The Scalable Engineering Applications (SEA) Group focuses on using state-of-the-art computational science to deliver scalable simulation tools for engineering and science applications. The group develops and applies new scalable algorithms for advanced exascale-class computing architectures, including the Frontier supercomputer at Oak Ridge National Laboratory (ORNL).
Members of the SEA Group leverage the latest advancements in numerical methods and performance portability for high-performance computing (HPC) to create simulation codes that can be deployed on a variety of heterogenous HPC architectures. The group also designs and creates modern production-level software packages at the confluence of modern engineering and computational science for a network of subject matter experts throughout ORNL, the Department of Energy (DOE), and the broader scientific community.
The SEA Group’s core areas of expertise include software and numerical methods for radiation transport as well as high energy physics in work supported by the DOE High Energy Physics and Advanced Scientific Computing Research (ASCR) programs. As the primary developers of Celeritas, a high energy particle physics Monte Carlo simulator for HPC, the SEA group supports detector modeling and simulation at the Large Hadron Collider (LHC). In pursuit of a 5× increase in simulation capability for future discoveries, a recent collaboration between ORNL and members of the team at A Toroidal LHC Apparatus (ATLAS) — the team at LHC that co-discovered the Higgs boson — yielded the first GPU-based simulation of the ATLAS detector.
Other areas of expertise within the SEA Group include computational fluid dynamics; computational mechanics, fracture, and heat transfer; and computational mesh generation. The group also leads the creation of scalable software tools that use finite element methods, particle methods, and hybrid particle-mesh methods. The technology developed by the group actively supports efforts funded by DOE’s Fusion Energy Sciences and ASCR programs to model and simulate fusion energy systems. This support includes facilitating work to simulate plasma-liquid metal interactions in plasma-facing components and breeding blankets of fusion power reactors to develop a better scientific understanding of the complex multiscale/multiphysics dynamics of transient scenarios. Other efforts in the group include the development of new simulation techniques for advanced manufacturing methods funded by DOE’s Advanced Materials and Manufacturing Technologies Office.
