the first Leadership Computing facility was established at Oak Ridge National Laboratory in 2004. A second Leadership Computing facility was established at Argonne National Laboratory in 2006. This expansion of computational resources led to a corresponding expansion of the INCITE program. In 2007, Argonne, Lawrence Berkeley, Oak Ridge and Pacific Northwest national laboratories all provided resources for the INCITE program.
Title: Advanced Simulations of Plasma Microturbulence at the Petascale and Beyond
Principal Investigator: William M. Tang, Princeton University
Co-Investigators:
- Mark F. Adams, Columbia University
- Stéphane Ethier, Princeton Plasma Physics Laboratory
- Scott Klasky, Oak Ridge National Laboratory
- Bruce Scott, Max-Planck Institute for Plasma Physics
- Weixing Wang, Princeton University
Scientific Discipline: Physics: Plasma Physics
INCITE Allocation: 8,000,000 processor hours
Site: Argonne National Laboratory
Machine (Allocation): IBM Blue Gene/P (8,000,000 processor hours)
Research Summary:
Worldwide energy consumption has risen twenty-fold during the 20th century and shows no sign of abating. Nuclear fusion presents tantalizing potential, promising a safe, clean, and sustainable way to meet a large portion of the world‘s energy needs on a continuous basis. However, the scientific and engineering challenges in designing a fusion reactor are formidable. Therefore, the first commercial power plants are not expected before the middle of the century. Computations carried out at the extreme scale of key plasma physics and materials science processes in fusion devices will be key to helping design, run, and interpret expensive large-scale experiments and are expected to significantly speed up the realization of fusion power plants.
A major plasma physics problem on the road to a working fusion power plant involves significantly improving the understanding, prediction, and control of large- and small-scale instabilities caused by unavoidable plasma inhomogeneities. One consequence is the occurrence of turbulent fluctuations associated with the significant transport of heat, momentum, and particles across the confining magnetic field. Understanding and possibly controlling the balance between these energy losses and the self-heating rates of the actual fusion reaction is key to achieving the efficiency needed to help ensure the practicality of future fusion power plants. The present INCITE project on advanced Particle-in-Cell (PIC) global simulations of plasma microturbulence at the petascale and beyond is motivated by this FES grand challenge.
Engaging the power of modern high-performance computers will help scientists understand how to control the balance between energy losses caused by microturbulence and the self-heating rates of the actual fusion reaction. This is essential to achieving the efficiency needed to help ensure the practicality of future fusion power plants.
Title: Cellulosic Ethanol: Simulation of Multicomponent Biomass System
Principal Investigator: Jeremy Smith, Oak Ridge National Laboratory
Co-Investigators:
- Xiaolin Chang, Oak Ridge National Laboratory
- Loukas Petridis, Oak Ridge National Laboratory
Scientific Discipline: Biological Sciences: Biophysics
INCITE Allocation: 30,000,000 processor hours
Site: Oak Ridge National Laboratory
Machine (Allocation): Cray XT (30,000,000 processor hours)
Research Summary:
Rational strategies for improving the efficiency of biofuel production from plant cell wall lignocellulosic biomass via cellulose hydrolysis require a detailed understanding of the structure and dynamics of the biomass. Lignocellulosic biomass is a complex material composed of cellulose microfibrils laminated with hemicellulose, pectin, and lignin polymers. To reduce biomass recalcitrance to hydrolysis by the improvement of pretreatment and the design of improved feedstock plants, a detailed understanding of biomass structure, mechanics, and response to pretreatment regimes is needed. During a previous INCITE award, we applied molecular dynamics (MD) simulation to understand the structure and dynamics of lignin aggregates and of lignin precipitation on cellulose fibers. During this time we also developed technology permitting the efficient simulation of multimillion atom biomolecular systems running on O(~10-100k) cores. This enables us to propose, for the present award, to extend the lengthscale of the systems under study to enable the simulation of full lignocellulosic biomass systems, consisting of cellulose, lignin and hemicelluloses, together in specific cases with hydrolyzing enzymes.
