Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Sibendu Som Project Name: CAT-engine-modeling Division: ES, MCS Project title: Simulation of Internal Combustion Engines with High-Performance Computing Tools Associated funding: Other Systems: CATERPILLAR Cluster, CONVERGENT Science Cluster Science: The primary goal of this project is to study scaling issues using the commercial code CONVERGE with detailed chemistry. As a part of this exercise, complete 3-D simulation of a single cylinder diesel engine will be conducted. A series of chemistry models with increasing complexity (species and reactions) will be included. The objective is to find the optimum reaction mechanism and cores that will provide useful design information not obtained from 2-D and reduced chemistry models. Project description: Traditional Lagrangian spray modeling approaches for internal combustion engines (ICEs) are highly grid-dependent due to insufficient resolution in the near nozzle region. This is primarily because of inherent restrictions of volume fraction with the Lagrangian assumption together with high computational costs associated with small grid sizes. A state-of-the-art grid-convergent spray modeling approach was recently developed and implemented by the co-PIs in CONVERGE software. The key features of the methodology include Adaptive Mesh Refinement (AMR), advanced liquid-gas momentum coupling, and improved distribution of the liquid phase, which enables use of cell sizes smaller than the nozzle diameter. This modeling approach was rigorously validated against non-evaporating, evaporating, and reacting data from the literature. The current numerical study focused on further demonstration of the grid-convergent modeling approach for simulating a single-cylinder Caterpillar compression ignition engine. Grid-Convergence of several critical engine performance and emission parameters such as pressure, heat release rate, ignition delay, and NOx and soot emissions was investigated. The goal of this study is to recommend grid-settings for accuracy/runtime trade-off for industrial partners. The use of grid-convergent modeling approach is possible by using grid sizes smaller than the nozzle diameter (d). A minimum resolution of “~0.5d” resulted in a peak cell count of 50 million. HPC in such an environment is quite challenging due to stringent demands on the load balancing algorithm. Traditional load balancing approach in CONVERGE resulted in poor scalability beyond 50 processors for the problems investigated. This bottleneck was circumvented by implementing the METIS algorithm during this project. Improved load balancing was observed up to 256 processors for cases with minimum resolution of “~d” with a peak cell count of about 16 million. The METIS algorithm will be available in the next version release of CONVERGE, thus highly benefitting the CONVERGE user community. To the best of our knowledge this is a pioneering study in the field aimed at demonstrating both grid-convergence and high-scalability of ICE simulations. Project URL: http://www.transportation.anl.gov/engines/multi_dim_model_home.html Current FY Hours Used: undetermined amount New FY Requested allocation: 495000 Q1: 125000 Q2: 125000 Q3: 125000 Q4: 120000 Justification: Thank You, The LCRC Accounts System