Back To Schedule
Wednesday, June 2 • 15:35 - 16:00
Composites and their forming processes in the era of Data and Artificial Intelligence: Composite Twins

Sign up or log in to save this to your schedule, view media, leave feedback and see who's attending!

  • Hybrid-Twins
  • Proper Generalized Decomposition (PGD)
  • Big data

Model order reduction -MOR- allows speeding-up complex calculations, by using reduced bases extracted in the “offline learning stage”, or constructed on-the-fly during the problem solution. The former procedure is at the origin of Proper Orthogonal Decomposition (POD) and Reduced Basis (RB) MOR methodologies. The latter, at the origin of the so-called Proper Generalized Decomposition (PGD) could seem, at first view, unattractive because of the fact that the reduced model is computed during the solution procedure itself. In fact, the PGD efficiency comes from the fact of solving the model by considering the model parameters as problem extra-coordinates, leading to a parametric solution. That solution, in order to circumvent the so-called curse of dimensionality, is expressed in a separated form.
Our recent researches leaded to a non-intrusive solver of parameterized partial differential equations that was successfully employed in a diversity of engineering problems of industrial relevance. Today the so-called sparse variant of the PGD solver allows the efficient computation of such non-intrusive parametric solutions at the very low-data limit. Problems involving moving fronts, localization or geometrical and topological parameters, usually encountered in RTM and SMC composite forming processes, require advanced interpolation schemes.
With such a robust parametric solution available, control, inverse identification, optimization and uncertainty quantification and propagation, all them operating under the stringent real-time conditions are attainable and were properly coupled with data-assimilation to produce efficient DDDAS (dynamic data-driven applications systems), that is, the so-called material and process digital and hybrid-twins.
Within the context of composites modeling, their associated forming processes and their in-service performances, we will prove that hybrid twins allow (i) a natural alliance between physics and data, mathematics and artificial intelligence, (ii) proceeding efficiently at the scarce-data limit and (iii) certifying designs.

Speakers, Jury Members & Final...
avatar for Francisco CHINESTA

Francisco CHINESTA

President of the Scientific Committee and Director of the Scientific Department, ESI Group
Francisco Chinesta is currently full Professor of computational physics at ENSAM Institute of Technology (Paris, France). He was (2008-2012) AIRBUS Group chair professor and since 2013 he is ESI Group chair professor on advanced modeling and simulation of materials, structures, processes... Read More →

Wednesday June 2, 2021 15:35 - 16:00 CEST
Conferences (Hall 6 - Room 611)