Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Justin W Thomas Project Name: NEAMS_Safety Division: NE Project title: CFD Support for Sodium-cooled Fast Reactor Safety Associated funding: DOE-NE Advanced Reactor Concepts (Note: Not actually NEAMS, the project name is a misnomer.) Other Systems: The NE Division cluster Eddy. Calculations will be shared between the Eddy and Fusion clusters, depending on availability. The requested allocation is roughly half of the actual expected computational expense associated with this effort. Science: The objective of this effort is to validate the use of computational fluid dynamics (CFD) for modeling and heat transfer in sodium-cooled fast nuclear reactor (SFRs) safety analysis by analyzing two transient tests performed at the Experimental Breeder Reactor-II (EBR-II) in the late 1980s. The Argonne liquid-metal reactor systems safety code SAS4A/SASSYS-1 is currently the flagship U.S. code for simulating transients in liquid-metal cooled nuclear reactors. There is potential to improve the 1-D thermal-hydraulics model of the primary and intermediate heat transport systems (outside the reactor core) by integrating 3-D CFD models in selected components. Of particular interest is the modeling of large pools during transients where the flow rate is decreased rapidly due to loss of pumping power. Thorough validation of CFD would be required before including CFD analysis in the safety case for an SFR license application, and the EBR-II tests provide a very rare experi mental data set of an SFR responding to severe accident initiators. A CFD model of the EBR-II reactor coolant tank will be integrated into the SAS4A/SASSYS-1 model of the core and the remaining cooling systems, and compare the computational results to the measured data. The successful completion of this project will include an evaluation of the applicability of CFD to such flows, as well as guidelines on CFD model development. Due to the size of the EBR-II coolant tank and the transient nature of the simulation, such an effort would be impossible without high-performance computing platforms like Fusion. Project description: EBR-II was a 62.5 MWth uranium-plutonium metal-fuelled sodium-cooled fast reactor operated by Argonne at the Argonne-West site in Idaho. The Shutdown Heat Removal Tests, specifically SHRT-17 and SHRT-45, tested the safety performance of EBR-II under severe accident initiators. SHRT-17 involved a complete loss of all pumping power to the plant while operating at full power and flow, followed by emergency shutdown (“scram”) of the reactor. SHRT-45 was even more severe: again pumping power was lost at full power and flow, but the emergency reactor shutdown system was not engaged. In both cases the reactor demonstrated its capability to remain cool and safe. The temperature and flow measurements of the coolant during the test provide a valuable data set of a nuclear reactor responding to rather severe accident initiators. During FY11, a preliminary model of the EBR-II reactor coolant tank was built with the commercial CFD code STAR-CCM+ and successfully linked to the existing SAS4A/SASSYS-1 model of the primary coolant system and reactor core. The first 120 seconds of the SHRT-17 test were analyzed on a mesh with 1.5 million cells using a maximum time step size of 0.025 sec. The feasibility of performing such integral calculations was shown, and the successful communication between the STAR-CCM+ and SAS4A/SASSYS-1 was demonstrated. Several parametric studies on discretization, parallel scalability and numerical behavior were performed. Unfortunately these studies indicate that the mesh refinement and tightening of convergence limits are required. In FY12, the project is now more mature. The initial development effort for linking SAS4A/SASSY-1 and STAR-CCM+ was completed for communication of flow through boundaries at the STAR-CCM+ model. Parametric studies have been performed to estimate the required computational expense of the SHRT analyses. Specific objectives for the continued effort in FY12 are: 1. Include heat transfer from solids in the coupling between SAS4A/SASSYS-1 and STAR-CCM+. This will require some software development during the first quarter. 2. Develop more detailed geometric models of the coolant tank that includes a representation of all of the large submerged components. This will require a finer mesh relative to the FY11 effort. This will require model development effort during the first quarter. 3. Develop an optimal mesh that captures both the laminar flow in lower regions of the plenum and turbulent flow near the flow boundaries. 4. Employ the volume-of-fluid method to model the sodium/gas free surface at the top of the tank. 5. Perform analysis of SHRT-45 and compare the predicted axial temperature distribution in the CFD model to the experimental measurements. In particular, use the measured axial temperature profile in SHRT-45 as a metric to evaluate the model, and apply any model revisions to the SHRT-17 model. 6. Perform necessary sensitivity analyses to determine the appropriate physics models. Project URL: Current FY Hours Used: undetermined amount New FY Requested allocation: 400000 Q1: 50000 Q2: 100000 Q3: 100000 Q4: 150000 Justification: The requested allocation in the FY12 renewal is substantially increased compared to the initial allocation request in FY11. This project began in FY11 and required substantial lead time to acquire detailed geometric data about the EBR-II plant, become proficient with the CAD tools in STAR-CCM+, develop the initial CFD model, and develop the software required to couple with SAS4A/SASSYS-1. Thus “production” runs of the coupled model were not performed until nearly the end of the 3rd quarter. In FY12, the project is much more mature and production runs would be expected in the 1st quarter. As mentioned in the Project Description, there is still some development effort required in the FY12 effort, which is the reason for the uneven allocation requests from quarter to quarter. Based on the calculations performed on 64 Fusion cores during FY11, the estimated computational expense for a full 250-second transient using the required tight convergence tolerances will be in excess of 80,000 core-hours. The requested allocation of 400,000 core-hours is sufficient for 5 such calculations, with the hope that parametric studies can be performed on smaller models. Unfortunately, much of the computational burden is caused by the transient nature of the simulation (and the inability to parallelize time), so these calculations will likely be performed on 32-128 cores for several days at a time. According to scalability studies performed in FY11, approximately 50,000 mesh cells/Fusion core is fairly optimal. Smaller models will be pursued to test numerics and for the development efforts. Thank You, The LCRC Accounts System