October 2014 - Allocations-admins - mailman3.cels.anl.gov

[LCRC Accounts] Project Request: HP_Li
by accounts＠lcrc.anl.gov 29 Oct '14

29 Oct '14

Hello, A new project on the LCRC cluster has been requested. Please forward the information on to the LCRC Allocation sub-committee. Applicant's name: Ying Li Applicant's institution: ANL Applicant's division: LCF Project Name: HP_Li Project title: High-Performance Lithium Battery Associated funding: Margaret Butler Postdoctoral Fellowship Other Systems: BGQ-Mira Science: Li battery is one of the most applicable energy storage devices with advantages of high energy density, stable discharging voltage, no charging memory effect, and environmental friendly. However, the most common difficulties of traditional Li battery during the discharge performance are layers of lithium slats precipitation on the surface of the anode and the limited reaction efficiency at the cathode. For example, one common issue is the solid-electrolyte interface (SEI) will be spontaneously generated across the surface of the anode, which might lead to further dendrite formation. This will be detrimental to the Li battery, even leading to short circuit or cell explosion. More advanced electrode design is requited to realize better performance in the form of <anodemetal|electrolyte|cathodeoxygen> Li-Air cells. The anode/organic electrolyte/membrane/aqueous electrolyte/cathode structure achieves the best performance. The most current limitations in Li Air batteries development are coming from the cathode. Mesoporous carbon, such as layered graphene has been used a cathode material with metal catalysts (such as Mn, Co, Pt and Au). There are still unsolved problems should be tackled before the real investigation of any real industrial experiments, such as the effect of pore size and pore size distribution on the cathode is still poorly understand. Molecular dynamics (MD) simulation is the ideal method to study the atomic interaction and understand the dynamics on cathode surfaces. After constructing a system approximately close to the geometrical structure of the porous graphene emerging in the aqueous electrolyte, with oxygen molecules and electrons involving, the reduction reaction on the cathode can be realized. Such as, 4e- + 2 H2O + O2 → 4 OH-. The effect of pore size of and pore size distribution on the cathode can be also studied here. Project description: The program used for the planned simulations is reactive force field (ReaxFF) MD, which was developed to support extremely scalable simulations on leadership computers. There are basically two tasks to accomplish in this project. Firstly, given the complexity of ReaxFF, we are planning to utilize parallel genetic algorithm (PGA) to derive the parameter set of ReaxFF potential as compared to ab initio quantum mechanic (QM) calculations from VASP in order to accurately describe the reaction mechanism. Secondly, adapt parallel replica dynamics (PRD) to accelerate the long time reaction, presumably needed around several nanoseconds, in the electrolyte and on the cathode. Due to interests on the nano-level of the interface structure is steadily increased, we anticipate the simulation systems will be comprised of ~5,000,000 atoms. Further, the PGA and PRD both are well suited for leadership-class computing resources because there are larger numbers of para meters sets and distributed replicas and the communication overhead between them is minimum. Integrating replicas of the system independently on each processor until the first transition occurs gives the correct transition-time distribution, and hence the correct dynamics. However, the proposed workflow needs to be implemented and optimized on Blues/Fusion. A~ 100,000 processor hours is requested for the proposed implementation and benchmark Industry partnership: N/A Project URL: Requested allocation: 400000 Q1: 100000 Q2: 100000 Q3: 100000 Q4: 100000 Justification: Storage requirements: The requester has used undetermined amount hours of their initial startup project. In addition to approving an initial amount, please specify a Category and Subcategory for this project. For a list of the current selection of approved categories, please see: https://wiki.lcrc.anl.gov/wiki/Processes/Categories Once the Allocation committee has approved the project, please go to the Project Management page to create it: https://accounts.lcrc.anl.gov/projects.php Thank You, The LCRC Accounts System

1 0

New LCRC report submitted for STOandPlus
by support＠lcrc.anl.gov 28 Oct '14

28 Oct '14

Full project report at https://accounts.lcrc.anl.gov/reporting/project/2014/?project=STOandPlus

