15th Presentation | Towards accelerating turbulent flow computations using deep neural networks | Georgios Momferatos
You can watch presentation’s recording here!
NanoAI welcomes Georgios Momferatos who will present his work, titled “Towards accelerating turbulent flow computations using deep neural networks”.
As always, an open discussion will follow.
When? June 29th, 15:30 Greece
Description:
Turbulence, loosely defined as chaotic, three-dimensional multi-scale vortical fluid motion is considered to be one of the most important open problems in physics and engineering. Numerical simulation is an indispensable tool in the prediction of properties of turbulent flows, which is crucial for a wide range of applications, from astrophysics to biofluid mechanics. However, the wide range of scales involved in turbulent motion makes direct numerical simulation (DNS) that resolves all scales prohibitive for most flows of practical interest. As a more viable alternative, large eddy simulation (LES) simulates the large scales of the flow directly while implicitly modeling the small scales. In most existing LES models, the small scales are modeled by the introduction of a turbulent viscosity coefficient. More recently, data-driven LES models have been proposed in order to overcome limitations of traditional approaches. In our approach, an LES model is trained on DNS data by leveraging a novel deep neural network architecture, yielding promising results towards the acceleration of turbulent flow simulations.
Georgios Momferatos was born in Athens, Greece. In his diploma thesis in the School of Mechanical Engineering of the National Technical Univesity of Athens he studied anastomosis flow, under the supervision of Prof. S. Tsangaris. In his master’s thesis in the School of Applied Mathematical and Physical Sciences of the National Technical Univesity of Athens he studied the motion of inertial particles in a turbulent flow, under the supervison of Prof. G. A. Athanassoulis. In his PhD thesis in the Department of Physics of Université Paris-Sud, he studied the structure of regions of extreme turbulent dissipation in the diffuse interstellar medium, under the supervision of Pierre Lesaffre. From February 2015 to October 2015, he was a postdoctoral researcher at the LERMA laboratory of the Paris Observatory, where he worked on star formation with Andrea Ciardi. He is currently Research Associate in Computational Fluid Dynamics at the Environmental Research Laboratory of NSCR “Demokritos”. Except of the ALIAKMON code (https://sites.google.com/view/aliakmon) for the direct numerical simulation of turbulence, he has also developed the post-processing code EXSTRUCT for the extraction and statistical analysis of the small-scale structure of a scalar field.
Subscribe to our mailing list: https://lists.iit.demokritos.gr/mailman/listinfo/nanoai
You can find more info on this upcoming talk as well as on our previous talks at http://nano-ai.iit.demokritos.gr/