PhD opportunities

[PROVIDED] Machine learning for parameter calibration of digital twins in crystal plasticity

Thesis proposal

Area of expertiseMechanics
Doctoral SchoolISMME - Systems Engineering, Materials, Mechanics, Energy
SupervisorM. David RYCKELYNCK
Research unitCentre of materials
ContactRYCKELYNCK David
http://www.mat.mines-paristech.fr/Accueil/Propositions-de-theses/
Starting dateOctober 1st 2021
KeywordsMachine learning, digital twin, constitutive equations, data-driven modeling
Abstract4D in situ tomography characterization methods now provide high-resolution multimodal data (microstructure, damage, crystallography).Massively parallel FFT methods make the transition from image to mechanical digital twin much easier. The direct comparison betweenexperiment and numerical results at the scale of plasticity and fracture offers an extraordinary new opportunity to validate local constitutivelaws and damage models. However, the complexity of such simulations added to the number of parameters of the models is still anobstacle to directly identify the models at the scale of crystal plasticity.The construction of latent spaces by auto-encoder makes it possible to interpolate simulation predictions by capturing their non-linearity.This makes it possible to automatically sample the simulation space with only a few high-fidelity mechanical calculations. Thanks to ametric qualifying the distance in simulation space, categories of models can be defined. High fidelity simulations are then performed onlyfor representative cases of each category of model. In the end, we will end up with an extremely efficient simplified modeling chain forquickly calibrating each digital twin. This generic method will be applied to the identification of crystalline plasticity laws of hexagonalmaterials (Titanium and / or Zirconium), focusing in particular on critical resolved shear stresses and interaction matrices.
ProfileEngineer and / or Master of Science - Good level of general and scientific culture. Good level of knowledge of French (B2 level in french is required) and English. (B2 level in english is required) Good analytical, synthesis, innovation and communication skills. Qualities of adaptability and creativity. Teaching skills. Motivation for research activity. Coherent professional project.

Prerequisite (specific skills for this thesis):
Programming in python, notion of statistical learning or knowledge of continuum mechanics, applied mathematics, good computer skills.

Applicants should supply the following :
• a detailed resume
• a copy of the identity card or passport
• a covering letter explaining the applicant’s motivation for the position
• detailed exam results
• two references : the name and contact details of at least two people who could be contacted
• to provide an appreciation of the candidate
• Your notes of M1, M2
• level of English equivalent TOEIC
to be sent to recrutement_these@mat.mines-paristech.fr
FundingContrat de recherche