Double-frame 3D-PTV using a tomographic predictor

Abstract

Thanks to the technological progress, 3D velocimetry techniques are becoming more popular. In particular, the time-resolved flow analysis by means of particle tracking is very attractive. Compared to double-frame recordings, higher seeding concentrations are feasible, yielding high spatial resolution results without bias errors due to strong velocity gradients. However, hardware restrictions still limit time-resolved measurements to rather small flow velocities and low magnifications. In aerodynamics, especially, this is a drawback, since often higher flow velocities are of interest. To conduct reliable 3D-PTV measurements from double-frame recordings, the well-established techniques tomographic particle imaging and 3D-PTV are employed for a novel processing approach. In this combined approach, the tomographic reconstruction is used as a predictor for the sensor locations of the corresponding particle images of the reconstructed particles. Furthermore, the reconstruction helps to identify and reject non-corresponding sets of particle images, reducing the amount of ghost particles to a minimum. A probabilistic tracking algorithm is then applied to estimate the flow field.