The proposed research aims at providing simulation models of biomass and biomass:enzyme interactions that will help us understand the physical origins of biomass recalcitrance, using atomic-detail computer simulation of biomolecular systems with the molecular dynamics (MD) method involving the stepwise integration of the equations of motion. The detailed multiscale structure revealed by these simulations will aid in understanding biomass recalcitrance to hydrolysis and in engineering efforts to improve second-generation biofuel yield.
Title: Climate-Science Computational Development Team: The Climate End Station II
Principal Investigator: Warren Washington, National Center for Atmospheric Research
Co-Investigators:
- Philip Cameron-Smith, Lawrence Livermore National Laboratory
- Scott Elliott, Los Alamos National Laboratory
- David Erickson, Oak Ridge National Laboratory
- Steven Ghan, Pacific Northwest National Laboratory
- James Hack, Oak Ridge National Laboratory
- Jim Hurrell, University Corporation for Atmospheric Research
- Rob Jacob, Argonne National Laboratory
- Philip Jones, Los Alamos National Laboratory
- Jean-Francois Lamarque, University Corporation for Atmospheric Research
- L. Ruby Leung, Pacific Northwest National Laboratory
- Bette Otto-Bliesner, University Corporation for Atmospheric Research
- Steven Pawson, NASA
Mark Taylor, Sandia National Laboratories
- Peter Thornton, Oak Ridge National Laboratory
Scientific Discipline: Earth Science: Climate Research
INCITE Allocation: 110,000,000 processor hours
Site: Oak Ridge National Laboratory
Machine (Allocation): Cray XT (70,000,000 processor hours)
Site: Argonne National Laboratory
Machine (Allocation): IBM Blue Gene/P (40,000,000 processor hours)
Research Summary:
The Climate Science Computational End Station (CCES) will project future climates using scenarios of anthropogenic emissions and other changes resulting from energy policy options. Climate change simulations and climate variability studies will directly inform national science policy, thereby contributing to the DOE, NSF and NASA science missions. Of particular importance are global high resolution simulations that will improve the scientific basis, accuracy, and fidelity of climate models. Continuing model development and extensive testing of the Community Earth System Model (CESM) to include recent new knowledge about ocean and land ecosystems is at the cutting edge of climate science research.
Title: High-Fidelity Simulations for Advanced Engine Combustion Research
Principal Investigator: Joseph Oefelein, Sandia National Laboratories
Co-Investigators:
- Jacqueline Chen, Sandia National Laboratories
- Ramanan Sankaran, Oak Ridge National Laboratory
Scientific Discipline: Chemistry: Combustion
INCITE Allocation: 60,000,000 processor hours
Site: Oak Ridge National Laboratory
Machine (Allocation): Cray XT (60,000,000 processor hours)
Research Summary:
Transportation by automobiles and trucks in the United States accounts for two-thirds of our oil use and one-fourth of our greenhouse gas emissions. Thus, the interdependent advancement of both fuel and engine technologies is a key component of the strategy to dramatically reduce both oil consumption and greenhouse gases.
The calculations proposed here aim to contribute to this goal through development of advanced predictive capabilities for turbulent combustion processes in internal-combustion (IC) engines. We will apply an optimal combination of large eddy simulations, direct numerical simulations, and molecular dynamics simulations to provide new insights with respect to key phenomenological processes and further refinement and validation of key sub-models. While the focus of the current effort is on IC-engines, it should be noted that the challenges and approach described here apply to any propulsion and power device. The collaborative effort is supported by a portfolio of five DOE funded projects spanning fundamental (Office of Science, BES) to applied research (Energy Efficiency and Renewable Energy, OVT) with strong coupling to a companion set of experiments. These projects directly address targeted research areas identified as part of a BES sponsored workshop entitled Basic Research Needs for Clean and Efficient Combustion of 21st Century Transportation Fuels. The major goals of the effort are 1) to provide new insights into the dynamics of turbulent combustion processes in IC-engines, and 2) maximize the collective benefits of these insights through synergistic collaborations between the sub-groups of researchers involved.
