Towards a "headache-free" flow region selection

Identifying different flow zones based on thermo-fluid dynamic properties is not a trivial task. In this paper, we propose a Machine Learning (ML) method based on the Gaussian Mixture Model (GMM) to consistently identify the boundary layer and shock wave regions without using any input parameter or topological information of the flow domain. This approach enables us to obtain a robust and fully automatic domain selection, which is essential for performing the drag breakdown into physical components (viscous and wave contributions) using far field methods. The most challenging issues we face are discriminating between the boundary layer and shock wave wakes and dealing with the high sensitivity of vortex- based methods to domain changes.