Particle position optimization of incompletely reconstructed particle clouds
P. Godbersen(1), D. Schanz(1), A. Schröder(1,2)
1: German Aerospace Center (DLR), Institute of Aerodynamics and Flow Technology, Göttingen, Germany
2: Brandenburgisch Technische Universität (BTU) Cottbus-Senftenberg, Germany
DOI:
A key aspect in Lagrangian particle tracking with Iterative Particle Reconstruction is the position optimization of triangulated particles. Such an optimization is conducted by slightly displacing the individual particles to better match their back projected images with those captured on the cameras. However, the quality of this optimization can be degraded in the early stages of particle reconstruction when only a small number of particles has been found yet. We investigate this effect, motivated by worse than expected performance of DLR multi-pulse STB code in our initial submisson to the 2nd LPT Challenge, and provide some initial fixes to this problem together with the introduction of a novel placeholder approach, capable of preventing such issues in the future. The proposed approach is useful in high seeding density situations, as well as measurement setups that image a not insignificant amount of particles outside of the common camera volume.
