Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Elia Merzari Project Name: SHARP_TH Division: NE Project title: SHARP Thermal-Hydraulics Development Associated funding: DOE-Nuclear Energy Other Systems: Intrepid. NE clusters. Science: To validate, develop and extend the thermal-hydraulics capabilities of the multi-physics nuclear reactor analysis platform package SHARP. This project comprises several activities in the SHARP thermal-hydraulics development and validation efforts in support of the Nuclear Energy Advanced Modeling and Simulation (NEAMS) program under DOE-NE: Item 1) Simulation of the SIBERIA experiment performed in Russia and part of a collaborative benchmark to test all shear stress predictions in case of partial blockage of the channels. The aim is to validate RANS models using STAR-CCCM+ using both LES data generated with NEK and available experimental data. Item 2) Using the SHARP toolset support activities related to SHARP in current NEUPs. In particular, the modeling of experiments related to nuclear fuel bundles responses to earthquake, currently being performed at George Washington University. This exercise is particularly important because it will help validated the SHARP toolset in a complex experiment. NEK (high fidelity CFD code) will be coupled through MOAB with DIABLO (a FE stress analysis code). Item 3) This project will support a potentially funded NEUP related to design of completely passive spent fuel pools. Both issues are key in addressing the issues raised by a Fukushima type scenario. Item 4) Using the commercial CFD code STAR-CCM+ we plan to study in further detail the Argonne MAX experiment. In previous years, preliminary calculations for the MAX experiment were performed. A data set has now been produced and validation studies on our tools will be performed. Project description: Item 1, (STAR-CCM+ simulation of “SIBERIA”) STAR-CCM+ employs the finite volume formulation of the Reynolds-Averaged Navier-Stokes (RANS) method to determine the local three-dimensional velocity distribution in the channel. The results are compared with available experimental data for Validation. Item 2, (Simulation of the George Washington University experiment) This experiment present complex fluid structure interaction (FSI) between fluid and water induced by a shaker table. A wide variety of options will be tested for FSI (i.e., NEK coupling with DIABLO). This calculations will be typically large and involve transient analysis. To simulate highly unstable turbulent flows the solver utilizes an LES approach, where the filtered equations of motion are solved. All scales smaller that the filter size, are modeled using a dynamic subgrid scale (SGS) model. The stress analysis and the corresponding moving boundaries and deformations will be handled by the coupling with DIABLO through MOAB. Item 3 (Support of Purdue NEUP) As part of this activity, several design iterations will be performed to design a reliable and. completely passive (i.e., operating in natural convection) spent fuel pool. Each iterations will be studied using the CFD code STAR-CCM+ in RANS mode. A sensitivity analysis on modeling options will be performed. Item 4, (MAX experiment support) Using insight from previous work steady state and full transient URANS calculations will be performed using the inlet data now available. Results will be compared with available experimental data and used to determine and guide the future experimental matrix and possibly an international benchmark. Access to the Fusion cluster would be crucial in performing all the proposed work. The STAR-CCM+ code and the Nek5000 code have already been installed on Fusion. Compilers, such as the Intel Fortran Compiler necessary to compile users subroutines are already installed. All jobs run on Fusion will be run in parallel. All calculations proposed require large computational resources due to the size of the problem. Two to Four users are expected for this project. Project URL: Current FY Hours Used: undetermined amount New FY Requested allocation: 814 Q1: 114 Q2: 200 Q3: 400 Q4: 100 Justification: At Argonne and elsewhere, the CFD codes STAR-CCM+ is routinely applied to simulations that use up to thousands of cores, and good parallel performance has been observed. Optimal performance is usually observed in steady state analysis is observed usually with 50,000 - 100,000 meshes per core. Nek5000, has proved scalability even below the ranges proposed. The typical size of the problems proposed are the following: Item 1) SIBERIA experiment : 10,000,000 meshes in steady state. This would require typically 128 cores to be performed. Considering that a problem would likely converge within 24h of continuous calculation and that we would like to perform at least 20 production runs with 3,000 h of development this would translate in approximately 64,000 core-hours. Item 2) GW experiment : Fusion will be used here for test runs rather than production runs. These will involve several (around 30) runs with 64 cores for 48 hours. For a total of approximately 100,000 hours. Item 3) Support of the Purdue NEUP: 100,000,000 for the largest cases in steady state. A large number of design testing through CFD will be performed. These will involve approximately 15 runs (only final production runs will be performed here) on 32 nodes for 24 hours. Costing approximately 400,000 including development time. Item 4) MAX simulation: 5,000,000 meshes. This would require typically 64 cores to be performed. Long transient are likely to be necessary with 200h of consecutive calculations. Considering that we would like to perform at least 20 production transient runs, with 3,000 h of development this would translate in approximately 250,000 core-hours. The uneven distribution of the request is due to the likelihodd that item n.3 work will be concentrated in the 3rd quarter due to testing of several design options. Thank You, The LCRC Accounts System