Continuous mass flow measurement of granular materials

This thesis reports on the development of a double inclined plane (DIP) transducer system based upon the principles of force reaction. This transducer had a design specification to measure the true mass of “free” flowing granular materials, primarily agricultural crops, with an accuracy of ± 2 % on total mass flowed for flow rates between 1 and 10 kg/s. Two absolute values are used to assess accuracy in this study, (i) the total mass flowed (kg), a measure of the mass accumulation and, (ii) the mass flow rate (kg/s), a ‘spot’ reading of flow rate. The performance of the novel system has been evaluated through theoretical, laboratory and field studies and consideration has been given to the commercial and business aspects of the manufacture, marketing and further development of the device. Following mathematical and laboratory studies of the sponsors current force reaction transducer - a single reaction device, angular variations were highlighted as having a significant effect upon output. A further study found that the least sensitive reaction plate angle was 55° To overcome this problem, the double inclined plane (DIP) concept was developed, 2 single reaction plates, joined along their apexes, angled at 55 degrees to the horizontal, mounted upon a horizontal strain gauged beam. Angular compensation when tilting the transducer was provided by generating a higher force from the shallower face and a lower force from the steeper face. A mathematical model of the new transducer allowed the output to be predicted to within 1.7 %. Initial calibration was undertaken in the laboratory and tested using a combine clean grain system simulation apparatus. In-situ machine studies were performed by mounting the transducer in a New Holland TF42 combine, firstly on an extended bubble up auger and finally in the drop box, between the clean grain elevator and bubble up auger. Initial pilot studies were conducted with the combine static to calibrate the system and finally a full harvest field trial was undertaken. Over the harvest field trial, the accuracy on accumulated mass was better than 0.9 % over 127 tonnes. Tramline effects upon transducer output were found to be self cancelling, as the resulting positive and negative ‘spikes’ in the signal, when summed over time approximated to zero. Changes in pitch angle, up to 10 degree caused between 1.5 and -2.8 % randomly distributed error. Roll angles up to 7.5 degrees, caused between -0.9 % and 1.7 % randomly distributed error in static trials. Field beans and oilseed rape required calibration constants 6.8 % and 3.1 % lower than that for 12.5 % moisture content wheat, but with adjusted calibration constants, gave excellent repeatable results. Increasing moisture contents of up to 30 % in wheat resulted in the transducer under reading by, on average, 1.6 %. An assessment of manufacturing costs was made and the unit cost was £154.12 each for 30 units reducing to £109.59 each for 3000 units. It is recommended a pricing objective of maximising sales growth is used which will position the device at the less expensive end of the market. Due to commercial sensitivity, a draft patent has been written to protect the DIP concept. The first stages of commercial adoption already being undertaken by a major multi-national agricultural machinery company, who are evaluating a pre-production prototype. This thesis provides the systems, data and principles required to create a novel, commercially practical transducer system, based upon the principles of force reaction. The problem of angular compensation has been overcome in a simple and effective manner offering a relatively inexpensive but accurate method of measuring mass flow rate, which has already received commercial interest.