Informationen zur Anzeige:

Thesis -Application of PINNs on 3D CFD Data
Graz (Österreich)
Aktualität: 19.06.2025

Anzeigeninhalt:

19.06.2025, AVL List GmbH
Graz (Österreich)
Thesis -Application of PINNs on 3D CFD Data
Aufgaben:
This thesis investigates the potential of Physics-Informed Neural Networks (PINNs) in the field of fluid dynamics. The focus is set on turbulence modelling allowing applications relevant to industrial contexts. The thesis should cover a comprehensive literature review, highlighting two promising use cases: (a) up-sampling and (b) acceleration of simulations. Up-sampling refers to enhancing the resolution of simulation outputs - such as velocity fields - from coarse to fine grids (also known as super-resolution). Acceleration encompasses methods aimed at reducing computational time for solving fluid dynamics problems. Based on insights gained from the literature, one promising approach will be applied in a case study using a fluid dynamic dataset provided by AVL. The implementation will be carried out in Python, leveraging deep learning frameworks such as TensorFlow and PyTorch to incorporate state-of-the-art functionalities. Literature Review on PINNs Application of PINNs on 3D fluid dynamics data
Qualifikationen:
Ongoing studies in the fields of Mechanical Engineering, Data Science, Informatics or similar Experience in python coding Experience in Data Science beneficial Experience in Fluid Dynamics beneficial

Fachrichtung

Art des Angebots

Standorte