(12e) Particle Velocity and Particle Concentration Determination in Granular Dilute Phase Flow Using a Single Layer of Optical Sensors

In various industrial processes granular material is pneumatically conveyed. To allow for flow monitoring and flow control in closed conveyor pipes, the particle velocity and the particle concentration are crucial parameters to be obtained. The possible abrasive behaviour, fast transportation velocities, and the lack of interference of the measurement equipment with material flow make a non-invasive measurement principle desirable. Due to the long free lengths of path of the particles in dilute phase flow, the individual particle velocities are quite different. To determine the flow accurately, a cross-sectional particle velocity profile has to be measured for dilute phase flow. Numerous measurement methods how to obtain flow parameters in dilute phase flow have been developed. Among others, capacitance based principles (e.g. ECT), charge based principles, and measurement principles based on the attenuation of acoustic sound, light, or radiation are commonly used. With cross-correlation flowmeters these typical quantities are determined in two measurement layers. The signals obtained from the upstream and the downstream sensor layer are then cross-correlated to derive the duration required for a natural disturbance of the flow to move from the upstream layer to the downstream layer. Particle velocity can hence be easily and reliably calculated. In this paper we present a transmitted light approach with the speciality that the two quantities, particle velocity and particle concentration, can be determined in one single measurement layer, perpendicular to the direction of flow.

We developed a measurement principle that comprises various single light dependent resistors (LDRs). LDRs change their electrical resistance depending on the light intensity received on their sensitive sensor surface. These LDRs are mounted on one side of a transparent sightglass in one single column. Each LDR is exposed to light from a spotlight that shines through the transparent pipe. Around the circumference of the measurement layer, two light sources and two inward oriented LDR sensor columns are mounted alternately so that each sensor column is exposed to the light of an opposed spotlight. Whenever a particle passes the measurement layer of LDRs, the particle blocks the light and throws a well-defined shadow on at least one light-sensitive sensors per column. The signal run over time is recorded for each sensor and applied to a special data post-processing algorithmic.

Emanating from the simple fact that slowly conveyed particles have a longer duration within the sensitive measurement layer and hence the affected light sensor is shadowed longer than for a fast particle, a measurement conception can be derived: To make the different ?shadow-times? comparable and easy and fast to process, the time domain of the signal run is transformed into the frequency domain by using the discrete Fourier transform (DFT). Simply stated, slow signal contributions correspond to low frequencies in the Fourier spectrum while fast particle velocities and hence fast signal contributions can be found at higher frequencies in the Fourier spectrum. The particle concentration is determined by means of counting the significant signal deviations whenever a particle passes the measurement layer. Due to a matrix shaped setup of LDRs, for the same time step a shadow and hence a signal deviation can be detected in both LDR columns, resulting in a distinct cross-sectional position of the particle. The concentration profile can be estimated by assigning the particle frequentness to each possible 2D position in the pipe.

The final paper will present the detailed measurement conception and will also show the applicability of the proposed method for particle velocity and particle concentration determination in granular dilute phase flow. A dynamic particle model will be used to demonstrate the advantages and restrictions of presented method and to estimate the achievable accuracy.