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

[LCRC Accounts] Yearly Allocation Request for GAS_injection
by accounts＠lcrc.anl.gov 30 Aug '17

30 Aug '17

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Lorenzo Bartolucci Project Name: GAS_injection Division: ES Project title: High-Pressure Gaseous Fuel Injection Research – in-nozzle flow and jet evolution Associated funding: Other Systems: Science: Current state of the art of fuel injection systems provides several studies about the in-nozzle analysis of liquid fuel injectors but a lack in knowledge is present in the gaseous fuel injector in-nozzle flow that determines the gas jet characteristics. The main goal of this project is to deeply understand the phenomenon occurring into the gaseous injector during the transients – both opening and closure – and how those are related with the following jet evolution, in order to obtain detailed information that can be used as input for ICE simulations. The analysis will be carried out in collaboration with the Argonne x-ray diagnostic group and the University of Rome Tor Vergata. Project description: This project focuses on investigating the in-nozzle flow for a gaseous inward opening injector and the consequent under-expanded jet evolution. This is aimed to provide a better understanding of the local phenomenon and how those affect the typical behavior of the under-expanded gaseous jet (such as Mach disk location and width, local thermo-fluid-dynamics properties and mixing with the surrounding mixture). At the current stage, an optimum mesh strategy and size have been found as a compromise between the numerical accuracy and the computational costs, the computational approach has been well set as well, based on an LES framework. The X-ray radiography data has been used as benchmark for validation. Preliminary results have been obtained for the steady state simulation showing a good agreement with the experimental data. However there is room for significant improvement of numerical results in terms of jet spread and mass distribution. In particular two different locations of high mass are present downstream of the Mach location that are not present in the experimental data and require further investigation. Furthermore, future work would aim to represent also the opening and closing transient to provide a detailed instrument of analysis of the high injection phenomena occurring during these unsteady phases. Simulations will be performed with the commercial code CONVERGE and computational grids of the order of 30 millions of cells which will require the use of 192-320 cores depending on the specific case. The project has one member (barlo) working full-time on it, with 1-2 cases running regularly. With an average estimate of 256 cores per run and 2 cases the expected amount of core-hrs needed for this project for each quarter is 96 cores x 24 hrs x 2 runs x 90 days = 415,000. Considering some queue and post-processing time our request is for 1,500,000 hrs total, equally distributed for each quarter. Industry partnership: Project URL: Current FY Hours Used: undetermined amount New FY Requested allocation: 1500000 Q1: 375000 Q2: 375000 Q3: 375000 Q4: 375000 Justification: Efficiency: Recent scalability test (April 2017) performed on Blues provided following results: Simulation of gas jet from inward opening pintle. Mesh size = 20,000,000 cells 1 node (16 cores) – walltime = memory problems 2 nodes (32 cores) – walltime = 45min 15sec - 100 % 3 nodes (48 cores) – walltime = 33min 18sec - 90 % 6 nodes (96 cores) – walltime = 18min 58sec - 79 % 12 nodes (192 cores) – walltime = 10min 52sec - 70 % Storage requirements: I kindly ask to keep the project folder to the current size of 4TB. Thank You, The LCRC Accounts System

1 0

New LCRC report submitted for GAS_injection
by support＠lcrc.anl.gov 30 Aug '17

30 Aug '17

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

1 0

New LCRC report submitted for GNR-Au
by support＠lcrc.anl.gov 30 Aug '17

30 Aug '17

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

1 0

New LCRC report submitted for Materials_Genome_SCs
by support＠lcrc.anl.gov 30 Aug '17

30 Aug '17

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

1 0

[LCRC Accounts] Yearly Allocation Request for anl-dcc
by accounts＠lcrc.anl.gov 30 Aug '17

