Speaker
Description
Two data-driven approaches for interpreting turbulent-flow states are discussed. On the one hand, multidimensional scaling and K-medoids clustering are applied to subdivide a flow domain in smaller regions and learn from the data the dynamics of the transition process. The proposed method is applied to a direct numerical simulation dataset of an incompressible boundary layer flow developing on a flat plate. On the other hand, a novel nonlinear manifold learning from snapshot data for shedding-dominated shear flows is proposed. Key enablers are isometric feature mapping, Isomap, as encoder and, 𝐾-nearest neighbors algorithm as decoder. The proposed technique is applied to numerical and experimental datasets including the fluidic pinball, a swirling jet and the wake behind a couple of tandem cylinders.
| Classification | Both methodology and application |
|---|---|
| Keywords | Dimensionality reduction; Clustering; Turbulent flows |