Title: High Fidelity Tokamak Edge Simulation for Efficient Confinement of Fusion Plasma
Principal Investigator: C.S. Chang, New York University
Co-Investigators:
- Scott Klasky, Oak Ridge National Laboratory
- Scott Parker, University of Colorado
- Linda Sugiyama, Massachusetts Institute of Technology
Scientific Discipline: Physics: Plasma Physics
INCITE Allocation: 50,000,000 processor hours
Site: Oak Ridge National Laboratory
Machine (Allocation): Cray XT (50,000,000 processor hours)
Research Summary:
Success of ITER and commercialization of the magnetic fusion reactors require good confinement of plasma energy at the plasma edge to form a plasma energy pedestal significantly above the cold, wall-interacting plasma of the -scrape-off- layer (this type of plasma operation is called -H-mode-). With the bifurcation of the edge plasma into the H-mode, the core plasma energy is quickly enhanced to the level of -fusion- condition for some unknown reasons. This is often called the -the-tail-wagging-the-dog- phenomenon in the fusion community. ITER program is based upon the H-mode pedestal formation and the tail-wagging-the-dog phenomenon, both of which are not well understood. A first-principles understanding of these phenomena is one of the main mission of the present INCITE project.
Large scale simulations, pushing the limit of Jaguar, of tokamak plasma in realistic device geometries will be performed in this INCITE project. An extensive study of edge pedestal physics will be performed. In order to understand the nonlocal edge pedestal effect on the core plasma confinement, multiscale simulations of the whole volume plasma will also be performed. The whole volume includes the magnetic separatix, magnetic axis and the material wall boundary. Nonlocal turbulence propagation and the self-organization of the plasma profile to a critical global state will be simulated in H-mode plasmas. These simulations will hopefully shed light to the understanding of the edge pedestal physics and the tail-wagging-the-dog phenomenon, long awaited problems in magnetic fusion physics. As the edge plasma pedestal becomes steeper, a large scale instability called -edge localized modes- (ELMs) crashes the plasma pedestal, limiting the core confinement and damaging the material wall. Control of large scale ELMs is an essential condition to the success of the ITER program. This INCITE will study what the edge localized modes are and how to control them.
Title: Investigation of Multi-Scale Transport Physics of Fusion Experiments Using Global Gyrokinetic Turbulence Simulations
Principal Investigator: Weixing Wang, Princeton Plasma Physics Laboratory
Co-Investigators:
- Mark Adams, Columbia University
- Stephane Ethier, Princeton Plasma Physics Laboratory
- Scott Klasky, Oak Ridge National Laboratory
- Wei-li Lee, Princeton Plasma Physics Laboratory
Scientific Discipline: Physics: Plasma Physics
INCITE Allocation: 20,000,000 processor hours
Site: Oak Ridge National Laboratory
Machine (Allocation): Cray XT (20,000,000 processor hours)
Research Summary:
The development of magnetic fusion as a secure and reliable energy source that is environmentally and economically sustainable is a formidable scientific and technological challenge in the 21st century. Understanding heat and particle losses caused by plasma turbulence in magnetic fusion devices is especially important for the next generation of burning plasma experiments such as international ITER reactor because the size and cost of a fusion reactor are expected to be largely determined by the balance between these energy losses and the self-heating rates of the actual fusion reaction. Accordingly, the control and possible suppression of turbulence caused by plasma microinstabilities is a major area of ongoing research of which advanced numerical simulations is a prominent component.
