Automatic Extraction of Hairpin Vortices in Turbulent Flows via Topology-Guided Segmentation and Physically Informed Merging

Hairpin vortices are key structures in wall-bounded turbulence, playing a major role in momentum transport, mixing, and energy dissipation. However, they often exhibit irregular shapes and complex interactions with nearby flow structures in high Reynolds number conditions, making accurate extraction difficult. In this work, we present a novel framework for automatic hairpin vortex extraction that utilizes their physical and geometric traits with a topology-based approach. The method begins by identifying vortical regions using the λ₂ criterion, followed by a modified contour tree-based method to decompose these regions into smaller segments. Segments with positive spanwise vorticity fluctuation (ω_y′ > 0), likely belonging to hairpin vortex heads, are used as seeds to trace vortex lines. Segments that overlap with the vortex lines are then merged to form candidate hairpin vortex regions. These regions are further refined by skeleton extraction. The curvature and the physical information along the skeletons are used to validate these hairpin vortices. Finally, smooth surfaces enclosing the confirmed hairpin vortices are generated for visualization. We demonstrate the effectiveness of our approach by applying it to turbulent channel flow from the Johns Hopkins Turbulence Database (JHTDB), significantly improving both the accuracy and efficiency over the previous techniques.