30 Aug '17

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Tsung-Shung Lee Project Name: anl-dcc Division: PHY Project title: Dynamical coupled-channel analysis of hadrons at Argonne Associated funding: DOE Other Systems: NERSC Science: This project will develop theoretical approaches for studying the spectrum and structure of excited baryon states within the fundamental theory of strong interaction; Quantum Chromodynamics (QCD). The results will provide information for advancing our understanding of strong interactions and structure of visible matter in universe. Project description: In 2016-2017, we have applied the ANL-Osaka 8-channel coupled-channel model of pi-N and gamma-N reactions to extract the N-N*(nucleon resonances) transition form factors from the Jlab's data of pion electro-production reactions. The results for the well-established N* with masses lower than about 1.6 GeV are robust for testing the predictions from LQCD and various hadron models. A paper describing our results is being prepared for a publication in Physical Review. Our analysis has also demonstrated that the data with pi-N, eta-N, and KN production channels are not sufficient to obtain reliable results for the higher mass N* (> 1.6 GeV) which couple strongly with the two-pions and omega production channels. In 2017-2018, we plan to extend our code to include the data of these two channels in our analysis, as a step to improve our results for the higher mass N* and also as a necessary step to analysize the next-generation data which will soon be obtained from the new experiments to be performed at JLab with 12 GeV upgrade and J-PARC in Japan. Accordingly, we will need much more computation resources than what were needed in the analysis in 2013-2017. The information on the higher mass N* is essential to establish the spectrum of baryons within QCD. We have used ANL-Osaka model to investigate the electroweak pion production reactions on the deuteron. Our results reveal a substantial final state interaction effects and hence must be taken into account in extracting the photo-neutron and the neutrino-neutron cross sections from the data on the deuteron target. Our predictions have provided information to assess the validity of various cut-procedures used by experimentalists to analyze the data from JLab and Mainz in Germany. We will continue this effort in 2017-2018, focusing on the polarization observables which are essential in determining the excitations of neutron to N* states. The neutron-N* transitions are needed to determine the isospin structure of the N* states for testing LQCD and various hadron models. A model for the Kd¯ → πY N reactions with Y = Λ, Σ is developed, aiming at establishing the low-lying Λ and Σ hyperon resonances through analyzing the forthcoming data from the J-PARC E31 experiment. The off-shell amplitudes generated from the ANL-Osaka dynamical coupled-channel model are used as input to the calculations. The predictions have been made for the forthcoming experimental tests. In particular, the question on the existence of a low-lying J(P) = 1/2(+) Σ resonance with a pole mass MR = 1457 − i39 MeV will be resolved. In 2017-2018, we will extend this investigation to also investigate the possible di-baryon states with strangeness S=-2 associated with K(+)d → K(-)YY. Industry partnership: Project URL: Current FY Hours Used: undetermined amount New FY Requested allocation: 470000 Q1: 200000 Q2: 100000 Q3: 90000 Q4: 80000 Justification: Storage requirements: Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Yearly Allocation Request for navistar_tbc
by accounts＠lcrc.anl.gov 29 Aug '17

