Tribology of rough surfaces: characterization and optimization by AI.

Laboratory: PPRIME Institute, Department: D3-GMSC, Team: Tribolub

ED SIMME Doctoral School proposes the following thesis subject:

Title of the thesis : Tribology of rough surfaces: characterization and optimization by AI.

This project would be under the supervision of : Dominique Souchet, D3-GMSC department of PPRIME Institute at Universite de Poitiers.

Co-supervisor(s) : Arthur Francisco /

Starting date: 10/2022

Thesis presentation:
A mathematical surrogate can replace advantageously a batch of complex elastohydrodynamic calculations: over a wide range of parameters, the surrogate instantaneously gives results very close to numerical results that are computationally expensive. It is then possible to search for optimal operating parameters using genetic algorithms for example.
The aim here is to go much further, using convolution networks (deep neural networks), to optimize the textural patterns of surfaces in a bearing or a pad. The model will be built with the open source software Keras and Tensorflow and will be trained with the team’s MUSST software for tribological calculations.

See the detailed description: SujetUP-2022-souchet-francisco-EN.pdf

For further information and to apply, please contact: : Francisco ().

Submission date: 01/20/2022 à 17 h 05 min

Direction de l'École Doctorale SIMME
ENSMA - Téléport 2 - 1 av. Clément Ader
BP 40109
86961 Futuroscope Chasseneuil Cedex