This petascale simulation project will investigate the physics of turbulence-driven momentum, energy, and particle transport, and their relationship to tokamak fusion experiments. The focus will be on the nonlinear physics occurring on multispatial and multitemporal scales involving both ion and electron dynamics. Our numerical studies will emphasize i) the physics validation of our simulation model against results from the three major fusion experiments in the United States, namely NSTX, DIII-D and C-MOD, and ii) the application of predictive capability in these simulation tools for assessing critical plasma confinement issues associated with ITER. Reliable predictions of the confinement properties in modern laboratory fusion experiments will require global kinetic simulations with multi-scale resolution—a true grand challenge that will require petascale computing capabilities.
Title: Magnetic Structure and Thermodynamics of Low Dimensional Magnetic Structures
Principal Investigator: Markus Eisenbach, Oak Ridge National Laboratory
Co-Investigators:
- Paul Kent, Oak Ridge National Laboratory
- Malcolm Stocks, Oak Ridge National Laboratory
Scientific Discipline: Materials Science: Condensed Matter and Materials
INCITE Allocation: 50,000,000 processor hours
Site: Oak Ridge National Laboratory
Machine (Allocation): Cray XT (50,000,000 processor hours)
Research Summary:
This project will explore the energy landscape of low dimensional magnetic structures (LDMS) that are free standing, adsorbed on surfaces, and embedded in the bulk. Within bulk magnetic materials, all defects are LDMS, having spin arrangements that differ from their surroundings. The properties of LDMSs lead to novel responses to electric, magnetic, and stress fields that may result in important logic, memory, optical, and structural applications. In many materials, typically steels, magnetic defects are important to strength and fracture toughness. The goal of this work on LDMS is to understand their low-temperature magnetic structure and their thermodynamic fluctuations at higher temperature. This understanding can lead to advances in energy and information applications and to stronger, lighter materials for increased energy efficiency. The work will advance the overall objectives of the Office of Science and will contribute to and benefit from modeling and experimental work in the Energy Frontier Center for Defect Physics in Structural Materials.
Title: Nuclear structure and nuclear reactions
Principal Investigator: James Vary, Iowa State University
Co-Investigators:
- Joseph Carlson, Los Alamos National Laboratory
- Pieter Maris, Iowa State University
- Hai Ah Nam, Oak Ridge National Laboratory
- Petr Navratil, Lawrence Livermore National Laboratory
- Witold Nazarewicz, University of Tennessee
- Steven Pieper, Argonne National Laboratory
Scientific Discipline: Physics: Nuclear Physics
INCITE Allocation: 43,000,000 processor hours
Site: Oak Ridge National Laboratory
Machine (Allocation): Cray XT (28,000,000 processor hours)
Site: Argonne National Laboratory
Machine (Allocation): IBM Blue Gene/P (15,000,000 processor hours)
Research Summary:
Developing a comprehensive description of all nuclei (stable and unstable) and their reactions requires investigations of rare and exotic isotopes with unusual proton-to-neutron ratios that are difficult to produce and study experimentally because of their short lifetimes. We perform state-of-the-art simulations to provide needed predictions where direct experiment is not possible or is subject to large uncertainties.
Predictions for the structure and reactions of nuclei, with assessed uncertainties, are important for the future of the nation's energy and security needs. Such calculations are relevant to many applications in nuclear energy, nuclear security and nuclear astrophysics, where rare nuclei lie at the heart of nucleosynthesis and energy generation in stars.
