Experimental and Numerical Evaluation of the Heating Behavior of Thermoplastic Composites Prior to Forming

The high strength of thermoplastic composites combined with the simple processability, and high wear resistance makes them very attractive, and the lightweight industry has a huge demand for recyclable, high performance materials. To meet those demands, high standards are imposed on material quality and lean, efficient processes. Thermoplastic composites are mostly supplied in semifinished sheet form. Each sheet is made up of several stacked consolidated layers of woven fibers embedded in a thermoplastic matrix system. Parts from thermoplastic composites are manufactured in stamp-forming processes. Therein semifinished sheets are transported to a heating station to be heated above the transition temperature of the polymeric matrix to make the material formable. Even though the thermal conductivity of the polymer is raised owing to the fiber reinforcement, the transversal thermal conductivity is relatively low. The fact that thermal deconsolidation takes place between the layers of woven fibers when the temperature passes the transition temperature of the polymeric matrix even adds to this phenomenon. In order to ensure process efficiency, process models must be developed. Existing models do not consider thermal deconsolidation and available analytical models for the anisotropic thermal conductivity are inaccurate. This work deals with the creation of models for the anisotropic thermal conductivity that are sensitive to changes in the fiber volume ratio and the composite temperature. In that sense, also thermal deconsolidation has to be considered. The first step is the creation of a unit cell to break down a composite sheet to a small repetitive unit. Heat flux balancing over the fiber direction and the transversal direction in the unit cell leads to expressions for the anisotropic thermal conductivity. The next step is to adjust the model to target the problem of thermal deconsolidation by applying a function for the coefficient of linear thermal expansion in transversal direction. Ultimately, also the temperature dependent anisotropic thermal conductivity can be calculated. The modeling is accompanied by validation experiments. Samples are prepared by parallel plate molding and tested for their specific heat capacity using differential scanning calorimetry, for density using Archimedes’ principle and X-ray computed tomography and for anisotropic thermal conductivity using the transient plane source method. The newly developed models are far more accurate than existing ones. Including thermal deconsolidation into the model shows a tremendous improvement to all other attempts made in the past. Since the models are very promising in a controlled environment, a computational fluid dynamics simulation of a real composite infrared heating step in a thermoforming press is conducted. The results are validated via heating tests in the infrared heating station of the thermoforming device. A direct comparison of different levels of complexity of the material properties shows, that only by inclusion of anisotropic properties in combination with thermal deconsolidation effects as presented in this work can lead to satisfying results that match the experimental data.