Ramanan Sankaran

Ramanan Sankaran received his PhD in Mechanical Engineering from the University of Michigan, Ann Arbor. He develops and applies HPC numerical simulation and modeling tools to explore novel concepts in energy conversion, utilization, materials, and engineering applications. He develops computational and data science tools for multiscale problems in physical sciences and complex reacting flows. He develops and uses some of the most powerful computers in the world, including the HPC systems at ORNL, such as Frontier.

Applications for multiscale flow simulations and foundation models

There is a growing need for developing and applying multiscale predictive simulation tools for a wide range of applications involving complex chemical flows and turbulence. In this talk, we will highlight a few application areas at Oak Ridge National Laboratory with a strong connection to real-world problems in energy and national security. Multiscale simulation tools and models are being developed and applied to problems such as high-temperature materials, chemical separations, and post-detonation debris flow. A sample of applications will be highlighted to show the ways in which computational tools are being exercised to support the laboratory’s core capabilities in these application areas. We will also highlight the recent developments in artificial intelligence and the development of foundation models for turbulence. We will introduce a multiscale hierarchical turbulence transformer, called Matey, that is being developed and trained at the laboratory building on a wide array of high-fidelity simulation datasets.