45847: Aerospace Engineer, Physicist, Computer Scientist or similar - Prediction of Boundary Layer Transition Using Machine Learning

German Aerospace Center (DLR)

Göttingen, Germany

Work group:

Institute of Aerodynamics and Flow Technology



Area of research:

PHD Thesis



Part-Time Suitability:

The position is suitable for part-time employment.



Job description:


Future aircraft will use laminar flow wings in order to reduce the fuel consumption and the ecological footprint of aviation. For the design of laminar wings, state-of-the-art CFD codes currently use empirical criteria, database methods, transport equation models or concepts based linear stability theory (LST) for transition prediction. The different methods vary significantly in the amount of physics modelled, and usually the consideration of more boundary layer physics leads to a significant increase in computational costs. For the design of aircraft, however, a physics-based transition model is required at low computational cost.


A large number of transition studies using LST theory have been performed at DLR in the last two decades. These studies cover a variety of different configurations at different flight Reynolds numbers, Mach numbers, angles of attack, etc. Due to this, a huge data base of laminar boundary-layer flow data with corresponding linear stability results and transition locations exists, already covering a wide range of flow parameters relevant for laminar-turbulent transition for future aircraft. These data can be used for the development and training of new transition prediction concepts based on machine learning (ML), either for predicting directly the transition location or for estimating the local boundary-layer instability characteristics as an alternative to the database methods.


This research project/PhD Position is envisaged to comprise the following steps:


selection of the most promising ML-based concepts; preparation of suitable training data sets; implementation and training of ML-based approaches for prediction of  transition location and estimation local stability characteristics for subsequent use in N-factor methods; comparison of the different concepts; assessment of potential and limitations of ML-based transition prediction

Please apply via recruiter’s website.

Quote Reference: 45847

Favorite