August 2017 - Allocations-admins - mailman3.cels.anl.gov

New LCRC report submitted for SnI4
by support＠lcrc.anl.gov 31 Aug '17

31 Aug '17

Full project report at https://accounts.lcrc.anl.gov/reporting/project/2017/?project=SnI4

1 0

New LCRC report submitted for grbdft
by support＠lcrc.anl.gov 31 Aug '17

31 Aug '17

Full project report at https://accounts.lcrc.anl.gov/reporting/project/2017/?project=grbdft

1 0

New LCRC report submitted for PACC
by support＠lcrc.anl.gov 31 Aug '17

31 Aug '17

Full project report at https://accounts.lcrc.anl.gov/reporting/project/2017/?project=PACC

1 0

New LCRC report submitted for KTP
by support＠lcrc.anl.gov 31 Aug '17

31 Aug '17

Full project report at https://accounts.lcrc.anl.gov/reporting/project/2017/?project=KTP

1 0

[LCRC Accounts] Project Request: DFT_UMo
by accounts＠lcrc.anl.gov 31 Aug '17

31 Aug '17

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: Zhigang Mei Applicant's institution: ANL Applicant's division: NE Project Name: DFT_UMo Project title: Fundamental thermophysical properties of U-Mo alloys by DFT Associated funding: DOE NNSA Other Systems: Science: U-Mo based alloys fuels are the most promising fuel for the future high-performance reactors with low-enriched uranium (LEU). As part of a global effort on nuclear non-proliferation, two types of U-Mo fuels with different Mo concentrations, i.e., U-7wt%Mo (U-7Mo) dispersion fuel and U-10wt%Mo (U-10Mo) monolithic fuel, are currently being developed by the European High Density LEU Fuel program and the US High-Performance Research Reactor Fuel Development program (HPRRFD), respectively. However, the accelerated swelling of U-Mo fuel at high burnup raises concern about its applications at high power and burnup. Due to the low solubility of fission gas atoms (mostly Xe and Kr) in U-Mo alloys, fission gases precipitate into gas-filled bubbles, which leads to the fuel swelling as well as overall loss of mechanical strength. Understanding the process of nucleation, growth, migration and coalescence of fission-gas bubble is, therefore, a key aspect to an acc urate prediction of the fuel swelling at high burnup. Mesoscale simulations using approaches such as phase field method plays an important role in the prediction of gas bubble evolution kinetics in U-Mo fuels. However, there are very few studies of the fundamental thermophysical properties of the U-Mo alloy system, which are required for the mesoscale simulations of gas bubbles in U-Mo fuels. Due to the difficulty of treating the f-electrons in uranium and the complexity of describing the disordered structure of the U-Mo alloys, there is no study of the elastic and surface properties of U-Mo alloys from either experiment or theory. To this end, we systematically investigate the elastic and surface properties of the U-Mo alloys using DFT calculations. Project description: We will use density functional theory (DFT) method as implemented in VASP to predict the elastic and surface properties of U-Mo alloys. The special-quasirandom structure (SQS) method as implemented in ATAT will be used to generate a set of system-independent 128-atom-bcc SQSs spanning the entire concentration range, i.e., five different concentrations, in order to simulate the binary random substitutional U-Mo alloys. DFT codes VASP has been extensively tested on LCRC machines, such as Blues and Bebop. Recent test of the latest version of VASP6 shows a good scalability using 16 Broadwell nodes or less on Bebop. My tests on Bebop show that a typical structural relaxation of 128-atom SQS of U-Mo takes 5 hours using 8 nodes and 1 hour for static calculations. For bcc U-Mo alloy, there are 6 symmetrically different surface orientations with a maximum Miller index of 2. For each surface orientation, eight different terminations will be calculated to get better statistics. Therefore, the total cpu-hours for surface energy calculations will be 36*8*6*6*8*5=400000 cpu-hours. In order to get the elastic properties of U-Mo alloys, stress-strain method will be used to calculated the elastic constants. 12 different sets of strains will be applied to the relaxed cubic U-Mo SQS structure to calculate the resultant stresses in order to obtain all the 21 stress tensors. 9 different concentrations of U-Mo alloys together with the two endmembers will be studied in this work. Therefore, the total cpu-hours for elastic property calculations will be 36*8*6*11=200000 cpu-hours. The total cpu-hours needed for this projects are expected to be 600000. Industry partnership: Project URL: Requested allocation: 600000 Q1: 150000 Q2: 150000 Q3: 150000 Q4: 150000 Justification: Recent tests on Bebop show that the latest version of VASP6 can be well scaled to up to 32 Broadwell nodes on Bebop. 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

New LCRC report submitted for Laser_ign
by support＠lcrc.anl.gov 31 Aug '17

31 Aug '17

Full project report at https://accounts.lcrc.anl.gov/reporting/project/2017/?project=Laser_ign

1 0

Re: [allocations-admins] [lcrc-core] Outstanding Allocations Questions 8/30/17
by Roberts, John E. 31 Aug '17

