Introducing OpenLPT: new method of removing ghost particles and high-concentration particle shadow tracking

We developed an open-source Lagrangian particle tracking (OpenLPT) based on the Shake-the-Box (Schanz, Gesemann, and Schröder, Exp. Fluids 57.5, 2016) method. The source code of OpenLPT is available on GitHub repository (@JHU-NI-LAB). The code features a new method that removes the majority of ghost particles at a high particle image density. The resulting percentage of ghost particles drops from 110% to 26% for image density at 0.125 ppp—nearly 84% of ghost particles are removed. Extensive tests of OpenLPT using synthetic data sets show that the code produces tracks with accuracy and processing time similar to the previously-reported values. In addition, OpenLPT has been parallelized to run on high-performance computing clusters to drastically increase its processing speed. To examine the code’s capability of tracking shadows of small tracers for backlit experiments, the blurred-particle effect was also included on synthetic images and OpenLPT was tested to process these noisy images. The results show that OpenLPT can also track shadows of a high-concentration of particles reliably in 3D. Based on the test, the optimal depth of field (DoF) and particle concentration for future experiments using Lagrangian shadow tracking are provided. For example, DoF controlled by the aperture should be set at around half of the size of the view area. At this DoF, most particles in the interrogation volume can be tracked, whereas particles outside the interrogation volume become too dim to affect results. 40 experimental data sets for a wide range of particle concentrations were also used for evaluating the code, and the results show a nice agreement with the synthetic tests.