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