July 8 (Mon)

Place 274, Building No.3, Grad. Sch. of Eng. (工学部 3 号館 274)
contact Center for Computational Science

Time 10:30-11:30
Jens Kasten
Leipzig University, Germany
Feature-Based Analysis of Time-Dependent Flow Fields
Visualization of fluid flows can be divided in different areas of research. First, direct visualization uses advanced tools from computer graphics to depict the flow in an easily comprehensible image. In contrast, feature-based analysis helps to focus on the important structures of the flow. For time-dependent flow fields, the concept of Lagrangian coherent structures (LCSs) became famous in the recent years. Typically, these structures are associated with the Finite-time Lyapnuov Exponent (FTLE). This feature indicator connects LCSs to manifolds of strong particle separation and convergence. However, also other structures such as vortices are often connected to LCSs. Here, the acceleration magnitude can help to identify areas of strong swirling motion. In my talk, I will present methods to extract both features. A special focus is placed on using robust combinatorial methods for feature extraction [1,2,3]. I will present applications to different time-dependent two-dimensional flow fields. An overview of the presented tools can be found in [4].

[1] J. Reininghaus, D. Gunther, I. Hotz, S. Prohaska, and H.-C. Hege. TADD: A computational framework for data analysis using discrete Morse theory. In Proceedings International Congress on Mathematical Software 2010, pages 198.208, 2010.
[2] H. Edelsbrunner, J. Harer, and A. Zomorodian: Hierarchical Morse Complexes for Piecewise Linear 2-manifolds. In Proceedings Symposium on Computational Geometry 2001, pages 70-79.
[3] U. Bauer, M. Kerber, and J. Reininghaus. PHAT: Persistent Homology Algorithm Toolbox. http://phat.googlecode.com
[4] J. Kasten: Lagrangian Feature Extraction in Two-dimensional Unsteady Flows. PhD thesis, FU Berlin, 2012.
  PDF file is available here.