An experimental study of reconstruction accuracy using a 12-Camera Tomo-PIV system

A tomographic PIV system composed of a large number of cameras is used to experimentally investigate the relation between image particle density, number of cameras and the reconstruction quality. The large number of cameras allows to determine an asymptotic behavior for the object reconstruction over a wide range of values for the tracer particles concentration. In order to quantify the effect of additional cameras, the quality of the reconstruction is evaluated through the reconstruction signal-to-noise ratio and the normalized intensity variance. The use of additional cameras even with small relative angular separation is shown to reduce the formation of ghost particles in the reconstructed volumes, resulting in a greater reconstruction signal-to-noise ratio and variance of the reconstructed object intensity. Object reconstructions are obtained for levels of the seeding density almost three times greater than a typical four-camera tomographic PIV system, based on evaluation of the reconstruction signal to noise ratio. This approach offers the perspective for future tomographic PIV measurements of higher resolution, particularly suited for the study of high-Reynolds number turbulent flows