29 Aug '17

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Sibendu Som Project Name: navistar_tbc Division: ES Project title: CFD study of dimensional variations on charge preparation and heat-release management in a multi-injector diesel engine Associated funding: In process Other Systems: TBD Science: Improving the fuel efficiency of diesel-engine can bring significant socio-economic benefits since, a significant fraction of US freight is transported by diesel-engine powered trucks. For a given engine architecture, the fuel-injection, air and after-treatment system determines the emissions and efficiency. A multi-hole, single injector per cylinder represents the traditional fuel-injection system. In this engine, fuel-distribution and mixing are restricted by the spray penetration and the strength of the in-cylinder charge motion. These limitations can be overcome by application of innovative concept of multiple-injector per cylinder design. This system provides enhanced flexibility for the fuel-air mixing process and the heat-release rate management for better performance and emissions. However, manufacturing process variations and tolerances in the injector and combustion system can impact the emission and performance. Therefore, this project proposes to computa tionally evaluate the effects of key manufacturing tolerances on the charge preparation process and thus on the engine performance Project description: A multi-hole, single injector represents the traditional fuel-injection system found in today’s production diesel engine. The injector is completely characterized by parameters such as the flow-rate, cone-angle, number of injector holes and the injector placement location. A common-rail injector can also have up to 5 injection events per cycle. In addition to the fuel-injection system variables, the inputs to the combustion system include parameters such as the swirl motion, intake and exhaust valve characteristics and the combustion system profile. Given such a large number of inputs, the combustion system has been traditionally optimized using computation fluid analysis (CFD). In the present study, the level of complexity is significantly higher as multiple number of injectors are implemented per cylinder. The central injectors are complemented with peripheral injectors. The peripheral injectors given their intrinsic location have different flow-char acteristics than the main injector and need to treated as a separate variable for the optimization process. The casting of the original cylinder head architecture will be extensively modified along with the machined features to accommodate the new injectors while providing the necessary cooling and durability to the engine. Given the spatially distributed nature of the heat-release process the new combustion is more susceptible to the manufacturing variations of various components, particularly cylinder head and the fuel injector. Therefore, accounting for all the design space for optimization needs to be carefully analyzed by high-fidelity engine simulations. Navistar would like to work with the Argonne National Laboratory (ANL) CFD team led by Dr.Sibendu Som for the analysis. Preliminary estimates indicate that several millions of core-hours of supercomputing will be needed to understand, optimize and characterize such a complicated combustion system. ANL with a proven tr ack record of implementing software innovations for supercomputing advances in the area of internal-combustion engine design and with a close proximity to the Navistar facility lends itself to be an ideal project partner for this endeavor. Industry partnership: TBD Project URL: http://www.transportation.anl.gov/engines/multi_dim_model_home.html Current FY Hours Used: undetermined amount New FY Requested allocation: 600000 Q1: 150000 Q2: 150000 Q3: 150000 Q4: 150000 Justification: Our recent publication (J. Kodavasal, K. Harms, P. Srivastava, S. Som, S. Quan, K.J. Richards, M. Garcia, “Development of stiffness-based chemistry load balancing scheme, and optimization of I/O and communication, to enable massively parallel high-fidelity internal combustion engine simulations,” Journal of Energy Resource Technology; JERT-16-1022, 2016) together with MCS and Convergent Science discusses the improvements to the Converge tool that has resulted in significant improvement in scaling. Currently, we are able to scale the code up to 4096 processors on Mira with about 70% scaling efficiency for a fixed mesh size. This was achieved due to the implementation of: (1) MPI I/O, (2) Improved Communication, (3) METIS load balancing scheme, (4) Development of a new chemistry load balancing scheme. The above changes are not only expected to benefit calculations on Mira but also help on computing clusters like Blues. Storage requirements: 5 TB Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Yearly Allocation Request for spray-modeling
by accounts＠lcrc.anl.gov 29 Aug '17

29 Aug '17

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Sibendu Som Project Name: spray-modeling Division: ES Project title: Modeling of nozzle flow and spray for heavy duty injectors Associated funding: Aramco Other Systems: Aramco Science: This project will use a fully Eulerian modeling of internal nozzle flows by means of Computational Fluid Dynamics (CFD) simulations performed with a software called CONVERGE. The focus will be placed on the modeling of all the physical phenomena that are associated with the high pressure injection of fuel in the combustion chamber of internal combustion engines. The in-nozzle cavitation will be modeled through the Homogeneous Relaxation Model (HRM) and will mostly focus on gasoline-like fuels such as Naphtha. Project description: The project will continue an ongoing activity and will address the evaluation of the behavior of gasoline-like fuels in internal nozzle flows when realistic injector geometry and needle motions obtained through x-ray imaging are involved. The project will apply previously validated methodologies and setups to a multi-hole heavy-duty injector. Experimental validation for the Diesel and Naphtha fuels by means of comparison against CFD results will be performed. Particular attention will be dedicated to the prediction of orifice-to-orifice and shot-to-shot variabilities associated with the use of real nozzle geometries. A possible outcome will be the development of an erosion model to assess the effect of cavitation on the nozzle internal surface. The model will identify those zones of the nozzle that could be more affected by the cavitation occurrence. Final key point of the collaboration will be to incorporate all the learnings from the internal nozzle fl ow simulation into actual combustion system optimization study. Industry partnership: Aramco Project URL: http://www.transportation.anl.gov/engines/multi_dim_model_home.html Current FY Hours Used: undetermined amount New FY Requested allocation: 1000000 Q1: 250000 Q2: 250000 Q3: 250000 Q4: 250000 Justification: Our recent publication (J. Kodavasal, K. Harms, P. Srivastava, S. Som, S. Quan, K.J. Richards, M. Garcia, “Development of stiffness-based chemistry load balancing scheme, and optimization of I/O and communication, to enable massively parallel high-fidelity internal combustion engine simulations,” Journal of Energy Resource Technology; JERT-16-1022, 2016) together with MCS and Convergent Science discusses the improvements to the Converge tool that has resulted in significant improvement in scaling. Currently, we are able to scale the code up to 4096 processors on Mira with about 70% scaling efficiency for a fixed mesh size. This was achieved due to the implementation of: (1) MPI I/O, (2) Improved Communication, (3) METIS load balancing scheme, (4) Development of a new chemistry load balancing scheme. The above changes are not only expected to benefit calculations on Mira but also help on computing clusters like Blues and Bebop. Storage requirements: 5 TB Thank You, The LCRC Accounts System

