Feasibility Study of Tomographic Particle Image Velocimetry for Large Scale Convective Air Flow

Abstract

A promising approach to capture the unsteady three-dimensional flow structures of large scale convective air flow is the recently developed Tomographic Particle Image Velocimetry (Tomographic PIV). In a test set-up the technique is applied to a measurement volume of 0.69 m x 0.42 m x 0.24 m, which is more than 1000 times larger than in previous applications. In order to test the feasibility of Tomographic PIV for large scale applications different calibration errors, which are important for the reconstruction quality of the three-dimensional intensity distribution, are analysed in detail. These errors are the reprojection and the triangulation error. The results reveal that the final calibration errors are of the order of 0.1 pixel from which one can conclude that large scale Tomographic PIV is practicable. The implementation of the reconstruction algorithm (multiplicative algebraic reconstruction technique, MART) is discussed. Furthermore, different attempts to reduce the calculation time of the reconstruction algorithm are presented, which are the pixel wise approximation of the calibration function and the local approximation of the lines of sight. In total a speedup of 60 percent has been achieved. In order to test the correct implementation of the algorithm, the three-dimensional positions of the markers on the calibration plate are reconstructed. Finally, the technique is successfully applied for evaluation of the three-dimensional displacement vector field of the test set-up.