Title: Performance Evaluation and Analysis Consortium End Station
Principal Investigator: Patrick Worley, Oak Ridge National Laboratory
Co-Investigators:
- David H. Bailey, Lawrence Berkeley National Laboratory
- Jack J. Dongarra, University of Tennessee
- William D. Gropp, University of Illinois at Urbana-Champaign
- Jeffrey K. Hollingsworth, University of Maryland
- Robert F. Lucas, University of Southern California
- Allen D. Malony, University of Oregon
- John Mellor-Crummey, Rice University
- Barton P. Miller, University of Wisconsin at Madison
- Leonid Oliker, Lawrence Berkeley National Laboratory
- Allan Snavely, University of California at San Diego
- Jeffrey S. Vetter, Oak Ridge National Laboratory
- Katherine A. Yelick, University of California at Berkeley
- Bronis R. de Supinski, Lawrence Livermore National Laboratory
Scientific Discipline: Computer Science
INCITE Allocation: 30,000,000 processor hours
Site: Oak Ridge National Laboratory
Machine (Allocation): Cray XT (20,000,000 processor hours)
Site: Argonne National Laboratory
Machine (Allocation): IBM Blue Gene/P (10,000,000 processor hours)
Research Summary:
To maximize the utility of Department of Energy leadership class systems such as the Cray XT4, Cray XT5, and IBM Blue Gene/P, we must understand how to use each system most efficiently. The performance community (performance tool developers, performance middleware developers, system and application performance evaluators, and performance optimization engineers) can provide the tools and studies to enable these insights, if they have adequate access to the systems. To provide further understanding of these high-end systems, this proposal focuses on four primary goals: (1) update and extend performance evaluation of all systems using suites of both standard and custom micro, kernel, and application benchmarks; (2) continue to port performance tools and performance middleware to the BG/P and XT4/5; (3) validate the effectiveness of performance prediction technologies, modifying them as necessary to improve their utility for predicting resource requirements for production runs on the leadership-class systems; and (4) analyze and help optimize current or leadership class application codes.
Title: Petascale Modeling of Chemical Catalysts and Interfaces
Principal Investigator: Robert Harrison, Oak Ridge National Laboratory
Co-Investigators:
- Edoardo Apra, Oak Ridge National Laboratory
- David Dixon, University of Alabama
- Karol Kowalski, Pacific Northwest National Laboratory
- William Shelton, Oak Ridge National Laboratory
- David Sherrill, Georgia Institute of Technology
- Bobby Sumpter, Oak Ridge National Laboratory
Scientific Discipline: Chemistry: Catalytic
INCITE Allocation: 75,000,000 processor hours
Site: Oak Ridge National Laboratory
Machine (Allocation): Cray XT (75,000,000 processor hours)
Research Summary:
We refocus our prior INCITE work on the central science theme of understanding, controlling, and ultimately designing catalytic chemical processes with special interest in surfaces that are relevant to diverse battery technologies, ultra-capacitors, fuel cells, environmental chemistries, and catalytic processes for sustainable energy production including biomass conversion. This is a grand-challenge and cross-cutting problem that, in close partnership with experiment and theory, requires sustained progress over the coming decade starting now with petascale and eventually with exascale computers. This fundamental science topic and the essential integration of experiment with the required new theory and advanced computational tools have been identified as proposed or cross-cutting research directions in multiple Basic Research Needs (BRN) reports including -BRN to Assure a Secure Energy Future, -BRN for Solid State Lighting-, -BRN for Electrical Energy Storage,- and -BRN Catalysis for Energy.- Theoretical simulations output of this proposal are structured to deliver -a significant increase in the rate of discovery, innovation and technological change'- (from BES report -New Science for a Secure and Sustainable Energy Future-) in energy storage and production. The cross-cutting simulation capabilities will, by design, have immediate impact also on biomass conversion, inorganic and organic photo-voltaics, and the heavy-element chemistries that are vital to diverse DOE missions.
Title: Three Dimensional Simulations for Core Collapse Supernovae
Principal Investigator: Anthony Mezzacappa, Oak Ridge National Laboratory
Co-Investigators:
- John Blondin, North Carolina State University
- Stephen Bruenn, Florida Atlantic University
- Christian Cardall, Oak Ridge National Laboratory
- William Raphael Hix, Oak Ridge National Laboratory
- Jirina Stone, Oak Ridge National Laboratory
Scientific Discipline: Physics: Astrophysics
INCITE Allocation: 60,000,000 processor hours
Site: Oak Ridge National Laboratory
Machine (Allocation): Cray XT (60,000,000 processor hours)
Research Summary:
Core-collapse supernovas are the death throes of massive stars. They are the single most important source of elements in the universe. Understanding how they occur is one of the crucial unsolved problems in astrophysics. The focus of this project is to perform multidimensional, multiphysics simulations of core-collapse supernovas in an effort to determine the supernova explosion mechanism.