1 0

[LCRC Accounts] Project Request: SmallWorlds
by accounts＠lcrc.anl.gov 27 Oct '14

27 Oct '14

Hello, A new project on the LCRC cluster has been requested. Please forward the information on to the LCRC Allocation sub-committee. Applicant's name: Mark Hereld Applicant's institution: ANL Applicant's division: MCS Project Name: SmallWorlds Project title: Small Worlds - Dynamic Snapshot 3D Microscopy Associated funding: DOE BER Other Systems: Science: The project aims to develop new high resolution microscopy capability that can use both holographic and interferometric modes to capture 3D images in a single multi-camera snapshot. This will allow us to study detailed dynamical interactions between molecules and structures within and around cells. We are working to demonstrate the technology on the organismal community in Frank Collart's tree, bacteria, funghi system. The scientific goal is to study exchange and transformation of key nutrients to better understand how the community collectively improves nutrient uptake efficiency. Project description: Image reconstruction requires computation for instrument modeling, sensitivity analysis, pre-analysis of training sets, and comparative analysis of reconstructed volumes with ground truth calibration data. Much of this computation will require parallelization for several reasons: in order to perform analysis in time to influence observations, to generate large libraries of synthetic reference imagery and training data, and to reconstruct dynamical interactions from large sequences of detector data produced at 10 frames per second. (*) Experiments to determine scaling properties of reconstruction algorithms including memory and time requirements as function of reconstruction volume and number of sources being modeled. (*) Dependence of reconstructed volume fidelity on number of fluorescent point sources in the detection volume. (*) Accuracy of algorithms for each of the two modes of operation: so-called interferometric, and FINCH. (*) Characterize instrument configurations that combine aspects of the interferometric and FINCH modes in order to determine optimal experiments for various sample configurations and scientific goals. (*) Limits to point position interpolation for highest resolution particle tracking. Many of these experiments will begin as ensemble runs requiring core-parallel but not node-parallel implementations. Depending on timing results, memory requirements, and ultimate problem size we will work on node-parallel implementations. These developments will pertain particularly to the high-speed volume reconstruction codes. Industry partnership: Project URL: http://smallworlds.cels.anl.gov Requested allocation: 80000 Q1: 15000 Q2: 20000 Q3: 25000 Q4: 20000 Justification: Storage requirements: 1 TB will be sufficient at this phase. The requester has used undetermined amount hours of their initial startup project. In addition to approving an initial amount, please specify a Category and Subcategory for this project. For a list of the current selection of approved categories, please see: https://wiki.lcrc.anl.gov/wiki/Processes/Categories Once the Allocation committee has approved the project, please go to the Project Management page to create it: https://accounts.lcrc.anl.gov/projects.php Thank You, The LCRC Accounts System

1 0

New LCRC report submitted for SRF-materials
by support＠lcrc.anl.gov 24 Oct '14

24 Oct '14

Full project report at https://accounts.lcrc.anl.gov/reporting/project/2014/?project=SRF-materials

1 0

[LCRC Accounts] Project Request: Wind_turbine_LES
by accounts＠lcrc.anl.gov 24 Oct '14

24 Oct '14

Hello, A new project on the LCRC cluster has been requested. Please forward the information on to the LCRC Allocation sub-committee. Applicant's name: Shashi Aithal Applicant's institution: ANL Applicant's division: CELS Project Name: Wind_turbine_LES Project title: High-fidelity LES simulation of wind-turbine flow-fields Associated funding: Other Systems: Science: The project is focused on the application and validation of a novel filtered turbulence model (DLRM) to the fluid dynamic simulation of the wind-turbine wake, in order to predict its characteristics (pressure coefficient as a function of time and power generated). Results of the DLRM model will be compared with results obtained from the RANS k-omega SST and from the WALE LES model and with experimental PIV measurements, that are available on the same geometry. Project description: We intend to use OpenFoam 2.2-x as the computational tool for this project. OpenFOAM (OF) is a parallel, finite-volume multi-physics open-source CFD code capable of modeling a wide-range of problems of interest to DoE. One of the strengths of OpenFOAM is that new solvers and utilities can be created by its users. In this project, new moving mesh-strategies and turbulence models (RANS, LES and hybrid LES/RANS) will be applied to study fluid-flow characteristics in the near-wake region of wind-turbine blades. The performance and scalability of OF has been extensively studied. Scaling studies on dsmcFoam and dieselFoam have shown good scalability. Small DSMC cases (1 million cells) have shown good scalability up to 192 cores on Fusion. DieselFoam has shown good scalability up to 512 cores with about 500K cells on Fusion. For fluid-flow cases, it is seen that roughly 50K cells are required per core on Blues to achieve good strong scaling. For this set of wind-turbine simulations, a typical case would have around 80 million-cells. Based on our previous experience good strong scaling can be obtained up to 1600 cores on Blues for such a problem. Several different case-studies are planned for this year. A full 80-M cell case is expected to need about 80-100K core-hours. We intend to study code scaling and performance with coarse grids initially before conducting full-scale production simulations. The anticipated core-hour usage is about 400K. There will be 3-4 members on this project. Industry partnership: Project URL: Requested allocation: 400000 Q1: 50000 Q2: 150000 Q3: 150000 Q4: 50000 Justification: Storage requirements: The requester has used undetermined amount hours of their initial startup project. In addition to approving an initial amount, please specify a Category and Subcategory for this project. For a list of the current selection of approved categories, please see: https://wiki.lcrc.anl.gov/wiki/Processes/Categories Once the Allocation committee has approved the project, please go to the Project Management page to create it: https://accounts.lcrc.anl.gov/projects.php Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Project Request: far_field_hedm
by accounts＠lcrc.anl.gov 23 Oct '14

