Development of a Solids Conveying Throughput Model for Grooved Barrel Extruders based on Discrete Element Simulations
Regarding the design of single screw extruders, the prediction of the throughput needs to be of high accuracy. The calculation of the solids conveying throughput is particularly important for extruders with a grooved feed section. In contrast to smooth barrel extruders the throughput of the entire plant is determined at the feed section. Various simplifications are necessary for an analytical modelling of the solids conveying, e.g. the classification into conveying cases, the assumption of solid block flow and the assumption of pressure anisotropy coefficients. In numerical simulations with the Discrete Element Method (DEM), which has recently been successfully used to describe solids conveying in smooth barrel extruders, the simplifications mentioned above are not taken into account. Here, the pellets are approximated as spherical particles or particles composed of spheres. Based on virtual overlaps, contact models and the solution of Newton's equations of motion, a more complex consideration of the pellet flow is possible. In each iteration step of the simulation, the particle velocities, contact forces and derived quantities, e.g. the mass throughput in the screw channel and the grooves or the radial pressure build-up along the grooved barrel can be evaluated. Therefore, a DEM simulation model can be investigated by means of statistical design of experiments in order to convert the target values into a metamodel by regression. Long computation times of numerical simulations will be avoided in this matter as well as too simple assumptions of analytical approaches. The influencing parameters to be simulated are divided into material, geometry and process parameters. Relevant material parameters are the coefficients of friction of the polymer-polymer and polymer-steel surface as well as the restitution coefficient and the particle diameter. The geometry is varied in the form of the screw diameter, the channel depth and pitch, the number of grooves and their width, depth and angle. By varying the angle, both conventional axial grooves and helical grooves are taken into account. Finally, the process parameters speed and backpressure are also considered in the simulations. In order to reduce the simulation effort, irrelevant parameters are identified in preliminary investigations.
Springer