1 0

[LCRC Accounts] Yearly Allocation Request for foam-EngineProcesses
by accounts＠lcrc.anl.gov 29 Aug '17

29 Aug '17

Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Sibendu Som Project Name: foam-EngineProcesses Division: ES Project title: SSpray-Wall Interaction Process for Internal Combustion Engines Application: a Numerical Investigation with Global and Local Characterization of the Spray Morphology Associated funding: DOE-Vehicle Technologies Office Other Systems: Michigan Technological University Science: This project will use the Eulerian-Lagrangian approach for Computational Fluid Dynamics (CFD) simulations through a software called CONVERGE. In particular the focus will be placed on modeling gas exchange, mixture and spray formation and on the investigation of the spray-wall interaction. Project description: This project involves Argonne National Laboratory as part of a multi-institutional team including also Michigan Technological University (MTU) and University of Massachussets at Dartmouth (UMassD). The goal of this work is to develop and validate an advanced spray-wall interaction and associated film formation and vaporization modeling approach via application of a VOF method with an integrated evaporation sub-model (eVOF). In the previous year, the computational effort was directed toward the validation of the state-of-the-art CFD models and to the implementation of post-processing tools for the extraction of spray morphology information near the impinging point. The implementation of non-evaporating models is already ongoing and will be completed in the first quarter of the current year. The following step will include a film vaporization sub-model whose development and validation will rely on the results of the DNS analysis of spray-wall impingement. The aim is to achieve accurate predictive simulations of sprays and their impingement without need of extensive parameter tuning. The project will also develop sub-models for droplet formation characteristics post-wall impingement via detailed DNS and LES models which are supported by accurate experimentation. Targeted experimentation of the spray-wall interactions and liquid wall film under conditions matching the thermodynamic charge state and surface temperatures to those of engines will be performed to support development and validation of the spray-wall interaction models. Industry partnership: None Project URL: http://www.transportation.anl.gov/engines/multi_dim_model_home.html Current FY Hours Used: undetermined amount New FY Requested allocation: 600000 Q1: 150000 Q2: 150000 Q3: 150000 Q4: 150000 Justification: Our recent publication (J. Kodavasal, K. Harms, P. Srivastava, S. Som, S. Quan, K.J. Richards, M. Garcia, “Development of stiffness-based chemistry load balancing scheme, and optimization of I/O and communication, to enable massively parallel high-fidelity internal combustion engine simulations,” Journal of Energy Resource Technology; JERT-16-1022, 2016) together with MCS and Convergent Science discusses the improvements to the Converge tool that has resulted in significant improvement in scaling. Currently, we are able to scale the code up to 4096 processors on Mira with about 70% scaling efficiency for a fixed mesh size. This was achieved due to the implementation of: (1) MPI I/O, (2) Improved Communication, (3) METIS load balancing scheme, (4) Development of a new chemistry load balancing scheme. The above changes are not only expected to benefit calculations on Mira but also help on computing clusters like Fusion and Blues. Storage requirements: 5 TB Thank You, The LCRC Accounts System

1 0

New LCRC report submitted for pozdyn
by support＠lcrc.anl.gov 29 Aug '17

29 Aug '17

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

1 0

New LCRC report submitted for Meso
by support＠lcrc.anl.gov 29 Aug '17

29 Aug '17

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

1 0