23 Oct '14

Hello, A new project on the LCRC cluster has been requested. Please forward the information on to the LCRC Allocation sub-committee. Applicant's name: Hemant Sharma Applicant's institution: ANL Applicant's division: XSD-MPE Project Name: far_field_hedm Project title: Realtime analysis of far-field high energy diffraction microscopy (HEDM) data Associated funding: Other Systems: orthros cluster at APS using 384 cores Science: Far-field High Energy Diffraction Microscopy (ff-HEDM) is a novel technique for non-destructive characterization of structure of polycrystalline materials using X-Rays. The brilliant x-rays produced at 3rd generation synchrotron sources, including Advanced Photon Source (APS), enable scientists to look at the constituent structure of materials at the micrometer level. Furthermore, the data can be used to refine the crystallographic orientation and strain state of the sub-structure (so-called 'grains') within the materials. Such knowledge can enable material scientists and modellers to understand, predict and enhance the macroscopic properties of materials for use in a wide variety of applications ranging from jet engines to nuclear materials. This project aims to enable realtime and post-mortem analysis of such diffraction data generate at beamline 1-ID of APS. Project description: We have developed a very efficient data-reduction algorithm for realtime analysis. The first step of computation involves the pre-processing of raw-diffraction images to reduce the size of the data and to find the diffraction peaks. The second step involves 'indexing' (grouping) the peaks belonging to the same 'grain' and then optimization of the centre-of-mass position, crystallographic orientation and strain state of each grain in the investigated volume of the sample. At the 1-ID beamline, we can investigate a sample volume every 3 minutes consisting of between 360 to 1440 diffraction images of 8MB each in size. The package is implemented in C99 and uses only the NLOPT package as a dependency. Parallelization at each stage is achieved using the SWIFT programming language. This allows for ease of implementation and scalability. The software is being used routinely on Orthros (the local HPC at APS) using 384 cores and has achieved upto 90% efficiency. Once we have finished testing (500 hours), we expect a different user every week to process data. The core team would be of 4 people involved in data analysis. Industry partnership: We have GE-Global Research, Air Force Research Laboratory as our main industrial partners and many more of our users from academia and industry will make use of the computation power. Project URL: Requested allocation: 980400 Q1: 230400 Q2: 250000 Q3: 250000 Q4: 250000 Justification: The software used to take about a month on a single CPU to run once and this time has been brought down to 3 minutes using the following improvements: 1. We have used the SWIFT parallel scripting language for parallelization. Even on the level of a single diffraction image, we parallelize the computation by 4x to achieve speedup. Furthermore, each grain is processed in parallel with no inter-process communication. 2. All the programming is done in C99. 3. The data volume is reduced from about 4-10 GB to about 100 MB. This has significantly eased data I/O problems. Storage requirements: 2 TB of data generated over the course of a day should be stored temporarily. We will accordingly clean up as new data is generated. The requester has used undetermined amount hours of their initial startup project. In addition to approving an initial amount, please specify a Category and Subcategory for this project. For a list of the current selection of approved categories, please see: https://wiki.lcrc.anl.gov/wiki/Processes/Categories Once the Allocation committee has approved the project, please go to the Project Management page to create it: https://accounts.lcrc.anl.gov/projects.php Thank You, The LCRC Accounts System

1 0

Re: [allocations-admins] New LCRC report submitted for glassforms
by John Valdes 22 Oct '14

22 Oct '14

glassforms has turned in their FY14 report: https://www.lcrc.anl.gov/jazz/CSAC/project-reports-2014/Progress/glassforms… Turning on their allocation. John

1 0

New LCRC report submitted for glassforms
by support＠lcrc.anl.gov 22 Oct '14

22 Oct '14

Full project report at https://accounts.lcrc.anl.gov/reporting/project/2014/?project=glassforms

1 0

New LCRC report submitted for SHARP_ARC
by support＠lcrc.anl.gov 21 Oct '14

21 Oct '14

Full project report at https://accounts.lcrc.anl.gov/reporting/project/2014/?project=SHARP_ARC

1 0

[LCRC Accounts] Project Request: gm_gasexchange
by accounts＠lcrc.anl.gov 13 Oct '14

