On-line Characterisation of Particulate Suspensions Using a Multi-Sensor Approach

Publication Reference: 
ARR-38-02
Author Last Name: 
Williams
Authors: 
Professor R A Williams with Dr Yanmin Zhang
Report Type: 
ARR - Annual Report
Research Area: 
Characterisation
Publication Year: 
2000
Publication Month: 
11
Country: 
United Kingdom

Following on our new sensing strategies to enable on-line measurements in concentrated and, using different approaches, for dilute flowing mixtures reported in the second year of our work, we have since concentrated on the dilute flowing system using the Particle Gymnasium approach.

Over the last year extensive work has been conducted on design, fabrication and testing of a particle flow system, a tube sensor and its on-line data acquisition system. Significant results have been obtained in terms of the relationship between the sensor signal and particle size and shape. The benefits of the revised approach using a 'sensor tube' rather than an orifice (Coulter) have been confirmed.

An on-line data acquisition system was designed and assembled, including means of reducing signal noise without affecting the signal response, as we intended to use the response time for particle length estimation. An experimental apparatus for particle flow was designed and built for examining the effect of particle flow behaviour on the measurement signals. Most importantly, a new Pt wire sensor of a ring shape was designed and fabricated, which can be easily fitted via flanges into the pipeline for measurements. Special designs for adding particles into either a horizontal or vertical pipe flow system were also made to examine the effect on signal response.

Using the newly designed experimental apparatus, it is possible to carry out both dynamic tests (adding particles into to the flow system) and static tests (particles fixed on a string to achieve a specific particle radial position and orientation) tests. Results revealed that the change of the particle radial position, up to 83% of the tube diameter, did not affect the peak value of the signal, while particle orientation had a great effect on the peak, giving a 63% difference for a cylindrical particle of 2.2x5.35mm (diameter x length). Particles moving close to the electrodes produced some more complex voltage signals. Particles that rotate during measurements also yielded more complex voltage perturbations. Therefore it is possible to use the orientation effect for particle shape estimation. Experimental and modelling results have also shown that for cylindrical particles of constant orientation both the particle diameter and length could be estimated using the Particle Gymnasium sensor.

Further work is needed to confirm the above finding using particles of arbitrary shape and size. It is also necessary to examine whether we can recognise the particle shape, other than only non- spheroid, from the signal profiles. Other data analysis methods, such as neural networks, may be appropriate especially when the relationship between the profile and size and shape becomes too complicated to be described by simple equations and when particulates can be classified into different user-defined categories.

Work on concentrated systems as well as multiphase flows has also continued, as summarised at the Annual Meeting and in this report.

A revised forward programme has been discussed to focus upon microstructure sensing of concentrated flowing mixtures.