31 Aug '17

Comments: NEMphyValidation, 80K: Vijay would like 80K for each quarter of FY18. He wants 50K for Q4 now to get started. IP-ML, 250K: Created for 50K this quarter. BEM_Uncertainty, 55K in Q4: I see 4 quarters of time requested. I’m guessing this is just for time in FY18. REMM, 483K in Q4: Actually looks like they requested this for Q1. ATLAS-HEP-group, 100K in Q4: : I see 4 quarters of time requested. I’m guessing this is just for time in FY18. Thanks! -John On 8/31/17, 3:23 PM, "Bair, Raymond A." <rbair(a)anl.gov> wrote: IP-ML 50K for Q4 seems fine. ----------------------------------------- Ray Bair Argonne National Laboratory and the University of Chicago On 8/31/17, 3:19 PM, "Roberts, John E." <jeroberts(a)anl.gov> wrote: IP-ML wants 50K for Q4 and 250K through FY18. Pinging the rest. Thanks! -John On 8/30/17, 11:11 AM, "lcrc-core on behalf of Bair, Raymond A." <lcrc-core-bounces(a)lists.lcrc.anl.gov on behalf of rbair(a)anl.gov> wrote: LCRC, Since I am traveling today, here are what I see as the open issues for LCRC Allocations this week. Radix, request to add 500K. Ray sent to Allocations Committee for decision. 3DDES, early user/new project, requests interim 1M in Q4 on Bebop. Ray sent to Allocations Committee for decision. See new project requests below. These did not come in as “year allocation”. Do they want FY17 start? Or, do we need to move these to the FY18 allocation requests? Please check with PIs. NEMphyValidation, 80K IP-ML, 250K BEM_Uncertainty, 55K in Q4 REMM, 483K in Q4 ATLAS-HEP-group, 100K in Q4 Regards, Ray _______________________________________________ lcrc-core mailing list lcrc-core(a)lists.lcrc.anl.gov https://lists.lcrc.anl.gov/mailman/listinfo/lcrc-core

1 0

New LCRC report submitted for PbTe_thermoelectric
by support＠lcrc.anl.gov 31 Aug '17

31 Aug '17

Full project report at https://accounts.lcrc.anl.gov/reporting/project/2017/?project=PbTe_thermoel…

1 0

New LCRC report submitted for CMFD-RANS
by support＠lcrc.anl.gov 31 Aug '17

31 Aug '17

Full project report at https://accounts.lcrc.anl.gov/reporting/project/2017/?project=CMFD-RANS

1 0

[LCRC Accounts] Yearly Allocation Request for DFT_ALD_SiC
by accounts＠lcrc.anl.gov 31 Aug '17

31 Aug '17

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Zhigang Mei Project Name: DFT_ALD_SiC Division: NE Project title: Computational Design of Novel Precursors for Atomic Layer Deposition of SiC Associated funding: LDRD Innovate Other Systems: Science: The goal of the proposed work is to use a multi-scale computational approach (DFT+KMC) to down-select novel chemical precursors for atomic layer deposition (ALD) of SiC and identify the ideal process of ALD, and to verify the proposed precursors and reaction routines using experiments. In addition to gaining a better fundamental understanding of the underlying kinetics of the precursor-surface reaction, this research will help to identify new precursors and optimize deposition conditions for ALD of SiC at low temperatures. Low temperature ALD is preferred, as it will limit interdiffustion of the substrate and coating, reduce thermal stresses within the coating, and will maintain the initial desirable microstructure of zircaloy. The implications of the proposed activities can be extended beyond nuclear materials to other areas of interest, such as semiconductors. Project description: Computational screening of precursors for ALD SiC: Due to the large chemical space for silicon and carbon containing molecules and complex nature of ALD surface reactions, our screening procedure of potential precursors for ALD SiC is composed of the following three steps: (1) Evaluate the thermodynamic feasibility of the formation of SiC from various Si and C precursors; (2) Predict the energetics and kinetics of the ALD half-reactions for the proposed precursors; (3) Validate the down-selected precursors by experiments. From the PubChem Compound Database, we found more than 900,000 silicon-containing molecules and significantly more molecules for carbon precursors. By filtering the PubChem database with proper descriptors, such as molecular weight and boiling point, there are about 200,000 silicon-containing molecules left. As the first step of the screening procedure, we are constructing a large chemical reaction database of silicon and carbon-contain ing molecules with thermodynamic properties calculated by high-throughput DFT calculations. To automate the construction of the database, we have developed a computational framework using python, which can (1) get molecular information from the PubChem Database, (2) extract Canonical SMILES description for each molecules by OpenBabel, (3) obtain initial geometries as input file for Gaussian by OpenBabel, (4) perform DFT B3LYP geometry optimization calculation by Gaussian, (5) perform TD-DFT calculations with 6-31+G(d) for optimized geometry using B3LYP/6-31G(d), and (6) upload TD-DFT/DFT B3LYP input/output files to online database. With everything automated, we expect that our database will have more than 100,000 entries of silicon and carbon-containing molecules by the end of FY17. With the large chemical reaction database, we will apply machine learning algorithm to develop chemical reaction prediction models using novel fingerprint for molecules. The developed machine lea rning model can be used to screen the entire PubChem Compound database for all the possible precursors. In the second year, we will continue to add more entries for our silicon and carbon molecule database using high-throughput DFT computations in order to identify potential new precursors for SiC using machine learning. Our initial tests show that the computational time for a typical precursor pair such as SiCl4 and CCl4, takes about 100 cpu-hours. Due to the huge number of precursors for Si and C, we expect to calculate 5000 pairs of precursors. The total computational time we requested for FY18 is 500,000 cpu-hours. Industry partnership: Project URL: Current FY Hours Used: undetermined amount New FY Requested allocation: 500000 Q1: 125000 Q2: 125000 Q3: 125000 Q4: 125000 Justification: Efficient DFT code Gausian will be used to do high-throughput DFT calculations. Storage requirements: 1 TB Thank You, The LCRC Accounts System

1 0