Further development of a simulation model for the description of the crystallization kinetics of semi-crystalline thermoplastics

Kurzfassung

During the shaping from the melt semi-crystalline thermoplastics form crystalline regions as the polymer solidifies, depending on the material properties and the cooling conditions. These regions change significantly the mechanical and technical properties of a component. Predicting the degree of crystallization resulting from a manufacturing process (fraction of crystalline regions) can therefore help to better predict the component properties obtained. The calculation of the degree of crystallization is shown here on the example of the additive manufacturing process Filament Fuse Fabrication (FFF). In this process, a polymer melt is deposited through a nozzle and the component is created layer by layer. In additive manufacturing with its complex cooling conditions, the degree of crystallization influences not only the material properties but also the component properties, such as the bonding (inter layer strength). The aim was to transfer a known crystallization model to the example of PEEK and implement the model in Abaqus, in order to be able to predict the degree of crystallization of semi-crystalline thermoplastics, based on the existing temperature conditions in the FFF process. Since crystallization depends on many process and material parameters, the prediction of crystallization kinetics is quite complex. The approach chosen is based on a formula that is a further development of the Avrami approach. In addition to temperature-time data, this also includes parameters determined by fitting differential scanning calorimetry (DSC) curves. DSC measurements with different cooling rates are considered and fitted with a single set of parameters. By using the different DSC data, the cooling rate dependent degrees of crystallization can be calculated. A Python script for fitting the parameters of semi-crystalline thermoplastics was developed and various approaches were tested. The completed formula is used in a Fortran subroutine that uses temperature data from an Abaqus simulation to calculate the degree of crystallization. The script is applied to a simplified 2D model of an FFF process. Individual beads are considered, which are activated sequentially to model the deposition process. Individual thermal boundary conditions are set for each activation step to model the cooling as accurately as possible.