This project will perform three-dimensional simulations to understand how stars more than ten times the mass of our sun die in catastrophic stellar explosions known as core-collapse supernovae. Core-collapse supernovae are the dominant source of elements in the universe, including all the elements between oxygen and iron and half the elements heavier than iron; life would not exist without these elements. These supernovae are complex, three-dimensional, multi-physics events, but there are as yet no three-dimensional models of sufficient realism. This is a significant void in supernova theory. The simulations described here will begin to fill this void. These simulations will be the first three-dimensional simulations performed with multifrequency neutrino transport, critical to realistic modeling of the neutrino shock reheating that is believed to be central to the supernova explosion mechanism. A complete understanding of the core-collapse supernova mechanism requires parallel simulations in one, two, and three spatial dimensions. The nuclei in the stellar core undergo a transition through a series of complex shapes that can be modeled only in three spatial dimensions. These modeling efforts will extend to three dimensions both the macroscopic and microscopic models of stellar core phenomena in core-collapse supernovae.
Title: Ultrascale Simulation of Basin-Scale CO2 Sequestration in Deep Geologic Formations and Radionuclide Migration using PFLOTRAN
Principal Investigator: Peter Lichtner, Los Alamos National Laboratory
Co-Investigators:
- Glenn Hammond, Pacific Northwest National Laboratory
- Richard Mills, Oak Ridge National Laboratory
Scientific Discipline: Earth Science: Environmental Sciences
INCITE Allocation: 15,000,000 processor hours
Site: Oak Ridge National Laboratory
Machine (Allocation): Cray XT (15,000,000 processor hours)
Research Summary:
This project brings petascale computing resources to bear on current environmental problems involving global warming and sequestration of greenhouse gases such as CO2 in deep geologic formations and migration of radionuclides from highly contaminated DOE legacy sites from WW-II and the Cold War at the DOE Hanford and Oak Ridge sites. We will apply petascale computing to extreme scale problems involving CO2 sequestration in large geologic basins to investigate the effects of displacing deep formation water brines by large volumes of CO2 on potential contamination of drinking water aquifers.
High performance computing will be employed to better quantify parameter uncertainty and conduct sensitivity analyses for groundwater flow and uranium transport at the Hanford 300 Area. The project will leverage new data being produced by the Hanford 300 Area IFRC project (funded by DOE-SC BER SBR) that better defines hydrostratigraphy, physical heterogeneity, and geochemical process models for the Hanford 300 Area. PFLOTRAN is capable of launching thousands of simultaneous simulations, each simulation depicting large high-resolution problems domain with an independent set or realization of uncertain parameters. Results generated from these ensembles of simulations help to characterize parameter sensitivity and quantify uncertainty. Leadership class computing that provides tens to hundreds of thousands of processor cores is required to perform these large ensembles of simulations.
At the Oak Ridge IFRC site, we will work with members of the ORNL IFRC project in constructing a watershed scale groundwater model with refined resolution using high performance computing for a watershed scale model with refined resolution for the S-3 ponds and the Bear Creek Watershed on the DOE Oak Ridge Reservation. This model will integrate multiple processes at multiple scales to investigate the influence of process interactions at small scales on the fate and transport of contaminants in the field, and the scale dependence of the controlling parameters such as dispersivity, attenuation, mass transfer and reaction rates.
