I-FENN: Accelerating finite element simulations with machine learning

Panos Pantidis
| October 25th 2022 – 11 am, NYUAD C1 (ERB) -045 |

For the past several decades, researchers and practicioners from all engineering fields have been relying on the finite element method to model and understand the physical world. Despite the method’s many strengths, the computational time of this method stands as a bottleneck as we attempt to model multi-physics, multi-scale and non-linear phenomena. The question is, can machine learning help us do better?

This presentation introduces I-FENN, a framework which integrates the conventional finite element method with neural networks. The objective is to solve nonlinear computational solid mechanics problems with the use of these machine learning algorithms, faster yet at least equally accurate as the conventional approach. The talk will delve into the mechanics of the framework and
discuss its significance, current capabilities, limitations, future outlook and more.

Speaker’s Bio

Dr. Pantidis obtained his BSc degree from the Civil Engineering Department of the Aristotle University of Thessaloniki, Greece, in 2015. He obtained his Ph.D. degree in 2019 from the Civil Engineering Department of the University of Massachusetts, Amherst. Consequently, he moved to New York City where he worked as a Senior Engineer in the engineering consulting firm Thornton Tomasetti. Dr. Pantidis is now a Postdoctoral Research Associate at New York University Abu Dhabi, working in the fields of computational solid mechanics and machine learning.