13 Oct '14

Hello, A new project on the LCRC cluster has been requested. Please forward the information on to the LCRC Allocation sub-committee. Applicant's name: Sibendu Som Applicant's institution: ANL Applicant's division: ES Project Name: gm_gasexchange Project title: Large Eddy Simulation of IC Engine Flows Associated funding: LDRD funding for Argonne, GM funding for GM and Univ. of Michigan Other Systems: GM R&D Science: To enhance engine fuel efficiency and lower engine emissions, new strategies such as multiple injections, stratified charge direct injection, and high levels of exhaust gas recirculation offer great potential. However, these strategies also render combustion stability issues and cause higher cycle-to-cycle variations. High cycle-to-cycle variations can lead to misfire as well. This is undesirable from engine robustness, efficiency, and emissions standpoint. The cycle-to-cycle conditions are significantly impacted by engine in-cylinder flows generated during intake and exhaust processes. This is determined by geometries of the intake port, intake and exhaust valves and combustion chamber shape including piston bowl. Computational fluid dynamics (CFD) and large eddy simulation (LES) approaches are proposed to investigate the cycle-to-cycle variations. The specific objective is to determine the best practice for performing LES simulations of IC engine f lows, and to provide design guidelines for engineers to effectively use the LES tools to accelerate new engine designs. Project description: CFD with RANS (Reynolds-Averaged Navier-Stokes) approach has been applied to engine development for many decades and are still the main tool for engine design today. Due to solving ensemble or phase-averaged mean quantities, RANS is very much limited for average cycles, hence cannot capture cycle-to-cycle variations. The high fidelity DNS (direct numerical simulation) is too computationally intensive and is still not feasible for full engine simulations in the near future. By solving spatially filtered Navier-Stokes equation, LES has been highlighted to have the potential for predicting the details of the transient in-cylinder flows and offer the essential information regarding the nature of cycle-to-cycle variations a decade ago. However, the computational requirements and need for multiple-cycle simulations required by LES hindered its further research and application for IC engines. Together with great leap of computer power in past decade, Argonn e’s and GM R&D’s expertise in LES calculations, and availability of high-fidelity experimental data from University of Michigan (funded by GM R&D), this proposal aims at establishing the best engineering practices for LES calculations in IC engines. The increased predictive capability with LES will enable more efficient and robust engine design. GM R&D has developed many engineering CFD practices for RANS and LES simulations. One example is the numerical setup of valve opening and closing events for in-cylinder flow simulations. In the past, pending on the meshing topology used, a minimum valve lift together with a crank angle position is arbitrarily chosen to mimic the valve opening and closing events. In most cases, the minimum valve lift is determined mainly from solution stability and simulation cost considerations. This is true especially when the meshing topology involves non-orthogonal skewed meshes. A lack of accurate flow field measurements for validation in the valve vicinity also hinders progress in this area. Recent meshing advancements in CONVERGE in conjunction with the availability of accurate high speed PIV data render a new opportunity for such study. Using these toolsets, GM R&D has established a strategy for accurate predictions of engine in-cylinder flows for RANS simulations. The next ste p is to test this RANS methodology for LES simulations. With this proposal, the following main components of LES for engine flows will be employed: (1) Use dynamic structure sub-grid scale (SGS) turbulence model on unresolved small flow scales; (2) Use the best CFD setup derived from RANS simulation on minimum valve lift, opening /closing crank angle position and minimum mesh size in the valve seat region; (3) Use a 2nd-order convective scheme; (4) Evaluate model accuracy against available measured data from University of Michigan; (5) Explore different data analysis methods like POD on LES results. In summary, this proposal will be executed in a manner such that CFD simulations will collaborate closely with experimental diagnostics. The high-speed PIV data for the TCC-III engine from University of Michigan will be available for LES model validation. The ultimate goal will be to use LES for IC engine simulations for in-cylinder processes involving engine flow, spray, combustion, and soot emissions. The specific methods used for these flow simulations with moving boundaries will include: 1) Finite volume schemes 2) RANS turbulence models 3) CONVERGE software is being used - 2 base and unlimited child licenses available Industry partnership: General Motors - Research and Development Group Project URL: http://verifi.anl.gov/ Requested allocation: 500000 Q1: 200000 Q2: 200000 Q3: 50000 Q4: 50000 Justification: The simulations will be performed with CONVERGE code. As reported by all the other projects of the PI (Sibendu Som), the scaling efficiency is quite good (more than 70%) up to 2000 processors for typical engine problems Storage requirements: 1 TB The requester has used undetermined amount hours of their initial startup project. In addition to approving an initial amount, please specify a Category and Subcategory for this project. For a list of the current selection of approved categories, please see: https://wiki.lcrc.anl.gov/wiki/Processes/Categories Once the Allocation committee has approved the project, please go to the Project Management page to create it: https://accounts.lcrc.anl.gov/projects.php Thank You, The LCRC Accounts System

1 0