Particle Tracking Thesis 2019

2D/3D PTV Application and Technique

Lagrangian Particle Tracking Thesis
Abstract: Geological sequestration of CO2 in saline aquifers is one of the solutions pursued for reducing greenhouse gas emissions. Understanding the interaction between the resident brine phase and the invading CO2 phase is crucial in predicting the safety and security of the storage process. In this work particle tracking velocimetry (PTV) is used to study CO2– water multiphase flow in a micro model at reservoir - relevant conditions. Int his work, using an open source PTV code (OpenPTV), the position, velocity, and acceleration of tracer particles dispersed in water was extracted. Demonstrated within the figures of this work, are plots of particle trajectories, velocities, and multiphase flow migration through the micro model. Different flow stages are identified during the displacement process. This thesis serves as a starting point for those unfamiliar with using OpenPTV. Additionally, supporting code is provided that addresses trajectory linking and repair of the OpenPTV data.
Keywords: FlowTracks; OpenPTV; Particle tracking velocimetry; Trajectory linking and repairing
Thesis PDF link:
Thesis Author: M. A. Santana
3D multi-view Particle Tracking Velocimetry Thesis
Abstract: In this study a new flow visualisation method the 3D Multi-View Particle Tracking Velocimetry has been developed. This multi-view vision based measurement system can capture large-scale flow structures with optical obstacles. In order to increase the tracking performance of this method, it is enhanced with a novel tracking algorithm, the multi-pass tracking algorithm with robust initialisation. The use of multiple webcams in this fluid visualisation method reduces the hardware costs significantly and makes it feasible to capture flow structures for various types of flow. The workflow developed for this measurement system enables a systematic execution of the measurements. The conventional multi-view camera calibration process with bundle adjustment algorithm is improved with an image filtering step, which reduces the resulting mean reprojection error by 90%. In order to find the camera set with the lowest triangulation error, two multi-view triangulation approaches are developed and compared. In this context, the results of the multi-view triangulation with precalculation were closer to the lowest possible triangulation error than the multi-view triangulation with elimination. In order to increase the capabilities of the development process, a software-in-theloop environment has been developed as well. Using this software-in-the-loop environment, the multi-pass tracking algorithm with robust initialisation is compared with a conventional tracking algorithm. The results show that this tracking algorithm delivers significantly higher tracking efficiencies which do not decrease dramatically with ascending seeding rates. Finally, experiments for an indoor flow case were carried out using this new measurement system. The comparison of the experimental results with a comparison measurement using hot-sphere probes showed that this measurement system delivers plausible results.
Keywords: fluid visualisation; particle tracking velocimetry; PTV; 3D-PTV; 3D-MVPTV; computer vision; multi-view; image processing; Strömungsvisualisierung; Mehrfachansicht; Bildverarbeitung
Thesis PDF link:
Thesis Author: T. Askan