Title: Uncertainty Quantification for Three-Dimensional Reactor Assembly Simulations
Principal Investigator: Thomas Evans, Oak Ridge National Laboratory
Co-Investigators:
- Kevin Clarno, Oak Ridge National Laboratory
- Matthew Jessee, Oak Ridge National Laboratory
- Wayne Joubert, Oak Ridge National Laboratory
- Scott Mosher, Oak Ridge National Laboratory
- Bradley Rearden, Oak Ridge National Laboratory
Scientific Discipline: Energy Technologies: Nuclear Energy
INCITE Allocation: 18,000,000 processor hours
Site: Oak Ridge National Laboratory
Machine (Allocation): Cray XT (18,000,000 processor hours)
Research Summary:
The performance of nuclear fuel in boiling water reactors (BWRs) is strongly dependent on the power distribution within a nuclear fuel bundle. Vendor software, as is typical in the nuclear industry, provides pin-averaged power distribution data on a 6-inch axial grid, which is much too large to be able to determine the within-pin power distribution near the control blades, before and after withdrawal. However, the problem is not simply single-physics radiation transport. The effective macroscopic nuclear cross-sections, which are used in a radiation transport solver, are linearly dependent upon the density of the materials. Before we proceed to develop a coupled thermal-fluid-dynamics and radiation transport solver, it is necessary to evaluate the required fidelity of each. Evaluating the sensitivity of the radiation transport solution to the spatial resolution of the coolant density would guide the software development of a coupled physics solver by determining the -fidelity requirements of the thermal-fluid-dynamics solver.
Recent work with the Denovo three-dimensional, discrete ordinates transport code has demonstrated the ability to model the power distribution for this problem on a •-inch axial grid with a 6 mil (0.006 inch) radial (x, y) mesh (using 60,000 cores on the Cray XT5 at ORNL). Therefore, it is clear that the solution is within reach. Uncertainty quantification is one of the most important, and least implemented, topics in computational science. As numerical algorithms and high-performance computing architectures have become more advanced and powerful, the ability to model complex physical phenomena has grown to the point where computer models are often used in place of experiment.
Title: Understanding the Ultimate Battery Chemistry: Rechargeable Lithium/Air
Principal Investigator: Jack Wells, Oak Ridge National Laboratory
Co-Investigators:
- Edoardo Apra, Oak Ridge National Laboratory
- Ray Bair, Argonne National Laboratory
- Peter Cummings, Vanderbilt University
- Alessandro Curioni, IBM Zurich Research Laboratory
- Paul Kent, Oak Ridge National Laboratory
- Teodoro Laino, IBM Zurich Research Laboratory
- William Shelton, Oak Ridge National Laboratory
- Winfried Wilcke, IBM Almaden Research Center
- Ye Xu, Oak Ridge National Laboratory
Scientific Discipline: Energy Technologies: Energy Storage
INCITE Allocation: 25,000,000 processor hours
Site: Argonne National Laboratory
Machine (Allocation): IBM Blue Gene/P (15,000,000 processor hours)
Site: Oak Ridge National Laboratory
Machine (Allocation): Cray XT (10,000,000 processor hours)
Research Summary:
Lithium-air cells consist of an anode (ideally lithium metal), an electrolyte - which can be either aprotic or aqueous - and the air cathode1-9. We will focus on aprotic systems, as these are known to be rechargeable and have the highest energy storage density. During the discharge process, Li-ions and atmospheric oxygen combine to form Li2O2 (and perhaps at higher currents some Li2O). The Li2O2 are solids and precipitate on the cathode. It has been demonstrated that it is possible, at least at low electrical current densities, to reverse this process and convert the Li2O2 back into lithium metal. This exciting proof-of-principle work still presents very big technical and engineering challenges before one can be confident that practical propulsion batteries can be based on the Li/Air system. The most important ones are to realize a high percentage of the theoretical energy density, to improve electrical efficiency of recharging, to increase the number of times the battery can be cycled, to limit the negative effects of moisture in the air, and to improve the power density. These five engineering goals depend critically on a deep understanding of four fundamental battery chemistry science issues. These are: (1) detailed mechanisms of the battery discharge and recharge, (2) role of catalysts, (3) solubility of lithium ions and solid lithium oxides and optimization of electrolyte, and (4) the role of the electrode-electrolyte interface in managing oxygen gas, liquid electrolyte, and solid lithium oxides in a nano-structured air-cathode.
