Application of wavelet filtering to the radioactive particle tracking technique

Abstract

The Computer Automated Radioactive Particle Tracking (CARPT) technique has been used for the investigation of fluid dynamics, material dispersion and mixing in various multiphase rectors. The accuracy of the CARPT measurement depends on the properties of the tracer particle, the algorithm and the technique for calibration and signal processing. In this paper, a filtering technique based on the wavelet theory for the removal of white noise from the data is presented. It is shown, experimentally, that the adopted wavelet-based filtering algorithm reduces the level of noise in the data by 80–90%. The suitability and necessity of wavelet filtering are further illustrated by performing a group of typical CARPT experiments in an air-water bubble column operated at different superficial gas velocities. It is demonstrated that the use of the unfiltered data results in significant over-estimation of turbulent parameters while the influences on the mean velocities are minor.