Workshop on Model Order Reduction and Probabilistic Model Learning for Materials and Structures

9:00-9:20

Connection to video conference services

9:20:9:30

Introductory statement

9:30-10:20

  (40 mn + 10 mn Q)

Keynote: Iuri B.C.M. Rocha, Delft University of Technology, On-the-fly reduction of concurrent multiscale models through probabilistic machine learning

10:20-11:00

 (30 mn + 10 mn Q)

Pierre Kerfriden, Mines Paris PSL / Cardiff University, Probabilistic learning for error estimation in large-scale, top-down multiresolution finite element models.

11:00-11:10

Break

11:10-11:40

  (25 mn + 5 mn Q)

Sami Hilal, EDF Châtoux / Mines Paris PSL, Bayesian calibration of mesoscale thermo-mechanical models for the reliable control of ALM by direct energy deposition.

11:40-12:10

  (25 mn + 5 mn Q)

Vasilis Krokos, Synopsis Ltd. / Cardiff University (UK), Bayesian convolutional neural network as probabilistic surrogates for the fast prediction of excess stress in structures with microscale patterns, local features or defects       

12:10-12:40

  (25 mn + 5 mn Q)

Pablo Alvarez Pereira, EDF Châtoux / Mines Paris PSL, Probabilistic modelling of a welding process via local reduced order models

14:00-14:50

Keynote: Gianluigi Rozza, SISSA (Italy), A localized reduced-order modeling approach for PDEs

14:50-15:30

  (30 mn +10 mn Q)

David Ryckelynck, Mines Paris PSL, Transfer learning via model order reduction

15:30-15:40

Break

15:40-16:10

  (25 mn + 5 mn Q)

Thomas Daniel, SafranTech / Mines Paris PSL, ROM-nets classifiers for local model recommendation

16:10-16:40

  (25 mn + 5 mn Q)

Hugo Launay, Mines Paris PSL, Hyper-reduced arc length algorithm for stability analysis in elastoplasticity