DT-HATS Doctoral Candidates Present Research at AI and Fluid Mechanics Conference

Share the Post:

From 31 August to 3 September, four DT-HATS doctoral candidates – Eirini Kefaloukou, Xiao Xiao, Dimitrios Chondroudakis and Tommaso Baffetti attended the AI and Fluid Mechanics Conference in Chania, Greece, chaired by Prof. Manolis Gavaises (City St George’s, University of London).

As part of the conference, Xiao Xiao, from IFP Energies nouvelles, presented the poster “Development of Deep-Active-Learning Neural Networks for the Real Fluid Modeling (RFM) Framework”. Her work investigates a deep-learning approach for the thermodynamic closure required in real-fluid CFD simulations, aiming to accurately reproduce the required fluid properties while reducing computational cost compared with conventional lookup-table methods.

Tommaso Baffetti, from Université libre de Bruxelles, presented his poster “Attention-based reduced-order forecaster of reactive flows dynamics”, focusing on a reduced-order modelling approach combining Proper Orthogonal Decomposition (POD) with a Transformer-based regressor for predicting flame dynamics. The results obtained so far show promising accuracy and stability for the considered test case.

These contributions highlight the ongoing research carried out within the DT-HATS project and its efforts to advance the application of machine learning to fluid-dynamics problems.