Development of a Flow Model for the Design of a Momentum Type Beet Mass Flow Sensor

Abstract

A lot of research is done to monitor yield during harvesting. Combined with a positioning system, this information can be used to produce yield maps. Based on the yields of successive years, site-specific field management can be done, aiming for higher local economic and ecological yield efficiencies. In this research, a curved plate mass flow sensor for sugar beet is designed with the aid of a mathematical model. The principle is based on measuring the impulse flow colliding with the plate. For granular materials, an accuracy of 2% is achieved. The mass flow sensor for sugar beet is integrated as a curved side rack in the cleaning channel of the harvester. The side rack is mechanically isolated from the frame. A theoretical model of the beet flow over the turbines of the cleaning unit is calculated and incorporates all physical parameters that influence the mass flow measurement by the momentum sensor. The flow density and friction coefficients with the rack and the turbine characterise the mass flow. From the model, the influence of the different parameters (slope, beet velocity, friction, length of the plate) on the momentum is investigated. The force exerted on the rack of the cleaning unit is measured, from which the momentum can be indirectly calculated. A measurement device is constructed, minimising the influence of the harvesting conditions and the material properties. After calibration, measurements are carried out on the cleaning unit of a Dewulf R6000T (two-phase system) and an Agrifac ZA 215 EH (one-phase system). Research is done to construct the sensor independent on friction properties between beets and the rack. Major influence comes from the velocity of the transported material. Variations in beet speed have to be registered continuously with a Doppler radar meter. Influences of a varying slope can also be corrected regarding the instantaneous beet velocities. More accurate measurements are done when the sensor is installed on bigger spinning wheels because of higher momentum due to a higher centrifugal force and hence a higher signal to noise ratio. To evaluate the model, only average mass flows and flow speeds could be used to predict the executed moments. When comparing the integrated measured momentum and the scale weight of the harvester bin, the error never exceeds 3% when using the derived equation.