Numerical process simulation for extrusion using correlated stochastic shear viscosity data

Abstract

Within polymer processing, nowadays, one has to deal with higher variations in material properties, for sure if one deals with recycled thermoplastic materials. Those waste streams typically can be of the same polymer type, but different products and or applications. To end up with a as homogeneous as possible material, typically solid mixing of the different batches is performed at large scale. Depending of the application and origin of the material, there is still a possibility to cope with larger variation within material properties. The most important material property for standard polymer processing as injection moulding or extrusion is the shear viscosity behavior in a large range. Within this investigation, at first, a strategy is developed to identify this variation in an industrial way. Next, a strategy is developed to use this characterized material data within a 1d approach valid for several types of the extrusion process. To end up with a physical reliable numerical tool coping with stochastics within data, there is a need to identify a correlation as function of time within the characterization of the shear viscosity. Within the process simulation, this correlation time is also used to implement stochastic information in a physically relevant way. This investigation can also be seen as a multiscale approach, as material characterization is done at meso scale; on processing scale (macro scale), local stochastic changes are compensated within the processing. In this investigation, the comparison is made between the processing of virgin material with the processing of recycled material. Finally, the potential of this approach is shown in a conceptual way.