Tomometric Approach Using Multi Sensor Data Fusion in Particle Segregation Studies

Urmila Datta1,2, Vidar Mathiesen1,2, Saba Mylvaganam1,2

1Telemark University College, Faculty of Technology, Kjølnesring 56, N-3914 Porsgrunn

2Tel-Tek, Kjølnesring, N-3918 Porsgrunn, Norway, saba.mylvaganam@hit.no

ABSTRACT

In conjunction with particulate segregation studies in pneumatic transport of particulates of varying sizes, a tomometric multi sensor scenario was realized using dual plane multi-electrode capacitance sensors, laser Doppler anemometric interrogation of particle velocity and flange mounted Acoustic Emission (AE) sensors. The laser anemometer gave the particle velocity directly, whereas the dual plane multi-electrode capacitance sensors and AE sensors gave a set of concurrent time series at different locations. The concurrent time series from these sensors were convolved to estimate different velocities of interest. The velocities obtained from these sensors based on different physical phenomena, were then compared with velocities from FLUENT® simulations. The different particles tend to segregate in the process and this segregation was confirmed by simulations and from results using Multi Sensor Data Fusion (MSDF). This paper presents the multi sensor suite, some analysis of their signals and simulation results. The paper also presents very briefly, the system integration excluding the laser, which was used mainly for verification purposes only.

Keywords Tomometry, segregation, Acoustic Emission (AE), LDA, MSDF, fluidised beds, pneumatic transport, Laser Doppler Anemometer (LDA)